key: cord-0033718-hdnru3tb
authors: nan
title: Physicians Poster Sessions
date: 2014-03-28
journal: Bone Marrow Transplant
DOI: 10.1038/bmt.2014.45
sha: b6f41fc94e683f77e6a97220dcac26271cf57e3e
doc_id: 33718
cord_uid: hdnru3tb

nan

Introduction: Allogeneic stem cell transplantation (HCT) is the only curative approach for patients with Fanconi anemia (FA) and associated myeloid malignancy (AML) . The data about the impact of chemotherapy and HCT are rare. Materials (or patients) and Methods: Between January 1993 and May 2013, 15 patients were identifi ed in the AML-BFM database (Germany) with FA and AML. AML was diagnosed at a median age of 11.6 yrs [0. 9-19.6 ]. In 10 patients,FA preceded the occurrence of AML by a median of 32.3 months [0-161]. One patient was further excluded, because she received HCT before AML. Results: Six of 14 patients underwent HCT achieving a 5 year OS of 47%. The median time to transplant after diagnosis of AML was 2 months [0. [3] [4] . One patient died after relapse, one due to severe graft versus host disease (GvHD), another of unknown cause. Of the three surviving patients++, two had chronic GvHD. They had received a radiation free preparative regimen (mainly busulphan based) including antibodies for intensive GVHD prophylaxis. Two surviving patients received low dose cytoreductive therapy (thioguanine and cytarabine) to bridge time to HCT. None of them achieved complete remission (CR) before HCT. Eight of 14 patients did not undergo HCT with median survival of 40 days, despite in some patients intensive AML driven chemotherapy. Discussion: Three of 14 patients survived and 2/3 surviving patients received low dose cytoreductive therapy for disease control before HCT. Intensive chemotherapy is associated with treatment related mortality by infections. Achieving CR pre transplant might not be necessary for survival. Disclosure of Interest: None Declared.

Introduction: In the absence of a fully matched donor, adult patients with Acute Leukemia (AML and ALL) in CR are usually off ered an allogeneic transplant from an alternative donor, but there has been thus far no demonstration that outcome when using an alternative donor is superior to outcome following an autologous stem cell transplantation (ASCT). We reported last year on a matched pair analysis comparing 113 AML patients transplanted in CR1 with an haplo-mismatch family donor to 226 autologous stem cell transplantations done in the period from January 2000 to December 2010 ( Blood 2013, 3093a). Improvements in both transplant approaches may have changed the outcome, justifying a new analysis. Materials (or patients) and Methods: In the present study, we considered the period from January 2007 to December 2011 and we collected information on patients with transplanted in fi rst (CR1) and second CR (CR2). Also, we considered only T repleted haplo mismatch transplants. We used patient age, diagnosis, status at transplant and interval from diagnosis to transplant ( < > 6 months for CR1; < > 18 months for CR2) and cytogenetics as matching factors. The pair match analysis was done on 130 haplo versus 235 ASCT. Results: The median follow up was 26 months (range, 2-66) for haplo and 20 months for ASCT. Patients allografted were transplanted more recently (median year of transplant : 2010 vs 2009, p<10e-4) , and they received less frequently total body irradiation in the pretransplant regimen (17% vs 29 %; P= 0.01). 7% of the patients in the group haplo received a second allograft, while Introduction: The impact of nutritional status on outcome of allogeneic stem cell transplantation (alloSCT) is controversial. This retrospective study investigates the infl uence of pre-transplant weight loss and serological indicators of nutritional homeostasis on relapse and death of acute myeloid leukaemia (AML) after alloSCT. Materials (or patients) and methods: Pre-transplant weight loss along with serum levels of total serum protein (TSP), albumin, C-reactive protein, and leptin were collected in a training cohort (n=149) and correlated with clinical outcome. Metabolic risk groups were defi ned and tested in an independent validation cohort (n=167). Results: We identifi ed pre-transplant weight loss exceeding 2% and TSP lower than 70 g/L, as strong independent predictors of relapse and death. Patients in the metabolic high risk group (low TSP and weight loss >2%) had an increased risk for relapse (P=0.0002) and death (P=0.002), but a similar risk for acute GVHD. Weight loss coincided with reduced pre-transplant serum leptin levels. The adverse infl uence of weight loss and high metabolic risk on relapse and overall survival could be confi rmed in the validation cohort. Multivariate analysis of both cohorts revealed a hazard ratio for relapse of 7.78 (2.59-23.36, P=0.0003) in the metabolic high risk group. Discussion: Altered nutritional homeostasis prior to alloSCT correlates with recurrence of AML after transplantation. Studies addressing pre-transplant nutritional interventions in order to reduce AML relapse rates are warranted. Disclosure of Interest: None Declared.

M. Guenova 1,* , P. Ganeva 1 , G. Balazenko 1 , A. Michova 1 , B. Spassov 1 , V. Varbanova 1 , G. G Mihaylov 1 , G. Arnaudov 1 1 

Introduction: The use of modern induction protocols allows for achievement of complete remission (CR) in a considerable number of patients with acute leukemia. Regardless of this the largest part of them relapse because of the persistence of leukemia cells, that cannot be revealed through conventional morphology methods of examination and which are defi ned by the term minimal residual diseases (MRD). In general relapse remains the leading cause for treatment failure, even after transplantation. Materials (or patients) and Methods: We present the results from studies of MRD in acute myeloid leukemia patients in the SHATHD Sofi a (CIC 859) Considerably larger evet free survival (EFS) and overall survival (OS) was established for patients with lower levels of MRD, assessed after induction treatment. Results: Analysis of MRD levels at the 100th day after allo SCTthrough 8-color fl ow cytometry in 24 acute leukemia patients reveals also considerably longer EFS (long rank test, P-0=003) and OS (long rank test, P-0=03 (inpatients with levels below 0=05% residual leukemia blasts. Discussion: The results suggest that MRD evaluation during treatment of acute leukemia patients is signifi cant for their prognostic evaluation both as OS and as time to relapse. On the other hand there is an active debate over the possibilities for risk stratifi cation and therapeutic decisions on ground of MRD levels assessment, and the introduction if novel approaches to overcome leukemia resistance to standard therapeutic regimens and agents as well as strategies aimed at modulation of the allogeneic GVL eff ect. 

NKp46 expression at diagnosis on clinical outcome in acute myeloid leukemia (AML) patients with allogeneic hematopoietic stem cell transplantation (HSCT). Materials (or patients) and Methods: NKp46 expression on peripheral blood NK cells was prospectively assessed by fl ow cytometry in 125 AML patients. Median age at diagnosis was 48 years (range 20-65). Among these 125 patients, 56 patients received allogeneic HSCT. The threshold for NKp46 expression was set on dispersion criteria of the population. Clinical outcome was evaluated with regards to NKp46 expression. Results: In the group of allografted patients, patients whose NK cells highly expressed NKp46 at diagnosis had better overall survival (OS) (P=0.0229; HR=3.062; 95% CI = [1.168-8.031] ) and relapse free survival (RFS) (P=0.0482; HR=2.881; 95% CI = [1.009-8.228 ]) than patients with low NKp46 expression. Patients with high NKp46 expression had improved survival probabilities at 2 years compared with patients with low NKp46 expression (90% vs 54%, respectively). No clinical benefi t was observed for non-allografted patients with high NKp46 expression compared to non-allografted patients with low NKp46 expression, suggesting that NKp46 expression is a predictive biomarker of graft outcome rather than a prognostic biomarker. Discussion: In this study we have identifi ed NKp46 as a potential biomarker predictive of HSCT outcome. Clinical translation of these results could fi nd applications in terms of identifi cation of patients at high risk of adverse outcome. Moreover, these results highlight the potential of therapeutic strategies aiming at maintaining high levels of NKp46 on NK cells after allogeneic HSCT. Finally, this biomarker could be used as a stratifi cation criterion in clinical trials. Disclosure of Interest: None Declared. Most of the patients (46/52) received 12Gy TBI based conditioning regimen in combination with VP16 or CY. The rest of the patients received BU based MAC regimen. For graft versus host disease profi laxis most of the patients (45 pts) received tacrolimus and short sirolimus (day -1 -day 30), the remaining patients received cyclosporin A + short methotrexate or tacrolimus + mycophenolate mofetil for GVHD profi laxis. Results: One graft failure occurred. 9/52 patients died before day 100. In 23/52 patients did not developed acute graft versus host disease, 7/52 grade 1., 17/52 grade2., 4/52 grede.3. and 1/52 grade.4. acute GVHD appeared. 20/43 were free from chronic GVHD, 9/43 patients had limited chronic GVHD, 14/43 patients had extensive chronic GVHD . Eleven patients relapsed, from that 9 patients died subsequently due to relapse, 2 patients still alive after salvage chemotherapy and with GVHD induced by donor lymphocyte infusion. The log rank statistical analysis of the risk factors for survival showed the most robust eff ect of chronic GVHD (Kaplan Meier estimated survival w/o cGVHD: 42% vs. lim.cGVHD: 89% P=0.007 or vs. ext.cGVHD: 57% P= 0.14; lim.cGVHD vs. ext.cGVHD hazard risk:2.1, P=0. 3) . Additionally important proved to be the EBMT risk score (KM survival with score 0-3: 56% vs. with score 4-6: 25% HR: 0.33 p= 0.02), the pretransplant remission status (KM survival CR1: 58%, CR2: 37%, relapse 1: 0%, advanced phase: 33%; only signifi -cant is the CR1 survival benefi t from the others), the use of VP16 in the conditioning regimen (KM survival 65% vs. 40%, P= 0.039) and the eff ect of acute GVHD (KM survival ac.GVHD0-2: 50%, acGVHD3-4: 20%, P= 0.034). Discussion: The most advantageous combination of diff erent factors to get cured are: disease in CR, TBI/VP 16 conditioning regimen, no acute GVHD, but limited chronic GVHD (the most important factor). The most robust eff ect of limited chronic GVHD of being cured gives room for less toxic conditioning regimens and chronic GVHD in combination. GVHD titration is still the most important and most challenging task of allogeneic stem cell transplantation. Disclosure of Interest: None Declared. 1,2 1 University of Calgary, 2 Oncology, Alberta Children's Hospital, Calgary, Canada Introduction: Juvenile myelomonocytic leukemia (JMML) is a myeloproliferative neoplasm diagnosed commonly in children under the age of 2 years. Signs and symptoms of this disease include anemia, fatigue, lymphadenopathy, hepatomegaly, splenomegaly and thrombocytopenia. Allogeneic hematopoetic stem cell transplantation (HSCT) is the only known curative therapy, with no standard conditioning regimen. While myeloablative conditioning is considered necessary to eradicate stem cell clones, treatment related mortality rates are high with busulphan, cyclophosphamide and total body irradiation (TBI)-based conditioning regimens, with many transplant physicians reluctant to administer TBI to young children. Recently, treatment of acute myeloid leukemias in adults with a less toxic regimen consisting of fl udarabine and myeloablative doses of busulfan (Bu/Flu) have proven to be eff ective in reducing the high rates of transplant related morbidity and mortality, while promoting engraftment and survival. Despite this growing evidence supporting the use of Bu/Flu in transplantation for myeloid leukemias, concerns persist that this regimen is not suffi ciently myeloablative to be curative in a disease such as JMML. Materials (or patients) and Methods: We report a case series of 3 children diagnosed with JMML between 7 months and 2 years of age who were conditioned with our institutional protocol of busulphan (16mg/kg total), fl udarabine (250mg/m 2 total), rATG (5.5mg/kg total) and 400 cGy of TBI followed by allogeneic stem cell transplant. Results: All patients presented with leukocytosis, monocytosis, thrombocytopenia, anemia, and splenomegaly. Patient 1 was diagnosed with JMML while she was admitted for an acute respiratory infection, whereas patient 2 presented with unexplained fevers and abdominal distension. In contrast, patient 3 was diagnosed during a routine examination, however he presented with more serious symptoms. In addition to leukocytosis, multiple cytopenias and massive hepatosplenomegaly, this patient developed respiratory symptoms soon after diagnosis, suggestive of pulmonary JMML infi ltration. Post transplant, patients 1 and 2 are cured of JMML with full donor myeloid chimerisms, 2 and 4 years post HSCT respectively. Both patients had peripheral blood stem products. Patient 1 had acute graft versus host disease of the skin and gut, and RSV and ventilator induced chronic lung injury. Patient 2 did not have any major complications following transplant. The third patient received an umbilical cord blood HSCT, and while initially engrafted with 86% whole blood donor chimerisms, eventually died of hemorrhage post liver biopsy. Post-mortem examination confi rmed recurrence of JMML in bone marrow in addition to a T-cell lymphoproliferative disease in the liver. Discussion: Novel conditioning regimens are needed to reduce transplant related mortality while maximizing tumor eradication and engraftment in patients with JMML undergoing HSCT. Our Introduction: Aggressive T-cell lymphomas represent 10% to 15% of non-Hodgkin's lymphomas in adults. Patients with relapsed or refractory disease are generally considered incurable with conventional therapies. ATG had been used in the conditioning regimen to reduce the incidence of GvHD for a long time especially in the matched unrelated donor HSCT. The early experiment result in our hospital showed that ATG inhibited the proliferation of lymphoid tumor cells in a dose-dependent manner especially in the T cell tumors. We used the ATG as the part of the conditioning regimen in the all patients and to evaluate the long-term antileukemia eff ect, the safety and complication in the patients with relapsed or high-risk T cell lymphomas. Materials (or patients) and Methods: 18 patients( male 9, female 9) were enrolled into this study. Median patient age at the time of transplantation was 28 years (range, 7-55 years) . At the time of transplant, 4 patients reached fi rst or subsequent complete response (CR) with conventional therapy or the salvage therapy, 4 patients had a partial remission(PR), 7 patients had relapsed disease not responding to salvage therapy or progressive disease, and 3 patients had primary refractory disease. Donors were 10/10 HLA matched related (5), 10/10 matched unrelated (3) , 8/10 matched unrelated (5) and mismatched related (5). The median number of CD34+ cells within the allografts was 9.31/kg body weight (BW) (range, 4.6-24.85/kg BW). Rabbit antithymocyte globulin (ATG 2.5 mg/kg×4 days) and total-body irradiation (10 Gy in fi ve fractions) were used in all 18 patients. All patients but one also underwent cyclophosphamide (120 mg/kg). The only one who did not receive CTX had experienced with autologous transplantation. Fifteen high risk patients were in additional use of VP 16 or VM26 30-40 mg/kg. Two patients in CR1 and one patient who were 55 years old had not received VP16. Graft-versus-host disease (GVHD) prophylaxis was cyclosporine based, usually in combination with methotrexate. Quantitative chimerism analyzes were performed using short-tandem-repeat-based polymerase chain reaction techniques at regular intervals for every 4 weeks after transplantation in bone marrow at the fi rst six months. Results: All patients but one achieved a complete remission in the fi rst three months after allogeneic HSCT. One patient achieved PR on month 1 and soon died from progressive disease. The CR rate after transplantation was 94.4%. At a median follow-up time of 13 months, fourteen(77.8%) patients are alive. OS at three years was 71%. Still four patients were died after transplantation, two from relapse and two from treatment related complications. Acute GvHD grades II-IV occurred in eight patients( 50%) and grades III-IV in four patients. Two patients suff ering from acute GvHD grade IV died due to treatment-related complications. The maximum cumulative incidence of cGVHD was 42.8%. The high levels of CMV DNA were detected in seventeen patients in the fi rst three months after transplant. Two of them gradually envolved into viral haemorrhagic cystitis. One patient suff ered from aspergillus pneumonia while another two suff ered from bacterial pneumonia. There was no case of venous occlusive disease. Discussion: Anti-thymocyte globulin (ATG) could improve the outcome of allogeneic hematopoietic stem cell transplantation in patients with aggressive T cell tumors. Disclosure of Interest: None Declared.

Introduction: Acute Myeloid Leukemias (AML) recurring after allogeneic stem cell transplantation (allo-SCT) have a very dismal prognosis. Conventional salvage treatments may include intensive chemotherapy eventually followed by donor lymphocyte infusions (DLI) or a second allo-SCT. Although with intensive chemotherapy a second complete remission (CR) may be achieved, its duration is often short and its toxicity may preclude any further consolidation or a second allo-SCT. On the other hand, DLI alone are often unable to control a leukemic recurrence. Azacytidine is a demethylating agent which has proven effi cacy in high-risk myelodisplastic syndromes and AML, especially those with low disease burden at onset (blast cells infi ltration 20 -30%). Clinical data suggest that Azacytidine may induce responses in up to 30% of these patients and biological data suggest that, apart from the direct demethylating, antiprolipherative and cytotoxic eff ects on the leukemic cells, it might also infl uence the donor immune system, enhancing the Graft versus Leukemia (GvL) eff ect. Materials (or patients) and Methods: We report here a series of 6 patients submitted to allo-SCT for AML in fi rst (n=4) or second (n=2) CR. Median age at transplant was 50 years (range 41-62). Three patients received a myeloablative conditioning regimen. The donor was a matched sibling in 2 cases and a matched unrelated in 4 cases. Two of these patients developed grade II aGVHD after allo-SCT, successfully treated with standard steroid. The disease recurrence was observed at a median of 7 months from allo-SCT (range: 3 -48) . At relapse, the disease presented with blast cell marrow infi ltration, not exceeding the 20% in all the cases. All the patients had already stopped the immunosuppressive treatment with the exception of one of those who developed aGVHD who was tapering the steroid. Results: Azacytidine was administered at the conventional dose (75 mg/sqm sc/day for 7 days every 28 days). The fi rst response evaluation was performed after the fourth cycle. Patients showing a response or a stable disease continued the treatment until disease recurrence or progression. The median number of cycles administered is 4 (range 1 -12) . Two patients are not evaluable for response and toxicity as the have just completed the fi rst two treatment cycles. Among the four evaluable patients, 1 achieved a CR that lasted for 18 months and is alive with active disease, 1 maintained a stable disease (SD) for 3 months and then progressed and died and 2 were non responders and died for disease progression. Overall the treatment was well tolerated. Two of the 4 evaluable patients (50%) developed a grade II-III WHO haematological toxicity which was treated with study drug dose reduction.

No grade III-IV WHO extra-hematological toxicity was observed. Discussion: Our experience, although limited, suggest that azacytidine may be eff ective and is well tolerated in patients with AML relapsing after allo-SCT. In particular, long lasting remission Introduction: It is a great problem to get complete response in patients with resistant acute myeloid leukemia (AML) before allogeneic stem cell transplantation. There are many patients with resistant AML treated with high dose chemotherapy unsuccessfully. On the other hand it is a possibility to decrease the somatic status and to get much more comorbidities after some course of aggressive chemotherapy which may exclude patient from the group of candidates to stem cell transplantation. It has been shown the capability of cytarabine (Ara-C) combined with purine analogues to improve the results of AML patients treatment. We used combination of low dose Ara-C with cladribine to prepare intensively pretreated patients with resistant AML before allogeneic stem cell transplantation. Materials (or patients) and Methods: We used the scheme of Ara-C 20 mg sc bid for 14 consecutive days combined with cladribine 5 mg/m/2 IV once a day during the fi rst 5 days. The therapy was begun after informed agreement was signed. Results: Six patients aged 24-58 years (Me 55 years) were treated with combine chemotherapy. One of the patients did not get any response after induction chemotherapy 7+3 and 2 courses of high dose chemotherapy. Other 5 patients had relapsed AML after diff erent period of complete response including a patient after autologous stem cell transplantation. Normal karyotype was detected in 2 patients. Other patients had diff erent chromosomal aberrations including a patient with complex karyotype with 2 monosomies which was detected during relapse for the fi rst time. Nobody of patients had FLT3, NPM1 and c-Kit mutations. Coexpression of CD4, CD22, and CD7 was determined on the myeloblasts of 3 patients. Combined chemotherapy was effective in 3/6 patients. The response was confirmed after the first course. The patients with response were treated with additional induction course. Then one patient was treated with high dose consolidation. Allogeneic stem cell transplantation fully matched related and unrelated donors was performed in 2 patients. The period from obtained response till the start of conditioning regimen was 2 and 4 months. Myeloablative regimen Bu + Cph was chosen for both patients. The number of bone marrow blasts was increased to the level higher than 5% in one patient before the initiating of conditioning regimen.Complete response with full donor chimerism is preserved during 3 and 16 months after stem cell transplantation. Discussion: We conclude that low intensity chemotherapy like combination of low dose Ara-C with cladribine may be a very eff ective kind of treatment in some patients with resistant variant of AML who were aggressively pretreated. We suppose that to save the response during the pretransplant period it is necessary to use high dose chemotherapy after complete remission was revealed. The mechanism of action of low intensity chemotherapy in patient with resistant AML is unknown. One possible explanation is absence of unfavorable mutations irrespective of cytogenetic fi ndings. Disclosure of Interest: None Declared. Introduction: The cytogenetic profi le is an important prognostic factor in the myeloid malignancies. Monosomy 7 and/or a complex karyotype are known to correlate with poor prognosis and poor disease-free survival when a conventional approach is undertaken. We sought to analyse the role and outcome of hematopoietic stem cell transplantation (HSCT) in poor risk AML and MDS in adults in our centre. Materials (or patients) and Methods: A retrospective case analysis was performed for 28 patients (10 female, 18 male) who were treated at Manchester Royal Infi rmary between Jan 2007 and Dec 2012. We have analysed the outcome of patients with cytogenetically defi ned poor-risk AML or MDS treated with chemotherapy followed by matched related (MRD), or voluntary unrelated donor (VUD) transplantation in CR1 or CR2. Most of our patients received 1 line of chemotherapy (n=16), 11 received 2 lines and only 1 patients required a 3 rd line. The median time from diagnosis to HSCT was 8 months (range 1-88). Poor-risk cytogenetics were defi ned as monosomy 7 and/or complex karyotype as per MRc AML criteria. Primary diagnoses were AML (n=13), therapy-related AML (n=8), MDS (n=5) and therapy-related MDS (n=2). The median age was 53 years (range 23 and 72). After initial chemotherapy, 20 patients received a reduced intensity conditioning (RIC) with Fludarabine Melphalan Campath (FMC) or Fludarabine Busulphan Antithymocyte globulin (FluBu-ATG)conditioning regimens and 8 received myeloablative contidioting (MAC) with diff erent conditioning regimens (BuCy; Cy-Total body irradiation(TBI); BuCyCampath or FluCyTBICampath). A total of 16 patients received VUD, 2 umbilical cord blood (UCB) and 10 MRD. Results: The median follow-up of our cohort was 26 months (range . The overall survival (OS) at 1 year, 2 years and 5 years was 89%, 63% and 36% respectively with a transplant-related mortality (TRM) of 11% at 1 year and 15% at 2 years. The TRM was accounted for by cytomegalovirus reactivation or sepsis with subsequent multi-organ failure. The non-TRM was predominantly secondary to leukaemic relapse. The progression-free survival (PFS) was 82%, 63% and 30% at 1, 2 and 5 years respectively. Of the total of 28 patients, 8 developed acute graft-versus-host disease (GVHD) of the skin, 3 of gastro-intestinal tract and 3 of the liver with no GVHD-related mortality. Discussion: Our fi ndings show that HSCT is a valid option in AML and MDS with adverse cytogenetics with potential long term disease remission and improved overall survival. Disclosure of Interest: None Declared.

Introduction: Improvements in genetic characterization of acute myeloid leukemia (AML) have allowed risk-adapted post remission treatment for AML. In this regard, the main disease aspects considered are cytogenetics and molecular profi le, mainly NPM1 and FLT3 gene mutations. We have analyzed the results of autologous hematopoietic cell transplant (AHCT) in patients with normal karyotype (NK) AML without FLT3 internal tandem duplication (FLT3-ITD) or other adverse features, depending on the presence or not of NPM1 mutation. Materials (or patients) and Methods: Patients had been enrolled in the AML-99 and AML-03 CETLAM trials. We included patients with NK, no FLT3-ITD, who had achieved complete remission (CR) after a single induction course and, in the most recent trial, without MLL rearrangements and minimal residual disease (MRD) after consolidation chemotherapy. In the two trials the intention in these patients was to perform an AHCT, regardless the availability of an HLA-matched donor. One-hundred eleven patients fulfi lled the mentioned criteria. Characteristics of these patients were: 54 male (49%) and 57 female (51%); 63 patients ≤50 years-old (yo) (57%) and 48 >50 yo. (43%); NPM1 mutation was observed in 53 patients (48%) and was absent in 58 (52%). Fifty nine of the 111 patients fi nally received the AHCT. Characteristics of transplanted patients were: 28 male (47.5%) and 31 female (52.5%); 39 patients ≤50 yo (66.1%) and 20 >50 yo (33.9%); NPM1 was mutated in 35 of these patients (59.3%) and non-mutated in 24 patients (40.7%). The causes of the 52 patients who did not receive the planned AHCT were: stem cell mobilization failure (n=25, 48%), allogeneic procedure (n=2, 4%), prolonged aplasia (n=4, 8%), consolidation chemotherapy due to NPM1 mutation (protocol deviation, n=4, 8%), relapse before AHCT (n=8, 15%) and other non-specifi ed causes (n=9, 17%). Results: Five-year overall survival (OS), disease free survival (DFS) and cumulative incidence of relapse (CIR) in the 111 patients (intention to treat AHCT) were: 61±5%, 51±5% and 45±3% respectively. When comparing patients with or without NPM1 mutation, OS was 71±7% vs 51±7% (p=0.068), DFS 61±7% vs 41±7% (p=0.008) and CIR was 35±7 vs 54±8 (0.004), respectively. Similarly, 5 years OS, DFS and CIR in the 59 patients who fi nally received the AHCT were 59±7%, 48±7% and 47±5%, respectively; the values according to NPM1 status were: OS 65±9% in patients with NPM1 mutated vs 48±11% in NPM1 not mutated (P=0.179), DFS was 55±9% vs 39±11% (P=0.071) and CIR was 38±8% vs 61±11% (P=0.031). Discussion: Patients with NK, NPM1 mutation and FLT3 wt should be considered as a favourable prognosis. Patients with this characteristics and AHCT as an intention to treat had better DFS and inferior CIR compared with those NPM1 not mutated. Furthermore, patients with NPM1 mutation who fi nally were transplanted had a signifi cant inferior CIR. AHCT had a role in intensifi cation therapy in these patients. Disclosure of Interest: None Declared. Introduction: Relapse of myeloid malignancies after allogeneic hematopoietic stem cell transplantation (alloHSCT) is associated with a very poor prognosis. The aim of this study was to assess the eff ect of 5-azacitidine (5-aza) administration as prophylaxis and treatment of relapse after alloHSCT. Materials (or patients) and Methods: Results of treatment with 5-aza after alloHSCT for 20 patients with myeloid malignancies (AML 80% (16), MDS 20% (4)) was evaluated. 2 patients were grafted from matched family donor, 16 -matched unrelated donor, 2-haploidentical). Median age was 26 years, range 4-56, 11 male, 9 female. Conditioning regimen was myeloablative in 4 (20%) cases, reduced intensity conditioning was used in 16 (80%) cases. In 50% (10) cases the administration of 5-aza was prophylactic, because of the high relapse risk: 6 patients had advanced disease at the moment of alloHSCT, the unfavorable cytogenetic was detected in 3 cases, 1 patient had the minimal residual disease at the moment of alloHSCT. Median time for administration of 5-aza was day+30 -day+60 after alloHSCT; the main criteria was the recovery of hemopoiesis; 5-aza was injected subcutaneously 35mg/m 2 /daily, 7 days of 28-day cycle, 4 cycles. In case of relapse (bone marrow blasts > 5%) the median number of 5-aza cycles administrated was 2, range 1-6. As the therapy of relapse 5-aza was combined with donor lymphocyte infusion or chemotherapy in 80% (8). Median time of administration was day+165 . Results: Median duration of follow-up in both groups was 336 days . In the group of prophylactic administration 80% patients are still alive in complete remission, 20% died because of late relapses, no facts of treatment related mortality (TRM) were registered. Median duration follow-up in this group was 351 day (133-1035). 5-aza was used as a treatment of relapse in 10 (50%) patients. Remission was achieved in 20% (2 patients), 8 patients had progressive disease. 2 patients are still alive, median of follow-up in this group is 322 (65-1270) days. 8 patients died: 6 (75%) from relapse, 2 (25%) from TRM (sepsis, myocardial infarction). 5aza therapy was well-tolerated, no severe hematological and nothematological toxicity was observed. Discussion: Based on our preliminary analysis 5-aza is eff ective for prevention of relapse and improvement of overall and disease-free survival in a high-risk group. Administration of 5-aza as a therapy for relapse had relatively low effi cacy. Further larger prospective, randomized trials are needed for more solid conclusions. Disclosure of Interest: None Declared.

Introduction: It has been assumed that HSCT recipients with an unrelated or HLA mismatched donor run a reduced risk of posttransplant relapse of acute leukaemia. We here suggest that this may be otherwise. In addition, we report that AML recipients with a high pre-conditioning CD4 T cell count have an increased relapse risk. Materials (or patients) and Methods: This is a single centre study on children and adults with AML (N=89) or ALL (N = 100) in 1 st or 2 nd remission who received non-T-depleted myeloablative HSCT between 1998 and 2006. Mean recipient age was 26 yrs. (range . Donors were either an HLA identical sibling (Id. Sib.) or an alternative (unrelated) donor. Alternative donors were either identical at 10/10 loci by high resolution HLA typing or had documented or non-excluded HLA mismatch. The pre-transplant conditioning was TBI 12 Gy (87%) or i.v. Busulfex (13%) with cyclophosphamide or VP16. Antithymocyte globulin (ATG, ATGAM 20 mg pr.kg pr. dose, or Thymoglobuline 2.5 mg pr.kg pr. dose) was given for a total of three doses day -3 to day -5 before HSCT. ATG was given to none of 77 patients with Id. Sib. donor, and to 93 of 112 patients with alternative donors. Blood CD4 and CD8 T-cells were assessed by fl ow cytometry prior to start of conditioning and their eff ects on relapse were analyzed as continuous variables using Cox multivariate regression.

Results: During analysis of risk factors for post-transplant relapse of acute leukaemia we noted an increased risk associated with recipient pre-transplant CD4 T-cell count (high worse than low), and with the use of an alternative donor. In contrast, the use of ATG during conditioning reduced the risk. We further analysed these eff ects in an unselected group of 189 recipients with AML or ALL in 1 st or 2 nd remission. Main results include: 1) The increased relapse risk associated with CD4 was restricted to AML recipients.

2) In recipients with AML and Id. Sib. donor (N = 44) the relapse hazard rate (HR) was 5.8 (P = 0.028). No similar eff ect was found in ALL. 3) In patients with any donor (Id. Sib. or alternative) the eff ect of CD4 was still seen in AML patients, but only in recipients who did not receive ATG during pre-transplant conditioning (N = 53. HR = 8.04, P = 0.004). The adverse eff ect was completely abrogated in ATG conditioned AML patients. In ALL patients no eff ect of CD4 was observed regardless of ATG (N = 100). 5) There was an increased relapse risk associated with donors with non-excluded or documented HLA mismatch. This adverse donor eff ect was present both in AML and ALL and was also abrogated by ATG (AML + ALL without ATG: N= 96, HR = 16.36, P = 0.0001). Discussion: High pre-conditioning recipient CD4 T lymphocyte count was associated with increased relapse in AML. This eff ect was abrogated by ATG during conditioning. An adverse eff ect was associated with alternative donors. We therefore suggest 1) that CD4 T lymphocytes of recipient origin survive conditioning regimens, if ATG is omitted. 2) Surviving recipient CD4 T lymphocytes may contribute to a reduced graft versus leukaemia eff ect in AML.

3) HLA mismatch may potentiate this adverse eff ect. 4) When mismatched donor HSCT in the past has been associated with a reduced relapse risk, this eff ect could in some cases have been a result of ATG rather than the mismatch. Disclosure of Interest: None Declared. S110 and Leipzig for refractory AML between 2006 and January 2013. All patients had active disease prior to HCT, 46 patients with a median of 25 (range 5.3 -90) % bone marrow blasts and three patients with extramedullary myeloid blast infi ltration or blasts in peripheral blood. Primary induction failure (PIF) was the reason for HCT in 20 patients while 29 patients had fi rst or higher refractory relapse. Thirty one patients had one or two lines of pre-treatment while 18 patients had more than two lines of pre-treatment. Conditioning for HCT was performed with chemotherapy consisting of fl udarabine (4x30mg/m 2 ), cytarabine (4x2g/m 2 ) and amsacrine (4x100mg/m 2 ) followed 4 days later by 4 Gy total body irradiation in 38 patients or busulfan (8x1mg/m 2 ) in 11 patients, combined with cyclophosphamide (120mg/m 2 ). Thymoglobine (3x2mg/m 2 ) was given when unrelated donors were used (FLAMSA-RIC). Results: Estimated overall survival (OS) and event free survival (EFS) at 3 years after a median follow up of 36 (range 6-71) months was 16% and 11%, respectively. Causes of death were relapse in 59%, infection in 14% and graft versus host disease (GvHD) in 10% of all patients. Twenty eight patients (57%) achieved CR four weeks after HCT while nine patients had partial remission (PR), nine patients had stable disease (SD) and three patients had either no evaluable bone marrow examination or died before due to infections. Another seven patients with PR and SD achieved CR (overall CR rate 71%) from four weeks to day 90 after HCT following reduction of immunosuppression. Risk factors for OS were blast count of ≥ 20% prior to HCT with an OS at 3 years of 4 % compared to 34% in patients with blast counts of < 20% (P=0.02) and a trend for patients with ≥3 vs. <3 lines of pre-treatment (5.6 vs. 22%; P=0.08). Pre-treatment lines were also a risk factor for EFS. Patients with <3 lines of pre-treatment had a better EFS of 18% than patients with ≥3 lines (0%; P=0.05). Limited chronic GvHD resulted in EFS of 29% compared to 15% in patients without chronic GvHD (P=0.06). This was due to a lower RI of 48% compared to 76% (P=0.06). Combining patients with low blast count and less than 3 lines of pre-treatment (n=15) resulted in an OS of 38% at 3 years, compared to 7% in the other group (P=0.02). Occurrence of acute GvHD and cytogenetic risk factors did not infl uence OS, EFS or RI signifi cantly. Discussion: In this study we document an OS of 16% at 3 years for patients with refractory AML after HCT with FLAMSA-RIC. Outcome is dependent on blast count prior to HCT, lines of pre-treatment and the incidence of chronic limited GvHD. Disclosure of Interest: None Declared. Results: Intermediate and unfavorable cytogenetics were found in 303 (88.3%) and 40 patients (11.7%), respectively. The median follow-up of survivors was 28 months. The 3-year OS for patients with intermediate and unfavorable cytogenetics who underwent allo-SCT were 66.7%(SE 3.6%) and 30.3%(SE 10.6%), respectively (P= 0.001) and for the patients who underwent auto-SCT in these two cytogenetic groups were not statistically signifi cant.The cumulative incidence function (CIF) of relapse at 36 months after allo-SCT in the intermediate and unfavorable groups were 26%(95% CI:20-32.5%) and 68 Discussion: Our study showed that outcome after allo-SCT in de novo AML differs depending on cytogenetic risk-group; and confirmed the classification of chromosomal aberrations into intermediate and unfavorable groups was a better predictor of relapse rate and OS than almost any other tested parameters.

In the unfavorable group, the relapse rate was extremely high, and allo-SCT should be performed in CR1. The impact of cytogenetic risk group was not seen on TRM in our study. Apart from cytogenetic group, chronic GvHD was significantly higher in the intermediate risk group that can decrease the probability of relapse and increase the survival in this cytogenetic group after allo-SCT. In conclusion, the presence of unfavorable cytogenetic abnormality versus those with normal karyotype or intermediate cytogenetic abnormality associated with worse outcome after allo-SCT in AML patients and their treatment remains an unmet medical need. Further studies analyzing the impact of particular chromosomal aberrations in AML on patient's outcome after allo-SCT would allow for refined risk stratification in adults with de novo AML and update indications for allo-SCT in de novo AML. Disclosure of Interest: None Declared. S111 Results: 65/69 pts (94%) received one or two cycles of standard induction chemotherapy, 43 pts (66%) obtained the complete remission (CR), treatment related mortality was 15% (10 pts), 33 pts then received at least one cycle with high-dose cytarabine, 10 pts further standard chemotherapy. Overall, 29 pts (38%) received a SCT, alloSCT 14/27 (52%), autoSCT 15/50 (30%). Disease status at SCT: CR1 22, persistence of disease 2, upfront 5. Causes for failure of the program (48 pts): death before induction 3 cases, death during chemotherapy 11, death during donor search 2, no donor 1, clinical reasons 3, patient refusal 1, disease refractory to induction 11, early relapse 4, failure of autologous leukapheresis (LK) 12. Day-60 transplant related mortality was 10% (3/29, 2 allo-SCT, 1 autoSCT). At last follow up 25 pts (32%) were alive, 19 (25%) in CR. Median and estimated 3 yrs OS of all pts (77) were 392 days and 28%, respectively, median and estimated 3 yrs OS of transplanted pts (29) were 1155 days (161-2437) and 55%, respectively. Discussion: our analysis revealed that alloSCT was feasible more frequently than autoSCT, that is in one-half vs one-third of pts. Three principal causes explained the failure of the autoSCT program in most pts (83%): disease refractory to induction, death during induction and failure of LK. Causes of program failure for alloSCT pts were more heterogeneous. Survival of pts who received the programmed SCT, allo or auto considered together, was encouraging. In conclusion, pts with advanced age and high risk AL/MDS can be treated and gain survival with an intensive therapeutic program including a SCT. To increase the feasibility of a transplant program in these setting it is crucial to reduce chemotherapy toxicities, improve the rate of complete remission and of successful autologous LK. Disclosure of Interest: None Declared. Introduction: Acute leukemia remains one of the greatest challenges in oncology. Increasing evidence indicates that the bone marrow microenvironment (BMM) plays a pivotal role in both the pathogenesis and relapse of leukemia. Recent studies confi rmed galectin-3 (gal-3), a multifunctional member of the β-galactosidebinding protein family, is critical in leukemia drug resistance and patient's prognosis. Here we determined the role of gal-3 in the promotion of BMM-induced acute leukemia cells' (ALCs') survival, and the mechanisms involved. Materials (or patients) and Methods: Our study applied the coculture system with mesenchyml stromal cells (MSCs) to mimic the leukemia BMM in vitro. We validated our hypothesis in different sub-types of ALC lines including Reh, Sup-B15, Jurkat and Kasumi-1. Bone marrow was obtained from healthy adult donors with informed consent and MSCs were cultured and identifi ed as previously reported. Results: CCK-8 test demonstrated MSCs improved viable ALC number with or without cytotoxic agents Idarubicin (IDA) or Etoposide . PI/Annexin V assay confi rmed reduced apoptotic level of MSCs-conditioned ALCs induced by IDA or Vp-16 (P<0.05) ( Figure 1A ). To clarify the underlying mechanisms we performed Western-blot and real-time polymerase chain reaction. We found that compared to ALCs cultured alone, both mRNA and protein expression of gal-3 were signifi cantly up-regulated in MSCs-conditioned ALCs (P<0.05). Increased gal-3 levels correlated with Akt phosphorylation, β-catenin stabilization and higher expression of β-catenin target gene cyclin D1 (P<0.05) ( Figure 1B and 1C ). Then we used gal-3 small interfering RNA (siRNA) to silence its expression in ALCs to determine whether gal-3 modulates the protective eff ects of MSCs in ALCs. Compared with MSCs-conditioned vector-transfected Kasumi-1, we detected an increased apoptotic percentage against IDA and a decrease of β-catenin protein level in gal-3 siRNA-transfected ones ( Figure 1D ).

[PH-P023] S113 (haplo-HSCT) performed during the fi rst complete remission (CR1) in adults with Philadelphia chromosome-negative (Ph-negative) acute lymphoblastic leukemia (ALL). Materials (or patients) and methods: Patients were classifi ed as having high-risk if they met one of the following criteria at diagnosis: (1) adverse cytogenetic [t(4;11) , or complex karyotype (≥5 abnormalities)]; (2) older age (≥35 years); (3) high-leukocyte counts (≥30 × 10 9 /L for B-precursor ALL or ≥100 × 10 9 /L for T-precursor ALL), or (4) delayed CR1 (remission required more than 28 days of induction therapy). All other patients were classifi ed as low-risk. Results: The median time from diagnosis to transplantation was 180 days. All patients achieved neutrophil engraftment within 30 days after HSCT, and by 100 days, the cumulative incidence of platelet engraftment was 88.5%. The 100-day cumulative incidence of grade II-IV and III-IV acute graft-versus-host disease (GVHD) was 40.2% and 6.7%, respectively. The 3-year cumulative incidence of total and extensive chronic GVHD (cGVHD) was 61.7% and 28.4%, respectively. The 3-year cumulative incidence of relapse and non-relapse mortality (NRM) was 18.3% and 19.9%, respectively. Overall survival (OS) and disease-free survival (DFS) at 3 years were 69.5% and 63.5%, respectively. Patients with limited cGVHD experienced a signifi cantly better OS and DFS than those without cGVHD or those with extensive cGVHD. The outcomes were comparable between the high-and low-risk groups (Table 1 ).

Discussion: Comparable transplantation outcomes between highand low-risk Ph-negative ALL CR1 patients may be related to the stronger graft-versus-leukemia eff ect, the shorter time required for donor selection, and a convenient repeated access to donor cells when donor-derived cellular therapy is needed for the relapse. Disclosure of Interest: None Declared. Hematology and Bone Marrow Transplant Unit, San Raff aele Scientifi c Institute, Milan, Italy Introduction: Allogeneic haematopoietic stem cell transplantation (ASCT) is the only curative option for patients (pts) with high risk (HR) Acute Myeloid Leukemia (AML) and Myelodysplastic Syndromes (MDS). Elderly subjects (> 65 yrs), who represent the majority of AML/MDS pts, are often excluded from ASCT programs due to the related high risk of toxicities and mortality. Recent development of reduced-intensity conditioning regimens and improvement of supportive measures have made ASCT a feasible option for selected older pts. Published data suggest that ASCT from HLAmatched related donors improves the outcome of this population. Materials (or patients) and Methods: We analyzed data from 72 consecutive pts with HR AML/MDS who underwent ASCT in our Institution between 08/2004 and 04/2013. Transplanted pts included in this study fulfi lled the following criteria: (1) cytologically proven diagnosis of primary or secondary MDS or AML; (2) age ≥ 65 years; (3) Performance Status (ECOG) ≤ 2; (4) completion of the transplant procedure.

Results: Diagnosis: 57 AML and 15 MDS. Median age was 68 years (range 65-76 years). Pts' and transplant characteristics are listed in Table 1 . Median follow-up (FU) after ASCT among surviving pts was 36 months (range, 8 -48 months) . 69 out of 72 pts (96%) were in CR at day +30 after ASCT. At the last FU 35 out of 72 pts (49%) were alive, 32 (44%) in CR, 36 pts (51%) died for the Introduction: Patients (pts) with refractory acute myeloid leukaemia (AML) have a particularly dismal prognosis using conventional supportive therapy and chemotherapy. In this double-center retrospective observational study, sequential high-dose melphalan (HD-Mel) was used as part of conditioning regimen for allogeneic hematopoietic stem cell transplantation (aSCT). Materials (or patients) and Methods: 162 adult pts (median age 55, range 17 to 71 years) with refractory AML were transplantat. For all pts pretransplant assessment of transplantation (PAM) score, European BMT (EBMT) score, and Hematopoietic Cell Transplantation-Comorbidity Index (HCT-CI) was calculated. 107 pts (66%) underwent a total body irradiation (TBI) based conditioning regimen and 55 pts (34%) a chemotherapy based regimen after a median interval of 5 days (range 1-18 days) following HD-Mel application. HD-Mel dose was 140 mg/m 2 in 150 pts, and 200 mg/m 2 in 12 pts. 48 pts (30%) were transplanted with an identical sibling donor (ISD), and 114 pts (70%) with matched unrelated Introduction: Regimens like the "FLAMSA-protocol" that combine intensive induction and dose-reduced conditioning regimens for allogeneic blood stem cell transplantation (alloBSCT) have shown to give promising results in the treatment of patients with highrisk myeloid leukemia (HR-ML). We retrospectively looked for the infl uence of in-vivo T-cell depletion using ATG on GvHD, posttransplant immune reconstitution, relapse and survival. Materials (or patients) and Methods: One hundred and fi ve patients (pts, 58m, 47f, age 52 years, 20-68) with HR-ML (52 de-novo AML, 26 sAML, 18 MDS, 9 MPS) received Fludarabine (4×30mg/m2), Amsacrine (4×100mg/m2), Ara-C (4×2g/m2, FLAMSA), age adapted HD-Mel (100-200mg/m2) and alloBSCT (103 PBSC, 1 CD34+CB, 1 BM, median CD34+cells/kg 7.8 × 106, 2-38) from 39 related and 66 unrelated donors. A total of 73 patients (70%), most of them receiving grafts from unrelated donors, received ATG Fresenius (60mg/kg in 61 and 30mg/kg in 12 pts). Twenty-eight pts (27%) had untreated MDS (16), sAML (9) or MPS (3) . Among 77 previously treated pts 22 (27%) were in 1st CR, 12 (16%) in 1st PR, 9 (12%) had primary refractory disease and 34 pts (44%) had relapsed (21 untreated, 13 treated: 5 in 2nd CR/PR, 8 refractory). Results: One hundred and one pts had donor reconstitution of hematopoiesis after a median of 14 days (8-46), 3 died early and 1 had graft failure. At day +28, 101 pts (96%) were in CR. Median follow-up is 3.7 years (0.5-9.8), Overall survival (OS) at 1&4 years was 65. 5% (60.8-70.2) and 58. 8% (53.7-63.9) . The respective event free survival (EFS) was 55.0% (50. 1-59.9 ) and 36.3% (31. 1-41.5) . aGvHD occurred in 52 pts (50%) and cGvHD was seen in 44 pts (47%) surviving more than 100 days after SCT. Estimated relapse rate (RR) at 1&4 years was 34. 4% (29.4-39.4 ) and 54. 6% (48.8-60.4) . Estimated non-relapse-mortality rate (NMR) at 1&4 years was 14.9% (11. 3-18.5 ) and 17.5% (13. 2-21.8) , respectively. Looking for prognostic factors we found that OS, EFS and RR were infl uenced by disease status, previous therapy and cytogenetics. Age and HCTCI were predictors for NRM. ATG reduced cGvHD (P <.001), but delayed hematopoietic and immune reconstitution (P<.001 for WBC and PLT reconstitution, P<.001 and p=.002 for lymphocyte count at day + 28 and +100, P <.001 for CD3CD4-cell count day 50-400) and increased RR (P=.014). Due to the fact that therapy of relapse was eff ective in many patients the diff erence in OS (4yOS 55% vs 68%) was not signifi cant. Discussion: Sequential conditioning using FLAMSA-Melphalan is safe and eff ective in patients with HR-ML. Final results are still compromised by disease recurrence in a considerable number of patients. Our results indicate, that in the setting of HR-ML, the benefi ts of high-dose ATG in terms of reduced cGvHD are accompanied by slow immune reconstitution and an increased relapse rate. Therefore studies are needed to tailor ATG-dose to disease risk. Disclosure of Interest: A. Groten Confl ict with: Fresenius, Travel support, T. Schroeder: None Declared, A. Dienst: None Declared, K. Nachtkamp: None Declared, E. Rachlis: None Declared, U. Germing: None Declared, R. Haas: None Declared, M. Kondakci: None Declared, G. Kobbe Confl ict with: Fresenius, Travel support. S116 Results: A median of 7 x10 6 CD34+/Kg and 2.7 x10 8 CD3+/Kg cells were infused. 5 pts were not evaluable for engraftment. The overall engraftment rate was 89%; median time to neutrophil and platelet engraftment was 17 and 16 days respectively. 2 pts experienced primary graft failure and 1 patient never reached platelet engraftment. Transplant related mortality (TRM) at 100 days, 1 and 2 years were respectively 15.4%, 22.6% and 27.15%. At the last follow-up 28/71 patients experimented relapse at a median of 5,7 months (and died because of it after a median of 15,6 months) and 23/71 were alive in CR with a median follow-up of 35.4 months. Acute Graft-versus-Host-Disease (GvHD) occurred in 21/71 pts of grade II-IV, in 14/21 of grade III-IV. Chronic GvHD occurred in 12 pts (according to NIH evaluation 7/12 moderate and 5/12 severe). 2 years OS and RFS were 42.7% and 36% respectively. In multivariate analysis only grouped complex plus monosomal karyotypes and phase other than CR correlated with inferior OS (P= 0.03, RR 2.77 and p= 0.03, RR 2.73 respectively) and RFS (P= 0.003, RR 3.59 and P= 0.04, RR 2.5 respectively) . No diff erences among donor types, ages and comorbidity scores were found. Discussion: Our study (which includes a relevant proportion of pts >60 years old) supports allo-SCT as a feasible and potentially curative option for sAML. Allo-SCT provides better outcome to very high risk pts in 1 st CR; it is mandatory to consider transplant in the treatment algorithm of such pts. Disclosure of Interest: None Declared.

M. Gramatzki 1,* , C. Kellner 1 , M. Peipp 1 , N. Schub 1 , A. Humpe 1 , M. Staudinger 1 1 

Introduction: In AML patients despite intensive chemotherapy and additional hematopoietic stem cell transplantation, residual leukemic stem cells (LSC) may lead to relapse. Therefore, elimination of LSC by targeted therapy may represent a promising therapeutic approach. Recently, CD96 was identifi ed as marker antigen on AML LSC (Hosen et al., PNAS 104:11008, 2007) . Here, strategies for engineering autologous stem cell grafts as well as for in vivo targeting of residual AML stem cells by addressing CD96 for magnetic cell sorting (MACS) or antibody dependent cellular cytotoxicity (ADCC) are described. Materials (or patients) and Methods: CD96 antibodies containing a human IgG 1 Fc portion optimized for Fc receptor binding and wild type or affi nity maturated CD96 binding domains were constructed by recombinant DNA technology. Parental CD96 antibody TH111 was raised in our laboratory (Gramatzki et al., Exp. Hematol. 26:1209 , 1998 as well as recombinant CD96 antibodies were enriched by affi nity purifi cation from supernatants of hybridoma TH111 or transgenic cell cultures, respectively. For MACS based cell sorting of CD96-positive target cells, parental CD96 antibody was biotinylated using a biotinylation assay targeting free amines. Analysis of antibody binding, and biotinylation as well as subsequent diff erentiation of cell populations before and after cell sorting was done by fl ow cytometry. The potential of stem cells to proliferate and diff erentiate was analyzed by colony forming assays. Lytic properties of recombinant affi nity optimized CD96 antibodies were evaluated using purifi ed NK cells of healthy donors and CD96-positive target cells in standard chromium release assays. Results: To evaluate the effi cacy of purging LSC by MACS technology, stem cell containing grafts (n=10) were spiked with CD96positive AML cells. Up to a 1000-fold depletion of targeted cells was achieved using biotinylated CD96 antibody TH111 in combination with anti-Biotin-microbeads (Miltenyi, Bergisch Gladbach, Germany). Viability, cell count and the potential of HPC to proliferate and diff erentiate were not aff ected by the cell sorting procedure. Eradication of AML stems cells is also an issue after allogeneic stem cell transplantation. To target CD96-positive AML-LSC by ADCC, chimeric antibodies containing wild type or affi nity maturated variable regions in combination with an optimized human IgG 1 Fc were generated. As shown by fl ow cytometry, the antigen binding affi nity of the recombinant maturated antibody was enhanced (EC50 0.6 μg/ml vs. 2 μg/ml). Moreover, also NK cell mediated lytic properties against CD96-positive target cells were found elevated (EC50: 0.02 μg/ml vs. 0.15 μg/ml) as analyzed in standard ADCC assays. Discussion: By removing AML-LSC this CD96 purging strategy may allow a revitalization of autologous hematopoietic stem cell transplantation for certain patients with AML. Furthermore, CD96 antibody optimized for antigen as well as Fc receptor binding aims at using CD96 in vivo, possibly also in the allogeneic situation. Disclosure of Interest: None Declared.

Introduction: Relapse after allogeneic hematopoietic stem cell transplantation (HSCT) is mostly associated with a dismal prognosis in patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). In fact, the considerable toxicity of conventional chemotherapy often limits the potential success of conventional reinduction therapy. A 10-day regimen with decitabine (DAC), a demethylating agent, has been recently shown to be associated with a considerable response rate in patients with relapse after conventional chemotherapy. This approach has not been reported after allogeneic HSCT so far. Materials (or patients) and Methods: We report on ten patients with either AML (n=9) or MDS (n=1) and a median age of 55 years (range 37 -64) relapsing a median of 13 months after HCST (range 4 -79) . The median number of prior therapies was 3 (range [2] [3] [4] [5] . Notably, four patients had even progressed to a prior therapy with azacytidine. All patients received an outpatient-based 10 day regimen of DAC at a dose of 20 mg/m2 every 4 weeks. Dose was adopted subsequently depending on response and toxicity. Donor lymphocyte infusion (DLI) was given on day 11 of the DAC schedule only in case of absence of GVHD. Results: A median of 2 treatment cycles (range 1 -5) were administered. In six patients DAC was followed by at least one DLI infusion while 2 patients underwent subsequent second transplantation following a conditioning with 2 Gy TBI. The median blood donor chimerism increased from 70 % (range 18 -95) before treatment to 98 % (range 64 -100) after the fi rst cycle of DAC. At a median follow up of 4 months (range 2-9) after the fi rst cycle fi ve out of ten patients (50%) are alive and leukemia-free. One patient died of progressive disease and two patients died of infectious complications (pneumonia, colitis) due to DAC-related neutropenia while two patients showed progressive disease. Except of neutropenia no other DAC-induced toxicity was observed including no aggravation of GVHD. Discussion: A 10-day schedule of DAC with or without DLI is associated with a rapid anti-leukemic eff ect in a considerable subset of patients with early relapse after allogeneic HSCT including prior failure to azacytidine therapy. Infectious complications due to sustained neutropenia require however careful surveillance of these patients. Disclosure of Interest: None Declared. S118 at day 30 and 100 were 7.5% (n=3) and 10% (n=4), respectively. Late death (>100 days) was observed in 7 patients (17.5%) and 3 of them were related to allo-SCT. aGVHD was seen in 20 (50%) patients and chronic extensive GVHD in 8 patients (20%) . Of those 24 patients who underwent allo-SCT in CR1, 18 patients (75%) were still in remission without any sign of disease. Nine of 16 patients (56.3%) who underwent allo-SCT beyond CR1 lost due to relapsed leukemia. Survival analysis revealed superior OS between patients in CR1 (median 901 days) and beyond CR1 (median 202 days) (log-rank, OS P=0.003). For disease free survival we have seen the same profi le (P=0.007). In CR1 patients (n=24) univariate analysis revealed no impact of standard-risk cytogenetics, high HCT-CI (≥2), high WBC at dx (>30k), sex (F to M), ABO mismatch and double induction for CR achievement. The only adverse factor for less OS in CR1 was past history of invasive fungal infection pre allo-SCT (P=0.038, 901d vs 553d). Discussion: Allogeneic SCT in AML patients should be undertaken in CR1 with matched sibling donor. Pre-tx invasive fungal infection was the only worst predictive factor for OS in CR1 AML patients cohort. Disclosure of Interest: None Declared. Introduction: The development of treatment-related acute myeloid leukemia (t-AML) is the most serious secondary event that can occur after chemotherapy or radiation therapy. The prognosis in this group of patients is very poor. Materials (or patients) and methods: Here, we report the outcome of allogeneic hematopoietic stem cell transplantation (alloHSCT) in 13 patients (11 women, 2 men) with t-AML treated in our department between 2004 and 2013. This is 9% of all alloHSCT performed in patients with AML in our department in this period of time. Patients were treated by chemo-or radiotherapy due to the following previous malignancies: breast cancer (n=5), ovarian cancer (n=1), Hodgkin lymphoma (n=4) and non-Hodgkin lymphoma (n=3). Patients with treatment-related myelodysplastic syndrome and patients with AML secondary to myelodysplastic syndrome were not included to the analysis. Due to primary malignancies six patients were treated by chemotherapy only, 3 by radiotherapy and 4 by both methods. Three patients were also treated by high dose chemotherapy and autologous HSCT. The median interval between previous chemo-or radiation therapy and diagnosis of t-AML was 4 years (0,5-8 years). Median age of patients at diagnosis of t-AML was 51 years (37-63 years). None of them presented with extramedullary involvement at diagnosis, seven of them presented with chromosomal changes (trisomy 8 n=2, inversion 16 n=1, complex changes n=3). All patients were treated by induction and consolidation chemotherapy due to AML. Results: Disease status at the time of HSCT was: fi rst complete remission (CR) (n = 8); second CR (n = 2); and non CR (n = 3). Patients received myeloablative (n=2) or reduced-intensive (n=11) conditioning regimen and peripheral blood stem cells (n=12) or cord blood cells (n=1) from human leukocyte antigen-matched sibling donors (n=5), unrelated donors (n=7) or mismatch cord blood (n=1). Graft-versus-host disease (GvHD) prophylaxis was performed with a combination of cyclosporine and methotrexate or mycophenolate mofetil, additionally 10 patients received anti-T lymphocyte globulin. Neutrophil engraftment was achieved in all but one patient. Acute graft-versus-host disease (GVHD) grade II-IV occurred in one patient and chronic GVHD was observed in 2 patients. Relapse of disease occurred in three patients (23%), and all these patients died due to relapse. Median time between HSCT and relapse was only 2 months. Only one patient (8%) died due to transplant related reason, multiorgan failure, before neu-trophil recovery. One patient died due to brain tumour which was revealed 9 months after HSCT, and one due to reason not related with t-AML and transplantation. After the median follow-up time of 23 months (1-105 months) seven patients (54%) were alive in complete remission of the disease. The one-year overall survival was 58% (95% CI 31-81). Discussion: In summary, this retrospective analysis indicates that alloHSCT is a curative therapy for some patients with t-AML following primary malignancies, with acceptable toxicity and transplant related mortality. The major cause of death is relapse of the disease. Disclosure of Interest: None Declared.

Introduction: Survival outcomes from hematopoietic stem cell transplantation (HSCT) in severe aplastic anaemia (SAA) have improved steadily over the past decade but haploidental HSCT remain challenges mainly associated with graft failure and GVHD. So it has been considered in the fi nal choice of treatment status for SAA. Materials (or patients) and Methods: We retrospectively evaluated the outcome of 27 children with SAA who received unmanuplated haploidentical HSCT in China Pediatric Bone Marrow Transplant Group between July 2002 and Nov 2013. All 27 cases failed to respond to previous immunosuppressive therapy (average 23 months, range from 1 to 72 months) and were heavily transfused before transplantation. 3 patients were even in active infection at the time of haploidentical HSCT because they failed fi rst HSCT. Results: For the HLA 12/27, 8/27 and 7/27 had 1/6, 2/6 and 3/6 loci mismatch respectively. Among the 27 patients, 25 achieved neutrophil and platelet recovery at a median of 13±1.8 days (range from 10 to 21 days) and 19±9.1 days (range from 7 to 41 days) respectively. Two patients (7.4%) failed to achieve engraftment, both received second same donor HSCT and 1 achieved sustained engraftment. The cumulative incidence was 33.3% for grade II-IV acute GVHD and 27.3% for chronic GVHD. After median 24months (1-138months) follow-up only 3 patients died because of acute GVHD, infection and post transplantation lymphoproliferative disease respectively. The 3 year OS and EFS was 86.9 ± 12.4% and 85.2%± 12.4% for all surviving patients, especially 100% for our 12 patients with 1/6 mismatched HSCT. Discussion: These excellent outcomes suggest that unmanuplated haploidentical HSCT is a feasible way for children with SAA without HLA-identical sibling donor. It should be considered to be used earlier especially for the patient with only 1/6 mismatched family donor. Disclosure of Interest: None Declared. S120 increased risk of cancer. Hematopoietic stem cell transplantation (HSCT) is a modality that is potentially curative for the bone marrow failure that these patients have. Here we report our Centre's experience of HSCT in adolescent and young adult (AYA) patients with FA, a cohort of patients who are older and more likely to be heavily pre-treated than the majority of patients in reported series. Materials (or patients) and Methods: We carried out a retrospective analysis of patients' data from medical records and electronic systems from 1988-2012. We included patients with confi rmed FA based on positive chromosome breakage study who underwent stem cell transplant at our institution aged >14. Results: A total of 12 patients with FA underwent HSCT. The median age was 17.7 years (range 14-26y) with female predominance of 75%. Patients' baseline characteristics are shown in table 1. Two patients had dysplasia on the bone marrow with abnormal cytogenetic. Prior to HSCT all patients were transfusions dependent; median ferritin level of 10 patients was 3500 ng/l (range 134-6998). Five patients (41%) received prior steroids (n=3) and/or androgens (n=3). One patient received prior ATG and cyclosporine. Conditioning included TBI in 5 patients (41%). ATG was included in the conditioning regime in 10 patients (83%). Eleven patients (91%) received a graft from an HLA identical donor; stem cell source was bone marrow in 83.3% cases. The Overall 5 year survival was 67%. Four patients (33%) died, all before day +100 of HSCT without evidence of engraftment. The causes of mortality were transplant related being infection, Acute Respiratory distress syndrome (ARDS) or grade 4 Graft versus host disease (GVHD). At a median follow up of 47 months (range 5-245) for surviving patients (67%), all the latter patients engrafted and remained transfusion independent. Two surviving patients developed aGVHD ≥grade II. One patient developed chronic GVHD.

Discussion: Our fi ndings support the feasibility of allogeneic HSCT in older and more heavily pre-treated patients with FA. Patients who engrafted had excellent long-term outcomes. The prevalent risk of early TRM highlights the need for optimization of pre-and peri-transplant supportive care and measures to reduce the risk of graft failure. Further studies of HSCT in AYA and older patients are warranted. Disclosure of Interest: None Declared.

Introduction: Stem cell source is the main issue in allogeneic hematopoietic cell transplantation (alloHCT). In these days, there are more and more alternative donors (AD) other than matched sibling donor (MSD) in which bone marrow (BM) was preferred to mobilized peripheral blood (PB) as stem cell source in adult idiopathic aplastic anemia (AA). However PB has some advantages over BM: not requiring general anesthesia; easy accessibility; probability of higher stem cells dose. There needs to answer which stem cell source is preferable in specifi c donor setting and how much stem cell dose is adequate. Therefore we wanted to investigate the clinical impact of stem cell source on alloHCT in AA according to donor type. Materials (or patients) and methods: We retrospectively analyzed the eff ect of stem cells on alloHCT in AA. Mismatched donors (MMD) included haplo-identical family donor, mismatched unrelated donor or partially matched family donor. AD referred either matched unrelated donor (MUD) or MMD. Cord blood stem cells were not included in this analysis. Results: Total 267 patients were included in this analysis. Donors were MSD in 172, MUD in 38 and (MMD in 58 patients. BM and PB were used as stem cell sources in 194 and 74 patients, respectively. There was no cord blood. BM was associated with low incidence of acute graft versus host disease (GvHD; P<0.001) but not in G3/4 acute GvHD (P=0.427), and this tendency was more dominant in MSD (P=0.011) than in AD (P=0.057). BM, however, had no impact on other transplantation outcomes including chronic GvHD (P=0.673) and primary/secondary graft failure (P=0.774). Higher stem cell dose (CD34+ cells >3*10 6 /kg) was also associated with higher incidence of acute GvHD (P=0.039) but not in G3/4 acute GvHD (P=0.805). This tendency was only observed in AD (p=0.094) and not in MSD (P=0.334). Higher stem cell dose had no impact on other transplantation outcomes except for low incidence of extensive chronic GvHD in MSD (P=0.025). Multivariate analysis on overall survival in MSD revealed that only Age at alloHSCT<31 years old (HR 3.234; 95% CI, P=0.010) and prior platelet transfusion less than 86U (HR=2.618; 95% CI 1.017-6.738; P=0.046) were the favorable prognostic factors. On the other hand, multivariate analysis in AD revealed that higher stem cell dose (HR=2.596; 95% CI 1.020-6.609; P=0.045) was the only signifi cant favorable factors on overall. BM stem cell was not 

Cyclosporine A based 91.6% -· CSA alone -· CSA + Methotrexate -· Other 6 (50%) 3 (25%) 3 (25%) GvHD grade ≥II -· Acute -· Chronic 3(25%) 1(8.3%) Introduction: Alternative donor HSCT for patients with acquired SAA who fail to respond to immunosuppressive therapy and lack a matched sibling or unrelated (UD) donor, using a haploidentical family donor, is a potentially curative option. A haploidentical donor is available for most patients, the cost of the procedure is cheaper than cord HSCT, and time to procure the graft is short. But, published data are limited, mostly restricted to children and report poor outcomes due to poor engraftment and high risk of GVHD. We present a single centre study of patients receiving haploidentical HSCT for SAA. Materials (or patients) and Methods: Eight patients were transplanted, who either had refractory acquired SAA (n = 4) or who failed to engraft following UD (n = 3, 2 of whom failed UD HSCT x 2) or cord blood (n = 1) HSCT. Two patients had secondary severe marrow aplasia: one following chemotherapy for Hodgkin's disease who failed UD HSCT and one after 2 failed UD HSCTs for MDS. Median age was 32 yr (range 19-57) and median disease duration 46 months (6-91). Family donors were siblings (n=5), parents (n=2) and children (n=1). Conditioning regimen was fl udarabine 30 mg/ m 2 (days −6 to −2), cyclophosphamide (CY) 14.5 mg/kg (days −6 and −5) and TBI 2 Gy (day −1). Unmanipulated peripheral blood stem cells (PBSC) were infused to maximize cell dose (median CD34+ cell dose 6.2x10 6 /kg, range 1.8-8. 3 . GVHD prophylaxis was CY 50 mg/kg/d (day +3 and +4) to deplete donor allo-reactive T-cells, tacrolimus 1mg/day for 9 months and taper between 9-12 months, and MMF 15 mg/kg until day 35. Median follow of survivors was 14.8 months (7.2-44.4) . Median (and range) Karnofsky and HCT-CI scores were 80% (50-90) and 3 (0-5), respectively. Results: Six patients had sustained neutrophil engraftment; median time to neutrophil engraftment 18.5 days (range 16-23) and 5 sustained platelet engraftment 26 days (range 21-27). Full donor chimerism in unfractionated cells, CD3 and CD15 lineages was achieved and maintained at last follow up. Two patients who failed to engraft died on days +60 and +137 from sepsis. Both had multiple HLA antibodies directed against the donor, which persisted at high level (MFI) despite treatment with rituximab and plasma exchanges pre-transplant. Two patients developed CMV viraemia, and there was no case of EBV post-transplant lymphoproliferative disorder. One developed Guillan Barre syndrome. There was only one acute GVHD (Gd II skin) and no chronic GVHD. Median follow up was 12.2 months (3.2-40.4) . The conditioning regimen was well tolerated and with no haemorrhagic cystitis. One patient is currently 30 weeks pregnant at day+ 470 day posttransplant. Discussion: We show in a small cohort that non-myeloablative haploidentical HSCT with post-graft high dose CY and PBSCs is a feasible option. Engraftment occurred after failed prior MUD or cord HSCT. Our patients had a high HCT-CI and poor performance status, but had minimal toxicity from the transplant procedure. The lack of GVHD (only one case of acute Gd II skin GVHD) despite using PBSC is of interest and warrants further exploration in larger studies. High levels of donor specifi c anti-HLA allo-antibodies explain the non-engraftment in two patients and warrants careful screening of family donors, which if present, likely preclude them from the procedure. Disclosure of Interest: None Declared.

Introduction: Immunosuppression (IS) with ATG and cyclosporine is the treatment of choice for severe aplastic anemia (SAA) patients not eligible for hematopoietic stem cell transplantation. Severity degree criteria at diagnosis are well defi ned; in contrast the defi nition of remission after IS is not totally satisfactory. For the defi nition of complete remission (CR) there are several criteria worldwide, however with minor impact on the clinical outcome. Partial remission (PR) includes all patients after IS who do not meet criteria for SAA any longer and are independent of transfusions. We hypothesize that partial remission status is more heterogeneous and therefore a subclassifi cation would be more informative for patients' outcome. Materials (or patients) and Methods: To confi rm our hypothesis we used the database from the randomized EBMT-G-CSF study (Tichelli et al, Blood 2011) . For the analysis we included all patients in PR (according to NHI criteria). We defi ned good PR all patients fulfi lling simultaneously hemoglobin (Hb) ≥80 g/L and platelets (Plt) ≥ 30 G/L. Patients with at least one of these blood values below the cut-off criteria were considered as poor PR. Neutrophils were not included in this analysis since all patients in PR had values above 0.5 G/L. We compared event free survival (EFS: loss of response, secondary myelodysplastic syndrome, death) and overall survival (OS) at six years of both groups considering platelet counts, hemoglobin and the type of PR (good versus poor) at day 90 and 180. From 205 patients of G-CSF study, at day 90 and 180, 45 and 54 patients respectively were in PR and evaluable. Results: Patients with Plt ≥ 30 G/L at day 90 showed signifi cant better EFS at 6 years (62%) than those below (37%; P=0.045). There was no diff erence between patients with higher or lower Hb (P=0.906). When compared good versus poor PR, there was no diff erence for EFS (P=0.245) neither for OS (P=0.365). Patients with Plt ≥ 30 G/L at day 180 showed signifi cant better EFS at six years (64%) than those below (20%; P=0.008). Patients with Plt above 30 G/L showed also a better OS (91%) than those with lower values (53%; P=0.011). There was no diff erence for patients with higher or lower Hb (P=0.232). When considered good versus poor PR, there was a trend for better OS in favor of good PR (63% versus 44%; P=0.081) whereas EFS showed not diff erence (90% versus 67%; P=0.10). Discussion: Platelet values at day 90 and 180 in PR patients seem to be the main prognostic factor for EFS. In this analysis Plt values on day 180 was the only prognostic factor for OS. Hemoglobin does not seem to be relevant; however the low number of patients included in this analysis might limit the interpretation of these results. The subclassifi cation of patients in good and poor partial remission seems to be meaningful for the outcome. Platelets appear to be the most relevant prognostic factor. Disclosure of Interest: None Declared.

Introduction: There have been sporadic reports of the development of delayed disease recurrence following bone marrow transplantation (BMT) for severe aplastic anaemia (SAA) despite sustained majority or full donor chimerism. This is termed "donortype aplasia" (DTA). Unpublished Japanese registry data from the period 1980-2010 reported DTA in 5.7% of 660 patients aged <20 years following family/unrelated donor transplants but a surprisingly high 10-year overall survival at 87%. 1 The study identifi ed statistically signifi cant correlations with use of fl udarabine in conditioning therapy (P < 0.0001), low infused total nucleated cell (TNC) dose (≤3 x10 8 /kg, P=0.008) and use of immune suppressive therapy (IST, P=0.04) before transplant. Replacement of conventional cyclophosphamide (Cy) 200 mg/kg conditioning therapy with Cy 100 mg/kg plus fl udarabine resulted in a 14% incidence of DTA compared to 0% in the Cy 200 group (P=0.04). Materials (or patients) and Methods: We describe another 7 children from fi ve institutions from Europe, Israel and the USA who developed DTA in the period 2001-2010. They had presented with SAA/VSAA and proceeded to transplant aged 5.8-15.8 years. Only one received upfront IST; this patient went on to receive 10/10 high resolution matched unrelated donor (MUD) BMT 25 months after presentation. The remaining six patients all had matched siblings (MSD) and underwent transplantation within 1-5 months of presentation.

Conditioning for MSD comprised Cy 200mg/kg total dose (6 patients) plus ATG (various formulations and doses, 3pts) or Alemtuzumab (0.9-1 mg/kg total dose, 3pts) and for MUD BMT Cy 120mg/kg, fl udarabine 150mg/m 2 plus Alemtuzumab 0.9 mg/kg. GVHD prophylaxis comprised ciclosporin/methotrexate in four MSD BMT and ciclosporin alone in the remainder. Infused TNC doses averaged 4.48 x 10 8 /kg (range 2. 1-8.4) . Results: DTA developed at 12-44 months post-BMT. Most recent white blood cell chimerism prior to developing DTA was 61, 72, 80, 90, 91, 93 and 100%. Parvovirus was a likely aetiological factor in three patients: in two DTA developed soon after primary parvovirus infections and a third patient went from 40% marrow cellularity to severely hypocellularity soon after secondary reactivation. Two patients developed Aspergillus infections requiring partial lung and kidney resections respectively as a consequence of their recurrent pancytopenia. One patient received two donor leukocyte infusions to combat mixed chimerism and persisting transfusion dependence but without benefi t. Two patients received CD34-selected stem cell top-ups, again without sustained response. Two patients are too early following development of DTA to present outcome data. The remaining fi ve patients underwent a second BMT, all from their original donors. Three are completely well with good counts at 8-124 months post-second transplant. One patient developed further DTA and required 5/6 matched cord blood transplantation and another developed atypical haemolytic uraemic syndrome and persistent thrombocytopenia. Discussion: In contrast to Japanese experience, DTA was skewed to patients who had received sibling BMT (6/7 cases) and only one Introduction: Eculizumab, humanized monoclonal antibody directed against complement component C5, has changed the treatment paradigm of patients with haemolytic paroxysmal nocturnal haemoglobinuria (hPNH). HSCT should be the preferred option for PNH patients on eculizumab who develop overt bone marrow failure (BMF) during their clinical course or presenting concomitantly with SAA and PNH. The optimal use of eculizumab in the peri-transplant setting to prevent haemolytic/thrombotic complications and also the ideal conditioning regimen to eradicate the PNH clone is yet undetermined. Materials (or patients) and Methods: We analyzed 7 patients with PNH and SAA who underwent HSCT at our Centre for their BMF.

Of these 5 patients were on eculizumab for a median of 8 months (range [4] [5] [6] [7] [8] [9] [10] [11] [12] for PNH prior to worsening of their aplasia necessitating regular blood product (red cells and platelet) support. Two patients were initiated eculizumab during the pre-transplant period for their coexisting haemolytic PNH and SAA, whilst donor search was undertaken.The median age 32 years (range 18-56 yrs) and median disease duration was 48 months (range 4-252 months). All patients had evidence of haemolysis, elevated LDH and a large granulocyte clone (median 95%, range 57-100).The conditioning received was a fl udarabine/BU-based RIC HSCT, of these 4 received T-cell depletion using ATG (anti-thymocyte globulin) and 1 with alemtuzumab (100% CD3+ cells co-expressed CD52). Two patients received a haploidentical transplant with post-transplantation cyclophosphamide. The source of stem cells in all 7 was GCSF mobilized peripheral blood stem cells.Eculizumab was continued until the transplant and the last dose was given on the start of the conditioning protocol. Results: Pre-transplant 3 patients successfully underwent embryo (n=1) and ova (n=2) cryopreservation with ovarian hyper stimulation protocol (supraphysiological doses of estrogen) with eculizumab used prophylactically to cover thrombotic complication, as they were severely thrombocytopenic and were unable to receive low molecular weight heparin. Neutrophil and platelet engraftment occurred at 16 and 21 days respectively. No increased infectious complications was observed. The PNH clone was undetectable on D+14 and none was seen after engraftment with normalization of LDH. No clinical evidence of venoocclusive disease (VOD) was seen. Acute GVHD (grade 2) and Chronic GvHD (skin and gut) were observed in one patient each. Two patients failed to engraft, of which one was successfully rescued with a haploidentical HSCT although the PNH clone disappeared following the conditioning used in the fi rst transplant for both patients. Full donor chimerism in unfractionated, CD3 and CD15 lineages was achieved at D100 post HSCT. Discussion: This data demonstrated the feasibility of using eculizumab in patients with PNH and SAA, during the peri-transplant setting. The use of myeloablation in conditioning facilitates the eradication of the PNH clone. The successful use of eculizumab in preventing the thrombotic risks associated with ovarian hyper stimulation protocol is also evident. Larger studies and uniform strategies are needed to evaluate the optimal use of anti-complement therapy for SAA/PNH patients undergoing HSCT. Disclosure of Interest: None Declared.

Introduction: Treating patients with severe aplastic anemia (SAA) who fail to respond to immunosuppressive therapy (IST) and do not have an HLA-matched donor is challenging. Moreover, patients with SAA who develop infection because of neutropenia or graft failure after hematopoietic stem cell transplantation (HSCT) require urgent HSCT. HLA haploidentical family donor could be considered as one of alternative donor sources because it's 1) easy to fi nd the donor in relatives, 2) fast to engraft and 3) possible to use donor cell derived cell therapies such as virus specifi c cytotoxic T lymphocytes (CTLs) and mesenchymal stromal cells (MSCs). We analyzed 25 children with aplastic anemia (AA) who underwent HSCT from HLA mismatched unrelated donor (mismatched UR-HSCT) on schedule or HLA haploidentical donor (haplo-HSCT) for an urgent need of SCT. Materials (or patients) and Methods: Nineteen children with SAA conducted mismatched UR-HSCT and six children with SAA received urgent haplo-HSCT between 2000 and 2013 in Nagoya University hospital. The conditioning regimen in mismatched UR-HSCT consisted of cyclophosphamide (200 mg/kg), antithymocyte globulin (ATG, 10 mg/kg), and total body irradiation (TBI, 5 Gy). In haplo-HSCT, patients received a conditioning treatment consisting of fl udarabine 30 mg/m 2 i.v. x 4 days, melphalan 70 mg/m 2 i.v.

x 2 day and TBI 5Gy in two fractions. BMT was conducted on day0 and G-CSF mobilized peripheral blood stem cells (PBSC) were infused on day 6 to boost numbers of stem cells in haplo-HSCT while ATG were given at 10mg/kg prior to BMT and at 5mg/kg one day before PBSCT. All 25 patients received GVHD prophylaxis consisted of tacrolimus and short term MTX. Results: The reasons of haplo-HSCT were very severe aplastic anemia with a neutrophil count of zero cells/mm 3 with infections in two patients, graft failure after the fi rst SCT in three patients and no suitable unrelated donor in one. The median age was 9 (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) years old in mismatched UR-HSCT and 10 (5-15) years old in haplo-HSCT. The engraftment was achieved in 17 out of 19 (89.5%) patients of mismatched UR-HSCT and the two rejected patients were rescued by the second urgent haplo-HSCT. One of the patients received donor bone marrow derived MSCs infusion on day 0 in combination of haplo-HSCT to enhance the engraftment. All the patients of haplo-HSCT achieved engraftment. Acute GVHD grade II or more and chronic GVHD were seen in 3 and 4 out of 19 mismatched UR-HSCT respectively and 3 and 3 out of 6 haplo-HSCT respectively. CMV reactivations were observed in 15 UR-HSCT and 6 haplo-HSCT and one patient developed ganciclovir/foscavir resistant CMV infection and rescued by CMV specifi c CTLs therapy. EBV reactivations were observed in 4 UR-HSCT and in 3 haplo-HSCT and the patients were treated by rituximab. One patient developed CD20 negative EBV associated lymphoproliferative disease and rescued by EBV specifi c CTLs therapy. All 25 patients are alive with the median follow-up period of 7 years (from 6 month to 11 years and 8 months). Discussion: HLA mismatched UR-HSCT is a useful option for children with SAA and unmanipulated haploidentical SCT with the option of cell therapies such as virus specifi c CTLs and MSCs was a feasible salvage therapy for children with SAA who don't have HLA matched related donor and need urgent HSCT. Disclosure of Interest: None Declared. 1 1 Hematopoietic Stem Cell Transplantation, Federal Center For Pediatric Hematology, 2 Hematopoietic Stem Cell Transplantation, Federal Center For Pediatric Hematology, 3 Transfusiology, 4 Molecular Biology, 5 Hematology, Federal Center For Pediatric Hematology, Moscow, Russian Federation Introduction: Hematopoietic stem cell transplantation from unrelated donors remains the only curative option for severe aplastic anemia patients refractory to ATG/CsA immunosuppression (IST).

Although the results of MUD transplantation in SAA have improved signifi cantly with molecular typing, fl udarabine in conditioning and improved GVHD control, especially in campath-based protocols, GVHD and infections remain a serious problem, associated with signifi cant morbidity and mortality. We investigated the role of new method of graft processing -TCR alpha/beta depletion as a way to improve the results of MUD transplants in SAA. Materials (or patients) and Methods: Ten patients with SAA were treated since November 2012 till August 2013. Median age at HSCT was 14 (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) years, 6 male/4 female. 5/5 pts were refractory/relapsed after at least two courses of ATG/CsA, 2 pts had concurrent severe hemolytic PNH. Time interval from diagnosis to transplant was 4,3(1-10) years. Donors were unrelated volunteers, 10/10 matched. Six pairs were gender mismatched. CMV status was D+/R+ in 2, D-/R+ in 8. Preparative regimen included cyclophosphamide 100 mg/kg, fl udarabine 150mg/kg, ATG(horse, ATGAM) 100mg/kg and 6Gy thoraco-abdominal irradiation. Two patients recieved alemtuzumab instead of ATG because of anaphylaxis. Post-transplant GVHD prophylaxis included Tacro till day 60 and Mtx on days +1,+3,+6. PBSC grafts were depleted of TCRalpha/beta cells and CD19 cells with CliniMACS device as recommended by the manufacturer. Patients recieved a median of 8,6(6,8-13,5)x10 6 CD34 per kg, 3(1-30) x10 4 TCRalpha/beta per kg. Results: All patients engrafted with a median of 14 days for WBC and 13 days for platelets. In one patient secondary graft failure developed. He was salvaged later with a second unrelated graft. Acute grade 2 skin GVHD developed in 1 patient. No case of grade 3-4 aGVHD was observed. Five (50%) patiens had CMV reactivation requiring therapy, median time to CMV clearance was 2 (1) (2) (3) weeks. Prolonged mixed chimerism in T-cells did not compromise graft function. One patient developed moderate steroid-sensitive pneumopathy of uncertain etiology. With this exception no sign of chronic GvHD was noted. With a median follow up of 8 (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) months all patients are alive, 8 of them off any IST. Discussion: In our experience the described combination of conditioning regimen and graft manipulation method provide a safe way (at least in the short term) of unrelated hematopoietic stem cell transplantation in severe aplastic anemia. 3 Laboratoire d'Immunologie et Plateforme Nancytomique, 4 Unité de Transplantation Médullaire Allogénique, 5 Faculté de Pharmacie, Département de Microbiologie-Immunologie, Nancy, France

Introduction: Viral infections are the major cause of morbidity and mortality for the patients following haematopoietic stem cell transplant (HSCT). Absence of effi cacy of anti-viral drug is often correlated to absence of anti-virus specifi c immunity. Moreover, anti-viral drugs sometime cause important side eff ects. An alternative therapy is specifi c immunotherapy by virus specifi c T cell (CTL: Cytotoxic T Lymphocytes). It has proved eff ective in clinical studies to control virus infections and to contribute to virus-specifi c immunity reconstitution. However we aimed to identify the T sub-populations in CTLs after IFNg immunomagnetic selection. Thanks to the recent discovery of stem cell like memory T cell (T SCM ), which is a long life memory subpopulation, with a high proliferative potential and an enhanced capacity for self-renew. We also identifi ed this sub-population in CTLs. Materials (or patients) and Methods: Freshly selected virus-specifi c T cells (EBV-CTL n=1 ; ADV-CTL n=2 ; CMV-CTL n=2) were stained by the LIVE/DEAD AQUA fl uorescent-reactive, anti-CD4, anti-CD8, anti-CD3, anti-CD45RA, anti-CD27, anti-CD197 (CCR7), anti-CD95, anti-IFN-g. Expression of CD27, CCR7, CD45RA and CD95 was analysed on CD3 + CD4 + IFN-g + and CD3 + CD8 + IFN-g + gated T cells. Naïve T cells (T N ) were defi ned as CD45RA + CCR7 + CD95 -; Stem cell like memory T cells (T SCM ) as CD45RA + CCR7 + CD95 + ; Central memory T cells (T CM ) as CD45RA -CCR7 + ; Eff ector memory T cells (T EM ) as CD45RA -CCR7and Eff ector T cell (T EFF ) as CD45RA + CCR7 -. Flow cytometric acquisition was performed on 500 000 events per sample on a Navios cytometer (Beckman Coulter). Kaluza software (v1.2) was used for analysis (Beckman Coulter) . Results: Within all the virus specifi c CTLs (ADV-CTL, EBV-CTL and CMV-CTL), all the T cell sub-populations (T N , T SCM , T CM , T EM , T EFF ) were observed among CD3 + CD4 + IFN-g + and CD3 + CD8 + IFN-g + T cells. TEM was the most important in CD3 + CD4 + IFN-g + and CD3 + CD8 + IFN-g + T cells (>80%). However, stem cell like memory T cell compartment (T SCM ) was also identifi ed in CD3 + CD4 + IFN-g + (0.028-0.45%) and CD3 + CD8 + IFN-g + (0.0-1.08%) except in ADV-CTL. T N was present mainly in CD3 + CD4 + IFN-g + T cells for EBV-CTL, but mainly in CD3 + CD8 + IFN-g + for ADV-CTL and CMV-CTL. Discussion: In anti-virus CTLs selected by IFN-g based immunomagnetic technology, all the T cell sub-populations described by then are present. Although the number of isolated CTL available for infusion to patients is low, the presence of Tscm allows for a high in vivo expansion potential on antigen stimulation. Disclosure of Interest: None Declared. Introduction: The infusion of selected virus-specifi c T-cells represents an option for the treatment of drug resistant CMV, EBV or other viral infections post transplantation. The re-stimulation for IFNγ secretion and selection of these cells can be performed via the CliniMACS® Cytokine Capture System® (CCS, Miltenyi Biotec GmbH, Germany) . Recently a novel device of the same manufacturer, the CliniMACS Prodigy® (Prodigy) was made available, promising a fully automated, GMP-compliant and highly effi cient process. Within an application for a GMP-manufacturing licence the performance of this machine in comparison with the routinely used CliniMACS® Plus instrument (Plus) has been evaluated. Materials (or patients) and Methods: Lymphaphereses (≥3×10 9 WBC, >10% T-cells, viability >98%) were collected from 3 healthy CMV positive individuals after obtaining a written inform consent. Starting with 1×10 9 WBC (refer to table for exception) three selections of CMV-antigen specifi c T-cells were carried out with the Prodigy and Plus simultaneously. The MACS® GMP PepTivator® HCMVpp65 (Miltenyi Biotec GmbH, Germany) was applied for re-stimulation.

The starting and target cell fractions were analysed using up to 9 colour single platform fl ow cytometry for viable CD3+ IFNγ secreting cells, their subsets, and contaminating lymphocytes. Results: The concomitant selection results for the target fractions are shown in the The volume of the fi nal target fraction was 7-9.6 ml with the Prodigy and 40-43 ml with the Plus, respectively. The hands-on duration including quality control analyses was substantially shorter when using the fully automated system (<7 vs. >12 h with the Plus). No microbial contamination was found. Discussion: The immunoaffi nity selection of antigen specifi c cells cannot be assigned as a routine procedure. However the Clini-MACS® Plus instrument is used for several years in clinical scale preparation thus may be the standard when new approaches have to be evaluated. Both machines seem to provide cell fractions with similar characteristics. In each procedure the minimum (specifi ed in advance) of 1 ×10 4 target cells was isolated and the contaminating T-cells were suffi ciently low to avoid a high risk of GvHD. The proportions of dead cells were relatively high in all instances, which seem to be expected in such preparations with low cell numbers. Neglecting some initial technical diffi culties the CliniMACS Prodigy® will be an alternative especially taking into account the signifi cantly reduced hands-on time.

Disclosure of Interest: None Declared. S126 frequent complications after allogeneic stem cell transplantation. In addition relapse of the underlying disease is the other major cause of treatment failure. Screening for minimal residual disease (MRD) allows identifi cation of impending relapse and pre-emptive immunotherapy such as the application of donor lymphocyte infusions (DLI) or the transfer of cytokine induced killer (CIK) cells may in principle prevent relapse in a group of high risk pediatric patients. Materials (or patients) and Methods: CIK cells are generated from peripheral blood mononuclear cells (PBMC) of CMV-or AdV-seropositive healthy donors according to the standard protocol. Besides the standard stimulations, IL-15 is supplemented and viral CMVor AdV-antigen in the form of a peptide pool is presented in the medium to increase the frequency of existing virus-specifi c cytotoxic T-lymphocytes (CTL) identifi ed by fl ow cytometry stained with specifi c MHC-Multimers. The cell-mediated cyto toxicity is evaluated in vitro targeting leukemia cell line and primary viral infected cells. Results: Besides the main population, containing CD3 + CD56cells, T-NK cells (CD3 + CD56 + ) an essential subpopulation is expanding which consist mainly of a terminal diff erentiated activated CD8 + TEMRA (CD45RA + CD62L -) or TEM (CD45RO + CD62L -) phenotype, respectively (d0 32.2±14.03x10 6 -d15 1485.14±1153.71x10 6 , mean of total cell numbers, n=7) . The incubation with viral antigen leads to an up to 11.0 fold donor-dependent increase regarding CMV-specifi c CD8 + cells. The simultaneous peptide stimulation during the culture period has no negative infl uence on the anti-tumor eff ect directed to the M4 AML subtype cell line THP-1 in vitro (CIK antigen pos 48.25±25.17% AML lysis, CIK antigen neg 53.50±11.36% AML lysis; mean of n=4; E:T ratio 40:1. In subsequent in vitro experiments the enhanced cytotoxic capacity of antigen-stimulated CIK cells targeting viral antigen-loaded T2 cell line is shown (CIK antigen pos 28.2±23.40% T2 lysis, CIK antigen neg 5.0±3.67% T2 lysis; mean of n=5; E:T ratio 5:1). In follow-up in vitro experiments donor-mismatched fi broblasts are infected with CMV and fi rst results may indicate to an increased lysis of the infected fi broblasts by CIK antigen pos cells compared to CIK antigen neg cells. Uninfected fi broblasts representing healthy recipient tissue and therefore function as a control for Graft-versus-Host disease (GvHD) are not killed at all. Discussion: The generation of dual specifi c cytotoxic CIK cells may be an improved immunotherapy after stem cell transplantation inducing cytotoxicity against leukemia cells and might help to clear specifi cally virus reactivation. Beyond that the minimal alloreactive potential and therefore the low risk of inducing GvHD disease, of the CIK cells is already shown in murine and human in vivo settings. Disclosure of Interest: None Declared. Introduction: Pre-emptive immunotherapy based on minimal residual disease (MRD) status with donor lymphocyte infusions (DLI) using cytokine-induced killer (CIK) cells may be benefi cial to prevent relapse without causing graft-versus-host-disease (GvHD). While CIK cells have shown potent in vivo activity against various cancer types such as lymphomas or colorectal cancer, their cytotoxicity against B-ALL, characterized by the expression of CD19, has been limited. Hence, retargeting of CIK cells using chimeric antigen receptors (CARs) to facilitate selective target cell recognition and enhance specifi c cytotoxicity represents a promising approach. Materials (or patients) and Methods: CIK cells were generated following an optimized protocol by ex vivo expansion of donor-derived peripheral blood mononuclear cells (PBMC) through the addition of interferon (IFN)-γ, anti-CD3 antibody, IL-2 and IL-15. Cells were transduced using CAR encoding VSV-G pseudotyped lentiviral vectors. CARs comprise an extracellular scFv antibody fragment as an antigen-binding domain, linked via a fl exible hinge region and a transmembrane domain to an intracellular signaling moiety such as CD3 zeta chain (fi rst generation CAR), or zeta chain fused to a costimulatory protein domain such as CD28 (second generation CAR). Transduction effi ciency and cytotoxicity were determined using fl ow cytometric methods. Results: We established an optimized protocol for transduction of CIK cells with CD19-specifi c fi rst and second generation lentiviral CAR constructs, and characterized cells for expression of an EGFP marker gene and CAR surface expression. Eff ects of exposure to lentiviral vector particles on the development of CIK cell subpopulations and CAR expression were monitored over four weeks of continuous culture. Thereby transduction did neither aff ect the relative proportion of CD3 + CD56 -, CD3 + CD56 + and CD3 -CD56 + CIK cells, nor their activity against CD19-negative targets. To investigate functionality of CD19-specifi c CARs and the contribution of CAR-mediated recognition of CD19 to cytotoxicity, we generated a CD19-expressing variant of MDA-MB453 breast carcinoma cells by retroviral transduction. While retargeted CIK cells had no signifi cant eff ect on parental MDA-MB453 cells (specifi c lysis of 3.75% ± 4.05), they effi ciently lysed MDA-MB453/CD19 cells (specifi c lysis of 47.8% ± 0.4) in the same assay. In subsequent in vitro cytotoxicity assays we could further demonstrate potent and selective cytotoxicity of retargeted CIK cells towards established cancer cell lines endogenously expressing CD19 and primary pre-B-ALL blasts. Thereby coincubation of cells for 14 hours at diff erent E/T ratios signifi cantly increased specifi c lysis of target cells. Discussion: Our results demonstrate potent and specifi c cytotoxicity of CIK cells expressing CD19-specifi c lentiviral CAR constructs against otherwise CIK-resistant established cancer cells and primary pre-B-ALL blasts. Ongoing work now aims at characterization of retargeted CIK cells in suitable in vivo models in NOD/SCID common γ chain knockout (NSG) mice. Disclosure of Interest: None Declared. Introduction: Decidual stromal cells (DSCs) isolated from fetal membranes of term placentas, are easily expanded and highly immunosuppressive in vitro. DSCs express mesenchymal stromal cell (MSC) markers, and have a high expression of T-cell inhibitory markers and integrins. In the present study, we introduce DSCs as a cellular therapy for cGvHD. Materials (or patients) and Methods: Three patients (1 (ALL), 2 (AML), and 3 (KML)) with severe extensive cGvHD were treated with DSCs (1-2.8 x 10 6 cells/kg). Patient 1 and 2 received two infusions and patient 3 received one dose. One third of DSCs administered to patient 1 and 2 were labelled with 111 Indium and the in vivo-distribution was tracked for 48h. Blood samples were obtained before and up to 4-10 weeks after the fi rst infusion. Samples were analysed by fl ow cytometry and luminex. Results: All patients had cGVHD of skin, liver and obstructive bronchiolitis. Patient 2 had malabsorption. Following DSC-treatment, patient 1 showed no eff ect on liver or skin. Esophageal varices disappeared and banding has not been needed since infusion of DSCs. Patient 2 had liver enzymes normalized and albumin stabilized, and a transient improvement of skin could be seen. Spirometry showed improvement after bronchial dilatation. DSC-treatment showed no eff ects on patient 3. Patients 1 and 2 are regarded as partial responders (PR) and patient 3 as a nonresponder (NR). Patients receiving 111 In-DSCs showed the same distribution pattern of the isotope over time. The isotope was initially located in the lungs, followed by dissemination to liver and spleen. The fl ow cytometry and luminex data are presented as the median frequency of data from all time points for each patient. Patient 3 had high frequencies of HLA-DR + cells within the CD3 + CD4 + cell population (Th) (median 72.9%, range 72.7-73.3%). The corresponding proportions in patients 1 and 2 were 21. 5% (17.6-21.9) and 36. 5% (25.8-50-8) , respectively. Among CD3 + CD8 + cells (Tc), the frequency of HLA-DR-expression was 33.6% (30.9-37.5), 60.5% (56. 7-68.1) and 80.6% (70.8-83.8) for patient 1, 2, and 3, respectively.

The percentage of Th-cells with a naïve (CD45RA + CCR7 + ) phenotype was 4.8% (3.6-6.3) in patient 3, but 24.4% (4.3-24.4 ) and 25.1% (11. 2-26.3) in patient 1 and 2, respectively. The proportion of terminally diff erentiated (CD45RA -CCR7 + ) Th-cells was 2.3% (2.1-2.6), 7.4% (2.4-8.7) and 12.7% (10.9-23.2) in patients 1, 2, and 3, respectively. The frequency of Tregs (CD4 + CD25 high CD127 low/-) was 11.5% (8. 63-15.9) for patient 3, whereas they were 6.4% (4.8-6.5) and 3.3% (2.5-4.8) for patient 1 and 2, respectively. Patient 3 had the highest proportion Th-cells with a Th17 (CD45RA -CXCR3 -CCR4 + CCR6 + ), Th1/Th17 (CD45RA -CXCR3 + CCR4 -CCR6 + ) and Th2 phenotype (CD45RA -CCR4 + CXCR3 -CCR6 -). Patient 3 also had the highest median plasma concentrations of IL-17, IL-4 and IFN-γ. Discussion: DSCs induced PR in two out of three patients. The distribution of DSCs do not diff er from what has been seen with MSCs, despite that DSCs express higher levels of integrins known to be important for homing to damaged tissue. The data suggests that patient 3 (NR) have a more activated/exhausted immune system. If this aff ects the DSC-treatment is diffi cult to determine in such a small study. However, luminex and fl ow cytometry data correlate regarding plasma concentrations of cytokines and presence of specifi c T cell subsets in peripheral blood. Disclosure of Interest: None Declared. Introduction: αOur previous studies have shown that TCR-γδ T cells homeostatically reconstitute in large numbers in approximately 25% of patients receiving haploidentical marrow grafts depleted of αβ T cells (αβ TCD). These γδ T cells exhibit potent anti-leukemia activity, and patients who recover from BMT with increased γδ T cells show 9-year leukemia-free survival (71% vs. 20% with normal γδ T cell recovery; P <0.0001) and without increased incidence or severity of GvHD (P = 0.96). Until recently, the lack of suitable reagents for αβ TCD or γδ T cell manufacturing and high risk of haploidentical transplantation slowed eff orts to build upon these initial fi ndings. In this study, we present preliminary fi ndings from our laboratory of two cGMP-compliant graft-engineering strategies for human clinical-scale γδ T cell-enriched donor innate lymphocyte immunotherapy (DILI).

Materials (or patients) and Methods: A total of three products were examined. All products were cultured in our GMP facility with ISO-7 classifi ed air fi ltration and exchange. Leukapheresis products were obtained from three volunteer donors, purifi ed by densitygradient centrifugation, divided, and cultured in 50%RPMI/50% Clicks supplemented with 10% HS, 2mM l-glutamine, 2mM Zoledronic Acid, and 100μ/mL IL2 either in standardized T150 cell culture fl asks or in a polycarbonate bioreactor chamber (Centri-Cult®; Miltenyi Biotec, Bergisch Gladbach, DE) . Cells were kept at a density of 1 x 10 6 /ml and the culture supplemented with IL-2 50u/ mL every other day until cell harvest at day +14 at which time the product was depleted of αβ T cells. Cells from both methods were evaluated for lymphocyte subsets, sterility and potency against standardized human leukemia cell lines. Results: Results from DILI manufacturing are detailed in the table below. All products produced γδ T cell expansions ranging from 74 to 200 fold. The Prodigy® CentriCult® bioreactor chamber showed no signifi cant impediments for use for this manufacturing protocol. Post-manufacturing release testing showed minimal residual αβ T cells with predominant composition of γδ T cells and a smaller NK population. Expanded γδ T cells expressed activation-associated T cell antigens and were cytotoxic against AML lines K562, KG1a, HL-60, AML193; the myeloma cell line U266/TIB-196; and glioblastoma lines U251MG and U87MG. Culture products tested negative for gram stain, 14-day anaerobic and aerobic bacterial cultures, endotoxin, and mycoplasma with the exception of product 1 manufactured in fl asks that grew out S. aureus during culture, likely due to manipulation required for multiple fl asks. Discussion: These results affi rm that clinical-scale manufacturing of cytotoxic ex vivo expanded/activated γδ T cells in a Donor Innate Lymphocyte Infusion product can be accomplished in a cGMP compliant manner. Manufacturing with the Miltenyi Prod-igy® is feasible, potentially allowing a simplifi ed protocol with less potential for microbiologic contamination, thereby expanding the use of DILI therapy for prophylaxis or treatment of resistant disease. Disclosure of Interest: None Declared. Introduction: Haplo-identical HSCT may resolve the lack of sufficient suitable completely HLA-matched donors for treatment of end-stage blood-cancer patients with a HSCT. Yet, current techniques do not allow for such procedure as presence of T-cells will cause severe GvHD and absence of T-cells (i.e. naked HSCT) will lead to occurrence of opportunistic infections. Materials (or patients) and Methods: Kiadis Pharma is developing ATIR, a T-cell immunotherapy based on a donor lymphocyte preparation selectively depleted of host alloreactive T-cells, through use of photo-dynamic therapy. In a dose-fi nding phase I/II clinical study (CR-GVH-001), the product was shown to enable safe and effi cacious haplo-identical HSCT. Currently, a subsequent phase II clinical study (CR-AIR-007) is ongoing. Results: Using recently developed analytical methods, the ATIR product from both studies was/is extensively characterized 

showing that only recipient-alloreactive T-cells were selectively depleted from the product while retaining reactivity against unrelated 3 rd party antigens and general potent T-cells. Additionally, ATIR was shown to have retained viral-specifi c T-cells, preserved the presence of memory and naïve T-cells and showed responsiveness to pathogens. Thereby, ATIR will provide mature immune cells that off er immune protection without eliciting severe GvHD. Discussion: The in vitro characterization data are supported by clinical data showing absence of TRM during 5-year follow-up (CR-GVH-001) over a broad dose range and no occurrence of severe GvHD/infections in the ongoing clinical study (CR-AIR-007). Together, these data show that using ATIR as an adjunctive medication, haplo-identical HSCT can be safe and effi cacious. Disclosure of Interest: None Declared. 3 Istituto Clinico Humanitas, Rozzano (MI), 4 

Introduction: Over the last decade the development of new treatments for advanced disease have improved survival rates for metastatic colorectal (mCRC) patients, but therapies are still associated with a poor prognosis and signifi cant morbidity. The functional behavior of natural killer (NK) cells makes them appealing potential eff ectors for cancer immunotherapy, however in mCRC patients only anecdotal cases have been reported so far both in the autologous and allogeneic setting. In this study we evaluated the capacity of autologous NK cells, freshly or after cytokine activation, to lyse mCRC cells and the expression of ligands for adhesion and triggering NK receptors involved in CRC recognition and killing to clarify the mechanisms involved in tumor susceptibility or resistance to NK cells. Materials (or patients) and Methods: After obtaining informed signed consent, 25 mCRC patients who underwent surgery to remove tumor or metastases have been enrolled to date. mCRC cells were isolated, in vitro expanded (after pathologic confi rmation of neoplastic origin) and analyzed for the expression of HLA class I and ligands for NK triggering receptors. NK cells were purifi ed and analyzed for the presence of receptors involved in NKmediated cytotoxicity. The expression of NK receptors and levels of cytotoxic activity against patients mCRC cells were also evaluated after overnight (ON) activation of NK cells with IL-2 and/or IL-15. Results: Tumor cells were successfully expanded from 21 of 25 samples. Results of experiments performed in 10 patients documented a substantial inability of patients freshly NK cells to lyse mCRC cells (<8% lysis at E:T ratio of 20:1). ON cytokine activation resulted in greatly increased NK cytotoxic potential in all patients evaluated (IL-2: mean 28%; range:10-71; and IL-15: mean 40%; range 16-76 at E:T ratio of 20:1). NKp30 and NKp46 were variably expressed by patient NK cells and could be up-regulated after ON activation, while mCRC cells expressed diff erent levels of most NK ligands. Additional days of NK cell activation was able to further enhance their lytic capability. Preliminary experiments, suggest that mCRC are susceptible to anti-EGFR-induced ADCC mediated by NK cells regardless KRAS mutation status. Discussion: Our data support the ongoing design of an immunotherapy approach with autologous NK cells for poor prognosis mCRC patients. Disclosure of Interest: None Declared. 1 1 Laboratory Medicine, Therapeutic Immunology, 2 Department. of Medicine, KAROLINSKA INSTITUTET, Stockholm, Sweden Introduction: We introduced mesenchymal stem cells (MSCs) for treatment of acute graft-versus-host disease (GVHD). The placenta and fetal membranes protects the fetus from the mother's immune system. Decidual stromal cells (DSCs) are easy to expand and give better immunosuppression in mixed lymphocyte culture (MLC) than other stromal cells. The DSCs were positive for CD29, CD44, CD73, CD90, CD105 and CD49D, but were negative for hematopoietic markers. DSCs are of maternal origin. Indoleamine2,3-deoxygenase, prostaglandin E2, PD-L1 and interferon-g participate in the immunosuppression by DSCs, as shown in blocking experiments in MLC. DSCs promote CD4+, CD25+, FoxP3+ regulatory T-cell expansion in a contact-dependent manner. We report our experience. Materials (or patients) and Methods: In vitro, in MLC and in vivo to study graft-versus-host disease (GVHD), we used a major histocompatibility mismatched mouse model, Balb/C mice as recipients and C57BL/6 as donors. Conditioning was busulfan and cyclophosphamide or 8 Gy of total body irradiation. Escalating doses of human DCSs, 1 x 10 5 -10 6 , were infused to recipient mice at day +1, +3, +5, +7 and +21 after hematopoietic stem cell transplantation (HSCT). Weight, mortality and histopathology were determined. DSCs were transduced with immunosuppressive genes, IL-10, prostaglandin E2 receptor (EP2), IDO and PDL-1 using a lenti virus vector. Eight patients with steroid-refractory grade III-IV acute GVHD were given 0.9-2.8 x 10 6 DSCs/kg at 15 infusions. Median age was 57 years. Results: In mice using Balb/c as responder and C57BL/6 as stimulatory cells in MLC, human DSCs induces a dose-dependent immunosuppression. In a xeno MLC setting, the proliferation of mouse (Balb/c) splenocyte, stimulated with human lymphocytes, were inhibited by human DSCs. In Balb/c mice conditioned and transplanted with BM and splenocytes from C57BL/6 donor (Mismatched GVHD model), human DSCs could attenuate the severity of GVHD and treatment on day +3 seemed the optimal infusion time. Human DSCs were transduced with lentivirus + GFP vector bearing IDO, prostaglandin E2 receptor (EP2), IL-10, IFNg and PDL-1. Transduced cells also inhibited mice MLC. The best eff ect was seen with PDL-1 transduced DSCs. In the patients there was no toxicity from infusion of DSCs. Two patients had complete response and four had a partial response. Melena stopped in three patients. Three patients survive from 1½ years to 3 years. Discussion: MSC therapy has been widely used. Some patients respond while others don't. A randomized trial in the US using industrial MSCs, showed no diff erence in overall response versus placebo. Experiments in mice models for GVHD suggest that MSCs need be licensed with interferon-g, nitric oxide or transduced with IL-10 to show an eff ect. DSCs may be a valid alternative to bone marrow-derived MSCs, because of better expansion potential and maybe better immunosuppressive eff ects. We have shown that human DSCs in a xenomodel prevent development of GVHD in mice, but survival was not signifi cantly improved. PDL-1 transduced DSCs seem to give the best eff ect. Clinically, DSCs were found to be eff ective with three out of eight patients being long-term survivors, as opposed to around 10% as expected in patients with steroid-refractory acute GVHD. In conclusion, DSCs may be successfully used for immune modulation and GVHD. Disclosure of Interest: None Declared.

Introduction: Colorectal cancer (CRC) is one of the most commonly diagnosed cancers worldwide. More than 1 million people are diagnosed with CRC each year, and a staggering 0.5 million died of the disease in the same time period.

Adoptive T-cell transfer (ACT) refers to an immunotherapeutic approach in which anti-tumor T lymphocytes, usually the tumor infi ltrating lymphocytes (TIL), are identifi ed, grown ex vivo and then re-infused into the cancer patients. This strategy has emerged as an eff ective treatment for patients with metastatic melanoma and other solid tumors, while at present, both active and adoptive immunotherapy do not play an important role in the treatment of advanced CRC. Aim: In order to develop an ACT protocol for CRC treatment, an experimental approach that does not require neither the defi nition of molecular defi ned tumor antigens, nor the availability of TIL was designed. It was based on the in vitro stimulation of patient's CD8+-enriched T-cells from peripheral blood mononuclear cells (PBMCs) with dendritic cells (DCs), pulsed with apoptotic tumor cells as a source of tumor antigens, to generate autologous CTLs with strong anti-tumor activity. Materials (or patients) and Methods: 78 CRC patients were enrolled. Tumor biopsies were obtained at surgery, together with 100 ml of heparinized peripheral blood (PB). Tumors were mechanically dissociated to a single-cell suspension and cultured to obtain a tumor cell line from each patient. DCs were generated from previously separated PBMCs, using a magnetic positive selection of CD14+ monocytes, cultured in presence of recombinant human and recombinant human Granulocyte-Macrophage Colony-Stimulating Factor (rh GM-CSF) for 6-7 days. Anti-tumor CTLs were elicited in co/micro-cultures using DCs as antigen-presenting cells, autologous apoptotic tumor cells as source of antigens and T CD8+ lymphocytes enriched eff ectors, in presence of weakly irradiated T CD4+ lymphocytes and PBMCs, and , with weekly stimulation. The immune monitoring of the diff erent cell populations was performed by fl ow cytometry analysis. CTLs Interferon-γ (IFN-γ) secretion was assessed by ELISpot assay, to evaluate their activation in response to autologous tumor. Results: Primary tumor cell lines were obtained from 20 out of 78 patients (25,6%). DCs were generated from 26 patients, and among them, 6 had the corresponding tumor cell line. This was the reason why co/micro-cultures were set up for 6 patients. ELISpot results showed that strong and signifi cant IFN-γ secretion was detected at the third, fourth and fi fth stimulations for one patient and at the second for another patient, whereas for three patients a weak secretion was detected during the second and third stimulations. T-cells from one patient did not react to the stimulations. Discussion: Despite the gut intestinal fl ora had adversely aff ected the establishment of primary tumor cell lines, an important success rate of 25,6% was obtained. In addition, although our immunological study must be performed on an increased number of CRC patients, our results suggested that the generation of tumorspecifi c CTLs could be useful for supporting an ACT approach in CRC. Disclosure of Interest: None Declared.

K. Miao 1,* , R. Lu 2 , J. Li 1 , S. Qian 1 1 First Affi liated Hospital of Nanjing Medical University, Nanjing, China, 2 Hematology, First Affi liated Hospital of Nanjing Medical University, Nanjing, China Introduction: Because many Chinese patients with hematologic diseases who need allogeneic hematopoietic stem cell transplantation (allo-HSCT) lack a human leukocyte antigen (HLA)-matched donor, haploidentical HSCT has provided an alternative option. Materials (or patients) and Methods: Here, we report the results of Haplo-HSCT with granulocyte colony stimulating factor (G-CSF) mobilized peripheral blood stem cells as the grafts without T-cell depletion in thirty-eight patients. Results: Thirty-eight patients with the median age of 30.5 years (ranging from 13 to 53 years) were enrolled in this study, and 29 cases were in high risk status. The patients received myeloablative preconditioning with or without total body irradiation and acute graft-versus-host disease (GVHD) prophylaxis consisting of basiliximab, cyclosporine A, methotrexate, mycophenolate mofetil and a rabbit anti-thymocyte globulin. Thirty fi ve patients attained successful neutrophil and platelet recovery. The median time for neutrophil recovery was 16 days (ranging from 10 to 23 days) and the median time for platelet recovery was 19 days (ranging from 10 to 66 days). During the follow-up at a median time of 33.1 weeks (ranging from 1.1 to 412.6 weeks), 11 (28.9%) patients developed aGVHD grade I -II, 7 (18.4%) patients developed aGVHD grade III -IV. The incidence of cGVHD was 27.6%. Nine (23.7%) patients died within the fi rst 100 days after transplantation. The cumulative survival proportions at one year and two years were 52.51±8.57% and 43.76 ± 9.11%. The disease-free survival rate longer than 3 years was 40.34±8.71%. Discussion: The results suggest that G-CSF-primed merely peripheral blood stem cell grafts without in vitro T cell depletion are an appropriate stem cell source for Haplo-HSCT, which may be an important HSCT strategy for patients without HLA full-matched donors. Disclosure of Interest: None Declared.

M. Bernardi 1,2,* , E. Albiero 1,2 , A. Alghisi 3 , K. Chieregato 1,2 , C. Lievore 3 , D. Madeo 2 , F. Rodeghiero 1,2 , G. Astori 1 1 Advanced Cellular Therapy Laboratory, Department of cellular therapy and hematology, San Bortolo Hospital, 2 Research Laboratories, Hematology Project Foundation, 3 

Introduction: in addition to their use in regenerative medicine, mesenchymal stromal cells (MSC) are used in the prophylaxis and treatment of Graft Versus Host Disease due to their strong immune-modulatory properties. Moreover, these cells have been shown to prevent graft failure and promote engraftment in hematopoietic stem cell transplantation (HSCT). MSC can be isolated from bone marrow, adipose tissue, cord blood or umbilical cord blood but for obtaining a suffi cient number of cells for clinical application, MSCs need to be ex-vivo expanded. Expansion is possible when fetal bovine serum (FBS), which promotes cell growth and proliferation is added to basal media but its use is discouraged by regulatory authorities, due the risk of zoonoses and immune reactions. Human platelet lysate (Hu-PL), containing several bioactive molecules and growth factors, has been proposed as a valid substitute of FBS. Growth factors release is commonly achieved after one or more overnight freezing/thawing cycles of platelet rich plasma; however the process is time consuming and diffi cult to standardize. We have developed a fast method to obtain PL from PRP by using ultrasound. Effi ciency was tested by expanding bone marrow (BM)-MSC in the obtained Hu-PL. Materials (or patients) and methods: sonication was performed by treating PRP bags in an ultrasonic bath at a frequency of 20 kHz for 30 minutes at room temperature. To evaluate platelet lysis we measured PDGF-AB release by ELISA. Effi ciency was tested by measuring cumulative population doubling time (cPD), diff erentiation capacity and immunogenic properties of BM-MSC expanded in D-MEM+10%Hu-PL and in D-MEM+10%FBS. Results: 74% of PDGF-AB was released after treatment. Hu-PL signifi cantly enhanced BM-MSC proliferation rate compared to FBS. The cPD of cells growth in Hu-PL at 10%, 7.5% and 5% was significantly better when compared to cells expanded in 10% FBS. BM-MSC expressed MSC markers and were able to diff erentiate into adipogenic, osteogenic and chondrogenic lineages. Immunosuppressive activity of BM-MSC expanded in Hu-PL was maintained when co-cultured with T-cells. Discussion: we conclude that Hu-PL can be quickly produced by sonication; moreover, Hu-PL reduce the cell PD time compared to FBS maintaining both cell diff erentiation potential and immunomodulatory capacities. Disclosure of Interest: None Declared.

M. Bernardi 1,2,* , V. Adami 3 , E. Albiero 1,2 , D. Madeo 1 , G. Astori 2 , F. Rodeghiero 1,2 1 Research Laboratories, Hematology Project Foundation, 2 Advanced Cellular Therapy Laboratory, Department of cellular therapy and hematology, San Bortolo Hospital, Vicenza, 3 High Throughput Screening core facility,CIBIO, University of Trento, Trento, Italy Introduction: Human platelet lysate (Hu-PL) represents an eff ective substitute of fetal bovine serum (FBS) for mesenchymal stromal cells (MSC) expansion. Compared to FBS, Hu-PL favors MSC proliferation signifi cantly shortening the population doubling time and avoiding the risks related to the use of animal derivatives. Growth factors contained in the platelet granules are commonly released upon disruption following one or more freezing/thawing (F/T) cycles. As an alternative to the F/T method, we have recently described the use of ultrasounds for the production of Hu-PL starting from platelet rich plasma (PRP). Since Hu-PL signifi cantly increases the growth rate of MSC compared to FBS and given the possibility of free radicals formation during the process of PRP sonication, we have investigated both the MSC chromosomal stability during expansion by karyotype analysis and Hu-PL safety applying the cytokinesis-block micronucleus assay (CBMA). Materials (or patients) and Methods: Hu-PL was produced by sonication of the PRP bags for 30 minutes at 20 kHz. MSC have been isolated from bone marrow (BM) samples and expanded in D-MEM+10% Hu-PL. Karyotyping by G-banding was carried out at the end of passage 6. For CBMA, the Chinese hamster ovary cell line (CHO-k1) lacking DNA repair mechanism was exposed to increasing concentrations of Hu-PL from 0.1% to 30% in three independent experiments and using diff erent Hu-PL batches. For the positive control, micronucleation was induced by exposing the CHO-k1 to Mitomycin-C. After 24h, cytokinesis was blocked by Cytochalasin B. Cells were fl uorescent stained with Hoechst and micronuclei were automatically detected using an high content imaging system (Operetta) analyzing at least 2000 binuclear cells/well. Results: In our experiments growth proliferation induced by the use of Hu-PL did not lead to micronuclei formation on CHO-K1 cells compared to negative control (P<0.01). Moreover, karyotype did not reveal genomic alterations on BM-MSC expanded in 10% Hu-PL. Discussion: micronuclei formation is considered a biomarker of chromosomal damage, genome instability and cancer risk. Our results suggest that MSC can be safely expanded in Hu-PL produced by sonication. Disclosure of Interest: None Declared.

B. Calmels 1,* , A. Drezet 2 , C. Huynh 1 , A. Autret 3 , R. Bouabdallah 4 , D. Coso 4 , C. Malenfant 1 , C. Lemarié 1 , C. Chabannon 1 1 centre De Therapie Cellulaire, Institut Paoli-calmettes, 2 UMR1068, INSERM, 3 Unité Biostatistiques -Département de la Recherche Clinique et de l'innovation, Marseille, France Introduction: Autologous hematopoietic stem cell transplantation (AHSCT) is still widely used for treatment of patients with hematologic diseases. Cryopreservation and storage are mandatory, allowing administration of high dose chemotherapy. Reinfusion of grafts thawed at the bedside might be associated with side eff ects attributed to cryoprotectant or cell lysis products. In order to remove such byproducts, biomedical devices like Cytomate or Sepax have been developed for washing out thawed grafts. However, washing of thawed grafts remains controversial since CD34+ progenitor cell loss is unavoidable. This is opposed to bedside thawing, where cell loss is supposed to be lower, while unknown owing to the absence of quality control. There are, to our knowledge, no reports comparing these 2 procedures for their relevant clinical endpoints, i.e. hematopoietic reconstitution in the autologous setting. In order to determine whether cell loss associated with thawing and washing has a detrimental eff ect on engraftment, we retrospectively compared two cohorts of matched patients, receiving either washed or unwashed autologous grafts at a single institution. Materials (or patients) and Methods: Between Jan. 2000 and Dec. 2004, 1,360 AHSCT were performed: we selected 364 of them for which a) grafts were thawed at the bedside and b) hematopoietic reconstitution raw data were available in the electronic data management system. We selected 218 AHSCT that fall within 4 diagnosis groups for which intensive chemotherapy regimens were homogeneous (Table 1) ; among this group, we retained the 65 AHSCT for whom cryopreserved CD34+ cell doses were within the therapeutic range of 2.0 to 5.0x10 6 /kg. We then sought to compare this group to a matched cohort of AHSCT washed before infusion: among the 1,570 AHSCT performed between Jan. 2005 and Dec. 2010 we selected 580 AHSCT for which: a) hematopoietic reconstitution raw data were available in the electronic data management system, b) patients received the same 4 intensive regimen than the unwashed group and c) grafts were homogeneously processed (dry thawing followed by washing on CytoMate); among this cohort, the 260 AHSCT with cryopreserved CD34+ cell doses between 2.0 and 5.0x10 6 CD34 + /kg were defi ned as the washed group. Results: In order to set up two comparable groups of AHSCT, we used a case-match design with one-to-two matching. Each of the 65 unwashed AHSCT was matched with 2 washed AHSCT issued from the 260 potential matches, using 4 baseline variables: sex, age at treatment, diagnosis and total cryopreserved CD34+ cell dose. The case-match analysis exhibits, as expected, no significant diff erence in terms of age, sex ratio, diagnosis and CD34+ cryopreserved cell dose (Table 1) . Median time to reach 0.5x10 9 /L circulating neutrophils is comparable between the two groups, with 12.4 and 12.5 days in the unwashed versus washed groups, respectively. S131 Discussion: Our approach allowed us to compare 2 matched cohorts issued from 2,930 AHSCT performed over 10 years at a single institution. This accurate matching leads us to conclude that washed autologous grafts do not compromise hematopoietic reconstitution as compared to bedside thawing. Moreover, using automated devices such as CytoMate or Sepax allow for better standardization, improved stability of thawed cell products as well as precise determination of infused CD34 + cells, and might thus be preferred over bedside thawing. Disclosure of Interest: None Declared. Introduction: Up to now immune responses following therapy of malignant melanoma with dendritic cells (DCs) appear to be only transient. Therefore the combination of DCs with antibodies, T cells and/or low-dose chemotherapy as possibilities to prolong the immunological eff ect and improve the clinical outcome is now under investigation. We conducted a clinical phase I/II trial to investigate the combination of autologous, tumor antigen specifi c mRNA loaded DCs with expanded autologous T cells in melanoma stage IV patients. Materials (or patients) and Methods: DCs were generated by enriching monocytes using the Elutra cell separator and diff erentiating them to immature DCs (iDCs) with GM-CSF and IL-4. iDCs were harvested on day 5 electroporated with hTERT and Survivin mRNA and matured by adding maturation cocktail as described by Jonuleit et al. After 24 hours the mature DCs were frozen in aliquots. Blood samples were taken before start of DC vaccination and at defi ned time points during vaccination. Patients who showed immune responses against hTERT and/or Survivin after initial treatment with DCs were off ered additional treatment with ex-vivo expanded T cells. A second leukapheresis was performed and T cells were enriched with the Elutra cell separator. After depletion of Tregs using CD25 Dynabeads, T cells were expanded with CD3/CD28 Dynabeads for 10 days in a WAVE bioreactor. Patients were non-myeloablative conditioned with Fludarabine and Cyclophosphamide and after removing of CD3/CD28 beads 3x10 10 T cells were transfused fresh. DC vaccination was continued the day after T cell transfer. Results: Three patients have been treated with this DC/T cell combination. All of them showed immune responses against hTERT peptides either already before start of DC vaccination or 3-5 months after the fi rst vaccination. Only patient three showed an immune response against Survivin peptides before DC-vaccination, which disappeared with start of vaccination. Patients one and two had a sustained response against hTERT peptides after treatment with T cells. In patient one this response dropped 20 months after beginning of DC-vaccination correlating with progression of the disease. In patient two the response dropped 29 months after start of DC-vaccination and was combined with an up-coming immune response against Survivin and a progression of disease from week 31 on. In patient three the response against hTERT peptides vanished at the time of T cell transfer and was combined with a strong response against Survivin peptides. The disease progressed after 11 months. Patients included in this study who did not receive additional T cells had a progression free survival (PFS) of 3 to 13 months with a median of 7 months. Patients who did not mount any immune response had a PFS of 3 to 10 months while patients who showed an immune response before start of vaccination or developed an immune response at a later time point against hTERT alone had a PFS of 8 to 13 months. Discussion: These data indicate that immune responses can be prolonged by additional T cell transfer. Prolonged responses against hTERT seem to result in longer PFS. The role of responses against hTERT and Survivin and their impact on clinical outcome in melanoma patients has to be further investigated. Disclosure of Interest: None Declared. Introduction: Junctional epidermolysis bullosa Herlitz (JEB-H) is caused by a congenital defi ciency in laminin-5, leading to lifethreatening skin and mucosal fragility. There is currently no established cure. Fetal membranes and cells from term placentas have wound healing capacities, and we have previously used decidual stromal cells (DSCs) from fetal membranes to treat acute GVHD. The immunogenicity of allogeneic stromal cells in vivo is poorly characterized. Materials (or patients) and Methods: Amniotic membranes (AM) from a term placenta were applied to severe wounds of an 11 months old JEB-H patient. The patient was treated with fi ve infusions of allogeneic DSCs (2 x 10 6 /kg) within a three-month period. The patient had received two units of blood, three and one month prior to the fi rst treatment with AM. Serum samples from the JEB-H patient and from SCT patients treated with DSCs for acute GVHD (n=8), chronic GVHD (n=3) and haemorrhagic cystitis (n=5) were analyzed by fl ow cytometric crossmatching (FCXM). The number of DSC doses ranged from 1-5 infusions, derived from 1-3 placental donors, and the patients were followed for at least four weeks.

[PH-P064] S132 Samples that were positive in FCXM were analyzed by Luminex to determine specifi city of anti-HLA antibodies. Peripheral blood mononuclear cells (PBMCs) from the JEB-H patient were stimulated with DSCs and the proliferative response was measured by 3 H-thymidine incorporation. The expression of laminin-5 on DSCs was confi rmed by fl uorescence microscopy. Results: One week after the AMs were applied improvement in healing was observed over the elbows. The patient was given DSCs two weeks after the application of AMs. Already after three days, lesions over the groins were improved. Subsequently, there were improvements in the face and the fi ngertips. Healing processes were seen in the middle of the blisters, as opposed to at the margin which may occur spontaneously in JEB-H. After three weeks, the wound over the left ear was completely healed and the face had further improved. A second dose of DSCs was given and the elbows were improved. After three more infusions of DSCs the improvements were transient. The patient survived 23.5 months, as compared to average fi ve months. After four infusions of DSCs from three diff erent donors, the JEB-H patient had developed anti-HLA antibodies specifi c to 38 HLA class I antigens. PBMCs from the patient had a higher proliferative response to DSCs than to third-party PBMCs (median stimulation index (SI) 6.3 (range 4.3-8.2) and 1.9 (range 1.7-2.7), respectively), which contrasts to the pattern observed in PBMCs from healthy donors (median SI 2.8 versus 6.8). Two out of 16 SCT patients showed a positive FCXM test. One of these had multispecifi c HLAantibodies before DSC infusion. The other patient showed a positive FCXM test four weeks after DSC infusion, but had no anti-HLA antibodies as confi rmed by Luminex analysis. Discussion: AM and DSC-infusions may improve the healing process of blisters in JEB-H patients, but the response appears to be transient. DSCs may induce multispecifi c anti-HLA class I antibodies in patients with a competent immune system. None of the SCT patients had developed anti-HLA antibodies after DSC infusion, indicating that the risk of alloimmunization by DSCs is low in immunocompromised patients. Disclosure of Interest: None Declared.

K. Chieregato 1,2,* , C. Zanon 1 , S. Castegnaro 1 , F. Rodeghiero 1,2 , G. Astori 1 1 Department of Cellular Therapy and Hematology, San Bortolo Hospital, 2 Hematology Project Foundation, Vicenza, Italy Introduction: Cytokine induced killer cells (CIK) comprise two main fractions: the fi rst, CD56 + , represented by NK-T cells (CD3 + CD56 + ) and NK cells (CD3 -CD56 + ), the second, CD56 -, characterized by T lymphocytes (CD3 + CD56 -). The CD56fraction is thought to be responsible of CIK alloreactivity, whereas the antitumor eff ect seems to be mainly restricted to the CD56+ fraction. Here, we investigated the CD56 + alloreactivity and cytotoxicity by evaluating the cytokine secretion profi le and lytic degranulation. Materials (or patients) and methods: CIK (n=6) were obtained by stimulating PB-MNC with IFN-gamma (day 0), IL-2, anti-CD3 monoclonal-antibody (day 1) and IL-2 every 3 days. At day 21, CD56 + and CD56fractions were immunomagnetically sorted. Cell alloreactivity of the CIK bulk and the sorted populations was tested by mixed lymphocyte reaction (MLR), cytotoxicity against K562 cells by calcein-release assay. Cell degranulation was quantifi ed by CD107a expression and Th1 (IFN-gamma, TNF-alfa, IL2) and Th2 (IL4, IL6, IL10) cytokines were then quantifi ed by ELISA in the supernatant after MLR and cytotoxic assay. Results: The CIK alloreactivity was mainly restricted to the CIK bulk and to the CD56-subset as confi rmed by the Th1/Th2 cytokines released in the supernatant. The CIK cytotoxicity was fully retained by the CD56+ cell subset, and appeared to be independent by lytic degranulation as confi rmed by the absence of diff erences in the expression of CD107a on NK, NK-T and by the cytokines released in the supernatant. This suggests that the CIK cytotoxicity could be mediated by an alternative mechanism probably dependent by FAS-L or CD3 pathways. Discussion: In a clinical setting, the immunomagnetic purging of the CD56-cell fraction by the CIK bulk could limit the risk of Graft Versus Host Disease maintaining the antitumor eff ect restricted to the CD56+ cell subset. Disclosure of Interest: None Declared. Introduction: Donor lymphocyte infusions (DLI) can produce lasting remissions in patients with relapsed chronic myeloid leukemia (CML), but are less eff ective in non-CML diseases. Chemotherapy-induced lymphodepletion prior to DLI, achieved with cyclophosphamide (Cy) and fl udarabine (Flu), has been shown to enhance activation of donor lymphocytes and cause signifi cantly more acute graft-versus-host disease (GVHD) than DLI alone. To safely balance the toxic versus benefi cial eff ects of activated donor lymphocytes, we combined lymphodepleting chemotherapy with the infusion of donor T cells engineered to carry a suicide gene to treat patients with aggressive hematologic malignancies. Materials (or patients) and Methods: Donor T cells were transduced with a retroviral vector expressing the Herpes simplex thymidine kinase (TK) suicide gene and the human Thy-1 selection marker and expanded in culture for 15 to 19 days prior to their infusion. In this phase I/II trial, the safety and effi cacy of Cy/Flu/DLI-TK was studied between February and December 2011 in 10 adults with relapsed multiple myeloma (n=5) or myelodysplasic syndrome/ acute leukemia (n=5). One had previously failed to respond to at least one standard DLI while others had not received any DLI before inclusion. All were free of immunosuppressive therapy at inclusion. Patients were infused with a mean cell-dose of 5 (range: 1-10) x10 6 CD3 + cells/kg of which 98% were TK + cells (range: 97%>99%). DLI products contained a mean ratio of 0.6% (range: 0.1-1.6) CD4 + FoxP3 + Helios + regulatory T-cells. This trial was registered at www.clinicaltrials.gov as NCT 01086735. Results: The Cy/Flu regimen was profoundly lymphodepleting and led to a mean 3-day period of neutropenia below 500 PMN/μL. Three patients developed acute GVHD following TK + cell infusion. One patient had a grade I cutaneous GVHD resolving with local steroids. In another patient with grade III cutaneous and digestive involvement, intra-venous administration of ganciclovir (GCV) led to the complete and lasting resolution of symptoms, correlated with clearance of TK + cells in peripheral blood. A third patient had a grade III liver GVHD. Due to only partially controlled leukemia at the time of DLI, GCV treatment was postponed for one week after GVHD onset in order to support a putative GVL eff ect. Treatment with GCV led to the rapid disappearance of TK + cells in peripheral blood and liver but without clinical improvement after 2 weeks, so that an additional immunosuppressive therapy should be instituted. Six patients, including one of the 2 treated with GCV, experienced relapse/progression of their malignancy and received additional anti-cancer therapy. With a median follow-up of 22 months after Cy/Flu/DLI-TK, 6 patients are alive, 5 of them being disease-free. In patients not treated with GCV, TK+ cells could still be detected in peripheral blood till at least 12 months after their infusion. Discussion: Lymphodepletion followed by suicide-gene-transduced T-cell infusion may help to safely enhance the immune activity of DLI, which could be further improved by regulatory T-cell depletion (Maury et al, Science Translat Medicine 2010). Disclosure of Interest: None Declared.

Introduction: Mesenchymal stromal cells (MSC) amplifi ed in vitro are able to inhibit steroid-resistant graft-versus-host disease (GVHD) occurring in some patents reconstituted with allogenic hematopoietic stem cells. However, it is not established whether the potent regulatory eff ect of MSC is restricted to alloreative T cells or whether it may also modulate residual disease relapse. We explored this issue in vitro using Jurkat T cells (Jurkat) as a model for residual disease. Materials (or patients) and Methods: MSC were purifi ed from femoral head remains after informed consent of patients undertaking hip surgery, and amplifi ed in the presence of platelet lysate. Amplifi ed MSC were plated in 24-well plates and Jurkat were seeded either directly on MSC or were deposited in transwells above the MSC. Cultures were performed for 48 hours in RPMI medium with 10% FCS, with or without 250 μ/ml γ-interferon (IFN). Cell cycle status of Jurkat, as well as L-tryptophan (Tryp) and kynurenine (Kyn) titers of supernatants were determined. The sensitivity of Jurkat toward Tryp depletion was assessed using Tryp-free RPMI and graded concentrations of the amino acid. Results: When incubated with MSC and IFN, apoptotic and G-phase of Jurkat were respectively increased by 6.3-fold and 1.3-fold, and SG2M phase represented 0.5-fold the value observed in cocultures with MSC in the absence of IFN (mean of 5 experiments). IFN alone, in absence of MSC had no eff ect on Jurkat cell cycle. MSC-induced inhibition was eff ective both in regular cultures and when Jurkat were seeded in transwells preventing cell contacts. Jurkat proliferation inhibition was associated with a decrease of Tryp and an increase of Kyn titers in the culture supernatants. Complementation experiments using Tryp-free RPMI confi rmed that Jurkat required Tryp for proliferation and showed that Tryp concentrations <3 μM altered cell cycle. Kyn did not mediate MSC inhibition as the complementation of Tryp-replete cultures with Kyn did not modify the cell cycle status. Discussion: This study shows that MSC regulatory eff ect extends to malignant cells. MSC-induced Jurkat inhibition was direct, required exogenous IFN and was mediated via a cell contactindependent mechanism that involved Typ catabolism. These observations suggest that MSC infused to a patient to fi ght the GVHD may, in addition of controlling the GVHD, inhibit residual disease progression as long as leukemic cells are sensitive to Tryp depletion. This may be the case for B lymphocyte or T lymphocyte derived malignancies, whose normal counterparts are highly sensitive to Tryp depletion. By contrast, if the residual disease is resistant to low Tryp titers, repetitive infusions of MSC should be avoided because they may also inhibit T cells specifi c for the residual disease and favor in this way its relapse. Altogether these data suggest that cells derived from the mesenchymal lineage exhibit a control upon tumor cells that has been largely underrated so far. It may well be that MSC, in the presence of IFN (known to inhibit tumor growth in many instances) may eradicate a signifi cant portion of Tryp-sensitive tumors without the help of adaptive immunity. However once facing a Tryp depletion-resistant tumor, MSC activation via IFN may lead to tumor escape. Further investigations addressing this issue are required before establishing the innocuousness of repetitive infusions of MSC in leukemic patients. Disclosure of Interest: None Declared. 3 Unit of Molecular Neuro-Oncology, IRCCS Neurological Institute C.Besta Foundation, Milano, Italy Introduction: In the last few years cell therapy based on dendritic cells (DC) pulsed with tumor lysate become a promising approach in addition with conventional therapy for the treatment of glioblastoma (GB)-aff ected patients. The success of this approach strongly depends on the possibility to generate routinely large amounts of high quality, functional mature DC (mDC) in compliance with Good Manufacturing Practices (GMP). Moreover, a high level of standardization is required to ensure maximum reproducibility of the mDC production process. Materials (or patients) and Methods: In Cell Therapy Production Unit (UPTC) of Neurological Institute C. Besta Foundation a protocol for mDC production under GMP conditions was validated for treatment of GB-patients. From 2010 to 2012 37 lots of mDC were produced. The aim of this study was to evaluate retrospectively: a) sterility, absence of mycoplasma, endotoxin and adventitious viruses, b) feasibility of producing a large number of autologous mDC, c) quality of mDC, in terms of viability, maturation status and potency. Compendial methods, according to European Pharmacopoeia were used to test sterility, mycoplasma, endotoxin and adventitious viruses. Non compendial methods were applied to evaluate viability, phenotype and potency. In particular viability was assessed by trypan blue exclusion test, phenotype by cytofl uorimetric analysis of the typical mDC markers CD80, CD83, CD86, HLA-DR and potency by Mixed Lymphocyte Reaction (MLR). Results: 37/37 lots of mature mDC were sterile, endotoxin levels were <2.86 UI/ml (cut off value) and in all lots mycoplasma and adventitious viruses were absent. With respect to the number of mature mDC obtained, 35/37 lots were conform for the total number of cells for ensure the full treatment of patients; in 2/37 lots the fi nal number of mDC was insuffi cient to ensure all the planned vaccinations. All mDC lots showed very high viability after thawing (Mean: 93,82%, SD:3,20%). Phenotype evaluation of mDC showed a marked up-regulation of the typical DC markers in comparison with immature DC. Moreover the values among the lots were very constant: CD80 (Mean: 79,5%, SD:5,5%), CD83 (Mean: 53,7%, SD:16,2%), CD86 (Mean: 94,5%, SD:2,1%), HLA-DR(Mean: 98,5%, SD:1,4%). MLR test for the evaluation of the potency of the mDC demonstrated that all lots were functional, with respect to their ability in the induction of lymphocytes response. Discussion: These results demonstrate that our protocol for DC production is highly reproducible and permits routinely to generate large amounts of safe and functional mDC for in vivo use in immunotherapy approaches. Disclosure of Interest: None Declared. S134 containing rhIL-2. The expanded populations were characterized by fl ow cytometry. Results: We performed 18 expansions starting from a mean of 10 ml peripheral blood from untreated CLL patients (range 2-30 ml) that contained a mean of 9.3% T cells. This method allowed reproducible expansion in about 22 days of mean 410x10 6 CD3 + T cells (range 71 to 2184x10 6 ) with a mean 224 fold expansion (range 4.4-1326) . Final products were depleted of CLL cells from days 7-14 onwards (mean 0% at day 18-25), the only signifi cant contaminant being NK cells (mean 18.5%). Only in one case, a signifi cant percentage of CLL B cells could be observed at the end of culture (day 24), but this was due to the particularly high percentage neoplastic cells in the starting population in this patient (98%), resulting in relatively late depletion of these cells, which took place between days 14 and 21, and therefore remained detectable at day 24 (3.8% CLL B cells at day 24). Despite the very low percentage of starting T cells in this specifi c patient (1.2%), 152x10 6 T cells could be obtained, with a 42 fold expansion. The resulting blinatumomab expanded T cells (BET) contained both CD4 + and CD8 + cells in varying proportion. Only in one case the fi nal product was composed predominantly of CD4 + cells (95%). Expanded T cells were polyclonal, as shown by TCR Vβ expression by fl ow cytometry. Indeed CMV specifi c clones, detected by CMV peptide (pp65 495-503 )-loaded HLA-A*0201 tetramer, were also expanded using this method. Final T cells were composed predominantly of the eff ector and central memory subsets. Th1 were slightly prevalent over Th2 cells (mean 20% and 10%, respectively), whereas Th17 and Treg were less than 1%. We also observed that the CD272 and CD279 synapse inhibitors diminished in BET compared to the starting CLL T populations (from 73% to 19% and 61% to 18%, respectively) . These data suggest that stimulation and expansion with blinatumomab and rhIL-2 has normalized expression of these regulators on CLL T cells. Indeed BET were highly cytotoxic against CD19 + targets cell lines or primary CLL cells, with 70-90% lysis at a 3:1 eff ector target ratio in presence of blinatumomab. Finally BET were compared to Xcellerated cells expanded using anti-CD3/CD28 Dynabeads and rh-IL-2. The 2 protocols showed equivalent effi ciency and comparable cell composition at the end of culture. Discussion: We conclude that the use of blinatumomab and rhIL-2 provides a reproducible, simple and GMP-compliant protocol, allowing expansion of large numbers of autologous polyclonal T cells from relatively small volumes of peripheral blood from CLL patients and effi cient simultaneous depletion of CLL cells. This procedure is an attractive option for adoptive therapy for these patients after immunosuppressive treatments. Disclosure of Interest: J. Golay: None Declared, A. D'Amico: None Introduction: Cytomegalovirus (CMV) plays an important role in the morbidity and mortality of patients after allogeneic hematopoietic stem cell transplantation (HSCT). Cellular therapies such as adoptive transfer of CMV specifi c cytotoxic T lymphocytes (CMV-CTLs) are currently applied to control therapy refractory CMV reactivation. However in-vitro expanded CMV CTL can lose their proliferative potential and their capacity to persist in the patient after adoptive transfer. A better understanding of the phenotype of in vitro expanded CMV-CTLs may help to further improve current CTL expansion techniques. Aims: Here we studied the impact of conventional CD14+ derived dendritic cells (convDCs) and self-diff erentiated myeloid-derived antigen presenting-cells (smartDCs) on the CMV CTL phenotype and the functional activity. The analysis comprised exhaustion and senescence markers and cytokine release of the expanded CMV-CTLs. Materials (or patients) and Methods: The convDCs were generated by culturing CD14+ cells in the presence of recombinant human IL-4 and GM-CSF for 5 days. Self-diff erentiated myeloid-derived antigen presenting-cells (smartDCs) were generated by transduction of CD14+ cells with a lentivirus encoding GM-CSF, IL-4 and the CMVpp65 protein and subsequent in-vitro culture for 7 days. Both convDCs and smartDCs were matured on day 5 (IL-1b, IL-6, TNF a and PGE 2 ) and cultured for 38-48 hours. Peptide loaded convDCs and endogenously expressing CMVpp65 smart DCs were used to stimulate CMV-CTLs for 7 days. The expanded T-cells cells were analysed based on the surface marker expression by a multicolour fl ow cytometry panel. Exhaustion was monitored by PD-1 (programmed death-1) and anti-Tim3 (T cell immunoglobulin domain and mucin domain 3) expression. Senescence was studied by CD57 expression. Results: In the cohort of three CMV seropositive donors (n=3) we have observed that memory distribution (CCR7+CD45RA-, CCR7-CD45RA-) of the expanded T-cells remained the same in CD8 T cells and CMV-CTLs regardless of the use of convDCs or smartDCs for stimulation. CD8 T-cells stimulated with con-vDCs loaded with CMV pp65 peptide pool and smartDCs endogenously expressing CMV pp65 showed significant increases of Tim3, PD1 in comparison to unstimulated CD8 T-cells (Tim3: P<0.0001, PD1: P<0.01). Similarly we found increase of Tim3 on the CMV tetramer positive subpopulation after stimulation with convDCs loaded with CMV pp65 peptide pool and smartDCs (P<0.05). Finally, we found a significant increase of CD57 on CD8 T cells stimulated with convDCs loaded with CMV pp65 peptide pool compared to unstimulated T-cells but not in CD8 T-cells stimulated with smartDCs endogenously expressing CMVpp65 (CD8-CD57: P<0.05). Interestingly, no increase of CD57 was found for the CMV tetramer positive subpopulation after stimulation with convDCs loaded with CMV pp65 peptide pool and smartDCs. Discussion: Our preliminary in-vitro data show an increase in the exhaustion markers (Tim3, PD1) on both CD8 T cells and CMV-CTLs stimulated with both conv DCs and smart DCs and CD57 on CD8 T cells stimulated with conv DCs. Further evaluation of these markers in CMV-CTL might help to improve adoptive transfer by defi ning a subpopulation leading to effi cient and prolonged control of viral reactivation. Disclosure of Interest: None Declared. S135 (AML). The preparative regimen consisted of intermediate doses of Cyklophosphamide (Cy), Fludarabin (Flu) and titrated doses of total lymphoid irradiation (TLI). The trial design excluded systemic IL-2 treatment to avoid expansion of regulatory T cells. Materials (or patients) and Methods: The fi rst 6 patients was treated with Cy/Flu and TLI followed by infusion of short-term (16hours) IL-2 activated NK cells. Four patients had high risk MDS or MDS-AML and two had relapsed primary AML. Results: The treatment was well tolerated and no severe noninfectious toxicity has been noted in the patients. Three of the patients had positive microchimerism, with detectable NK cells of donor origin at day 7, but none of the patients reached the primary endpoint (>100 donor NK cells/μl at day 14). Nevertheless, all patients displayed reduced tumor burden 1 month after therapy. Interestingly, two patients achieved complete remission that lasted at least 3 months and one of these has proceeded to allogeneic stem cell transplantation. Two additional patients had partial remission with stable blast cell counts for 3 and 6 months in the absence of additional therapy, respectively. Two patients with minor response and progressive disease died in infections within three months of therapy. Discussion: Although the long-term effi cacy needs to be evaluated, the results suggest that NK cell therapy may induce remission in patients with chemo-refractory disease and provide a bridge to allogeneic stem cell transplantation. Disclosure of Interest: None Declared. Introduction: We applied prophylactic CD8-depleted (CD8 depl ) donor lymphocyte infusions (DLI) in the setting of T-cell depleted allogeneic hematopoietic stem cell transplantation (HSCT) in a phase I/II trial. T-cell depletion was carried out by the use of high-dose Alemtuzumab (100mg or 60mg for unrelated or sibling donor transplantation, respectively). We have demonstrated the feasibility of this approach after having treated 23 patients in this protocol (Meyer et al. Blood 2007) . Materials (or patients) and Methods: From 2004 to 2011 patients with diff erent hematologic diseases were included and followed for a median observation time of 2.6 years after transplantation (range, 1.5-9 years). Median age was 56 years (range, 19-71). Stem cell source was peripheral blood from either matched siblings (n=23), matched unrelated (n=62), or donors with single HLA mismatches (unrelated n=48, related n=1). Tapering of cyclosporine A was started in the 6 th week after transplantation. Subsequently, CD8 depl DLI were administered prophylactically in escalating doses starting with 1x10 6 CD4 T cells/kg bodyweight. Results: 53 patients received at least one dose of DLI. Among 81 patients who did not qualify for DLI, 60 patients had primary mild GVHD. Following DLI, acute GVHD was the major reason for withholding subsequent DLI-doses (66%), mainly because of acute GVHD ≤ 2°. Extensive chronic GVHD was diagnosed in 9.4% of the patients. Overall survival after 1 and 3 years was 60% and 43%, respectively. Survival signifi cantly diff ered between the DLI and non DLI group after 3 years (63.8% vs. 30.1%, P=0.002). The inferior outcome of the non-DLI group was very similar when only those patients who did not receive DLI despite the absence of GVHD were considered (28.6%). The largest patient cohort in our trial were patients with AML and MDS (n=21 with and n=33 without DLI). In these patients, the survival benefi t for the DLI group after 3 years was signifi cant (75.8% vs 28.9%, P=0.0012). Interestingly, the relapse rate did not diff er between DLI and non-DLI patients. The presence of GVHD at any time was associated with a reduced relapse rate (57.9% vs. 31.4%, P=0.0015), independent of DLI-application. However, GVHD had no impact on overall survival. Decreasing donor T-cell chimerism (TCC) was found in 34 patients who subsequently received DLI and 13 who did not. Following CD8 depl DLI, 31 patients (88%) converted to full donor. In contrast, only 2 of the patients with decreasing TCC in the non-DLI group (15.3%) converted spontaneously. All patients with mixed TCC relapsed later on. Discussion: In summary, we observed that the application of prophylactic CD8 depl DLI was associated with a survival benefi t and the re-establishment of full donor T-cell chimerism in the context of T-cell depletion. Still, this non-randomized trial is not suffi cient to demonstrate any causality. In AML/MDS-patients the improved survival seems to be associated with lower treatment-related mortality rather than a reduced relapse-rate. Our data strongly ask for randomized trials comparing prophylactic application of CD8 depl DLI vs. no DLI as well as CD8 depl vs. non-manipulated DLI in a preemptive setting. Disclosure of Interest: None Declared. Introduction: The impact of donor lymphocyte infusion (DLI) on its outcome is well known in patients with chronic myeloid leukaemia but still limited in patients with other haematological disease relapsed after allogeneic stem cell transplant (SCT). Aim. To analyze the effi cacy of DLI either in patients relapsed after SCT and as pre-emptive purpose. Materials (or patients) and Methods: Fifty -seven high -risk haematological were included in the retrospective analysis: 20 (35.0%) acute myeloid leukaemia (AML); 17 (30.0%) lymphomas; 8 acute lymphoblastic leukaemia (ALL); 8 multiple myeloma (MM); 4 chronic myeloid leukaemia (CML). Forty-four (77.0%) patients had relapsed or a progressed disease at fi rst DLI; 6 (5.5%) had molecular relapse; 3 (5.5%) had a chronic phase CML, 2 (3.5%) patients were treated for mixed chimerism and 2 (3.5%) as a pre-emptive program. A dose escalated program of DLI was performed with an initial DLI CD3+ cell dose per kilogram of recipient body weight equal to 0.5x10 7 /kg for sibling transplant and to 0.5x10 6 /kg for unrelated transplant. Results: A median of 3 (1-9) DLI was performed. Twenty-two (38.5%) patients developed acute graft -versus -host disease: grade III in 7 (12.0%) patients. No DLI related deaths were observed. With a median follow-up of 12 months , 30 (52.5%) patients are alive: 24 (428.0%) in complete remission. All the patients treated for molecular relapse acute leukaemia or for chronic phase CML (9 -15.5%) are alive in molecular remission. Fourteen (32.0%) of the 44 patients treated for relapsed/progressed disease are alive in complete remission. Full donor chimerism was documented in the two patients treated for mixed chimerism. One -year overall and progression -free survival are 72.0% and 70.0% respectively. Discussion: a lot of open questions still remain regarding DLI, but our results support the use of DLI for relapsed/progressed patients, with a low toxicity. Moreover, as expected, high -rate of complete remissions were observed in early relapses, confi rming the relevance of molecular monitoring after transplant. For high -risk patients, DLI should be considered as pre-emptive program after transplant. Disclosure of Interest: None Declared. S136 1, 2 , A. Perez Corral 1, 2 , C. Pascual 1, 2 , J. Gayoso 1, 2 , P. Balsalobre 1, 2 , D. Serrano 1, 2 , C. Muñoz 1, 2 , I. Buño 1, 2 , J. Anguita 1, 2 , 2 1 Hematology, Hospital General Universitario Gregorio Marañón, 2 Instituto de Investigación Sanitaria Gregorio Marañón, Madrid, Spain Introduction: Little consistent data on the benefi cial eff ect of NK cells has been reported in patients receiving umbilical cord blood (CB) transplantation. Our aim is to evaluate the infl uence of donor and recipient KIR genes content, in the setting of haplo-cord SCT (stem cell transplant) consisting of single CB in combination with CD34+ cells from a third party HLA-mismatched donor. Materials (or patients) and Methods: 31 consecutive patients with haematological malignancies who received haplo-cord SCT between 2004 and 2012, were included ( 2 Hematology Division, General Peripheral Hospital of Patras "Agios Andreas", Patras, Greece Introduction: Major obstacles in using FOXP3 + T regulatory cells as T-cell based immunotherapy against GVHD after allogeneic hematopoietic cell transplantation are their low numbers in the circulation and the lack of specifi c cell surface markers for efficient purifi cation. DNA methylation has been considered to play a role in the regulation of T-cell eff ector function and cytokine gene expression, indicating a promising role of hypomethylating agents for immunomodulation. Recently it was shown that in-vitro treatment of conventional T-cells with hypomethylating agent azacitidine (aza) induced FoxP3 expression and converted CD4 + CD25cells into immunosuppressive T-cells the suppressor function of which is independent of FOXP3 expression (Blood 2010; 116:129-139) , suggesting that aza induced suppressor function depends on the modifi cation of other hypomethylated genes. Human leukocyte antigen-G (HLA-G) is a non-classical HLA class I molecule, shown to exert immunoregulatory functions, the [PH-P076] S137 expression of which is epigenetically regulated. We investigated whether hypomethylating agent aza can induce HLA-G + immunoregulatory T cells. Materials (or patients) and Methods: Negative selected CD3 + T cells from peripheral blood of healthy individuals were stimulated with anti-CD3 + plus anti-CD28 + coated beads and then treated for 72 hours with aza (0.5-15 mM) in the presence of 50U/ml inteleukin-2 (IL-2). Phenotypical characterization of in-vitro aza treated cells was performed with fl ow cytomentry. Aza-induced HLA-G + T cells, were irradiated and then used as third party cells in CFSE based suppression assay, in which anti-CD3/CD28 beads were used as stimulators. For the analysis of the in-vivo eff ect of aza on HLA-G expression, peripheral blood mononuclear cells (PBMC) of patients with myelodysplastic syndrome (MDS) were isolated at baseline and after Vidaza treatment and were analyzed by FACS for HLA-G expression. Results: In-vitro treatment of CD3 + T cells with aza increases the percentage of HLA-G + cells. The optimum aza concentration for maximum HLA-G induction with the lowest toxicity in CD3 T cells at protein and mRNA level, was determined at 5 mM (HLA-G + :6.88±3.9%, P=0.0022). Maximum HLA-G induction was observed on CD4 + CD25 + population (n=2, 9.21±5.5%). Aza treatment of FACS-sorted CD4 + CD25 neg HLA-G neg cells induced CD4 low CD25 + HLA-G positive cells, revealing that aza induced HLA-G + cells are not the result of a selectively expanded proexisting HLA-G + population. Strikingly aza induced CD4 low CD25 + HLA-G + are FoxP3 negative. In-vitro Aza induced HLA-G + cells display HLA-G dependent suppressive function. The % of CD4 + HLA-G + and CD8 + HLA-G + peripheral blood lymphocytes of healthy donors was 1.56±0.5 and 3.03±1.2 respectively while in MDS-Vidaza treated patients was 4.25±0.7 and 2.28±0.3 respectively. Although preliminary data do not show a signifi cant increase in % and absolute number of CD4 and CD8 lymphocytes of MDS-Vidaza treated patients, CD4 + CD25 high HLA-G + cells showed a 5 fold increase on day 15 post Vidaza treatment. Discussion: In conclusion, we generated a CD4 low FoxP3 neg immunoregulatory population, which expresses extracellulary HLA-G and therefore can be easily isolated for adaptive T-regulatory therapies. On the other hand, our results indicate that the use of aza post-transplant as a mean of reduction of disease relapse risk is questionable since aza may impair the GvL eff ect. Disclosure of Interest: None Declared.

V. Volpin 1,* , N. Cieri 1,2 , C. Bonini 1 1 Experimental Hematology Unit, Division of Regenerative Medicine, Gene Therapy and Stem Cells, San Raff aele Scientifi c Institute, Milan, Italy Introduction: Costimulation plays a critical role for T-cell activation. Recent studies have highlighted that costimulatory signals are required not only for effi cient naïve T-cell priming but also for secondary recall responses. Nevertheless the costimulatory requirements of memory responses have been much less studied than those of primary responses, especially in humans. Nonetheless, the study of the costimulatory requirements of memory T cells is of high relevance considering that memory T cells play a critical role in several diseases, including infections, cancer and autoimmunity. Here, we evaluated the ability of the costimulatory molecules of the CD28-family (CD28 and ICOS) and of the TNFRfamily (4-1BB, OX40, GITR, CD27, CD30, HVEM) to ex vivo expand human memory T-cell subsets, namely central memory (T CM ), eff ector memory (T EM ) and the recently identifi ed memory stem T cells (T SCM ), for immunotherapeutic purposes. Materials (or patients) and Methods: To evaluate the eff ects of costimulatory signals on memory T cells, we activated purifi ed memory T-cell subsets with coated anti-CD3 antibody and soluble agonistic antibodies directed against costimulatory molecules. Retroviral transduction was performed to formally prove the pro-liferation and postmitotic status of analyzed cells, and low doses (5ng/ml) of homeostatic cytokines were added to cultures. The effi cacy of costimulatory molecules was measured in terms of T-cell expansion and preservation of the original surface phenotype. Results: In resting conditions, we observed that the large majority (more than 90%) of memory T cells were positive for CD28, CD27 and HVEM costimulatory molecules, while expression of ICOS, 4-1BB, OX40, GITR, and CD30 was up-regulated upon TCR triggering. Interestingly, we found that each human memory T-cell subset best responded to distinct costimulatory signals: CD28 best supported the expansion of T SCM coupled to the preservation of original phenotype, while CD27 best expanded T CM and T EM lymphocytes. In addition, CD27 and CD28 imprinted distinct functional programs, fostering gIFN and IL-2 production respectively in all subsets analyzed. We next assessed IL-7 Receptor alpha (CD127) expression, a marker of T-cell fi tness. Among T SCM cells, CD28-costimulation preserved CD127 expression on CD4+ T cells signifi cantly better than 4-1BB-, CD27-and CD30-mediated costimulation. Discussion: Altogether, our data suggest that tailoring the costimulatory signals to the T cell subset to be expanded could improve the ex vivo manipulation protocols necessary for adoptive immunotherapy of cancer and infectious diseases. Disclosure of Interest: None Declared. Introduction: Dendritic cells (DCs) have the unique ability to prime naïve T-cells, thus representing a fundamental tool in cancer therapy and especially in vaccine-or antitumor specifi c T cell-based immunotherapy. However, the laborious and costly methods to generate and expand DCs from the peripheral blood (PB) or monocytes, limit their potential for broad clinical applications. In addition, although DCs can be produced from cord blood (CB)-derived CD34+ cells, the fi nal DC yields are low to be clinically translated. New methods and sources are needed in order to obtain suffi cient numbers of DCs (10 6 -10 8 DCs). Materials (or patients) and Methods: In this study we sought to investigate the potential of large scale generation of DCs from CBderived CD34+ cells, using the G-Rex10 bioreactors (Wilson Wolf Corporation) instead of the standard culture dishes or fl asks. We additionally asked for the optimal culture conditions and tested the capability of produced DCs to induce Th1 responses upon stimulation with a mixture of Toll Like Receptor ligands (TLR-Ls). Non transplantable CB units were used after a signed informed consent from the parents. CD34+cells were enriched to more than 90% purity from CB units (n=5) by an immunomagnetic separation method. The CD34+ cells were cultured in the presence of ABblood group serum (ABS) and culture media supplemented with SCF (50ng/ml) and GM-CSF (100ng/ml) for 2, 3, 4, or 5 weeks (wks) and with IL-4 (50ng/ml) plus GM-CSF (100ng/ml) for 1 additional wk. The cells were initially plated in a 6-well plate (10 5 cells/well) and once the cells expanded up to 5x10 6 , half were transferred into G-Rex10 bioreactor and half were cultured in "conventional" culture plates in the presence of the above mentioned conditioned medium. In culture plates the cells were splited at confl uency while in G-Rex10 every 3 days. Results: The highest median absolute number of myeloid DCs (CD33+/CD11+) in the bioreactor was obtained by 5-wks culture over the 3-, 4-, and 6-wks culture (1.5x10 9 vs 0.017x10 9 vs 0.8x10 9 vs 1.2x10 9 , respectively). The number of myeloid DCs that were conventionally cultured under the same conditions did not exceed a median number of 10x10 6 cells. To evaluate the impact of the serum origin in DCs expansion, we cultured CB-derived CD34+ cells into G-Rex10 bioreactor, in the presence of either autologous CB serum (ACBS) or ABS. At the end of the culture, we identifi ed 1.5x10 9 DCs with myeloid characteristics in the ABS culture while only 0.011x10 9 in the ACBS culture (P<0.05). To address whether DCs expanded from CB-derived CD34+ cells have the ability to produce Th1 responses, we stimulated DCs with a mixture of TLR-L 3 (Poly I:C: 20μg/ml) and TLR-L 7/8 (R848: 4μg/ml) for 48 hours and measured the levels of IL-12p70, TNF-α, IL-6 and IL-10 by ELISA. Increased levels of IL-12p70, TNF-α and IL-6 were detected, while the IL-10 levels were low to undetectable, indicating that the produced DCs with myeloid features have a strong potentiality for Th1 responses. Discussion: Overall, we report for fi rst time, over a 10 4 fold myeloid DCs expansion from CB-derived CD34+cells, by using the new generation G-Rex10 bioreactors and describe optimal culture conditions. This large scale myeloid DC generation could signifi cantly contribute to the clinical applicability of DCs in cancer immunotherapy. Disclosure of Interest: None Declared.

Introduction: Myelodysplastic syndromes (MDS) are a heterogeneous group of clonal hematopoietic stem cell diseases characterized by ineffi cient haematopoiesis with frequent progression to acute myeloid leukaemia (AML). Chronic myelomonocytic leukaemia (CMML) is a clonal haematopoietic malignancy characterised by the features of both myeloproliferative neoplasm and MDS, such as monocytosis, variable degree of leukocytosis, anaemia, thrombocytopenia and dysplastic changes in diff erent haematopoietic lineages, and has poor prognosis with median survival of barely 1 to 3 years. Both MDS and CMML are disorders of elderly and, with the exception of allogeneic bone marrow transplantation (ASCT), have no curative treatment options. Because of the older age, only a minority of CMML patients are eligible for ASCT. In recent years, active immunotherapy (AIT) using dendritic cell (DC) vaccines transfected with various leukaemia-associates antigens has become an exciting concept for treating these patients. Moreover, recent reports have indicated eff ectiveness of combining chemotherapy with AIT that leads to a longer overall survival of cancer patients. Materials (or patients) and Methods: The purpose of this study was, fi rst of all, to fi nd out whether myeloid-derived DCs generated from MDS patients acquire phenotypic and functional characteristics of mature DCs. Secondly we were interested in how many of investigated patients could be potential candidates for DC vaccine treatment taking in account their clinical characteristics and general condition. Importantly, monocytes of MDS patients are often a part of malignant clone. In our experiments we used PB MNCs from 10 CMML patients, 5 non-CMML patients and 5 healthy controls. Median age of CMML vs. non-CMML patients was 69 and 68, respectively. Mean absolute monocyte count was 4,0x10 9 /l and 1,0x10 9 /l for CMML vs. non-CMML patients. Mean haemoglobin was 13 vs. 12 g/dl for CMML and non-CMML patients. Four CMML patients had abnormal karyotype by conventional cytogenetics (i.e. +8) or mutational analysis (RUNX1, JAK2), while 3/5 non-CMML patients had various cytogenetic abnormalities. Two CMML patients were previously treated with either hydroxyurea or azacitidin, while all 5 non-CMML patients were under active treatment at the time of sampling. All patients had ECOG status 0-1. Results: Mature DCs were developed using a novel production protocol consisting of two days of culturing peripheral blood monocytes in the presence GM-CSF and IL-4 followed by one day of maturation using GM-CSF, IL-4, TNF, INF-gamma, PGE 2 , IL-1beta and R848. Our results show that we were able to generate mature functional DCs, assessed by phenotype, Signal-3 and/or migration assays, in 9/10 CMML cases and all non-CMML patients. There were insignifi cant variations in down-regulation of CD14 and expression level of CD80, CD86, CD40, CD83, CD274, CCR7 and HLA-DR for CMML and MDS patients compared to normal controls. Discussion: Taken together, the results of our study show that all but one patient in the CMML group and all patients in non-CMML group could be relevant candidates for vaccination within a clinical protocol using DCs transfected with WT1 and PRAME mRNA (protocol under preparation Introduction: More than 300 NIH registered clinical trials applied mesenchymal stem cells (MSCs) to treat patients with a variety of diff erent diseases e.g. myocardial infarction, stroke and graft-versus-host disease (GvHD). Initially, MSCs were thought to replace lost cells in damaged tissues. Despite controversial reports regarding the outcome of MSC treatments, they rather seem to exert their benefi cial eff ects by the secretion of immunosuppressive factors. In this context small extracellular vesicles (80-140 nm), exosomes, were identifi ed to mediate the immunosuppressive eff ects of MSCs. Indeed, by treating a steroid-refractory GvHD patient with MSC-exosomes containing high concentrations of anti-infl ammatory cytokines (TGF-b1, IL-10, HLA-G), we gained evidence that they can eff ectively suppress GvHD symptoms. Assuming that MSC-exosomes of diff erent donors diff er in their immunosuppressive potential, we now have compared the impact of diff erent MSC-exosome preparations on the proliferation behavior of stimulated T cells. Materials (or patients) and Methods: Carboxyfl uorescein succinimidyl ester (CFSE)-labeled PBMCs were stimulated with phytohaemagglutinin (PHA) in the presence of exosomes enriched from supernatants of diff erent donor-derived MSCs. After 5 days, the CFSE intensity-as a marker for lymphocyte proliferation -was assessed by fl ow cytometry and analyzed for CD3-positive cells. Results: CFSE-intensity-as an indicator for lymphocyte proliferation -of CFSE-labeled PBMCs which were stimulated with PHA in the presence of exosomes enriched from supernatants of diff erent donor-derived MSCs varies from 0.2% to 72%. Controls of CFSElabeled PBMCs without exosomes show a CFSE-intensity of 0.3% without and 61.7% after stimulation with PHA (s. Figure 1 ). Discussion: Our results confi rm the assumption that MSC-exosomes of diff erent donors diff er in their capability to modulate the proliferation stimulating impact of PHA on CD3 positive human T cells. Since in most clinical applications, especially in the treatment of steroid refractory GvHD patients, a strong immunosuppressive eff ect of MSC-exosomes is desired, we are currently searching for surrogate markers allowing the fast identifi cation of MSCs whose exosomes exert strong immunosuppressive impact and which can eff ectively be used in the clinical setting. Disclosure of Interest: None Declared.

Introduction: Mesenchymal stromal cells (MSCs) are multipotent cells that exert immunomodulatory eff ects; however, the mechanisms underlying these eff ects have not been completely clarifi ed. Aim of this study was to compare in vitro the immunomodulatory properties of MSCs with those of microvesicles (MVs) released in supernatants from the same MSCs. Materials (or patients) and Methods: MSCs were generated from bone marrow of 12 healthy donors (HDs) and MVs were isolated from their supernatant by serial ultracentrifugation. Both MSCs and MVs were characterized by fl ow cytometry and co-cultured with peripheral blood mononuclear cells (PBMCs) of 12 HDs and stimulated in vitro with both PHA and CpG to induce T and B cell proliferation, respectively. Growth factors and cytokines were quantifi ed in the supernatants by ELISA. Results: MVs were identifi ed as 1mm particles positive for CMFDA, CD107 and CD13 (suggesting a mechanism of membrane budding from MSCs). MSCs inhibited PHA-induced T-cell proliferation up to 80% (SD ±8.16; P=0.0001 as compared with condition PBMCs/ PHA) and 60% (SD ±20.86; P=0.0001) with MSC:PBMC ratios 1:2 and 1:10 respectively, whereas MVs reduced it up to 30% (SD ±39.32; P=0.01 as compared with condition PBMCs/PHA; MV dilution of 1:2 in co-culture fi nal volume). MSCs reduced CpG induced B-cell proliferation up to 70% (SD ± 2.82; P=0.002 as compared with condition PBMCs+CpG) and plasma cell activation up to 50% (SD ±10.59; P=0.001; MSC:PBMC ratio 1:10), whereas MV-induced inhibition was up to 60% (SD ±9.95; P=0.02 as compared with condition PBMCs+CpG) and 30% (SD ±13.53; P=0.02), respectively (same diluting conditions). In both T-and B-cell cultures, MSC coculture induced an increase of IL-6, IL-10, TGFb and a decrease of IL-2 and IFN, whereas in MV co-culture no diff erences were revealed in cytokines levels. Discussion: Our data indicate that MSC-derived MVs display a lower immunomodulatory eff ect in vitro, as compared to their cellular counterpart. Caution should be employed when considering the potential use of MV in substitution of MSCs in therapeutic approaches, especially when aiming at treating immune-mediated disorders. Disclosure of Interest: None Declared.

Introduction: Mesenchymal stromal cells (MSCs) represent a heterogeneous cell population concerning their proliferative, diff erentiation and allosuppressive potential. This is a donor-dependent property, which leads to the controversial clinical data. To avoid this great inter-donor variability, we established a GMP-compliant, serum-free MSC-Bank from the pool of bone marrow mononuclear cells (BM-MNCs) of eight allogeneic healthy donors. Materials (or patients) and Methods: Phase I: Pivotal experiments. In this phase we performed pivotal experiments, which led to the development of standard operating protocols of the MSC-Bank. We tested the BM-MNCs of 4 donors either individually or after pooling them for: clonogenicity, effi ciency to give rise to MSCs, proliferative and allosuppressive potential. Phase II: BM-sample collection and generation of MSC-Bank. Written informed consent from eight healthy bone marrow donors was obtained for collection of up to 230 ml bone marrow. The blood samples were tested for markers of transfusion transmitted diseases (TTD). BM-MNCs were isolated in GMP-facility and frozen separately in cryobags. After thawing, BM-MNCs from 8 donors were washed, pooled and cultured in DMEM supplemented with 5% platelet lysate (PL) until generated MSCs reached 80% confl uence. After harvesting, the MSCs were frozen in 209 cryovials 1.5x10 6 MSCs, which represent the MSC-bank. Phase III: Validation of MSC-bank. Three randomly-chosen MSC-cryopreserved vials were thawed, analyzed for their viability and expanded for 12-14 days in DMEM+10% PL until the end of second passage (end product). To test the quality of frozen end products, three diff erent bags containing expanded MSCs were thawed and tested for their viability, sterility, phenotype, proliferation/senescence, diff erentiation capacity, and their allosuppressive potential in mixed lymphocyte reaction (MLR). Results: The major message of phase I experiments was that MSCs generated from the pool of BM-MNCs of 4 donors were more allosuppressive in MLR than MSCs from each donor individually. All donors for MSC-Bank were negative for markers of TTD. From pooled BM-MNCs of eight donors we generated 320x10 6 MSCs which were frozen in 209 cryovials. To generate end products for validation of MSC-bank three randomly-chosen MSC-cryovials were thawed and expanded until the end product (P2), obtaining ~ 432x10 6 MSCs/vial (about 8.5 population doublings-PD). End products demonstrated a typical MSC-phenotype and revealed the fi rst signs of senescence starting at P4 and ceased their proliferation at P10 or P11 (68 days). All three end products diff erentiated in osteoblasts and adipocytes. Surprisingly, in MLR the allosuppressive eff ect of MSC-end product generated from the pool of 8 BM-MNCs was signifi cantly higher (P<0.04) than the mean eff ect of MSCs from 8 donors individually as well as the pool of MSCs from 8 donors. So far, we expanded 13 MSC-cryovials of MSC-bank and obtained 6,5x10 9 MSCs. Until now, 36 MSC-end products were used in 15 patients for the treatment of aGvHD with no adverse eff ects. Discussion: To our knowledge, this MSC-bank represents the fi rst validated xeno-free bank generated from the pool of 8 BM-donors. As these MSCs are from the same source, their use may ensure more consistency and reproducibility of their eff ect and therefore may be advantageous for clinical studies. Disclosure of Interest: None Declared. Introduction: Dasatinib is a dual bcr/abl/src-kinase inhibitor used for the treatment of bcr/abl+ leukemias. Originally it was designed as immunosuppressant and has been shown to inhibit eff ector T-cell activation. We recently identifi ed a previously unknown stimulatory eff ect of dasatinib on maturing human dendritic cells, leading to enhanced IL-12 production in response to TLR4-agonists (Wölfl et al., Blood 2013) . Furthermore dasatinib also had strong stimulatory eff ects on other cells of myeloid origin. Materials (or patients) and Methods: Here studied the eff ect of dasatinib on the diff erentiation of M1/M2-macrophages, using a cytokine-defi ned in vitro system. The diff erentiation in M1 and M2 macrophages is important, as these subtypes display either infl ammatory (M1) characteristics versus leukemia/lymphoma enhancing capacity in M2-macrophages. Human CD14+ monocytes were diff erentiated to macrophages, using GM-CSF (M1), M-CSF(M0), M-CSF and IL-4(M2a) or M-CSF and IL10(M2c) and evaluated phenotypically and functionally in response to an LPS-stimulus. Results: When dasatinib was present at the time of LPS-mediated activation, increased IL-12 production was observed in M1-macrophages. More importantly a shift of M2a-macrophages towards higher infl ammatory activity was achieved. However, incubation with IL-10 lead to terminal diff erentiation of macrophages with an M2c-type function, regardless of dasatinib-treatment. Discussion: Diff erentiation to the unfavorable M2-macrophage phenotype may be counteracted by treatment with src-kinase inhibitors during the diff erentiation process. However, once terminally diff erentiated, M2c-macrophages remain unaltered by dasatinib. These fi ndings lay the basis for modulating the microenvironment using src-kinase-inhibitors to direct the innate immune response towards infl ammatory, anti-tumor activity. Disclosure of Interest: None Declared. Introduction: Reduced-intensity allogeneic stem cell transplantation (RIT) can improve treatment results among patients (pts) with high-risk or advanced chronic lymphocytic leukemia (CLL). However, most of pts with CLL are (pts) older than 60 years and published data concerning RIT outcome of this group of elderly pts are limited. To evaluate the potential role of RIT in pts older than 60 years with CLL and the role of use of unrelated donor we retrospectively analysed the results of those pts transplanted in our centre during last 10 years. Materials (or patients) and Methods: From 9/2003 to 9/2013 33 consecutive pts with median of age 63 years (range: 60-70 years) with high-risk or advanced CLL (17p-/p53 mutation, chemoresistant CLL, relapse ≤ 24 moths, ≥ 3 treatment lines) underwent RIT (33% HLA identical related, 45% HLA matched unrelated, 22% HLA mismatched unrelated). Source of stem cells was peripheral blood and the median of infused CD 34+ cells was 5.1x10 6 /kg (range: 1.2-12.2x10 6 /kg). The conditioning regimen consisted of fl udarabine and melphalan (55%) or fl udarabine and cyclophosphamide (45%). Cyclosporine and methotrexate were administered as GVHD prophylaxis. Pts undergoing unrelated RIT had comparable prognostic variables as pts undergoing related RIT except of younger age of donors (P=0.0001) in unrelated RIT. Results: 91% of patients engrafted. None of 6 pts in CR before RIT progressed at day +30 after RIT and among 27 pts beyond CR before RIT 89% of them achieved at least PR at day +30 after RIT. 18 pts (54%) developed acute GVHD (3 pts grade III-IV) and among 26 evaluable pts 15 (58%) of them developed chronic GVHD (5 limited, 10 extensive). With median follow-up 43 moths (range 3-110 months) 8 pts (24%) are alive in CR. 9 pts (27%) relapsed or progressed with median time to progression 3 months (range 1-30 months) and died with median time after progression 23 months (range 1-52 months). 16 pts (49%) died due to NRM (94% infections). NRM till day +100 after RIT was 15%. The estimated probabilities of 3-years PFS and OS are 27% and 36%. There were no statistical diff erence between results of related and unrelated RIT. Discussion: In spite of relatively small number of evaluated pts and retrospective type of analysis our data show that RIT even in use of unrelated donor in pts over 60 years of age (usually heavily pretreated and with serious comorbidities) achieves in about 30% of them long lasting disease control of high-risk or advanced CLL. On the other hand, characteristics of our patient cohort suggest a later indication for RIT in elderly pts with CLL. Later indication for RIT means higher pretreatment and selection of aggressive CLL clones, which probably also leads to higher risk of mortality and relapse incidence after RIT. The role in later indication for RIT could play the fact that concerns about risks of transplant mortality in older pts led to eff orts to control the disease using standard chemotherapy, even in the presence of known adverse prognostic factors of CLL. Earlier indication for RIT could improve transplant results in elderly pts with prognostically unfavourable leukemia. Disclosure of Interest: None Declared. Introduction: Reduced-intensity allogeneic stem cell transplantation (RIT) could improve treatment results in patients (pts) older than 60 years with high-risk hematologic malignancies. Acute myeloid leukemia in elderly pts has poor prognosis and despite of higher risk of transplant mortality indication for RIT is usually performed early after diagnosis of AML. Diff erent situation may be among elderly pts with CLL, which is considered as a chronic disease, and concerns about the risks of transplant mortality can often outweigh known risk factors of CLL, which may lead to delay of indication for RIT. To evaluate the potential role of primary diagnosis on outcome of RIT in pts older than 60 years, we retrospectively analysed and compared the results of RIT in elderly pts transplanted for CLL with elderly pts undergoing RIT for AML in our centre. Materials (or patients) and Methods: From 1/2003 to 9/2013 33 consecutive pts older than 60 years (CLL group) usually with highrisk or advanced CLL and 50 consecutive pts older than 60 years (AML group) with usually high-risk AML underwent RIT. There were no diff erence in prognostic variables between both groups concerning age, PS, HCT-CI, type of donor, number of infused CD34+ cells, type of GVHD prophylaxis. There was only one diff erence in the type of conditioning regimen (CLL group -51% FLU/MEL, 49% FLU/CYl; AML group -100% FLU/MEL). Results: CLL group: with median follow-up 43 moths (range 3-110 months) 8 pts (24%) are alive in CR. 9 pts (27%) relapsed or progressed with median time to progression 3 months (range 1-30 months) and died with median time after progression 23 months (range 1-52 months). 16 pts (49%) died due to NRM. NRM till day +100 and +365 after RIT were 15% and 33%. The estimated probabilities of 3-years PFS and OS are 27% and 36%. AML group: with median follow-up 39 moths (range 5-105 months) 21 pts (42%) are alive and 20 pts (40%) are in CR. 12 pts (24%) relapsed or progressed with median time to progression 5 months (range 3-49 months) and 11 of them died with median time after relapse 2 months (range 0-10 months). 18 pts (36%) died due to NRM. NRM till day +100 and +365 after RIT were 8% and 20%. The estimated probabilities of 3-years PFS and OS are 48% and 53%. Comparison of these two groups showed a statistically signifi cantly better PFS (P=0,04) and a trend toward better OS in the group of patients transplanted for AML. Discussion: our data show better RIT outcome in elderly pts transplanted for high-risk AML in comparison with pts transplanted for high-risk or advanced CLL even though AML is generally considered to be more unfavorable disease in comparison with CLL. Potential explanation for these "surprising" results could be later Introduction: Allogeneic HSCT is the only curative treatment for CLL and is performed in patients (pts) with high-risk features at diagnosis (17p/p53 deletions) or advanced disease. Although only 25-30% of patients have matched siblings, alternative stem cell sources such as umbilical cord blood allow extending HSCT to pts lacking a conventional donor. Materials (or patients) and Methods: We analyzed 68 pts (49 males; 19 females) who underwent a single (n=16) or double (n=52) HLA mismatched umbilical cord blood transplantation (UCBT) between 2004 and 2012 in 34 EBMT centers. Median age at UCBT was 57 years (yrs) . At diagnosis, 56 pts had Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma, 8 Prolymphocytic Leukemia and 4 Richter Syndrome. Considering Binet staging, 28% had stage A, 57% B and 15% C; 54% had abnormal (27% del17p/p53, 15% del13q, 5% del11q and 7% others) and 21% normal karyotype (25% data not available). Median time from diagnosis to UCBT was 58 months . Thirty four pts were in PR at UCBT, 22 in CR and 9 had refractory disease. EBMT-score at transplant was 3-4 in 26% of pts, 5-6 in 71% and 7 in 3%. Sixty-six percent of pts received ≥3 chemotherapy regimens prior UCBT, 34% were refractory to purine analogues and 16% had previous autologous HSCT. The Minnesota RIC regimen (CyFluTBI) was given to 56 pts and ATG, at conditioning, to 15. GVHD prophylaxis was CsA+MMF in 59 pts. Median TNC collected and infused were 3.7x10 7 /Kg (1.8-7.1) and 2.7x10 7 /Kg (1-6.4) for single UCBT and 5x10 7 /Kg (2.6-9.7) and 3.9x10 7 /Kg (0.9-7.8) for double UCBT. Units were HLA matched to the recipient at 5-6 loci in 28% of pts and at ≤4 loci in 72%.

Results: Median follow-up was 30 months . OS and PFS at 3 yrs were, respectively, 55±7% and 47±7%. The cumulative incidence (CI) of neutrophil and platelet engraftment was 79±5% and 62±6% with a median time for engraftment of 21(5-67) and 43 days (6-189), respectively. CI of aGVHD at 100 days was 43%±6 for grade II-IV and 21±5% for grade III-IV with a median time of onset of 23 days . Three yrs CI of cGVHD was 33±6% with a median time of onset of 130 days (58-393). CI of relapse and TRM at 3 yrs were, respectively, 13±5% and 40±7%. In univariate analysis use of CyFluTBI regimen was associated with better OS (65% vs 15%, P<0.001), PFS (55% vs 15%, P<0.001), NE (86% vs 45%, P=0.03) and TRM (35% vs 70%, P=0.03). Age at UCBT≤57ys was associated with higher OS (71% vs 41%, P=0.02) and lower TRM (26% vs 52%, P=0.03). Less than three chemotherapy regimens prior to UCBT was associated with better OS (85% vs 48%, P=0.01), PFS (68% vs 38%, P=0.04) and lower TRM (15% vs 48%, P=0.02). Patients in CR prior UCBT had lower incidence of aGVHD grades II-IV (24% vs 49%, P=0.02). No risk factor was associated with cGVHD or relapse. Multivariate analysis was not performed due to low number of patients. Thirty pts died, 25 of transplant related causes (5 infections, 5 PTLD, 3 aGVHD, 2 cardiac and 1CNS toxicity, 1 rejection, and 8 other causes) and 5 of relapse. Discussion: UCBT is an alternative treatment option for pts with high risk or advanced CLL. Lower relapse incidence suggests that UCBT might be associated with a benefi cial GVL eff ect. The use of CyFluTBI as a conditioning regimen in this setting seems to be associated with good outcomes. Disclosure of Interest: None Declared.

A. Zavrelova 1,* , L. Miriam 1 , J. Radocha 1 , P. Belohlavkova 1 , P. Zak 1 1 

Introduction: Purpose: To evaluate allogeneic stem cell transplantation results of patients with myelofi brosis at our transplant center. Materials (or patients) and Methods: Methods: Between years 2002 and 2012 we performed 10 allogeneic stem cell transplantation for myelofi brosis at our center. Median age at transplantation was 57 years (49-62year). Median time from diagnosis to transplantation was 24 months (6-248 months). We used mainly reduced intensity conditioning with targeted busulfan dose (Bu 8mg/kg) except 1 patient (fully ablative regimen). 8 patients had unrelated donor and 2 patients had matched related donor. All patients with unrelated donor were given T cell depletion with thymoglobulin. According to dynamic international prognostic scoring system patients score was low intermideate 1, intermediate 2 and high risk in 2, 4, 2 and 2 patients, respectively. One patient had unfavorable karyotype and other patients had normal karyotype or unclassifi ed changes. Results: Results: Overall survival in patients was 90% in median observation time of 18 month (1-138 month) . Only 1 patient died early after stem cell transplantation from peracute EBV posttransplantation lymfoproliferative disease. 67% experienced aGVHD with quick response to corticosteroid therapy and 67% patients developed chGVHD, which was not debilitating them. We did not register any relapse of disease. Discussion: Conclusion: Allogeneic stem cell transplantation for myelofi brosis is feasible procedure with extremely good results and very low relapse rate. Due to these results it is possible to do this procedure early in the disease course to decrease transplant related complications and mortality. It is warranted to use T cell depletion with extremely low relapse rate in this kind of patients. Disclosure of Interest: None Declared.

Introduction: CML is a rare disease in children with an incidence 3-5% of all paediatric leukaemias. Similar to adults, stem cell transplantation is the only curative treatment approach for children with CML. Since the introduction of Tyrosine Kinase Inhibitors (TKI) haematopoietic stem cell transplantation (HSCT) is no longer the fi rst choice of treatment for patients with early phase CML. Imatinib has been approved for use in children and adolescents, and diff erent national CML treatment protocols have emerged. However life-long medication is necessary in most cases, and about 30% discontinue TKI due to acute side eff ects, growth impairment or become non-compliant when adolescent. (Millot et al. 2011) . Progression free survival after 5 years without HSCT is 65% (Champagne et al. 2011 ) For children with CML HSCT still remains an important treatment option. Materials (or patients) and Methods: The IBFM-protocol (EUDRA_ CT 2008-000569-50) for HSCT in paediatric patients with CML consists of a highly immunosuppressive reduced intensity conditioning regimen with the aim to reduce the transplant related mortality and to preserve fertility.The conditioning regimen consists of fl udarabin 4 x 40 mg/m2, thiotepa 2x 5 mg/kg, melphalan 1 x 140 mg/m2 and thymoglobuline 3 x 2,5 mg/kg. So far 14 patients in CP1 have been included after a median time of TKI treatment of 14 months. Results: Indications for HSCT were primary insuffi cient response (n=5), secondary response loss (n=3), additional rearrangement (n=1) and patient choice (n=5). Disease status prior to HSCT was bcr/abl< 0.0001 (n=2), < 0.001 (n=1), < 0.01 (n=7) and > 0.1 (n=4), one patient had 9% bcr/abl. Donors were in 50% matched siblings and matched unrelated respectively. Grafts were bone marrow in 12 and peripheral blood stem cells in 2 cases. There was no TRM, 2/14 patients experienced a-GvHD > Gr II and 1/14 patients developed extensive c-GvHD, which resolved completely. After a median follow-up of 21 months all 13/14 patients are bcr-abl negative. One patient has given birth to two healthy babies after HSCT. Discussion: Reduced intensity conditioning for children with CML with a low disease load seems to be feasible and may preserve fertility. If these fi ndings can be confi rmed in a larger patient cohort HSCT following a reduced intensity conditioning may constitute an alternative to the life-long continuation with TKI treatment in children and adolescents. Disclosure of Interest: None Declared. Hamburg, Germany Introduction: In up to 50% of myelofi brosis (MF) cases the JAK2-V617F mutation can be detected, making the JAK protein an interesting target for treatment. Ruxolitinib (Ruxo) a pan-JAK inhibitor was approved for the treatment of MF and showed effi cacy in disease treatment, irrespectively of the JAK2V617 mutation status. Beside allogeneic stem cell transplantation (allo SCT) remains the only curative treatment option for MF by using an allo-immune vs MF eff ect, to improve transplant outcome Ruxolitinib may be used in the context of allo SCT. However the impact of Ruxolitinib on the immune system, especially on T-cells, is poorly understood. Therefor we investigated the eff ects of Ruxolitinib on T-cells in vivo and in vitro. Materials (or patients) and Methods: Healthy donor T-cells were isolated by magnetic cell sorting to pan CD3+, CD4+ and CD8+ fraction. All three diff erent cell subsets were cultured with diff erent dosages of Ruxolitinib (100, 250, 500, 750nM and 1μM) for additional 48h. Thereafter cells were analysed for cell growth, cell death, mRNA expression and killing capacity against myeloid cancer cells. Additionally, immune profi les of 9 patients were analysed for the changes of the T-cell compartment as well as changes in mRNA expression during the treatment with Ruxolitinib over a period of 3 weeks. Results: T-cells with Ruxolitinib showed a delay in cell growths compared to control cells, however when analysing the CD4 and CD8 subset, mainly the eff ector CD8 cells showed a delay after 48h of culturing (2.8 x10 6 cells/ml vs. 1.9 x10 6 cells/ml, P =0.03). Surprisingly the rate of AnnexinV positive cells did not diff er in both subsets (37.8% vs. 34%). Analysing cell cycle inhibitors we found that CD8 cells express 1.5 times more p21, p27 and p53 when treated with Ruxo compared to their CD4 counter partners, arguing that CD8 cells were more likely to undergo growth arrest. We next investigated the function of T-cells against myeloid cancer cells. We found co-culturing of Ruxo treated CD3 cells with myeloid target cells resulted in a signifi cant decrease in their killing capacity (50.4% vs 37.2% P =0.001). To rule out an eff ect of Treg suppression of T eff ector cells we repeated the experiment with purifi ed CD8 cells again Ruxo treated cells showed a signifi cant decrease in their killing capacity (61% vs. 49,1%, P= 0.01). To explain the impairment in T-cell function we found that pro-infl ammatory cytokines like IL12A and IL23 to be signifi cant reduced (10.4 vs. 7.3 fold expression, P< 0.05 and 6.9 vs. 3.7 fold expression, P=0.003 respectively). To confi rm the in vitro data we analysed nine patients treated with Ruxolitinib. Strikingly to our previous fi nding we observed a decrease in total CD3 cells after three weeks of treatment (1560/ μl vs. 401/μl, P =0.01) and CD8 cells (630/μl vs. 203/μl, P=0.01). Analysing the mRNA level of cell cycle inhibitors we again found that p53 and p21 were signifi cant up regulated (P=0.005 and P= 0.03 respectively). Arguing that the growth inhibition is mainly mediated by p53 / cip cascade. Further we could confi rm that proinfl ammatory cytokines like IL7 and IL12A were signifi cant down regulated (P= 0.03) however IL23 reduction failed a signifi cant level. Discussion: We conclude that treatment with Ruxolitinib infl uences T-cell function by inhibition of cytokine expression leading to growth arrest and reduction of T-cell mediated killing. Disclosure of Interest: None Declared. 1 1 Internal Medicine V, University Hospital Heidelberg, Heidelberg, Germany Introduction: Allogeneic hematopoietic stem cell transplantation is a potentially curative therapy option for patients with high-risk CLL, defi ned by EBMT criteria (17p-/TP53mut, fl udarabine resistant, early relapse, and Richter's syndrome). Structured information about the course and outcome of patients who relapse after HSCT is limited. Materials (or patients) and Methods: In a single centre retrospective analysis, effi cacy of immunotherapeutic rescue strategies and outcome were assessed in those patients who underwent HSCT for high-risk CLL between June 2005 and April 2013 and had subsequent clinical disease progression or recurrence. Results: Altogether 77 patients with a median age of 53 (37-68) years were allografted in the 8-year period. Of these, 18 experienced clinical relapse or progression after a median time interval of 10 (3-60) months after HSCT, translating in to 4-year relapse incidence and progression-free survival of 27% and 59%, respectively. The primary rescue strategy was donor lymphocyte infusions (DLI) + rituximab (R), which was administered to 11/18 patients. (DLI could not be given to 7/18 patients because of preceding graft failure (3), early relapse of transformed CLL (3), or donor pregnancy (1)). Of the 11 patients who received DLI, 4 had been given DLI already prior to clinical relapse preemptively upon MRD persistence. The other 7 patients could not be treated pre-emptively because CLL recurrence occurred before withdrawal of systemic immunosuppression. Although all 10 patients who received DLI+R and have informative follow-up showed at least some temporary stabilization, a sustained MRD-negative complete remission (CR) was observed in only two patients (20%). Chronic GVHD subsequent to DLI developed in 3/10 patients (including one of the two long-term responders). As secondary salvage regimen, revlimid + R was given to 5 of the 8 DLI failures, resulting to CLL regression in 2 of them including one durable MRD-negative CR. With a median observation time of 20 (1-50) months, 10/18 patients are alive according to a survival probability of 51% 2 years after relapse. Discussion: Although individual patients relapsing after HSCT for poor-risk CLL can be put into durable MRD-negative remission with DLI+R, the overall effi cacy of this strategy seems to be limited. The addition of revlimid deserves further study, but the exploration of novel agents in this setting is eagerly awaited. With a median survival of 2 years, however, the prognosis of patients relapsing after HSCT does not to be worse than that of patients with high-risk CLL without prior transplant, suggesting that HSCT also in case of subsequent relapse improves the outcome by "resetting the clock" of the disease. Disclosure of Interest: None Declared. Introduction: High expression of EVI1 is a negative prognostic indicator of survival in acute myeloid leukemia (AML). More recently, negative eff ect of EVI1 overexpression in patients with chronic myeloid leukemia (CML) who received second-generation tyrosine kinase inhibitors (TKIs) after imatinib failure was detected. We report a retrospective study in which the impact of pre-transplant level of EVI1 on survival after allo-HSCT for CML was evaluated. Materials (or patients) and Methods: Overall there were 27 patients with CML in chronic phase (n=18), accelerated phase (n=6) and blast crisis (n=3) who received a related or unrelated HSCT after developing resistance to TKI. EVI1 expression was evaluated on peripheral blood before HSCT by the ratio of EVI1 transcripts to ABL transcripts on a real-time qPCR analysis. Patients were divided into two groups according to their EVI1 levels, below (EVI1-low) or above (EVI1-high) the median value. Results: Before HSCT the median EVI1 expression value was 0.38 EVI1/100ABL (range, 0.00-38.42). Although there was a trend for increased EVI1 expression in advanced phase, blast crisis and second or third chronic phase patients in comparison with those in fi rst chronic phase, this did not reach statistical significance (median, 0.86 [range, 0.00 -38.42] vs 0.15 [range, 0.03 -5.20 ], P=0.068). There was no association of EVI1 expression level with BCR-ABL expression level or BCR-ABL kinase domain mutations. After HSCT there was no statistically signifi cant diff erence between the EVI1-high and EVI1-low groups in complete molecular response rate, however, patients in EVI1-high group had signifi -cantly increased incidence of cytogenetic relapse compared with patients in EVI1-low group (64% vs. 15%, P=0.018). Gratwohl score, disease stage, age, time from CML diagnosis to HSCT, conditioning regimen, GVHD prophylaxis, donor type and EVI1 expression level were evaluated as predictors of diseasefree survival (DFS) and overall survival (OS). Only high Gratwohl score (P=0.029), advanced disease (P=0.037) and high EVI1 level (P=0.044) were associated with a shorter DFS in a univariate analysis. High EVI1 level and advanced disease stage were confi rmed to be independent predictors of DFS by multivariable Cox regression analysis. The hazard ratio (HR) for the EVI1-high group as compared with the EVI1-low group for DFS was 3.93 (95% CI: 1.17, 13.2) (P = 0.026). Gratwohl score was not included in the model because only 4 patients were in the group with low Gratwohl score (1 or 2) . Only advanced disease stage was found to be signifi cant for OS in a univariate analysis, and multivariate analysis for OS was not performed. High EVI1 expression level had no infl uence on OS in this patient cohort. Discussion: High EVI1 expression is a well-known marker of chemotherapy resistance in AML patients, associated with poor prognosis in about 10% AML cases. Our data suggest that high EVI1 level in CML patients refl ected changes in gene expression pattern during disease progression and aff ected relapse rate and disease-free survival, but not overall survival and transplantrelated mortality. Disclosure of Interest: None Declared. Introduction: Primary and secondary myelofi brosis are chronic myeloproliferative stem cell disorders curable exclusively by allogeneic HSCT. The signifi cance of splenomegaly and the question whether or not to conduct splenectomy (SE) prior to HSCT still is controversely discussed. Materials (or patients) and Methods: 95 patients (pts) (51 male, 44 female; median age at HSCT 51 years) with primary (n=60), post-PV (n=15) or post-ET (n=20) myelofi brosis underwent allo-HSCT after myeloablative conditioning containing total body irradiation (TBI) (n=45), chemotherapy regimen (n=28) or reduced intensity conditioning (n=22). Donors were HLA-identical (n=30) or mismatched (n=3) siblings and matched (n=44) or mismatched unrelated (n=18). Transplants consisted of unmanipulated peripheral blood stem cells (n=87), bone marrow (n=6) or purifi ed CD34+ cells (n=2). GvHD-prophylaxis was performed with CSA + MTX (n=51), 31 pts received anti-thymocyte-globulin (ATG) or alemtuzumab (n=13). Results: DIPSS plus scores were allocated to low (n=8), int-1 (n=19), int-2 (n=40) and high (n=20) risk groups. 24 pts had splenectomy (SE) prior to SCT, 13 allograft recipients were splenectomised subsequently after SCT (median 40 days post SCT) due to thrombocytopenia. Of the 13 pts with post-transplant SE, 7 were transplanted for primary, 4 for post-ET and 2 for post-PV myelofi brosis. 55 pts (58%) presented with splenomegaly prior to SCT. Median follow-up was 70 months among surviving pts (n=47), 1-year TRM was 27% and 5-year (5-y) cumulative relapse incidence was 12%. 5-y overall survival (OS) was calculated 49% and 5-y relapse-free survival (RFS) 50%. Primary graft failure occurred in 4 cases: 2 pts with pre-SCT SE and 2 pts with splenomegaly at SCT. 5-y RFS was signifi cantly superior in not-splenectomised stem cell recipients (56% vs. 40%, P=0,042). One year after SCT 83% of the not-splenectomised pts were alive but only 67% of pre-SCT splenectomised and 47% of post-SCT splenectomised pts. However, recipients with post-transplant SE had the highest proportion of DIPSS plus high risk pts. Post-transplant SE resulted in improved platelet counts in all cases and was not associated with perioperative mortality or morbidity. Cumulative 5-y relapse incidence was signifi cantly (P=0,039) lower in pts with splenomegaly at time of SCT (5,4%; 95% CI: 1,8-16,4%) compared to recipients with regular spleen size and splenectomised pts (20%; 95% CI: 11-37%). Distribution of risk groups according to DIPSS plus stratifi cation was comparable in pts with and w/o splenomegaly prior to SCT. Discussion: Our data point out that splenomegaly at time of SCT may not be considered as a risk factor for increased relapse rates but is on the contrary associated with reduced relapse risk and superior RFS. Introduction: To compare the curative eff ect of imatinib and allogeneic hematopoietic stem cell transplantation in the treatment of chronic myeloid leukemia. Materials (or patients) and Methods: 292 patients of chronic myeloid leukemia received imatinib and 141 patients took allogeneic hematopoietic stem cell transplantation. The clinical data of these patients were retrospectively analyzed so as to compare with the event free survival (EFS) and overall survival (OS) between these two groups with patients in the chronic phase and in advanced phase (including in the accelerate and blast phase). Results: (1) The EFS, OS, expected 5-year EFS and expected 5-year OS of the group of imatinib(278 patients in the chronic phase) were all higher than the group of allogeneic hematopoietic stem cell transplantation (120 patients in the chronic phase) (88.5% against 70.0%, 93.2% against 80.0%, 84% against 75.0% and 92% against 79.0%). The diff erences were statistically significant (P values were < 0.05). (2) The EFS of the group of imatinib(14 patients in the accelerate and blast phase) was 42.9%, and the OS was 42.9%. The EFS of the group of allogeneic hematopoietic stem cell transplantation (21 patients in the accelerate and blast phase) was 47.6%, and the OS was 57.1%. To compare the EFS and OS, there were no signifi cant diff erences between the two groups (P values were >0.05). Discussion: The EFS and OS of the group of imatinib were significantly higher than that of the group of allogeneic hematopoietic stem cell transplantation for the patients of CML in the chronic phase. Imatinib and allogeneic hematopoietic stem cell transplantation have the similar effi cacy for the patients of CML in the accelerate and blast phase. Disclosure of Interest: None Declared. Before 2003, 19 patients (group1) with sibling donor and different disease phase (6 CP1, 11AP or CP2, 2BC), received allogeneic bone marrow transplantation. The median age was 24 years (range; 8-34). The median time from diagnosis to transplant was 6 months (range; [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] .The EBMT score was mainly low of 0-2 (n=15, 78%). The median number of CMN was 2.5x 10 8 /kg (range; 1.15-3.87) . Myeloablative conditioning regimens (Bu-Cy, TBI-Cy or Bu-Cy-VP16) was used. Since 2003, 15 patients (group 2) were transplanted after a firstline therapy by tyrosine kinase inhibitors [Imatinib (n=11) and/or 2 nd TKI (n=4)]. Indications for transplantation were AP ± TF (n=7), BC (n=7) and CP1. The median time from diagnosis to transplant was 21 months (range; 5-30). The EBMT score was mainly high of 3-4 (n=11, 73%) . Only 5 patients achieved major (MMR) or complete molecular response (CMR) with TKIs before transplant. Conditioning regimen are Bu(iv)-Cy and TBI-VP16. Peripheral blood stem cells was the main source of SC (n=1) with a median number CD34+ cells of 4x10 6 /kg (range; 1.8-7.62). GVHD prophylaxis associated cyclosporine A and short course of methotrexate. Results: Group1: Mortality was mainly due to transplant related toxicity (n=9, TRM=47%) or hematologic relapse (2 patients in BC, 10.5%). Three patients with cytogenetic relapse (30%) received escalating doses of donor lymphocyte infusions (DLI) and could restore durable complete molecular remission (CMR). After a median follow-up of 42 months (range; 1-178), 8 patients are alive with CMR. The overall survival rate is of 43% at 3 years. Group2: TRM was signifi cantly lower (n=3; 20%). One patient died from relapse with BC (6%). Two patients with molecular relapse (16%) received Imatinib and could restore CMR. After a median follow-up of 38 months (range; 3-87 months), 11 patients (73%) are alive with CMR. Only 1 patient has an extensive chronic GVHD. The overall survival rate is of 74% at 3 years and signifi cantly higher than of the group1 (P=.001). Discussion: Transplantation for CML patients is associated with high mortality. First-line therapy by TKIs, even in the advanced phases and high EBMT score, signifi cantly improved results. Pre-emptive use of TKI after transplantation could restore CMR without risk of cGVHD. Disclosure of Interest: None Declared. Introduction: A recently published disease specifi c prognostic score (DIPSS) predicts survival of patient with myelofi brosis (MF) (Blood 2012 (Blood , 119:2675 . We have previously reported a transplant specifi c score (TS) based on 3 variables: spleen size, transfusions pre-transplant and donor type (BMT 2010 Mar; 45:458) , which predicts transplant outcome in MF patients. Aim of the study: The aim of the present study is to assess whether our TS can predict the outcome of patients with MF, after stratifi cation for DIPSS. In other words the question is whether the 2 scores are independent predictors of survival after an allogeneic transplant. Materials (or patients) and Methods: Patients. We have studied 96 MF patients grafted between 1992 and 2013. The median age was 57 years (23-69); median interval from diagnosis to transplant 866 days (66-3626); median circulating CD34 + cells 100/cmm (0-5280). The conditioning regimen was myeloablative (MA) in 36 patients (37%) and reduced intensity (RIC) in 60 (62%). The donor was an identical sibling (n=52), family mismatched member (n=18) or an unrelated donor (n=26). The DIPSS score was as follows: low (n=1), int1 (n=19), int2 (n=39), high (n=36). The Transplant score (TS) was as follows : low (n=47), high (n=49). Median follow up for all patients is 535 days (8-6630). S145 Results: Results. Engraftment was achieved in 90 patients (94%); six patients failed to engraft and died with pancytopenia, one despite a second transplant. Acute GvHD II-IV was recorded in 32% of the patients ; moderate/severe chronic GvHD in 19%. The overall actuarial 10 year survival is 33%; the transplant related mortality (TRM) is 22%, the relapse related death is 33%. In univariate analysis, negative predictors of survival were DIPSS (P<0.0001), TS (P<0.001), spleen size over 22 cm (P=0.004), transfusions >20 (P<0.0001), alternative donor (P=0.04). Age of patient, intensity of the conditioning regimen, interval diagnosis transplant were not predictive. In multivariate analysis DIPSS (RR 3.0 for high score (P=0.0004) and TS (RR 2.5, for high scores, P=0.005) both maintained their predictive value.

The 10 year actuarial survival for patients with DIPSS low, int1,int2, was 73% for low TS (n=36) and 23% for high TS (n=23) 3 University Hospital, Essen, 4 University Hospital, Bonn, 5 University Hospital, Göttingen, 6 University Hopsital, Hamburg, Germany Introduction: We investigate early outcome after allogeneic SCT in 22 patients -male (n = 13) and female (n = 9) -with myelofibrosis who received ruxolitinib prior to transplantation in order to reduce spleen size and constitutional symptoms. Materials (or patients) and Methods: The median age of the patients was 59 years (r: 42 -74 y) and ruxolitinib was given at doses between 2 x 5 mg (n = 5), 2 x 15 mg (n = 5), and 2 x 20 mg (n = 12) before fi rst (n = 19) or second (n = 3) fl udarabine-based reduced intensity conditioning from related (n = 2), and matched (n = 14), or mismatched (n = 6) unrelated donor. Thirteen patients had primary myelofi brosis and 9 post ET/PV myelofi brosis. Before ruxolitinib the patients were classifi ed according to DIPSS as intermediate-1 (n = 3), intermediate-2 (n = 14), or high risk (n = 5). Stem cell source was PBSC (n = 21) or bone marrow (n = 1) with a median CD34+ cell count of 7.1 x 10 6 /kg. Before ruxolitinib 21 patients (96%) had constitutional symptoms and all patients had splenomegaly. The median time from start of ruxolitinib to allogeneic SCT was 133 days (r: 27 -324) and the median treatment duration was 97 days (r: 20 -316). Most patients (n = 82%) received ruxolitinib until start of conditioning therapy. Four patients (18%) discontinued ruxolitinib between 28 and 167 days before transplantation due to progressive disease or no response (n = 3) or cytopenia (n = 1). Results: At time of transplantation 86% had improvement of constitutional symptoms and 41% had major response (>50% palpable) of spleen size, 14% had response of spleen size which was less than 50%, and 45% had no response/ lost response or progressive spleen size after ruxolitinib treatment. After discontinuation of ruxolitinib at fi rst day of conditioning regimen no "rebound" phenomenon was seen. One patient transformed to sAML before transplantation despite response of spleen size and constitutional symptoms. After busulfan (n = 16), treosulfan (n = 3), or melphalan (n = 3) dose reduced conditioning no graft failure was observed and the median time for leukocyte and platelet engraftment was 15 days (r: 10 -66) and 17 days (r: 8 -122) respectively. Acute GvHD I-IV was seen in 38% of the patients which was severe (III/IV) in 27%.

During follow-up 4 patients died, 1 patient with sAML at time of transplant due to relapse on day 102 and 3 patients due to therapy-related mortality. One female patient who received a second unrelated HLA-matched transplantation after treosulfanbased regimen died of CMV pneumonitis on day 75. She did not response to ruxolitinib regarding spleen size and constitutional symptoms. A second patient with iron overload and liver fi brosis died of liver toxicity on day 47. This patient initially responded to ruxolitinib but progressed regarding spleen size prior to transplantation. One patient who responded to ruxolitinib regarding constitutional symptoms and spleen size (< 50%) died of GvHD on day 77. The estimated 1-year OS and PFS was 81% (95% CI:72-90%) and 76% (95% CI:67-85%) respectively . Discussion: Ruxolitinib reduces spleen size and constitutional symptoms in the majority of patients before allogeneic stem cell transplantation. Ruxolitinib did not negatively impact engraftment after transplantation. More patients and a longer follow-up will be presented at the meeting. Disclosure of Interest: None Declared. Introduction: Studies on the long-term outcome of patients with chronic myeloid leukemia (CML) who have undergone an allogeneic hematopoietic stem cell transplantation (HSCT) have shown very rare hematological relapses (HR) after 10 years. However, long-term molecular assessment of the disease has never been closely analyzed. In our experience, late HR was observed in 4% of patients at a median of 13 years after HSCT (range 11-15) 1 . We were not able to detect an earlier molecular disease recurrence because no long-term monitoring of the BCR-ABL1 fusion gene was performed. To evaluate whether patients with a long followup (more than 10 years) after HSCT, apparently free of leukemia, may have residual leukemic cells we carried out a molecular monitoring. Materials (or patients) and Methods: Since January 2010 a quantitative real-time polymerase chain reaction (RQ-PCR) analysis was performed, to evaluate the presence of the BCR-ABL1 fusion gene, in all 65 CML patients surviving in continuous complete hematological remission (CCR) at least 10 years after HSCT at our Center. All molecular data are expressed using the International System. Overall, the analysis was performed in 46 patients; 7 refused to participate to the study, 7 were not traceable and 5 were followed at other hospitals. Results: Of the 46 patients analyzed, at a median follow-up of 220 months from transplant (range133-363), RQ-PCR resulted positive in 9 (19.5%). A patient with a BCR-ABL1 ratio of 62 IS , at 23 years after HSCT, presented a previously undetected hematological relapse 1 . A patient with a BCR-ABL1 ratio of 33 IS showed a concomitant cytogenetic relapse. The other 7 patients, with a median number of 0.2 IS (range 0.006 IS -0.7 IS ), with a normal karyotype and normal blood cell counts, were monitored by RQ-PCR every 3 months: the disease evolved into a HR in 1 patient 5 months later, persisted at the molecular level in 3 and became undetectable in 3. Discussion: Our study shows that BCR-ABL1 + cells identifi able by RQ-PCR may be detected in CML patients long after a HSCT. However, the prognostic signifi cance of residual BCR-ABL + cells at low molecular levels and the defi nition of "low molecular level" after HSCT remains poorly defi ned; we in fact observed the disappearance of the rearrangement in some patients, as well as the persistence or the evolution into an overt HR in others. The uncertain clinical signifi cance of these molecular fi ndings raises ethical and clinical concerns with regard to disclosing adequate information Introduction: In the current management of CML, SCT is reserved for patients who have failed tyrosine kinase inhibitors (TKIs) and advanced phase (AP) disease. Accelerated phase patients are a heterogeneous group in whom SCT is not mandatory as many patients have durable responses to TKI therapy alone. However, since survival clearly correlates with disease phase, TKIs and/or chemotherapy are employed to induce a second chronic phase (CP2) in AP patients in order to maximise the success of allo-SCT. Here we analyse the impact of return to CP on the outcome of CML patients transplanted in the era of TKI therapy. Materials (or patients) and Methods: We have analyzed 5732 CML patients who underwent allo-SCT reported to the EBMT registry between 2000 and 2011. Of these, 1247 had been treated with a TKI prior to SCT. Imatinib alone had been received by 777, in 418 imatinib had been followed by a second-generation TKI (2G-TKI) and 52 had received a 2G-TKI alone pre SCT. The TKI study population had the following characteristics: male n=772 (61.9%), female n=475 (38.1%), median age 43.9 years (range 18 -75). At transplant 635 (50.9%) were in fi rst chronic phase (CP1), 314 were > CP1 (25.2%), 168 (13.5%) were in accelerated phase and 130 (10.4%) were in blast crisis. Post TKI, all patients (n=1247) received allo-SCT of which 475 patients (38%) were from HLA fully matched siblings. Results: The best outcome for allo-SCT remains in patients transplanted in CP1, irrespective of TKI therapy, and long-term OS can also be achieved in patients with AP at time of SCT. Despite TKI therapy, a trend towards a diff erence in 5-yr OS was only documented in patients with CP2 (P=0.06, Table) . However, in comparing all patients in CP>1 (CP2 and CP3) there was a benefi t in 5-year OS for patients who had received TKI pre-SCT (49.5% vs. 41.4%, P=0.024). Discussion: Since the introduction of TKI therapy, there has been little change in the outcome of patients according to disease phase, with a trend towards a diff erence in improved survival only noted in CP2 patients who have received TKI pre-SCT. Therapy with TKI to CP>1 is associated with a signifi cantly improved outcome post SCT, but disappointingly, advanced phase patients continue to have a poor prognosis. Disclosure of Interest: None Declared. 1 University Hospital, Hamburg, 2 University Hospital, Frankfurt, 3 University Hospital, Cologne, 4 Medical School, Hannover, Germany Introduction: Bone marrow fi brosis is a hallmark of primary or post ET/PV myelofi brosis. Recent studies have shown that allogeneic stem cell transplantation induces rapid bone marrow fi brosis regression suggesting that fi brosis is more a dynamic rather than a static process. So far, no data are reported if fi brosis regression infl uences outcome after transplantation. Materials (or patients) and Methods: We studied 94 patients with myelofi brosis who received allogeneic stem cell transplantation at the University Medical Center Hamburg between 2002 and 2010. In 57 patients the complete bone marrow histology prior transplantation and/or on day+30 and +100 after transplantation were available. The median age was 57 years (r: 33-73) and patients were classifi ed as IPSS low-risk (n = 1), intermediate-1 (n = 5), intermediate-2 (n = 18), and high-risk (n = 33). Donor source was unrelated (n = 46) or related (n = 11) and of these 38 were HLAmatched while 19 were HLA-mismatched. 41 had primary and 16 post ET or post PV myelofi brosis. The median number of blasts was 1% (r: 0-17%) and gender was male (n = 32) and female (n = 25). All patients received busulfan-based dose-reduced conditioning regimen. Bone marrow fi brosis was graded according to the European consensus and WHO (MF-0 -3), respectively, and evaluated by experienced hematopathologists. Results: Before transplantation 41 patients (72%) had MF grade 3 and 16 (28%) MF grade 2. After engraftment on day+30 (± 10 days) (n = 48), 3 (6%) had complete regression (MF-0) and 7 (15%) near complete regression of bone marrow fi brosis while 17 (35%) had MF-2 and 21 (44%) had MF-3. On day+100 (± 20 days) complete regression (MF-0) was noted in 11 (25%) and near complete regression (MF-1) in 13 (29%) while 12 (27%) and 8 (18%) had still MF-2 or MF-3, respectively. Patients with complete and near complete regression (MF-0 and MF-1) on day +30 had a 5-year estimated overall survival of 100% and those with MF-2 and MF-3 of 75% (P = 0.09). Patients with complete or near complete regression on day+100 had a 5-year estimated survival of 95% in contrast to 60% for those with MF-2 or MF-3 (P = 0.04). If analyzed by regression by grade at day+100 7 (16%) had a reduction of 3 grades (e.g. MF-3 to MF-0), 12 (27%) of 2 grades, 16 (36%) of 1 grade, and 9 (21%) had no grade reduction at day+100. Reduction of 2 or 3 grades at day+100 resulted in 95% survival at 5 years while reduction of 1 or no grade had a survival of 70% however this diff erence did not reach statistical signifi cance (P = 0.1). There was no diff erence of fi brosis regression at day+100 between high risk IPSS and low/intermediate risk patients. Furthermore, in those patients with JAK2V617F mutation and complete or near complete regression 42% still had detectable JAK2V617F mutation level in peripheral blood. In contrast, 81% had complete donor cell chimerism if bone marrow fi brosis was MF-0/MF-1 while only 31% had complete chimerism at day+100 if fi brosis was classifi ed as MF-2 or MF-3. Discussion: This data on fi brosis regression after allogeneic stem cell transplantation suggests that a more rapid regression -independently of IPSS risk score -resulted in a favorable survival and may be used as an early predictive factor for excellent survival. Disclosure of Interest: None Declared. Introduction: Blastic plasmacytoid dendritic cell (BPDC) neoplasm is a rare and clinically aggressive tumor with poor prognosis.

Though majority of the patients initially respond to combined chemotherapy, relapses with subsequent drug resistance occur in virtually all patients, resulting in a median overall survival of only 1 year. There's no standardized therapeutic approach for BPDCN, but durable responses have only been reported in patients who underwent allogeneic hematopoietic stem cell transplantation (HSCT). Although a HLA-identical sibling donor is preferred, such a donor is unavailable for many patients, especially in China. Haploidentical donor has emerged as a successful alternative source of stem cells. Here we fi rst reported four BPDCN patients allografted as consolidative treatment with an haploidentical donor. Materials (or patients) and Methods: All four patients were diagnosed according to 2008 WHO classifi cation, without a HLA-matched donor. They all received modifi ed Bu/Cy/ATG as conditioning, including of cytorabine 4g/m 2 /day × 2 days (on day -10 and -9); busulfan 3.2 mg/kg/day intravenously × 3 days (on day -8 to -6); cyclophosphamide 1.8 g/m 2 /day × 2 days (on day -5 and -4) and semustine 250 mg/m 2 on day -3, along with rabbit antithymocyte globulin (ATG; thymoglobulin; Sanofi ), 2.5 mg/kg/d × 4 days (on day -5 to -2 days). The grafts were G-CSF mobilized bone marrow combined with peripheral blood stem cells. All patients received cyclosporine A, mycophenolate mofetil, and short-term methotrexate as prophylaxis for graft-versus-host disease (GVHD). Results: The median time of neutrophil recovery (>0.5×10 9 /L) and platelet recovery (>20×10 9 /L) were 12 and 14.5 days. All patients achieved full donor engraftment and attained complete remission at day 30. Acute graft versus host disease (GvHD) was happened in 3 of 4 patients, 2 in grade II and 1 in grade III. Chronic GvHD was observed in 3 of 4 patients, including 2 extensive GvHD. Three of 4 patients had CMV antigenemia at day 27,17 and 28, respectively, but there was no CMV disease diagnosed. No treatment related mortality was happened. All patients were still in ongoing CR with disease-free survivals of 570, 430, 217 and 215 days post transplantation respectively (details in table) . Discussion: The present study was, to the best of our knowledge, the fi rst report to successfully use haploidentical donor in BPDCN patients and modifi ed Bu/Cy/ATG was an eff ective and safe conditioning. Haploidentical HSCT was an attractive alternative treatment in BPDCN cases and may lead to long-term remission. Disclosure of Interest: None Declared. Discussion: These data suggest that obtaining T-lymphocytes clinically relevant doses for therapeutic use from a small sam-ple of thawed CB is feasible respecting the GMP conditions in all Ex vivo steps. T-lymphocytes expansion is obtained with a combination of IL-2, IL-7 and magnetic beads coupled to CD2/CD3/ CD28 mAbs that significantly enhance proliferation and activation of thawed CB T cells. The function of thawed expanded T cells must be assessed and compared to the function of thawed T cells obtained from peripheral blood by cytapheresis and used in routine for DLI. These results will show whether thawed expanded cells can be used as a therapeutic tool after HSCT (prevention/treatment of leukemia relapse, viral priming to treat viral infection) or whether they need to received additional Ex vivo treatments. Disclosure of Interest: None Declared.

Introduction: In haploidentical-SCT (stem-cell-transplantation) male-patients with female-donors have better prognosis compared to female-to-male-combinations due to Y-encoded minorhistocompatibility-antigens recognized by female-allo-immune eff ector-lymphocytes in the context of a graft-versus-leukemia-(GvL)-eff ect. We provide data in a dog-model that the minor-histocompatibility-antigen UTY might be a promising target to further improve GvL-immune-reactions after allogeneic-SCT. Materials (or patients) and Methods: Canine (c) purebred-beagledogs' PB and BM were studied. T2-cells (HLA-A2+, TAP-defi cient) were used. These human-(h)-UTY-sequence-derived HLA-A2binding-peptides were investigated: W248 (WMHHNMDLV), T368 (TLAARIKFL), K1234 (KLFEMIKYC). In vitro: Autologous-cDCs were generated with best of three DC-methods (Calcium-Ionophore, Picibanil, Cytokines). Generation cUTY-specifi c-CTLs: CD3+T-cells were co-cultured with autologous-mature cDCs+hUTY-peptides (weekly restimulation for 21 days; +hIL-2, +hIL-7). Cytotoxicity and antigen-specifi city were determined by [ 51 Cr]-release-and cIFN-g-ELISPOT-assays. Cells were quantifi ed day 0 and of harvest using anti-cmAbs/hmAbs (FACS), UTY-mRNA-expression via RT-PCRanalysis. In vivo: A female-dog was immunized with PBMCs from a DLA-identical-male-dog (day 0 and 14). PB-derived T-cells were harvested 35 days post 2nd-injection followed by analysing UTYspecifi c-reactivity. Results: Female cUTY-specifi c-CTLs were stimulated in vitro using autologous-DCs loaded with three HLA-A2-restricted UTY-derivedpeptides (≤2.9-fold-expansion) and specifi c T-cell-responses were determined in 3/6 female-dogs. CTLs specifi cally recognized/ lysed autologous-female peptide-loaded-DCs (900 spots/100,000 T-cells (median)/≤47.9%), but not naive autologous-female-DCs and -monocytes (p≤0.026). They mainly recognized BM and to a lower extent DCs, monocytes, PBMCs and B-cells from DLA-identical-male-littermates and peptide-loaded T2-cells in an MHC-Irestricted manner (up to p≤ 0.046). UTY-mRNA was only expressed in male-cells. A UTY-/male-specifi c-reactivity was also obtained in vivo after stimulation of a female-dog with DLA-identical-male-PBMCs. Discussion: We demonstrated natural UTY-processing/presentation in dogs. Female-dog-CTLs were specifi cally stimulated by HLA-A2restricted-UTY-peptides, thereby enabling recognition of DLAidentical-male-cells, mainly BM-cells. These observations suggest UTY as a promising candidate-antigen to improve GvL-reactions in the course of immunotherapy. Next-generation-sequencing and specialised-bioinformatics-algorithms are now focus for human-individualised-leukemia-treatment (T-cell-receptor-Introduction: Problem: Allogenic SCT/DLI are promising T-cell based therapies to cure AML-pts. Antileukemic T-cell-reactivity has to be improved/re-established in pts in vivo. Ex vivo leukemia-derived DC (DC leu ) are the most eff ective antileukemic T-cellstimulators. Materials (or patients) and Methods: Aim and Methods: We generated DC leu ex vivo from AML blasts from heparinised whole blood ('WB-DC' , to simulate the in vivo situation) from 65 AML-pts in active stages of the disease using standard methods ('Picibanil' , 'MCM-Mimic' , 'Ca-ionophore' , 'IFNa') or 11 minimalized cocktails ('WB-minicock-DC' , combinations of 1-3 selected cytokines, antibiotics, bacterial lysates, or other clinically approved responsemodifi ers) and to correlate proportions of DC-or T cellsubsets and cytokine profi les with results with their ex vivo stimulatory capacity for antileukemic T-cells and the pts' response to immunotherapy (SCT/DLI). Results: 1. Generation of DC: we could identify 4 of 11 minicocks, that allowed the generation of DC/DC leu from blastcontaining WB-samples with at least one of the methods. Some of the cocktails induced ex vivo blast-proliferation in individual pts. Proportions of DC-subtypes (e.g DC/DC leu /mature DC) were comparable to proportions generated with standard DC methods. 2. Antileukemic functionality: In 21 cases T-cells stimulated with 1 to 3 'WB-minicock-DC' resulted in 56% cases with blastlysis; in 6 pts 2-3 cocktails could be studied in parallel and in at least one of the cocktails a blastlysis could be achieved. Blast lysis (vs non-lysis) correlated with higher proportions of DC-subtypes: (DC,DC leu blastconversion to DC leu ), higher proportions of T-cell-subtypes (viable, CD8 Tcells), higher concentrations of IL-12 and IFNg but lower concentrations of IL-6 and IL-8 3. Clinical correlation: AML-pts successfully responding to immunotherapy (SCT or DLI therapy) presented with higher proportions of DC, DC leu and CCR7 + mature DC compared to pts without successful immunotherapy. Discussion: DC/DC leu can be generated regularly from MNC or WB and with at least 1 to 4 of 11 minicocks containing combinations of 1-3 selected, clinically approved responsemodifi ers. T-cells stimulated with 'WB-minicock-DC' achieved antileukemic function, although not with every cocktail. A patient-individual testing of the best cocktail as well as the achieved antileukemic (ex vivo) function can contribute to defi ne cocktails of response modifi ers to be applicated to AML pts to achieve or sustain remission. Disclosure of Interest: None Declared. Introduction: Hurler syndrome (MPS-IH) is a rare autosomal recessive lysosomal storage disease caused by mutations in the IDUA gene, resulting in the defi ciency of alpha-L-iduronidase (IDUA) enzyme activity with a consequent intracellular accumulation of glycosaminoglycans (GAGs). Among a broad spectrum of clinical manifestations, MPS-IH is characterized by a range of skeletal abnormalities known as dysostosis multiplex. The introduction of hematopoietic stem cell transplantation (HSCT) has significantly improved the life-span of Hurler patients. However, the musculoskeletal manifestations are only partially responsive to HSCT. Since MPS-IH is a progressive storage disorder that aff ects the skeleton during childhood, it may be necessary to undertake transplantation in MPS-IH infants (identifi ed by prenatal or neonatal diagnosis). Therefore, in the present study, we have fi rstly analyzed the skeletal phenotype of MPS-I mice and then investigated whether HSCT in newborn mice prevents the subsequent accumulation of storage material in the lysosomes and the derived bone damage. Materials (or patients) and Methods: 1-to 2-day-old mutant mice were conditioned using a single administration of 20 mg/kg busulfan and then injected via the superfi cial temporal vein with 2 x 10 6 bone marrow cells from wild type donors. Bones of transplanted mice were examined radiographically and microscopically at 37 weeks of age. Age-matched WT and untreated mutant mice were used as controls.

Results: Untreated MPS-I mice at 37 weeks of age develop a fl attened facial profi le and thickening of the digits and palmar regions. Radiographs reveal thickness of the zygomatic arches and long bones. Femurs from MPS-I mice were wider at 1.36 ± 0.10-fold normal (n = 8; P < 0.01) and more sclerotic compared to WT siblings. Similar features were observed for the tibia (1.23 ± 0.10-fold normal; n = 8; P < 0.01), humerus (1.46 ± 0.08-fold normal; n = 8; P < 0.01) and radius/ulna (1.19 ± 0.05-fold normal; n = 8; P < 0.01). Furthermore, pathological evaluation of cross sections demonstrated that femurs of MPS-I mice have abnormal osteocytes and a thick cortex. Transplantation of normal murine bone marrow cells into preconditioned MPS-I neonates led to a high engraftment level at 37 weeks after HSCT (peripheral blood, 58.9 ± 40.3%; n = 14) and a long-term multilineage donor hematopoiesis recovery. Similar engraftment levels were obtained in transplanted WT neonates (62.1 ± 33.7% n = 8; P > 0.05), as expected. Neonatal HSCT allowed restoration of IDUA activity in peripheral organs. Indeed, IDUA activity levels detected in heart, lungs, spleen, kidneys and liver were 18%, 22%, 70%, 35% and 57%, respectively, of those detected in unaff ected animals. Radiographic analysis has shown that the width abnormality of femurs was signifi cantly normalized (1.14 ± 0.08; n = 6). This was confi rmed also for the tibia (1.05 ± 0.13-fold normal; n = 6), humerus (1.08 ± 0.10-fold normal; n = 6) and radius/ulna (0.99 ± 0.08-fold normal; n = 6). Preliminary histological and microCT analyses of long bones seem to confi rm the correction of the bone phenotype in transplanted MPS-I mice.

Discussion: We conclude that neonatal HSCT would be an attractive therapeutic option to prevent severe bone dysplasia in MPS-IH. Disclosure of Interest: None Declared.

Introduction: Background: Cellular traffi cking (e.g. invasion through matrix barriers and penetrations of blood vessel walls) in healthy organisms is highly regulated and restricted. Matrix metalloproteinases (MMPs) mediate the proteolytic breakdown of these barriers. MMP-9 degrades collagen type IV and denatured collagens, and processes TGFß, IL-8 and IL-1ß into their active forms. A surface associated 82-kDa variant of MMP-9 was found to be expressed in leukemic cells (Ries et al, Biochem J, 2007) . The role of MMP-9 variant expression on leukemic cells is still unclear. Materials (or patients) and Methods: Aim and Methods: We generated a monoclonal antibody that specifi cally recognizes the aberrant MMP-9 form. By use of this antibody we analysed the expression of MMP-9 variant on blasts from patients (pts) with AML (n=22), ALL (n=20) and CLL (n=21) compared to healthy controls by FACS analyses and correlated our fi ndings with prognostically relevant subgroups. Results: 1. Results in AML: we could identify a variable coexpression of MMP-9 on blasts in diff erent FABtypes, however a significantly higher expression on pooled 'favorable' (30% coexpression) compared to unfavourable FABsubtypes (20%, P<0.05) and favourable (40%) vs unfavourable NCCN risk group (8%, n<0.005). Expression was also higher in pts studied at fi rst diagnosis (19%) vs pts at relapse/persisting disease (10%, P<0.05), in pts <60 years (25%) vs pts>60 years (16%, P=ns) and in responders to initiated induction therapy (21%) vs nonresponders (8%, P=ns). A cut-off value of coexpression (11%) could be evaluated that allowed a separation of pts in those with a more favourable (>11%)/unfavorable prognosis (<11%, NCCN). 2. Results in ALL: we could identify a variable coexpression of MMP-9 on blasts in diff erent EGILsubtypes, however a lower expression on pooled 'favorable' (8% coexpression) compared to unfavourable GMALLsubtypes (24%, P=ns) risk types. Signifi cantly lower coexpression was found on blasts from female (13%) vs male pts (28%, P<0.05) and on pts with primary (9%) vs secondary ALL (28%, P<0.05). Expression was also lower in pts studied at fi rst diagnosis (21%) vs pts at relapse/persisting disease (42%, P=ns). A cut-off value of coexpression (9%) could be evaluated that allowed a separation of pts in those with a more favourable (<9%)/unfavorable prognosis (>9%, GMALL). 3. Results in CLL: we could identify a low coexpression of MMP9 on CLL-blasts (10-12%) independent of diff erent subtypes and risktypes (data not shown). A predictive cut-off value could not be evaluated. Discussion: These fi ndings demonstrate opposing roles of MMP-9 variant in AML and ALL cells. It can be speculated, that in AML MMP-9 variant-mediated processing of regulatory factors on the surface of blast cells contributes to a favourable outcome, whereas in ALL degradation of matrix barriers by MMP-9 variant may increase the invasive capabilities of ALL blast cells and thus explain a rather poor prognosis. Our data suggest usefulness of MMP-9 variant as a prognostic marker in certain AML and ALL subgroups. Disclosure of Interest: None Declared. Introduction: The outcome of allogeneic stem cell transplantation (Allo-SCT) is closely related to graft versus host disease (GvHD) and graft versus leukemia (GvL) eff ects which, in part, are mediated by minor Histocompatibility Antigens (mHAgs). Twenty-six mHAgs have been identifi ed and reported to be diff erently and variably correlated with GVHD or GVL. These antigens are: UGT2B17 (correlated with GvHD), ACC-1, ACC-2, ACC-6, C19orf48, HB-1, LB-ADIR-1, LB-LY75-1K, LB-MR1-1R, LB-MTHFD1-1Q, LB-PTK2B-1T, LRH1 (correlated with GvL), HA-1, HA-2, HA-8, CD31 (correlated with both GvHD and GvL) and ACC-4, ACC-5, CTL7A7, DPH1, DRN7, HA-3, HEATR-1, P2RX7, LB-ECGF-1H UTA2-1 (with not well determined clinical signifi cance). Nowadays a simultaneous method to genotype a so large panel of mHAgs has never been employed. The aim of this work has been to develop a feasible method to genotype all the 26 mHAgs described so far and to test them for their correlation with GVHD and GVL in a group of donor/recipient pairs submitted to allo-SCT. Materials (or patients) and Methods: For the multi-genotype of 23 mHAgs we designed 3 multiplex and samples analyzed by mass-spectrometry through Maldi-Tof IPlex Gold technology. This assay is relatively fast and it requires a small amount of DNA. For the other three mHAgs we performed other three assays: two based on capillary sequencing of PCR products (for LB-MR1-1R and LRH1) and the last one based on PCR alone (for UGT2B17).

By these methods, we tested the 26 mHAgs in 70 donor/recipient pairs at least 6/6 matched at 4-digit high resolution typing, that underwent allo-SCT (sibling or MUD) because of Philadelphia positive CML (n=46) or ALL (n=24). Results: Maldi-Tof IPlex Gold technology proved a high degree of effi ciency. Out of a total of 3220 SNPs an evaluable genotype was obtained in 3176 (98.6%). Also the other assays were effi cient (417/420, 99.3%). As expected, sibling pairs showed most identity of MUD pairs. Notably, donor/recipient mismatch on ACC-5, UGT2B17, DPH1 and LRH1 can drive a pathogenetic mechanism that leads to GvHD increase (P<0.05). Next we identifi ed that LB-ADIR1 can improve RFS (P<0.05) as GvL eff ect. This is potentially important (P=ns, for the low patients number) especially for ALL-Ph+ patients because this mismatch can enhance GvL in a subgroup that is otherwise less or un-responsible to allo-immunotherapy. Discussion: Our data generated by a multi-genotype technique confi rm the role of mHAgs in addressing graft reactions; in some cases only GvL without GvHD. This suggest that a study of mHAgs (particularly ACC-5, UGT2B17, DPH1, LRH1 and LB-ADIR1) could be performed at the donor recruitment time in order to better and prospectively investigate the role of the known and new mHAgs involved in GvHD and GvL eff ects. Introduction: Life-threatening GVHD is still a major complication of allogeneic haematopoietic stem cell transplantation (allo-SCT). αβT-cells are considered to be main contributors in GVHD. Depletion of αβT-cells have so far been reported in haploidentical stem cell transplantations. First results have been promising with good engraftment and low incidence of GVHD. In this study we aimed to assess the feasibility of αβT-cell depletion in the context of MUD and MRD. Main purpose is to develop transplantation protocols which result in a more profound expansion of both NK and γδ-T cells. These cellular subsets, which are at the border of the innate and adoptive immune system, are considered to contribute to the graft-versus-leukemia (GvL) eff ect as well as to control viral reactivations after allo-SCT. In this way, we hope to decrease GvHD whilst preserving GvL eff ects in allo-SCT in patients with poor risk leukemia. Materials (or patients) and Methods: Initial proof runs for the generation of αβT-cell and CD19 depleted grafts have been performed in 4 healthy donors. In the fi rst 5 patients (cohort I) grafts for transplantation have been depleted with GMP-grade anti-αβTCR and anti-CD19 antibodies. The subsequent grafts for patients in cohort II (n=4) and III (n=3) have been selectively depleted with GMP-grade anti-αβTCRs antibodies only. Three conditioning regimens have been investigated (I): fl udarabine 120 mg/m 2 + cyclophosfamide 4800 mg/m 2 , (II): fl udarabine 120 mg/m 2 + busilvex AUC=90 and (III): ATG (Genzyme®) 4 mg/m 2 + fl udarabine 120 mg/ m 2 + busilvex AUC=90 followed by abT-cell depleted grafts from matched related or unrelated donors. No additional immune suppression was given after allo-SCT. Results: Products for 12 patients have been successfully processed and used for αβT-cell depleted allo-SCT between 2011 and 2013. A ~4 log depletion of αβT-cells has been observed in the product with a recovery of ~75% of CD34+ cells. In cohort I, primary engraftment (chimerism > 95%) was 40%. Engrafted patients showed a rapid reconstitution of γδT-cells and αβT-cells with a broad αβT-cell repertoire as determined by spectratyping. Due to the rapid reconstitution of αβT-cells in engrafted patients, CD19 depletion was omitted in further cohorts. To further improve engraftment, the condition regimen of the next cohort was dose-intensifi ed (cohort II). 75% of cohort II showed a swift engraftment. Again a dominance of γδT-cells was observed which associated with a rapidly reconstituting αβT-cell repertoire. In order to further increase engraftment cohort III was additionally treated with an early application of ATG (day -10/-9) and no graft failure has been observed so far. In the CD19-depleted group, two EBV reactivations occurred. No other viral re-activations have been observed. Discussion: αβT-cell depletion is feasible in the context of MRD and MUD. An intensifi ed conditioning with additional host T-cell depletion seems to be benefi cial for a profound engraftment. abTcell depletion associates with a swift and dominant reconstitution of γδT-cells as well as a rapidly restoring αβT-cell repertoire that displays a broad reactivity. Cohort III is currently expanded to confi rm whether this regimen results in solid engraftment. Subsequently we aim to use αβT-cell depleted allo-SCT as a platform for post-transplant immune interventions. Disclosure of Interest: None Declared.

EXTRACELLULAR HMGB1 PROMOTES THE MIGRATION OF CORD BLOOD CD34+ CELLS VIA SDF-1/CXCR4 AXIS X. Zhu 1,* , L. Yang 1 , Z. Sun 1 , X. Wang 1 , X. Liu 1 1 Hematology, Anhui Provincial Hospital, Hefei, China Introduction: This study was aimed to investigate the eff ect of HMGB1 (high mobility group protein) and/or CXCL12(stromal cell derived factor) on the migration of cord blood CD34 + cells, and to explore whether HMGB1 promotes cord blood CD34+ cells migration via CXCL12/CXCR4 axis. Materials (or patients) and Methods: Cord blood mononuclear cells were isolated by Ficoll-Paque density centrifugation. CD34+ cells were collected by a positive immunoselection procedure(CD34 MicroBeads) according to the manufacturer's instructions. The purity of the isolated CD34+ cells was detected by fl ow cytometry.

In vitro chemotaxis assays were performed using Transwell cell chambers to detect cells migration. 1×10 5 /well cord blood CD34+ cells were added into the upper chambers. Diff erent concentrations of HMGB1 and/or CXCL12 (0,10,25, 50,100,200 ng/ml) were used respectively to detect the optimal concentrations of HMGB1 and/or CXCL12 for inducing migration of cord blood CD34+ cells.

Freshly isolated cord blood CD34+ cells express CXCR4(CXCL12 receptor), and HMGB1 receptors TLR2,TLR4 and RAGE. To explore which receptors were required for the synergy of HGMB1 and CXCL12 on cells migration, we used anti-CXCL12, anti-CXCR4, anti-RAGE, anti-TLR2 and anti-TLR4 antibodies to detect the eff ect of HGMB1 alone or with CXCL12 on cord blood CD34+ cells migration.

Results: The purity of CD34+ cells isolated from cord blood mononuclear cells by magnetic cell sorting was 97.4%. 25 ng/ ml CXCL12 did not induce migration of cord blood CD34+ cells, whereas optimal migration was observed at 100 ng/ml. HMGB1 alone did not induce migration up to 50 ng/ml. We determined, by dose fi nding experiments, that the the best synergistic concentrations for cells migration are 100 ng/ml HMGB1 combined with 50 ng/ml CXCL12. The blocking experiments showed that both the anti-CXCL12 (4μg/ml) and anti-CXCR4 (5μg/ml) antibodies could block cell migration induced by HMGB1 alone or combined with CXCL12. But cord blood CD34+ cells in the presence of anti-RAGE,anti-TLR2 and anti-TLR4 antibodies, did not modify the response to CXCL12 in the presence of HMGB1. Discussion: Both HMGB1 and CXCL12 can induce cord blood CD34+ cells migration. HMGB1 enhances CXCL12-induced migration exclusively via CXCR4 and in a RAGE-and TLR-independent manner. The exact mechanism needs to be further explored. Disclosure of Interest: None Declared. Introduction: Umbilical cord blood transplantation (UCBT) is increasingly used and gives comparable results to transplantation with other stem cell sources. Donor lymphocyte infusion (DLI) is an eff ective treatment option for relapsed malignancies after hematopoietic stem cell transplantation with other stem cell sources, but is not available to UCBT recipients. In vitro cultured cord blood DLI present an attractive putative option for UCBT recipients. Materials (or patients) and Methods: In this study, we explored the use of in vitro cultured T-cells from the cord blood graft as an alternative to conventional DLI for UCBT recipients. T-cells from 5% of a cord blood unit used for UCBT were cultured for 7-11 days with either 600 IU IL-2/ml or 100 IU IL-2/ml and 20 ng IL-7/ml. The cultured cord blood DLI (CB DLI) was given to four patients as treatment for mixed chimerism with threatening rejection (n = 2), minimal residual disease (MRD, n = 1), and for a patient with graft failure with severe infection. The patients had received a UCBT due to AML, ALL or PNH/MDS. The patient receiving treatment for MRD was given T-cells cultured with IL-2 and IL-7, while the remaining three were given cells cultured with IL-2. To our knowledge, in vitro cultured and activated UCB T-cells produced in this manner have not been used for clinical DLI treatment previously.

In this study, we address the safety and feasibility of this treatment and study the outcome in the fi rst four patients to receive the cultured CB DLI. Results: No adverse reactions were seen at transfusion. Graft-versus-host disease (GVHD) after CB DLI could not be excluded in one patient (Patient 1), while the remaining three did not show any signs of GVHD. In the patient treated with CB DLI due to MRD (Patient 4), the malignant cell clone was undetectable for several months after infusion (see fi gure). In one patient with mixed chimerism, Patient 3, the percentage of recipient cells decreased in temporal association with DLI treatment (see fi gure Introduction: The Colony forming unit assay (CFU) is a progenitor cell cultivation assay used to approximate stem cell content and potency in donated cord blood units (CBU), we investigated the accuracy of the routinely performed viability measurement by Trypan blue staining prior to cultivation. Materials (or patients) and Methods: CBU were processed to Cord blood buff y coat (CBB) using the Sepax instrument (Biosafe) and 0. 5 ml of CBB was sampled and frozen in a separate vial in liquid nitrogen. The material was briefl y thawed and viability was measured by adding 0.1% Trypan Blue (T8154, Sigma), 100-200 cells were manually counted by microscopy and the fraction of stained cells was determined. For each CBB two diff erent cultivation protocols were performed in duplicate; one with 50.000 mononuclear cells (cNoT) and the other with 50.000 viable mononuclear cells as adjusted for the fraction of viable cells measured by Trypan Blue (cWiT) per ml of medium (Methocult GF H84434, Stem Cell Technologies) 37 CBU were investigated and calculations performed using SPSS (IBM) version 21.

Results: After 14 days in an incubator the growing Granulocyte-Monocyte colonies (CFU-GM) in both protocols were counted. Since only viable CFU-GM cells can form colonies, the number of growing colonies in the cNoT protocol was expected to be lower compared to the cWiT protocol relative the fraction of viable cells in the corresponding sample given that the viability measurement by Trypan Blue was accurate. We therefore theoretically calculated the number of expected colonies (cCalc) in the cNoT protocol as given by; cCalc CFU-GM colonies = (fraction of viable cells by Trypan Blue * cWiT CFU-GM colonies) The cCalc results were correlated to the number of actually growing colonies in the cNoT protocol. The Pearson correlation coeffi cient between the cCalc colonies and the cNoT colonies was 0.93 (P<0.0001). Discussion: A highly accurate viability measurement is necessary to obtain correct CFU-GM assay results. In this work we show that the fast, cheap and simple Trypan Blue viability stain is an accurate method to determine the viability of CFU-GM cells in CBU prior to release for transplantation and thus eliminating the need to implement more expensive and time-consuming methodology such as fl ow cytometry. Disclosure of Interest: None Declared. **Died of PTLD, GvHD controlled. S157 Materials (or patients) and Methods: The current study targeted high risk neuroblastoma from April 2008 to November 2013. Thirty days before launching our therapeutic modality, MIBG scintigraphy was performed. According to MIBG scan results, MIBG avid patients were consecutively enrolled in the study. Myeloablative chemotherapy comprised of etoposide (1200 mg/m 2 ), carboplatin (1500 mg/m 2 ) and melphalan (210 mg/m 2 ) two weeks following administration of therapeutic MIBG (12 mCi/kg) . In addition all patients received 13-cis retinoic acid after autologous stem cell transplantation (ASCT). Results: Thirteen enrolled patients had demonstrated various responses to previous treatments including 10 patients with very good partial response (VGPR), and 3 patients with partial response (PR). Mean age at diagnosis was 42.5 months (range, 17-65) and Mean age at transplantation was 60.2±21.3 (range, 34-92) months in patients. The median time to neutrophil engraftment after ASCT was 10 days (range, 9-13 days) and median time to platelet engraftment was 13 days (range, 10-20 days). None of the cases failed to engraft after ASCT. In studied patients, 3-y-OS was 66% ± 21% while 3-y-EFS was 53% ± 20%. Discussion: These fi ndings may underline the effi ciency of pre-ASCT MIBG scintigraphy in high-risk neuroblastoma patients. Patients with MIBG avid lesions at pre-ASCT stage may benefi t from exploitation of therapeutic MIBG combined with high dose chemotherapy. Furthermore, therapeutic MIBG may not be a necessary component of pre-ASCT regimen in MIBG non-avid patients. Disclosure of Interest: None Declared. Introduction: Posterior reversible encephalopathy syndrome (PRES) is a well-known complication of hematopoietic stem cell transplantation (HSCT). The impact of PRES after HSCT in pediatrics is not fully recognized. This study investigates the radiologic spectrum of PRES and its clinical associations in post-HSCT pediatric patients. Materials (or patients) and Methods: Seventy-four children (45 Males, 29 females) with mean age of 84 (4-180) months underwent HSCT between March and November 2013 in our center. The underlying disease was inherited abnormalities of RBC in 30, leukemia in 18, bone marrow failure syndromes in 15, primary immunodefi ciency syndromes in 8 and inborn errors of metabolism in 3. The most common pre-HSCT conditioning regimen included Busulfan and Cyclophosphamide and none of the patients received TBI. Cyclosporine was used in all patients for GVHD prophylaxis. Patients have been closely followed up with the mean follow-up time of 4 months. Brain CT-SCAN and subsequent MRI were performed in all children with neurologic symptoms. Follow-up MRI was performed two months later in patients with the diagnosis of PRES. Results: Of 74 patients, currently 62 patients (83.7%) are alive. 12(16.22%) post-HSCT children (6 males, 6 females) with mean age of 111(36-180) months developed neurologic symptoms with the most common being seizure (12) and headache(8). Brain imaging showed PRES in these patients; of whom 6 had Major Thalasemia, 4 Fanconi Anemia, one AML and one Diamond-Blackfan Anemia. Hemorrhagic foci were seen in MRI of two patients with PRES. Follow-up MRI of 4 patients showed resolution in three, while persistent fi ndings in one. Short-term Mortality rate was significantly higher among subjects with PRES (6 patients, 50%, p value: 0.003).

Discussion: PRES is a serious neurologic complication following HSCT in children and it is associated with increased mortality rate. Patients with certain underlying diseases and transplant conditions are more likely to develop PRES. Disclosure of Interest: None Declared. Introduction: Central nervous system (CNS) tumors are the second most common pediatric malignancies with an about 30% 5-year overall survival rate in high-risk group. The aim of this study was to assess the eff ectiveness of single or tandem high-dose chemotherapy (HDCT) with autologous hematopoietic stem-cell transplantation (auto-HSCT) in this patient group. Materials (or patients) and Methods: From October 2006 to December 2013, 37 pediatric patients with high-risk or relapsed medulloblastoma (N=22), supratentorial PNET (N=5), germinoma (N=4), pineoblastoma (N=2) and atypical teratoid rhabdoid tumor (N=3), choriocarcinoma (N=1) received HDCT with auto-HSCT after induction chemotherapy, radiotherapy and surgical treatment. At the moment of HDCT 15 patients were in complete remission (CR), 16 patients were in partial remission (PR) and 6 patients had stable disease (SD). Patients with germinoma received single auto-HSCT, patients with medulloblastoma, pineoblastoma, ATRT or supratentorial PNET received single or tandem auto-HSCT depending on age. The conditioning regimen for single auto-HDCT consisted of cisplatin, etoposide, and ifosfamide, or carboplatin, etoposide and thiotepa +/-intraventricular etoposide, or thiotepa and temozolomide. In tandem HDCT, the fi rst conditioning regimen was carboplatin and etoposide with intraventricular/intrathecal metotrexat, the second was thiotepa and cyclophosphamide with intraventricular/intrathecal metotrexat. Bone marrow (N=21), peripheral blood stem cells (N=11) or both (N=5) were used for stem cell sources. The mean transplanted CD34+ cell dose was 5.5 x 10 6 /kg (range, 1.0-11.3 x 10 6 /kg). Results: The median follow-up is 20 months (range, 1-95). The median time to engraftment was 15 (range 12-30) after auto-HSCT. A half (N=3) of the patients with SD at the moment of auto-HSCT died of disease progression. Thirteen of 31 patients with CR or PR relapsed 1-23 months after HDCT, the other 18 patients are currently in CR. The following therapy toxicity was observed: liver toxicity grade 3-4 (N=16), skin toxicity grade 3-4 (N=9), severe mucositis grade 3-4 (N=22), nausea/vomiting grade 3-4 (N=11), infectious complications grade 3-4 (N=23) . Four patients died of toxicity. 8-years Overall survival (OS) in all groups was 52% and 8-years disease free survival (DFS) was 49%. DFS was signifi cantly better among high-risk patients in 1 st CR compared to patients in 2 nd or following CR: 65% and 36%, accordingly (P=0.02). Patients in CR or PR at the moment of HDCT had better DFS rate than patients in SD: 61%, 53% and 25% (P=0.00), respectively. Discussion: HDCT with auto-HSCT in pediatric patients with highrisk CNS tumors may be a feasible option for patients in CR or PR after induction chemotherapy. It is ineff ective as a salvage therapy in refractory patients. Disclosure of Interest: None Declared. 1 1 Hematology, Nanfang Hospital, Southern Medical University, Guangzhou, China Introduction: Prior chemotherapy had been reported to aff ect the effi ciency of peripheral blood stem cell (PBSC) mobilization. As one of the major chemotherapeutic drugs for AML, the impact of cytarabine on mobilization remains unknown. Herein, we retrospectively analyzed AML patients performed PBSCs apheresis to explore the relation between prior medium-dose cytarabine (MD-AraC) chemotherapy and mobilization. Materials (or patients) and Methods: We retrospectively analyzed 90 patients with de novo AML, underwent mobilization in the haematology department of Nanfang hospital from August 1999 and November 2012. Median age at mobilization was 38 years (range, 12-60 years). Before mobilization, all patients had received various chemotherapeutic regimens, including induction and consolidation or intensive chemotherapies. According to the course of MD-AraC chemotherapy, patients were divided into group I, II III. With a complete response, patients received the same mobilization regimen, EA (cytarabine 1.0 g in IV, every 12 hours × 3 ~5 days; etoposide 0.1~0.2 g in IV, every 12 hours × 3~5 days) chemotherapy combined granulocyte-colony stimulating factor(G-CSF) 5~10 μg/kg/d. G-CSF starting after completion of chemotherapy 4 to 5 days when WBC down to the bottom until the end of the collection. Apheresis was scheduled to start when the WBC count recovered ≥4.0 × 10 9 /L or the CD34 cells ≥ 0.01% WBC of PB. After 24~48 h following high-dose conditioning regimens (mostly busulfan/cyclophosphamide), grafts were infused. The effi cacy of PBSC mobilization, hematopoietic reconstitution, survival rates were assessed for each AraC group. Results: The median doses of CD34 cell in those three AraC groups were 4.7×10 6 /Kg, 2.8×10 6 /Kg, 2.2×10 6 /Kg, respectively (P=0.006). In addition, patients collected ≥2.0×10 6 /kg numbers of CD34 cells in groups I need the lowest leukapheresis, total blood volume processed, G-CSF total dose and days (P<0.05). A signifi cantly greater proportion of good mobilization (≥2.0×10 6 CD34 cells/kg with at most 3 leukapheresis procedures) in the group I (39/46, 84.8%) compared with the group II (13/22, 59.1%) and group III (10/19, 52.6%) ( χ 2 =8.918, P=0.012). The sex, age, cytogenetic risk, the pior chemotherapy courses, the prior courses of the various chemotherapeutic agent drugs except MD-AraC did not correlate with mobilization response. Multivariate analysis revealed the course of prior MD-AraC chemotherapy was an independent pre-dictive factor for HSC mobilization [OR 0.627, P=0.022] . However, the MD-AraC chemotherapy has no eff ect on hematopoietic reconstitution and survival in AML patients treated with auto-HSCT. Discussion: Exposuring to the bone marrow toxic drugs is the major factor negatively eff ected mobilization, there mitoxantrone, fl udarabine, lenalidomide, platinum, alkylating agent, carmustine, nucleoside analogue, melphalan were reported. As in previous studies, prior exposure to MD-AraC also was an independent negative predictor for mobilization and without survival benefi t. The conventional chemotherapy not accurately presented a signifi cant infl uence in the mobilization in AML. In preparation for Auto-HSCT, MD-AraC must be taken into account. Disclosure of Interest: None Declared. Introduction: Choosing the perfect cord blood unit (CBU) for transplantation is based on tissue typing and the stem cell potential of the unit. CBUs are routinely assessed at collection with regards to total nucleated cells, CD34+cell count and progenitor cell cultivation protocols such as the time-consuming Colony-Forming-Unit assay. Eff orts are however continuously made towards fi nding better ways of defi ning true stemness of CBUs. Side population phenotype (SP) and activity of the Aldehyde Dehydrogenase enzyme (ALDH) are functional markers of stemness that can be assayed using fl ow cytometry. Here, we have developed a protocol for the simultaneous determination of CD34+, SP and ALDH+ populations in relation to immature leucocytes, i.e. the weakly CD45+ population of cells with low side scatter properties (CD45dimSSClow), inspired by the International Society of Hematotherapy and Graft Engineering (ISHAGE) guidelines for determination of CD34+ cell count. Materials (or patients) and Methods: 30 randomly selected CBUs were investigated in this study. 200 μl cord blood buff y coat was sampled from each CBU. After lysis of erythrocytes, 1.5×10 6 viable cells were stained with Hoechst 33342 (Life Technologies, Carlsbad CA, USA) for 90 minutes at 37°C. After a few minutes on ice, cells were stained with ALDH reagent according to manufacturers´ instructions for 30 minutes at 37°C (Stem Cell [PH-P122] S159 Technologies, Vancouver, Canada). Subsequently, the cells were put on ice and stained with the viability stain 7-AAD and antibodies directed against the CD45 and CD34 (BD, Franklin Lakes NY, USA) antigens for 30 minutes. Samples were analyzed on a FACSAriaII (BD, Franklin Lakes NY, USA) equipped with a 375nm near UV-laser. All calculations performed using IBM SPSS Statistics version 21. Results: There was no overlap between the SP and ALDH+ populations in any of the investigated units. The majority of SP cells were CD34 negative, whereas the ALDH+ population was dominated by CD34+ events. The mean size of the CD45dimSSClow population was approximately 10% of the total number of 7-AAD negative i.e. viable events. The CD34+ and ALDH+ populations amounted to a few percent of the CD45dimSSClow population and were approximately of the same size. The SP population was signifi cantly smaller Discussion: Taken together, we show that simultaneous staining for CD45, CD34, SP and ALDH+ cells is feasible using small amounts of cord blood buff y coat in a time frame possible to implement in routine laboratory work. In our study, the sizes of the ALDH+ and CD34+ populations in relation to the number of CD45dimSSClow events were approximately the same whereas the SP was smaller. There were also diff erences in immunophenotype between the SP and ALDH+ populations inferring that they represent cells with diverse biology. Since the stem and progenitor pool in each CBU is small the implementation of the CD45dimSSClow gating strategy enhances the relative size of the populations and therefore possibly the sensitivity of the measurement. In conclusion this study illustrates the heterogeneity of the stem and progenitor pool in CB both in quantity i.e. population size and quality i.e. immunophenotype and functional properties. Disclosure of Interest: None Declared. Introduction: Regulatory T cells (Tregs) may play an important role following haematopoietic progenitor cell (HPC) transplantation. In allogeneic recipients, higher Treg numbers have been shown to protect against graft-versus-host-disease (GVHD) but in both allogeneic and autologous recipients can also suppress immune responses to tumour antigens leading to higher rates of relapse. Previous studies have shown that Treg levels in the graft and also during immune reconstitution may aff ect transplant outcome. Materials (or patients) and Methods: This study considered CD3 + CD4 + CD125 high CD127 low FoxP3 + Treg levels in autologous HPC apheresis harvests (n=109) following one of three mobilisation regimens: G-CSF(Lenograstim), G-CSF and cyclophosphamide and G-CSF and Plerixafor (n=11, n=82 and n=16 respectively). Additionally Treg levels were compared in 75 allogeneic harvests from sibling (n=22) and unrelated (n=43) donors and 10 non-mobilised donor lymphocyte donations. Tregs were measured using multi-colour fl ow cytometry and expressed as Tregsx10 6 /ml, Tregs as a ratio of CD34 or CD3 positive cells and Tregs as a percentage of CD4 cells. Treg levels on two consecutive harvest days were also compared. Results: No signifi cant diff erences were observed in absolute Treg numbers/ml or ratio quantifi cations in autologous harvests mobilised by either G-CSF or G-CSF plus cyclophosphamide. Harvests mobilised by G-CSF/Plerixafor showed signifi cantly higher levels of Tregs/ml (0.42x10 6 /ml +/-0.09) than those mobilised by G-CSF alone (0.18x10 6 /ml +/-0.05) (P=0.034). Also a signifi cant increase in Tregs was observed when comparing the Treg:CD34 ratio following G-CSF/Plerixafor compared with G-CSF/cyclophosphamide regimens (0.66 +/-0.15 and 0.28 +/-0.04 respectively) (P=0.026). The levels of Tregs as a percentage of CD4 + cells did not diff er signifi cantly with any mobilisation protocol. Day of harvest did not aff ect Treg levels collected regardless of mobilisation regimen. In allogeneic harvests, no signifi cant diff erence was noted between Tregs/ml in the non-mobilised and mobilised harvests (P=0.21). However, expressing the data as either Treg:CD3 ratios or Tregs as a percentage of CD4 cells showed lower levels in products from G-CSF mobilised donors. These results which verge on signifi cance (P=0.057 and P=0.058 respectively) may be due to proportionately higher levels of CD3 and CD4 cells following G-CSF. Discussion: Overall these data indicate that neither the G-CSF or cyclophosphamide components of autologous mobilisation regimens elevate Treg levels. However, the higher absolute levels noted with G-CSF/Plerixafor may be associated with the higher white cell counts observed with this regimen. As the majority of patients in this cohort (93/109), received G-CSF or G-CSF/cyclophosphamide, it is reassuring to surmise that such products may not be associated with tumour suppression. It may however be worthy of note that Plerixafor mobilisation induces higher Treg levels/ml and as a ratio of the CD34 cells collected in the autologous setting. G-CSF used in allogeneic donors does not appear to aff ect absolute Treg levels which may be of clinical relevance to tumour tolerance and/or GVHD. Disclosure of Interest: None Declared.

Introduction: Almost 30 years have passed since the fi rst use of donor lymphocytes infusion (DLI) after allogeneic HSCT. Nevertheless, the therapy still remains poorly recognized. Mainly due to numerous indications and small number and huge diversity of patients (pts.) treated with DLI. The effi cacy and safety of DLI in pediatric pts. is particularly poorly studied. The aim of our study was to describe experience of PPGHSCT in respect of DLI. Materials (or patients) and Methods: The study included 51 pts., who underwent DLI in the years 1993-2011. The study group comprised 23 girls and 28 boys, aged median 9 (1-22) years, grafted because of oncohematological (45) and nonproliferative disease (6). The indications for DLI were as follows: 1) increasing recipient chimerism after nonablative HSCT (n=18), 2) modulation of reduced intensity conditioning regimen (n=2), 3) persistence or progression of minimal residual disease (MRD; n=3), and 4) relapse after HSCT (n=28). Results: DLI was carried out in median 6.0 (0.5-79.0) months after HSCT. The time of initiation of DLI was diff erent depending on the indications type (P=0.04). Median time from completion of immunosuppressive therapy to DLI was 49 (-28-1758) days. No diff erences were demonstrated in respect of interval between the completion of immunosuppressive treatment and start of DLI depending on the indications type (P>0.05). The source of DLI was peripheral blood cells collected by apheresis: nonstimulated and unselected (n=22), or G-CSF mobilized (n=29). The DLI was administered as a single dose (n=19), or in fractions (n=32). In case of fractionated DLI, it was administrated an average of 2 (1-11), on median each 5 weeks. The dosage of DLI was determined based on the count of CD3(+) cells. The total dose of CD3(+) cells was a median 28.0 (0.1-730.0) x10 6 /kg recipient's body weight (k.b.w.). The number of administrated CD3(+) cells per a dose was a median 6.0 (0.1-240.0) x10 6 /k.b.w. The total dose of CD3(+) cells and the number of doses were not diff erent between the indications (P>0.05). Thirty six pts. received only DLI and 15 pts. DLI with cytoreductive (n=12) or specifi c (n=3) therapy. The effi cacy of DLI was assessed based on clinical parameters and imaging studies, as well as by testing a donor chimerism by PCR and/or monitoring of MRD. The time for assessment of DLI effi cacy ranged from 0 to 70 (median 3) months. During the evaluation, 18 (35%) pts. experienced complete remission, in 3 (6%) pts. progression of the disease slowed down, 19 (37%) pts. showed relapse and progression of disease and 11 (22%) pts. rejected the graft. DLI was found as eff ective in 41% of cases. The effi cacy of DLI was aff ected only by the total dose of CD3 cells (P=0.02). Other factors like: time from HSCT to DLI, method of application (one dose vs. fractionated), source of cells, occurrence of GvHD after HSCT or type of the indication were not important to the outcome of DLI (P>0.05). In 18 (35%) pts. after DLI, complications of diff erent intensity were found (including sever bone marrow aplasia -5 pts. and GvHD -13 pts.), due to which 2 (4%) patients died. Discussion: DLI shows its effi cacy in a signifi cant percentage of children undergoing HSCT. It has low mortality rate due to complications. However, this method requires standardization. Disclosure of Interest: None Declared. Introduction: With sequencing technology improvements, the gut microbiome has been a topic of study in many diseases with an immunological component. In adult haematopoietic stem cell transplant (HSCT) patients, a change in the gut microbiome through HSCT has been associated with GvHD-related intestinal infl ammation and bacteraemia. We performed a longitudinal prospective pilot study to explore changes in gut microbiome in severe combined immunodefi ciency (SCID) patients through HSCT. Materials (or patients) and Methods: Stool samples were analysed longitudinally from 3 SCID patients (Babies A, B and D) based on occurrence of clinical events (conditioning, HSCT, neutrophil engraftment, pyrexia, antibiotic therapy, GvHD). Bacterial profi les were generated based on the 16S ribosomal DNA gene using both denaturing gradient gel electrophoresis (DGGE) and Illumina MiSeq next generation sequencing. Results: Combining all 3 patients in a partial least squares discriminant analysis (PLS-DA) showed that pre-and post-HSCT samples grouped separately (Figure) . In all 3 patients a low Shannon diversity index was observed which fl uctuated throughout HSCT. Baby A was 8 months old on admission and 9 months at HSCT. Pre-HSCT, the bacterial profi le was a mix of Staphylococcus and Clostridium XIVa. Post-HSCT the bacterial profi le was dominated by Staphylococcus. Pyrexia on day -3 occurred in conjunction with the appearance of Escherichia in stool. Baby B was 8 days old on admission and 6 weeks old at HSCT. Prior to conditioning samples were dominated by Bifi dobacterim and Streptococcus. Post-HSCT, Enterococcus and Staphylococcus dominated. The change between pre-and post-HSCT coincided with pyrexia, central line infection and antibiotic use. Growth of Enterococcus post-HSCT coincided with disseminated adenovirus infection. Baby D was 6 days old on admission and 5 weeks old at time of BMT infusion with no conditioning. At 3 weeks of age there was a dominance of Enterococcus in stool, which gradually declined over time post-HSCT. Pyrexia, GvHD and disseminated CMV developed from day 33 post-HSCT, when Enterococcus, Klebsiella and Eschericha were dominant in stool samples. Discussion: Despite the individual nature of the gut microbiome in each patient we observe a clear diff erence in the bacterial community following the HSCT process. The diversity was shown to be unstable over the duration of the study. This small study represents the fi rst such study to explore the gut microbiome in HSCT in SCID but eff ects of the microbiome on complications post HSCT in SCID and other primary immunodefi ciencies warrants further work. Disclosure of Interest: None Declared. Introduction: Autologous hematopoietic stem cell transplantation is a well established and almost inevitable step in the treatment of patients with multiple myeloma, proven by the improvement of survival rates presented in a sea of data from an enormous number of trials. Cryopreservation of harvested peripheral blood stem cells is an everyday routine in almost every transplant center. But is it possible to avoid additional costs, and DMSO intoxication of the patients without infl uencing the eff ectiveness of the procedure. We are caring to present our modest results of using noncryopreserved hematopoietic stem cells for transplantation of patients with multiple myeloma. Our motive was infl uenced by the fact that the viability of stem cells is not signifi cantly changed if stored at 4 ° C for at least 5 days. Materials (or patients) and Methods: Knowing the fact that the duration of the conditioning regimen for transplantation of patients with multiple myeloma (melphalan 200 mg/m 2 ) is short, we anticipated that the use of noncryopreserved hematopoietic stem cells in this setting would be more than appropriate and a good opportunity for simplifying the transplantation procedure without infl uencing it ' s eff ectiveness. We have evaluated 14 patients diagnosed with multiple myeloma, 9 males (64%) and 5 females (36%), with an average age of 51.9 years (range 39-65). Results: The majority of patients (64.3%) were treated according to the CTD (Cyclophosphamide, Thalidomide, Prednisone) regimen, and the reevaluation of the disease status before stem cell transplantation (SCT) showed that 9 patients (64.3%) were in very good partial response (VGPR) and 4 patients (28.5%) with complete response (CR). The median time from diagnosis to SCT was 7.8 months (range [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] . The average number of apheresis procedures was 1.7 (range 1-3), and the average number of collected cells was 3.1x10 8 /kg TT mononuclear cells (range 4.0-2.0). G-CSF mobilizing regimen was used in most of the patients. The number of days to confi rmed engraftment in our group of patients was 11.5 (range 9-15). The amount of blood transfusions was on average 0.7 (range 0-4), and transfusion of thrombocytes (mainly pooled thrombocytes) 24.7 units (range 0-82). When we compared this results with the control group of patients, searched retrospectively, where cryopreserved cells were used, we couldn`t confi rm any signifi cant diff erence in the duration of aplasia (median time to engraftment 11.6 days (range 8-20)), nor any signifi cant diff erence in transplant related mortality or post transplant complications. There was a diff erence in the amount of transfused blood products but that was related to complications not related to the transplant procedure. Discussion: We must conclude that the use of noncryopreserved hematopoietic stem cells in transplantation of patients with multiple myeloma represents a safe, equally eff ective and relatively simple procedure, that has, among other positive values, costeff ective advantages and eliminated DMSO intoxication of the patient. However, for performing an optimal SCT in this manner, a good coordination and timing between teams for stem cell collection, administration of high-dose chemotherapy conditioning regimen, storage and application of noncryopreserved stem cells, and good supportive treatment of the patients in the transplant unit is essential. Disclosure of Interest: None Declared. Introduction: BEAM is the standard conditioning regimen for lymphoma patients undergoing autotransplant. Fotomustine could be used as substitute of BCNU in this regimen Materials (or patients) and Methods: Here we present the results in 34 consecutive patients treated with FEAM compared with 54 treated with BEAM. Patients were similar in terms of diagnosis, median age, number of previous therapies, disease status at transplant, number of CD34+cells infused (3.6x10 6 /kg). Results: Hematopoietic engraftment and treatment related toxicity are shown in table1. Thirty-four (77%) BEAM patients required more than 20 days to recovery platelets vs 11(46%) FEAM patients, (P=0,01). Mucositis (98% vs 82%, P=0.019) and diarrhea (98% vs 85%, P=0.044) seem occurred more frequently in the BEAM group. Almost all patients in the two groups had fever during neutropenia. The median days of hospitalization were 22 in FEAM and 25 in BEAM group:54% of patients treated with BEAM and 35% with FEAM needed more than 24 days of hospitalization, P=0.06. The incidence of TRM was 9% in BEAM and 12% in FEAM cohort, P=0.9. At univariate analysis, older age was only factor infl uenced negatively TRM. In patients <65years, TRM was 7% in BEAM and 4% in FEAM, P=0.05. The 90-day overall response rate (ORR) was 86% (71% CR) and 85% (65% CR) in patients treated with FEA-Mand BEAM, respectively. No diff erence was observed regarding 90 day ORR between the two groups.After 31 months of median follow-up , the estimated 2y EFS was 57%, without diff erence among the two groups. Discussion: In our experience FEAM ensured a reduction of mucositis, diarrhea and a more rapid platelets engraftment with a reduced hospitalization. We observed a higher incidence of TRM in both groups, being 32% (FEAM) and26% (BEAM) of our patients older 65years. In younger patients the TRM did not diff er from that reported in the literature and in patients treated with FEAM the TRM occurred less frequently.In terms of effi cacy, ORR and EFS of FEAM was comparable to BEAM, however longer follow-up is needed to evaluate fully its effi cacy and long term safety. Disclosure of Interest: None Declared.

Introduction: Hematological malignancies are a heterogeneous group with variable incidence and prognosis. The overall survival and cure rates of patients with HMs have improved dramatically for ex through stem cell transplantation either autologous or allogeneic. To achieve these results there is a need for prompt intensive treatment, but it aff ects fertility and bone density. The risk of developing premature ovarian failure is a major concern in longterm survivors and there is therefore a great need for information on fertility issues. This descriptive study, conducted in the Stockholm region, assessed bone mineral density and reproductive health in hematological cancer survivors after haematopoietic stem cell transplantation (HSCT). Materials (or patients) and Methods: Thirty-seven premenopausal women having undergone autologous or allogeneic HSCT were consecutively included in the study between1998 and 2010; they were followed annually until 2011. The diagnoses were acute lymphoblastic leukaemia (n=6), acute myeloid leukaemia (n=9), chronic lymphocytic leukaemia (n=1), chronic myeloid leukaemia (n=12), Hodgkin lymphoma (n=4), non-Hodgkin lymphoma (n=5). Fertility preservation options were off ered before cancer treatment to most women. The mean age at diagnoses was 27 and at the fi nal evaluation 39 years. Twenty-nine women (78.4 %) with a mean age of 25 (21-43) years had never been pregnant. All women were in a menopausal state (POF) due to given therapy. Hormone substitution was given and bone mineral density was measured repeatedly. Results: Twenty-six patients received allogeneic HSCT and eleven autologous HSCT. Before allogeneic HSCT, nineteen patients received myeloablative conditioning; seven had reduced-intensity conditioning. Eleven patients got total body irradiation. Eight patients were transplanted with grafts from an HLA-identical sibling donor while 18 had unrelated donors. Three patients developed venous thrombosis after the HSCT; two malignancies occurred after allogeneic HSCT, one malignant brain tumor, one cancer in situ cervices. During follow-up, bone mineral measurements showed a slight increase in the spine while no decrease was seen in femoral neck and total hip values over time. Resumption of menstruations occurred in three women allowing spontaneous pregnancies, two babies were born and one pregnancy was terminated. Additionally, three pregnancies were achieved with oocyte donation, surrogacy, and adoption respectively. Discussion: The risk of infertility has not been extensively studied after HSCT. The type and intensity of cytotoxic agents and the cumulative doses are, besides age, the most important factors determining the likelihood of gonad failure. A resumption of menstrual cycles does not guarantee normal fertility. However, the oestrogen substitution is crucial in women with POF. This study highlights the effects on bone density and fertility after HSCT for hematological malignancies. Since the number of patients who regained their menstruation and/or became mothers is very small, it is difficult to draw conclusions regarding important background factors. Early preservation of fertility is important issue, as the management of general health in these groups. A multidisciplinary collaboration between hematologists and Assisted Reproductive Techniques teams is needed. Disclosure of Interest: None Declared. Introduction: Second allogeneic stem cell transplants are uncommon but can deliver a potent graft versus malignancy eff ect and re-establish normal haematopoesis in patients who are fi t enough to undergo the procedure. Materials (or patients) and Methods: Our retrospective study included 27 patients (M: 16, F: 11) with a median age at second transplant of 43 (range: 19-66). The most common indication for transplantation was acute leukaemia in 13 patients (52%). Second transplantation was performed for relapsed disease in 11 patients (40%) and graft failure in 14 patients(52%) whereas a secondary haematological malignancy was the indication for transplantation in two patients (8%). Matched sibling donors were used for 16 patients (59%) of second transplants while the remaining 11 (41%) were fully matched unrelated donor transplants. 12 transplants were T depleted with either alemtuzumab in vivo (8/12) or Alemtuzumab ex vivo (4/12) . 7 transplants (26%) were carried out using myeloablative conditioning whereas reduced intensity conditioning was used in 20 patients(74%) . The same donor was used for both transplants in 7 (26%) cases. 9 patients received a T deplete protocol for their fi rst transplant and subsequently underwent T replete second HSCT. Results: The median time for neutrophil engraftment was 14.5 days (range 2-52 days). Severe grade 3 or 4 acute graft versus host disease occurred in 15% patients whereas chronic graft versus host disease was observed in 26%. Median overall survival was 12 months (Range: 0 to 101) with 55% of patients surviving more than 12 months after HSCT. 63% of deaths were related to non relapse mortality (NRM) whereas 32% were due to progressive disease. Discussion: Our single centre retrospective analysis suggests second allogeneic transplantation can achieve reasonable overall survival although as expected non relapse mortality is signifi cant. In view of the smaller number of patients in our single centre study this needs to evaluated from the registry data to identify which group of patients derive signifi cant benefi t from second HSCT. This information will enable clinicians with informed decision making for patients undergoing second allogeneic HSCT. Disclosure of Interest: None Declared. Introduction: Optimization of pre-transplant risk assessment is a crucial issue to improve the allo-HSCT decision making process. To date 2 major algorithms are in use in clinical practice: the EBMT risk score and the Hematopoietic Cell Transplantation Comorbidity Index (HCT-CI) score. Recently a disease risk index (DRI) -Armand et al, Blood 2012;120:905 -was defi ned to calibrate HSCT outcome across studies and centers. Here we are presenting the results of a retrospective study designed to evaluate the strength of the 3 aforementioned score in stratifi cation of transplant outcome after an haploidentical HSCT (haplo-HSCT). Materials (or patients) and Methods: We included 183 adult patients (pts) who underwent an haplo-HSCT for hematologic malignancies, between 2004 and 2013 and were reported to the local database. Risk assessment score and outcome analysis included all consecutive pts receiving an haplo-HSCT as 1 st allogeneic transplantation. Pts receiving haplo-HSCT as 2 nd or 3 rd HSCT were excluded from the present analysis. Their median age was 48 (range, 15-77) years. The cohort included a broad representation of diseases (116/183 acute leukemia, 22 Hodgkin lymphoma) and disease status. Donor source was mobilized peripheral blood stem cell in all the cases. Only 5 pts were conditioned with a nonmyeloablative regimen. The median follow-up for survivors was 25 months. Results: The overall survival (OS) at 2-y was 30% and the transplant related mortality at 100-days 23%. The 2y OS according to EBMT / HCT-CI / DRI risk score are reported in table. The evaluation of the HCT-CI impact after DRI stratifi cation was still able to show a signifi cant diff erence in outcome showing better survival for pts with low DRI score and low HCT-CI score as expected. Discussion: DRI score and HCT-CI score predict survival after haplo-HSCT. The integrated application of DRI and HCT-CI may improve the defi nition of transplant eligibility for pts candidate to allogeneic HSCT form alternative donors including family haploidentical source. Disclosure of Interest: None Declared. [PH-P131] S164 with decreased viability of cells. Trypan blue assay show the percentage of all survived cells without possibility of determining the number of survived CD34+ cells. Furthermore, it is assumed that the stem cells are more resistant to cryopreservation and it is assumed that not survived cells belong to granulocytes. The aim of this study was to determine infl uence of reinfused CD34+ cell number and cell viability performed with trypan blue assay, on the speed of hematopoietic recovery. Materials (or patients) and Methods: We analyzed 114 PBSC products with leukapheresis performed on cell separator in the period from January 2011 until November 2013. Absolute numbers of CD34+ cells were enumerated on fl ow cytometer. PBSC products were cryopreserved with 10 % DMSO. Before freezing, cell viability was determined by trypan blue assay. Products were gradually frozen with programmed average speed of 2.3 ° C/min to a temperature of -196 ° C and stored in liquid nitrogen. PBSC autologous transplantation was performed in 114 patients with median age of 56 years (19-72) and diagnosed multiple myeloma (n= 51), non-Hodgkin's lymphoma (n= 37), Hodgkin's disease (n= 22) and acute myeloid leukemia (n= 4). On the fi rst posttransplantation day each patient received pegylated fi lgrastim 6mg sc. Time elapsed from transplantation to hematopoietic recovery (leukocytes >1x10 9 /L, neutrophils >0.5x10 9 /L and platelet count >20x10 9 /L at least two days after platelet transfusion) as well as number of transfused blood cell products was measured. Results: The median number of CD34+ cells infused was 5.6x 10 6 / kg of body weight (range 1. Blood samples were drawn before the infusion and at 0.5 -8.0 h following a start of the drug administration. After collection the blood samples were immediately adjusted by 1 M citric acid to pH below 5 to avoid artifi cial ex vivo activation of TREO. Plasma concentrations of TREO and S,S-EBDM were determined by an HPLC method with tandem mass spectrometry detection. Pharmacokinetic parameters were calculated in WinNonlin 6.2. Results: Changes in plasma concentrations of TREO were best described by a two-compartment model, whereas levels of S,S-EBDM were best fi tted by a one-compartment model. Values of C max and area under the curve (AUC) of TREO as well as S,S-EBDM increased with the dose of TREO. Following 2 h infusion of TREO at dose of 12 g/m 2 (n = 8) and 14 g/m 2 (n = 8) the mean C max values of the parent drug were 1926 ± 906 and 3070 ± 2129 μM, whereas the mean C max values of S,S-EBDM amounted to 14 ± 11 and 18 ± 11 μM, respectively. In turn the mean AUC of TREO after the dose of 12 g/m 2 and 14 g/m 2 was 5664 ± 2458 and 8644 ± 4849 μM × h, whereas the values obtained for S,S-EBDM equaled to 43 ± 26 and 59 ± 34 μM × h, respectively. The mean biological half-lives of TREO and S,S-EBDM were 1.77 ± 0.51 h and 1.83 ± 1.15 h (n = 26), respectively. Discussion: The changes in S,S-EBDM concentrations in the pediatric patients' plasma followed TREO levels, though concentrations of the epoxide were two-order lower in comparison to the parent drug. This might be explained by the fact that formation of S,S-EBDM from TREO in vivo proceeds simultaneously with its elimination. Importantly, biological half-lives of the parent drug and its epoxy-transformer were comparable, therefore S,S-EBDM is supposed to be completely eliminated from the patients' blood within relatively short time, similar to TREO. Disclosure of Interest: None Declared. Introduction: Peripheral blood stem cells are the major source of stem cells for autologous and allogeneic hematopoietic stem cell (HSC) transplantation. In June 2012, we implemented the cell separator Spectra Optia for stem cell harvest, and since May 2013 we have used it exclusively for apheresis procedures at our center. As a standard, donors are mobilized with G-CSF, and stem cell mobilization in patients is performed with G-CSF after chemotherapy. However, in some patients, particularly after heavy chemotherapeutic pretreatment, this approach fails to yield the target number of HSC. In such cases, plerixafor is added to the G-CSF regimen. We evaluated the performance of the Spectra Optia device in the allogeneic and autologous settings and in plerixafor mobilized patients. Materials (or patients) and Methods: In this retrospective study, 46 autologous mobilized patients underwent 124 leukapheresis procedures with Spectra Optia (Terumo BCT). Seven patients, who poorly mobilized or failed at G-CSF induced peripheral blood stem cell collection, plerixafor was used as mobilization agent (in 18 apheresis procedures). We also performed 10 stem cell collections in 8 allogeneic donors. Procedure performance was evaluated on CD34+ cells collection effi ciency (CD34 CE2) (calculated as 100 x (CD34+ cells/kg collected x body weight)/ (CD34+/μL x blood volume processed))). Data are presented as median (min-max). Results: In the standard mobilization regimen cohort, median CD34+ cell precount was 36/μL (5-545/μL) with 9x10 9 /L leukocytes (0.4-75.4x10 9 /L). CD34+ cell yield was 2.8x10 6 /kg (0.3-18.3x10 6 /kg) corresponding to a CD34 CE2 of 62% (21-139%). Sixty-eight procedures (64%) yielded a CD34+ cell dose above 2x10 6 /kg in a single apheresis. A median of 9944 mL (3109-16833 mL) whole blood was processed (2.3 times total blood volume (TBV) (0.5-3)) in a median of 201 min (107-277 min). Despite a comparatively lower CD34+ cell count in the plerixafor cohort (10/μL (2-22/μL)) before the procedure, a median yield of 0.8x10 6 /kg CD34+ cells was collected in a single apheresis with Spectra Optia. We reached a similar CD34 CE2 (64% (18-144%)) as for standard mobilization. A median of 11902 mL (3550-17545 mL) of whole blood was processed (2.4 TBV; in 220 min (91-252 min). The median run time was slightly longer (261 min (148-278 min)) in the donor cohort. A median of 13211 mL (5743-18903 mL) of blood was processed (2.8 TBV; , and the CD34 CE2 (54%; 18-84%) was not statistically diff erent to the one observed in autologous HSCT patients. Discussion: We observed good clinical performance of the Spectra Optia device in harvesting allogeneic and autologous stem cells.

Our experience showed that Spectra Optia delivered median CD34+ cell collection effi ciencies above 50%, even in poor mobilized patients additionally treated with plerixafor. Disclosure of Interest: None Declared. Introduction: Major ABO-incompatible allogeneic stem cell transplantation may lead to acute and delayed haemolytic reactions, delayed red blood cell (RBC) engraftment or pure red cell aplasia. Preventive measures to overcome these risks are graft manipulation by in vitro RBC depletion or in vivo absorption of recipient's anti-A/B antibodies by slow infusion of donor-type RBCs pre transplantation. Furthermore, the use of an anti-A/B immunoadsorption column allows specifi c depletion of anti-A/B-titers before transplantation. Materials (or patients) and Methods: We report on an 18 year old male patient with Fanconi anemia who underwent allogeneic HSCT with an unrelated matched (10/10) bone marrow transplant after reduced intensity conditioning with fl udarabine, cyclophosphamide, ATG, and methylprednisolone. Immunosuppressive protocol consisted of cyclosporine A (started on day -7) and MMF (started on day -7). Because of major ABO incompatibility (donor AB positive, recipient 0 positive) and impossibility of red cell depletion in the stem cell product, immunoabsorption with Gly-cosorb®-ABO column processing twice the plasma volume with COBE Optia was performed on days -5 to -1 after administration of rituximab once on day -21. Results: Neutrophil engraftment was on day +28. Bone marrow on day +30 showed complete remission and full donor chimerism. Bone marrow on day +100 showed complete remission, but mixed donor chimerism (85% donor type). Anti-A and anti-B isoagglutinine titers pre-and posttransplantation were as shown below. There was normal reticulocyte engraftment without signs of posttransplant pure red cell aplasia or haemolysis. Discussion: Antigen-specifi c immunoadsorption with Glycosorb®-ABO column allowed an eff ective reduction of antibody titers in a patient with major ABO-incompatible allogeneic HSCT for Fanconi anemia. Further evaluation is needed to defi ne the most useful approach to apply this technique in major AB0-incompatible allogeneic HSCT. Disclosure of Interest: None Declared. There were no treatment-related deaths. Toxicities were manageable and reversible: mucositis (48% G2, 39% G3), skin (22% G2, 7% G3), self-limited transaminitis (34% G2, 11% G3) and hyperbilirubinemia (27% G2, 18% G3) (no cases of VOD). No cardiac,lung, renal or CNS toxicities. Neutrophils and platelets engrafted at median d+10 (8-12) and d+12 (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) , respectively. Response and CR rates were 82% and 68%, respectively. and logistics, and a committed donor. This procedure has shown promissing results in patients diagnosed with relapsed or refractory Hodgkin´s lymphoma (Burroughs LM et al. Biol Blood Marrow Transplant 2008; 14:1279 -1287 . Materials (or patients) and Methods: We retrospectively evaluate the results of HAPLO-HSCT with RIC regimens (Fludarabine 30 mg/ m 2 x 5 days (-6 to -2), Cyclophosphamide14,5 mg/kg x2 days (-6 to -5), Busulfan IV 3,2 mg/kg x 1-2 days (BUX, days -3 to -2) or 200 cGy TBI on day -1) and GVHD prophylaxis based on HD-CY (50 mg/kg on days +3 and +4) and a calcineurin inhibitor plus mycophenolate from day +5 performed in GETH centers to patients diagnosed with relapsed or refractory Hodgkin´s lymphoma. Results: From March-2009, 29 HAPLO-HSCT have been performed in patients diagnosed with relapsed or refractory Hodgkin´s disease in 11 GETH centers. Median age was 31 years (18-53), 19 were males and all were in advanced phases of their disease. Autologous HSCT was previously employed in 90% of them, and allogeneic HSCT in 10%. Disease status at HAPLO-HSCT evaluated by PET was complete remission in 8 (28%) and persistent disease in 21 (72%). Bone marrow was the stem cell source in 15 (52%) and peripheral blood in 14 (48%), without T-cell depletion in all cases. The haploidentical donor was the patient´s mother (13), father (2), brother (8), sister (5) or daughter (1) . The RIC regimens employed included 1 dose BUX (11), 2 doses BUX (14) or 200cGy TBI (4). Median neutrophils engraftment was day +17 (11-44) and platelets >20K was day +26 (11-150). Main toxic complications were grade II-III muchositis in 50%, febrile neutropenia in 75% and CMV reactivations in 58% with a transplant related mortality rate of 7% (2/29) at day +100 and 17% (5/29) at 6 months post-transplant. Acute GVHD grade II-IV aff ected to 7/28 patients at risk (25%), with grade III-IV in 3/28 (11%). Chronic GVHD was present in 3/19 (16%), being extensive in 1/19 (5%). After a median follow-up of 9 months (0.3-49), 13/22 (59%) remain alive and in complete remission. Relapse or progression occurred in 6/28 (21%). Immune reconstitution was fast and complete in those evaluated. Discussion: Haplo-HSCT with PT-Cy is a useful tool in the treatment of patients with relapsed or refractory Hodgkin´s lymphoma, rendering long-lasting remissions with limited toxicity, low GVHD incidence and early immune reconstitution. Disclosure of Interest: None Declared.

Introduction: The aim of this registry-based retrospective study was to analyze the outcome of second alloSCT (alloSCT2) performed in patients with malignant lymphoma who had relapsed after a fi rst alloSCT (alloSCT1). Another purpose was to identify prognostic factors which might infl uence outcome after allo-SCT2. Materials (or patients) and Methods: Primary endpoint was disease-free survival (DFS) measured from alloSCT2; secondary endpoints were overall survival (OS), non-relapse mortality (NRM), and incidence of relapse (REL).

Eligible were patients >=18 years who were registered within the EBMT and had received an alloSCT2 for relapse of lymphoma between 2000-2011 with a minimum interval of 3 months between alloSCT1 and alloSCT2. Centers with eligible patients were contacted to provide additional treatment and follow-up information.

Statistical analysis was descriptive and employed log rank comparisons for univariate assessment of the impact of baseline characteristics on survival endpoints. Results: 229 patients were identifi ed who fulfi lled the inclusion criteria. Additional information upon center request was provided for 84 them, resulting in exclusion of 19 patients who had been re-grafted for reasons other than relapse. The fi nal study sample of 65 patients had a median age of 39 (18-67) years, a median interval from diagnosis to alloSCT1 of 21 (4-107) months, and a median interval from alloSCT1 to alloSCT2 of 18 mantle cell lymphoma, four patients with an anaplastic T-cell lymphoma, four patients with follicular lymphoma and one patient with an undefi ned NHL. There were twenty male and eleven female patients with a median age of 52 years (range 22-70), with 20% above the age of sixty. The median lines of previous therapies were 2 (range: [1] [2] [3] [4] . 29 patients were treated with BeEAM and 2 patients with mantle cell lymphoma received additionally Zevalin. Results: All patients had chemosensitive disease and before transplantation twenty -six patients (84%) were in complete (CR) and 5 (16%) in partial remission. A median number of 4,12*10 6 CD34+ cells/kg (range: 2,20-9,96) was infused. All patients showed engraftment with a median time to achieve an absolute neutrophil count > 1*10 9 /L of 10 days (range 7-13) and to platelets >20*10 9 /L of 11 days (range 5-26). The median time of fever was 6 days (range: 0 -22). The most common grade 3 and 4 toxicity during the whole treatment period were diarrhoea (n=10), mucositis (n=7) and febrile neutropenia (n=6), followedby nausea (n=4) and cardiologic toxicities (n=3). There were no pulmonary toxicities observed and no transplant related mortality occurred. The median time of follow-up is 15 months. After at least day +100 20 patients (65%) were still in CR, while 11 patients (35%) showed progression after a median time of 5 months after transplantation (range 2-14), Until today fi ve patients (16%) died (4 DLBCL, 1 HL), all due to lymphoma progression. The 1-year PFS is 68% and the 1-year OVS 87%. Discussion: Thus Benda EAM seems to be feasible with a promising response rate und a randomized trial comparing BendaEAM with BEAM is warranted. Disclosure of Interest: None Declared. 3 Hematology, CHU BRABOIS, Nancy, 4 Hematology, CHU, Reims, 5 Hematology, CHU, Tours, 6 Hematology, CHU ST ELOI, MONTPELLIER, 7 Hematology, CH, Argenteuil, 8 Hematology, CHU, TOULOUSE, 9 Hematology, CH, Saint etienne, 10 Hematology, CHU, angers, 11 Hematology, CHU, PIERRE BENITE, 12 Hematology, Hopital Saint Louis, Paris, France

Introduction: Patients with refractory HL or early and disseminated relapse have a very poor outcome with a 5-years EFS of 46% with tandem autotransplantat. 1 Reduce intensity conditioning (RIC) followed with allogeneic transplantation in this setting was associated with a 4-year PFS at 24% 2 mainly because lack of stable remission at the time of allotransplant. Fort these reasons we decided to propose to this high-risk group tandem transplantation with a second allotransplant if an HLA matched donor was available. A donor research was initiated at registration including HLA-matched sibling donor and HLAmatched unrelated donor (10/10). Materials (or patients) and Methods: Patients received second-line chemotherapy to obtain response (assessed by PET scans) then a BEAM regimen followed 2 or 3 months later by RIC transplantation (fl udarabine and busulfan regimen) or a second autotransplant (BAM regimen). From November 2010 to October 2013, 94 patients (from 25 French centers) have been included and we will focus on the 75 fi rst patients with available data. Patients characteristics are: median age at 28 yrs , sex ratio at 1, 48 % of patients had primary refractory disease (biopsy proven in 50%) and 52% relapse with a median time to relapse at 6 months. First line chemotherapy was ABVD in 56 cases (intensifi ed with BEACOPP/VABEM in 12 cases), BEACOPP in 19, eighteen patients received radiotherapy. Second line was mostly MINE, DHAP or ICE and 42% of patients received one second line chemotherapy before the BEAM regimen, the remaining received 2 to 4 lines (with Brentuximab vedotine in 9 patients ) and before BEAM.

Results: 41% of patients had and HLA donor (familial in 39%), six patients did not receive any transplant for refractory disease and 16 patients progressed early after BEAM and received chemotherapy (with brentuximab vedotin in 8 cases). 70% of patients received the second transplant( allogeneic in 62%) (in tandem in 80% of cases or after CT for early relapse). 22 patients did not receive the second transplant (toxicities n = 5, early progression n = 4 insuffi cient stem-cell collection n= 5 or physician/patient decision n = 9). On intent to treat analysis 30% of patients had at least one progression after inclusion. Twelve patients died in the study, 3 deaths from sepsis in patients in CR after the protocol (2 in tandem auto/allo). 12 patients had AGVH grade 4 in 2 cases in patients progressive before the transplant). One further patient progressed after autotransplant, received further CT and secondary allo and died from sepsis. Eight patients died from HL, three patients having received no transplant due to refractory disease, 2 having received only one autotransplant and rapid progression and 3 relapsing after the procedure (auto/allo). Discussion: This protocol is still ongoing up to 100 patients. We confi rm the poor prognosis of these patients with refractory/early relapse of HL despite intensive fi rst-line CT. The procedure is feasible without unexpected toxicity of the procedure, the main cause of death remaining progressions, Brentuximab vedotin may help to increase the number of patients achieving CR for transplant. 1 Introduction: Uncertainty remains in regards to the role of highdose therapy (HDT) and autologous hematopoietic cell transplantation (auto-HCT) in the front-line treatment of aggressive lymphomas. Accordingly, this systematic review/meta-analysis aims at evaluating the totality of evidence pertaining to the efficacy of HDT and auto-HCT as treatment intensifi cation after frontline therapy with rituximab-based chemotherapy regimens in aggressive lymphomas. Materials (or patients) and Methods: A total of 1956 references were identifi ed through a systematic search of PUBMED/MEDLINE (n=1616) and abstracts from relevant conference proceedings or trial registries (n=340) performed on September 27, 2013. Four references were considered eligible for inclusion in our analysis; however, one of these represented an ongoing clinical trial (Le Gouill et al. J Clin Oncol 29: 2011 (suppl; abs 8003) ). Data was meta-analyzed for benefi ts: complete remission (CR) rates (data from 2 studies, total n=485), progression-free (PFS) (data from 2 studies, total n=652), event-free (EFS) survival (data from 1 study, total n=248), and overall survival (OS) (data from 3 studies, total n=900); and harms: grade 3/4 adverse events (AE) (data from 1 study, total n=399) including infection-related AEs (data from 1 study, total n=253), and treatment-related mortality (TRM) (data from 3 studies, total n=900) using the random eff ects model. Results: Three studies enrolled a total of 900 subjects in the chemoimmunotherapy-only (CIT) (group 1, n=455) or the CIT followed by HDT/auto-HCT (group 2, n=445) arm. Introduction: Splenectomy before allogeneic stem cell transplantation (ASCT) is usually performed in some individuals diagnosed with lymphoma and myelofi brosis. However, much is still unknown regarding the impact of splenectomy on the outcome of ASCT. Previous work has suggested an association between acute and chronic graft versus host disease (GVHD) and splenectomy prior to ASCT. Conversely, more recent work failed to see this correlation and instead showed an association between splenectomy and relapse after chronic myeloid leukemia. Materials (or patients) and Methods: Therefore, we performed a retrospective analysis on the impact of splenectomy on clinical outcome in patients transplanted at Karolinska University Hospital between 1995 and 2012. A total number of 1158 patients were included in this study (31 with splenectomy and 1127 without). Diff erent clinical parameters were analyzed after ASCT.

Results: The only incidence that was signifi cantly diff erent between patients with or without splenectomy was the occurrence of post transplant proliferative disease (PTLD, P=0.001). No diff erence could be observed regarding transplant related mortality or overall survival. In a multivariate analysis, including parameters that were signifi cantly diff erent between the two patient groups, splenectomy was the only factor signifi cantly associated to PTLD (P= 0.01).

Discussion: This fi nding might argue for a distinct role of the spleen in the control of Epstein Barr virus (EBV) associated PTLD.

The results also highlight that there might be a potentially higher risk for EBV-PTLD in patient sub-categories treated with anti-thymocyte globulin (ATG) that have undergone splenectomy. Further, multi-center studies on patients treated with ATG is needed in order to verify our fi ndings on the role of the spleen and EBV biology. Disclosure of Interest: None Declared. All patients received unmodifi ed grafts from a matched related (n=12), matched unrelated (n=10) or mismatched unrelated (n=7) donor. Progression-free (PFS) and overall (OS) survival were calculated from the time of allo-HCT. Kaplan-Meier survival curves and a permutation-based logrank test were used to compare PFS and OS based on alemtuzumab use and the sIPI. Results: All but one patient engrafted with full donor chimerism. The cumulative incidences (CI) of grade II-IV acute GVHD at days +100 and +180 were 36% (95%CI: 19-53%) and 46% (95%CI: 27-64%), respectively. The CI of chronic GVHD at 1 and 2 years was 20% (95%CI: 7-38%) and 29% (95%CI: 12-49%), respectively. The CI of progression of disease and non-relapse mortality at 2 years were 32% (95%CI: 15-51%) and 19% (95%CI: 7-37%), respectively. With a median follow-up in survivors of 40 months (range 2-80 months), the 2-year OS and PFS are 64% (95%CI: 47-86%) and 49% (95%CI: 32-73%), respectively. In vivo T cell-depletion with alemtuzumab was associated with a markedly reduced 2-year PFS (0% vs 64%, P=0.008; Figure 1 ). Conversely, a sIPI at transplantation <1 was associated with a much improved 2-year PFS compared to sIPI >1 (65% vs 13%, P=0.021; Figure 2 Introduction: Thiotepa is an alkylating agent approved for highdose therapy for autologous HSCT. Because of its excellent capacity to cross the blood-brain barrier, it is regularly used for autoHSCT for primary CNS lymphoma (PCNSL). However, although thiotepa-based myeloablation might have benefi ts over traditional BEAM preparation because of better brain availability and less pulmonary toxicity, clinical level information about thiotepabased autoHSCT outside of the PCNSL fi eld is sparse. The purpose of the present retrospective study was to provide information on the potential risks and benefi ts of thiotepa-based preparative regimens in autoHSCT for distinct subtypes of lymphoma outside of the PCNSL setting. Materials (or patients) and Methods: PRIMARY OBJECTIVE was to compare the outcome of thiotepa-based autoHSCT (TT) with that of BEAM autoHSCT (BEAM) separately for diff use large B-cell lymphoma (DLBCL; excluding PCNSL), follicular lymphoma (FL), peripheral T Cell lymphoma (PTCL) and Hodgkin's lymphoma (HL). PRIMARY ENDPOINT was progression-free survival (PFS), secondary endpoints were overall survival (OS), non-relapse mortality (NRM), and incidence of relapse (IR Introduction: We hypothesized that combination of dose-adjusted (DA)-EPOCH +/-R followed by ASCT as a consolidation in the fi rstline therapy provide high effi cacy with acceptable toxicity in patients with poor-risk, aggressive NHL. Materials (or patients) and Methods: From May 2007 to December 2012, a total of 23 patients with poor-risk, aggressive NHL defi ned as IPI≥3 or T-cell NHL (except ALK+) or Ki-67+≥90%, were planned to receive total of six to eight cycles of DA-EPOCH+/-R followed by consolidation with ASCT as a part of fi rst-line therapy. In DA-EPOCH doxorubicin, vincristine and etoposide are infused over 96 hours, cyclophosphamide and prednisone are administered on a bolus schedule, and doxorubicin, etoposide and cyclophosphamide are pharmacodynamically dose-adjusted based on the neutrophil nadir. On sixth day all patients received pegylated granulocyte colony-stimulating factor (peg G-CSF, Neu-lastim®). Criteria needed for transplant eligibility were: age <65, Ki-67+≥90% or IPI≥3 or T-cell NHL (except ALK+), and absence of transplant-limiting comorbidities. Myeloablation with BEAM regimen was performed in all patients. Three patients discontinued DA-EPOCH+/-R due to prolonged cytopenias, fatal reactivation of hepatitis B, and patient refusal, respectively. The remaining 20 patients proceeded to ASCT. Ten (50%) were males, median age was 42,5 years (range 21-58). There were 11 (55%) patients with DLBCL, fi ve had T-cell NHL (3 ALCL/2 ALK-/ and 2 T-NOS), two had mantle cell lymphoma, and two had Burkitt lymphoma. A total of 14 patients (70%) had increased LDH, 15 (75%) were Ann Arbor III/IV, and 13 (65%) patients had IPI≥2. Prior to ASCT, 15 (75%) patients were in complete remission, fi ve in partial remission. Results: After median follow up of 36 months, 3-year OS and PFS were 73% and 75%, respectively (Table 1) . According to original protocol, median dose escalation was three times per patient (172%). In 15 (75%) patients, last cycle of DA-EPOCH was successfully used for stem cell mobilization. During treatment and follow-up period there was no unexpected toxicities, cardiovascular complications or, although too early, secondary malignancies. There was no transplantation-related mortality. Discussion: CHOP+/-R is not optimal therapy for patients with poor-risk, aggressive NHL, especially for subgroup of younger patients. Eff orts to improve outcome with dose-dense regimens showed similar results, whereas dose-intensifi cation strategies showed improvement at the cost of higher toxicity. DA-EPOCH, due to its continuous low-dose drug exposure and dose modification according to patients' hematopoietic capacity, allows us to apply highest acceptable doses of drugs. ASCT is not established as a part of fi rst-line therapy, however, therapeutic benefi t was observed in high-risk group of patients in some studies. Our results showed that ASCT as a consolidation after DA-EPOCH +/-R is acceptable and effi cacious therapeutic option in the treatment of patients with poor-risk, aggressive NHL, without signifi cant additional toxicities. Disclosure of Interest: None Declared.

Introduction: Patients with relapsed and refractory Hodgkin's lymphoma (HL) unable to undergo high dose chemotherapy and autologous stem cell transplant (HDC ASCT) have a reported survival of 15-20%. Those failing to respond to fi rst-line salvage chemotherapy (1 st -salvage) may be off ered second line salvage chemotherapy (2 nd -salvage) and may receive HDC ASCT. This study aimed to determine the outcome of HL patients who failed to respond to 1st-salvage and received 2nd-salvage prior to HDC ASCT.

Materials (or patients) and Methods: We identifi ed 31 patients who required 2 nd -salvage prior to ASCT from 1996 to 2012. Patient were grouped as (1) relapsed-refractory, defi ned as patients with prior CR and on relapse found refractory to 1 st -salvage and required 2 nd -salvage and (2) refractory-refractory, defi ned as patients refractory to primary treatment and also refractory to 1 st -salvage and given 2 nd -salvage. Kaplan-Meier (KM) method was used for overall (OS) and progression free survival (PFS). Results: 278 patients with HL received HDC ASCT. Of these, 31 patients received both 1 st -salvage and 2 nd -salvage prior to HDC ASCT. Male:Female 18:13, median age at HDC ASCT was 22 years (14-44 years). Response to 2 nd -salvage was complete response (CR) 16%, partial response (PR) 71%, and stable disease 13%. After HDC ASCT, CR: PR: progressive disease was observed in 16(52%): 4(12%): 9(29%) patients respectively. Overall response (CR+PR) in 20 patients (65%). 15 patients progressed within 12 months ASCT. 11/31 patients died; (3/11 died within 12 months, 8/11 from 12 to 60 months). Five years OS for entire cohort: relapsed-refractory: refractory-refractory groups is 57: 73: 56% (P=0.5) while PFS is 51: 73: 40% (P=0.1) respectively, P-value not signifi cant due to small sample size. Both median OS and PFS have not yet been reached in any subgroup except refractory-refractory group whose median PFS is 6 months. Discussion: Though long-term outcome of patients requiring two salvage regimes was inferior to patients requiring one salvage regimen, it appears better than literature reports with chemotherapy without ASCT. We conclude that HDC ASCT is valid approach for HL patients even when more than one salvage chemotherapy regimens is required for tumor reduction. relapsed-refractory seems to have better outcome than refractory-refractory in our study. Importantly, using this didn't lead to excessive toxicity such as delayed engraftment or chemotherapy related toxic deaths. Disclosure of Interest: None Declared.

Introduction: To examine the role of hematopoietic stem cell transplantation (HSCT) for pediatric patients ≤ 18 years old at the diagnosis with refractory or recurrent lymphomas as a multicenter survey in Turkey.

Materials (or patients) and Methods: To evaluate the role of HSCT we retrospectively analyzed the outcome of 122 patients from 14 centers in Turkey who underwent autologous (n=90) or allogeneic (n=32) HSCT because of refractory or recurrent lymphomas. Phase II studies suggest to consolidate response achieved after front-line treatment with autologous stem cell transplant (SCT) and the role of fi rst-line allogeneic SCT is currently under investigation. However, no comparative data is available in this setting. Materials (or patients) and Methods: We retrospectively evaluate the impact of front-line SCT consolidation in 209 patients treated from January 1990 to December 2012 at our Center. Patients treated with a palliative intent and primary cutaneous T-cell lymphomas were excluded from the study. Histologic revision was performed in all cases. Specifi c diagnosis were PTCL not otherwise specifi ed (32%), anaplastic large-cell lymphoma (ALCL) anaplastic lymphoma kinase (ALK) positive (34%) and negative (17%), angioimmunoblastic T-cell lymphoma (10%), enteropathy-associated T-cell lymphoma (5%) and others (2%). Results: Median age was 49 years (range 15-85) with a prevalence of male sex (61%), advanced stage (68%) while IPI was >2 in 44%. Primary treatment were MACOP-B (39%) CHO(E)P (39%), intensive regimens (18%) or others (4%). Complete response to primary treatment (i.e. before SCT consolidation) was 60% (5% partial remission). Outcome of primary responders was good, with a 3-year overall survival of 74% (82% in ALCL ALK+ and 69% for the other histologies). By multivariate analysis a better overall survival was signifi cantly associated with IPI<2 (P=0.001), primary response (P=0.000), and ALCL ALK+ histology (P=0.012). We then focus on the 126 responding patients that either received (32%) or not (68%) a SCT consolidation. With the exception of age, the clinical characteristic of those patient were similar. In patients receiving SCT the stem cell source was mostly autologous (N=41/44). The multivariate analysis performed on responders, showed that only IPI was predictive of a better survival (P=0.006) while ALCL ALK+ histology (P=0.083) and performing SCT (P=0.303) were not. Discussion: Response to primary treatment rather than post-remissional programs is the crucial determinant of PTCL outcome. Disclosure of Interest: None Declared. Discussion: FT is a promising preparative regimen for allogeneic SCT in pts with lymphoid malignancies. Dose intensity is important in lymphoid malignancies. FT and FM are both dose intensive regimens as refl ected by lower relapse rates than FB2. FT is associated with lower NRM than FM, in similarity to the favorable profi le of FB2. These eff ects result in a more favorable OS with FT, in particular in advanced disease. A subset of pts with refractory disease could also be salvaged. Introduction: Evidence from the DLBCL setting suggests that allogeneic stem cell transplantation (alloSCT) is a reasonable salvage therapy for patients progressing despite undergoing autologous SCT (autoSCT and even as an alternative to autoSCT. Data regarding the eff ects of alloSCT in patients with primary mediastinal B cell lymphoma (PMBCL) are very limited. The objective of the current study was to investigate the outcome of alloSCT in patients with PMBCL who were previously treated with rituximab-based regimens and autoSCT. Materials (or patients) and Methods: All EBMT registered patients diagnosed with PMBCL who had undergone alloSCT between 2001 and 2011 as second transplant after a previous autoSCT were eligible. Patient characteristics, disease-and transplant-related data were collected from MED-A forms. Centers with potentially eligible patients were contacted to provide additional treatment and follow-up information, including a written histopathology report. Results: 12 patients, 5 females (42%) and 7 males with confi rmed PMBCL, were included. The median age was 27 years, ranging from 21-to 61 years. Median time from diagnosis to alloSCT and from autoSCT to alloSCT were 14 (IQR 11-17 months and 7 months (IQR 4-9) respectively. Fifty-eight percent (N=7) had received at least 3 chemotherapeutic regimens prior allograft; 11 (92%) had been treated with rituximab-containing regimen and 67% (N=8) received mediastinal irradiation prior allosct. Evaluation of disease status at transplantation revealed that 42% (n=5) were in CR1/PR1 (alloSCT was performed here as part of a tandem transplant program), 17% (n=2) were in CR/PR>1 and 42% (n=5) were transplanted with refractory disease. 50% (n=6) were transplanted with residual mediastinal mass larger than 5 cm. Within a median follow up of 36 months, only 3 patients remained alive, whereas 9 patients died due to disease progression (n=3) or non-relapse mortality (n=6). Discussion: The current preliminary study does not suggest that alloSCT for PMBCL in the rituximab era can benefi t a large proportion of patients. Larger studies are warranted to further defi ne the role of allograft in this specifi c indication. Disclosure of Interest: None Declared. Introduction: Enteropathy-associated T-cell lymphoma (EATL) is a rare T-cell Non-Hodgkin Lymphoma. Based on clinical presentation EATL can be divided into two subgroups; EATL can arise in celiac disease (CD) patients without a known history of celiac disease (EATL de novo) or EATL manifests in adult patients with previously diagnosed (refractory) celiac disease who clinically deteriorate (secondary EATL). Currently, there are no standardized treatment protocols and prognosis of EATL remains poor. We evaluate extended follow-up of EATL patients referred to our tertiary celiac disease center (1999) (2000) (2001) (2002) (2003) (2004) (2005) (2006) (2007) (2008) (2009) (2010) (2011) (2012) (2013) . Materials (or patients) and Methods: A total of 61 patients with EATL were included. The diagnosis of EATL was established according to the WHO. We evaluated treatment strategies and overall survival (OS). Treatment response was assessed as complete remission (CR), stable disease (SD) or progressive disease (PD). Results: In 31/61 patients (51%) EATL was diagnosed in patients without previous history of (refractory) celiac disease (EATL de novo). The remaining 30/61 (49%) patients suff ered from secondary EATL (RCDII n=19. CD n=11). Nineteen patients (31%) were treated with chemotherapy combined with surgery. Treatment consisted of systemic chemotherapy in 12 (20%), surgery in 12 (20%), chemotherapy and resection with autologous stem-cell transplantation (auSCT) in 5 (8%) and chemotherapy and resection with allogenic stem-cell transplantation (alloSCT) in 2 (3%). CR was achieved in 23 patients (38%), SD in 5 patients (8%), and PD in 33 patients (54%). CR was mainly achieved after SCT therapy.

Overall the relapse rate in CR patients was 52% with a median follow-up of 13 months after therapy (range 1-54 months). Patients treated with chemotherapy and resection with auSCT showed a relapse rate of 40%. With a median survival of 6 months (range 0 -142 months), 50/61 patients died (82%). One, three-and fi veyears OS was 37 %, 16 % and 10 % respectively. Patients who were treated with chemotherapy and resection with auSCT showed a median survival of 15 months. Patients with EATL de novo showed better survival compared to patients with secondary EATL (P = 0.006).

Discussion: Best treatment response and the lowest relapse rate is achieved after intensive therapy (auSCT Introduction: Bendamustine has demonstrated effi cacy as single agent in several lymphoproliferative disorders, including Hodgkin's lymphoma (HL). Despite the wide use of this compound, alone or in combination, there are no published data regarding its mobilizing activity. In 2011, we started a phase II open-label prospective study with Bendamustine, Gemcitabine and Vinorelbine (BeGEV) to evaluate the effi cacy of this induction regimen before high dose chemotherapy plus autologus stem cell transplant (ASCT). One of the study objectives was to detect the role of Bendamustine as part of a mobilizing regimen for peripheral blood stem cell (PBSC) collection. Materials (or patients) and Methods: Between August 2011 and September 2013, 36 consecutive patients with relapsed/refrac-tory HL were enrolled in a Phase II open-label prospective study with BeGEV followed by ASCT. The treatment schedule was: Bendamustine (90mg/sqm, days 2-3), Gemcitabine (800mg/sqm, day 1 and 4) and Vinorelbine (25mg/sqm, day 1) plus G-CSF 10mcg/Kg beginning on day 7 continued daily until the target yield would be reached. PBSC collection was planned starting from cycle 1 or from cycle 3 in case of bone marrow involvement. Three million CD34/Kg were considered as the minimum cell dose established for a safety rescue. Other than successful rate of harvest, we evaluated the absolute number of collected CD34+ cells/Kg, the number of procedures performed per cycle, preleukapheresis circulating CD34+ cells/mcL, white blood cells (WBC) count and the day of fi rst collection. Adverse events were also recorded. All patients provided written informed consent at the time of study inclusion. Results: All patients were able to mobilize readily and all achieved the primary end point with at least 3.6 x 10>6 CD34+/Kg collected in a single patient. The median yield of CD 34+/Kg collected was 8.25 x 10>6 CD34+/Kg (range, 2.3-15) after a median of 1 procedure (range, 1-2). The median preleukapheresis circulating CD34+/mcL and WBC count/mcL were 94/mcL (range, 15-237) and 21750/mcL (range, 11200-87080), respectively. The median day of fi rst collection was 12 (range, [9] [10] [11] [12] [13] [14] [15] . Seventeen pts underwent leukapheresis at cycle 1, 10pts after cycle 2 and 9 after cycle 3 (due to logistic reasons in 18pts and to Cytomegalovirus reactivation in 1). Hematologic and non-hematologic side eff ects were acceptable and no toxic deaths occurred. One patient developed blood-pressure decrement during the apheresis, but she was able to complete the procedure without sequelae. To date, 21 patients (58%) underwent ASCT with prompt engraftment. Data about neutrophils and platelets engraftment will be presented in the fi nal analysis added to comparison with historical IGEV published data (Magagnoli et al, BMT 2007) . Discussion: In our knowledge this is the fi rst biggest prospective study evaluating Bendamustine as mobilizing agent in resistant HL pts before ASCT. These results confi rm that BeGEV regimen, combined with G-CSF support, can be successfully and safely used to mobilize PBSC. Disclosure of Interest: None Declared. Introduction: At diagnosis, a peripheral blood lymphocytes to monocytes ratio (LMR) lower than 2.6, identifi es a group of DLBCL patients with a poor prognosis when treated with R-CHOP (Rambaldi et al, ASH 2012) . No data are available for patients treated upfront with Rituximab containing high dose sequential S179 chemotherapy programs (R-HDS) and autologous stem cell transplantation (AST). Aim of this study was to investigate whether LMR ratio may identify a high-risk patient subgroup that benefi ts from a primary high-dose program with ASCT. Materials (or patients) and Methods: We analysed LMR ratio at diagnosis in a series of DLBCL patients enrolled into a trial comparing R-CHOP 14 with R-HDS associated with AST (R-HDS 0305, Clinical Trials.gov.number NCT00355199 by GITIL). Patients characteristics: DLBCL without CNS involvement, with an age between 18-60 years and an High IPI (stage > II B-bulk with ECOG-PS=0-3 and age adjusted IPI (aaIPI) 2-3 or age 61-65 years with ECOG-PS = 0-2 and IPI > 3). R-CHOP 14 (8 cycles) or R-HDS regimen and AST were carried out as previously reported ( Introduction: Allogeneic hematopoietic stem cell transplantation (AlloSCT) represents a potential curative option for relapse or refractory lymphoproliferative disorders, based in part on "graft versus lymphoma eff ect" (GVLE). However the role of GVLE enhancement, and the adequate way to implement it in relapse or persistent disease after AlloSCT remains unclear. Our goal was to evaluate how the GVLE enhancement, through tapering immunosuppressive treatment (IST) and/or DLI, is able to control the disease in patients with lymphoid malignancies who relapse after AlloSCT or have disease persistence at day +100. Materials (or patients) and Methods: Twenty-six patients with post-AlloSCT relapse or persistent disease were retrospective analyzed from a series of 112 patients who consecutively underwent alloSCT in our centre. Results: In 19/26 (73%) patients, GVLE was enhanced by tapering immunosuppressive treatment (IST) and/or donor lymphocyte infusion (DLI), achieving response in 13 (68%), with 11 complete remissions (CR). Median of days from immune-manipulation to response was 105 (39-343), and 9/19 (47%) of patients were alive and disease-free at last follow-up. Concerning histologies, immune-mediated response was observed in 6/6 patients with non-Hodgkin lymphoma (NHL), 5/7 with chronic lymphoid leukemia (CLL) and 2/6 with Hodgkin lymphoma (HL). Graft versus host disease appeared in 16/19 (84%) with a median of 36 days (-3 to 114) from IST withdrawal or DLI; only one death due to GVHD was observed. The remaining 7 patients in whom GVLE enhancement was not possible due to active or previous GVHD received conventional chemotherapy +/-radiotherapy, achieving 2 CR and 1 PR. Among them, 2 patients were alive in CR at last follow-up. With a median follow up 56 months (range 11-138), estimated four years progression free survival (PFS) and overall survival (OS) for the whole series of 26 patients were 32% and 45% respectively. In those patients who achieved CR after GVLE appeared, 4 years PFS and OS were 40% and 50% respectively. Main cause of death whatever the histologic diagnosis or treatment group was progression or relapse. On the univariate analysis the development of cGVHD after immune manipulation (4 years PFS 32% vs 0%; P=0.017), severe cGVHD (4 years PFS 51% vs 25%; P=0.05) and diagnosis different of HL (4 years PFS 35% vs 17% months; P=0.016) were the variables with signifi cant infl uence on PFS between patients with immune manipulation. Discussion: Disease response to immune manipulation proves the existence of GVLE, which appear in all histological subtypes, although it seems less frequent in HL. Trough GVLE enhancement we are able to achieve maintained responses, achieving PFS and OS similar to those described in non-relapsing patients. Therefore, immune approach should always be considered as treatment in relapse/refractory patients. Disclosure of Interest: None Declared. Introduction: This report is a retrospective analysis of patients with Hodgkin lymphoma (HL) who relapsed or progressed after fi rst autologous stem cell transplant (autoSCT). The aim of the study was to determine the response rate to conventional salvage chemo-and/or radiotherapy and prognostic factors infl uencing treatment outcome. We also intended to analyze overall survival (OS), and progression-free survival (PFS) for patients treated with conventional salvage therapy alone, and for those who proceeded to second autoSCT or to allogeneic stem cell transplant (alloSCT). Materials (or patients) and Methods: The study group consisted of 103 patients who relapsed at the median of 9 months (1-138) after autoSCT. The median age of patients was 30 years (20-68) at relapse. Four patients died due to rapid disease progression before salvage treatment. Ninety nine patients received conventional salvage chemotherapy +/-radiotherapy. Following salvage therapy 29 patients proceeded to allogeneic and 24 to second autoSCT. Results: The overall (complete or partial) response rate to salvage therapy was 64%. On multivariate logistic regression the variables that were associated with lower response rate were age >= 40 years (OR 13,3; 95% 1.9-90.9; P=0.008), and clinical stage 4 at relapse (OR 8.2; 95% 2.4-28.1; P=0.001). After the median followup time of 36 months , the 3-year OS and PFS was 44% and 26%, respectively. Within a group of patients with complete or partial response (PR) after conventional salvage therapy, those patients who proceeded to second autoSCT showed signifi cantly S180 better OS than those who proceeded to alloSCT or who did not undergo second transplant. The respective 3-year OS was 72%, 45% and 58% (P=0.006). The median survival time for patients who underwent second autoSCT was not reached in comparison to 23 months (95% CI 3-44) for patients who proceeded to alloSCT, and 49 months (95% CI 34-63) for those who did not undergo second transplant. Within the group of patients with response worse than PR after salvage therapy the respective median survival time was 41 months (95% CI 16-66), 17 months (95% CI 12-23) and 11 months (95% CI 7-15) (P=0.006). In univariate analysis age (>= 40 years vs < 40 years), B symptoms, and bulky disease at relapse (>= 5cm vs < 5 cm), time to progression after fi rst autoSCT (=< 12 vs > 12 months), the best response to conventional salvage therapy (worse than PR vs at least PR), and second autoSCT were significant for OS (P < 0.05). In the multivariate analysis only response to salvage therapy (HR 4.0; 95% CI 2.3-7.0; P<0.001), second autoSCT (HR 0.4; 95% CI 0.2-0.8; P=0.01) and time to progression after auto-SCT (HR 1.8; 95% CI 1.0-3.2; P=0.051) remained signifi cant for OS. Discussion: Our results indicate that age and clinical stage are independent predictors of response to salvage chemo-and/or radiotherapy for patients with HL who relapsed after fi rst auto-SCT. Second autoSCT following salvage therapy seems to provide longer survival than salvage therapy alone or followed by alloSCT. The outcome of second autoSCT for patients with at least partial response after salvage therapy is satisfactory. The treatment strategy for patients who did not respond to conventional salvage regimens remains an area for further studies. Disclosure of Interest: None Declared. Introduction: Relapsed and refractory Hodgkin's lymphoma (HL) patients may experience long-term survival after allogeneic transplant (alloSCT), but disease recurrence represents the main cause of treatment failure. PET (positron-emission tomography) -positive patients after alloSCT have a dismal outcome. Serum TARC (thymus and activation-regulated chemokine) is produced by Reed-Sternberg cells and may be a marker of disease. Our study had the following aims: i) to assess whether serum TARC levels were correlated to disease status; ii) to correlate TARC levels with PET results after alloSCT; iii) to evaluate whether the combined results could increase the ability to assess or predict relapse. Materials (or patients) and Methods: Twenty-two patients (M/ F=14/8) were evaluated in a prospective observational study. The median age was 31,5 (range, 18-45). Twenty patients had nodular sclerosis HL. Disease status before alloSCT was complete remission n=8, partial remission n=9, stable/progressive disease n=5. Patients received reduced intensity conditioning regimen and peripheral stem cells from a sibling (matched n=7, haploidentical n=2) or unrelated (n=13) donor. Median follow-up was 28 months (range, 2-48). TARC was assessed before and after alloSCT with a median time interval of 47 days (range, 7-700). PET was performed every 3-6 months after alloSCT. Results: Before alloSCT, the median TARC level was 720,6 pg/mL (range, 208,6-1332) in PET-negative patients, and 2542,5 pg/mL (range 94-13870) in PET-positive patients. After alloSCT, TARC was 1218 pg/mL (range, 31-4388) in persistently PET-negative patients compared to 22397 pg/mL (range, 602-106578) in PETpositive ones (P < 0.0001). In 6 patients who relapsed after allo-SCT TARC increased progressively before PET became positive, with a median fold increase of 2,93 (range, 1,5-7,11) at the time of relapse. Considering TARC values done on the day of PET scans, the ROC curve showed that the cut-off value of 1726 pg/mL had a sensitivity of 93% and specifi city of 84%. To address this question, we compared patients with MCL who have received rituximab maintenance treatment after autoSCT with those who were followed after autoSCT without any further treatment until progression in a single-centre retrospective study. Materials (or patients) and Methods: Eligible for this analysis were all patients who underwent autoSCT for MCL at our institution between 2000 and 2012. Patients with MCL who received rituximab maintenance therapy after autoSCT within a prospective phase II trail on the value of rituximab maintenance in B cell lymphoma (every 3 months for 2 years, NCT number 01933711) were compared with patients who were transplanted during the same time period within or outside the aforementioned trial but did not receive rituximab maintenance therapy. The impact of maintenance therapy on PFS and OS was analysed by multivariate cox regression. Rituximab maintenance was considered as a time-dependent event.

Results: A total of 72 patients met the inclusion criteria. Median age was 60 years (30-74). Prior to autoSCT all patients had been exposed to rituximab-based induction and/or salvage therapy, and 45 patients had been treated with high dose cytarabine (HD-ARA-C). AutoSCT was performed after fi rst line treatment in 51 patients. A complete response (CR) before autoSCT was achieved by 27 patients. Twenty-two patients from the phase-2 trial were randomized to receive post-transplant rituximab maintenance for two years, whilst the 50 remaining patients were followed with observation only. Patients with and without rituximab maintenance therapy did not signifi cantly diff er with regard to age, upfront autoSCT, remission status and HD-ARA-C exposure prior to autoSCT. However, there was a trend that patients who received rituximab maintenance therapy were transplanted more recently (control group: 2000-2012, rituximab maintenance group: 2002-2012, P=0.06). Median observation time after autoSCT was 56 months. Two-year PFS and OS from autoSCT of the control group were 65% and 84%, respectively, as compared to 90% and 90%, respectively, of the rituximab maintenance group. By univariate landmark analysis rituximab maintenance therapy was associated with signifi cantly better PFS (HR 0.21, P=0.014) but so far not OS. Multivariate adjustment for age, year of transplant, achievement of CR prior autoSCT, upfront autoSCT and high dose ARA-C treatment confi rmed the benefi cial impact of rituximab maintenance therapy (P=0.02 HR 0.23). Discussion: Our observation that rituximab maintenance therapy improves PFS in MCL patients after autoSCT extends the fi ndings of a recent report that showed a benefi t for rituximab maintenance after R-CHOP in elderly MCL patients and provide a rationale for studying maintenance approaches for eligible MCL patients after autoSCT in a comparative prospective trial. Disclosure of Interest: None Declared. Introduction: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) has been used as a curative option for adult T-cell leukemia/lymphoma (ATLL). However the suitable pre-transplant prognostic index is not well known. Here we intend to clarify the relation between pre-transplant prognostic index and day 100 overall survival (OS) and OS after allo-HSCT in our single institute located in an endemic area of ATLL. Materials (or patients) and Methods: There were 87 ATLL patients undergone allo-HSCT at Imamura Bun-in Hospital from June 1998 to July 2013. We analyzed the relation between pre-transplant prognostic indices such as HCT-CI, EBMTscore, the prognostic index for acute-and lymphoma-type adult T-cell leukemia/lymphoma (ATL-PI) and day100 OS and OS after allo-HSCT retrospectively. OS was analyzed with Kaplan-Meier method. Cox regression analysis about OS was also performed. Statistical signifi cance defi ned P<0.05 in log-rank test. All statistical analysis was performed with EZR (R commander). Results: In 87 patients (52 male, 35 female), median age was 52 (range: 32-69) years. Performance status (PS) at HSCT was 0-1 in 66 patients and 2 in 11 patients. Disease status at HSCT was 28 complete remission (CR) and 59 non-CR (15 partial remission (PR), 15 stable disease (SD), 29 progressive disease (PD)). Fiftyone patients received BMT, 20 PBSCT, and 16 CBT, respectively. Fifty-fi ve patients had received myeloablative conditioning (MAC) and 32 reduced intensity conditioning (RIC). HCT-CI was 0 in 19 patients, 1 and 2 in 55 patients, and over 2 in 13 patients, respectively. EBMT score was 1 and 2 in 4 patients, 3 in 20 patients, 4 in 34 patients and over 4 in 29 patients, respectively. ATL-PI at HSCT was 0 and 1 in 23 patients, 2 in 42 patients, 3 in 13 patients and over 3 in 9 patients, respectively. In univariate analysis, HCT-CI ≥3 (P<0.01), EBMT score ≥4 (P<0.01), and ATL-PI at HSCT ≥3 (P<0.01) contributed to inferior OS. EBMT score ≥4 (P<0.01) also contributed to inferior day 100 OS, however, HCT-CI ≥3 (P=0.31) and ATL-PI at HSCT ≥3 (P=0.12) did not contribute to inferior day 100 OS. In multivariate analysis, EBMT score ≥4 contributed to inferior day 100 OS and OS (hazard ratio :2.11 and 8.76 respectively). Discussion: Among pre-transplant prognostic indices, EBMT score ≥4 only contributed to inferior day 100 OS and OS after HSCT in ATLL patients. The fact that 1/3 of mortality happened within 100 days after HSCT takes into consideration, our results suggest that EBMT score would be most important pre-transplant prognostic index when planning to undergo allo-HSCT for ATLL patients. Disclosure of Interest: None Declared. (CR1), 25% were in CR 2 and 29% had active disease; 40% received a full intensity conditioning and 60% got reduced intensity one. As cell source, 35% were bone marrow, 53% peripheral blood and 12% cord blood cells. Donors were related in 53% of the cases (45% were 10/10 HLA matched) and unrelated in 47% of cases (20% were 10/10 HLA matched). MFC was performed using BM samples with a sensitivity of 0.01%. Chimerism analysis was performed on marrow and/or blood samples using PCR with an accuracy of ± 5%, a mixed chimerism was defi ned by having 5% or more of recipient cells. Results: After transplantation, all patients engrafted, the cumulative incidence of acute GVHD at 3 months was 19.9% (95% CI: 16.2-20.6) while the cumulative incidence of chronic GVHD reached 26.7% (95% CI: 22.9-30.5) at 1 year. After a median follow-up of 16 months (range: 3-77), the median OS was 66 months (65-NR) with a 3 years probability of 64% (95% CI: 56-73), the median PFS was 32 months (13-NR) with a 3 years probability of 50% (95% CI: 37-58) while the transplant related mortality rate reached 13.6% (95% CI: 10-16) at 2 years. The 3 months chimerism evaluation (n=137) showed a mixed chimerism in 12 (9%) Introduction: In young patients with high risk myelodysplastic sindrome (MDS) or high risk acute myeloid leukemias (AML), allogeneic stem cell transplantation (HSCT) is the only potentially curative therapeutic strategy. In patients who are candidate to undergo to this procedure, disease relapse represents the major reason of failure. The current options for relapsed disease after allogeneic HSCT are salvage chemotherapy, donor lymphocyte infusion (DLI) or a second allogeneic transplant. Unfortunately the majority of patients who relapsed after allogeneic HSCT shows a very short survival. DLI can induce a second remission for this subgroup of patients, but a severe graft versus host disease (GvHD) can occur, resulting in signifi cant mortality. In recent years, the monitoring of chimerism and minimal residual disease (MRD) by fl ow cytometry analysis facilitated the detection of residual burden of disease, even though the kinetics of disease recurrence are very rapid. Recent studies documented as Azacitidine, a nucleoside analog of cytidine, induces FOXP3 expression in naïve T-cells, which in turn induces a regulatory T-cell population that mitigates GvHD eff ects, while preserving a graft-versus-leukemia (GvL) eff ects. Here we report the results of the effi cacy of azacitidine in the setting of MDS or AML patients who showed a drop in donor chimerism and/or MRD positive after allogeneic HSCT to prevent relapse of disease. Materials (or patients) and Methods: We analyzed data from 10 patients (5 MDS and 5 AML) who received azacitidine 75 mg/m 2 every 28 (Vidaza, Celgene Corporation, Summit, NJ, USA) subcutaneously on days 1-7, every 28 days. A limited number of azacitidine cycles were administered (range 1-6) because of possible induction of GvHD. The median age of patients was 49 years (range 38-55). Three patients received HSCT from Cord Blood (mm 4/6), 3 from sibling donor, and 4 from MUD 8/8. Nine patients received myeloablative conditioning regimen, only one RIC received. Three patients were also treated with infusions of DLI after 3 cycles of azacitidine, 2 of them in partial response (mixed chimerism) and in absence of GvhD and one in complete remission and full donor chimerism, in absence of GvHD. Results: The PFS is 40% after azacitidine initial treatment with a maximum follow up of 30 months. Four patients died because of relapse of disease, 2 of them are actually in complete remission, MRD negative and full donor chimerism. Two patients are in mixed chimerism with MRD positive. We also compare data analyzed from an historic cohort of 12 patients (median age 46 years) aff ected by AML who received DLI infusions because of drop of chimerism, and/or appearance of positivity of MRD to delay the relapse. Six patients received HSCT from sibling donor and 6 from MUD source, 5 were treated with RIC regimen, while 7 myeloablative conditioning regimen received. Only two patients are alive with a follow up of 48 months form DLI infusion. Discussion: In conclusion these data demonstrated that MRD treatment with azacitidine seems to be eff ective to prevent or to delay the relapse of disease with an acceptable profi le of safety in patients with high risk MDS and high risk AML after allogeneic HSCT. These data provide the rationale for the combination of azacitidine and DLI, using more and continuous cycles to maintain the response. Disclosure of Interest: None Declared. Introduction: Recent studies suggest that immunophenotypic remission (IR) may be a relevant prognostic factor in patients with multiple myeloma (MM), however data after allografting are lacking. Materials (or patients) and Methods: At our center, between January 2000 and December 2011, 80 consecutive multiple myeloma patients underwent an allograft. Sixtynine/80, with a follow-up of at least 3 months were included in this study. Three of them were further excluded because of incomplete data or death before disease restaging. Thus, 66 patients, median age 54 years (35-66), were evaluated for IR compare to conventional complete remission (CR). Bone marrow aspirates had to contain at least 13000 cells/microL for fl ow cytometry studies with high-sensitivity immunophenotyping for IR investigation. Plasmacells quantifi cation was obtained by 4 to 6-colour staining with the following monoclonal antibodies: CD38, CD138, CD56, CD19, CD45, cyKappa, cyLambda. IR was defi ned as less than 0.01% monoclonal plasmacells in bone marrow aspirate detected by multiparameter fl ow cytometry, and CR according to standard criteria. The times of observation were censored on 01/10/2013. Results: Conditioning regimen was non-myeloablative in 55 patients, reduced intensity in 10 and myeloablative in 1 patient. Post-grafting immunosuppression consisted of cyclosporine with mycophenolate mofetil or methotrexate. Donors were HLA identical siblings in 58 patients and unrelated in 8. Only 1 patient received bone marrow as source of stem cells. Allograft was part of the fi rst line treatment in 35/66 (53%) patients. All patients included had adequate bone marrow samples for IR evalutation. After a median follow-up of 85 months (range 31-158), the incidence of acute and chronic graft-versus-host disease was 45% and 52%. Three-year treatment related mortality was 14% in the overall population (N=80) and 9.1% among patients who survived at least 3 months and included in the analysis (N=66). At followup, 24 patients achieved CR and IR (CR/IR group), 21 achieved IR but not CR because of persistence of urine/serum M-component (noCR/IR group), and 21 did not achieve either CR or IR (noCR/ noIR group). Median overall survival (OS) and event-free survival (EFS) in patients who achieved IR were 96 and 41 months versus 36 and 6 months in those who did not (P<0.001). In details, median OS and EFS were not reached and 59 months in the CR/IR group, 64 and 16 months in the noCR/IR, and 36 and 6 months in the noCR/noIR, respectively (P<0.001 for both EFS and OS). In univariate analysis, being in the CR/IR group was the only signifi cant predictor for prolonged OS and EFS (P<0.001). Of note, cumulative incidence of extra-medullary disease at fi rst disease relapse after the allograft was 4% in the CR/IR, 38% in the noCR/IR and 14% in the noCR/noIR groups (P<0.001) at 4 years. Discussion: The achievement of IR showed a signifi cant impact on clinical outcomes, also among patients not in CR. Discrepancies between IR and CR and a higher incidence of extramedullary relapse observed in the noCR/IR group, suggest that myeloma cells may escape immune control outside the bone marrow. In this group, imaging studies such as positron emission tomography may be indicated to allow detection of early relapse. Disclosure of Interest: None Declared. Introduction: Relapse represents the main cause of treatment failure in patients with acute myeloid leukemia(AML) undergoing allogeneic stem cell transplantation (allo-SCT). The identification of minimal residual disease (MRD) prior to allo-SCT strongly predicted the recurrence while results came from posttransplant monitoring did not show the same. Anyway, only few studies were reported after allo-SCT and MRD was evaluated in small cohorts of patients using different techniques as well as different timing of assessment, cut-offs and methods of analysis. In our study we investigated MRD before and after allo-SCT by multiparameter flow cytometry (MFC) and Wilm's tumor 1 levels (WT1). Our purpose was firstly to compare the diagnostic performance of MFC and WT1 in the identification of relapse, secondarily to evaluate the impact of MRD on survival when it was studied by both methods at the optimal cut-off values and timing of assessment. Materials (or patients) and Methods: Fresh BM samples from 42 pts (20 males, 47.6%;mean age 45 ys, SD: 14). were investigated before and one month after the allo-SCT by six-color MFC and RQ-PCR WT1 mRNA. MRD was evaluated in pts achieving a complete remission (CR). Thirty-two out of 42 pts were in their 1 st or 2 nd CR before the transplant whereas 40 pts were in CR after this procedure (2 0f 42 died for transplant related mortality). At least 250.000 events were acquired for MRD analysis by MFC while samples containing less of 1000 copies of ABL were evaluated as degraded and inadequate for analysis by WT1. Results: Area under curve of the ROC curves were analyzed . Only MRD evaluated after allo-SCT by both MFC and WT1 achieved a fair accuracy (AUC≥0.700) in predicting the relapse. Indeed, posttransplant was chosen as the optimal timing of assessment and 0.05% as well as 101.4 x 10 4 ABL copies were considered the most predictive cut-off s for MFC and WT1, respectively. Despite the greater sensitivity of MFC compared to WT1 (80.0% vs 75.0%), less specifi city (66.7% vs 87.5%) and positive predictive value (44.4% vs 66.7%) were showed.Pts with positive MRD by MFC had signifi cantly higher risk to relapse compared to pts with a negative test (disease free survival (DFS): 86% vs. 40%; Cox regression crude P = 0.013). Similarly, patients with positive MRD by WT1 showed a lower DFS compared to negative ones (DFS: 87% vs. 29%; Cox regression crude P = 0.003).No signifi cant diff erences were observed in overall survival. At multivariate analysis, post transplant positive MRD identifi ed by both MFC and WT1 was signifi cantly related to a shorter DFS after adjusting for age, gender and cytogenetics (P = 0.016 and P=0.017, respectively). Discussion: In this study, pre and post transplant MRD were evaluated by MFC and WT1. Although both methods showed a moderate sensitivity to identify the relapse, a lower specifi city characterized the MFC. These data confi rmed concerns recently reported on some leukemia -associated phenotypes (LAP ). These pitfalls should increase the risk of false -positive results. On the other hand, the choice of post-transplant as the most predictive timing of assessment may refl ect a major uniformity of BM status at this time. Finally, the strong impact of post-transplant positive MRD on DFS, independently of method used, should encourage a wider use of MRD in this setting. Disclosure of Interest: None Declared. Introduction: Recent results from our prospective multi-center study identifi ed patients with t (8; 21) AML as high-risk according to their MRD status after the second consolidation chemotherapy and allo-HSCT can benefi t this part of patients; however, the relapse rate was reported to be 22% even after allo-HSCT for those high-risk patients. In addition, allo-HSCT tended to lower the relapse rates of KIT-mutated patients. To date, the respective value of MRD and KIT mutation to further distinguish between patients with low and high risks of relapse in allo-HSCT setting has not been assessed.

Materials (or patients) and Methods: one hundred consecutive adult AML patients with t (8; 21) receiving allo-HSCT in CR were enrolled between January 2006 and June 2013. KIT mutation was screened at diagnosis in 80 patients. Serial MRD monitoring by RQ-PCR post HSCT was done for all patients. The impact of MRD monitoring and KIT mutation on transplant outcomes was assessed. Results: Achieving > 3 log reduction at 1, 2 or 3 month after HSCT in transcripts from diagnosis, is associated with a signifi cant lower cumulative incidence of relapse (CIR) (20% vs 34%, P=0.037; 12% vs 100%, P<.001; and 10% vs 43%, P<0.001, respectively) and higher probability of LFS at 2 or 3 month (67% vs 0%, P<.001; and 72% vs 0%, P<.001, respectively). KIT mutation at diagnosis, is associated with a higher relapse rate (32% vs 14% for KIT+ vs KIT-, P=0.022; 32% vs 14% vs 26% for KIT+ vs KIT-vs KIT unknown, P=0.074) while there was no statistically signifi cant diff erence in LFS with respect to KIT mutation (64% vs 50% for KIT+ vs KIT-, P=0.11). Furthermore, in multivariate analysis, MRD remained the sole prognostic factor for CIR. In addition, the serial monitoring as well as the combination of MRD and KIT also allowed identifi cation of relapse risk after HSCT. Discussion: this study was the fi rst one to evaluate the impact of MRD monitoring and KIT mutation and their relative importance on transplant outcomes. Results showed that MRD monitoring by RQ-PCR at regular early time points post HSCT had more prominent signifi cance in identifi cation of patients at high risk of relapse, as compared to KIT mutation among adult t (8; 21) AML after allo-HSCT and could now be incorporated in clinical trials to evaluate the role of risk directed prophylactic/preemptive therapy. Disclosure of Interest: None Declared. Introduction: Mixed-lineage-leukaemia (MLL) gene is a recurrent chromosome change in acute leukaemia. In a prospective, multi-centre cohort study conducted in 2012, we demonstrated that allo-HSCT could be a valuable treatment choice for patients with MLL+ acute leukaemia. To date, no studies have focused on the impact of monitoring MLL expression before and after HSCT. Therefore, we performed a prospective cohort study to evaluate the prognostic value of MLL gene expression for predicting relapse in patients with MLL-rearranged acute leukaemia following allogeneic haematopoietic stem cell transplantation (allo-HSCT). Materials (or patients) and Methods: From Oct 2007 to Oct 2012, forty consecutive patients diagnosed with MLL+ acute leukaemia undergoing allo-HSCT were enrolled. These patients were part of a multi-centre clinical trial, registered at www.chictr.org as # ChiCTR-ONC-12002739, whose clinical outcome was previously reported in 2013. Bone marrow (BM) MLL transcript levels were monitored serially by real-time quantitative polymerase chain reaction (RQ-PCR) at designated time points in 40 MLL-rearranged acute leukaemia patients who were treated with allo-HSCT. The patients were followed up for a median of 24.5 months (range: 12-60 months). A total of 236 BM samples were collected and analysed. Of these, 230 were concurrently monitored for MRD by FCM for leukaemia-associated aberrant immune phenotypes (LAIPs) and by RQ-PCR to assess WT1 gene expression. Detectable MLL expression at any level (MLL >0.0000%) was defi ned as MLL positive. MRD positivity (MRD+) was defi ned as being MLL positive at any time during the fi rst year post-HSCT, while the absence of MLL positivity was defi ned as MRD negativity (MRD-). Results: The 3-year cumulative incidence of relapse of patients who experienced MRD+ (n=9) post-HSCT was 93.5% (CI: 87%>100%), compared to 12.5% (CI: 5.6%>19.4%) for MRD-patients (n=31) (P<0.001). For these patient groups, the 3-year overall survival was 12.5% (CI: 0.8%>24.2%) and 77.8% (CI: 68.4%>87.2%), respectively (P<0.001), and the 3-year LFS was 0% and 72.2% (CI: 61.1%>83.3%) (P<0.001), respectively. MLL positivity was associated with a higher relapse rate (HR=18.643, 95% CI: 3.449-100.757, P=0.001) and lower DFS (HR=11.05, 95% CI: 3.169-38.533, P<0.001) and OS (HR=14.438, 95% CI: 3.638-57.297, P <0.001), as determined by Cox multivariate analysis. A good correlation was found between MLL and WT1 expression. MLL gene expression had a higher specifi city and sensitivity than WT1 or MRD monitored by FCM for predicting relapse in these MLL+ AL patients undergoing allo-HSCT. Discussion: Our data indicate that MLL expression is a valuable and essential marker for MRD monitoring in MLL+AL patients following HSCT. The MLL expression during follow-up is highly predictive of leukaemia relapse and has better specifi city and sensitivity than WT1 and MRD monitored by FCM. Although our results need to be confi rmed in a large-scale, prospective study, we consider the current qualitative assessment of MLL expression to be a very useful test for further risk stratifi cation to guide the prevention and management of early relapse in MLL+AL patients following HSCT. Further research should focus on suitable MRD monitoring-directed interventions after HSCT, which could improve the clinical outcomes of these patients. Disclosure of Interest: None Declared.

Introduction: Myelofi brosis is a rare bone marrow disorder characterised by increased megakaryocyte numbers leading to bone marrow fi brosis. It is a heterogeneous disorder with some patients requiring no treatment while others can have life threatening cytopaenias or even transformation to Acute Myeloid Leukaemia (AML). The optimum time for Haematopoietic Stem Cell Transplantation (HSCT) is not known. Here we report our single centre experience at Manchester Royal Infi rmary. Materials (or patients) and Methods: A retrospective case analysis was performed for 14 consecutive patients (8 male, 6 female) who had undergone allogeneic HSCT for Primary Myelofi brosis (n=7) or transformed from Primary Polycythaemia (n=4) or Essential Thrombocythaemia (n=3). Patients received an allogeneic marrow (n=2) or peripheral blood stem cells (n=12) from HLA-matched siblings (n=6), matched voluntary unrelated donor (VUD, n=6) and single mismatched VUD (n=2). These were conducted between July 2008 and November 2011. The median age was 52 years (range 24-64). At transplantation 9 patients had a DIPSS-Plus score of INT-2 or High and 5 had INT-1 (one patient had developed AML and another MDS RAEB-2). 6 out of 14 patients received a full intensity conditioning regimen, 1 patient had total body irradiation/Cyclophosphamide (CY) and 5 patients Busulphan (BU)/CY. 8 patients received a reduced-intensity conditioning regimen with Fludarabine (FLU)/BU. RIC with ALG (VUD, n=4 and Sibling, n=2) or alemtuzumab (mismatched VUD, n=2) were used. These patients also received cyclosporin. All other patients received cyclosporin and methotrexate as GvHD prophylaxis. Splenectomy was performed in 3 patients prior to HSCT. 12 patients had a Karnoskfy score of 80% or greater and 2 a HCT-CI score of 2 or greater. Results: The median time to neutrophil engraftment was 17 days (Range 11-27 days). Acute GvHD Grades 2 to 4 was observed in 8 patients. Thrombosis was the most common post transplant complication (n=4). 5 patients have died, 4 from infection and one from transformation to AML. 9 patients are still alive with a median survival of 30 months (range 13 to 68 months). 2 of these S185 patients have been alive for over 60 months since transplantation. Both had PMF and were transplanted over the age of 50 years with minimal splenomegaly, and a DIPPS-Plus of Introduction: Allogeneic hematopoietic stem cell transplantation (allo-SCT) is the only curative option for patients with myelodysplastic syndromes (MDS). The role of treatment prior allo-SCT and in particular induction chemotherapy is a matter of an ongoing debate, since it might on the one hand reduce the likelihood of relapse by remission induction, but is also potentially associated with several toxicities. Materials (or patients) and Methods: We compared the outcome of a group of 110 patients suff ering from MDS (n= 89) or secondary AML, (n=21) who had been transplanted between 2001 and 2012 at our center, in accordance with their pretransplant treatment strategies. Of those, 54 patients were directly transplanted (upfront group) without any prior therapy to transplantation including 74% of them receiving a sequential conditioning regimen using the so-called FLAMSA protocol. Of 44 patients having received induction chemotherapy prior to transplantation 20 were in complete remission at the time of transplantation, while 24 had still active disease. The chemotherapy group and the upfront group were well balanced with regard to disease parameters and transplant characteristics with the exception, that there were more therapy-related MDS in the upfront group. (16% vs. 5%; P=0.002) In addition, we also compared the results of 12 patients having received disease modifying drugs like Valproat or 5-Azazytidine prior to transplantation. Results: The 5-year overall survival (OS) of the entire group was 43%. The 5-year OS of the patients in the upfront group was significantly higher in comparison to the chemotherapy group (67 % vs. 43%; P=0,032). This inferior outcome of the chemotherapy group was mainly attributed to those patients who were refractory after induction chemotherapy (5-year OS 33%; P=0.006), while the OS of those patients who achieved CR (55%, P=0,462) was comparable with the upfront group. Relapse-free survival after 5 years was signifi cantly lower in the chemotherapy group in comparison to the upfront group (34% vs. 56%; P=0.034), which was related to a low RFS in the group of patients refractory to induction chemotherapy (5-year RFS 17%). The RFS of patients in CR after induction chemotherapy (55%) was similar to that of the upfront group. Non-relapse mortality (NRM) was 25% for the entire group with no diff erences between the diff erent subgroups. Discussion: Despite the limitations of retrospective analyses our data suggest that upfront transplantation, for example using the sequential FLAMSA approach, results in promising results and might be a relevant alternative for patients with advanced MDS. Disclosure of Interest: None Declared. Introduction: Allogeneic stem cell transplant is performed in the treatment of patients with both hypoplastic MDS as well as to S186 rescue those high risk MDS-AML patients rendered hypoplastic due to induction chemotherapy. Whether the risks of rejection, disease progression or relapse diff er in these patient groups and consequently the appropriate duration of immunosuppression post-transplant remains to be determined. Materials (or patients) and Methods: Retrospective data including diagnostic and pre-transplant histology and chimerism for all sequential patients with MDS/AML transplanted between 2008-2013 with a hypocellular bone marrow in pre-transplant bone marrow biopsies were analysed. The patients were divided into those who presented with a hypocellular marrow or had a cytogenetic abnormality associated with hypocellular MDS (13q14) (hypo MDS) and those who were normocellular/hypercellular on presentation, but became persistently hypocellular following treatment (acquired hypoplasia). Clinical information including dates of taper of immunosuppressive therapy (IST), GvHD, survival status and cause of death were also collected. Results: Thirty one patients (20 with hypo MDS, 11 with acquired hypoplasia) were transplanted (36 transplants, 1 st transplant n= 31, 2 nd transplant n=4, 3 rd transplant n=1 ), 34 were t-cell depleted reduced intensity conditioned (25 with Alemtuzumab, 9 with ATG), and 2 with ric haplo peripheral blood stem cells with post-transplant Cyclophosphamide. In the hypo MDS group mean cellularity at presentation was 19% (range 5-30%), falling to 12% pre-transplant (the exception was a patient with 13q deletion who had an original cellularity of 60%, falling to 5% post Azacytidine and pretransplant). There were 7 episodes of graft failure/non-engraftment, with no episodes in the acquired hypoplasia group (P=0.0756). The mean time to death or last follow up was 682 days (59-1692), with overall survival of 68.4%, and 1 episode of disease relapse.

In the acquired hypoplasia group mean cellularity at presentation was 70% (35-90%), falling to 15% (0-25%) pre-transplant. The mean follow up was 609 days , with overall survival of 46%. There were 4 (31%) episodes of disease relapse. In both groups the initiation of IST weaning was very variable (day 28-1852). During the fi rst month following transplants which resulted in non-engraftment or graft failure 73% of tests performed had cyclosporine levels <300ug/l or tacrolimus levels <10μg/l. Discussion: Our data suggest that whilst patients with hypoplastic MDS have very low rates of relapse post-transplant and excellent overall survival their rate of graft failure is high. In contrast for patients with acquired hypoplasia disease relapse limits overall survival. We suggest that hypoplastic MDS patients may benefi t from intensive (dose/duration) immune suppression akin to that in HSCT for aplastic anaemia. In contrast for patients with acquired hypoplasia, the higher risk of relapse may be attenuated by early withdrawal of immunosuppression. Disclosure of Interest: None Declared. Introduction: Myelodysplastic syndromes (MDS) are a heterogeneous group of clonal stem cell disorders associated with worsening cytopenias and a variable risk of progression to acute leukemia, these disorders also leading to reduced survival and a compromised quality of life, especially in transfusion-dependent patients. Allogeneic BMT is considered the only curative approach for patients with MDS, while bone marrow transplantation clearly has a role in the treatment of MDS, the decision to proceed to transplantation is not always easy and the optimal approach has not been clearly defi ned. Materials (or patients) and Methods: From February 2008 to December 2012, twenty three pts with a diagnosis of de novo MDS according to the WHO classifi cation (RA: 6, RAEB1:7, RAEB 2: 10) received an HSCT from an HLA identical sibling donor. The median age of the series is 38 years old (25-52), the sex ratio is 1,5. Twenty pts (86, 9%) were stable on their cytopenias, their percentage of blasts on the bone marrow and their transfusion requirements, three of them were in progression. All of them received only blood and platelets transfusions and none received others therapy as growth factors, hypomethylating agents or chemotherapy. The median duration of the disease before the transplant is 18 months . The conditioning regimen used was an association of Fludarabine (200mg⁄m 2 ) and Busilvex (12. 8 mg⁄Kg) in a daily injection, for 4days for twice. GVHD prophylaxis consists in the association of Ciclosporin (CSA) and Methotrexate (short course Seattle), All the patients received peripheral stem cell transplant with a median rate of CD34:5 X10 6 ⁄Kg (3, (2) (3) (4) (5) (6) (7) (8) (9) . Results: At the 15th of August 2013, the median follow up was 25 months . Median time of neutropenia is 10d (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) . All pts have both, blood and platelet needs (3 blood units⁄pt) and (2 platelet units⁄pt).Twenty two pts (95, 6%) are still alive, 20 pts in CR (86.9%) and two pts in relapse (8.6%). Transplant related mortality is about 4.3%. Two pts (9%) have an Acute GVHD (II-IV) and 15 pts (68.1%) developed a Chronic GVHD witch is extensive in 36% pts. The overall survival (OS) and the Disease free survival are respectively 89% and 86%. Discussion: According to our good results in TRM (4.3%), in relapse (8.6%) and OS 89% and DFS 86%, this conditioning regimen F-BU4 seems a good option in MDS. Disclosure of Interest: None Declared. Introduction: The only curative option for MDS is allogeneic HSCT. New therapeutic options for controlling MDS for example the use of demethylating agents such as azacytidine have been introduced during recent years. Various conditioning regimens have been used. At Karolinska University Hospital, fl udarabine + treosulphan (fl u+treo) as conditioning regimen has been used mainly in elderly patients while busulphan (po) + cyclophosphamide (bu+cy) has been used in younger patients. Fludarabine + busulphan (po; fl u+bu) has been used in patients >60 ys either with low risk disease or co-morbidities. The aim of this analysis was to analyze outcome using the current strategy for selecting conditioning regimen. Materials (or patients) and Methods: 53 patients with MDS were transplanted since 2009. The median age was 58 (36-68) ys. The IPSS risk scores were: INT-1 = 12; INT-2 = 16; CMML = 10; High = 5, MDS-AML = 7. Thirty patients were treated with azacytidine prior to HSCT, 20 received induction chemotherapy (of whom 9 also had received azacytidine), 4 patients got cytokines only (epo+G-CSF), while 7 patients were untreated before HSCT. Bu+cy was used in 20 patients; fl u+bu in 10; and fl u+treo in 23 patients, with ATG given to patients receiving unrelated donor grafts. The median age in the three groups was 50, 64, and 60 ys, respectively. The IPSS scores were comparable with the exception that no patient with MDS-AML received fl u+bu. 13 patients received grafts from sibling donors and 40 from unrelated donors. The stem cell source was bone marrow in two patients, PBSC in 49, and double cord blood units in two patients. Results: The estimated overall survival (OS) for the entire cohort at two years is 72% with a DFS of 62%. Nine patients have died from transplant related causes and 8 from disease relapse. The 2-year OS was 69% for bu+cy, 62% for fl u+bu, and 80% for fl u+treo. The corresponding 2-year DFS were 63% for bu+cy, 52% for fl u+bu, and 68% for fl u+treo. DFS was 66% for int-1, 74% for int-2, 38% for CMML, and 59% for high/AML. Discussion: We conclude that fl u+treo +/-ATG is an eff ective and safe preparative regimen for MDS comparable to bu+cy also in higher-risk patients. New strategies are needed to improve outcome of allogeneic HSCT in CMML. Disclosure of Interest: None Declared.

Introduction: The prognosis of myelodysplastic syndromes (MDS) is critically infl uenced by cytogenetic abnormalities, which can be associated with a higher risk of relapse after hematopoietic stem cell (HSC) transplantation. In this multicenter, retrospective study, we assessed the impact of the R-IPSS cytogenetic score on the outcome of MDS patients transplanted from HLA-identical siblings or HLA-matched unrelated donors. Introduction: To investigate the use and outcome of thiotepa in combination with fl udarabine or other drugs we screened 4852 patients in EBMT database and found 225 patients with a median age of 56 years (range 18 -71) who received thiotepa-based regimen. Materials (or patients) and Methods: Male/female distribution was138/87. Diagnoses were RA/RARS in 29 (16%), RCMD in 16 (9%), RAEB in 76 (41%), RAEB-t in 28 (15%), and sAML/sMDS in 35 (19%). At transplantation 51 (25%) were in CR1, 105 (51%) did not respond to chemotherapy (no CR), and 49 (24%) were transplanted without prior treatment.

Reduced intensity conditioning was used in 144 (64%) and MAC in 81 (36%) of the patients prior to allogeneic stem cell transplantation from HLA-identical sibling (39%), matched unrelated (54%), and mismatched related or unrelated (6%) donor or other relatives (1%). Bone marrow or peripheral blood were used in 35% and 65%, respectively. Results: The median time to leukocyte engraftment was 17 days (9 -46) . The non-relapse mortality at 1 year was 36% and CI of relapse at 3 years was 21%. The 3-year relapse-free and overall survival were 40% and 42%, respectively. We further analysed a comparison between thiotepa plus fl udarabine (n = 42) vs. thiotepa/fl udarabine plus other drugs such as busulfan, cyclophosphamide, or melphalan (n = 86). A non-significant (P = 0.63) trend for more NRM at 1 year (39% vs. 31%) and a trend for higher risk of relapse at 3 years (25% vs. 14%) was seen for the thiotepa/fl udarabine plus other drugs group resulting in a favorable trend in survival at 3 years for the thiotepa/fl udarabine combination (51% vs. 32%, P = 0.28).

Within the thiotepa group signifi cant factors for improved survival were age less than 60 years (P = 0.01), and RA/RARS vs. others (p < 0.001). Introduction: The only cure for the hematological abnormalities of Fanconi anemia (FA) remains allogeneic hematopoietic cell trans-plantation (HCT). Few reports are available on outcomes after HCT in FA patients (pts) with myeloid malignancies. We analyzed data of the outcome of 33 Japanese FA patients with myeloid malignancies. Materials (or patients) and Methods: Between 11/1991and 11/2013, 33 FA pts received HCT. These included 16 males and 17 females aged 1.1-37.4 (median age 11.2 years). Twenty-three pts had myelodysplastic syndrome (MDS) in refractory anemia (RA) (N=12), refractory anemia with excess blasts (RAEB) (N=11), while ten pts had acute myeloid leukemia (AML). All pts with RAEB or AML and 8 pts with RA had cytogenetic abnormalities involving: chromosome (Chr) 1(N=13), Chr 3 (N= 10) and Chr 7 (N=12), with 13pts having complex abnormalities. Donors were related for 16 pts: matched (N=7) or mismatched (N=9) and unrelated for 17 pts: matched (N=8) or mismatched (N=9). Twenty-five pts received bone marrow cells (BM), two received peripheral blood stem cells (PB), three received BM + PB cells, and three received cord blood grafts. Nine pts received radiation-cyclophosphamide (CY) based regimens, and 24 received fludarabine-based regimens. Three pts received cyclosporinebased GVHD prophylaxis, six received CD34 positive selection or T-cell depleted grafts, and 24 received tacrolimus-based GVHD prophylaxis.

Results: Twenty-eight pts had neutrophil recovery by day 28. Two pts received CD34 positive selected grafts had rejection, and three pts died of infection or bleeding by day 28. Two pts had secondary graft failure by one year after engraftment. With a median follow-up of 3.8 years (range 0.4-19.6), 21 pts are alive with leukemia free, and 12 pts died; causes of death were relapse (N=4), graft failure (N=1), infection (N=2), bleeding (N=1), multiple organ failure (N=1), lymph proliferative disorder after HCT (N=1), donor-type leukemia (N-1) and secondary cancer (N-1). Acute graft-versus-host disease (GVHD) of grades II or more developed in fi ve pts and chronic GVHD developed in 12 pts. Survival probabilities at 3 years of RA, RAEB and AML were 91.7% (95% CI, 76% to 100%), 80.8% (95% CI, 57% to 100%), and 26.7% (95% CI, 0% to 56%), respectively. Discussion: Our study indicates that long-term survival for FA patients with myeloid malignancies is achievable, and HCT may be necessary before developing of advanced MDS (RAEB) or AML. Cytogenetic abnormalities involving chromosome 1, 3, 7 and complex karyotype may be associated with increased risk of MDS and AML development for FA patients. Disclosure of Interest: None Declared. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is eff ective curative treatment of childhood MDS, but long-term survival depends on a lot of factors. Among them impact of acute graftversus-host disease (aGVHD) on overall survival is controversial.

To analyze the infl uence of aGVHD on the outcome of childhood MDS after allo-HSCT.

Materials (or patients) and Methods: Allo-HSCT were performed in 36 patients (pts) (19 boys; 17 girls) with following diseases: refractory cytopenia of childhood -9 pts (25%), refractory anemia with excess blasts -11 pts (30,5%), refractory anemia with excess blasts in transformation-14 pts (39%), juvenile myelomonocytic leukemia in 2 pts (5,5%). Cytogenetic analysis showed: normal karyotype in 8 pts; monosomy 7-11 pts; structurally compleх abnormalities of karyotype-6 pts; other aberrations-11 pts. The median of age was 10 years (1-19 years) . Unrelated allo-HSCT was done in 26 pts (72 %), related -in 6 pts (17 %), haplo-in 4 pts (11%). Myeloablative conditioning regimens (MAC) were used in 18 pts (50%); reduced-intensity conditioning (RIC) in 18 pts (50 %). MAC consisted of Busulfan (Bu) 16 mg/kg + Cyclophosphamide 120 mg/kg. RIC included Fludarabine (Flu) 150 mg/m 2 + Melphalan (Mel) 140 mg/m 2 , Flu 150 mg/m 2 + Bu 8mg/kg. The bone marrow (BM) was used in 22 pts (61,1%), peripheral blood stem cells (PBSC) in 12 pts (33,3%), combination of BM and PBSC in 2 pts (5,6%). Results: 5-year overall survival (OS) was 55%. OS after MAC allo-HSCT -64%, after RIC allo-HSCT -45% (P=0,25). Engraftment was on day + 19 (range 11-43). In group of patients who achieved engraftment (n=36) aGVHD I-III grade developed in 19 pts (55%) (gr I -6 pts,gr. II -4 pts.,gr.III-9pts). 12 pts had no signs of GVHD. 5-years OS in group of pts with grade I-III aGVHD (n=19) was 76% (MAC -72%, RIC -52%; P=0,214). 5-years OS in group of pts without signs of aGVHD (n=12) -20% (P=0,003). The main reason of pts mortality in aGVHD group -4 pts gr IV aGVHD, 2 pts infectious complications, 2 pts progression of disease, without aGVHD-3 pts infectious complications, 4 pts progression of disease. Discussion: Allogeneic HSCT -eff ective treatment for children and adolescence with MDS. Our data demonstrate that aGVHD reliably infl uence on OS of pts with childhood MDS and further studies should be conducted. Disclosure of Interest: None Declared. Introduction: Extramedullary hematopoiesis, as consequence of bone marrow microenvironment dysregulation, is a key feature of advanced stage disease in Philadelphia negative myeloproliferative neoplasms (Ph-neg MPN), and in particular of myelofi brosis.

In patients with myelofi brosis, extramedullary hematopoiesis occurs mainly in the spleen, where a microenvironment that provides a residence for circulating hematopoietic progenitor/stem cells (HSCs/HPCs) may be derived from endogenous splenic cells and/or from the mobilization of bone marrow mesenchymal stromal cells (MSCs). Materials (or patients) and Methods: In order to investigate Phneg MPN splenic microenvironment, we evaluated the possibility to in vitro isolate MSCs from the spleen of 23 patients with myelofi brosis undergoing therapeutic splenectomy for progressive splenomegaly and of 7 healthy donors (HDs) undergoing splenectomy for trauma surgery. Written informed consent was obtained from both patients and HD. Following the standard procedure for BM-MSC expansion, we were able to isolate and in vitro expand MSCs from the spleen of 9 patients with myelofibrosis (39%) and of 3 HDs (43%). MSCs were characterized for morphology, clonogenic effi ciency (CFU-F), proliferative capacity (population doubling, PD), immunophenotype (fl ow-cytometry), osteogenic and adipogenic diff erentiation potential (histological staining), and ability to reach senescence. Moreover, the capability to support hematopoiesis of HD-derived CD34+ cells by cocultures on feeder layers of irradiated spleen MSCs from both patients and HDs will be evaluated Results: Preliminary data suggest that: i. spleen MSCs from both patients and HDs show a morphology typical of aged cells, and they enter senescence phase at earlier passages (p) (median value: p4, range: p2-p10 and p4, range: p4-p10, respectively), in comparison with BM-MSC; ii. CFU-F number is higher in patients than in HDs (median: 0.07, range: 0.03-0.1 and median: 0.03, range. 0.03-0.04/10 6 seeded cells, respectively) showing that higher number of MSC precursors are present in myelofi brosis splenic tissues;

iii. patients MSC proliferative capacity is lower than that of HDs. Discussion: Experiments aimed to assess the diff erentiation potential and the support to in vitro hematopoiesis are in progress. How these characteristics of spleen-derived MSCs from patients may aff ect local microenvironment and support extramedullary hematopoiesis need further investigations. Disclosure of Interest: None Declared. Introduction: This study was designed to investigate the involvement of intracellular ROS in adipocyte hyperplasia induced by Ara-c. Materials (or patients) and Methods: C57BL/6J female mice (6-8 weeks, 20g, n=80) were divided into 4 groups. Ara-C group animals were administered 0.5 g/(kgμd) Ara-C (Sigma, USA) via intraperitoneal injection for four consecutive days to induce hematopoietic stress. NAC group animals were administered 0.1 g/(kgμd) N-acetyl-L-cysteine (NAC, Sigma, USA) via intraperitoneal injection for four consecutive days and 1g/(Lμd) NAC drinking water were given for 28 days. Ara-C+NAC group animals were given both reagents as described above. Control group animals were injected with the same volume PBS. Intracellular ROS levels were measured using an H2DCFDA probe (Molecular Probe, USA) 10mM. To observe the changes of adipocyte in bone marrow, tibias were collected and detected by histopathology once a week. PPARγ and adiponectin protein levels were assessed by western blotting and mRNA levels were assessed by qRT-PCR. Data were presented as mean± S.D. Statistical diff erences between two groups were evaluated by Student's t test. For multiple group comparisons, data were analyzed by one-way analysis of variance (ANOVA). Results: We found that adipocyte hyperplasia could be induced by Ara-C treatment. Compared to control group, the sinuses of Ara-C treated mice were widely dilated, hyperemic and composed of discontinuous endothelial cells. A signifi cant increase of adipocyte counts was also observed in the tibias of Ara-C treated mice. Moreover, the gene expression and protein levels of major adipogenic transcription factor PPARγ and its target gene adiponectin were signifi cantly increased. Next, we investigated whether ROS is involved in the adipogenesis induced by chemotherapy. The fl ow cytometry analysis on bone-marrow derived mesenchymal stem cells revealed that Ara-C was able to induce ROS generation with a signifi cant increase compared to control group whereas NAC reduced the production of ROS. In addition, adipogenesis in long bones following Ara-C treatment is successfully inhibited by NAC. Decreased numbers of adipocyte were observed and the expression of PPARγ and adiponectin was suppressed by treatment with NAC. Discussion: Recent studies reveal that adipocyte may play a negative role in hematopoiesis. However, the cause of adipocyte hyperplasia after chemotherapy remains unknown.

Mesenchymal stem cells (MSCs) could diff erentiate into adipocyte. ROS plays an important role during the early stage of adipocyte diff erentiation of MSCs in vitro. However, whether ROS involves in the adipogenesis induced by chemotherapy in vivo is unclear.

In the present study, we demonstrated that Ara-C treatment is able to induce ROS generation that in turn infl uences key factors involved in adipocyte diff erentiation of MSCs in vivo. We show that adipocyte hyperplasia is diminished in the presence of ROS scavenger NAC. The use of antioxidant may be a potential way to inhibit the generation of adipocyte induced by chemotherapy that in turn improves the hematopoietic recovery.

In addition, several reports demonstrate the role of ROS in other kinds of stem cells. Future studies will shed light to whether an increase in ROS is a general requirement for stem cells diff erentiation and the common mechanisms by which ROS initiate stem cell diff erentiation. Disclosure of Interest: None Declared. ). The set of up-or down-regulated genes between CD271-MSCs and PA-MSCs were hierarchically clustered and displayed in heatmap images using Multiple Experiment Viewer software. Reporters identifi ed in the discriminatory genes analysis were annotated with information from Gene Ontology (GO), which provides information on molecular function, as well as various pathway resources for information on involvement in biological signalling pathways. Diff erences between the sample group means of CD271-MSCs and PA-MSCs were assessed with Student's t-test (two-tailed, equal variance). Reporters were considered as diff erentially expressed when they passed the fi ltering criteria of an uncorrected P-value of 0.05 or less, and fold change diff erence of at least 1.5-fold up-or downregulation of the mean of the CD271-MSC sample group compared to the PA-MSC sample group. Results: In CD271-MSCs were identifi ed 287 genes with higher expression and 204 genes with lower expression compared to PA-MSCs. Preliminary results using Functional Grouping Analysis showed that the most prominent associations of upregulated genes were related to immune response. The majority of associated genes with 'T-cell immunity' overlap with partially redundant categories of 'Innate immunity' (34 genes, P=3.9e-07), 'Response to toxins' (52 genes, P=9.7e-06) and 'Receptor signaling' (52 genes, P=5.5e-05). Interestingly, many of these genes are involved in antigen presentation and cell-mediated immune responses: HLA-Class II, CIITA, invariant chain (CD74), alpha-2-microglobulin and genes important in peroxisome function. In addition, genes of innate immunity were also upregulated e.g. several members of the defensin gene family, complement factors and iNOS. Genes identifi ed with 'Receptor signalling' were mainly related to developmental processes including cell proliferation and diff erentiation. Only the set of down-regulated genes involved members of signifi cant pathways. Consistent with the fi ndings of the functional processes, signalling pathways such as that of TGF-beta and Wnt-pathway were aff ected which may correlate with more alterations in cytoskeleton and proliferation potential. In addition, cytokine/ chemokine signalling pathways were signifi cantly enriched, confi rming the aforementioned expression of immunoregulatory molecules by these MSCs. Discussion: Taken together, these results may explain the genetic basis for the functional diff erences between CD271-MSCs and PA-MSCs concerning their proliferative, diff erentiation and engraftment-promoting properties. Disclosure of Interest: None Declared. Introduction: Mesenchymal stem cells (MSCs) are multipotential cells which are capable of diff erentiating into a variety of cell types. Due to their immunomodulatory properties, MSCs have been used as cell-based therapy to reduce grafts-versus-host disease (GVHD) after allogeneic hematopoietic stem cell transplantation (HSCT). As reactivation of cytomegalovirus (CMV) constitutes a frequent problem after HSCT, we studied whether MSCs may also have an infl uence on CMV-specifi c T cell responses. Materials (or patients) and Methods: To study the eff ect of MSCs on alloantigen-induced proliferation a mixed lymphocyte reaction (MLR) was conducted. CD8 negative peripheral blood mononuclear cells (PBMCs) from donor A were co-incubated with CD8 positive cells from donor B (APC: T cell ratio=8:1) for 6 days. MACS®-purifi ed CD8 positive T cells from buff y coats of HLA-A2/ CMV seropositive healthy volunteers were stimulated with immunodominant CMVpp65-and infl uenza-derived peptides in a mixed lymphocyte-peptide culture (MLPC). Bone-marrow derived MSCs from human platelet lysate (PL) or fetal calf serum (FCS) based medium were co-cultured with the MLR and MLPC. Flow cytometry and enzyme-linked immunospot (ELISPOT) assays were used to study the immunomodulatory eff ect of MSCs on virus-specifi c CD8 positive T cells. Results: We confi rmed in a MLR that third party MSCs suppress alloantigen-induced proliferation. We demonstrated that MSCs do also inhibit proliferation of CMVpp65 (495-503)-specifi c CD8 positive T cells. Inhibition was strictly dependent on the number of MSCs but independent of the medium. We could corroborate our fi ndings with other immunodominant T cell specifi city towards CMVpp65 (417-26) and infl uenza matrix protein. Thus, our data are not in line with the notion that MSCs have a diff erential eff ect on alloantigen-and virus-specifi c T cells. Discussion: MSCs have strong immunosuppressive eff ects on alloreactive T cells and infusion of MSCs could be a promising immunotherapy for GVHD. However, MSCs can also inhibit CMV specifi c CD8 positive T cells. Therefore attention must be paid to early detection and pre-emptive treatment of CMV reactivation when patients are undergoing therapy with MSC. Disclosure of Interest: None Declared. Introduction: Mesenchymal stem cells (MSCs) have been applied to treat refractory chronic graft-versus-host disease (cGVHD) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) because of their potent immunomodulatory eff ects. However, the infl uence of cGVHD on MSCs is unknown. Here we analyzed the characteristics of MSCs derived from patients with cGVHD and without cGVHD. Materials (or patients) and Methods: Bone marrow MSCs were isolated from 22 patients with cGVHD (median age 29, range 15-47) and 18 patients without cGVHD (median age 26, range 18-43). Chimerism was analyzed by short tandem repeat (STR)-PCR. MSC frequency was evaluated by counting colony forming unit fi broblasts (CFU-F) in 10 6 bone marrow mononuclear cells. Senescence-associated β-galactosidase (SAβ-gal) assay and real time PCR were used for evaluating senescence of MSCs. Apoptosis and immunoregulatory functions were assessed by fl ow cytometry, and supernatant cytokines was detected by ELISA. The migration potential was evaluated using transwell chambers. Results: Both MSCs from patients with or without cGVHD were of host origin, and positive for CD73, CD90, CD105; negative for CD11b, CD19, CD34, CD45, HLA-DR. The frequency of MSCs did not diff er signifi cantly between patients with cGVHD (19.35±4.16) and without cGVHD (23.77±10.21) (P=0.635). They showed similar morphology, population doubling times and diff erentiation capacity. SAβ-gal assay revealed that the percentage of senescenct cells was 8.82±1.17% in patients with cGVHD and 11.64±3.17% in patients without cGVHD (P=0.415). The expression of senescenceassociated genes such as p53, p21, p16 were also comparable. No remarkable diff erences of apoptotic cells were observed between two groups. Importantly, the immunoregulatory functions of MSCs were not aff ected by cGVHD. They strongly inhibited the proliferation of phytohemagglutinin-activated peripheral blood mononuclear cells (PBMCs). Cocultured with CD4 + T cells, they signifi cantly increased the percentage of CD4 + CD25 hi CD127 lo/-Treg cells compared with CD4 + control cells (7.81±0.53% versus 1.64±0.19% P<0.001), and no signifi cant diff erences compared with MSCs from patients without cGVHD (8.59±0.88% P=0.788).

In addition, secretion of immunosuppressive cytokines IL-10, TGFβ, HGF, PGE2 was similar between two groups. To test whether migration of MSCs was infl uenced by cGVHD, we used transwell assay and results showed that this function was slightly impaired but did not reach signifi cance. Discussion: These results suggest that MSCs of patients with cGVHD preserve similar characteristics and functionality as those from no cGVHD controls. Hence, they could be considered in an autologous setting for those patients whose MSCs were not impaired before hematopoietic stem cell transplantation. Disclosure of Interest: None Declared. Introduction: For clinical application of cells derived from human pluripotent stem cells (hiPSCs), the safety and cost-eff ectiveness of culture system for hiPSCs should be guaranteed. Current ex vivo feeder-free culture system for human pluripotent stem cells requires the gelatin conditioned with mouse embryonic fi broblast (Matrigel®) and continuous supplementation of basic FGF. In this culture system, Very high cost is required and contamination of animal product is unavoidable. In this study we evaluate the long-term effi cacy of newly developed hiPSCs culture system using human pMSCs-conditioned media (PCCM) without bFGF and Matrigel®. Materials (or patients) and Methods: Surgically isolated placental chorionic plates from healthy women who had undergone abortion at 6-8 weeks of gestation were minced and incubated. At approximately 2 weeks after inoculation, colonies of fi broblast-like cells were collected. For purifi cation of pMSCs, CD44+CD34-cells was obtained by fl owcytometric isolation. For generation of PCCM, These cells were cultured by chemically defi ned culture media for 48 hours and then culture media (PCCM) was collected. bFGFsupplemented feeder-free culture media(mTeSR1®) was used as control. The hiPSCs were maintained on culture plate coated only by gelatin using PCCM with or without bFGF. Stemness of cultured hiPSCs was identifi ed by immunostaining for alkaline phosphatase (ALP), fl owcytometry for stage specifi c embryonic antigen (SSEA)-1, SSEA-4, tumor rejection antigen (TRA)-60, TRA-81, and RT-PCR for Oct-4, Nanog, and Rex-1 at every 10 th passage. Embryoid body (EB) formation was induced from cultured hiPSCs at every 10 th passage. For detection of the presence of three germ layers within the formed EB, RT-PCR analysis were performed on day 21 (Desmin for mesoderm, AFP for endoderm, TUJ1 for ectoderm). To defi ne the composition of PCCM, cytokine array and ELISA was performed. Results: Over the 26 passages for 6 months, the undiff erentiated morphology of colonies of hiPSCs was observed only in condition using PCCM without bFGF. Stemness markers of morphologically undiff erentiated colony were well expressed in this condition. EB formation from this colonies was successful and expression of three germ layer specifi c markers were well observed. Cytokine array and ELISA identifi ed higher concentration of IL-8, MCP-1, GRO, and GRO-a and lower concentration of bFGF in PCCM compared with mTeSR1®. Discussion: This newly developed culture system using PCCM without exogenous bFGF supplementation and gelatin coating plate showed to be eff ective in long-term ex vivo maintaining stemness of hiPSCs. The mechanism by which PCCM supports stemness of hiPSCs is considered to be bFGF-independent. However, exact pathway should be investigated. Results: The median time of neutrophil engraftment (above 0.5x10E9/L) was 16 days, all pts engrafted. The most frequent toxicities were grade III/IV infections according to common toxicity criteria in 11 of 15 pts and gastrointestinal toxicities (grade III in 8 of 15 pts). Incidence of acute GVHD was evaluated in 14 pts: 50% (7/14) of pts had GVHD (grade I+II in 4 pts, grade III in 3 pts). Incidence of chronic GVHD was evaluated in 13 pts, 46% (6/13) of pts had GVHD (limited in 3 pts, extensive in 3 pts). Nonrelapse mortality (NRM) after 1 year and 2 years was 7% and 13%. Causes of death were refractory GVHD (n=1) and infection (n=1). Complete remission was achieved in 13 of 15 pts (87%), progression was presented in 2 pts (13%). Complete chimerism was achieved in 67% of pts (10/15). With median follow-up from SCT 30 months (range 6-65), 83% of all pts (11/15) were alive (9 pts in remission, 2 pts with relapse), 4 pts died (2 deaths from NRM, 2 deaths from progression of MF). Two relapses (13%; 2/15) occurred in intervals of 6 and 18 months after SCT. Discussion: FLAMSA-RIC protocol represents a promising approach to the treatment of high-risk or intermediate-2 risk myelofi brosis with high response rate (87%); progression-free survival and overall survival at 2 years from SCT were 62% and 83%, respectively. It provides a combination of eff ective disease control, low non-relapse mortality, and acceptable toxicity. Other prospective clinical trials are needed to confi rm the results of this novel therapeutic strategy. Disclosure of Interest: None Declared. 

Results: After RIC2, 201 patients (88,5%) engrafted with the median time to ANC>500/μL of 15 (range, 1-40) days. Grade II-IV acute GVHD after RIC2 occurred in 31,9% of patients. With a median follow-up of 21 (range 1,5-79) months after RIC2, 56 patients were still alive. At 2 years, the rates of OS, LFS, NRM and RI were 21 +/-3%, 14+/-2%, 22+/-3%, and 64+/-3%, respectively. Duration of remission following RIC1 and disease status at RIC2 were found to be strongest predictors of patients' outcome -3 distinct groups were identifi ed: those who relapsed more than 7 months after RIC1 and were in CR at time of RIC2, those with either early relapse after RIC1 or no CR at RIC2, and those with early relapse after RIC1 and no CR at RIC2. The overall survival rates in those groups were 40+/-8 vs. 26+/-5% and 7+/-3%, respectively. Discussion: Second reduced intensity conditioning allogeneic transplant is feasible in acute leukemia patients who relapse after RIC1. Due to the lack of other eff ective therapeutic options, this approach remains a valid option for patients with delayed relapses following RIC1 as well as for those achieving CR prior to RIC2. Disclosure of Interest: None Declared. Introduction: Fanconi anemia is a genetic disorder associated with diverse congenital abnormalities, progressive bone marrow failure, and increased risk of leukemia and other cancers. Aff ected persons often die before 30 years of age. Bone marrow transplantation is an eff ective treatment. Since 1997, when we started our transplant program at Nasser Institute (Cairo, Egypt), Fanconi anemia patients, subjected to BMT using high dose cyclophosphamide had an inferior outcome with overall survival near to 16%. Thus, in this study, we investigated the use of lower dose cyclophosphamide and its impact on the outcome and overall survival. Materials (or patients) and Methods: In the time period between December 2007 and February 2012, thirty one patients with Fanconi anemia (FA) with evidence of bone marrow (BM) aplasia underwent allogeneic BM transplants (BMT) from matched sibling donors using PBSCs at the BMT unit of Nasser institute, Cairo, Egypt. Median age at BMT was 11.7 years. Conditioning consisted of low-dose cyclophosphamide (CY; 5 mg/kg x 4 days) and a total dose of fl udarabine of 120/m 2 . Graft-versus-host disease (GVHD) prophylaxis using cyclosporine-A starting dose 3 mg/kg then tailored according to trough level. In addition anti-thymocyte globulin (ATG) was administered in the pre-transplant period to promote engraftment and in the post-transplant period for additional GVHD prophylaxis (30 mg/kg total dose). Results: Thirty one Fanconi anemia patients with median observation time 1.9 years, the overall survival was 64.5%. Engraftment occurred rapidly (mean, 13.6 days for an absolute neutrophil count > or = 0.5 x 10 9 /L; mean, 16.1 days for platelet count > or = 50 x 10 9 /L). Twenty patients have sustained engraftment and are transfusion-independent . Two patients developed secondary graft failure after initial engraftment and one had primary graft failure. Five patients developed acute GVHD (15.6%),and the incidence of chronic GVHD did not exceed 3.2%. Discussion: PBSCs transplantation is a feasible option for FA patients having matched sibling donors and should be performed early in the course of the disease, before the development of complications. We believe that the use of low-dose CY in addition to ATG in the conditioning regimen was responsible for improvement in the survival of FA patients undergoing BMT. The regimen was well tolerated and was associated with a low incidence of complications including GVHD. Disclosure of Interest: None Declared. Blood 2010), secondary AML, and patients requiring 2 induction courses to obtain CR. The chemotherapy sequential regimen consisted in fl udarabine 30 mg/m2, high-dose cytarabine 2 g/m2, and amsacrine 100 mg/m2 from days -12 to -9 (FLAMSA). After 3 days of rest, RIC consisted of 4 Gy TBI on day -5, cyclophosphamide (40 mg/kg with HLA-identical sibling, 60 mg/kg for unrelated or mismatched donors) on days -4 and -3, and rabbit antithymocyte globulin (ATG, Genzyme) (5 mg/kg total dose) from day -3 to day -1. As a new experimental approach, we replaced TBI by iv. busulfan (BU) 3.2 mg/kg/d during either 4 or 2 days according to patient age (>55 years) (from day -7 to -4 or from day -5 to -4). GvHD prophylaxis consisted in ciclosporine from day −1, and mycophenolate mofetil (15 mg/kg bid), starting from day 0. Except for cord blood transplantation, patients received 3 prophylactic increased doses of donor lymphocyte infusions (DLI) if they were in CR and GvHD-free at day +120 or 30 days after discontinuation of immunosuppressive agents starting at 1x10 6 CD3+ cells/kg. Results: Between August 2010 and March 2013, 26 consecutive AML patients in CR1 were included; 11 males and 15 females with a median age at allo-HSCT of 55 years (range: 24-67), 19 (73%) were de novo AML and 7 (27%) secondary AML. According to cytogenetics and molecular markers, 22 (85%) were unfavourable and 4 (15%) were in intermediate II category. Before allo-HSCT, to reach CR1, 20 (77%) patients received one induction chemotherapy and 6 (23%) needed 2 inductions. Stem cell source was PBSC for 23 (88%) patients, CB for 2 and BM for 1 patient. Donors were 10/10 HLA matched siblings in 9 (35%) patients, 10/10 HLA matched unrelated in 8 (31%) patients and HLA mismatched for the rest of patients [unrelated 9/10 (n=7), CB 4/6 (n=2)]. For ABO compatibility, 13 (50%) were compatible, 5 (19%) had minor incompatibility and 8 (31%) had major incompatibility. For conditioning, 6 (23%) patients received TBI, 13 (50%) received 4 days BU and 7 (27%) received 2 days BU. After transplantation, 23 (88%) patients engrafted. At day 90 post-allo-HSCT, 18 (78%) showed total donor chimerism and 5 (22%) had mixed chimerism and all patients were in CR. There were 6/23 patients with acute GvHD [2 gr I, 2 gr II and 2 gr III] and 5/23 chronic GvHD [4 limited and 1 extensive], all before DLI. After a median follow-up of 9 months (range: 0.03-35), the 2-years probability of overall survival (OS) for the whole population was 58% (47-69) and the 2 years cumulative incidence of relapse was 18% (17) (18) (19) . No statistical diff erence in terms of OS and relapse incidence was found between the 3 types of conditioning. Discussion: FLAMSA-RIC regimen followed by allo-HSCT showed promising results in high-risk CR1 AML patients. Because of some early severe infections, an effi cient prophylactic anti-infectious strategy is recommended. and 58%). Cumulative incidence of non-relapse mortality (NRM) was at day +100 7% (RIC=3%, MAC=17%) and at 3 years 28% (RIC=27%, MAC=31%), respectively. Engraftment of neutrophils (ANC>500/μL) was signifi cant diff erent with a median of 16 (range 9-31) days after MAC compared to 23 (range 13-60) days after RIC (P= 0.0051). Platelet recovery (PLT>20.000/μL) occurred at a median of 24 (range 11-51) days after MAC and 21 (range 8-41) days after RIC (P= 0.2866). Graft rejection was rare, occurring only in 3 patients after RIC. The cumulative incidence of acute GVHD was 29% (n=17/59, grade I n=13/17, ≥ grade II n=4/17). The risk for aGVHD for patients after MAC was signifi cantly higher compared to patients after RIC (71% (n=12) versus 29% (n=5), P< 0.0001). The cumulative incidence of chronic GVHD was 54%. Extensive cGVHD was seen in 44% and limited cGVHD in 56%. Survival of patients with an extensive cGVHD was poor compared to patients with a mild or limited cGVHD (18% versus 39%, P= 0.2059). Discussion: Allogeneic HCT following RIC is a feasible and safe treatment option for patients with primary and secondary MF with a 3-year overall survival of 64%. Even allogeneic HCT with MMUD may off er curation to an older and more comorbid patient population. Disclosure of Interest: None Declared.

Introduction: Systemic Light-chain (AL-) amyloidosis is a rare protein folding and deposition disorder which is caused by a monoclonal plasma cell or B cell disorder with poor prognosis. Based on small series of patients and case reports allogeneic transplant (allo SCT) has emerged as potentially eff ective (EBMT retrospective data: Schönland et al., Blood 2005 , DLI data: Haematologica, 2008 . However, TRM was 40%: Therefore, a more formal proof of concept of using allogeneic hematopoietic transplantation for treatment of AL Amyloidosis is lacking. Materials (or patients) and Methods: The primary endpoint of this NIS is effi cacy (best hematological remission (HR) and organ response). Secondary endpoints are acute and chronic GvHD, TRM, non-hematological toxicity, event-free and overall survival. We selected those centres that in the past performed any allogeneic HSCT for AL amyloidosis. We approached 24 centres, of which 4 are participating in the study today. So far 10 patients have been included and were transplanted between 2006 and 2012. Age at allo SCT was 50 years in median (range, 42 -60 years). Five patients had cardiac, 7 kidney, 5 liver and 2 patients nervous system involvement. As underlying disease one patient had symptomatic multiple myeloma, 9 patients a clonal plasma cell dyscrasia or another B cell disorder. All patients were in a good performance status measured by Karnofsky Index (3 patients 80%, 6 patients 90%, 1 patient 100%). Previous chemotherapy included high-dose melphalan with autologous stem cell transplantation in 8 patients as well as bortezomib, lenalidomide, melphalan and steroids. Disease stages at allo SCT were as follows: 1 CR, 1 VGPR, 3 PR, 3 stable disease, 1 progression. Two patients received myeloablative and 7 patients received RIC conditioning using TBI 2 Gy / fl udarabine (2 patients with ATG). High-dose melphalan 200 mg/m 2 was applied prior to the one syngeneic SCT. Donors were: 5 sibling matched donors, 1 mismatched relative donor, 1 syngeneic donor, 2 matched unrelated donors and 1 mismatched unrelated donor (MMUD). Source of stem cells was peripheral blood in 9 and cord blood in 1 patient, respectively. Results: All patients engrafted. Acute GvHD grade II-IV occurred in 6 patients and chronic GvHD in 6 patients (3 patients with limited and extensive disease, respectively). The patient who received a graft from a MMUD died of steroid-refractory acute GvHD (Grade IV) of the gut. Best HR after allo SCT was CR in 4, VGPR in 1 and PR in 2 patients. Nine patients are alive with a median follow-up of 23 months. Two patients relapsed or progressed. Discussion: Preliminary analysis of the fi rst 10 patients of this NIS showed that allo SCT is feasible and eff ective in patients with AL amyloidosis. In opposite to our retrospective analysis we have observed a low TRM using mostly reduced-intensity conditioning with TBI 2 Gy and fl udarabine. Allo SCT might be a reasonable treatment option in young and medically fi t and heavily pretreated patients. Disclosure of Interest: None Declared. Introduction: Because of the changes of the demographic development and long-livity there will be more patients (pts) in their 7. th and 8. th decade, which will be diagnosed with a myeloid malignancy such as AML/MDS. Further, the personal fi tness of these elderly has changed with more of them fi t enough and asking for receiving a more intensive therapy for cure. After the introduction of reduced-toxicity conditioning we transplanted from 1999 to 2012, 250 consecutive pts with AML/MDS aged ≥ 60 years (yrs). Materials (or patients) and Methods: The 144 male and 106 female pts with a median age of 66 yrs (range 60-77) were transplanted for de novo AML (n=95), s/tAML (n=104) and MDS (n=51) with mainly unfavorable cytogenetics. The donor was matched/mismatched unrelated in 74% and in 26% related in 26%. Only 16% were transplanted in CR1/2, 84 % with advanced or untreated disease. The conditioning regimen was the FBM protocol (Fludarabine 4x30mg/m2, BCNU/carmustine 2x150mg/m2, Melphalan 110mg/ m2; Bertz et al., JCO 2003) in 98%, and the graft in 97% PBSC. For GVHD prophylaxis in 91% a combination of cyclosporine A plus alemtuzumab or ATG-F™ was administered.

Results: At a median follow up of 57 months (3-157) 37% of the pts are alive; main causes of death were relapse (n=62), infection (n=35) and age-related diseases (n=13 Introduction: Allogeneic hematopoietic stem cell transplantation (HSCT) is an established therapy for malignant and nonmalignant hematologic disorders. Reduced-intensity conditioning (RIC) regimens have expanded the use of HSCT to elderly and higher risk patients. However, HSCT still remains associated with a signifi cant mortality and morbidity and the careful assessment of risks and benefi ts before transplantation is essential. Major factors which infl uence non-relapse mortality (NRM) and overall survival (OS) after HSCT are diagnosis, type of transplant, remission status and the patient's risk profi le, which includes age and presence of comorbidities. The use of the hematopoietic cell transplantation-specifi c comorbidity index (HCT-CI) has been proposed to predict the probability of non-relapse mortality (NRM) and overall survival (OS) following HSCT. However, the HCT-CI usefulness for older patients receiving a reduced intensity allogeneic HSCT remains unclear.

Materials (or patients) and Methods: A retrospective medical record review was performed to collect data from patients who underwent an allogeneic HSCT at a large, urban, NCI Results: In vitro isolated and expanded human MAB proved poorly immunogenic in resting conditions and intrinsically resistant to T-cell killing. However, upon exposure to INF-g or diff erentiation in myotubes, MAB acquire the ability to promote expansion of alloreactive T cells and become sensitive to T-cell killing. Resistance of mesoangioblasts to T-cell killing is largely due to the expression of the intracellular serine protease inhibitor PI-9. Strikingly, this intrinsic mechanism of stem cell immune evasion does not interfere with the development of T-cell based immune responses against stem cell progeny. In patients receiving MAB infusions, despite standard corticosteroid plus tacrolimus continuous treatment, the numbers of circulating lymphocytes, the relative proportion of naïve/memory T-cell subset and of Tregs were comparable to that measured before treatment. Thus, the treatment does not perturb T-cell homeostasis, and accordingly, infectious adverse events were not recorded. We described allo reactive T-cell responses in three out of fi ve patients, and an inverse correlation between alloreactivity and clinical benefi t. Discussion: Altogether these results suggest that hypoimmunogenic MAB might become immunostimulatory in the infl ammatory milieu encountered in dystrophic muscles, justifying and recommending the use of immunosuppressive and/or antiinfl ammatory drugs in allogeneic cell therapy of DMD. Disclosure of Interest: None Declared. Our 10 years experience with use of marrow derived mononuclear cells for revascularization of critically ischemic limbs prompted use to use these cells to stimulate revascularization of aff ected bones. Materials (or patients) and Methods: For this procedure 5 patients (4 males, 1 female, age 25-35 years) were enrolled. All suff ered from nontraumatic osteonecrosis. In two of them circulating lupus anti-coagulant was detected what was previously associated with deep venous thrombosis. In one of these two the procedure was performed in both hips. The diagnosis was confi rmed by radiology followed by magnetic resonance imaging evaluation. Clinical observation was based on Harris, Ficat and Mitchel scores. The procedure was performed as follows: (i) bone marrow harvested from posterior iliac crest under a general anesthesia in a volume of 500 mL was enriched in mononuclear cells (BMMNC) with the use of a Cobe Spectra separator, (ii) on the same day orthopedic surgery was performed: 4 channels were drilled with the use of Kirschner wire starting from the greater trochanter area, the channels (4.5 mm in diameter) penetrated to the aff ected head of the femur and were fi lled with the marrow mononuclear cell population in a volume of 30 mL. (iii) patients were kept in bed for one day and then they could move with the use of crutches. In a few days the patients started passive and then active rehabilitation in the out-patient clinic. Results: In the transplanted inoculum were: CD34-CD45-CD90+ cells: 0.026%, CD34+CD45-VEGFR+: 0.012%, CD16+CD14+ cells having physical parameters of monocytes 0.027%, which included 70% Tie2+ cells. The observation following the surgery included magnetic resonance imaging and the clinical scoring. In the patient with two hips aff ected, one hip had to be replaced. Histopathology revealed an increased number of small collagen IV positive vessels with erythrocytes. The second hip 18 months after the marrow mononuclear cells implantation is stabilized in function and MRI documented the presence of bright areas corresponding to the repairing zones. Also other cases being from one year to 8 months after the cellular treatment enjoy stabilization of the aff ected hips. Discussion: The procedure seems to be feasible off ering some improvement in the function of aff ected hips with rather low personal and fi nancial costs as compared to total hip replacement. were exaggerated prior to the cellular therapy. These patients constituted the fi rst cohort (experimental) and were followed for 4 to 6 yrs. After that 10 patients enrolled the second cohort (validation). Materials (or patients) and Methods: Procedure employed for the fi rst and the second cohort was the same and included: (i) bone marrow harvest from posterior iliac crest in a volume of 500 mL, (ii) mononuclear cells enrichment with the use of a Cobe Spectra separator which ended with a volume of 100 mL, (iii) 70 mL of the cell suspension were injected in 0.5 mL portions to calf muscles of the aff ected limb. Results: Analysis of the implanted cells revealed the presence of (mean±SEM): (i) hematopoietic progenitors: CD34+, CD45+: receptor on NK cell receptors was analyzed by fl ow cytometry in 11 metastatic primary cell lines: 319M, 531MII, 588M, 595M, 654M, 685MII, 699B, 709M, 716M, 722MII, 728MII. All cell lines derived from patients were lung metastases except for the 531MII, which was one-rib metastases, and for the 699B, which was a primary tumour. Mean Fluorescence Intensity (MFI) was calculated by MFI of the specifi c staining relative to the MFI of the appropriate isotype control staining. Activated and Expanded NK (NKAE) cells were obtained by co culture of PBMCs from healthy donors with the K562mbIL15-41BBL cell line (previously irradiated with 100Gy) in a ratio 1:1.5 for three weeks in RPMI supplemented with 10% human AB serum (Sigma), 100 IU/mL IL-2 (Miltenyi), and 1% PS. The NK cell purity was assessed by fl ow cytometry. The cytotoxicity of NKAE/NKAE IL-2/IL-15 cells was monitored using a conventional 2-h europium-TDA release assay (Perkin-Elmer Wallac, Turku, Finland). The number of NK cells was calculated by multiplying the lymphocyte counts to the percentage of CD3 -CD56 + NK cells. To explore the importance of the NK cell receptors and their ligands interactions for a NK cell elimination of osteosarcoma, cytotoxicity assays were performed in the presence of diff erent blocking antibodies. Results: At least 3/5 NKG2D ligands and HLA class I on primary osteosarcomas had a low expression (MFI/MFI isotype ≤ 10) in 10/11 and 8/11, respectively. However, Fas and CD112 were highly expressed (MFI/MFI isotype ≥15) in 8/11 of the primary osteosarcomas. Cytotoxicity assays revealed primary osteosarcoma were resistant to criopreserved NK cell-mediated cytolysis, although cytolytic activity was strongly enhanced by IL-15/IL-2 overnight activation, and when fresh NKAEs were used as eff ectors. The contribution of each ligand to NK cell-mediated cytolysis was studied by specifi c antibody blockade, depending on the ligands expression. We also explored diff erent ways to enhance NK cell cytotoxicity for each osteosarcoma. We did not fi nd a common pattern; each osteosarcoma seemed to be dependent on diff erent pathways. Discussion: NKG2D and HLA class I ligands have a low expression on primary osteosarcoma, probably underlying a NK cell evasion mechanism. However, Fas/FasL and DNAM-1/CD112 pathways could be important for NK cell recognition. Although primary osteosarcomas were resistant to NK cell cytolysis, cytotoxic activity was strongly enhanced by IL-15/IL-2 overnight stimulation, and with activated and expanded NK cells. However, each osteosarcoma depended on diff erent pathways. This fact shows the heterogeneity of osteosarcomas and underlies the need for personalized strategies to treat these tumours. Disclosure of Interest: None Declared. (n=13), irradiation (n=3), surgical treatment (n=9), high-dose chemotherapy with auto-HSCT (n=9). In 10 of 13 pts allo-HSCT was performed as salvage therapy. The donors were haploidentical (n=10), matched unrelated (n=2) or matched related (n=1). The graft source was G-CSF -primed bone marrow (n=6), PBSC (n=3) or both (n=4). The median CD 34 count in the graft was 10.6 х 10 6 , CD3 5.5 х 10 7 . Conditioning regimen consisted of fl udarabin and melphalan (n=8) or busulfan and thiotepa (n=5). The graft-versus-host disease (GVHD) prophylaxis consisted of tacrolimus or cyclosporine A with addition of ATG (n=7) or post-transplant cyclophosphamide (n=6) 

The expression of the most important ligands was evaluated by cytofl uorimetric analysis and gene expression on cultured tumor cells and in immunohistochemistry sections of the original sample embedded in paraffi n. Results: Gene expression of ligands was performed by realtime quantitative PCR (relative quantification) An endogenous control was used to account for differences in the amount and quality of total RNA added to each reaction. PVR, MICA/B, were highly and uniformly expressed in tumor cells (TC) of patients analyzed (mean value: 5.57, 6.17, 6.5 respectively), while, lower levels of ULBP1/3 were detected ( Expression of phenotypic markers of each lineage was evaluated by immunohistochemical markers (HIC). HLA typing, evaluating KIR mismatches and cytotoxicity assays were performed. Results: Cells were isolated from tumoral tissue derived from biopsy, cultivated under the appropriate conditions until confl uence.The cells had to be characterized by HIC. The HIC evaluation was performed on the basis of tumoral markers at diagnosis. 7 NB were non tumoral with a low expression of neoplastic markers and only 3 were neoplastic with an espression of cancer markers above of 40%, only one didn't grow. 3/9 RB were neoplastic with an expression of tumoral markers on 70% of the cells, 1/9 didn't grow in vitro and 5 weren't tumoral. As concern ES only one was neoplastic. 4/6 WT were neoplastic with an high expressions of tumoral markers on average 70% of the cells. These data confi rm the technical diffi culty to obtain tumoral primary cultures from surgical biopsy especially in NB. Q-PCR was performed to determine the mRNA expression of tumoral markers detected at onset:preliminary data show relevant levels of mRNA, in keeping with HIC revaluation. We are now in the process of developing a multiparametric fl ow cytometric screening for solid. With this approach heterogeneous cell populations can be analyzed and this can provide informations on the functional status of regulatory processes by the simultaneous measurement of multiple key elements. Moreover patients candidate were assessed for HLA compatibility by serology and by high-resolution molecular analysis. NK alloreactivity assay against patient's blast was performed.

On the basis of our pre-clinical studies, a child aff ected by stage IV ES in partial remission, after completing the whole of the AIEOP protocol, underwent haploidentical transplantation of hematopoietic stem cells (HCT) from his KIR alloreactive uncle.The transplant consisted of 1.5x10 6 /kg of CD34+ and 8x10 4 /kg of CD3+. After 4 months from HCT, patient was in complete remission. He had a overall survival of 3 years after transplantation respect 1-2 months expected. An additional child aff ected by stage IV NB in partial remission after 3 lines of chemotherapy was grafted from her KIR alloreactive father and survived in partial remission for 5 months. Discussion: Alloreactive NK-cell mediated antitumor eff ects might provide useful perspectives with the aim to design innovative new cell therapy approaches against solid tumors in children. Disclosure of Interest: None Declared. Introduction: DSRCT is a rare and highly aggressive mesenchymal tumor. Despite aggressive multimodal treatment, the prognosis remains poor. We performed a mono-institutional retrospective study to analyze treatment and outcome of pts with DSRCT. Materials (or patients) and Methods: We retrospectively reviewed the medical records of pts with DSRCT treated from May 1997 to July 2012 at Humanitas Cancer Center. Demographic data, tumor characteristics, treatment and response rate (RR) according to RECIST criteria were collected; median Progression-Free Survival (mPFS) was analyzed with Kaplan-Meier method. Results: We treated 26 pts with DSRCT (M/F ratio, 24/2), median age at presentation 30 (range 15-65). Primary tumor site was abdomen/pelvis (88%; n= 24) and mediastinum (12%; n= 2). Sixteen pts were metastatic at diagnosis (62%). Six pts underwent debulking surgery elsewhere, of whom 4 with local disease (LD) and 2 with metastatic disease (MD). Pts received induction chemotherapy (CT) as follows: 10 VAI (vincristine, adriamycin, ifosfamide) with 6 PR, 2 SD and 2 PD; 6 miniCARBOPEC (etoposide, carboplatin and cyclofosfamide) with 5 PR and 1 PD; 4 modifi ed P6-regimen with 2 PR and 2 SD. 6 pts received diff erent adriamycin-based CT of whom one was lost to follow-up. Overall, the overall response rate (PR + SD) was 73%. 19 responding pts were treated with highdose chemotherapy (HDC) of whom 11 with MD and 8 with LD. At day+100 after HDC, the disease status was CR 1, PR 8, SD 5 and PD 5. Seven pts underwent surgery after HDC and 3 had post-surgery radiotherapy (RT). The mPFS of pts who had ASCT was 11.61 months (range, 5.07-97.2). Discussion: DSRCT management requires multimodality treatments. Although investigational, HDC seems to obtain long-term disease control, even in the metastatic setting. The role of HDC in DSRCT needs to be prospectively evaluated. Disclosure of Interest: None Declared. Introduction: Classical tumor therapy is generally based on surgery combined with radio-and chemotherapy. However, recently additive immunotherapy has gained in impact. Diff erent strategies have been developed with the aim to strengthen the antitumor capacity of the immune system. It has further been demonstrated that treatment with chemotherapeutics or use of ionizing radiation (X-ray) contributes to a benefi cial tumor microenvironment. Materials (or patients) and Methods: In order to gain deeper insight into how X-ray in combination with adjuvant hyperthermia treatment results in immunological responses, we examined the role of T-, B-and NK-cells by use of RAG1 -/mice as well as NK1.1-depleting antibodies in the following two scenarios.

(1) Immunization with whole tumor cells that have been exposed to X-ray and hyperthermia (2) Local in situ treatment of solid tumors with X-ray and hyperthermia Results: (1) In vitro treatment of tumor cells by X-ray in combination with hyperthermia induces a mixture of apoptotic and necrotic cells. After subcutaneous injection of these cells as vaccines we could generate very eff ective immune activation and tumor protection. Furthermore, our in vivo results suggest that the impact of NK cells on antitumor immunity signifi cantly diff ers dependent on the phase of tumor progress and treatment. Importantly, during the antitumor immunization period, NK cells mediate a negative infl uence on the generation of the favored antitumor response in our model. (2) The transfer in a therapeutic immunization protocol of established tumors in situ showed similar eff ects. Depletion of NK cells during the period of local treatment with X-ray and hyperthermia provided a signifi cant better outcome than in non-depleted mice.

Discussion: Summarized, we demonstrated that the treatment with X-ray in combination with hyperthermia induced the best immunogenic eff ects. Moreover, our preclinical studies indicate importance of immune editing eff ects of NK cells. These experiments are of high relevance for the further development of multimodal immunotherapeutic protocols combining X-ray and hyperthermia with cell therapeutic approaches for the treatment of solid tumors. Disclosure of Interest: None Declared. of EBV-specifi c autologous polyclonal CTL therapy. Human papillomavirus (HPV) 16 is a prognostic marker for enhanced overall and disease-free survival in oropharyngeal cancer, but its uses as a predictive marker or a therapeutic target have not yet been proven in the specifi c setting. Materials (or patients) and Methods: The aim of this study was to evaluate the feasibility of expanding HPV-specifi c CTL from 10 healthy donors and 5 patients with HPV+ oropharyngeal squamous cell carcinoma (OSCC), to be employed as targeted therapy for OSCC. We conducted experiments to validate an in vitro culture method to expand HPV-specifi c CTL, by peripheral blood mononuclear cell (PBMC) stimulation with 15mer peptide pools derived from the HPV16 E6 and E7 proteins.

Results: T-cell lines (TCL), that included a median 75% CD4+ and 22% CD8+ T lymphocytes, were successfully generated from 7/10 healthy individuals and 4/5 OSCC patients. While only 70% of TCL expanded from healthy donors presented specifi c cytotoxic activity against PHA blasts pulsed with HPV E6/E7 peptide mix, all TCL obtained from OSCC patients exerted specifi c cytotoxicity, also at very low eff ector to target ratio. Some of the T-cell lines showed HPV16-specifi c INFg production in Elispot assays, with those from OSCC patients yielding consistently higher frequencies than healthy subjects (median 90 SFU/10 6 cells vs. 24 SFU/10 6 cells). Noncultured PBMC from the same individuals did not show any specifi c cytotoxicity, nor a measurable HPV16-specifi c IFNgproducing cell frequency. Introduction: The lower morbidity and mortality of reduced-intensity conditioning (RIC) regimens have allowed allogeneic hematopoietic cell transplantation (HCT) in older patients. However, there are only limited data on the feasibility and outcomes of URD HCT in elderly patients. The aim of the study was to compare the outcome in OS and PFS for patients transplanted using unrelated donor (URD) in patients age 60 or older. Materials (or patients) and Methods: We retrospectively analyzed outcomes in 62 consecutive hematologic malignancy patients aged > or =60 years (median, 62 years; range: 60-70 years) undergoing reduced intensity conditioning regimens (RIC) from URD. In this study, URD was used only when a MRD was not available.Then we compared the outcome of 17 elderly patients (age >65 years) with 44 younger patients aged between 60 and 65 years. Results: No patients experienced graft rejection. The median HCT comorbidity index score was 2 (range, 0 to 6). With a median follow up of 36 months (range, 5-74), the cumulative incidence of grades II to IV acute GVHD was 28% and of grades III to IV acute GVHD, 13%. At 2 years, the cumulative incidence of chronic GVHD was 27%, progression-free survival (PFS) was 62%, overall survival (OS) was 63%, and relapse was 14%. Non relapse mortality (NRM) was 24% at 2 years. The cumulative incidence of grade II-IV Acute GVHD was 43% for the younger group and 17% for the older group (P = 0.056). The cumulative incidence of chronic GVHD was not diff erent between the two groups (23% vs. 45% (P=0.3), respectively). Two-year OS and PFS was 57% versus 86% (P = 0.059) and 55% versus 86% (P = 0.03), in the younger and the older group respectively. The 2-year NRM and relapse was 26% versus 14% (P = 0.4) and 19% versus 0% (P = 0.04), in the younger and older group respectively. Discussion: This retrospective study suggest that RIC HCT from URD is a safe and effective option for patients aged > or =60 years or older, and in the absence of suitable related donors, well-matched URD may offer a very reasonable alternative, and that does not appear to be associated with a detrimental outcome. However these results are encouraging showing once again that with an adequate selection, age is not a definitive limitation. Disclosure of Interest: None Declared.

Introduction: HLA compatibility between the donor and the recipient in unrelated bone marrow transplantation is the key parameter for the clinical outcomes. Although both serological and allelic disparities have been reported to be associated with the post-transplant complications and survival, reports on direct comparison between antigen and allele mismatched UBMT are limited. We analyzed the JMDP registry data to compare the outcomes between antigen and allele mismatched UBMT. Materials (or patients) and Methods: 6183 UBMT recipients transplanted as fi rst HSCT for AML/ALL/CML/MDS between 1933 and 2011 in Japan were included in this analysis. UBMT before 1993 (n=62), or those reported with no GVHD prophylaxis (n=14) reduced-intensity conditioning, and the use of anti-thymocyte globulin. Discussion: The main problem in using a female donor for a male recipient in HSCT has been the increased risk of acute GVHD. This has contributed to worse survival. In the present study, we found that unrelated male donor transplants to male recipients resulted in signifi cantly more acute GVHD of grades II-IV than when using a female sibling donor. This was seen despite the fact that MUD patients were more often treated with ATG, which has been shown to reduce the risk of acute GVHD in MUD patients to a similar extent to that seen in HLA-identical sibling transplants without ATG.

To conclude, male patients with acute leukaemia who received grafts from MUD donors had a higher risk of acute GVHD and the same survival and LFS as using grafts from female HLA-identical sibling donors. Therefore, a female sibling donor should be preferred. Disclosure of Interest: None Declared. Introduction: The HLA genes are the most polymorphic in the human genome and are characterized by a large number of alleles and haplotypes. Nowadays a variety of methodologies are available for HLA typing both at the protein and at nucleic acid level, but ambiguity can aff ect the possibility of the right call for each HLA-locus. In particular, phase ambiguities arise from the incomplete genomic coverage or the contemporary Sanger sequencing of two heterozygous alleles that determines diff erent haplotypes.

The new generation sequencing (NGS) technologies have the potential to perform HLA-typing in a simply, rapid and accurate way without phase ambiguities. In this study we performed high-resolution HLA-typing using an NGS platform based on pyrosequencing and subsequent bioinformatic analysis and we evaluated his feasibility, reliability and robustness in 40 samples.

Materials (or patients) and Methods: Fourteen amplicons for sample were synthesized using two custom assay. The output fi le was then uploaded into JSI SeqPilot software to align all sequences with the reference database (ref 3.9 2012). Results: The PCR reactions generate 560 amplicons who correspond to HLA-A/B/C exons 2, 3 and 4, DQB1 exons 2 and 3 and DPB1, DQA1, DRB1/3/4/5 exon 2. Using Multiplex Identifi er (MID) tag method, we pooled amplicons from diff erent samples to analyze them contemporary. We pooled our 40 samples into 8 pools (5 diff erent samples in each one) and performed 8 sequencing runs. We have obtained over of 150 reads for most amplicons. The assignment of unambiguous genotype was possible on 45.5% of alleles. The ambiguities were related to the assay design (above all for class-II). Notably, some ambiguities on the locus B and C have a little biological importance because the genomic diff erences between the two alleles were located on the transmembrane domain coding region, so both alleles code for the same peptide binding domain, instead. Ten cases analyzed in this study were also genotyped using conventional strategies resulting in a concordance of 100%. Discussion: A rapid and accurate HLA-typing method without phase ambiguities is necessary to obtain a quickly and deep HLA compatibility analysis between donor a recipient. Using a NGS platform we obtained a high-resolution HLA-typing for the most important loci of the samples without phase ambiguities. We discriminated very well the alleles that are responsible of differences on the peptide binding domain in three days, so clonal amplifi cation and pyrosequencing seems a feasible and promising approach. . Graft composing was assessed by the determination of NC, CD34 and CD3 cell by local standard immunophenotyping and compared between the two study groups. Other main variables were analyzed: donor weight, patient weight and the pair ratio (donor weight/patient weight) in order to quantifi ed large discrepancy between donors and recipients weights (ratio <1) and consequently to evaluate the risk to no collect the minimum nucleated cell (NC) yield 2x108/kg required.

Results: The weight of infant donors was signifi cantly less then the young donors and consequently the bone marrow volumes harvest were signifi cantly smaller (see the Introduction: The Italian Bone Marrow Donor Registry was established in Italy in 1989, with the role of providing an unrelated volunteer with immunogenetic characteristics to haematological patients waiting for a transplant and who do not have an identical sibling. The minimum eligibility criteria for a prospective donor must fi t the Italian law for transfusion activity. The Donor Center (DC) is responsible for protecting donor safety: at every meeting with the donor, a medical evaluation of the volunteer donor is performed, to verify donor eligibility and to evaluate his suitability. In case of donors partially compliant with eligibility criteria, a "second opinion" SIMTI Committee, designated by the pertinent scientifi c society, helps DC to establish the suitability of the donor, with the aims to protect donor welfare and safety. The SIMTI Committee evaluates not only the suitability of donor partially compliant with eligibility criteria, but also the donors follow-up and management of HSC collection and minimum Results: Median duration of G-CSF administration was 6 days for both Zarzio® (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) and Granocyte® (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) . Both groups had a median of one apheresis and the group mobilized with Zarzio® achieved the goal of more than 2x10 6 /kg CD34+ cells per body weight in 57 patients with one session and in 17 with two sessions (one of them co-administered plerixafor) of apheresis, while 6 patients failed to mobilize (1 co-administered plerixafor). In the originator group 44 achieved the goal with one session of apheresis, 9 with two, while 8 failed (1 co Introduction: Peripheral blood stem cell collection (PBSC) by apheresis has become the preferred source of stem cell grafts in hematopoietic stem cell (HSC) transplantation setting. The apheresis process involves establishment of an extracorporeal circulation, separation of whole blood into its components, and selective collection of stem cell-rich fraction for use as a stem cell graft. A novel apheresis system, the Spectra Optia-a modern electronic automated device-has been recently introduced into clinical practice in most transplant centers to overcome some shortcoming such as the high amount of irrelevant white blood cells (WBC) in the fi nal product and the high variability in CD34 collection efficiency (CE2 -calculated by CD34 + /ml product x Poduct Volume)/ CD34 + /ml pre x Total Processed Blood Volume). Materials (or patients) and Methods: We have recently introduced a new Optia device into our collection facility and consequently the aim of this study was to assess and compare the quality of the apheresis process and the stem cell product in HSC donors subjected to the procedure by Optia and Cobe Spectra devices. We have tested 31 Optia and 49 COBE Spectra procedures or PBSC collection. Results: The following results were obtained: · Signifi cantly higher volume of product with the Optia, median 298 ml (108-440) vs. 180 ml (135-280) with the Spectra. · Signifi cantly lower absolute number of WBC with the Optia, median 551 x 10 8 (167-1032 ) vs. 841 x10 8 ( 216-2141) with the Spectra. · CD34 collection effi ciency (CE2): median CE2 is similar in both instruments (45.3% by the Optia vs. 47.6% by Spectra), and average CE2 is lower with the Optia (45.9% +/-13.3 by Optia vs. 59.4% +/-30.5 by the Spectra). · In both instruments a slight decrease in CE2 was noted as a function of increasing CD34 cells pre collection, although this decrease was less pronounced with the Optia instrument. · The process collection with the Optia requires about one hour more. Discussion:

1. Both Spectra and Optia devices satisfi ed CD34 yield effi ciently although a greater variability in collection (CE2) was observed with the COBE Spectra (median vs. average CE2). S213 2. High product volume using the Optia is signifi cant and is of concern in respect to the number of frozen bags to be stored and later on transplanted to patients particularly due to DMSO exposure.

3. The lower total WBC in the product is clearly an advantage of the Optia.

4. The longer time needed with the Optia instrument is inconvenient for both the donors and the operating staff . The results presented here should be taken with some precautions since the operator experience and performance with the Optia are considerably less than with the COBE Spectra in our institution. At this point in time, both devices were found suitable for PBSC collections Yet, the high product volume obtained using the Optia, should be resolved possibly by reducing plasma fl ush volume. Disclosure of Interest: None Declared.

Introduction: The optimal mobilization strategy prior to autologous stem cell transplantation (auto-SCT) for patients with lymphoma is yet to be determined. Materials (or patients) and Methods: We reviewed our institutional experience using chemomobilization with high-dose (HD) etoposide (1.6 gr/m 2 ) and G-CSF (300 μg/day) in 79 patients with lymphoma. The majority (76%) had received at least two prior regimens of chemotherapy and 12 (15.2%) patients had previously failed to mobilize following HD cyclophosphamide or DHAP or ICE with G-CSF. Results: HD etoposide and G-CSF chemomobilization resulted in successful collection (>2 x 10 6 CD34+ cells/kg) in 82.3% of patients within a median 2 (1-6) apheresis days. Patients had stem cells collected between days +8 and +15, with a median +12 day. Median total CD34+ cells/kg collected was 5.95 x 10 6 (0. 1-36.8 ). Seventy -one percent of patients yielded > 2 x 10 6 CD34+ cells/ kg in ≤ 2 days of apheresis, and were defi ned as good mobilizers. While median CD34+ cells/kg collected for good mobilizers was 7.6 x 10 6 , it was 3.9 x 10 6 for poor mobilizers (P<0.001). This regimen was safe with a low rate of febrile neutropenia (7.6%) and acceptable rates of RBC (40.5%) and platelet transfusions (22.8%). Hematopoietic recovery after auto-SCT was achieved on expected time. Therapy -related myelodysplastic syndrome/acute myeloid leukemia occurred in only one patient (1.3%) with in a median follow-up of 16 months after chemomobilization. Patient's characteristics, effi cacy and safety of the chemomobilization are summarized in Table 1 -4. Discussion: In conclusion, HD etoposide (1.6 gr/m 2 ) and G-CSF (300 μg) chemomobilization in the majority of heavily pretreated lymphoma patients appears to result in effective, tolerable and safe stem cell collection. This regimen seems to overcome the effects of most of the prior high-risk features frequently associated with poor mobilization. Large, prospective, randomized studies are needed to analyze whether 'on-demand' use of plerixafor in chemomobilized lymphoma patients with etoposide decreases mobilization failure rate without increase in cost. Disclosure of Interest: None Declared. Introduction: Autologous peripheral blood stem cell transplantation is a 'rescue' of patient's self hematopoietic stem cells (HSC) from myeloablative eff ects of chemotherapy or irradiation. These HPSC are either cryopreserved or liquid preserved and reinfused to repopulate the patient's marrow after myeloablation. Cryopreservation of HSC using 10% dimethyl sulphoxide (DMSO) as a cryoprotectant under liquid nitrogen storage conditions (-196°C) for long-term usage is a well-established procedure while liquid preservation of stem cells is usually done for storage at 4°C for 36 to 96 hours. Aim of the study was to compare the outcome of autologous stem cell transplantation using cryopreserved or non-cryopreserved stem cells. Materials (or patients) and Methods: Twenty adult patients younger than 60 years were enrolled in this study. They had undergone autologous peripheral stem cell transplantation. They had been stratifi ed into 2 groups: Ten patients received cryopreserved autologous PBSCT and the remaining ten patients received non cryopreserved autologous PBSCT. With a follow up period of 6 months to 55 months. Results: No signifi cant diff erence was detected in engraftment of neutrophils and platelets between both treatments groups. When we compared cryopreserved and non cryopreserved groups for clinical outcome, there had been a longer overall survival (OS) and disease free survival ( DFS) in favor of the cryopreserved group but this had not been translated into statistically signifi cant values. In addition, there was no signifi cant diff erence as regard recurrence of disease. However, non cryopreserved group had shorter hospital stay and was less costly than cryopreserved, 8,149 $ versus 8,900 $ respectively. This could be attributed to the fact that liquid nitrogen was not used, shorter hospital stay (P value 0.000) thus lower cost for medications, laboratory tests and procedures; however this was not translated into a statistically signifi cant value. However, cryopreserved was found to be cost-eff ective as regards keeping the patients alive for those who cannot tolerate to proceed to autologous stem cell transplant following mobilization immediately. Discussion: Non cryopreserved therapy required less number of aphaeresis sessions, is associated with shorter hospital stay and is less costy than cryopreserved therapy and give us the possibility of transplantation of patients with hepatitis without the need for buying a special liquid nitrogen tank. Disclosure of Interest: None Declared. S214 lymphoma. Peripheral blood stem cells have become the main source for the ASCT worldwide, because of its advantages over bone marrow. Several risk factors have been identifi ed for poor stem cell mobilization, and diagnosis of lymphoma is one of the most important ones, with an inadequate stem cell harvest in 4 to 25% of the cases. Even though stem cell mobilization in relapsed lymphoma patients can be relatively diffi cult, mobilization strategies have not been standardized and there is a signifi cant variation amongst centers. The aim of this study was to review the mobilization strategies used by EBMT centers in relapsed lymphoma and to evaluate the failure rates. Materials (or patients) and Methods: EBMT centers were invited to participate in this non-interventional prospective clinical study that was started in 2010. Centers were requested to collect data on all consecutive patients with relapsed lymphomas considered to be candidates for an ASCT. Data collected included age, sex, diagnosis, number of prior chemotherapy regimens, mobilization regimen, collected CD34+ cells, marrow harvests. Results: In total, 217 patients with relapsed lymphomas from 30 EBMT centers were included in this study; 135 patients (62%) with non-Hodgkin's lymphoma (NHL) and 82 patients (38%) with Hodgkin's lymphoma (HL). There were 124 males and 93 females with a median age of 54 (range 19 -77) years. Median number of chemotherapy lines received before this relapse was one (range 1-8). Twohundred-and-seven patients (95%) were mobilized with chemotherapy and cytokines, being DHAP (36%) and ESHAP (13%) the most frequent protocols at fi rst mobilization, and 11 patients (5%) were mobilized with cytokines alone. All patients used G-CSF at the fi rst mobilization. Ten patients (5%) were at fi rst mobilized with G-CSF, but switched to plerixafor (PLX) during the fi rst mobilization. These were all patients that were mobilized with chemotherapy as well. Twenty-seven patients (12.4%) failed to mobilize adequate stem cells (<2 x 10 6 CD34+ cells/kg) during fi rst mobilization. Four of those patients received PLX. The median number of stem cells collected at fi rst mobilization was 5 x 10 6 CD34+ cells/kg (range: 0 -82). In 198 patients (91%) only one mobilization course was given, 17 patients (8%) had two mobilization courses, 2 patients (1%) underwent three courses. Three patients had a mobilization failure after only G-CSF; they all were successful in a second attempt after chemotherapy. Five of the patients failing the fi rst mobilization with chemotherapy received PLX at second mobilization, but only three of them were successful. One patient failed both fi rst and second mobilization and received PLX at third mobilization without success. Nineteen patients (8.7%) still had an inadequate amount of stem cells after those mobilizations. Of those, only 4 patients (2%) underwent bone marrow harvest. Discussion: Mobilization strategies for patients with relapsed lymphoma are very diverse but more than 90% of them are mobilized using the combination of salvage chemotherapy plus G-CSF. PLX was used in less than 10% of the mobilization procedures during the time period analyzed. In our experience, the failure rate was 12.4% after the fi rst mobilization attempt and 8.7% after several attempts. Disclosure of Interest: None Declared. Introduction: Biosimilar Granulocytre Colony-Stimulating Factor (G-CSF) has been approved on the basis of comparable quality, safety and effi cacy as the originator product. So far, biosimilar G-CSF Zarzio® has been approved also for autologous peripheral blood stem cell (PBSC) mobilization and for prophylaxis of sever neutropenia duration following conditioning chemotherapy and stem cell infusion. However, there is still general skepticism about safety and effi cacy of Zarzio® in these setting of patients. Materials (or patients) and Methods: From March to November 2013, 22 consecutive adult patients with hematologic malignancies (acute leukemia n=5, lymphoma n=13 and multiple myeloma n=4) underwent autologous PBSC mobilization after administration of chemotherapy associated to biosimilar Filgrastim (Zarzio®) in our Institution. Zarzio® was administered according to the study protocol in which patients were enrolled for acute leukemia (5 mcg/Kg/day in 2 patients and 10 mcg/Kg/day in 3 patients), at dosage of 10 mcg/Kg/day for multiple myeloma and 5 mcg/Kg/day for lymphoma. The target of CD34+ cell dose was 4 x 10 6 /Kg recipient body weight. This cohort of 22 patients was retrospectively compared with 53 consecutive patients (acute leukemia n=2, lymphoma n=28 and multiple myeloma n=23) who underwent autologous PBSC mobilization after administration of chemotherapy associated to Lenograstim (Myelostim®) at the same dosage from March 2011 to February 2013. Results: The two groups of patients were similar as baseline clinical features, including sex (P=1), age and body weight at leukapheresis (P=0.124 and 0.357, respectively), bone marrow involvement and disease status at leukapheresis (P=0.451 and 0.501 respectively), previous radiotherapy (P=0.551), previous chemotherapy lines (P=0.977) and mobilization regimes of chemotherapy received (P=0.198), with the only exception for diagnosis distribution (high rate of acute leukemia in the Zarzio® group; P=0.014). As for PBSC collection data, median days of G-CSF administration, median CD34+/mcL number at leukapheresis and median number of CD34+ x 10 6 /Kg collected at fi rst leukapheresis were similar between two groups of patients. However, in group of patients who received Zarzio®, we observed an higher rate of mobilization failures (22.7% compared to 3.8% of Myelostim® group; P=0.02), an higher rate of patients unable to reach the target of CD34+ cell planned dose (40.9% vs 7.5%; P=0.001) and an higher (but not statistically signifi cant) rate of patients needing Plerixafor administration (13.6% vs 1.9%; P=0.076). We not observed any adverse eff ect directly related to Zarzio® administration. Discussion: Despite the limitation due to the low number of patients, our data suggest that Zarzio® could be less eff ective when compared with Myelostim® for PBSC mobilization after chemotherapy in adult patients with hematologic malignancies and more expensive considering the high rate of patients needing Plerixafor administration. Further studies on a larger number of patients are warrant to better evaluate the role of Zarzio® in this setting. Disclosure of Interest: None Declared. 3 Hematology/Oncology, University Chidren's Hospital, Jena, Germany, 4 Hematology/Oncology, University Children's Hospital, Graz, Austria, 5 Scientifi c Development, Fresenius Biotech GmbH, Planegg Martinsried, Germany Introduction: T and B cell depletion of haploidentical peripheral stem cells with CD3/CD19 coated magnetic microbeads eff ectively prevents from GvHD and allows to coinfuse large numbers of donor NK cells and other accessory cells Additional in vivo depletion of the graft with serotherapy is not mandatory. Thus, a dosing scheme of ATG is needed, which provides a rejection prophylaxis by depleting lymphocyte subsets of the patients but which does not harm the stem cell graft. We present data with reduced ATG doses given at the beginning of the conditioning regimen in order not to impair cotransfused NK cells and immune recovery. Materials (or patients) and Methods: A total of 33 pediatric patients received 3x10 mg/kg ATG-F (Fresenius) starting at day -12, followed by fl udarabine (160 mg/m 2 , d -8 to -5), thiotepa (10 mg/m 2 d -4) and melphalan (140 mg/m 2 d -3 to -2). Most patients received MMF untill day 30-60. ATG serum levels were measured in 19 patients by fl ow cytometry (amount of ATG binding to the Jurkat cell line, defi ned as T cell specifi c rabbit IgG). Apoptosis/ necrosis of donor NK cells was assessed with Annexin V/PI staining and fl ow cytometry. Results: Median time to ANC>500 and to independence from platelet transfusion was 9.5 and 13.5 days. Graft rejection occurred in 7/33 patients (21%). All rejectors could be rescued with reconditioning and 2 nd stem cell donation or with infusion of autologous back ups. Acute GvHD grade I and II-IV was observed in 12/26 (46%) and 2/26 (8%) patients without rejection, respectively. Limited (extensive) chronic GvHD occurred in 5/24 (1/24) evaluable patients. Median peak levels of 22.1 μg/ml specifi c rabbit IgG were reached between day -8 and -6 and dropped to 2.9 μg/ml at day 0. In vitro incubation of NK cells from healthy donors with a comparable dosage of ATG-F (1 μg/ml) resulted in 26 % apoptosis and 0.2 % necrosis (70% vital cells) after 24 hours. Moreover, NK cells were incubated with patient´s serum, taken after ATG treatment and adjusted to a concentration of 1 or 2 μg/ml. Cell death occurred in 20% of NK cells each. Functional activity was measured against K562 targets. Specifi c lysis of decreased from 83% to 71% (corresponding to a loss of 15% activity) after incubation with 2μg/ml and to 11% (87% loss of activity) with 1000μg/ml. Immune recovery was monitored and compared with a historical group receiving OKT3 and the same chemotherapy (n=34). Recovery of CD56+ NK cells was fast with a mean number of 414 vs. 232 cells/μl (ATG group vs. OKT3 group, P<0.01) at day +14, 252 vs. 342 cells/μl at day +30 (P=0.1) and 189 vs. 217 cells/μl at day +90 (P=0.3). CD3+ T cells reached 36 vs. 12/μl at day +30 and 189 vs. 225/μl at day +90 (no signifi cant diff erences for all data pairs). Discussion: Conclusions: ATG-F was started at an early time point, resulting in low serum levels of specifi c ATG at day 0 and in a fast NK cell recovery. In vitro results suggested, that the majority of NK cells will not be damaged herewith. Immune recovery of T and NK cells was comparable to that of a historical control group who received OKT3. However. The rejection rate was higher than expected and an increase of the ATG dose has to be considered. This approach will be also of interest for other transplantation strategies in which various components of the grafts and additionally given T cells have to be preserved. Introduction: Biosimilars of fi lgrastim have been available since 2008 and are now in widespread use for the prevention of chemotherapy-induced neutropenia and stem cell mobilisation. However, some professional bodies have raised concerns over their use in healthy donors for allogeneic mobilisation given the lack of clinical data in this setting. Here we report the use of biosimilar fi lgrastim compared with a matched historical control group in allogeneic transplant. Materials (or patients) and Methods: A total of 26 healthy relateddonors (parent or sibling) received biosimilar fi lgrastim (Zarzio®) for allogeneic stem cell mobilisation at a single centre between 2011 and 2013. Donors and recipients were compared with a matched historical control group (n=48) who had been treated with original fi lgrastim (Neupogen®) at the same centre between 2005 and 2011. Donors and patients in both groups were treated according to the same clinical protocol, with G-CSF 10 μg/kg/day administered to donors on days 1-5 with leukapheresis performed on day 5. Results: Donor and recipient characteristics (age, gender, body weight) were similar in both groups. Both the biosimilar and originator groups had similar donor mean white blood cell counts at baseline (6.13 vs 6.24 x 10 9 /l) and on day 5 (46.9 vs 45.3 x 10 9 /l). Mean donor CD34+ cell count on day 5 was 92/μl in the biosimilar group and 88/μl in the originator group (P=0.713). Median number of CD34+ cells per recipient body weight was 9.7 x 10 6 in the biosimilar group and 8.00 x 10 6 in the originator group (P=0.437). Occurrence and intensity of bone pain was similar in donors in both groups. The majority of recipients in both groups had acute leukemia, myeloma, lymphoma or congenital immunodefi ciency syndrome and around half underwent a non-myeloablative transplant. Median time to neutrophil engraftment (>500/μl) was similar in both the biosimilar and originator groups (16.5 [range 11-44] vs 15.0 [range 9-23] days; P=0.078), as was platelet recovery (>20 g/l) (12.5 [range 8-28] vs 12.5 [range 3-38] days; P=0.990). Acute graft-versus-host disease (GVHD) occurred in 27% of patients in the biosimilar group and 38% patients in the original group, while chronic GVHD occurred in 15% and 19% of recipients, respectively. Discussion: Biosimilar G-CSF is as eff ective and well tolerated as originator G-CSF for related-donor allogeneic stem cell mobilisation. Long-term follow-up of donors is required to confi rm the safety of biosimilar and originator G-CSF. Introduction: Plerixafor, a CXCR4 antagonist, can be used with G-CSF or chemotherapy plus G-CSF mobilization (chemomobilization) to enhance the mobilization of hematopoietic stem cells. Previous studies have shown that plerixafor have impact on cellular composition of the collected graft. Some characteristics of the blood stem cell grafts (e.g. stem cell primitivity and lymphocyte content) have been linked with engraftment and patient outcome after ASCT. Materials (or patients) and Methods: Thirty-seven patients with NHL were included in this prospective GOA study (Graft and Outcome in Autologous Stem Cell Transplantation). All patients received chemomobilization. Fourteen patients (38 %) mobilizing poorly received plerixafor to enhance mobilization (plerixafor group). Twenty-three patients served as controls (control group). The samples of each cryopreserved leukapheresis products S216 were studied by fl ow cytometry. CD3/CD8/CD45/CD4 and CD3/ CD16+CD56/CD45/CD19 antibodies with BD Trucount tubes were used to determine the absolute counts of T, B, and NK cells as well as the CD4 and CD8 subpopulations of T cells. Stem cell primitivity was investigated by using the following antibodies: CD34, CD38, CD133, CD19, CD117 and CD45. Results: The median number of aphereses was two in the plerixafor group and one in the control group (P = 0.006). The median of total stem cell yield was 2.8 x 10 6 /kg CD34 + cells (range 1.9 -5.1 x 10 6 /kg) in the plerixafor group and 3.9 x 10 6 /kg CD34 + cells (range 2.0 -16.8 x 10 6 /kg) in the control group (P = 0.017). The median proportion of the primitive stem cells (CD34 + CD133 + CD38 -) cells from all CD34 + cells was signifi cantly higher in the plerixafor group when compared to the control group (3.0 % vs. 2.1%, P = 0.049) but there was no signifi cant diff erence in the total amount of these cells (medians 0.07 x 10 6 /kg vs. 0.05 x 10 6 /kg, P = 0.914). The median amounts of total CD3 + T cells (152 x 10 6 /kg in the plerixafor group vs. 64 x 10 6 /kg in the control group, P = 0.001), helper (CD3 + CD4 + ) T subsets (76 x 10 6 /kg vs. 38 x 10 6 /kg, P = 0.001), suppressor (CD3 + CD8 + ) T subsets (73 x 10 6 /kg vs. 22 x 10 6 /kg, P = 0.002) and NK (CD3 -CD16/56 + ) cells (19 x 10 6 /kg vs. 5 x 10 6 /kg, P <0.001) in the graft were all signifi cantly higher in the plerixafor group when compared to the control group. CD19 + cells were apparent only at in a few patients and there was no diff erence between the groups (P = 0.998). Discussion: Plerixafor added to chemomobilization in poor mobilizers results in collection of higher amount of lymphocytes. In addition, the proportion of the most primitive stem cells in the graft is higher. The clinical relevance of these observations will be studied in regard to hematopoietic and immune reconstitution as well as progression free survival in this prospective GOA study. Disclosure of Interest: None Declared. Introduction: Since 1994 we have been collecting over ten thousand peripheral blood stem cells (PBSC) in an allogeneic and autologous setting using the COBE Spectra apheresis device, mainly using the MCN program. As our large center embraces innovation, we have recently transitioned to Spectra Optia in March 2012 and compared both machines in a similar setting. Materials (or patients) and Methods: 68 Patients were mobilized with chemotherapy + G-CSF 7 days before apheresis. 38 of them underwent procedures using Spectra Optia; 30 patients were put on COBE Spectra. In addition, PBSC from 255 normal donors, receiving only G-CSF, were collected on the Spectra Optia and compared with 30 donors on COBE Spectra. For all subjects, precounts and post-counts of total blood cells and CD34+ cells were obtained. From apheresis products, the same quality control parameters were obtained. Both apheresis devices were run in a consistent fashion, depending on the setting (auto or allo). With Spectra Optia, a collection preference of 20 (in allo) and 20 to 40 (in auto) was used with a chamber fl ush of 16 ml and chamber chase of 2 ml. Collection phases were carried out every 500 to 700 ml in an allogeneic setting and in the autologous setting were determined by the device. For COBE Spectra, a collect Hct of 2-3 % was maintained and collections were done at a collect fl ow of 0.8 ml/min. All results were analyzed using a non-parametric Mann-Whitney U test. Results: Pre-procedure and post-procedure blood counts for all patients and donors were comparable between the two device groups, except for the recipients body weight which was higher in the COBE Spectra group (76 (7-130) kg vs 85 (58-105) kg; P=0.0068). In the allo setting we could process less blood on Spectra Optia (11.3 (3.5-15.4 ) L vs 13.2 (7.3-17 .0) L) to get to the same CD34+ target dose (7.5 (0.13-224) x 10 6 vs 8.1 (1.5-21.9) x 10 6 c/kg). However, the collection effi ciency (CE1) was similarly high (82 (5-204) % on Optia; 83 (49-151) % on Spectra). In the auto setting, we obtained similar results: less blood was processed to get to a higher transplant dose. However, the CE1 was similar, also in this setting: 85 (48-239) % on Optia vs 85 (46-184) % on COBE Spectra). Interestingly, we observed a much higher CD34+ dose/ kg obtained per liter that we processed on Spectra Optia, than on COBE Spectra (0.67 (0.10-4.6) vs 0.46 (0.06-2.5) x 10 6 /kg/L; P=012) in the autologous setting. The RBC contamination of Spectra Optia products was much lower, both in the allo (7 (1-22) ml vs 15 (3-52) ml) and auto setting (5 (2-16) vs 6 (3-13)). The large diff erence in the allogenic setting would be a benefi t for ABO-incompatible transplants. Discussion: The high collection effi ciency values on both devices enabled us to get high transplant doses in a short time and by processing relatively low amounts of blood. Spectra Optia appeared superior in collecting higher doses per liter of blood processed, especially in the autologous setting. Product cross cellular contamination was low, especially in the amount of RBC collected on Spectra Optia. Taking all these data together, we intend to move forward only with Spectra Optia in our large transplant setting. Disclosure of Interest: None Declared. Introduction: Plerixafor is approved for autologous peripheral blood stem cell mobilization in patients with Non-Hodgkin Lymphoma (NHL) or Multiple Myeloma. This agent may reduce the failure rate and/or the number of apheresis procedures required without increasing the toxicity, and this may reduce total transplant costs. With this background, the aim of this prospective non-interventional analysis is to assess resource utilization to document provider costs associated with peripheral stem cell mobilization and apheresis. Of note, the study aims to evaluate the impact on the time and eff ort associated and costs to the hospital when using plerixafor (P) with a primary analysis to compare measures of time/eff ort from patients drawn from the Pre-P versus P eras. Materials (or patients) and Methods: The study population includes NHL patients undergoing peripheral blood stem cell mobilization. Part I of the study is a retrospective medical record review study of 200 NHL patients from 7 centers in France and Germany. Selected patients will be evenly divided between two eras: 1) prior to approval of plerixafor = Pre-P era (until June 1, 2009) 2) after approval of plerixafor = P era (July 1, 2010 and onwards). Part II of the analysis is an ongoing prospective study consisting of time/motion evaluation of actual apheresis -20 events at each center. The actual apheresis events are being measured in consecutive patients scheduled to be candidates for peripheral blood stem mobilization. Outcome measures include number of visits for administration of mobilizing agents; duration (days) of administration of mobilizing agents; agents used as mobilizing agents; adverse events detected during mobilization; number of apheresis sessions; hours of apheresis sessions; attainment of CD34+ target (yes, no); days until CD 34+ target level was met. In addition, time-motion assessments will be obtained retrospectively (Part I) and concurrently (Part II) and included the total time to prepare the patient, perform apheresis and manage adverse events. Costs will be evaluated and quantifi ed in terms of micro-costing group interviews with local hospital administration. The primary study end point is diff erence in mean time to perform apheresis (including apheresis related adverse events, if any) and costs to the hospital in terms of micro-costing per patient, between patients in the Pre-P versus P eras. Results: At the time of abstract submission, data collection is ongoing and results will be presented during the meeting. The key fi ndings of this study will demonstrate the favorable impact of novel interventions on the number of apheresis procedures required to reach a target peripheral blood stem cell, and failure rate of mobilization, thus translating into reduced total transplant costs without increasing the toxicity. Discussion: The fi nancial implications for transplant centers would be signifi cant and would pave the way for further studies aiming to optimize staff time and resource utilization related to apheresis in real-world practice. Introduction: Allogeneic hematopoietic stem cell transplantation (HSCT) is a successful treatment strategy for hematopoietic malignancies and inborn errors of metabolism or the immune system. In about two thirds of the patients no suitable related donor can be identifi ed. A suitable unrelated HLA-matched donor can be found through large international donor registries. Cell grafts from unrelated donors are almost always collected at distant collection sites and storage and transportation of cell grafts becomes a crucial link in the transplantation process. Several studies have investigated the impact of factors infl uencing the graft quality, such as liquid storage and transportation Materials (or patients) and Methods: We studied the infl uence of graft quality on clinical outcome in 144 patients treated with allogeneic SCT. As a measurement of the graft quality we used the viability measured by 7AAD on a frozen/thawed sample from the PBSC graft in %.

Results: There was great variation in the viability of the frozen/ thawed samples (median viability 64%; range 24-96%). When comparing the viability using 7AAD on freshly arrived PBSC grafts compared to frozen/thawed vials no clear correlation could be observed. More patients who received PBSC with inferior quality (viability <64% on the frozen/thawed sample) developed acute graft versus host disease (aGVHD) of any grade than patients receiving grafts with better quality, 71% and 57% respectively (P=0.025). This also is refl ected by the fact that Transplant related mortality (TRM) were 22% in the group receiving grafts with a viability of <64% on the frozen/thawed vial compared to only 8 % in the patient group with better graft quality (P=0.03).

To measure the impact of graft quality on viral complications with a relatively early onset after HSCT we decided to analyze EBV-PTLD and CMV reactivation. Of the patients receiving grafts with inferior quality, 61% of the patients suff ered from cytomegalovirus (CMV) reactivation as compared to 44% in the group receiving grafts of a better quality (P=0.05). Discussion: We have found that poor graft quality, measured as viability on frozen/thawed samples from PBSC grafts, is associated with increased occurrence of acute GVHD. Also, patients receiving grafts with inferior graft quality have more CMV infections. There is a need for better analyses for assessing graft quality in routine transplantation care. Our study also suggests that clinical follow-up of cell collection and processing as measurements of quality needs to be more elaborate, Engraftment-and survival data is not suffi cient as quality markers. Disclosure of Interest: None Declared. S218 suggest SCT can be performed safely on patients (pts) well into their 70s, which is critical since the median age of pts with MM is 70 and in NHL it is 66. Materials (or patients) and Methods: We review results of 2 Ph 3 trials (RCTs) utilizing P upfront + G vs. G alone to collect stem cells and results of the compassionate use protocol (CUP) of P+G to mobilize stem cells in those who failed prior collections. We compare collection results and safety in pts ≥65 or <65 years of age (yrs Overall, there were more women in the pts ≥65; other baseline characteristics were similar. As seen in Table 1 , 84% of MM pts ≥65 yrs collected suffi cient cells to proceed with single SCT and was similar to pts <65 yrs (86%). Median days of apheresis to reach this target was 3 (range 1-8) and 3 (range 1-7). The number of MM patients who proceeded to SCT was similar between the 65< and ≥65 yrs pts (83 vs 85%) and the number of NHL pts who proceeded to SCT was similar for pts 65< or ≥65 yrs (79 and 76%, respectively). In NHL pts, 70% of pts <65 and 64% of pts ≥65 yrs collected suffi cient cells to proceed with single SCT. Median days of apheresis to reach this target was 3 (range 1-10) and 3 (range 1-8). Overall, there was a higher percentage of pts who mobilized suffi cient cells to proceed to SCT when treated with P+G upfront rather than after failing mobilization using standard techniques regardless of age. There was no diff erence in engraftment to WBC or Plts between those ≥65 or <65 yrs. In the RCTs, ≥Gr 3 febrile neutropenia was slightly higher in pts ≥65 yrs that received P+G (7.6 vs. 4.0% G alone), but similar to those <65 yrs who received P+G (6.9%). Other adverse events (AEs) were similar between the <65 and ≥65 pts in the RCTs. AEs were similar in the CUP pts 65< and ≥65 yrs. Discussion: Although age has traditionally been a risk factor for poor mobilization, these data suggest that with the use of P+G there is no diff erence in the ability to mobilize or transplant MM and NHL pts ≥65 yrs. Engraftment seems no diff erent for pts who proceed to SCT. Generally, AEs appear similar when comparing age and P+G vs G alone. Age alone should not be a consideration for SCT in pts with MM or NHL. Introduction: Depletion of CD45RA+ naïve T cells from donor lymphocyte infusions (DLI) aims at the reduction of alloreactivity while preserving immunocompetent memory T-cells to provide pathogen-specifi c reactivity. Here, we investigated the performance of the CliniMACS system for depletion of CD45RA+ T cells from DLI intended for therapeutic or pre-emptive antiviral boost, following TCRαβ+ depleted haploidentical hematopoietic stem cell transplantation (HSCT). Materials (or patients) and Methods: Whole-blood-derived buff y coats were obtained from two HLA-haploidentical donors. The cells were labelled with CD45RA reagent and processed with the CliniMACS device using the D3.1 program with the DTS (Deple-tion Tubing Set). In fl ow cytometric analysis viable CD45RA+ T cells were defi ned by their cell scatter properties of lymphocytes, positivity for CD3 and CD45RA, and negativity for propidium iodine staining, as well as lack of CD13-, CD19-, CD14-and CD45RO-expression. CD3+CD45RA-and CD3+CD45RO+ T cells were further defi ned as CD4+ or CD8+, respectively. Aliquots of CD45RA-depelted products containing 25 × 10 3 CD3+ cells per kilogram of recipient weight were prepared for fresh DLI and cryopreservation. Results: The log 10 CD3+CD45RA+ cell depletion rate was 3.84 and 3.85, corresponding to 0.0049 % and 0.0071 % residual CD3+CD45RA+ cells, respectively, in the target fractions. The recovery of CD3+CD45RA-CD45RO+ cells was 51% and 54% and the ratio of CD4+ to CD8+ cells was increased from 2.5 to 9.7 and 1.8 to 6.8. Two patients following TCRαβ+ depleted haploidentical HSCT with symptomatic (patient#1; CMV, VZV) and asymptomatic (patient#2; CMV, AdV) viral infection, received a fresh CD45RA+ depleted DLI of 25 × 10 3 CD3+ cells /kg from their original HSCT donor, at day +84 (patient#1) and +70 (patient#2), respectively. Patient#2 received a second dose of 50 x 10 3 CD3+ cells /kg on day +91. At the time of DLI, peripheral blood CD3+CD4+ and CD3+CD8+ levels were < 10/ul. Following the fi rst DLI, the time to clearance of VZV (patient#1) and AdV (patient#2) viremia was 2 and 8 weeks, respectively, whereas decreasing levels of CMV in peripheral blood were detected at 3 (patient#1) and 5 weeks (patient#2). Peripheral blood CD3+CD4+ and CD3+CD8+ levels reached > 100/ul at 3 (patient#1) and 9 (patient#2) weeks post DLI. There was no GvHD before or after DLI. Discussion: The ClinMACS system (D3.1 program, DTS) enables extensive depletion of naïve CD45RA+ T cells from DLI products, which can be safely used for therapeutic or pre-emptive antiviral boost following TCRαβ+ depleted haploidentical HSCT. Disclosure of Interest: None Declared. Introduction: In 2009, Plerixafor was granted Marketing Authorization by EMA, to be used in patients with lymphoid malignancies and who "…mobilize poorly …" in response to conventional mobilization treatment. Since there is no consensus across collection and transplant facilities on the defi nition of "poor-mobilization", and since the price tag of plerixafor is high and signifi cantly adds on the cost of the collection and transplant procedure, French competent authorities as represented by the "Comité Economique des Produits de Santé" (CEPS), commissioned a nationwide survey on the use of plerixafor in France. Genzyme-Sanofi promoted constitution of the registry where data were collected on mobilization and collection; data analysis was conducted by a committee of experts, who here report on the results of this observational study. Materials (or patients) and Methods: Contacts were established with haematologists/transplant physicians and their corresponding apheresis practitioners at all transplant programmes that annually report their activity to the "Agence de la Biomédecine" (ABM), the public health authority that monitors and supervises cell, organ and tissue transplantation in France. Results: Thirty-three programmes/collection facilities (out of 37 registered with the ABM) reported on 262 patients (Male: 148, Female: 114; median age: 57, range: 2-72) who were prescribed plerixafor between November 2011 and November 2012. Patients were informed and provided written consent to access their health records. As expected, the vast majority of enrolled patients were treated for lymphoid malignancies; however, a small proportion of patients (n= 21 or 8%) received plerixafor for other types of diseases where high-dose chemotherapy was deemed an indication. One hundred and sixteen patients had previously failed at least one attempt for mobilization/collection, while 146 patients received plerixafor in the course of their fi rst attempt for mobilization/collection, using a "pre-emptive" strategy to immediately overcome poor or insuffi cient mobilization. Enumeration of circulating CD34 + cells was widely used as a surrogate marker for stem cell mobilization: 67% and 78% of patients treated with plerixafor in the "failed collection" and "pre-emptive" subgroups had < 15 CD34 + cells/mL of peripheral blood immediately before introduction of plerixafor. Patients experienced a median 3.89 (0.3; 52.0) fold increase in the number of circulating CD34 + cells. Two hundred and fi fty fi ve treated patients reached the targeted number of collected cells; a median number of 3.66x10 6 CD34 + cells/kg (0.1; 16.4) was collected after introduction of plerixafor (and was eventually combined and transplanted with previously collected cell products). Discussion: This "real-life" survey -that includes a signifi cant proportion of candidate patients for plerixafor at a national level -suggests that prescriptions have not exceeded the expected fi gures, despite variations in criteria used to defi ne poor-mobilization; plerixafor improved upon the mobilization status, and allowed for effi cient collection of a blood graft for most treated patients. 3 Clinical Immunology and Transfusion Medicine, University and Regional Laboratories, 4 University of Lund, 5 Hematology, University Hospital Lund, Lund, Sweden Introduction: Processing of peripheral blood progenitor cells (PBPC) for clinical transplantation or research applications aims to eff ectively isolate or deplete specifi c cell populations. We have previously reported the use of a novel ultrasound-based sorting technology (acoustophoresis) for sorting of platelets and CD4+ cells from PBPC products. Utilizing a further improved sorting device, this study aimed to optimize affi nity bead acoustophoresis for smaller cell populations, in this case CD8+ cells. Materials (or patients) and Methods: PBPC samples (n=16) were obtained from patients and healthy donors. Following density gradient centrifugation, mononuclear cells were labelled with anti-CD8 microbeads (Dynal) and sorted either on an acoustophoresis-microchip or magnetically. PBPC samples, target and waste fractions were analysed for purity, recovery, T-cell function and progenitor cell content. Results: PBPC products contained a mean 11.6 ± 7.1% CD8+ cells before sorting. Purities obtained with acoustic sorting of CD8+ lymphocytes were 93.3 ± 6.8% compared to 94.4 ± 8.6% for magnetic sorting (n=16). Viabilities of sorted cells were 97.0 ± 3.9% (acoustic) and 97.5 ± 3.5% (magnetic). Recovery of acoustic sorted CD8+ cells was 57 ± 19% of the total CD8+ cells compared to a median recovery of magnetic sorted CD8+ cells of 43 ± 17%. Leukocyte subpopulation analysis performed after CD8 selection showed a relative increase of CD4 cells in the non-target fractions due to the removal of CD8 cells. Functional testing of sorted CD8+ cytotoxic T cells showed unimpaired mitogen-induced proliferation capacity with proliferation rates of 87.5 ± 7.8% and 64.3 ± 7.7% (acousto) compared with 92.2 ± 12.2 and 69.5 ± 15.2 (magnetic) after 6-day stimulation with CD3/CD28 and ConA, respectively. Furthermore, hematopoietic progenitor cell assays showed a preserved colony-forming ability of post-sorted non-target cells (CFU-C/5,000 seeded cells: 14.4 ± 6.5 pre-sort versus 11.3 ± 5.9 post-sort). Discussion: Acoustophoresis is a promising technology to efficiently sort bead-labelled lymphocyte populations from PBPC samples with high purity and recovery without impairing lymphocyte function. Affi nity-bead acoustophoresis is, thus, an interesting technology for PBPC processing. Disclosure of Interest: A. Urbansky: None Declared, A. Lenshof: None Declared, J. Dykes: None Declared, T. Laurell Confl ict with: co-founder and share holder of Acousort AB, S. Scheding Confl ict with: co-founder and share holder of Acousort AB. Introduction: The receptor (uPAR) of the urokinase-type plasminogen activator (uPA) plays a key role in cell migration processes. uPAR is expressed on the surface of various cell types, both in full-length and cleaved forms, lacking the N-terminal DI domain (DIIDIII-uPAR). uPAR binds uPA and vitronectin (VN) and regulates integrin activity. The receptor can be shed from the cell surface, generating full-length and cleaved soluble forms, suPAR and DII-DIII-suPAR, respectively. Both forms of soluble uPAR have been detected in human plasma and urine. suPAR is still able to bind uPA and VN, whereas DIIDIII-suPAR, when exposing the chemotactic SRSRY sequence (aa 88-92) at its N-terminus, is able to bind receptors for the formyl-methionyl-leucyl phenylalanine peptide (fMLF). We previously demonstrated the involvement of uPAR soluble forms in G-CSF-induced human CD34+ hematopoietic stem cell (HSC) mobilization. Further, we demonstrated that DII-DIII-suPAR can induce mobilization of hematopoietic stem/progenitor cells in mice. Since HSC mobilization and homing are specular processes which utilize same mediators and similar signaling pathways, we investigated whether the soluble forms of uPAR could be also involved in HSC homing. Materials (or patients) and Methods: We examined suPAR and DIIDIII-suPAR expression in cultures of human marrow stromal cells. We then evaluated the levels of both suPAR forms in sera from fi ve healthy donors and from fi ve patients before and after chemotherapy-based conditioning regimens. We also examined the potential eff ects of the diff erent soluble forms of uPAR in long term cultures (LTC) of G-CSF-mobilized CD34+ HSCs, in the presence of suPAR or of the uPAR84-95 peptide, containing the SRSRY chemotactic sequence of DIIDIII-suPAR. Results: Interestingly, stromal cells produced suPAR and high amounts of the chemotactic DIIDIII-suPAR. We found a signifi cant increase in DIIDIII-suPAR levels in sera from patients before conditioning, as compared to healthy donors; however, the chemotherapy-based conditioning regimen signifi cantly decreased circulating DIIDIII suPAR levels. Both suPAR and the uPAR84-95 peptide increased the number of adherent clonogenic progenitors in LTC of G-CSF mobilized HSCs. Discussion: Our results suggest that marrow stroma produces soluble forms of uPAR which could contribute to the engraftment of marrow CD34+ HSC. According with our previous observation on the mobilizing eff ects of DIIDIII-suPAR, the circulating cleaved suPAR seems to be lowered by chemotherapy-based conditioning regimen, likely allowing CD34+ HSC homing. Disclosure of Interest: None Declared. Introduction: Neutropenia after stem cell transplantation may contribute to the severity of infectious complications and may have a signifi cant impact on overall morbidity and mortality in this setting. Granulocyte colony-stimulating factors (G-CSFs), fi lgrastim, is recognised to be useful in accelerating engraftment after autologous stem cell transplantation. Several forms of biosimilar non-glycosylated G-CSF have been approved by the European Medicines Agency, with limited published data supporting the clinical equivalence in peripheral blood stem cell mobilisation and recovery after autologous stem cell transplantation. Materials (or patients) and Methods: With the aim of evaluate the use of G-CSF after autologous stem cell transplantation, we studied retrospectively 124 patients (pts) consecutively treated with a biosimilar of fi lgrastim, at a dosage of 5 mcg/kg/day, given intravenously from day 7, in alternate days, to absolute neutrophil count of 500/mmc for three days. Hospitalization period, number of days with fever, and time to neutrophils and platelets recovery (neutrophils> 500/1000 cells/mm 3 and platelets > 20000 cell/ mm 3 ) were evaluated as primary endpoints. Results: In order to assess the eff ect of fi lgrastim (biosimilar) on alternate days, beginning in D+7, on neutropenia duration and clinical outcome after transplant, 124 patients were included. The median age was 54 (range 18-69) years with 52.4% male and 47.6% female. The baseline malignancies were: Multiple Myeloma (57.3%), Non-Hodgkin's Lymphoma (35.5%), Hodgkin's Disease (6.5%) and others (0.8%). The most frequently received chemotherapy regimens were: high dose Melphalan200mg/m 2 (54%), FEAM (28.2%), BEAM (13.7%), Melphalan140mg/m 2 (2.4%), Melphalan80mg/m 2 (0.8%) and Carbopec (0.8%). The median of CD34+ cells infused were 2.93x10 6 (1-12) cells/kg. Median time to neutrophils> 500 cells/mm 3 12d (9-32); neutrophils> 1000 cells/ mm 3 12d and platelets > 20000 cell/mm 3 12d (6-42); days of fever: median 4d (0-18). Length of hospital stay: median 16d . No deaths related to transplant. The comparison with our previous practice (fi lgrastim (Neupogen 5 mcg/kg/day intravenous from day 7 each day to absolute neutrophil count of 500/ mmc for three days.) Discussion: Although the study population was not matched with previously analyzed patients (discrepancy regarding the predominance of each sex, age, prevalent diseases, and conditioning regimens used) the results of hematopoietic recovery were overlapping. The median of CD34+ cells infused was higher, but hospitalization was similar. The diff erence found in time for recovery of neutrophils and platelets appear to have no signifi cant clinical impact. Filgrastim (biosimilar) on alternate days appears to be as eff ective as daily fi lgrastrim (Neupogen) in the context of autologous hematopoietic stem cell transplantation. Disclosure of Interest: None Declared. Introduction: Recombinant G-CSF fi lgrastim (Neupogenâ) has been widely used for mobilization of peripheral blood progenitor cells (PBPC), alone or in combination with chemotherapy. Although developing new cost eff ective mobilization regimens is important, the experience with biosimilar G-CSF agents for this indication has been limited by concerns about biological equivalence and unidentifi ed functional diff erences, and scarce data have been published regarding the use of biosimilars in this context. In 2012, after a 2-month experimental period, we incorporated biosimilars in our mobilization protocols with the aim of prospectively assessing their effi cacy and safety. Materials (or patients) and Methods: From April 2012 to June 2013 fi lgrastim biosimilars (Zarzio®, Sandoz and Nivestim®, Hospira) were used in 70 consecutive PBPC collections in 64 patients (pts), with no other change in mobilization regimens nor in equipment. Collection and engraftment parameters were prospectively registered. A series of 72 consecutive collections in 64 pts mobilized with Neupogen® from April 2011 to June 2012 were used as the comparator group. A collection is defi ned as the PBPC harvested from a single mobilization procedure, regardless of the days of apheresis required to collect them. Results: The study and control cohorts were well matched for median age (56/55), underlying disease (multiple myeloma 47/45%, lymphoma 48/45%), retransplant candidates (3 pts vs 5) and number of previous lines of therapy (2 ou more 41/53%, n.s.). Most pts were mobilized with cyclophosphamide + filgrastim (81/71%, n.s.); poor mobilizers received etoposide (4/8%) or plerixafor (10/14%) + filgrastim. Side effects attributable to filgrastim were minimal and transient in both groups, consisting of transient myalgias/arthralgias (which were grade 3 in 6/4%) with no reports of unexpected events. The number of days of apheresis required to reach the target CD34+ count did not differ (with a single apheresis in 58/46%, n.s.). The collection was considered inadequate in 7/15%. CD34+ counts in pre-apheresis PB were similar (median 32/24/ul) as were total CD34+ collections (median 3.79 vs 2.51 x10 6 /kg, both n.s.). In the study group 59 pts were transplanted (92%) as compared to 54 pts in the control group (84%). Time to engraftment was identical in the 2 groups, with median time to reach 0.5x10 9 /l neutrophils and 20x10 9 /l platelets of 12 and 17 days respectively. The difference in costs was calculated at approximately 80,000€/year. Discussion: Our results indicate that mobilization of PBPC with original and biosimilar fi lgrastim has equivalent results in the key parameters of collection and engraftment. Although we must acknowledge the fact that long-term toxicity data are not yet available, changing to a biosimilar appears safe and effi cient, while allowing signifi cant cost advantages for a transplant unit. Disclosure of Interest: None Declared.

Filgrastim n=124 (biosimilar) p CD34+ cells infused 2.59x10 6 (1-9) 2.93x10 6 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) .04 neutrophils> 500 cells/mm 3 11d (9-23) 12d (6-42) .20 neutrophils>1000 cells/mm 3 12d (9-23) 12d In an open-label, multicenter phase 2 study, 23 patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL) or myelodysplastic syndrome (MDS) will undergo an immunotherapeutic approach consisting of donor lymphocytes selectively depleted of host alloreactive T-cells through the use of photodynamic therapy (ATIR). The myeloablative regimen consists of TBI (1200 cGy), thiotepa (5 mg/kg) and fl udarabine (200 mg/m 2 ). Patients fi rst receive a stem cell graft that undergoes in vitro immunomagnetic T cell depletion using CD34+ positive cell selection. ATIR is infused between 28-32 days after a haploidentical CD34-selected HSCT. No posttransplant GVHD prophylaxis is used. Enrolment is expected to be completed by early 2014. Selectivity of the photodepletion process and reactivity toward recipient and third party cells is assessed using a CFSE-based proliferation assay. Data from this assay is analyzed by Modfi t LT™ software leading to identifi cation of dividing cell populations. From this analysis, a proliferation index, refl ecting cell divisions, is calculated. Results: As of December 12, 2013, 7 patients have been treated with ATIR (mean age 44, range 21-61), AML=4, ALL=3, CR1=2, CR2=5, with a mean follow-up of 4 months post HSCT (range 2-7 months). Two additional grafts have been treated ex vivo and 3 patients are being prepared for HSCT.Stable engraftment was achieved at a median of 11 days 11 (8-18) for neutrophils and 12 days (9-23) for platelets in all patients. No patient experienced graft rejection. None of the patients treated to date developed acute GVHD (any grade). In addition, following ATIR infusion no severe infections (grade 3 or 4) were reported, no patient has relapsed nor died. Based on the 9 batches manufactured to date, ATIR mainly consists of T-cells (>90%). Remaining cells (≤10%) are B lymphocytes and NK cells. Donor cell proliferation toward donor, recipient, and third party cells, as well as CD3/ CD28 is measured and used as release criteria. Results of donor cell proliferation before the photodepletion process and in the ATIR cell product show elimination of anti-host reactivity and preservation of immune reactivity toward third-party cells. In addition, the pan-T-cell activator anti-CD3/CD28 causes robust proliferation in both populations. Moreover, donor cytotoxic T-lymphocyte precursors (CTLp) toward the host are decreased by more than 95%. These results confirm the selectivity of the photodepletion process. Introduction: Allogeneic haematopoietic progenitor cell (HPC) transplant is a widespread treatment in many oncological diseases. The HPC used can be collected directly from the bone marrow or mobilized to the peripheral blood (with the use of aproprieted growth factors) and then collected by apheresis. During the last 2 years, our service updated the aphaeresis equipment from the Cobe Spectra to the Spectra Optia (both from Terumo), which features a new algoritm of collection, in theory obtaining a graft of higher cellular purity. Materials (or patients) and Methods: In order to evaluate actual diff erences in the grafts collected, we review data from 10 4 collections performed in healthy adult donors in our service in 2012 and 2013 in either one of the two equipments. The graft composition was determined by fl ow cytometry in a FACSCanto II (BD). We compared the volume of collection, WBC, CD34 and T lymphocytes collected. Results: For each parameter evaluated, the mean and standard deviation was calculated, and diff erences accessed with the T test.

The results are presented in table 1. Discussion: There are signifi cant diff erences between grafts collected by aphaeresis in the 2 equipments. The Optia Spectra collects cellular products with larger volumes, although much more diluted, resulting in a lower cellular concentration. Regarding the total CD34 cells collected, which are the target cells of the procedure, no signifi cant diff erences were found. We can conclude that although both cellular separators are suitable for HPC collection,

S222 the Optia Spectra produces grafts with less WBC but higher percentage of CD34 cells, confi rming its higher purity. Disclosure of Interest: None Declared. (1.82-6.19 ) and a signifi cant diff erence (P=0.003) with the control group (4.0; 1.52-6.77). There was no signifi cant diff erence in terms of neutrophils and platelets engraftment (>500 and >20000/ μL, respectively). Discussion: In this study we confi rm the effi cacy of AMD to mobilize and collect an adequate dose of CD34+ cells in pts with malignant lymphoma and MM without signifi cant diff erence in term of engraftment in respect of pts mobilized with CHT and G-CSF. Based on our data, we developed and so apply in clinical practice an algorithm for optimal utilization of plerixafor ( Figure 1 ). Disclosure of Interest: None Declared. Introduction: Autologous peripheral blood stem cell (a-HPC-A) collection and transplantation are standard procedures in the treatment of patients with chemosensitive hematological malignancies. Abundant collection of a-HPC-A is assured by the combination of stem cell mobilization by growth factors (GF) and chemotherapy and apheresis collections. In case of major contamination of a-HPC-A collections by mature cells (polymorphonuclear leukocytes co-mobilized by GF) a putative detrimental eff ect on engraftment has been hypothesized. Materials (or patients) and Methods: A patients series including 127 subjects suff ering from lymphomas, myelomas and AML were retrospectively analyzed in platelet (plt) engraftment by an empiric algorithm that normalizes the time to plt > 20,000 after transplantation for the infused CD34 cell dose per kg. The algorithm is a simple mathematical tool which works as follows :pltEngNorm :[(pltEng*pltEng/100)+pltEng]*CD34kg/10, where pltEng is the time (days) to plt > 20,000 and CD34kg is the a-HPC-A dose in millions per kg of body weight. pltEngNorm gives, in this way, a new engraftment time that has been normalized for the CD34 dose and that may identify some other factors infl uencing engraftment. The application of the algorithm to our series gave access to a late engraftment series which had some common characteristics : diagnosis of non-Hodgkin lymphoma and a total nucleated cell (TNC) infused dose higher than 40 x10 9 . Starting from these evidences, Cox analysis and Kaplan Meyer curves have been employed to confi rm factors (CD34, TNC, gender, diagnosis, age) infl uencing engraftment. Results: Our 127 patients received a median of 5.81x10 6 CD34 cells/kg (range 2.01-13.02), experienced an actual engraftment after a median of 12 days (range 7-33) for polymorphonuclear leukocytes (PMN) and after 15 days (range 9-108) for plt. In this series, the median pltEngNorm was 10.62 (range 3.63-69.63) and 5 patients had a pltEngNorm > 30 (this threshold of pltEngNorm has been calculated adding, to the observed median value, a numeric fi gure calculated as the double of the observed standard deviation). As anticipated in the Method section, the patients' series with a delayed pltEngNorm had a very high dose of TNC infused with graft and, in 4 cases out 5, suff ered from non-Hodgkin lymphoma. In the whole series, Cox analysis identifi ed a dose equal to 38.5 x10 9 TNC or higher as a factor that reduces the probability of the engraftment by 50 %, while a dose equal to 4.5 CD34 x 10 6 /kg CD34 or higher has been found able to increase by 50 % the likelihood of engraftment (Prob at chi 2 = 0.0010). No infl uence by diagnosis, age and gender has been observed. Kaplan Meyer curves plotted for these two variables show the eff ect of the combined infl uence of CD34 and TNC on plt engraftment (see fi gure). Discussion: Our data indicate that an empiric algorithm is able to identify cases of autologous transplantation with an anomalous plt engraftment and may represent a useful tool to analyze the quality of graft and transplantation in an individual recipient. Cox analysis confi rmed the reliability of the algorithm and has identifi ed TNC content in the graft as a detrimental factor for plt engraftment, irrespective of CD34 cell dose infused. Future strategies to mobilize and collect a-HPC-A should take into account TNC contamination of the graft as an additional parameter able to infl uence engraftment, besides CD34 cell dose. Disclosure of Interest: None Declared. Introduction: HSCT is a curative option for many patients (pts) with hematological malignancies. Significant advances in supportive care and conditioning regimens over the past decade have allowed the extension of this therapy to older individuals. Information regarding the outcomes of this older subset of pts undergoing HSCT is limited, especially those undergoing haploidentical (HI) HSCT. We report on the outcomes of pts 60 years of age or older undergoing HI and matched related (MR) HSCT. Materials (or patients) and Methods: We did a retrospective review of pts 60 years of age or older enrolled on our 2 step HI or MR HSCT trials. Conditioning consisted of a myeloablative regimen (MA) with 12 Gy of TBI over 4 days, or a reduced intensity regimen (RIC) with Fludarabine/Cytarabine or Thiotepa followed by 2 Gy TBI. Immediately after the TBI in both regimens, a large fi xed dose of allogeneic T cells were infused (2 x 10 8 /kg), resulting in an alloreaction limited to pts undergoing HI HSCT, and characterized by high fevers and in some cases, rash and diarrhea. Two days later, cyclophosphamide (CY) was administered daily x 2 to eradicate alloreactive T cells and induce bidirectional tolerance. Using this method, non-alloreactive T cells remain, forming the basis for early immune reconstitution and avoiding signifi cant GVHD A CD34 selected stem cell product was infused after CY. Results: Of 182 pts undergoing HSCT on one of the 2 step protocols, we identifi ed 62 pts (40 males, 22 females) who were ≥ 60 years of age. The majority of pts received a HI HSCT (90%) with 34 pts ≥ 65 years old (55% (1pt; n=110) . The adaptation of the new risk score concerned HLA and sex matching. HLA matching showed patients with at least 4/6 HLA matching on HLA-A, -B and DRB1 without any mismatch on DRB1 alleles between the recipient and each UCB unit and the 2 units together (0 pt; n=77); patients with at least 4/6 HLA matching on HLA-A and -B between the recipient and each unit but without complete matching on DRB1 or with less than 4/6 HLA matching between the 2 units (1 pt; n=51), and fi nally patients with more than 2 mismatches on the three loci (2 pts; n=8). Donor-recipient sex combination separated all others (0 pt; n=73) from the male recipient with one female and one male cord blood unit (1pt; n=45) and male recipient with two female cord blood units (2 pts; n=18) . Hence, we found 5 patients (4%) in score 1, 12 (9%) in score 2, 23 (17%) in score 3, 26 (19%) in score 4, 30 (22%) in score 5, 26 (19%) in score 6, 11 (8%) in score 7 and 3 (2%) in score 8. We pooled patients with scores 1-2; 3-4 and 5-8 as their outcomes were found to be similar. Results: After a median follow-up of 49.5 months, the 3 years probabilities of overall and progression-free survivals for the whole population were 41% and 35% respectively. with a two-year cumulative relapse incidence of 28% (24.0 -31.8). Interestingly patients with score 1-2 showed better overall survival rates than those with score 3-4 and 5-8 with a 3 years probability of 77%, 42% and 32% respectively (P=0.002); this was associated with a 3 years relapse incidence of 19%, 35% and 55% respectively (P=0.021), while the 3 years transplant related mortality (TRM) was 6%, 39% and 36% respectively (score 1-2 versus all others, P=0.02). Discussion: In conclusion, we found that the EBMT risk score can perfectly be applied to double UBCT with a signifi cant impact on diff erent outcomes; its use enables a better selection of patients who will benefi t the more from this treatment strategy. Disclosure of Interest: None Declared. 3 Hematology, LUMC, Leiden, Netherlands Introduction: In contrast to peripheral blood stem cell grafts, number of total nucleated cells (TNC) per kg is used internationally when a request for a bone marrow graft is made, because TNC has been shown to be an important prognostic factor for bone marrow transplantation. At our stem cell laboratory two techniques are used to concentrate cells or reduce volume from bone marrow harvests: 1) manual concentration (centrifugation of fl asks) or 2) mechanical concentrating by the COBE® Spectra Apheresis System. The fi rst method is used for small volumes (<500 ml) in combination with a small recipient. The second method is used for large volume grafts or major ABO-incompatibility. The manual concentration method hardly infl uences the amount of TNC, whereas using the mechanical concentration method results in enrichment of mononuclear cells (MNC). We questioned whether MNC counts post-processing would be superior to TNC with respect to prediction of engraftment and would therefore be a better parameter for quality control of the processing procedure. Materials (or patients) and Methods: We retrospectively analyzed all allogeneic bone marrow harvests processed in our lab between January 2011 and November 2013. The following graft parameters were collected: volume, TNC, TNC/kg (before and after processing), MNC, MNC/kg CD34, CD34/kg, and engraftment data in the recipients: days of neutrophil recovery > 0.5 x 10 9 /l and platelet recovery > 20 and > 50 x 10 9 /l. Results: Eighty patients (73 children and 7 adults) were transplanted with bone marrow grafts. In 54% bone marrow was manually concentrated. The total TNC count was reduced from a median of 8.4 before to 7.1 x 10 9 after concentration. Using mechanical concentration the TNC decreased on average with 58.9% from a median of 13.1 to 5.1 x 10 9 ; the MNC decreased with 12% and CD34 cells with 9.3%. TNC and CD34 cells in the harvest correlated well (Pearson correlation coeffi cient 0.75; P<0,0001).

In the evaluable patients (n= 73) 4 died before engraftment. Two non-engraftments were observed. All other patients engrafted with a median neutrophil recovery of 22 days. In 8 patients platelet recovery (> 20 x 10 9 /l) was not reached. Of these, 5 received another transplant. The median platelet recovery in the remaining patients was 26 (> 20 x 10 9 /l) and 31 days (>50 10 9 /l). The graft TNC and MNC/kg counts after processing were found not to correlate with engraftment. However, pre-processing TNC/kg showed correlation with platelet recovery: when TNC was >= 5,5 x10 8 /kg (26% of patients), 100% reached platelet recovery > 20 x 10 9 /l compared to 82% in the group with TNC counts < 5,5 x 10 8 /kg. This was also shown with CD34 cells using a cut-off value of 3.5 x 10 6 /kg infused. Discussion: We show that the amount of TNC in the graft is greatly reduced when mechanically concentrated, whereas MNC and CD34 are hardly aff ected. However, post-processing TNC and MNC/kg are not correlated with engraftment parameters. The pre-processing TNC counts and CD34/kg are the only predictors for platelet engraftment. Therefore, we recommend to use preprocessing TNC counts as quality parameter of the bone marrow harvest, as is current practice and to use CD34 pre-and postprocessing as quality control for the concentration procedures. Disclosure of Interest: None Declared. Results: Median age was 63.5yrs (range:45-67). Disease indications-AML-8,FL-1 and CLL-1.Median total cell dose infused (cord 1+2) for TNC was 5.5 x10 7 /kg and for CD34 was 2.3 x 10 5 /kg. The 'winning' unit had a median unit:recipient match of 5/6 (range 4-5/6), with a median TNC of 2.35x10 7 /kg (range 1.42-3.50) and CD34 of 1.3x10 5 /kg (range 0.37-3.20).The 'losing' unit had a median unit:recipient match of 4/6 (range 4-5/6), with a median TNC of 2.25x10 7 /kg (range 1.67-3.70) and CD34 of 0.97x10 5 /kg (range 0.28-2.00). The degree of HLA matching was statistically significant for the winning unit (P= 0.033, paired t test).Median duration of follow up is 11 mnths (range 4-11). All patients engrafted. ANC recovery to >0.5 x 10 9 /L occurred at a median of 21.5 (range 10-28) days and platelets to >20 x10 9 /L at a median of 39 (range 25-49) days. 7/10 patients developed mild grade1-2 acute GvHD. Of these 2/7 developed grade 1-2 acute GvHD of the gut controlled with topical (1/2) and oral (1/2) steroids. All 7/7 patients received topical steroid to control skin GvHD. Only 3/10 patients had CMV reactivation requiring treatment.Chimerism analysis revealed that the winning unit could be predicted as early as D14. By D21 all 10/10 patients were 100% donor chimeric. However Introduction: During the last decade, peripheral blood stem cells (PBSC) has become the most common graft source in unrelated donor stem cell transplantation (URD-SCT). However, few studies are available that compares the outcomes of PBSC vs. bone marrow (BM) transplants in the setting of URD-SCT. Furthermore, an analysis of the long-term outcomes of URD-SCT according to the graft source has not yet to be reported in adults with acute lymphoblastic leukemia (ALL). In order to determine which graft source is superior in adults with high-risk ALL, we compared the long-term outcomes of URD-SCT between PBSC and BM. The strengths of this study include suffi cient follow-up duration and restriction to a single disease with a uniform pre-and post-transplantation treatment strategy. Materials (or patients) and Methods: To determine which graft source is superior in adult high-risk ALL, we analyzed the longterm outcomes of 106 consecutive transplants from 8/8-matched or 7/8-matched URD (38 PBSC vs. 68 BM). All patients received a uniform strategy of pre-transplant chemotherapy (modifi ed hyper-CVAD regimen), myeloablative conditioning (total body irradiation plus cyclophosphamde), and an identical graft-versushost disease (GVHD) prophylaxis (tacrolimus plus methotrexate). Results: The median patient age was 23 years (range, 15-48 years).

All patients had at least one high-risk factor. Eighty-three patients Discussion: Our data suggest that "total body irradiation-fl udarabine-cytarabine conditioning double CBT" is a feasible approach for adults with high-risk ALL who lack a suitable related or unrelated voluntary donor. Disclosure of Interest: None Declared. shown that a higher number of CD34 cells infused leads to a more rapid engraftment. However, as long as CD34 cell doses infused were above the cut-point associated with more rapid engraftment, we were not able to discern specifi c advantages of doses higher than that threshold. Relapse was not aff ected by CD34 cell dose in any group. This is in contrast to a previous CIBMTR study of HLA-identical sibling MAC HCT, where CD34 cell doses >6x10 6 / kg correlated to decreased relapse risk. We have no explanation for this discrepant fi nding, since a high CD34 cell dose is expected to be more important in RIC than in MAC HCT. Despite this, CD34 cell dose was of importance of OM. Earlier studies of RIC HCT have shown a correlation between high cell doses and better survival, but at cost of more chronic GVHD. In this study we did not fi nd a CD34 cell dose above which acute or chronic GVHD risks were higher. However, threshold levels of CD34 cells associated with increased GVHD incidence in previous studies have been high, >107 cells/kg. Hence, it is possible that such high cell doses have been avoided in many centers in recent years.

To conclude, in RIC HCT for AML or MDS, the PBPC CD34 cell dose should be delivered in excess of 4x10 6 /kg for HLA-matched siblings, and in excess of 6x10 6 /kg using MUD. However, no data on accessory cells infused were analyzed in this study, which may hamper the interpretation of the results. Disclosure of Interest: None Declared. Introduction: The identifi cation of a donor has always limited the extent of allogeneic HSCT. Recently it has been shown by diff erent teams that a haploidentical donor could be a valid option to perform allo HSCT given adapted immunosuppression. Notably the use of PT-HDCy), after T-replete HSCT following reduced intensity (RIC) or non-myeloablative (NMAC) conditioning, has been associated with promising results. However little data exist concerning elderly population when this population is characterized by a lack of HLA matched sibling and a higher incidence of severe GVHD and non-relapse mortality. Materials (or patients) and Methods: Using this strategy in a merged program between our two institutions we recently transplanted 30 patients over the age of 60 years (70% received the strict Baltimire conditioning while 30% were prepared with more aintense treatment. all received the same PTHDcy schedule. they were compared with patients of the same age transplanted from a MRD or MUD prepared with our sandard Fluda (5 days) -Iv Bu(2 days) and ATG (2 days Patients with haplo had a trend for higher HCT-CI and DRI. They received more often a NMAC (P<0.05). and ext cGVHD were less frequent after haplo (P<0.05). Early NRM and Relapse occurrences were similar after haplo and MRD transplants. Early relapse was lower after MRD but population had better DRI. Discussion: In conclusion haplo HSCT based on HD-PTCy is associated with lower severe GVHD incidences and no superior NRM and relapse occurrences. Longer follow-up is needed to fully evaluate this strategy and will be presented. In the same time period the following procedures were carried out within trials of regenerative medicine: 3. volume reduction by a simple buff y-coat (BC) separation for a protocol of bone regeneration in the orthopaedic setting (n=54) and 4. density-gradient mononuclear cells separation in a trial for the treatment of critical limb ischemia (CLI) (n=13). For each protocol specifi c software (table) have been used and/or settled to diff erent clinical targets. Results: Initial and fi nal volume, volume reduction, recovery of both Total Nucleated Cells (TNC) and CD34+ cells are reported below (median and range). Data are divided for Allo-HSCT, Auto-HSCT, BC for regenerative medicine and density-gradient separation (DGS). No side eff ects were reported in the clinic; in particular, no transfusional reactions were reported in the fi rst group except for one patient who developed a reaction after the infusion of 20 mL of the product, possibly due to the presence of irregular Abs.

In the latter case, the graft was rescued by MNC separation with DGS protocol and the infusion was carried out without any side eff ect. In case of very large volumes, the product was split in two or three aliquots and processed in parallel in diff erent devices, in order to spare time. Time to engraftment of transplanted patients was in the normal range. No microbiological contamination due to the manipulation process was reported. Discussion: The relatively low volume of the separation chamber required multiple processing of large volume products. However all the procedures were run in a fully automated setting through specifi c software aimed to diff erent clinical targets, resulting in a modest time involvement of the operators. Clinical effi cacy of the graft was in line with historical controls. BM manipulation with Sepax was shown eff ective, operator independent and could be applied for a broad range of clinical applications. Introduction: Donor lymphocyte infusion (DLI) is frequently used to harness the graft-versus-leukemia (GVL) activity in patients who relapse after allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, whether the cell composition of this G-CSF mobilized DLI would play an essential role in the outcomes of patients after mDLI still remained unclear. In this study, in this study, our purpose was to elucidate the clinical signifi cance of cell composition of mDLI in diff erent type of allo-HSCT including HLAidentical and haploidentical transplant. Materials (or patients) and Methods: 194 patients with hematological malignancies undergoing allo-HSCT from January, 2007 to March, 2011 were enrolled in this study and received mDLI for various clinical reasons. The infused cellular components of mDLI were examined by fl ow cytometry. We analyzed the possible risk factors of patient clinical outcomes and focused on the diverse infused cells in mDLI. Results: The objectives in this study was composed of patients undergoing HLA-identical (n=60) and haploidentical allo-HSCT (n=134). The median of MNCs infused in mDLI was 1.0×108 (range: 0.75-2.18×108) cells/kg. The results showed that infusion of a higher number of CD34+ cells was associated with a lower DLI-related mortality (DRM, P=0.038) and a better overall survival (OS, P=0.002) in all patients. In HLA-identical transplant patients infused a lower number of CD14+ cells (<0.33×10 8 /kg) was the independent risk factor of the occurrence of II-IV°aGVHD (HR=0.104, P=0.032). In addition, a dose of CD14+ cells less the 50 th percentile was associated with a lower incidence of hematological relapse after mDLI (HR=0.193, P=0.007). However, in contrast to the results of HLA-identical transplant. It showed that a higher number of CD14+ cells was the independent risk factor of II-IV°aGVHD (HR=1.758, P=0.034) in haploidentical allo-HSCT.

which had the capacity to regulate alloreactive T-cell responses. Due to the immunosuppressive eff ect of MDSCs, several recent studies proved the correlation between the occurrence of aGVHD and MDSCs. Another surprising fi nding of our study was that a higher number of CD14+ cells in mDLI were associated with the lower incidence of relapse. Nausch et al. has demonstrated that CD14+HLA-DRlow/neg cell levels correlated with activated CD25+ NK-cell proportions and a ligand for activating NKG2D receptor on NK-cells termed as RAE-1 was expressed on murine MDSCs. Based on these data, this monocytic CD14+ cells would also help to enhance the graft-versus-leukemia (GVL) eff ect through modulating the activity of NK cells. While, in haploidentical transplant the infused CD14+ cells in mDLI displayed the opposite role and they were associated with a higher incidence of aGVHD. We postulated that maybe in haploidentical transplant the balance between these two subsets CD14+ cells was diff erent from that of HLA-identical transplant. CD14+HLA-DR+ monocytes might take the lead role in this mismatched allo-HSCT. In conclusion, the fact that CD14+ cells would act as distinct regulators in diff erent types of transplants might provide a new sight for the cellular therapy through manipulating the component of infused cells. Disclosure of Interest: None Declared. Introduction: Graft sources such as an HLA-mismatched unrelated donor (MMUD), unrelated umbilical cord blood (UBC), or an HLA-haploidentical related donor (haplo-RD) are viable alternative treatment options for patients who lack an HLA-matched donor. Here, we evaluated the outcome of these diff erent graft sources, while making a distinction between the use of a T-cell replete graft (TCR) and a combination of a T-cell replete and a T-cell deplete graft (cTCR/TCD) in haplo-RD transplantation.

March 2011 alternative donor graft sources were used in 183 patients, while 196 transplantations were conducted at our institution. Graft source was a MMUD in 75, UCB in 23 and a haplo-RD in 98 recipients, respectively. In the HLA-haploidentical setting a TCR graft followed by high dose cyclophosphamide (Cy) was used for 24, while a combined T-cell replete and deplete graft (bone marrow on day 0 plus CD6+ deplete PBSCs on day +6) was used for 74 transplantation procedures. Median age was 43 years. Acute leukemia was the most frequent underlying disease (MMUD: 73%; UCB: 87%; haplo-RD: cTCR/TCD: 78%; haplo-RD TCR: 64%), while B-cell lymphoma was diagnosed in 5% (MMUD), 9% (UCB), 16% (cTCR/TCD) and 25% (TCR). In haplo-RD transplantation disease stage was advanced in 100% of the TCR group including 42% S233 second allo-transplantations, whereas in only 16%, 9% and 8% of the cTCR/TCD, UCB and MMUD group a second allo-transplantation was performed. In the TCR group 75% had adverse cytogenetics in AML (46% cTCR/TCD; 40% UCB and 44% MMUD). Results: With a median follow up of 33 (MMUD), 17 (UCB), 48 (cTCR/TCD), and 6 (TCR) months estimated one-year OS and DFS were 55% and 48% for the MMUD, 54% and 49% for the UCB, 44% and 43% for the cTCR/TCD, and 51% and 35% for the TCR recipients. One-year OS was 79% for patients receiving a fi rst allo-transplantation in the TCR group. CI of acute GvHD was 76% for the MMUD, 47% for the UCB, 46% for the cTCR/TCD and 46% for the TCR group, while CI of chronic GvHD was 52% after MMUD, 22% after UCB, 34% after cTCR/TCD, and also 34% after TCR haplo-RD transplantation, respectively. One-year CI of NRM was 36% (MMUD), 30% (UCB), 30% (cTCR/TCD) and 18% (TCR) with no signifi cant diff erence. Despite remarkable diff erences in risk contributions within the four groups, in particular regarding the TCR group including more high risk features, OS and DFS were not signifi cant diff erent between the groups. CI of acute GvHD was signifi cant higher in the MMUD group compared to the other groups (P=0.00001), and cGvHD was signifi cant lower for the UCB compared to MMUD recipients (P=0.04). Patients with diagnosis of lymphoma, AML with adverse cytogenetics, time from initial diagnosis to transplantation >1 year and second allo-transplantation had a signifi cant lower OS (P=0.01; P=0.02; P=0.03 and P=0.01). Using a cTCR/TCD graft was associated with a high incidence of PTLD (25%). Discussion: Our results confi rm the utility of alternative donor graft sources, suggesting that UCB and haplo-RD transplantation, in particular using T-cell replete grafts, is equivalent to the use of MMUD grafts, while resulting in less GvHD. However, in our cohort a longer follow up for the TCR group is needed. Disclosure of Interest: None Declared. In the setting of cord blood transplantation (CBT), the problem of low cellularity was brilliantly solved by the intrabone injection of CB stem cells, with good engraftment even in adult patients.

In cases of poor haplo-donor stem cell mobilization and insufficient number of CD34 + available, we decided to adopt the same strategy of intrabone injection of part of the positively selected CD34 + cells. Materials (or patients) and Methods: From September 2009 to April 2013 14 children aff ected by malignant or non-malignant hematological disorders and given aploHSCT received part of the CD34 + stem cells as intrabone injection because of a low stem cell dose collected from the donor (11 cases) or in case of a second haploHSCT after a previous graft rejection (3 cases).

The intrabone infusion was carried out at the patient bedside under general anesthesia as previously described (Frassoni F. et al. Lancet Oncol 2008; 9:831-39) . For patients receiving the intrabone injection as part of a fi rst aploHSCT the median number of CD34 + cells infused was 9.7 x 10 6 /kg (range, 5-12) and the median number of CD3 + lymphocytes was 0.7 x 10 4 /kg (0. [3] [4] [5] [6] [7] [8] [9] [10] [11] . For children receiving the intrabone injection as part of a second aploHSCT after a fi rst rejection the median number of CD34 + cells infused was 20 x 10 6 /kg (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) and the median number of CD3 + lymphocytes was 1.8 x 10 4 /kg (0.8-2.8). About 1/3 of the stem cell inoculum, corresponding to a total volume of 20 ml, was given intrabone, while the remaining stem cell portion was infused intravenously. Results: No complication occurred during or after the intrabone injection.

Eleven of the 14 patients achieved complete donor engraftment, while 3 patients rejected the transplant. The median time to neutrophil recovery was 13 days (range, [12] [13] [14] [15] [16] [17] [18] [19] [20] , while the median time to achieve platelet engraftment was 14 days (range, 13-40). Only 1 patient developed grade II acute GVHD and only 1 limited chronic GVHD. Three patients died, 2 for transplant related causes and 1 for disease progression. The overall survival probability was 76% (52-100), and the cumulative incidence of transplant-related mortality was 14% (4-51). Discussion: Our data suggest that the intrabone infusion of positively selected CD34 + cells, in the context of an haploHSCT, can be safely used in case of poor donor mobilization and low number of available stem cells, with the aim of minimizing the risk of graft rejection or poor engraftment. Our data need to be confi rmed in a larger number of patients and compared with those obtained with conventional intravenous administration of comparable numbers of cells. Disclosure of Interest: None Declared.

Introduction: Unrelated cord blood transplantation (UCBT) has been used increasingly in adults in the UK for over a decade. Two national prospective clinical trials have been recruiting (RIC UCBT and MAC UCBT) and national consensus guidelines have been produced (Shaw et al, Bone Marrow Transplantation 2009; 44:7-12) . As UK experience has not been appraised outside of these trials, we conducted a retrospective analysis to summarise demographic data and outcomes in pts >18 years treated with UCBT using Eurocord and BSBMT databases. Materials (or patients) and Methods: From 2000 to 2012, there were 176 registrations with corresponding cord blood bank data S234 from 23 centres. 28 pts in national prospective clinical trials were excluded from further analysis. Outcomes were analysed for 148 pts with acute leukaemia (n=81), myeloproliferative disorders/ myelodysplastic syndrome (MDS) (n=42), lymphoproliferative diseases/myeloma (n=20) or aplastic anaemia (n=5). Results: Pts had a median age of 40.8 years (18-72) and weight 69.5 kg (38.5-130). Various conditioning regimens were used, with 52% receiving a reduced intensity conditioning regimen. Most (74%) received double cord blood units (dCBU) and the remainder a single CBU. Recorded median total cell dose infused was 3.58 x10 7 /kg (0.41-34.35) for total nucleated cell count (TNC) and 1.55 x10 5 /kg (0.13-14.97) for CD34+ cells. Engraftment of neutrophils to >0.5 x10 9 /L occurred at a median of 22 days and platelets to >20 x10 9 /L at a median of 38 days Introduction: Graft failure (GF) is a major concern after cord blood transplantation (CBT). Primary graft failure (pGF) in CBT is associated with considerable morbidity and mortality. Salvage hematopoietic stem cell transplantation(HSCT) can rescue pGF patients. However, the stem cell source and the preconditioning regimen in salvage transplantation are yet to be detemined. Because the patients who received CBT almost always lack both matched-related and -unrelated donors during the clinically relevant period, haploid donor is considered to be suitable as stem cell source for salvage transplant.

Materials (or patients) and Methods: In this study,we retrospectively analyzed 17 patients with hematologic malignancies who receiced haploidentical transplant with non myeloablative conditioning as salvage therapy for pGF in CBT.At the first transplant, 12 patients received single unit CB, while 5 patients received double units CB. The first transplant was administered with myeloablative conditioning. The median interval between the two transplants were 36 days. The non-myelo ablative conditioning regimen for salvage transplant consisted of Flu(total 120mg/m 2 ), antithymocyte globulin (ATG, total 10mg/kg), cyclophosphamide (50mg/kg.d for one day) and low dose TBI (3Gy). G-CSF mobilized PBSCs and BM were transfused intravenously to the recipients just after the completion of the collection on days 0 and 2 of transplantation, respectively.The median number of infused total nucleated cells dose was 8.02 ×10 (1.2-5) and of 0.7x10 6 CD19+/kg (0.1-1.5). Twelve/14 pts achieved a complete donor engraftment from d15 to d42 (med d15). The med time to 500 PMN's and to 20 000 plat was 9 days (2-30) and 23 days (15-45). Two pts with CMML failed to engraft the exp product as well as further unmanipulated CBU. AGVHD grade II, III, IV was observed in 3, 1, 1 pts. Two pts died of TRM (1 with RIC neuro tox, 1 grade IV AGVHD). Five pts had relapsed (2/2 who did not engraft and 3/12 who engrafted exp CBU). One pt rejected the exp CBU at 3 m and is AW after rescue.Overall, after a med FU of 20 m (3-47 m) the 2y OS and DFS is 50%. Full donor chimerism maintained until relapse, death of TRM or last FU in 11/14 pts who received the exp graft. Discussion: Ex-vivo expansion of a single CBU is feasible and reproducible. Transplantation of the exp CBU with the CD34-counterpart of the same CBU produces rapid, complete and sustained donor engraftment after RIC in adults. Disclosure of Interest: None Declared. The parents were counselled on multiple occasions before proceeding with transplantation. Issues discussed were related to the risk due to general anaesthesia in an infant, the need for transfusion for the donor as the amount harvested for an older sibling would necessitate a top up transfusion and diffi cult iliac crest punctures in an infant as the posterior crest is mostly cartilaginous at that age. The children were anaesthetised by an experienced paediatric anaesthetist at our centre. All transfusions for the donor came from directed donations from the families and the samples were screened using nucleic acid testing in addition to antibody screening. The harvests in these infants were all performed by paediatric transplant physicians directly to ensure safety of the procedure. The posterior crest punctures were done using marrow harvest needles that were changed after 15 punctures. Results: All 15 children tolerated the general anaesthesia without any complications. They received top up transfusions during their donation. An average of 5 to 7 ml per kilogram of the recipient body weight of bone marrow was harvested from each of these donors. The volumes were low compared to the usual harvests done in older children. The total nucleated cell count and CD34 counts per ml of bone marrow were far higher in infant donors. The average yield was 7.5 x 10 6 /kg with a range from 4 to 23 x 10 6 /kg of the recipient body weight. There was no haematoma at puncture sites and the babies regained their mobility soon after recovery from anaesthesia. Discussion: Our series highlights that the process of haematopoietic stem cell donation is safe for an infant in a paediatric unit. The stem cell yield is extremely high resulting in rapid and engraftment in all recipients. Disclosure of Interest: None Declared.

Introduction: Cord blood (CB) is increasingly used as an alternative hematopoietic stem cell source and the results of unrelated CB transplantation (CBT) have improved in the last decade. However, graft rejection has been a major concern in unrelated CBT.

Although it is common practice to select the cord unit with the greatest total nucleated cell (TNC) counts within a given HLAmatch grade, the essential cell dose ensuring engraftment with acceptable probability was not established. We analyzed the impact of prefreeze TNC dose, CD34-positive cell (CD34C) dose based on per body, per actual (ABW), or per standard body weight (SBW) upon engraftment after unrelated CBT. Materials (or patients) and Methods: A total of 5668 patients who received single-unit CBT between 2000 and 2011 were analyzed based on the data reported to the Japan Society for Stem Cell Transplantation registry. A crude analysis was performed in whole cohort and then 787 patients were selected according to the following criteria: (1) AML in 1CR or ALL in 1CR; (2) . According to BST protocol, Cord was reconstituted with 1:1 dilution with a hyperosmolar solution of 7.5% Dextran/albumin using the Sepax device. Immediate infusional-related events were evaluated after infusion. In addition, we analyzed engraftment and main clinical outcomes. Results: A total of 145 patients underwent a single UCBT during the study period (median age 18 years, range 1-56) Most of patients received myeloablative conditioning regimen based on busulfan (6.4 mg/kg iv), thiotepa (10 mg/kg), fl udarabine (150 mg/m 2 ) and anti-thymocyte globulin (6-10 mg/kg). The diagnosis were: AML (n=47, 32%), ALL (n=47, 32%), NHL (n=14, 10%), primary immunodefi ciency diseases (n=12, 8%) and other (n=25, 18% Introduction: High dose chemotherapy followed by autologous stem cell transplant (ASCT) represents the standard of care for several hematologic malignancies. The failure in peripheral blood stem cell (HSC) mobilization is a crucial issue, because it results not only in the impossibility of carrying out the transplant, but also in an increased toxicity and risk of infectious complications following the mobilization attempts. Purpose of our study was to retrospectively analyze the results of HSC mobilization in patients with hematological malignancies, at our Center since 2011, focusing in particular on factors that may have infl uenced the mobilization. Materials (or patients) and Methods: We analyzed 272 patients who underwent a HSC mobilization process. Two were aff ected by amyloidosis, 30 by Hodgkin's lymphoma, 94 by non-Hodgkin's lymphoma, 50 by acute myeloid leukemia, 15 by acute lymphoid leukemia and 81 by multiple myeloma. The median age was 51 years and the M/F ratio 137/135. The primary outcome was CD34+ cell yield; secondary outcomes included number of aphereses, proportion of failures, possible factors negatively aff ecting the mobilization. Among the 272 patients, 63 met the criteria for the defi nition of predicted poor mobilizer according to the GITMO proposal; in particular, 56 satisfi ed major criteria and 7 met 2 minor criteria, while 209 had no adverse prognostic factors for mobilization. The mobilization strategies were: chemotherapy combined with G-CSF in 90% of cases, G-CSF alone in 3% and Plerixafor (PLX) with G-CSF in 7%. Results: In the entire patients' population, the median CD34+ cell yield was 7.1 x 10 6 /kg (range 2-21) and the median number of aphereses required to reach the target was 1 (range 1-3). According to the algorithm employed at our Institution, the median White Blood Cell (WBC) count to start the apheretic procedure was 13. x 10 9 /L (range 2-81) and the median number of CD34+ cells/μL in the peripheral blood was 43 (range 12 -232). The failure rate was 14% and the main negative prognostic factors for HSC mobilization in univariate analysis were: AML diagnosis (P=0.001), use of cytotoxic compounds such as melphalan and fl udarabine or clofarabine at least in one of the previous therapeutic regimens (P<0.001), more than 2 previous therapeutic lines (P<0.001). Among the 63/272 patients defi ned as predicted poor mobilizers, 28 were actually proven poor mobilizers, because they reached a median peak of 1.45 CD34/μL (0-13) and 12x10 9 /L WBC (0.2-45) at a median of 10 days from the start of a mobilizing procedure based on chemotherapy and G-CSF in 64% of cases, G-CSF alone in 7% and PLX with G-CSF in 29%; in 35 patients, a median of 4.5 x 10 6 /kg CD34+ cells (range 2-11) were collected using the following methods: chemotherapy and G-CSF in 77% of cases, G-CSF alone in 3% and PLX with G-CSF in 20%. Discussion: HSC collection is a challenge in patients eligible for an ASCT. In our experience, HSC mobilization was successful in the majority of patients (86%) and a considerable percentage of those failing with standard procedures could be successfully rescued with PLX. For these patients, the identifi cation of reliable predictors of mobilization failure is mandatory, with the aim of determining which patients might benefi t from the pre-emptive addition of PLX and in order to optimize the cost-eff ectiveness of the procedure. Disclosure of Interest: None Declared. The median percentage of pDC on the graft was 0.20% of non erythroid nucleated cells (range: 0.02-0.67) and no differences were noticed among sources. Cumulative incidence (CI) of grade II-IV acute GVHD at 100 days was higher on patients receiving grafts with higher percentages of pDC (22% for patients with less than 0.15%pDC graft content vs. 52% for patients receiving grafts with higher counts, P=0.026). There was no impact of graft pDC content on mortality, relapse, chronic GVHD or overall survival. In a multivariate analysis, graft pDC content remains an independent risk factor for acute GVHD [HR: 3.0, CI (95%): 1.15-7.8].

Discussion: Higher pDC graft percentages were associated with increased risk of acute GVHD, but had no impact on non-relapse mortality or survival. The precise role of pDC on immunity after HSCT deserves further investigation and might be related not only to pDC biology itself but to the chronology of the pDC interaction with host antigens. Disclosure of Interest: None Declared. The median number of leukapheresis performed was 1.5 (1-3); the median number of blood volemias processed was 3 (2-5); and the median number of CD34+ cells collected was 3.7 (0.44-22.6) x 10 6 /kg of patient body weight. After the fi rst mobilization, 11 children done 25 leukapheresis and 4 weeks later 5 bad mobilizers performed more 11 procedures. The most frequent adverse reaction described with this type of PBSC collection is oral paresthesias; so, in order to prevent symptoms associated with hypocalcemia in the most small child (N=8), we administered Calcium Gluconate in continuous perfusion during 1 hour. Five patients needed platelets transfusion at the end of the apheresis. Discussion: We can conclude that harvesting PBSC with Spectra Optia proved to be a safe and eff ective technique, even in patients with a body weight less than 10 kg. Most of them did the collection with just one apheresis in a out-patient regimen and without needing the support of pediatrician or specialist in intensive care. Disclosure of Interest: None Declared.

Introduction: Mobilized peripheral blood stem cells (PBPC) are the preferred stem cell source for allogeneic transplantation. Beside PBPC, transplants also contain further therapeutic active cells, e.g. NK cells. For collection mostly the Cobe spectra apheresis device (SPECTRA) was used. Recently its successor (Spectra Optia MNC apheresis system (OPTIA)) was introduced. Instead of continuous collection of the buff y coat during low spin centrifugation, the OPTIA collects target cells into an intermediate elutriation chamber during hard spin centrifugation. After processing a certain volume of peripheral blood, the chamber is fl ushed into the fi nal product. During fl ushing no substantial further blood is processed. The peripheral blood volume processed between two fl ushes (BVpbF) can vary, fl ushing itself is either triggered automatically or manually by the operator. We questioned whether the performance of the devices and the composition of the allogeneic grafts diff er between SPECTRA and OPTIA. Materials (or patients) and Methods: At the 5 th day of 10μg Lenograstim/kg BW PBPC were collected, starting apheresis 2h after the last G-CSF injection. 245 PBPC SPECTRA and 64 OPTIA runs (2010) (2011) (2012) (2013) 

Full diff erential blood as well as the PBPC count was obtained from PB before apheresis (as well as from the product itself. For platelets (plt) and PBPC collection effi ciency (CE2) was calculated by yield/(amount pre apheresis x processed blood volume). Further performance measurements were target cell collection rate (CR, 10 6 ) per fold processed total blood volume and kg BW donor and throughput (TP2) Major sites of involvement were skin (71%), and gastrointestinal (GI) track (upper: 52.6%, lower: 44.7%), whereas liver was involved in 29% of cases. Acute GvHD responded to steroids in the majority of patients (92%), but recurrence was observed in 59%. Of note, 9/11 patients with grade III-IV acute GvHD died of NRM. The development of aGvHD did not correlate with age, intensity of the conditioning regimen, TNC or CD34+ cell dose, and the degree of HLA match of the highest mismatched unit or the engrafting unit (by either low-or high-resolution typing at 6 or 10 loci). The CI of chronic GvHD was 39% (extensive: 24%,), with median time of onset of 6 months (range: 3.5-9). Affected organs included skin (73%), GI (40%), liver (20%), mouth (20%), serosae (13%), and eyes (6.7%). The occurrence of extensive chronic GvHD was associated with less than 6/10 HLA match of the engrafting unit (HR: 4.0, P=0.05). Among the 15 patients with chronic GvHD, 8 have discontinued immunosuppression, 2 are still under treatment, and 5 died of complications related to GvHD (n=4) or relapse (n=1). Discussion: A considerable incidence of grade III-IV acute GvHD was observed in DUCBT with an associated adverse eff ect on NRM.

Despite multiple HLA disparities, development of chronic GvHD after DUCBT was less frequent compared to the reported incidence in alloSCT from adult matched unrelated donors. Disclosure of Interest: None Declared. Introduction: Immune-and especially T-cell-recovery after profound lymphopenia is a major challenge in many clinical situations, such as allogeneic hematopoietic stem cell transplantation (allo-HSCT). Recovery depends on hematopoietic lymphoid progenitors emerging from the bone marrow (BM). Materials (or patients) and Methods: In this study, we characterized CD34 + Lin -CD10 + lymphoid progenitors in the peripheral blood of 39 patients by fl ow cytometry. T-cell potential of FACS sorted cells was assessed in 9 patient using OP9-DL1 diff erentiation assay in limiting dilution and their gene expression in 21 patients using RT-qPCR. Finally chemokine and cytokine concentration was assessed using Luminex or ELISA assay on plasma from 40 patients. Results: Our data demonstrate a strong recovery of this population 3 months after transplantation. This rebound was abolished in patients who developed acute graft-versus-host disease (aGVHD). A similar recovery profi le was found for both CD24 + B-cell committed progenitors and CD24circulating Thymus-Seeding Progenitors (TSP). We found a similar T cell potential for TSP isolated either from patients or from normal Healthy Donors (HD). Quantitative expression profi le of 22 genes involved in T-cell diff erentiation and homing, in TSP from patients with aGVHD did not diff er from those from patients without GVHD or HD. However, TSP from patients not presenting aGVHD had reduced CXCR4 gene expression, consistent with enhanced egress from the BM. CCR7 gene expression was reduced in all patients after allo-HSCT, as were its ligands CCL21 and CCL19. This reduction was particularly marked in patients with aGVHD, suggesting an impaired T-cell homing potential. Discussion: Thus, the data presented here identify the TSP population as an important target when seeking to enhance immune reconstitution in humans. Disclosure of Interest: None Declared. Introduction: NK cells are the fi rst lymphocytes to recover after allo-HSCT and play a major role in innate immunity. This recovery follows the NK diff erentiation pattern with an early expansion of the NK CD56 bright subset followed by NK CD56 dim CD16 + . In addition, relatively small cohort studies suggest that the kinetics of NK-cell reconstitution and cytomegalovirus (CMV) reactivations are closely intertwined. NK-cells that express the activating receptor NKG2C seem crucial in the resolution of CMV episodes. Materials (or patients) and Methods: We investigated prospectively in a single center the kinetics and profi les of NK cell reconstitution in a large cohort of allo-HSCT adult patients (n=439) treated (2005) (2006) (2007) (2008) (2009) (2010) (2011) with non T-cell depleted stem cells from bone marrow (76%) or peripheral blood (24%) and reduced intensity (61%) or myeloablative conditioning (39%). NK-cell subsets were analyzed at months M3, M6, M12 and M24 post-transplantation on freshly collected blood samples. Results: Data were analyzed with respect to conditioning regimen, source of stem cells, underlying disease, occurrence of Graftversus-Host Disease (GvHD), and profi les of CMV reactivation. We showed, from multivariate analysis, that acute GvHD impaired reconstitution of total and CD56 dim NK cells at M3 (P=0.006 and 0.002 respectively). An effi cient NK cell reconstitution at M3 was associated with a lower incidence of chronic GvHD, independently of a previous episode of acute GvHD and stem cell source. With regard to CMV, CD56 dim and total NK counts were lower at M3 in the group of patients reactivating CMV between M0 and M3 (P=0.03 and 0.01). Whereas there was no statistical diff erence between counts of total NK, CD56 bright and CD56 dim at M3 and a CMV reactivation detected during the M3-M6 period, NKG2C expressing total NK cells and the CD56 dim NK subset were significantly increased at M3 in patients who did not experience further reactivation (P=0.011 and 0.007). This favors a direct role of NKG2C expressing NK cells in the early control of CMV reactivation. Discussion: These data highlight the interest to monitor NK cells at the subset rather than at the level of the whole population and highlight in a large cohort of adult patients the benefi cial impact of a potent early NK reconstitution, especially of the CD56 dim NK subset. Disclosure of Interest: None Declared.

Introduction: To evaluate the prognostic signifi cance of quantitative chimerism for monitoring minimal residual disease and predicting relapse in patients with acute leukemia (AL) following allogeneic hematopoietic stem cell transplantation (allo-HSCT). Materials (or patients) and Methods: the quantitative chimerism levels of 129 AL patients were measured using real-time quantitative PCR based on 29 sequence polymorphism(SP) markers at designed time points after allo-HSCT. All patients were divided into diff erent groups based on the quantitative chimerism levels and intervention application.

Results: The ROC curve results indicate that the optimal cutoff point to predict an inevitable relapse is 1.0%, When this tentative cutoff value was used, chimerism levels had a sensitivity of 96.2% and specifi city of 79.6%, the relapse rate of patients with chimerism >1.0% at two years was 55.0%, while the relapse rate of the patients with chimerism <1.0% was 0%, P=0.000,quantitative chimerism>1.00% was indicative of a higher probability of relapse.The Cox multivariate analysis indicated only quantitative chimerism >1.00% were associated with lower disease-free survival(hazard ratio(HR)=10.825; 95% confi dence interval (CI) =4.704-24.912, P=0.000) and lower overall survival (HR=8.681; 95% CI=3.728-20.212, P=0.000). Of the 47 patients with quantitative chimerism >1.00%, 24 patients received modifi ed donor lymphocyte infusion (mDLI) immunotherapy, This group had a signifi cantly lower relapse rate (37.5%) than the 9 patients who did not receive mDLI immunotherapy (100%; P =0.001). Discussion: the quantitative chimerism is an independent prognostic factor that can predict clinical outcomes for AL patients after HSCT and provide a guide for suitable interventions. Dynamically monitoring the level of quantitative chimerism after HSCT, however, is more important for determining whether intensive intervention should be performed. Disclosure of Interest: None Declared.

Introduction: Immune status assessment after allogeneic hematopoietic stem cell transplantation (HSCT) relies primarily on complete blood counts and immunophenotypic analyses of T and B lymphocytes. However, this kind of approach does not analyze their functional capacity. Characterization of potential functional immune defects following allogeneic transplantation may contribute to predict clinical complications after transplant. Materials (or patients) and Methods: . Blood samples from a cohort of 77 transplanted patients were collected at fi xed time points (30, 100 and 365 days) after HSCT. Patient's samples were analyzed for lymphocytes subsets (CD3, CD4, CD8, CD56) T cell proliferation, CD4 intracellular ATP content, T cell cytotoxicity, neutrophil phagocytic capacity and granulocyte oxidative burst. High risk acute leukemia was the main indication for HSCT (61% of the patients). Eighty percent of the patients received a nonmanipulated matched unrelated graft. As preparative regimen, 66 patients received a fl udarabine-based toxicity-reduced regimen; while the rest of the patients, a conventional high-dose conditioning protocol was used. For graft versus host (GvHD) prophylaxis, in addition to cyclosporine, all the patients with unrelated donor received either alemtuzumab or rabbit anti-T lymphocyte globulin. Results: The GvHD (grade III and IV) and relapse incidence was 5 and 31 percent respectively. Complete blood count was reduced below normal limits until day 100 and normalized one year after transplantation. All lymphocytes subsets were below normal levels one year after HSCT. CD3 proliferation and CD4 intracellular ATP content was significantly reduced at day 30 and 100 following HSCT. CD4 intracellular ATP content remained reduced for one year after transplant. T cell degranulation, assessed as CD107 expression, did not show significant changes at all time points tested. Neutrophil phagocytic capacity remained unchanged during the first year after transplantation, while the oxidative burst was significantly reduced until day 100 returning to normal levels one year after transplantation. Patients (n=12) with low and stable levels of CD4 intracellular ATP content between day 30 and 100, showed an increased incidence of post-transplant complications (CMV reactivation, infections and relapse) in comparison with patients that increased the CD4 intracellular ATP levels in the same time period. Reduced T cell proliferation was not associated with low CD4 ATP intracellular ATP content. Transplanted patients with reduced T cell proliferation and low oxidative burst activity did not showed an increase in the incidence of post-transplant complications such as infections or relapse. Ex vivo assays of γδ + T cells were performed weekly until Hospital discharge and monthly until 6 months after HSCT. Phenotype of γδ + T cells was analysed using fl ow-cytometry; the following mAbs (CD3, CD45, pan-γδ, anti-Vδ1, -Vδ2, -Vγ9, CD45RO, CD45RA, CD27, CD16, CD56) were used to allow the identifi cation of the main γδ + T-cell subsets (naïve, central memory CM, eff ector memory EM, and terminally diff erentiated TD). Functional studies were performed using γδ + T cells both shortly after collection from pts and after in vitro expansion with zoledronic acid (ZOL) and IL-2. The cytotoxic activity of γδ + T cells was tested against primary leukemia cells; for these experiments, we used the CD107a degranulation assay and/or standard 51 Cr-release assay. Results: Until day +30 after HSCT, circulating T cells were consistently γδ + (>90%); subsequently, αβ + T cells began to increase over time. γδ + T cells included both Vδ2 + Vγ9 + and Vδ1 + Vγ9 +/cells, as well as, marginally, the Vδ1 -Vδ2 -Vγ9population. Of note, the Vδ1 + subset was signifi cantly higher than the Vδ2 + population in pts with cytomegalovirus (CMV) reactivation (P<0.01). On day +20 after HSCT, 10.5% of Vδ2 + cells on average were naïve, 53% CM, 23.5% EM and 13% TD. Similarly, 22.3% of Vδ1 + cells were naïve, 49% CM, 5.6% EM and 23% TD. The proportion of the different Vδ2 + and Vδ1 + T cell subsets changed signifi cantly over time, when comparing samples collected on day +20 with those obtained on day +90 or +180 after HSCT. In particular, within both Vδ2 + and Vδ1 + T-cell subsets, naïve and TD increased, whereas CM decreased over time; EM cells did not change. In pts experiencing CMV reactivation, γδ + T cells were mostly of the Vδ1 + phenotype (49% were TD, 10% EM, 21% CM and 20% naïve). We investigated the eff ect of exposure to ZOL and IL-2 on gd T cells in 21 pts; γδ + T cells consistently expanded in vitro, the resulting Vγ9Vδ2 population mainly comprising EM cells. These Vγ9Vδ2 cells lysed primary leukemia cells of both myeloid and lymphoid origin (as shown in a classical chromium-51 release assay in 8 pts); this eff ect was even more pronounced when leukemia targets were pre-treated with ZOL. Such activity was unambiguously TCRVγ9-mediated and dependent on phosphoantigen recognition. Finally, γδ + T cells obtained from 30 pts showed CD107a degranulation when challenged in vitro with primary blasts obtained from 12 pts with either lymphoid or myeloid acute leukemia. Discussion: We provide the fi rst detailed characterization of circulating γδ + T cells reconstituting in children given this type of T-cell-depleted haplo-HSCT. We demonstrate that γδ + T cells are important eff ector cells, which can be expanded and activated after exposure to bisphosphonates and IL-2 with the aim of improving their killing capacity against leukemia cells. Disclosure of Interest: None Declared. 

Introduction: Graft-versus-host disease (GVHD) after allogeneic hematopoietic stem cell transplantation (alloSCT) has been associated with low numbers of circulating CD4 + CD25 + FoxP3 + regulatory T-cells(T regs ). Because T regs express high levels of the IL-2 receptor, they may selectively expand in vivo in response to doses of IL-2 insuffi cient to stimulate T-eff ector T-cell populations, thereby preventing GvHD. Materials (or patients) and Methods: We prospectively evaluated the eff ects of ultra low-dose (ULD) IL-2 injections on T reg recovery in pediatric patients after alloSCT and compared this recovery with T reg reconstitution post alloSCT in patients without IL-2. Sixteen recipients of related(n=12) or unrelated(n=4) donor grafts received ULD-IL-2 post HSCT (100,000-200,000 IU/m 2 3xweekly), starting <day30 and continuing for 6-12weeks. Results: No grade 3/4 toxicities were associated with ULD-IL-2. CD4 + CD25 + FoxP3 + T regs increased from a mean of 4.8% (range, 0-11.0%) pre IL-2 to 11.1%(range,1.2-31.1%) following therapy, with the greatest change occurring in the recipients of MRD transplants. No IL-2 patients developed grade II-IV aGvHD, compared to 4/33(12%) of the comparator group who did not receive IL-2. IL-2 recipients retained T-cells reactive to viral and leukemia antigens, and in the MRD recipients, only 2/13(15%) of the IL-2 patients developed viral infections versus 63% of the comparator group (P=0.022). Discussion: Hence, ULD IL-2 is well-tolerated, expands a CD4 + CD25 + FoxP3 + T reg population in vivo, and may be associated with lower incidence of GvHD and viral infections. Disclosure of Interest: None Declared. Introduction: Although rates of engraftment have improved since the advent of hematopoietic stem cell transplantation (HSCT), graft failure confers high mortality due to diagnostic delay. Therefore, novel approaches to predict engraftment are of critical importance. We have shown that a novel nuclear medicine test FLT PET/CT could predict engraftment in rats; we hypothesized that this strategy would predict engraftment after HSCT. Materials (or patients) and Methods: Thirteen adult patients with acute leukemia received myeloablative conditioning with TBI/CY followed by (7/8 or 8/8) SD or UD BM or PBSC allogeneic HSCT. FLT imaging was performed peri-HSCT (at day -1 and day +5-12) (on NCT01338987). Analysis of images was performed using MIM Software, with standardized regions of interest of 10mm size used to capture standardized uptake value (SUV) in medullary spaces. Results: All 13 patients engrafted (ANC >500) at a median of 15 days for recipients of PBSC grafts and a median of 20 days for BM recipients. There were no toxicities attributable to FLT. FLT revealed a consistent pattern of marrow activity, fi rst to thorax, followed by cervical/lumbar, then sternum, and fi nally to extremities. This was shown by slope changes in SUV between day -1 and day +5-9 of: 0.1, 0.06, 0.04, and 0.03 respectively. Using thoracic peak SUV for those without relapse (n=10), FLT SUV inversely correlated with days to ANC >500 as a log slope of -0.29 (R2 of 0.822), predicting engraftment 3 to 25 days before ANC>500. The magnitude of FLT SUV distinguished between non-and early engraftment with at day -1 and at 5-6 days post-HSCT, of 0.65 and 1.1 respectively (P<0.01). Although all patients engrafted, 1 experienced secondary graft failure on day 60 and FLT SUV predicted the recovery without additional HSC infusion (0.93, outside the range of myeloablation).

Discussion: Hence, these data reveal that the kinetics of engraftment may be defi ned using FLT imaging as early as 5 days post-HSCT. In addition, these data provide unique insights into stem cell biology and homing post HSCT. In summary, our data support the safety and feasibility of FLT imaging for sensitive diagnosis of engraftment early after HSCT, thereby providing support for a future study in cord blood recipients. Disclosure of Interest: None Declared. diff erences in relation to diagnoses, relapse, various degrees of acute graft versus host disease (aGvHD) or concomitant GvHD and relapse. However, investigation of the impact of aGvHD on DC at diff erent time points revealed at an early time point after allo-SCT a signifi cant diff erence with regard to the immune recovery of mDC (P=0.015) and pDC (P=0.003) with faster immune recovery for the patients developing aGvHD. Survival rates for patients with a slow DC immune recovery (mDC<5/μl; pDC<4/μl) and the ones with a fast pattern of immune reconstitution (mDC≥5/μl; pDC≥4/ μl) after allo-SCT were investigated by using Kaplan-Meier survival curves. Myeloid DC recovery in the PB diff ered signifi cantly between patients with a slow and those with a fast recovery on day +120 (P=0.0463) and day +180 (P=0.0128), indicating better survival rates for the latters. Discussion: Knowledge about immune recovery pattern of DC subsets after allo-SCT, especially in comparison with DC in the PB of healthy children might contribute as an additional piece of the immunological puzzle in order to improve risk stratifi cation for patients who underwent allo-SCT for malignant diseases. Future analysis should also include monoclonal antibodies for HLA typing in order to allow a fast fl ow cytometric chimerism analysis between donor and recipient DC origin. Disclosure of Interest: None Declared. T cell full donor chimerism at D+60 and D+180 was signifi cantly associated with better overall survival (50 months vs 19 months p:0.04), 2 years disease free survival of (53% vs 32% p:0.03) and higher incidence (82%) of acute GvHD (p:0.01) and chronic GvHD (p:0.02) in the high risk AML group on both univariate and multivariate analysis. In the intermediate risk group though full donor chimerism was associated with improved disease survival on univariate analysis but this wasn't confi rmed on multivariate analysis and this was likely secondary to the small number of patients. In both univariate and multivariate analysis patients with full donor chimerism at 60 days were almost 80% more likely to have acute GVHD than patients with fully donor chimerism (p=0.04) and patients with full donor chimerism at D+180 were almost 85% more likely to achieve chronic GvHD (p:0.02).

Preemptive DLI infusion for mixed chimerism post D+180 was given to patients without prior severe GvHD after stopping Ciclosporin and was associated with improved disease free survival and a trend for improved overall survival for both risk groups by potentially augmenting the GvL eff ect. The 2 year incidence of graft-versus-host disease post DLI was only 29%. Discussion: Although the small number of patients included can be a limiting factor it seems that high risk AML group of patients will benefi t from acquisition of full donor T cell chimerism. In the intermediate risk group this eff ect was not demonstrated although molecular risk stratifi cation with fl t-3 and NPM-1 is needed to draw safe conclusions. Disclosure of Interest: None Declared. Flow-cytometry was used to prospectively assess CD19 + , CD20 + B-cell reconstitution on fresh blood samples before and at 1, 2, 3, 6, 9, 12 and 24 months after HSCT. Results: B-cell counts were already decreased before HSCT (median 16 cells per ml blood, range 0-343) and dropped even further within the fi rst month post-transplant (median 1/ml, range 0-297). Starting from the second month (median 4/ml), Bcell counts slowly but steadily increased (6/ml at 3 months, 68/ml at 6 months, 143/ml at 9 months, 151/ml at 12 months) to reach values in the normal range by month 24 (median 258/ml, range 22-625 Introduction: Natural Killer Cells are innate immune system cells important in host defenses against viruses and tumor cells. Two subpopulations are well described: NK CD56bright CD16neg (NK56++16-, lower frequency on peripheral blood-PB, high cytokine production) and NK56dim16pos (NK56+16+, higher frequency on PB, high cytotoxic activity). They are activated through a balance between signals given from activating and inhibitory receptors (KAR and KIR, respectively). The ligands of KIRs are the MHC molecules and in the absence of compatible MHC, NK cells are activated. In the allogeneic hematopoietic stem cell transplantation (HSCT), recent studies showed that NK cells recovery is important on infection control and, in the presence of a KIR-MHC mismatch, they may be important on graft versus host disease(GVHD) and graft versus leukemia eff ects. However few studies evaluated NK subpopulations recovery and HSCT endpoints. Materials (or patients) and Methods: NK (CD3-,CD56+) subpopulations (NK56++16-and NK56+16+) were quantifi ed by multiparametric fl ow cytometry at 9 sequential time points (before conditioning, at engraftment, and at days 3, 7, 14, 21, 60, 100 and 180 after engraftment). Overall, 111 patients, from 4 HSCT centers (65% male, median age 17 years, range1-74), receiving bone marrow (BM, 46%), umbilical cord (UCB, 32%) or peripheral blood (PB, 22%) from unrelated (n=90) or related donors (n=21) were studied. The most common diagnosis was acute leukemia (AML 36%, ALL 31%, MDS 9%, CML 9%, Aplastic anemia 8%). Most patients received myeloablative conditioning (MAC) regimens (60%). Antithymocyte globulin (ATG) was used in 44 patients (40%) and total body irradiation (TBI) in 56 (51%). Median follow up time was 14 months (range 4-35). Results: Eighty-six patients presented sustained allogeneic recovery (no diff erences among sources). Of these, median time to neutrophil engraftment was 18 days (range: 8-52). The cumulative incidence (CI) of non-relapse-related mortality (NRM) was signifi cantly higher in those with lower counts of NK56++16-during fi rst 3 weeks after HSCT (34% at 1 year for patients with less than 30 cells/uL at day 21 vs 11% for patients with higher counts, P=0.03). Overall survival was signifi cantly worse in patients with lower counts of NK56++16-subpopulations in the day 21 after engraftment (86% at 1 year vs 54% for patients with less than 30cells/uL -P=0.003). CI of grade II-IV acute GVHD and relapse were not signifi cantly aff ected by NK counts. The number of NK56+16+ cells did not aff ect any endpoint studied. Cell source, age and conditioning regimen did not aff ect any of the NK subpopulations counts. In multivariate analysis, NK56++16-counts lower than 30 cells/uL at 21 and 60 days after engraftment remain an independent risk factor for non relapse mortality Introduction: Takayasu's arteritis is a rare autoimmune vasculitis that aff ects medium and large vessels, especially the aorta and its main branches. Disease pathogenesis includes infl ammation of artery walls and subsequent narrowing of vessels with progressive impairment of blood fl ow. Materials (or patients) and Methods: We report outcomes of three patients with refractory Takayasu's arteritis treated with autologous hematopoietic stem cell transplantation (AHSCT), followed for 127, 43 and 15 months, respectively. Case 1: 41 yearold female, with a history of dizziness, claudication of upper and lower limbs, non-palpable right radial pulse and transient visual impairment for over 13 years. Arteriography showed irregularity and stenosis of abdominal aorta, right and left iliac arteries, left subclavian and carotid arteries. She had been unsuccessfully treated with steroids, methotrexate, cyclophosphamide, chlorambucil and mycophenolate mofetil. Case 2: 30 year-old female, with chronic history of intermittent visual defi cits, upper limb claudication, subclavian steal syndrome and low blood pressure measurements. Within the last year before enrolment, she had presented three episodes of transient ischemic strokes. Similarly to the fi rst patient, she had been previously treated with several immunosuppressive drugs, including steroids, azathioprine, cyclophosphamide and mycophenolate mofetil, besides a stent placed in the right carotid artery. Case 3: 39 year-old male with intestinal claudication, renovascular hypertension, upper limb claudication and dizziness. Magnetic angioresonance images evidenced right renal and mesenteric artery stenosis, and bilateral critical narrowing of carotid, subclavian, brachial and axilary arteries, as well as irregularities along the aorta. Patients were enrolled for AHSCT, mobilized with 2g/m 2 cyclophosphamide plus G-CSF, and had their hematopoietic stem cells harvested by leukoapheresis and cryopreserved, non-manipulated. In sequence, each patient received 200mg/kg IV cyclophosphamide and 4.5mg/kg rabbit anti-thymocyte globulin (ATG) over fi ve consecutive days, followed by infusion of the previously cryopreserved cells.

Results: The procedure was well tolerated by all three patients. Side eff ects included fever, rash, nausea and vomiting and neutropenic fever. None of the patients presented severe side eff ects during or after AHSCT. One patient presented positive CMV antigenemia after transplant and was preemptively treated, without any clinical manifestations of disease. On long-term follow-up, patients #1 and #3 presented remission of disease activity, with improvement of symptoms, normalization of acute phase reactant (C-reactive protein and ESR) levels and partial reversal of vessel stenosis. Patient #2 remained symptomatic and maintained high levels of acute phase reactants, indicating refractoriness to AHSCT. After transplant, she also failed etanercept injections, but disease activity was further controlled with tocilizumab infusions, started at approximately 2 years post-AHSCT and maintained until last follow-up. Discussion: Very few cases of vasculitides treated with AHSCT have been reported in the literature. To the best of our knowledge, this is the largest reported series of transplanted Takayasu's patients. Although response to transplant was not universal, benefi cial eff ects were verifi ed in two out of three patients, who remained in remission for long periods. Disclosure of Interest: None Declared. Introduction: Systemic sclerosis (SS) is one of autoimmune diseases in which autologus hematopoietic stem cells transplantation (ASCT) may be potential curative treatment, which may reverse of skin and organ fi brosis, and what is more can stop progression of the disease. Materials (or patients) and Methods: 16 patients-median age 48 (range 24-68 years) with a mean predominance (9 male and 7 female) with the rapidly progressive SS were treated with AHSCT between 2003 and 2013. All patients received previously treatment. Mobilisation of PBSC was achieved with Ctx with granulocyte-colony stimulating factor in 9pts, or G-CSF alone in 7pts. Conditioning regimen used combinations Ctx with Gemtuzumab ozogamycin in 15pts or Melphalan with Thymoglobulin in 2pts. Medium 4.0x10 6 CD34 positive cells/kg was transplanted and all patients received G-CSF since +1 days after transplantation until regeneration white blood cells (WBC>1G/L). Results: Mobilization of stem cells was successful in all cases. Engraftment was successful in 12 cases (2 patients died because of acute circulatory failure, one of them during conditioning regimen. 2 patients died before 30-days after transplant procedure, during aplasia period because of pneumonitis and respiratory failure). One patient died 6 months after transplantation because of circulatory failure. In one case disease progression occurred 1.5 years after transplantation requiring systemic treatment. The patient died 3.5 years after transplantation. One patient is lost to follow up. Nine patients are alive. All of them achieved disease improvement after autoHSCT. Major improvement was seen in skin (mRSS) and disability index. Organ function remained stable. Autoantibodies are still detectable. Median observation time is 4.7 years (range: 3-10 years). Discussion: Our preliminary results indicate that Auto-HSCT has favourable infl uence on course severe SS. The extension of the study by both increasing the number of included patients and prolonging follow-up time is required to confi rm current preliminary results. Disclosure of Interest: None Declared.

Introduction: In addition to physical complications, psychological aspects should also be considered in multiple sclerosis (MS) patients. The low quality of life and life expectancy of these patients justify the application of very aggressive ther apies such as high dose chemotherapy, immunotherapy and/or radiotherapy, with or without hematopoietic stem cell support. Because hematopoietic stem cell transplantation (HSCT) is an innovative approach in the treatment of autoimmune diseases (AID), there is an urgent need for studies that can as sess not only the eff ectiveness of the technique, but also its impact on patients' lives. The goal of our study was to evaluate the impact of HSCT in the Health-Related Quality of Life (HRQoL) of MS patients, two years after the procedure. Introduction: Autologous hematopoietic stem cell transplantation (AHSCT) for severe Systemic Sclerosis (SSc) has been developed in the past fi fteen years and shown as an eff ective therapy, based on its capacity of resetting the immune response (IR) and inducing de novo tolerance after transplant. However, no analysis of the very long term hematopoietic and immune reconstitution after AHSCT for SSc or for other autoimmune diseases has yet been performed. We therefore designed the present study to analyse long term IR after AHSCT for severe SSc. Materials (or patients) and Methods: Twelve SSc patients, who had been treated by AHSCT in a single transplant center hospital using the same mobilization procedure with cyclophosphamide (CYCLO) 4 g/m 2 and fi lgrastim 10 μg/kg/day and then conditioning, 6 weeks to 3 months later, with CYCLO (200 mg/kg total dose) without (n=5) or with (n=7) rabbit antithymocyte globulin (7·5 mg/kg total dose) followed by reinfusion of CD34+ selected AHSCT, were prospectively analysed. All patients were routinely followed using repeated measure of Fonctionnal Status, Rodnan modifi ed skin score, heart, kidney and lung functions for at least 2 years up to 11 years after AHSCT. Two groups were retrospectively constituted according to whether patients had a favorable clinical response (group A, responders; n = 5) or no response or relapse (group B, non responders; n=7) as previously published (Farge D et al, Arthritis and Rheumatism 2005; 52:1555-63). IR was studied before and at least yearly after AHSCT, using repeated clinical examination plus simultaneous extensive lymphocyte immunophenotyping, T cell receptor (TCR) diversity and thymic function analyses by spectratyping and T cell excision circles (TREC) quantifi cation. Results: Patients had similar characteristics at study entry. Before AHSCT, absolute numbers of CD4+ T cells and B cells were lower in group A comparing to Normal Controls (NC) (median TCD4+=450 microL -1 versus 680 microL -1 in NC; median BCD19+=70 microL -1 versus 143 microL -1 in NC), whereas an important increase of B-cells in group B was detected (median BCD19+=350 microL -1 ). A normalization of B-cell counts was observed from the second year after HSCT in group A (median BCD19+=170 microL -1 ), whereas a trend of B hyperlymphocytosis remained in group B (P=0.04) at 96 months after AHSCT. After 5 years, group A patients had shown higher T cell reconstitution compared to group B. The analysis of the T-cell repertoire showed the reconstitution of an overall broader clonal diversity in 4/8 SSc patients with sustained oligoclonal abnormalities in the others, irrespectively of clinical outcome. In our hands, the study of the thymic function was not informative to evaluate the immune reconstitution in this context. Discussion: To our knowledge, this is the fi rst study analyzing the long-term immune reconstitution of patients with SSc after AHSCT. We showed substantial diff erences between responders and non-responders according to the distribution of lymphocyte subsets, which may have further clinical implications for improving the conditioning regimen for SSC patients. However, the persistence of abnormalities even in good responders requires further investigations such as the analysis of lymphocyte subsets including B-cell diff erentiation and maturation. Disclosure of Interest: None Declared.

Introduction: Neuromyelitis optica (NMO) is an infl ammatory and demyelinating disorder of the central nervous system, recently recognized as a distinct disease hallmarked by pathogenic antiaquaporin 4 antibodies (AQP4). Currently NMO carries a poorer prognosis than multiple sclerosis. Use of Autologous stem cell transplantation (ASCT) has been reported worldwide for refractory autoimmune diseases (AD Median time between NMO diagnosis and transplant was 20 months. Before ASCT, the median EDSS (the Kurtzke Expanded Disability Status Scale) was of 6.5, 10/16 patients were positive for AQP4 antibodies and 11/16 had active lesions on magnetic resonance imaging (MRI). Fifteen patients were successfully mobilized with cyclophosphamide and G-CSF, one with G-CSF alone. The conditioning regimen consisted of BEAM plus anti-thymocyte globulin (9/16) or Thiotepa-Cy (3/16) or Cy and anti-thymocyte globulin (4/16). Hematopoietic recovery was documented in all patients within 10 days (range 3-25) after ASCT, both for neutrophils and platelets. Infectious complications required specifi c treatment in 9/16 patients (6 febrile neutropenia, 5 CMV and 2 VZV reactivations, 1 aspergillosis). All patients initially stabilized their disease. Disease progression was observed in 9/16 patients at a median of 10 months after ASCT. Relapse occurred in 13/16 patients at a median of 7 months after transplant, requiring further treatments in thirteen cases. Relapse, necessitating further treatments, occurred in 13/16 at a median of 7 months after ASCT. In the eight patients evaluable for AQP4 antibodies in the followup phase, the pathogenic autoantibodies remained positive after ASCT. Disease-free survival at 3 and 5 years was 31% and 10%, respectively, while progression-free survival at 3 and 5 years was 48%. No secondary malignancy was documented. All patients, but one patient who died from disease progression, are alive at a median follow up of 44 (14-128) months after ASCT. Discussion: This fi rst retrospective survey of the EBMT ADWP in 16 refractory NMO patients shows the potential of ASCT to reduce the highly infl ammatory picture typical of NMO in the short term, despite a tendency to relapse in the long term, together with a low incidence of toxicities. Disclosure of Interest: None Declared. Introduction: Neuromyelitis Optica (NMO) is a rare neurological autoimmune disorder characterized by recurrent attacks of optic neuritis (ON) and longitudinally extensive trasverse myelitis (LETM), associated with peculiar pathogenic auto-antibodies to aquaporin-4 (AQP4). Immunosuppression can halt disease progression, but some patients are refractory to multiple treatments. Thus, new therapeutic strategies for NMO are warranted. Here we update the successful follow-up of two patients with refractory NMO after allogeneic hematopoietic stem cell transplantation (HSCT). Materials (or patients) and Methods: Both treated patients had aggressive forms of NMO and serum positivity for AQP4 antibodies. Upn#1 is a 30 yrs old male aff ected by severe relapsing LETM, with paraparesis and thoracic spinal cord lesion on magnetic resonance imaging (MRI). Upn#2 is a 28 yrs female with recurrent attacks of ON and LETM. Both had received several lines of treatment without benefi t, including high-dose corticosteroids, cyclophosphamide, rituximab, natalizumab, alemtuzumab, plasma exchange, high-dose thiotepa/cyclophosphamide and/or BEAM with Anti-Thymocyte Globulin (ATG) and cyclosporine followed by autologous HSCT rescue. Before HSCT the two patients showed a Kurtzke Expanded Disability Status Score (EDSS) of 6 and 8.5 respectively, while both of them presented active contrast enhancing lesions on MRI and positivity for the pathogenic AQP4 auto-antibodies. Upn#1 underwent allogeneic HSCT from his HLA-identical sibling, while Upn#2 from a matched unrelated donor. Conditioning regimen consisted of full-dose treosulfan and fl udarabine. Graft versus host disease (GvHD) prophylaxis combined ATG-Fresenius with cyclosporine and a short course of methotrexate (Upn#1) or mycophenolate and rapamycin (Upn#2); B cell depletion of both patients and graft was obtained by rituximab. Results: Hematopoietic recovery occurred in both patients within day 30, and was accompanied by rapid achievement of full donor chimerism. Each patient experienced an episode of febrile neutropenia and one of them a CMV reactivation, both responsive to medical therapy; no serious adverse events were reported. None of the patients experienced neither acute nor chronic GvHD. MRI demonstrated a sustained disappearance of infl ammatory lesions, without any sign of disease progression. The immunological data obtained from the follow-up of these two patients suggest that allogeneic HSCT can alter the course of NMO through several concurring mechanisms: the eradication of autoreactive cell clones by high-dose chemotherapy and by the in vivo T and B cell depletion used for conditioning, elimination of long-lived plasma cells producing anti-AQP4 antibodies by putative donor T cell-mediated alloreactivity, the re-establishment of thymic central tolerance and renewal of the immune repertoire. Strikingly, these results correlated with a marked and sustained improvement of neurological functions in both treated patients. In UPN#1 EDSS dropped from 6 to 3.5, while in UPN#2 EDSS decreased from 8.5 to 7.5. At last follow-up (55 and 42 months after HSCT, respectively), both patients are alive, well and relapse-free. Discussion: Allogeneic HSCT can provide long-term disease control in NMO and should be considered as a treatment option for patients with aggressive and refractory forms of disease. Disclosure of Interest: None Declared. Introduction: RCDII can be defi ned as a low-grade intraepithelial lymphoma. This entity frequently transforms into an aggressive enteropathy-type-associated T cell lymphoma (EATL) with dismal prognosis. Current treatment strategies include cladribine (2-CDA) and autologous stem cell transplantation (AST) but their eff ectiveness in prevention of EATL is not well documented. Here we evaluated long-term follow-up and survival of 2-CdA monotherapy and 2-CdA-AST combination therapy. Materials (or patients) and Methods: Diagnosis of RCDII was based on persisting or recurring clinical symptoms and small intestinal villous atrophy, in patients with celiac disease, despite a strict GFD for over 12 months and a clinically validated cut-off value of more than 20% aberrant intraepithelial lymphocytes (IELs) detected by fl ow cytometric analysis. Patients received either 2-CDA monotherapy (n=25) or combination therapy (n=11) 2-CDA followed by AST. AST was performed in patients <70 years. Overall survival, EATL development and change in clinical, histological and immunological parameters were monitored. Results: Median age of the diagnosis RCD II was 63 years (range 42-78). Overall, 16 out of 36 patients (44%) died during a median follow-up of 59 months. In the monotherapy group, 14 out of 25 patients died (56%). Eight out of 14 (57%) died due to an EATL and 2 due to progressive refractory state. Progression into EATL developed after a median follow-up of 22 months after RCDII diagnosis (range 4-72 months). In the 2-CdA-AST combination therapy group, only 2 out of 11 patients died (18%), one as the consequence of an EATL. Overall survival in the monotherapy group was 47 months (range 3-106 months) compared to 87 months (range 18-190 months) in the combination-therapy group (P < 0.05). One, three-and fi ve-years OS in the combination therapy group was 100%, 91% and 81% respectively, compared to 81%, 66% and 56% in the monotherapy group. No signifi cant diff erences could be identifi ed in clinical characteristics, histological or immunological response in those who developed EATL and those who did not. Discussion: Both monotherapy and combination therapy induce improvement in the majority of patients, however progression to EATL was found and dismal outcome was found more often in the monotherapy group. These observations argue for an aggressive therapeutic approach for those RCDII patients eligible for AST. Disclosure of Interest: None Declared. Two patients died during follow-up of SSc from pulmonary and cardiac complications of the disease. One patient died from interstitial pneumonia at day + 65, leading to a TRM of 5.8% (1/17). Discussion: This study confi rms that AHSCT in selected patients with severe diff use cutaneous SSc results in sustained improvement of skin thickening and stabilisations of organ function up to 10 years after transplantation, so leading to a global clinical improvement, as showed by the persistent reduction in the EScSG clinical activity score. According to other experiences, TRM resulted reasonable, as a possible result of patients selection. These fi nding are in keeping with the view that AHSCT is eff ective in improving the infl ammatory and active phase of SSc, while letting unchanged and stable the fi brosclerotic one. Further studies are needed to evaluate the importance of CD34 selection, the need of immunosuppressive therapy post-AHSCT and the best timing of HSCT in the treatment of SSc patients. Disclosure of Interest: None Declared. (n=14), Connective Tissue disease (n=12), Immune cytopenia (n=5), Crohn's disease (n=3), infl ammatory demyelinating polyneuropathy (n=2), vasculatis (n=2), lupus (n=2), and rhumatoid arthritis (n=1). By the date of august 2013, median follow up was 52 months (range 1-142). At last follow up, 70 patients were Alive, 3 were lost of follow up and 24 were dead. Death was related to disease activity for 13 patients and related to AHSCT for 7 patients (2 of them had total body irradiation conditioning regimen). Thirteen of the 24 deaths occurred before the year 2001 (for a total of 34 AHSCT) whereas 11 death occurred after 2001 for a total of 63 AHSCT. Discussion: This retrospective national analysis provides useful informations regarding frequency, indications and broad outcomes in the context of translational and development phases of this treatment in poor prognosis and refractory SADs. There is therefore a case for a french national network (MATHEC) linking AHSCT centres with relevant regional level autoimmune disease specialists who are able to identify poor prognosis patients with refractory disease, before. Advanced, irréversible organ damage and chronic immunosuppression potentially compromise outcomes of AHSCT. National guidelines, written beyond the french society of bone marrow transplantation (SFGM-TC), will help centres dealing with general status for AHSCT, infection prophylaxis, supportive care and follow up after AHSCT. Disclosure of Interest: None Declared. Introduction: Autologous haematopoietic stem cell transplantation (HSCT) has recently been used to treat refractory Crohn's disease (CD). We analyze the long term eff ects of a protocol using unselected stem cells autotransplantation in refractory CD. Materials (or patients) and Methods: Eleven patients (4 male, age 22-46 years) with moderate-severe CD (median CDAI 336, range 236-395; four with perianal disease), refractory to multiple drugs including immunomodulators, were enrolled. Unselected PBSCs were collected after mobilisation with CTX 1.5 g/m 2 and G-CSF 10 mg/kg. The conditioning regimen included CTX 50 mg/kg on days -5 to -2 and rabbit ATG 2.5 mg/kg on days -4 to -2. Toxicity, clinical remission (CDAI < 150), endoscopic remission (SES-CD), extramucosal response (ultrasound sonography) and quality of life (IBD-Q) were assessed after mobilisation and at 3, 12 months and then every year after HSCT. Blood samples for immunologycal analysis were collected at baseline, after mobilisation, and 3, 6 and 12 months after transplantation. Immunologycal analyses evaluated: 1) CD4+/CD25high+/FoxP3+ regulatory T cells (Tregs); 2) Toll-like receptor 2-(TLR2) and TRL4-expressing monocytes (CD14+ cells); 3) IL-12, IL-10, TNF-alpha-production by mitogenstimulated CD14+ cells and IFN-gamma production by CD4+ T cells. Immunological results were compared with healthy donors and associated with clinical and endoscopic response during 12 months of follow-up. Introduction: AB0 incompatibility is not considered as an obstacle in hematopoietic stem cell transplantation (HSCT). However, might be associated with specifi c complications: delayed erythropoesis recovery, higher level of transfusion support required and pure red cell aplasia (PRCA) in major incompatibility, acute haemolytic reactions during fi rst three months after HSCT in minor incompatibility. The aim of this analysis was to assess if AB0incompatibility leads to delayed engraftment, higher incidence of graft versus host disease (GvHD), disease-free survival and overall survival after unrelated allogeneic HSCT. Introduction: Complications associated with allogeneic hematopoietic stem-cell transplantation (allo-SCT) can lead to critical illness with multiple-organ failure and allo-SCT recipients are considered as poor candidates for intensive care unit (ICU) admission. The decision to admit patients to ICU following allo-SCT remains a matter of debate and prognostic markers are lacking to help in the decision making process. Moreover the impact of pre-ICU admission factors on long-term survival have been rarely analysed. Materials (or patients) and Methods: We retrospectively analysed the factors aff ecting short and long term survival in a single centre cohort of allo-SCT patients admitted to ICU. 

S252 inclusion of high-dose CY in transplant regimens may sometimes incidentally lead to immediate death. However, the mechanism underlying this phenomenon has not yet been elucidated. In addition, the occurrence of acute heart failure following treatment with high-dose CY is unpredictable.

In this study, we investigated the cardiotoxic mechanisms of CY and evaluated the protective eff ects of antioxidants in order to determine the preliminary mechanisms that prevent the occurrence of CY-induced cytotoxicity in H9C2 cells. Materials (or patients) and Methods: A rat cardiac myocardial cell line, H9C2, was exposed to CY or CY metabolites, 4-hydroxycyclophosphamide (HCY) or acrolein (Acr), with or without various antioxidants (N-acetylcysteine (NAC), dexrazoxane, silymarin, isorhamnetin (ISO)). The degree of cytotoxicity was then evaluated using a MTT assay, lactate dehydrogenase (LDH) release measurements and reactive oxygen species (ROS) production measurements. In addition, we observed the cytotoxicity of CY and CY metabolites and the protective eff ects of various antioxidants using a real-time, live-cell imaging system (Incucyte™). Using the leukemia cell lines HL-60 and U-937, we confi rmed whether each of the above antioxidants inhibit an antileukemic eff ect. We also investigated how the myocardial cellular eff ects of CY was changed by S9 mix. S9 mix is rat liver homogenate (S9) with co-factors. We thereafter developed and validated an assay using liquid chromatography coupled with electrospray tandem mass spectrometry (LC/MS/MS) for the quantifi cation of CY and CY metabolites, as well as HCY and o-carboxyethylphosphoramide mustard (CEPM). We then assayed the culture supernatant of CY plus S9 mix. Results: The MTT and LDH release assays showed that treatment with CY (125-500 μM) and Acr (10-100 μM) did not cause cytotoxicity. In addition, neither of these compounds increased the ROS production. However, HCY (1-10 μM) exhibited myocardial cytotoxicity in a concentration-dependent manner and increased the ROS levels, including superoxide, hydrogen peroxide (H 2 O 2 ), hypochlorous acid (HOCl) and the hydroxyl radical (·OH). Both NAC and ISO inhibited HCY-induced cytotoxicity; however, the other antioxidants did not. A real-time live-cell imaging system also confi rmed acute cytotoxicity induced by HCY and the inhibition of this cytotoxicity. Neither NAC nor ISO inhibited the HCYinduced antileukemic eff ect. CY (250 μM) plus S9 mix induced cytotoxicity more severely than HCY (10 μM). In addition, the concentration of HCY was approximately 10 μM when 250 μM of CY was metabolized for two hours using the S9 mix. Both NAC and ISO inhibited CY plus S9 mix induced cytotoxicity. Discussion: CY itself did not exert a cardiotoxic eff ect. The cardiac toxicity of CY may thus be primarily caused by HCY, not Acr. Furthermore, HCY induced cardiotoxicity via variable ROS generation. However, treatment with CY plus S9 mix induced cytotoxicity more severely than treatment with HCY alone. Therefore, various metabolites of CY additively, rather than HCY alone, induce cardiotoxicity. Since both NAC and ISO inhibited CY plus S9 mix induced cytotoxicity, these compounds potentially serve as novel cardioprotective agents that can prevent the occurrence of highdose CY-induced cardiotoxicity. Disclosure of Interest: None Declared. Introduction: Almost all patients undergoing hematopoietic stem cell transplantation (SCT) develop some form of mucositis due to conditioning chemotherapy. This results in decreased food intake and often requires invasive nutritional support in the form of tube feeding or parenteral feeding. Most centers in The Netherlands prefer parenteral feeding in this patient category due to the perceived risk of gastrointestinal bleeding and/or perforation during tube placement. Due to lack of evidence concerning this issue and extensive experience in duodenal tube feeding in patients with mucositis at St. Antonius Hospital Nieuwegein (The Netherlands), we performed a retrospective study to evaluate the eff ectiveness of duodenal tube feeding patients with mucositis. Materials (or patients) and Methods: A retrospective database study was conducted in which all patients who underwent autologous SCT and received a duodenal tube during hospitalization between January 2010 and March 2013 were included. Primary endpoint was the safety of duodenal tube feeding as measured by the number complications that occurred. Secondary endpoint was the eff ectiveness of duodenal tube feeding expressed in weight loss and failure. All required data were retrieved from computerized medical records. Results: There were 63 patients included in this study. No gastrointestinal bleeding or perforation occurred. Complications were reported in 39.7% of all patients and were mainly of mechanical nature, notably dislocation. Mean weight loss at discharge was 3.0 kg (SD ± 5.5) and in 42 patients (68.9%) , weight loss at discharge was no more than 5% of initial weight at admission. A switch to parenteral feeding (failure) was required in 14 patients (22.2%). Discussion: Our study shows that duodenal tube feeding is a safe method for providing nutritional support in SCT patients with mucositis. It may also have an acceptable eff ectiveness in the prevention of weight loss and is relatively well tolerated by patients. Disclosure of Interest: None Declared. OS was 50% at 3 years, it was 55% in "Defi brotide Treated" and 45% in "No Defi brotide" groups (P=0.6). No patient suff ered of severe Liver toxicity, clinically mild SOS was diagnosed in 3.9% of patients in "Defi brotide treated" group and in 5.3% in "No Defi brotide" group (P=0.7); Bilirubin above 2.0 mg/dl was registered during the fi rst 30 days in 21% of patients in "No Defi brotide" group and in 15% of "Defi brotide treated" group (P=0.4). Discussion: Prophylaxis using Defi brotide after allogeneic hematopoietic transplantation using a myeloablative conditioning determines, in adults patients, a signifi cant reduction of TRM. Reduction of Liver toxicity did not reach statistical signifi cance in the sample we have studied. Thus, benefi cial eff ect of Defi brotide on TRM is independent from any lessening of liver toxicity and may depend from modulation of allo-reactivity. Disclosure of Interest: None Declared.

Introduction: It is known that the levels of cardiac troponins can be raised as a consequence of cardiotoxic action of chemotherapeutic agent used in oncology. In the last few years high-sensitive assay for cardiac troponin measurement became widely used in cardiological practice. However, there is no data to support the predictive value of cardiac troponin levels in patients with malignant lymphomas (ML) during the high-dose chemotherapy (HDC) combined with autologous hematopoietic stem cells transplantation (auto-HSCT).

It is known that the levels of cardiac troponins can be raised as a consequence of cardiotoxic action of chemotherapeutic agent used in oncology. In the last few years high-sensitive assay for cardiac troponin measurement became widely used in cardiological practice. However, there is no data to support the predictive value of cardiac troponin levels in patients with malignant lymphomas (ML) during the high-dose chemotherapy (HDC) combined with autologous hematopoietic stem cells transplantation (auto-HSCT). The goal of the study was to compare the changes in the troponin T («conventional troponin») and troponin I levels measured by the high-sensitive assay (hs-cTnI) at diff erent time points before and after the HDC and auto-HSCT in ML patients. Materials (or patients) and Methods: 73 patients with ML were enrolled in the study. Troponin T was measured in 52 patients (group 1), hs-cTnI was measured in 21 patients (group 2). The groups were matched for age and sex. Serum troponin levels in both groups of ML patients were evaluated before the HDC, directly after HDC (D0), on day 7 and 12 after auto-HSCT (D+7 and D+12 S255 respectively. They were treated with intrapulmonary administration of 50 μg/kg rFVIIa by bronchoalveolar lavage (BAL), concurrently with methylprednisolone pulse thereapy, fresh frozen plasma and maintenance of platelet count >50,000/mm 3 . Results: Complete and sustained hemostasis after a single dose of rFVIIa was observed in both patients. No toxicity or adverse events were observed with rFVIIa treatment. The oxygen capacity (PaO 2 /FiO 2 ratio) increased signifi cantly, and rapid clinical and radiological improvements were observed. However, they suff ered from subsequent infection, one with MDS died from respiratory syncytial virus pneumonia, and the other infant with HLH died of multidrug resistant Acinetorbactor baumanni infection at day 51 and 70 post-transplantation. Discussion: Our experience indicates that intrapulmonary administration of rFVIIa is an eff ective and safe treatment option for DAH after HSCT in children; Although high-dose corticosteroids has been the mainstay of treatment of DAH, high-dose corticosteroids deteriorate immunity and increases the risk of opportunistic infections. Therefore, new treatment modalities for DAH are desperately needed, not only to aid in the cessation of the acute bleeding episode, but also to spare the use of steroid. Although this study is limited by small numbers of patients, and further clinical studies are needed, intrapulmonary rFVIIa could be successfully used in pediatric patients with DAH after HSCT and it might spare the use of steroid. Disclosure of Interest: None Declared.

S. Christy Lindgaard 1 patients had a level 26-500, 33 (21%) had a level 500-2500, and 94 (59%) >2500. There was no signifi cant correlation between the diff erent ferritin levels, disease kind and status at HSCT. Results: After transplantation, 23 patients received iron chelating agents after a serum ferritine level of 1000 microg/l and stopped when the level decreased below 1000. The cumulative incidence of acute GVHD ≥ II at 3 months was 14% (11-16.5) with 10.5% (8-13) for grade III and 7% (5-9) for grade IV; the 1 year cumulative incidence of limited and extensive chronic GVHD were 4% (2-6) and 12.4% (9-16) respectively. After a median follow-up of 18 months (1-106), the 5 years OS probability was 65% for patients with ferritin level below 500 microg./l, 39% for level between 500 and 2500 microg./l and 28% for level >2500 micog./l, [Hazard ratio= 3.5 (1. 3 Pediatrics and Community Health Sciences, University of Calgary, 4 Pediatrics, Alberta Children's Hospital, 5 Cellular Therapy Lab, Foothills Medical Center, Calgary, Canada Introduction: Adverse reactions during hematopoietic stem cell infusion are common in patients undergoing stem cell transplantation (SCT), mainly due to high white blood cell (WBC) content as shown in adult studies and red blood cell replete cord blood stem cells. We studied a large cohort of pediatric patients undergoing SCT to determine the incidence, grade and predictors of adverse reactions. Materials (or patients) and Methods: All patients undergoing autologous and allogeneic SCT at our institution from 2004 to 2012 were included. We reviewed clinical data from medical records to collect adverse reactions during the infusion of stem cells on Day 0 of SCT, as well as for 24 hours following the end of stem cell infusion. Adverse reactions were graded according to the National Cancer Institute grading criteria (NCI CTCAE version 4.0). Covariates of interest included patient-related data, transplant type, stem cell source, and cellular content (total nucleated cell, CD34, CD45, CD3, granulocyte) of the infused product. Statistical analysis was performed using t-tests, chi-square and multivariate logistic regression. Results: In the nine year period, 120 allogeneic and 93 autologous patients received SCT. The majority of allogeneic SCT were for leukemia while autologous SCT were mostly for brain tumors and solid tumors. Some patients received 2 or more transplants (15 allogeneic and 37 autologous patients) and others received multiple stem cell infusions in a single transplant (7 allogeneic and 30 autologous transplants), contributing to a total of 361 infusion episodes. Grade 2 reactions occurred in 109 (30%) infusion episodes and grade 3 reactions occurred in 28 (7.7%). There were no grade 4 or 5 reactions. Results: One houndred-eightyeight and 125 HSCT were performed in oncologic and non oncologic patients, respectively, for a total of 313 consecutive procedures in 287 patients (162 male). Twenty-six out of 313 HSCT procedure (8,3 %) were complicated by the development of PRES. Univariate analysis identifi ed the following risk factors for PRES: age at HSCT > 2 years, hemoglobinopathy, G-CSF administration, MMUD, CB stem cells source and fl udarabine-based conditioning. Multivariate analysis confi rmed the use of CB as stem cells source to be an independent risk factor for the development of PRES (Tab. 1). Eight-year OS of patients with PRES did not signifi cantly diff er from those of patients who did not experience this complication (35.1% compared to 56.1%, P: 0.27). Discussion: Our result showed a strong association between CB HSCT and PRES. In this setting, steroid therapy used for GVHD prophylaxis may partially explain the association with PRES onset through exacerbation of hypertension which is a renowned risk factor for PRES. Interestingly, fl udarabine based regimen seems to be associated with PRES development as previously published in adults 1 outcomes including non-relapse mortality (NRM) after allogenic SCT. However, the post-transplant course of BMI and its eff ect on the outcome of allogeneic SCT has not been well studied. We hypothesized that the longitudinal course of BMI may provide helpful information regarding the risks of complications and NRM during the fi rst three months and the fi rst year after allogeneic SCT. Materials (or patients) and Methods: Data from 98 adult patients, who underwent allogenic SCT in our centre between 2005 and 2013, were retrospectively analysed. Body weight and relevant laboratory parameters (total protein, albumin etc.) were recorded before conditioning as well as on day +30, day +120, day +180 and day +365 after SCT. Data were censored due to death, relapse or progress of disease or reached day +365 past SCT. BMI was calculated by standard formula and categorized as follows: underweight <18.5kg/m2, normal weight 18.5-<25kg/m2, overweight 25-<30kg/m2, obesity >30kg/m2. The longitudinal course of weight and BMI-category were evaluated in the following time intervals: start of conditioning until day +120 (IN1) and day +120 until day +365 past SCT (IN2 Discussion: After allogeneic SCT the majority of patients lose weight in the fi rst post-transplant interval until day +120. In contrast, the second interval from day +120 until day +365 seems to be characterized by a diff erential course of patient weight. Our preliminary data suggest that this diff erential weight course may correlate with NRM in the second interval while weight loss during the fi rst interval is not associated with NRM in the second interval. More detailed data on the association of weight changes as well as of changes in laboratory data with NRM for the entire patient group will be presented at the meeting. Disclosure of Interest: None Declared. [3] . Diff erences in frequencies of discrete variables were tested using two-sided Fisher's exact. Univariate and multivariate analysis was conducted on the following risk factors: age at HSCT, sex, high risk, type of donor and type of diagnosis, conditioning regimen, total body irradiation, busulfan based regimen, acute and chronic GVHD, donor sources, ABO blood group incompatibility, Rh incompatibility, HLA, CMV reactivation. The median follow up time was of 60 months (range has the potential to identify diagnostic biomarkers and provide novel insights into the mechanisms underpinning its pathogenesis. Our overall objective is to characterize the global protein expression of IPS to identify novel pathways that diff erentiate IPS from infectious lung injury in HSCT recipients. Materials (or patients) and Methods: Patients with lung injury within 180 days of Allogeneic HSCT were classified as IPS based on criterion outlined in 2011 American Thoracic Society statement. For determining the global protein expression profile BALF was depleted of high abundance proteins, treated with trypsin and labeled with iTRAQ® 8-plex reagent for mass spectrometry (MS). The complex mixture of iTRAQ® labeled peptides was analyzed by 2D capillary LC-MS/MS on an Orbitrap Velos system in HCD mode for data dependent peptide tandem MS. Protein identification and relative quantification was performed using a target decoy strategy. Relative abundance of the proteins was determined with reference to pooled BALF from patients with no history of HSCT and respiratory failure (controls). The proteins that were differentially expressed were identified using false discovery rate of the t-test empirically using the random permutation testing. To determine the biological relevance of the proteins identified, Gene Ontology enrichment analysis was performed using the Functional Annotation Clustering algorithm in Database for Annotation, Visualization and Integrated Discovery (DAVID). Results: BALF was available on 21 patients with infectious lung injury after HSCT and 11 patients with IPS. The mean duration from transplantation to BALF collection was not diff erent in the two groups (56.9 ± 53.3 vs. 59.6 ± 39.4 in the two groups, respectively). No statistically signifi cant diff erence was seen in age, serum creatinine, BALF leucocyte count, neutrophil and lymphocyte count. BALF protein profi le is available from 4 patients with infectious lung injury and 2 patients with IPS. We identifi ed 542 proteins at a global FDR of 1%. These proteins represent diverse biological processes such as programmed cell death, proteasome-ubiquitin dependent protein catabolism, cellular ion homeostasis, response to oxygen radicals, carbohydrate catabolism, actin fi lament polymerization, plasma protein involved acute infl ammation and protein remodeling. Of these, 16 proteins demonstrated diff erential expression in subjects with IPS when compared to infectious lung injury. These proteins will be explored as potential diagnostic biomarkers for IPS. Discussion: High-resolution mass spectrometric platforms provide extended proteome coverage in BALF. A protein expression profile in BALF likely represents the pathophysiologic mechanisms involved in development of IPS. Early findings suggest presence of candidate biomarkers in the BALF for rapid diagnosis of IPS. Disclosure of Interest: None Declared. Results: Overall, the majority of long-term survivors were working or studying at follow-up. Nevertheless, a substantial number was on sick leave/disability pension presumably associated to sequelae from their underlying disease or treatment. At a median of 17 (3-28) years following the HSCT, 55% were working (44% full-time, 11% part-time), 28% studying (24% full-time, 4% part-time), 7% unemployed (5% full-time, 2% part-time) and 10% on full-time sick leave (>3 months) or had disability pension. In total were 19% (10% full-time and 9% part-time in combination with work, studies or unemployment) on sick leave or disability pension. Decreased physical and mental/social work capacity due to the HSCT was reported by 18% and 19% of the participants' , respectively, and 32% reported economic diffi culties. The reduced physical and mental/social work capacity was significantly associated with being on sick leave/disability pension (P = 0.01 and P = 0.02, respectively), while there were no such associations with having economic diffi culties. Being on part-or full-time sick leave/disability pension was associated with selfreported multi-morbidity (≥2 co-morbidities: OR 12.6, CI, 1.99-80.09, P = 0.01). Discussion: Overall, the majority of long-term survivors were working or studying at follow-up. Nevertheless, a substantial number was on sick leave/disability pension presumably associated to sequelae from their underlying disease or treatment. Disclosure of Interest: None Declared. 

Introduction: The attainment of transfusion independence after transplant is sometimes hampered by a combination of factors, ranging from infections to the need of combined therapy for clinical complications, as well as control of GVHD. Leucopenia and thrombocytopenia could be frequently observed. Pancytopenia obviously complicates the management of the post transplant period. Iron overload is frequently observed in hematological patients before and after hematopoietic stem cell transplantation (HSCT), as multiple red blood cell (RBC) transfusions are generally administered, owing to the presence of disease and marrow suppression after cytotoxic treatment. Moreover, iron overload is considered a signifi cant contributor to treatment related mortality, in thalassemias as well as in leukemias and myelodysplastic syndromes (MDS). Whereas several reports have focused on iron overload before transplant, little is known, up to now, on the eff ects of iron overload on the recovery of hematopoiesis after transplant. Materials (or patients) and Methods: we report on 10 patients, transplanted for hematological diseases (9 acute leukemia, 1 aplastic anemia) heavily transfused before transplant that, considering the iron overload, were treated after HSCT with deferasirox. Before starting deferasirox, the patients were fully engrafted and in complete remission (acute leukemia), although transfusion dependent, and with incomplete hematological reconstitution after allogeneic HSCT. Patients were selected according to the following inclusion criteria: 1) transfused pretransplant with more than 20 RBC units; 2) incomplete haematological recovery; 3) transfusion-dependence; 4) serum ferritin > 1800 ng/mL; 5) normal creatinine value. All patients received an initial dose of deferasirox 10 mg/kg/day, later adjusted according to side effects. Results: All patients experienced an increase in haemoglobin levels, with a reduction in the frequency of RBC transfusions, followed by transfusion independence (median time: 23 days from the fi rst dose of deferasirox). In addition, it was promptly (median time: 26 days) associated with haematological improvement, with sustained values and no further platelet support or growth factors administration. Moreover, ferritin values were progressively reduced with deferasirox treatment. The workup for other aetiologies resulted negative; no concomitant infection was documented (CMV: negative; HHV-6: negative; EBV: negative). No relevant modifi cations with immunosuppressive or myelosuppressive drugs were made during deferasirox treatment. Deferasirox was well tolerated. Discussion: The role of iron overload post transplant is not completely understood. No reports, in our knowledge, have up to now focused on the possible effect of iron chelators after transplant on the restoration of normal hemopoiesis and, in particular, transfusion independence. Basing on our results, we think that deferasirox determined stimulatory, and/or derepressive effects on hematopoiesis after allo-HSCT. In conclusion, this clinical experience raises the possibility of a potential additive benefit on hematopoiesis after transplant following iron chelation therapy with oral deferasirox. Further long term studies, in larger cohorts of patients are needed to confirm these data and design an efficient strategy to reduce iron loading after transplant.

We present our preliminary experience with intravescical hyaluronic acid for the treatment of HC in patients receiving an allogeneic HSCT. Materials (or patients) and Methods: Between december 2011 and november 2013, 12 patients with hematological malignancies presented with HC after receiving an allogeneic HSCT from a matched sibling (n=1), an haploidentical relative (n=1) or an unrelated donor (n=10). Stem cell source was peripheral blood in 11 patients and bone marrow in 1 patient. All patients received myeloablative conditioning regimens including cyclophosphamide (Cy) in 7 cases. Haploidentical HSCT recipient received post-transplant Cy as in vivo T-cell depletion. In all cases the administration of Cy was combined with 2-mercaptoethane sodium (MESNA) and hyperhydration. Results: HC developed at a median of 44 days (range 27-119 days) post-HSCT. HC has been classified as grade II (macroscopic haematuria) in 4 patients, grade III (with the presence of clots) in 4 patients and grade IV (requiring instrumentation for clot evacuation or leading to urinary retention) in 4 patients. The median platelet count at the time of HC onset was 70 x 10 9 /l (range 24-234). Ten evaluable patients had BK viruria (viral titres: 5 x 10 6 -1 x 10 8 copies/ml) and 3 of these were treated with intravescical (n=1) or i.v. cidofovir (n=2). Seven patients had concomitant CMV infection requiring antiviral treatment, and 6 patients had grade II-IV acute GVHD at the time of HC onset. Slow instillation over 10-20 minutes of 50 ml of hyaluronic acid (Cystistat) was performed through a single use hydrophilic transurethral catheter. Patients were invited not to void the bladder for at least one hour. In 5 cases propiverine 15 mg p.o. bid was associated with the instillations. Hyaluronic acid was administered at a median of 7 days (range 1-27 days) from the onset of HC: patients received instillations 1-2 times/week on the basis of the clinical status, for 2-4 weeks. Patients received a median number of 4 intravescical instillations (range 2-10). In 10 cases (83%) instillations were performed as outpatients. A complete response, with improvement of urinary symptoms and resolution of gross haematuria, has been reported in 8 patients (67%); two patients did not respond to the instillations; one patient experienced a relapse of HC following early discontinuation of the treatment and responded to hyperhydration with bladder irrigation a second course of intravescical instillations. One patient had initial response to the first three doses and is still receiving the treatment. One patient did not tolerate the intravescical treatment; no other adverse events were reported. Discussion: In conclusion, our results suggest that intravescical instillations of hyaluronic acid may represent a safe and minimally invasive treatment for patients with post-HSCT HC.

Interesting to note that the therapeutic procedure has been performed in the outpatient setting in a consistent proportion of patients. Further studies are mandatory to confirm our preliminary results. Disclosure of Interest: None Declared. Introduction: Graft failure is a rare but life-threatening complication after allogeneic transplantation (SCT). Second allogeneic SCT are the main therapeutic options, and immunopressive agents rather than myeloablative agents were adopted in conditioning regimen. In this retrospective study, we report the results of second allogeneic SCT containing fl udarabine-based regimen for secondary graft failure. Materials (or patients) and Methods: Nine patients with secondary graft failure who received second transplantation in Peking University Institute of Hematology from 2002 to 2013 were retrospectively reviewed. Secondary graft failure was defi ned as persistent neutropenia and thrombocytopenia after initial reconstitution. Results: Nine patients received second transplantation with fludarabine-based regimen, including 4 male and 5 female. The median age was 28 (3-39) years old. There are 5 malignant disease and 4 aplastic anemia. The onset of secondary graft failure was 6 (2-36) months after first transplantation. The conditioning regimen was fludarabine plus basiliximab or busulfan or cyclophosphamide or total body irradiation. As for the graft, 7 was peripheral blood, 2 was combination of bone marrow and peripheral blood. Only 2 patients received second transplanta-tion from a different donor. Seven of 9 patients acquired fast engraftment, with the median neutrophil engraftment time was 14 (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) days and the median platelet engraftment time was 12 (7) (8) (9) (10) (11) (12) (13) (14) days. One patients occurred grade IV acute GVHD, and 4 patients with extensive chronic GVHD. Finally, 5 patients survived with median survival of 45 Results: All patients in this cohort had hematological malignancies, the median age was 52 years (range, 28-66). Eight patients (50%) had a matched related donor, 7 patients (43.75%) had a HLA-10/10 matched unrelated donor and 1 patient underwent alloHSCT with two cord blood units. Seven patients (43.75%) received a myeloablative regimen with 12 Gy TBI and 9 patients (56.25%) a reduced intensity regimen. The stem cell source was bone marrow for 7 patients (43.75%), PBSC for 8 patients (50%) and cord blood for one patient. The status at transplantation was CR1 for 6 patients (37.5%), CR2 or more for 6 patients (37.5%) and partial response for 4 patients (25%). The median follow-up of the cohort was 7 months (range, [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] . Non-parametric tests such as exact Wilcoxon Mann-Whitney test or Kruskal Wallis were performed for the analysis of the physiological parameters. The median of serum creatinine was 112 μmol/L (range, 51-262) with CsA and 87 μmol/L (range, 50-125) with tacrolimus ER (P <.0001). The reduction of serum creatinine level was correlate with an improvement of GFR for all patients in the study. The median of potassium level was 4.2 (range, 3.5-5.2) with CsA and 4.0 (3.4-4.5) with tacrolimus ER (P =.001). The median time of switch was 45 days (10-162) and the median of serum creatinine in which the switch was realized was 138 μmol/L (range,110-262). The median blood trough level was 300 μg/L (range,100-438) for CsA 14 days after starting and 7.2 (range, [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] for tacrolimus ER 20 days after starting. The cumulative incidence of aGvHD grade >1 at 3 months was 25% (95%CI,14-36). After the switch for tacrolimus ER, no patient developped aGvHD. Two patients in this study developed severe cGvHD after the discontinuation of prophy-laxis (5 months and 10 months). The survey demonstrated that patients were satisfied with this formulation of tacrolimus and the adherence was reinforced. Discussion: The conversion from CsA to tacrolimus ER was followed by a signifi cant improvement in kidney function and in adherence to immunosuppressive treatment. We have now enlarged the study to several centers to confi rm these encouraging observations. Disclosure of Interest: None Declared. Introduction: Sinusoidal obstruction syndrome (SOS) has been a common and potentially life-threatening complication after allogeneic hematopoietic stem cell transplantation (HSCT). Incidence of SOS has decreased in recent years according to several reports, in particular severe and fatal cases. This improvement has been thought to be related to increased use of less toxic and reducedintensity conditioning regimens. Materials (or patients) and Methods: In the present study we specifi cally focused on patients undergoing myeloablative preparatory treatment with busulfan (16 mg/kg) and cyclophosphamide (120 mg/kg) (BuCy) before allogeneic HSCT during the period of 1990-2013. In total 397 patients were given BuCy. All patients received oral busulfan. Since the end of 1990's dose adjustment based on pharmacokinetic analysis has been performed for busulfan. Before 1995 at our centre, females were treated with norethisterone to delay menstruation and had an incidence of SOS of 27% as opposed to 3% in females without this treatment (P=0.007). Therefore norethisterone therapy was discontinued in 1998.The most common diagnosis was AML (n=181) followed by MDS (n=65) and chronic leukemias (n=56 All patients fulfilled the diagnostic criteria for VOD, including hepatomegaly and weight gain. All but two patients had elevated bilirubin levels (>2 mg/dL). 8/16 (50%) patients had evidence of reversal of portal vein blood flow on Doppler ultrasonography. All patients were treated with corticosteroids and thirteen patients (81%) were treated with Defibrotide, with treatment commencing on average 3 days (range 2-7 days) post appearance of VOD signs and symptoms. Defibrotide treatment was continued for an average of 11 days (range 4-21 days). Ten patients (62.5%) had severe VOD, all were treated with Defibrotide and corticosteroids, six (60%) died of VOD-related multiorgan failure. We have recently used defibrotide prophylaxis in two HSCT patients with several risk factors for development of VOD. Discussion: VOD is a severe complication of HSCT, with high rates of morbidity and mortality despite early initiation of defibrotide and corticosteroid treatment. The use of defibrotide for prophylaxis of VOD in identified high risk patients has been recommended, and seems promising. Large scale studies are needed to evaluate the efficacy of VOD prophylaxis with defibrotide. Disclosure of Interest: None Declared. , FWB (mean=20.5; SD 5.2, range: 2-28) showed a value in the upper 30% of the possible evaluation. In multivariate analysis, the inferior QOL score was reported for patients with aGVHD (P=0.0016), cGVHD (P=0.0006), QOL decreased with increasing age (P=0.048) and increased with time elapsed since allo-HSCT (P=0.0002). Discussion: Allogeneic HSCT represents an important intervention into the overall integrity of the organism. In particular, the development of GVHD can cause very serious organ, but also mental problems which can signifi cantly reduce QOL. All patients should be well informed, not only about the risks of treatment failure, organ complications or possible death, but also about chronic psychological consequences and the expected deterioration in the QOL, which can be fi nally perceived worse than the actual disease. Introduction: The recipients of solid organ transplants are infrequent candidates for allogeneic hematopoietic stem cell transplantation (alloSCT). Since the reported clinical experiences have been anecdotal, outcomes of alloSCT after SOT are not well known. In order to document the feasibility of alloSCT after SOT, we decided to perform a retrospective analysis of patients reported to the EBMT registry. Materials (or patients) and Methods: In 18/101 centers, which responded to the survey, we identified 28 patients who fulfilled the criterion: alloSCT performed after SOT. This group included 12 patients with kidney, 13 with liver and 3 with heart transplant. The indications for alloSCT were: acute leukemia (N=10), chronic leukemia (N=4), lymphoma (N=3), MDS/MPN (N=4), bone marrow failure (N=4) and inherited disorders (N=2). The median time from SOT to diagnosis was 23 months (-206 to 294), in 8 patients the diagnosis was known before SOT. Patients underwent 31 alloSCT (3 double alloSCT after SOT). The median year of alloSCT was 2006 (1990-2012) and the median time between SOT and alloSCT was 35 months . The conditioning was standard (N=16) or reduced (N=14) and included TBI in 12 cases, but the regimens were heterogeneous. The SCT was performed with either PBSCs (N=20), BM (N=9) or both (N=1) obtained from MRD (N=12), MUD (N=11), mMUD (N=3) or haploidentical (N=5) donors. T cell depletion was performed in 20/29 cases (in vivo N=15, ex vivo N=2, in vivo and ex vivo N=3). The most common GvHD /SOT rejection prophylaxis was CsA and methotrexate (N=12) followed by CsA and MMF (N=5), tacrolimus and MMF (N=5) and tacrolimus with methotrexate (N=1), but frequently included other immuno suppressive drugs, especially steroids. Results: Hematopoietic engraftment was achieved after 25/30 transplantations including 2 cases where the graft was lost. In evaluable patients, time to neutrophil recovery (>0.5 G/L) was 13.5 days (9-37, N=22) , in 3 patients the platelet count never fell below 20 G/L and the rest recovered after median of 17 days (10-103). Acute GvHD was observed in 17/31 cases including 10 with grade II-IV disease. Terminal solid organ transplant failure was observed in 10 of 30 alloSCTs (1/3 heart, 3/13 liver and 6/11 kidney transplants) after median of 6 weeks (0-569) after HSCT (8/10 in early post-HSCT period). Six of 10 failure cases have been described as SOT rejections (4 kidneys, 1 liver, 1 heart). After median of 14 months from alloSCT (0,3-244) 13/28 patients were alive (46%, 8/13 patients after liver, 5/12 after kidney and none after heart transplantation). The estimated probability of survival after HSCT was 75% at 3 months, 60% at 12, 45% at 36, and 40% at 60 months. In one third of patients (5/15), death was directly related to transplanted solid organ failure. Discussion: In summary, a signifi cant proportion of patients enjoy long term survival without graft loss, especially after liver transplantation. Therefore our results suggest that alloSCT is a feasible treatment option for patients after SOT. However, the risk of transplanted solid organ failure is high, including the possibility of solid organ rejection. Introduction: Palifermin, recombinant human keratinocyte growth factor, is used for mucositis prevention in adults following autologous and allogeneic hematopoietic stem cell transplantation (HSCT). It is known that palifermin decreases length of initial hospital stay, mean number of days of total parenteral nutrition and the use of opioids for pain control in oral mucositis in adults. There are no data evaluating palifermin use in children following autologous HSCT. The aim of this study was analysis of effi cacy and safety of palifermin in children and adolescents following autologous HSCT. Materials (or patients) and Methods: Ninety-nine consecutive patients were included into the study. Results of effi cacy of palifermin in 18 patients were compared to data of 81 patients not treated with palifermin. Palifermin was administered on compassionate-use basis. Results: Palifermin decreased the incidence of severe oral mucositis (grade 3-4) by 21% (44% vs 65%), however it did not contribute to the length of oral mucositis and total parenteral nutrition. There were no differences between palifermin and non-palifermin groups in opioid use, incidence of neutropenic fever, severe infection, hematological recovery and gastrointestinal hemorrhage. Introduction: Iron overload is associated with increased morbidity and mortality among stem cell recipients. It has been reported that high pre-transplant ferritin level adversely impacts clinical outcome in transplant setting and is associated with early complication after HSCT. Growing body of evidence however suggest, that serum level of hepcidin, the main regulator of iron homeostasis, correlates better with tissue and plasma iron than serum ferritin. We evaluated the risk of NRM, infections and organ toxicity and acute GVHD within 100 days after HSCT with respect to pretransplant serum ferritin and hepcidin levels and transferrin saturation. Materials (or patients) and Methods: A total of 51 patients (median age 40; range 20-65) were analyzed. Twenty four pts were treated for myeloid malignancies (AML, MDS, CML, MPD) and 27 for lymphoid neoplasia (ALL, NHL, HL, MM). Twenty six pts had advanced disease. Forty pts underwent allogeneic PBSCT (28) or BMT (12) from sibling (17) or unrelated (23) donors. In this group reducedintensity conditioning was applied to 15 pts. Autologous transplant was performed in 11 pts with use of PBSC. Elevated serum ferritin level >1000 ng/mL was found in 30 (59%) pts and >2000 ng/mL in 14 (27%) pts. Hepcidin level <50 ng/mL was seen in 42 (82%) pts. High transferring saturation (>50%) was detected in 25 (49%) pts. Results: Neutrophil recovery occurred in 48 pts at 14 (11-34) days; 3 pts died before hematologic reconstitution due to infection (2) or multiorgan failure (1) . Infection complications up to 100 days occurred in 49 (96%) pts including: FUO (36), bacterial (12) , fungal (9) ; CMV (13), infl uenza (2) and toxoplasmosis (1) infections. Any organ dysfunction was seen in 22 (43%) pts, however severe renal, hepatic, or lung toxicity was diagnosed in 5 (10%) pts. GVHD>2° occurred in 8 (20%) pts from allogeneic group. Seven (14%) pts died up to 100 days due to bacterial infection (1), viral infection (2), toxoplasmosis (1), GVHD (2) or multiorgan failure (1). In univariate analysis we did not fi nd statistically signifi cant infl uence of ferritin level, hepcidin level or transferin saturation on infection risk (including fungal infection), organ toxicity, GVHD and NRM. However, among pts with hepcidin level >50 ng/mL there were no early deaths and the risk for infection was the lowest (ns). A simple model based on ferritin level >1000 ng/mL, hepcidin level <50 ng/mL and transferring saturation >50% was created, which identifi es pts with iron overload and correlates signifi cantly with higher non-relapse mortality (P=0.037, HR=6.0, 95%CI=1.1-32).

Discussion: Iron overload is common complication in HSCT recipients, especially when assessed with serum hepcidin measurement. A simple model based on ferritin, hepcidin levels and transferrin saturation indentifi es patients with iron overload and higher risk for early non-relapse mortality both in autologous and allogeneic settings. Disclosure of Interest: None Declared. Table 1 . Any signifi cant diff erence with respect to sex was not identifi ed. When patients were stratifi ed according to age groups, patients who were older than 10 years old at the time of transplant had a tendency for increased risk of VOD compared to younger than10 (P=0.05). VOD incidence was similar in malign and non-malign diseases (11.5% and 13.4%, respectively), however it was higher in thalassemia major patients (20.6%)(P<0.01). Although there was a decreased tendency for VOD in patients receiving graft from MSD compared to other donor types, it did not reach to statistical signifi cance (P=0.057). With respect to stem cell source, VOD frequency was found 16.3% in BM, 12.6% in PBSC and 13.6% in CB recipients (P=0.637). We lost 12 of 67 VOD patients (17.91%) but only in two cases the causes of deaths were solely attributed to VOD. Discussion: Risk factors identifi ed in our study group are; transplant age older than 10 years old, a diagnosis of thalassemia major and a transplant from non-sibling donors. We also suggest that refractory thrombocytopenia may be another predictive factor for hepatic VOD. Disclosure of Interest: None Declared. Introduction: The endocrine system is one of the most frequent target of complications after autologous (auto) and allogeneic (allo) hematopoietic stem cell transplantation (HSCT). Materials (or patients) and Methods: We evaluated for endocrine abnormalities a retrospective cohort of 100 consecutive patients (median age, 32; range, 20-52) who underwent auto-(n=50) and allo-(n=50) HSCT with a median follow-up of 6 years (range, 1-15) and a disease-free survival of at least 1 year post-HSCT. Primary diseases were acute (n=44) or chronic (n=13) myeloid leukemia, Hodgkin disease (n=17), non-Hodgkin lymphoma (n=12) and multiple myeloma (n=14). Results: All women experienced ovarian insuffi ciency, manifested as secondary amenorrhea associated with hypergonadotrope hypogonadism and reduced volumes of ovaries and uterus. In allo-HSCT patients, serum 17beta-estradiol, delta-4-androstenedione, circulating androgens and dehydroepiandrosterone levels were signifi cantly decreased, especially in women developing cGVHD. In auto-HSCT patients, only serum 17beta-estradiol levels were decreased. Impaired spermatogenesis damage was observed in all transplanted patients. Lower sperm counts were observed in patients aff ected by cGVHD when compared to unaff ected patients. Testosterone was reduced in about 30% of patients up to 1 year after HSCT, particularly during acute and chronic GVHD. The onset of adrenal insuffi ciency (a total of about 20% of cases in the auto-and allo-setting) was always related to the duration (more than 100 days) and cumulative dose (greater than 10 gr/m 2 ) of corticosteroid treatment. Cortico-adrenal failure recovered in all patients after 3-12 months of short acting steroid substitution therapy.

All allo-HSCT recipients conditioned with BU/CY regimen showed growth hormone (GH) levels within the normal range, insulinlike growth factor (IGF)-I levels were lower in 38% of recipients aff ected by cGVHD, whereas IGF-1 resulted in the normal range in only 7% of subjects cGVHD-free. Sub-clinical hypothyroidism was found up to 5 years after allo-HSCT and the "low T3 syndrome" after 12-48 months, especially in patients with extensive cGVHD. In the auto-HSCT setting, we detected subclinical hypothyroidism in 12% of cases at 12 months and the "low T3 syndrome" in about 30% at 3 months, but none at 12 months. The incidence of hypothyroidism was higher in patients previously treated with neck/thoracic radiotherapy than in untreated patients (50% vs 1.3%, respectively). Bone mass density (BMD) at lumbar spine, femoral neck and phalanges were signifi cantly reduced (Z score mean values: -0.4 vs -0.9, -0.6 vs -1.4 and -1 vs -1.5 in auto-and allo-HSCT recipients, respectively) and GVHD development was associated with a more severe reduction in all bone sites. Eight patients (6 allo-and 2 auto-HSCT) developed avascular necrosis, 1 to 15 years (median, 28) following HSCT. Discussion: The underlying diseases, pre-transplant therapies, the age at HSCT, total body irradiation (TBI)-and high-dose chemotherapy-based conditioning regimens were the main risk factors of endocrine disorders after auto-and allo-HSCT. Our analysis further provide evidence that auto and allo-HSCT recipients show higher incidence of endocrine disorders suggesting that their early identifi cation may greatly improve the quality of life of longterm survivors after HSCT. Disclosure of Interest: None Declared.

Introduction: Nutrition plays an essential role in the processes of maintaining health. And the malnutrition is common in patients waiting for Hematopoietic precursor cell transplantation and represents a risk factor for post-transplant morbidity. Patients at any stage of the transplantation process are at high nutritional risk and should undergo careful nutritional assessment for the early identification of nutritional support requirements. The parenteral nutrition (PN) is treatment modality remains an artificial feeding technique that can give rise to numerous complications of varying severity. Therefore, appropriate selection of those patients who will derive a true benefit from PN is essential. Materials (or patients) and Methods: Prior to November of 2011, in our centre, we have administered through stipulated protocols Parenteral Nutrition from day +1 regardless of the individual characteristics of each patient. Since then, we have not used this universal strategy, we have administered only enteral nutritional supplements. 108 consecutive patients under ASCT have been performed in our center, 64 patients have received NP (from Jan 2008 to Oct 2011), and 46 belong to Non-NP group (from Nov 2011 to Nov 2013). The characteristics of both groups are shown in Table 1 . There are no signifi cant diff erences between both groups according to gender, age, diagnosis and mucositis incidence. We used baseline Serum Albumin and Total Proteins (TP) as parameters to assess the nutritional status of patients. Subsequently we have studied these values 5 and 10 days post-transplant to evaluate the impact of use parenteral nutrition. Results: For statistical analysis we used the test of T student. We found no signifi cant diff erences in baseline values of both determinations, so both groups are comparable. There are no diff erences (P>0.05) between the values of albumin and TP on day +5 and day +10. The values are listed in Table 1 . Discussion: 1.There are only a few prospective, randomized, controlled trials that have investigated the role of nutritional support in organ transplantation. 2.Total proteins and Albumin is a very good index of the status of the hepatic synthesis, but has a half life as very long (21 days) it takes to change with disorder and nutritional therapy recovered, hence, prealbumin determining to have a life shorter half (2 days) is much more eff ective to assess acute malnutrition and response to treatment. 3.The individualized prescription of parenteral nutrition, according to current recommendations established by scientifi c societies, helps to improve the safety and eff ectiveness of this nutritional therapy. Disclosure of Interest: None Declared. Introduction: Fludarabine has been used in combination with standard doses of intravenous injection of busulfan, thus reducing the toxicity previously observed with cyclophosphamide/ busulfan regimens. However, some studies have documented that fl udarabine based regimen was associated with an increased risk of relapse, especially in patients with active disease at the time of transplantation. However, it is unclear whether the additional use of anti-thymocyte globulin (ATG) to decrease the risk of acute graft versus host disease (GVHD) leads to improved clinical outcomes in this condition. Materials (or patients) and Methods: We evaluated the impact of anti-thymocyte globulin (ATG), to decrease the incidence of acute graft versus host disease, on the clinical outcome including transplant related mortality and relapse rate in patients received myeloablative conditioning based on fl udarabine and busulfan. Donors were fully HLA matched (6 out of 6 match for sibling donors). The dose of ATG was 4.0 ~ 6.0 mg/kg body weight for 2 or 3 consecutive days.

Results: Totally 66 patients undergoing allogeneic hematopoietic stem cell transplantation using peripheral blood stem cells were evaluated. The median follow-up time was 36 months. The incidence of moderate to severe acute GVHD (≥ grade 2) was not signifi cantly changed by the addition of ATG (ATG group; 27.8% versus non-ATG group; 29.2%). Transplant related mortality seemed to be lower in cases treated with ATG. The incidence of relapse was signifi cantly higher in patients received ATG (ATG group; 38.9% versus non-ATG group; 18.8%). 5 year disease free survival seemed to be lower in ATG group and 5 year overall survival was not signifi cantly diff erent between two groups. Discussion: Conclusively, adding ATG to fl udarabine based conditioning may not guarantee signifi cant benefi t of clinical outcomes in a setting of transplantation from a sibling donor. Therefore, a careful consideration of the use of ATG based on the patient's condition and the risk factors of the transplantation setting should be made. Disclosure of Interest: None Declared. Introduction: Chronic graft-versus-host disease (c-GVHD) is an immune-mediated disorder that occurs frequently after allogeneic hematopoietic cell transplantation (HCT). C-GVHD most often involves skin but lung can be involved and it is recognized as the major risk factor for reduced lung function. Bronchiolitis obliterans (BO) is a severe pulmonary manifestation characterized by a nonspecifi c infl ammatory injury and is strongly associated with c-GVHD, suggesting that BO is a pulmonary manifestation of c-GVHD. BO initially aff ects terminal and respiratory bronchioles, a region largely unexplored by spirometry, which is only altered in advanced disease. In contrast, the Impulse Oscillation System (IOS) and the nitrogen multiple breath washout (N2-MBW) are techniques characterized by a high sensitivity to peripheral airway changes and potentially more suited to early detection of small airways disease. Materials (or patients) and Methods: In a cross sectional study, a total of 161 patients (pts), divided into 4 groups: healthy controls (41), HCT candidates (47), HCT recipients (65) and pts with chronic obstructive pulmonary disease (COPD; n=8), were assessed by IOS, N2-MBW, spirometry, body plethysmography and diff using capacity for carbon monoxide (DLCO) in order to describe respiratory function changes in post-HCT pts without pulmonary GvHD and in order to characterize the pattern of peripheral airway changes in BO. Results: All subjects were able to perform IOS and N2-MBW without diffi culty. HCT, even without respiratory complications, does not aff ect spirometry but appears to cause an increase in air trapping, a reduction in DLCO and enhanced ventilation inhomogeneity both in conductive (Scond*VT) and acinar (Sacin*VT) airways.

In the cohort of transplanted pts 33 were diagnosed with c-GvHD, 15/33 categorized as severe and 8/15 presented lung involvement. Immune-reconstitution analysis show no diff erence across the 3 sub-populations. Pts with lung GvHD were characterized by further DLCO reduction, increase in oscillometric indices sensible to peripheral airways involvement and a further threefold increase in Sacin*VT. Compared to patients with BO, COPD patients with the same degree of spirometric obstruction (FEV1/ FVC<0.7, FEV1 50% predicted) showed only half the increase in predicted Sacin*VT (P0.03).

Discussion: At a cut off of 321% of predicted value, Sacin*VT could distinguish the subjects with BO from pts without lung impairment post HCT with good accuracy (87%), sensibility (87.5%) and specifi city (89.5%). IOS and N2-MBW are simple tests able to detect changes following HCT as well as those specifi c to BO. Further exploitation of this technique could provide improvement in diagnosis of lung GvHD in order to start appropriate therapy promptly, minimize symptoms and prevent irreversible organ damage. Disclosure of Interest: None Declared. Introduction: ICU clinicians are often reluctant to admit haematological patients due to the assumption of poor prognoses, as the survival of this patients has improved in recent years due to the advances in chemotherapy regimens and in order to achieve better survival rates, we consider crucial to identify prognostic factors which may predict their outcomes during their Intensive Care Unit (ICU) admission. Materials (or patients) and Methods: Evaluate the prognostic factors and long-term survival of patients with haematological malignancies who were transferred to ICU due to a lifethreatening complication during the first 100 days of autoHSCT. We performed a retrospective analysis between 2002 and 2012. Results: From a total of 144 patients, we have registered 11 (7.5%) transferred to the ICU in 13 episodes. Six of them were male and 5 female, the average age at the diagnostic was 58 years (25-66) and the average days to admission were 8 days (5-100) from the infusion. The haematological underlying diseases were: 4 NHL, 4 MM, 1 AML, 1 HL and 1 AL. Those patients are the 7% of lymphoma, 8% of myeloma, 6% of leukemia, 6% of Hodgkin's lymphoma and 50% of amyloidosis of the global of autoHSCT performed during this period. Intensive therapy regimens consisted of 5 patients with melfalan, 5 with BEAM and 1 with BuCy. Most of them in response to their haematological disease at the time of infusion (91%; 6CR, 4PR). Just one patient had received multiple transfusions. The most common reason precipitating the ICU transfer were septic status, in 38% (5/13) of patients due to respiratory and in the same number of cases due to central venous catheter sepsis. We obtained microbiological isolates in 8 cases (61.5%). Seven of them were bacterial (50% GNB, 50% GPB) and one viral. The average rate of neutrophils at the admission was 0x10 9 /L (0-8.2). Analitic data associated to infection were reviewed, aiming high levels of C-Reactive Protein (CRP) in 100% cases (12/12) of available data (12/13) and procalcitonin in 86% cases (6/7) of available data (7/13), with an average APACHE score of 15.5 points (5-23). On the other hand, we obtained a low acute renal failure rate among our patients (just 3 episodes, 23%). During the fi rst 24h of admission, in 4 occasions (30%) orotracheal intubation was needed, just one patient required non-invasive ventilation, in 8 patients (61%) vasoactives drugs were started and just in one patient continuous venovenous hemodiafi ltration was performed. The median ICU stay was 11 days . 5/11 patients died in the ICU (45,5%) with a median of 36 days (9-84) since de admission day. The death cause in 2/5 (40%) of patients was hemorrhagic complication and in 3/5 (60%) it was secondary to septic process. Discussion: Our ICU admission rate is low with a mortality intra-ICU rate of 45.5%, both of them similar to the ones reported on literature. Neither the type of haematological malignancy or the status of the disease seem to be related to a higher rate of transfer to ICU, although our limited number of patients make it not completely assessable. Our patients have income gravity data and medium APACHE scores, 30% of episodes are requiring orotracheal intubation and in more than half vasoactive drugs within the fi rst 24h of admittance. Severe neutropenia and high levels of procalcitonin and CRP should be considered as heavy variables when assessing the severity of the infection. Disclosure of Interest: None Declared. Introduction: The aim of this study is to compare the safety of two administration schedules of intravenous busulphan (4-times-daily versus once-daily dosing) as part of the conditioning therapy for HSCT. Materials (or patients) and Methods: We studied all the patients who received intravenous busulphan four times a day (BU-4 group) or once a day (3 hours infusion) (BU-1 group) during the conditioning therapy for HSCT at our Center during the last six years. In total, 66 patients received intravenous busulphan during the period of the study. Acute leukemias or myelodysplastic syndromes were the most common indications for HSCT (83.3% of the patients). Forty patients underwent allo-HSCT (half of them from unrelated donor) and 26 auto-HSCT. The SC source was PBSC in 47 cases, BM in 17, and UCT in 2. The distribution of patients ) . Results: The median hospitalization stay was 30 days (15-67), with no diff erences depending on the busulphan pattern of administration. Recovery of neutrophils and platelets counts varied depending on the conditioning regimen employed, but not on the two diff erent schedules of busulphan. The mean requirement of platelet and packed red cells (PRBC) transfusion was also similar in both groups (PRBC: 5,54 in BU-4 group vs 5,04 in BU-1 group; Platelet concentrates: 9,8 in BU-4 group vs 8 in BU-1 group). Sinusoidal occlusive syndrome (SOS) was developed in 4 cases (6.1%), 3 of them in patients who received BuCy conditioning regimen. Although there were no signifi cant diff erences in the incidence of SOS between the two groups, there was a trend to be superior in Bu-4. There were no diff erences in transplant mortality between the two group of patients (see table) . Discussion: This study shows that safety of once-daily administration of intravenous busulphan is at least similar to traditional 4-times-daily. Considering the advantages for the hospital Pharmacy and for the nurses of once a day schedule, it might be considered the method of choice for the administration of intravenous busulphan. Disclosure of Interest: None Declared. Introduction: The impact of smoking on health has been extensively studied in epidemiology. This retrospective monocentric study aimed at understanding whether smoking aff ects the survival outcomes of hematologic patients undergoing reduced intensity allogeneic stem cell transplantation (RIC alloSCT). Materials (or patients) and Methods: We searched for information on the smoking history (packs-year and duration of exposure) of 272 patients who received RIC alloSCT between 2001 and 2011 in our division, reviewing the clinical charts and by telephone interviews. We analyzed the impact of smoking and hard smoking (defi ned as smoking >=1 pack per day) on non-relapse mortality (NRM) by Cumulative Incidence method, on overall survival (OS) and progression free survival (PFS) by log-rank method. We ran a Cox multivariate analysis for NRM, OS and PFS using pre-transplant disease status, donor type, and smoking as covariates. Results: We had complete data of 172 patients. All the patients received RIC alloSCT for lymphoma (58%), multiple myeloma (26%), or acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS) (16%). The median age at alloSCT was 49 years (range, 18-66). Donor were HLA-identical siblings (45%), mismatched siblings (3%), unrelated (41%), or haploidentical (11%). At alloSCT, 54% of patients were in complete response (CR), 33% in partial response, 3% of patients were stable and 10% were in progression. Forty percent of patients were former smokers (21% hard smokers), 8% were smokers at admission (5% hard smokers). The exposure to smoking was a median of 1 pack per day (range, 0.1-3) for a median of 16 years (range, 1-40), and smokers had quit smoking by a median interval of 5 years before transplant (range, 0-40). The median follow-up after alloSCT was 6.3 years (range, 0. 4-11.4) . The NRM at 100 days was 5%, at 1 year 9%, at 5 years 12%. Fiveyears OS and PFS were 76% and 49%. In univariate analysis, smoking and hard smoking increased NRM (P=0.017 and P=0.020), and reduced the OS (P=0.035 and P=0.003, respectively). The multivariate analysis showed that smoking increased the NRM by a factor of 3.6 (P=0.009), and signifi cantly impacted OS increasing the risk of death by a factor of 1.8 (P=0.044). Overall survival was also impacted by alternative donor (P=0.018) and pre-transplant disease not in CR (P<0.001). Hard smokers had a more striking reduction of OS, with an increased risk of death by a factor of 2.8 (P=0.002) and an increased risk of NRM by a factor of 3.5 (P=0.016).

Smoking or hard smoking did not aff ect PFS, which was impacted by disease status (P<0.001). Six patients had a second cancer after alloSCT (3%), a patient was a non-smoker (breast carcinoma), 5 patients were smokers or former smokers (AML, MDS, bladder carcinoma, tongue carcinoma and squamous-cell skin carcinoma). The crude incidence of second cancer was 1% for non-smokers and 6% for smokers (trend, P=0.080). Introduction: Extramedullary relapse (EMR) of Acute Leukemia (AL) following allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a contributor to post-transplant mortality and remains a poorly understood, especially the diff erent characteristics of EMR between acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) patients. In order to investigate the incidence, risk factor and clinical outcome of EMR for AML and ALL, we performed a retrospective analysis 362 patients with AL who underwent allo-HSCT at the First affi liated Hospital of Soochow University from January 2001 through March 2012. Materials (or patients) and Methods: We retrospectively studied 362 patients with AL including including 208 AML patients, 147 ALL patients and 7 hybrid acute leukemia (HAL) patients who underwent allo-HSCT at our center during these 10 years.

Results: 1.The 10-year cumulative incidence of overall relapse of AL was 27.0%, EMR was 7.9%. Compared with AML, ALL patients had a higher incidence of EMR (12.9% vs 4.6%, P=0.009). 2. EMR sites of AL post-HSCT included central nervous system (CNS) (n=18), testis (n=5), skin (n=2), soft tissue (n=2), bone (n=2), lymph nodes (n=1), nasopharynx (n=1), and the peritoneum (n=1). Five of 26 (19.2%) patients presented with EMR in multiple sites. The most common type of EMR in ALL was CNSL. 3. Multivariate analysis results showed that the risk factors of EMR for AML patients included: advanced disease status at HSCT, hyperleukocytosis at diagnosis, a history of extramedullary (EM) leukemia before HSCT and conditioning regimen (total body irradiation [TBI]-based).

While the risk factors of EMR for ALL patients included hyperleukocytosis at diagnosis, adverse cytogenetics, and peripheral blood stem cell (PBSC) as stem cell source. 4. The prognosis of EMR of AL was poor. The 3-year overall survival (OS) of isolated EMR and EMR with concurrent BMR was 18.2% and 8.0% respectively. EMR of ALL is associated with better survival than that of AML. Discussion: Compared with AML, ALL patients had a higher cumulative incidence of EMR post transplantation. PBSC as stem cell source was a independent risk factor for EMR in ALL patients but not in AML patients, this maybe suggested that the GVL eff ect was less eff ective in EM sites of ALL patients. Disclosure of Interest: None Declared. Introduction: Recurrence is a major cause of treatment failure after allogeneic hematopoietic stem cell transplantation (allo-HSCT) for acute leukemia, and subsequential treatment options are very limited. We evaluate the effi cacy and toxicity of cytarabine and aclarubicin combined with granulocyte colony-stimulating factor priming (CAG regimen), consisting of concurrent use of granulocyte colony-stimulating factor (G-CSF) with low-dose cytarabine and aclarubicin, as a salvage therapy for acute leukemia patients who relapsed after allo-HSCT. Materials (or patients) and Methods: Fifty-nine patients (32 male and 22 female) with acute leukemia, with a median age of 27 years, relapsed post allo-HSCT and received salvage chemotherapy. Twenty seven patients received CAG regimen while 32 patients received non-CAG regimen such as intensive chemotherapy. Results: The overall response rate (ORR) of CAG and non-CAG groups were signifi cant diff erent (55.6% vs 28.1%, P = 0.033). With regard to disease type, comparing with non-CAG group, ORR of AML patients in CAG group was signifi cantly higher (64.3% vs 26.7%, P = 0.025). However, ORR of acute lymphocytic leukemia (ALL) in CAG group was similar as that in non-CAG group (50% and 41.2%, respectively; P = 0.471). Median overall survival (OS) from the starting of CAG chemotherapy and 2-year OS of CAG group were 9 (1-27) months and 16.1%. Meanwhile, median OS and 2-year survival of non-CAG group were 4 (1-49) months and 8.8%. Moreover, the median duration of neutropenia and thrombocytopenia of CAG group were signifi cantly shorter than that of non CAG group, 6 (1-12) vs 11 (5-28)days (P=0.000) and 8 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) vs 14 (7-35) days (P=0.000). For the patients who received donor lymphocyte infusion (DLI[JW1] ) as a subsequential therapy, two-year OS of CAG and non-CAG group were 17.2% and 12.5%, respectively (p[JW2] =0.577). Treatment related mortality (TRM) was seen in 2 cases in CAG group compared with 10 cases in non-CAG group. For CAG group, impact on overall response rate was signifi cantly associated with leukocyte level, and medullar blast percentage[JW3] at relapse (P = 0.005 and P = 0.000 respectively).

Furthermore, multivariate analysis showed response to chemotherapy was the only factor that correlated with better survival (P=0.032, HR 0.461, 95% CI (0.227, 0.937)). Discussion: CAG regimen as a salvage chemotherapy for relapsed acute leukemia post allo-HSCT could eff ectively reduce tumor burden with mild toxicity, especially for hypoplastic acute leukemia patients. For certain relapsed acute leukemia patients post allo-HSCT, CAG regimen may be an optimal choice as the bridge therapy followed by DLI or second stem cell transplantation (SCT Introduction: Organ transplantation is an increasingly used medical procedure for treating otherwise fatal end stage organ diseases with 107,000 transplants performed worldwide in 2010.

Post-transplant lymphoproliferative disorders (PTLDs) are serious, life-threatening complications of solid-organ transplantation (SOT) and bone marrow transplantation leading to a high mortality (30-60%). The incidence of PTLD, ranging from 1% to 20%, clearly relates to the type of transplanted organ, intensity of immunosuppression (IS), underlying disease, age, viral infections including EBV, cytomegalovirus and hepatitis C virus (HCV). Materials (or patients) and Methods: A retrospective study of 48 patients diagnosed with PTLDs from 1998 to 2013, in our center. We analyzed the incidence, clinical features, and outcome in a 48 patients diagnosed with PTLDs in our unit between 1998 to 2013. The data analysis was performed with the SPSS 17.1 program. Results: We analyzed 1340 transplants procedures performed in our Center during the study period. 48 patients (3.5%) developed PTLDs. The highest incidence 11% was reported in lung transplantation. A total of 48 subjects were analyzed in the study, 69% males, 31% females, 31% children. 27.1% (n=13) received pulmonary transplant, 20.8% (n=10) renal, 20.8% (n=10) hepatic, 16.7% (n=8) allogeneic bone marrow transplantation, 10.4% (n=5) heart, 2.1% (n=1) renal-heart and 2.1 (n=1) lever-renal. The median age at transplant was 36 years . The mean time between diagnosis and transplant was 34 months . 27% (n 13) developed PTLD early (<1 year). Histological classifi cation and immunophenotype: 92% were CD20+ and 63% EBER+. The main subtype was monomorphic PTLDs (M-PTLD) (58%) and the most common was DLBCL (47%). 80% had EBV active infection (demonstrated by PCR) after transplantion. Calcineurin inhibitor (tacrolimus n= 34, 71%) was the main immunosuppression regimen used, either as monotherapy or with corticosteroids (50%, n = 24). After the diagnosis of PTLD 84% of patients reduced immunosuppression, and 31% (n=15) presented signs of transplanted organ. 69% of patients received specifi c PTLD treatment. Several therapeutic approaches were currently used. 50% received Rituximab, 27% chemotherapy with Rituximab, 16% surgery and 7% other treatment. Median follow-up was 18 months (0-117meses). 1-year and 2-year was 56% and 50 ± 8% respectively. The response rate to standard therapy was 65%. The immunosuppression was discontinued in 29 patients and the survival was higher in this group (P = 0.008). The presence of M-PTLD and late-onset was associated with a worse prognostic, although it was not statistically signifi cant (P = 0.56) (ns). We have not found any association between OS and the time to relapse, immunosuppression regimen, staging, presence of EBER+ or therapy administered. The use of Rituximab was associated with a higher OS, without statistical signifi cance (ns). The results are shown in the following table. Discussion: In our study we didn´t fi nd signifi cant diff erence in OS between histological factors, staging, EBER + and treatment regimen used. The presence of M-PTLD and late-onset was associated with a worse prognostic and the use of Rituximab was associated with a higher OS, in both cases without statistical signifi cance (ns). Discontinue immunosuppression was associated with a better overall survival (P=0.008). Disclosure of Interest: None Declared. Introduction: Fortunately, the number of long-term survivors after allogenic stem cell transplant (allo-SCT) is growing. Thus, attention increasingly focuses on late complications such as avascular necrosis (AVN). AVN aff ects negatively quality of life and requires hip replacement when evolves. The main factor risk described is post-SCT long-lasting high-dose steroids. Materials (or patients) and Methods: We reviewed the number of patients diagnosed of symptomatic AVN after allo-SCT in our centre between 2008 to 2013. Disease, conditioning regimen, type of allogeneic transplant, acute and chronic graft versus host disease (GVHD) and time and dose steroids were analyzed. In cases of GVHD, the initial dose prednisone was 1.5 mg/kg/24h. AVN was diagnosed by nuclear magnetic resonance imaging (MRI) following had joint pain complaints. Patients were treated with calcium, vitamin D, zoledronic acid and were assessed by the Traumatology Service for autologous bone marrow mononuclear cells therapy indication. Results: 81 medical records were reviewed. The number of patients with symptomatic AVN was 5 (6.17%). All of them had acute GVHD which required treatment with high doses of corticosteroids. Discussion: All patients with symptomatic AVN required post-SCT long-lasting high-dose steroids, fact that supports other publications that consider steroids the main risk factor. However, the incidence in our center has been lower than in other works. Despite cell therapy is not the treatment of choice of AVN, in several published series and in our experience this treatment prevents the collapse of the femoral head in the early stages of AVN and avoids the need for inserting a prosthesis. Disclosure of Interest: None Declared. Introduction: The eff ects of certain risk factors on the survival of adults undergoing allogeneic hematopoietic stem cell transplantation (HSCT) have been the subject of research for many years. The impact of graft source, donor type, and iron parameters like ferritin has already been examined closely. However, observations of pediatric populations considering those factors remain rare and no score in this regard for children with HSCT is available yet. Materials (or patients) and Methods: We retrospectively analyzed the eff ects of patient age, patient sex, disease risk, donor age, recipient-donor sex match status, donor HLA match, graft source as well as ferritin, albumin, total bilirubin, C-reactive protein (CRP), aspartate transaminase (AST), alanine transaminase (ALT), gamma glutamyl transpeptidase (GGT), cholinesterase (CHE), and lactate dehydrogenase (LDH) taken at the time of transplantation on the 5-year-overall survival of 132 children undergoing allogeneic HSCT between 2001 and 2011 in a single center. The graft source was either bone marrow (n=82) or peripheral blood stem cells (n=50). The patients had the following underlying diseases: ALL (n=44), AML (n=29), CML (n=5), myelodysplastic syndrome (n=16), non-Hodgkin lymphoma (n=7), solid tumor (n=4), severe aplastic anemia (n=7), myelofi brosis (n=2) and genetic disease (n=18). Conditioning regimen was myeloablative in all cases. The disease risk was formed by dividing the patients into two groups according to their clinical risk. Patients with genetic disease, severe aplastic anemia, refractory cytopenia, myelofi brosis, leukemia and lymphoma in fi rst or second remission as well as chronic myeloid leukemia in chronic phase were low risk, while patients with solid tumor, advanced myelodysplastic syndrome as well as leukemia and lymphoma in more than second remission or in relapse were high risk. For statistics we used Kaplan-Meier-method for univariate analysis and Cox regression for multivariate analysis. Results: In univariate analysis, 5-year-overall survival decreased signifi cantly in patients with high disease risk (38.3% versus 74.7%, P<0.001), peripheral blood stem cells as graft source (47.1% versus 72.2% for bone marrow, P=0.001) as well as in patients with ferritin >1500 μg/L (40.8% versus 78.8%, P<0.001), CRP >10 mg/L (54.6% versus 69.4%, P=0.017), LDH >6 μmol/L·s (22.2% versus 66.8%, P=0.001), GGT >1 μmol/L·s (43.2% versus 67.9%, P=0.032) and CHE <60 μmol/L·s (35.7% versus 70%, P=0.002). Other factors did not show a signifi cant correlation. We subsequently developed a score of those parameters that were signifi cant in multivariate analysis, i.e. disease risk (HR=3.744, P=0.035), ferritin (HR=6.860, P=0.002) and CHE (HR=4.556, P=0.043), dividing the patient population into three groups: low with no risk factor, intermediate with one risk factor and high with two or three risk factors. For this score we found a 5-year-overall survival of 92.3% for the low risk group, 66.2% for the intermediate risk group and 17.4% for high risk group (P<0.001). Discussion: Our data show that disease risk, ferritin, and CHE are factors that decisively infl uence the prognosis after allogeneic HSCT in children. They should be evaluated in further trials as well as our proposed risk score. The characteristics that showed up signifi cant in univariate but not in multivariate analysis appear to have an infl uence as well and might show a stronger correlation in larger trials. Disclosure of Interest: None Declared. Introduction: It is important to prevent the formation of free radicals formed as consequence of infl ammatory processes, directly or indirectly, in patients undergoing stem cell transplantation. Glutation-S-Transpherase (GST) and Cytochrome P-450 (CYT P-450) enzymes are detoxifi cation enzymes that play a role in and are responsible for reduction of glutation by conjugating reactive chemicals. In this way, in the future it seems to be possible to make a prediction in terms of preventing possible complications by evaluation of gene polymorphism. We aimed to study the relationship between GST and CYT P-450 gene polymorphisms, and sinusoidal obstruction syndrome (SOS), mucositis, liver toxicity, febrile neutropenia, graft versus host disease (GVHD) and transplant related mortality, which are tightly related to the clinical course after HLA matched allogeneic bone marrow transplantation. Materials (or patients) and Methods: We included 62 allogeneic stem cell transplanted patients and their healthy donors as controls group in our study. Polymerase chain reactions were used to detect GSTM1 and GSTT1 polymorphisms, whereas PCR-RFLP (restriction fragment length polymorphism), for GSTP1 polymorphism and CYP1A2(C734A)(intron 1) and CYP2E1(5'-). All major transplant related complications were recorded. Results: Compared to GSTM1(-) polymorphisim, patients with GSTM1(+) polymorphmisim were mostly in remission at transplantion ( P= 0,001). Among early complications such as SOS, mucositis, liver toxicity, febrile neutropenia, GVHD there was no signifi cant diff erence between GSTM1, GSTT1 null genotype, GSTP1 and CYP1A2, CYP2E1 mutant allel (P>0.005). There was a trend to increase for SOS in patients with GSTT1 (-) genotype and CYP12A C which was not found statistically signifi cant. In patients with GSTM1 (+) genotype severe GVHD ( ≥ 2 grade) seemed to be lower but statistically not signifi cant. In addition, a numerical but not statistical signifi cant association was between graft failure and GSTM1 (-) GSTT1 (-) genotype. On the other hand, patients with GSTP1 GG genotype seemed to experience more severe GVHD compared to the patients with GSTP1 AA genotype which could not documented as statistically signifi cant. Similar status was for patients with CYP1A2 C allel compared to CYP1A2 AA wild genotype. Discussion: GSTA1, GSTP1, GSTM1 and CYT P-450 genotyping prior to allogeneic hematopoetic stem cell transplantation may allow better prediction of the outcome and the need for intervention as thereby improving clinical outcome. Disclosure of Interest: None Declared. Introduction: Following stem cell transplantation (SCT) patients are at increased risk of accumulating cardiovascular risk factors (CVRF) and developing a metabolic syndrome compared to the normal population. These parameters are associated with an increase risk of cardiovascular death. In the general population there is an association between the metabolic syndrome and obesity and the latter has also been implicated as a risk factor for cardiovascular complications after transplantation. In the UK the prevalence of obesity is relatively high with 2010 data from NHS National Statistics classifying 26% of adults over the age of 16 as obese. It is our impression, however, that the majority of our patients post SCT are non-obese despite their tendency to accumulate CVRF. Materials (or patients) and Methods: In this retrospective study we have reviewed case note data from all patients attending a monthly dedicated clinic for long term (>10 years) survivors of SCT over the time course of one year (Jan-Dec 2013). Data was collected on the frequency of dyslipidaemia, hypertension, diabetes or overt ischaemic cardiac disease. In addition indicators of obesity were collected including body mass index (BMI), and waist-hip ratios. Obesity was indicated by BMI>30 kg/m 2 and overweight defi ned as a BMI of 25-29 kg/m 2 . A waist-hip ratio >1 was considered abnormal. Results: 53 patients (29 male) were evaluated a median of 21 years (range 10.9-34y) post SCT. The sources of stem cells were sibling (n=37), unrelated (n=12), haplo-identical (n=1) and syngeneic (n=3). Underlying disease was as follows: CML n=42, AML n=5. ALL n=3, AA n=3. One patient had overt ischaemic cardiac disease and a further 30/53 patients had at least one cardiovascular risk factor (CVRF). Fifteen had one CVRF (11 dyslipidaemia, 4 hypertension), ten had two CVRF (8 dyslipidaemia plus hypertension) and fi ve had 3 CVRF (dyslipidaemia, hypertension and diabetes). Data on BMI was available for 45/53 (85%) with 5/45 (11%) defi ned as obese and 14 (31%) overweight. These data are lower than for the normal population. 39/53 (74%) had waist/hip ratios measurements. and7/39 (18%) had abdominal obesity. The relationship between these parameters and cardiovascular risk factors is shown in Table 1 . S278 Discussion: The prevalence of cardiovascular risk factors was high at 53%. Despite this the prevalence of obesity was approximately half that of the general adult population. Data in Table 1 does indicate a relationship between BMI, waist -hip ratios and the number of CVRF. Although the numbers are small the percentage of patients who are obese, overweight or have a raised waist-hip ratio are increased in patients with 3 CVRF compared to those with less. These data suggest that recommendations for ideal weight may be diff erent in the post-transplant population compared to the normal population. Disclosure of Interest: None Declared. (n=528) or autologous (n=1455) and conditioning was with (n=556) or without TBI (n=1427). Donor was sibling (n=302), matched unrelated (n=220) or cord blood (n=6). Source of stem cell was marrow (n=322), PBSC (n=1627), both (n=28) or cord blood (n=6). GVH prophylaxis included Campath in 203 cases. Of all the patients 1774 received single transplant but 209 received more than one transplant. Data was analysed as of 01/12/2013 using competing risk methods with death as competing risk for developing second cancers.

Results: Patients follow-up was more than 10 years in 382 cases (19%), between 5 to 10 years in 328 (17%), 1 to 5 years in 667 (34%) and less than 1 year in 606 cases (31%). Second solid cancers developed in 70 patients with the incidence of 1% at 5yr (95% CI: 0.5-1.6), 2.2% at 10 yr (95% CI: 1.6-3.3),4.8% at 15yr (95%CI: 3.6-6.8) and 8% (95% CI: 5.9-10.5)at 20 years. Site of second malignancy was brain (n=2), breast (n=15), cervix (n=3), GIT (n=11), genitourinary (n=9), lung (n=3), skin (n=17), head & neck (n=7), thyroid (n=3), non EBV related lymphoma (n=3). In univariate analysis 10 yr. probability of developing SMN was not infl uenced by gender, stage of disease, primary diagnosis, type of HSCT, use of TBI, type of donor or year of transplant. It was signifi cantly higher with use of PBSC (1.4% vs. 2.6%, P=0.02) and age above 65yr. (1.5% vs. 11%, P=0.001). In multi-variate analysis age above 65yr. (RR: 1.8, 95% CI: 1.1-2.9, P=0.02) and PBSC (RR:9.4, 95% CI:1-99, P=0.05) were independently associated with increased risk of SMN. 19 patients have died due to SMN (27%) and signifi cantly shorter with Gi, genitourinary and lung cancers. Analysis will be extended to identify the role of gvhd, TBI dose, fractionation and use of additional radiotherapy on the incidence of organ specifi c second solid malignancies. Discussion: This single centre analysis shows that the risk of developing SMN increases with longer follow-up and the survival is poor. Long term survivors of stem cell transplants need follow-up probably for life in speciality clinics. Early detection, surveillance and advice regarding avoidance of known carcinogens should be encouraged through patient education. Disclosure of Interest: None Declared. Introduction: Acute graft-versus-host disease (aGVHD) remains the major cause of non-relapse mortality (NRM) following allogeneic hematopoietic cell transplantation. The prognosis of patients with steroid-refractory aGVHD is very poor, although several salvage treatments such as mycophenorate mofetil (MMF) and antithymocyte globulin have been tried. It is speculated that a high topical concentration of corticosteroids for gastrointestinal (GI) GVHD may enhance the eff ect and overcome refractoriness caused by down-regulated steroid receptor as occurs in cases of ulcerative colitis. Two prospective studies reported the effi cacy and safety of intra-arterial steroid infusions (IASI) for severe GI GVHD (Shapira et al, 2002; Weintraub et al. 2010) . However, despite promising responses seen in these studies, it is unclear whether or not IASI improves survival. We therefore prospectively examined the effi cacy and safety of IASI, and compared the outcomes with those of historical controls. Materials (or patients) and Methods: We assessed consecutive patients with hematological disorders who developed systemic steroid-therapy refractory acute GI GVHD and who had signed written informed consent at our institution between 2008 and 2012. Patients with GVHD involving multiple organs were excluded. The control group consisted of 14 consecutive patients between 2001 and 2008 who had received second-line treatment including increased dose of steroids, MMF and infl iximab. The primary endpoint was set as treatment response rate of aGVHD at day 28. Enrolled patients were treated with infusions of 2 mg/kg methylprednisolone into the superior and inferior mesenteric arteries for lower GI GVHD, and/or 1 mg/kg methylprednisolone into gastroduodenal and left gastric arteries for upper GI GVHD. Results: A total of 19 transplant subjects aged 31-67 years (median 52) were enrolled. Seventeen patients received myeloablative conditioning. Donor sources consisted of HLA-matched related peripheral blood (rPB) (n=2), haploidentical rPB (n=6), unrelated bone marrow (n=8) and cord blood (n=3). Of 18 evaluable subjects, fourteen (78%) showed an overall response at 28 days, twelve (67%) achieved a complete response and two (11%) achieved partial response after a median of one IASI (range 1μ4). With a median follow-up of 23 months (range 7μ56), 1-year NRM was signifi cantly lower and 1-year OS tended to be higher in the study group than in controls (11% versus 50%, Gray's test, P=0.046; 67% versus 36%, Log-rank test, P=0.106, respectively). There were no serious complications related to IASI. Discussion: Consistent with previous reports, promising responses were also observed in this study. Of note, our data suggests the possibility that IASI can improve OS via a decrease of NRM in patients with systemic steroid-therapy refractory acute GI GVHD. The favorable responses to GVHD without worsening infectionrelated complications might contribute to a decrease of NRM since IASI might enhance response to GVHD and prevent untoward eff ects related to systemic immunosuppressive treatment.

In the future, a larger adequately powered prospective study will be required to validate our results. Introduction: Galectin-1 (Gal-1), a glycan-binding protein belonging to the growing family of animal lectins has an important role in immune cell activation, diff erentiation and homeostasis. Multiple experimental models have shown the role of Gal-1 in suppressing chronic infl ammation and autoimmunity. Gal-1 therapy demonstrated immunoregulatory properties in a murine model of graft-versus-host disease (GvHD). However, little is known about the possible impact of Gal-1 on outcome and occurrence of graft-versus-host disease in humans receiving non-myeloablative haemopoietic stem cell transplantation (NM-HSCT). Materials (or patients) and Methods: The purpose of this study was to investigate the impact of Gal-1 serum levels measured before NM-HSCT on i) incidence and severity grade of acute and chronic GvHD, ii) progression free (PFS) and overall survival (OS), iii) transplant related mortality (TRM) in patients with diff erent haematologic malignancies (HM). Fifty-eight patients with haematologic malignancies treated with NM-HSCT between Mar 2009 and Dec 2012, were included. Median age was 57 (range 17-71). There were 37 males (64%) and 21 females (36%). A total of 35 patients (60%) were treated for myeloid and 23 (40%) for lymphoid malignancies. All patients were conditioned with fl udarabine 90 mg/m 2 and total body irradiation (TBI) 2-4 Gy. Postgrafting immunosuppression consisted of calcineurin inhibitor and mycophenolate mofetil. Seventeen patients (30.3%) had HLA identical sibling donors, 35 (60.3%) matched unrelated donors (10/10 match) and 6 (10.3%) mismatched unrelated donors (9/10 match). Peripheral blood haematopoietic stem cells mobilised with granulocyte-colony stimulating factor (G-CSF) were used as stem cell source. The median follow-up was 428 days (range 73-1238). Acute GvHD (aGvHD) was defi ned as a complex of symptoms occurring within 100 days after transplantation and graded according to international consensus criteria. GvHD developing after day 100 was considered as chronic GvHD (cGvHD) and graded as (i) none/not needing systemic therapy (NNST) or (ii) needing systemic therapy (NST). Serum samples frozen at the time of tissue typing before NM-HSCT were used to perform Gal-1 analysis. Serum galectin-1 levels were assessed according to a standard time-resolved immunofl uorometric assay (TRIFMA) protocol. Results: High levels -defi ned as the 50% above the median levelof Gal-1 correlated with a higher incidence of cGvHD NST. After 8 months the incidence of cGvHD in patients with low vs. high Gal-1 levels was 57.5 % (95 % CI: 38.7-76.3) and 86.1 %, (95 % CI: 71.3-1.00), respectively (P=0.019). High levels of Gal-1 seemed also associated with improved OS, HR 0.40 (0.15-1.00), P=0.051. The 1-year TRM was lower in patients with high vs. low Gal-1 levels, i.e. 5.9 % (95 % CI:-0.7-19.0) vs. 33 % (95 % CI: 14.8-51.3), respectively (P=0.018). In a multivariate model adjusting for the type of donor (unrelated vs. sibling), the association between high levels of Gal-1 and a favourable impact on OS (P=0.003) and PFS (P=0.032) was further enhanced. There were no correlations between Gal-1 levels and aGvHD (P=0.422). Discussion: Gal-1 is a biomarker able to predict (i) incidence of cGvHD, (ii) outcome and (iii) risk of TRM in patients with HM undergoing NM-HSCT. Confi rmation of these fi ndings in an independent data set is warranted. Disclosure of Interest: None Declared. Introduction: Extracorporeal photopheresis (ECP) Software Version 4.1 replaced V3.0 in 2012 for the THERAKOS® CELLEX® Photopheresis System Instrument, with the introduction of variable centrifuge speeds that adjust to the collect fl ow rate, aimed at reducing heat build-up in the centrifuge chamber, and improved temperature monitoring. We compared the buff y coat collections, alarm rates and durations of treatments performed using CELLEX® instruments running V4 software with those that used V3. Materials (or patients) and Methods: ECP was performed on 2 consecutive days for each treatment cycle. 196 V3 treatment cycles (from 54 adults and 10 children) in 3 months were compared with 192 V4 cycles (from 61 adults and 9 children) conducted a year later. Treatments utilising a blood prime (BP) were analysed as a separate group. BPs were used with the majority of the paediatric treatments, and infrequently with 4 adults. Analyses of treatment times were split into single-and double-needle modes. A separate analysis was done of 23 adults who received 1 treatment cycle that included a diff erent software version on Day 1 to Day 2, to enable a direct comparison of V3 and V4. Total white blood cell (WBC) counts and cell diff erentials were recorded from samples taken from the cellular harvest obtained during treatment, and used to calculate the cell dose returned to the patient. Results: Treatments in single-needle mode without a BP took signifi cantly longer using software V4 than V3 (P<0.0001, V3 n=202 treatments, mean=136 min; V4 n=210, mean=146 min), and for double-needle mode treatments with a BP (P<0.0001, V3 n=60, mean=137 min; V4 n=37, mean=158 min). No signifi cant diff erence was found in treatment times when double-needle mode was used without a BP (V3 n=95, mean=115 min; V4 n=96, mean=117 min). No diff erence in treatment duration was found for the 23 adults who received paired V3 and V4 treatments. No signifi cant diff erence was found in WBC, neutrophil, lymphocyte, or monocyte cell doses in any of the groups. The frequency of red cell pump alarms was signifi cantly lower with V4 than with V3, in both the BP group (P<0.0022, 26% vs 58%, V3 n=62, V4 n=38), and the non-BP group (P<0.0004, 7.4% vs 16.4% V3 n=317, V4 n=338). The frequency of system pressure alarms was also signifi cantly lower with V4 than with V3 in the non-BP group (P<0.0001, 1.2% vs 11.6%). No signifi cant diff erences were found between alarm rates in the paired treatment group, but there was a trend for fewer alarms with V4. Interestingly, a signifi cantly higher bag collection volume was observed with V4 than with V3, for the BP, and non-BP, treatments (P<0.0001). Discussion: We found that when using the CELLEX® device with V4 software, the incidence of red cell pump and system pressure alarms was reduced compared to with V3, during the 3 month periods examined. The improved regulation of fl ow rates may account for this, and the variation of the centrifuge speed as a function of the fl ow rate may prevent packing of red blood cells which can cause alarms during treatments with a BP. Treatment durations were longer with V4, which may result from the increased frequency of pauses observed during treatment. The cellular harvest returned to the patient appears to be unaff ected by the software upgrade. Disclosure of Interest: None Declared. Introduction: Treatment of chronic graft-versus-host disease (cGvHD) after allogeneic hematopoietic blood stem cell transplantation (HCT) remains a challenge. Mouse models indicate that adoptive transfer of regulatory T cells (Treg) may suppress GvHD while preserving graft-versus-leukemia (GvL) reactions. In this study we aimed to develop a protocol for the efficient isolation and in vitro expansion of regulatory T cells and to conduct the first trial testing the toxicity and therapeutic efficacy of Treginfusion in five patients with otherwise treatment-refractory cGvHD. Materials (or patients) and Methods: Allogeneic Tregs were isolated from unstimulated leukapheresis products of the corresponding HLA-matched donors by Ficoll density centrifugation following magnetic activated bead sorting (MACS). To increase the amount and purity regulatory T cells were cultivated in the presence of rapamycin, IL-2 and anti-CD3/anti-CD28 beads for 7-12 days. Purity and functionality of Tregs was assessed during the purifi cation and expansion process using suppression assays and fl ow cytometry. Tregs were infused after a median time of 35 month (range 26-34month) after HCT. The kinetic and suppressive capacity of regulatory T cells in the peripheral blood was monitored weekly after adoptive transfer using fl ow cytometry and clinical grading of GvHD organ manifestations. 3/5 patients received low-dose IL-2 for 8 weeks after Treg infusion in order to support Treg expansion. Results: Final products contained regulatory T cells with a mean purity of 84.7% (of total cells; range: 77.7% -94%) and a mean quantity of 2.4x10 6 Tregs per kg BW (range: 0.52x10 6 -4.45 x10 6 ). All isolated cell products showed in vitro suppressive activity. Transfusion was well tolerated by all patients. Upon transfusion two of five patients showed a clinical response with improvement of cGvHD symptoms. The other three patients showed stable cGvHD symptoms for up to 21 month. In 3/5 patients immunosuppressive treatment could be reduced. Increased counts of Tregs were detectable in 4/5 patients upon Treg infusion. Suppression of activation marker expression on CD8 T cells was observed in 3 of 5 patients. Transfusion was well tolerated by all patients. With a median follow up time of 19 month after infusion no relaps of hematologic malignancy occurred. However, one patient developed malignant melanoma and another patient Bowen´s skin cancer 4 month and 11 month after Treg infusion, respectively. Discussion: This data show (i) a feasible and reproducible approach of isolating functional Tregs in high quantity and purity for clinical application and (ii) opportunities and risks of adoptive Treg transfer into patients with chronic GvHD. Disclosure of Interest: None Declared.

Introduction: Allogeneic transplantation is the only curative option for patients with high risk hematologic malignancies. Only one third of them have an HLA identical donor and around 60-70% will fi nd an unrelated donor, that´s why HAPLO-HSCT off ers a therapeutic option to most of these patients with the advantages of quick availability, easy programation and logistics, and a committed donor. Materials (or patients) and Methods: We retrospectively evaluate the results of HAPLO-HSCT with reduced conditioning or myeloablative regimens and GVHD prophylaxis based on PT-CY (50 mg/ kg on days +3 and +4) and a calcineurin inhibitor plus mycophenolate from day +5 performed in GETH centers. Results: From Dec-2007, 80 HAPLO-HSCT have been done in 14 centers. Median age was 37 years (16-66), 67.5% were males and all were in advanced phases of their disease or presented high risk features (29 Hodgkin´s, 22 AML, 9 ALL, 8 MDS, 5 NHL, 4 myeloma and 2 myelofi brosis). Previous HSCT has been employed in 65%, autologous in 38 and allogeneic in 15 (5 siblings, 3 unrelated and 7 cord blood transplants), and in 35% the HAPLO-HSCT was their fi rst transplant. Disease status at HAPLO-HSCT was CR in 45%, with persistent disease in 55%. Bone marrow was the graft source for 51% and peripheral blood for 49%, non T-cell depleted in all cases. The haploidentical donor was the patient´s mother (21), father (7), brother/sister (35) or off spring (17). Non-myeloablative conditioning was employed in 77.5% and myeloblative in 22.5%. Median neutrophils engraftment was reached at day +18 (13-45) and platelets >50K at day +27 (11-150). Main toxic complications were grade II-III mucositis in 36%, febrile neutropenia in 75% and CMV reactivations in 62%, with a transplant related mortality rate of 12.5% at day +100 and 19% at 6 months post-transplant. Acute GVHD grade II-IV aff ected to 24/73 patients at risk (33%), with grade III-IV in 10/73 (14%). Chronic GVHD was present in 12/51 (24%), being extensive in 6/51 (12%). After a median follow-up of 9 months (0.3-49), 26/80 patients have died due to relapse in 13, infections in 10 and GVHD in 3 cases. Event-free survival and overall survival at 1 year were 48% and 60% respectively. Immune reconstitution was fast and complete in those evaluated. Discussion: HAPLO-HSCT with PT-CY is a useful tool in the treatment of high risk hematologic malignancies, rendering long-lasting remissions with limited toxicity, low GVHD incidence and early immune reconstitution. Disclosure of Interest: None Declared.

Introduction: The standard risk factors for acute graft-versus-host disease (aGVHD) after allogeneic stem cell transplantation (allo-SCT) from related or unrelated donors are well defi ned and include HLA mismatch or unrelated donor, older recipient age, and female donor for male recipient (FM). The steroid-refractory (SR) forms of aGVHD are important to consider because they often have a major deleterious impact on transplant outcome. Unfortunately, the specifi c risk factors for SR aGVHD are less clearly defi ned. To characterize these risk factors after allo-SCT from matched related or unrelated donors, we undertook a retrospective analysis of adult patients transplanted at our center between 01/01/2000 and 12/31/2012. Materials (or patients) and Methods: Steroid-refractory aGVHD was defi ned as aGVHD progressing after 3 days of treatment, or unchanged after 7 days, or in incomplete response after 14 days. GVHD occurring after donor lymphocytes infusion were excluded. Cumulative incidences (CI) were used for SR aGVHD in a competing risk setting with death as a competing event. The Gray test was used to compare CI curves. Results: Six hundred and thirty four patients were identifi ed and included in the present study. The median age was 50 years (18-67). Diseases were AML (n=230), ALL (n=104), myeloma (n=80), NHL (n=74), Hodgkin's disease (n=18), MDS (n=47), CLL (n=29), CML (n=18), aplastic anemia (n=19), and MPS (n=15). Status at transplant were CR1 or PR1 or chronic phase (n=260), > CR1 or PR1 (n=237), refractory (n=101), or untreated (n=36). Conditioning regimens were RIC (n=405) or MAC (n=229). Rabbit ATG was administered to 327 patients, of whom 298 received a RIC regimen. Donors were MRD (n=360) or MUD (n=274). Sources of stem cells were PB (n=452), BM (n=177), missing data (n=5). The prophylaxis of GVHD was CsA+metho for 339 patients.

In the whole population, 71 patients presented a SR aGVHD at a median time of 29 days (8-137) after transplant, representing a CI of 11.2% ± 1.2%. Their OS at 1 year post-transplant was 27% ± 5%. In univariate analysis, the risk factors for SR aGVHD were MAC (P=0.02), MUD (P=0.02), no ATG (P=0.01), and a trend for FM (P=0.07). Other variables considered in univariate analysis were recipient age (P=0.6), female vs male donor (P=0.6), recipient CMV status (P=0.7), status at transplant (P=0.2), PB vs BM graft (P=0.9), number of CD34+ cells (P=0.4), and GVHD prophylaxis (P=0.6). In multivariate analysis, the risk factors for SR aGVHD were MUD (HR=2.5, 95%CI: 1.5-4.1, P=0.0003), FM (HR=2, 95%CI: 1.2-3.4, P=0.008), and no ATG (HR=2.1, 95%CI: 1.3-3.4, P=0.002). Patients were then divided into 3 groups. The CI of SR aGVHD was 2.7% ± 1.6% in the low-risk group (no MUD + no FM + ATG; n=112), 21.8% ± 3.4% in the high-risk group (MUD + FM +/-ATG, or MUD + no FM + no ATG; n=147), and 9.6% ± 1.5% in the intermediate-risk group (n=375); P=10 -6 . Discussion: We conclude that MUD, FM, and no ATG were independent risk factors for SR aGVHD in adult patients after allo-SCT from MRD or MUD. To avoid a high risk of SRaGVHD, our study suggest that in unrelated transplants, one should avoid a female donor for a male recipient and use ATG if the recipient is a female or if both recipient/donor are males. Disclosure of Interest: None Declared.

Introduction: Graft-versus-host disease (GVHD) remains a major problem following HSCT. Aim of this study was to ascertain the role of Extracorporeal Photopheresis (ECP) in 76 patients with steroid refractory or dependent acute and chronic GVHD (aGVHD and cGVHD). Materials (or patients) and Methods: A total of 37 patients were treated for aGVHD (median age 12 years, median follow-up for surviving patients 4 yrs [2 months-12 yrs]) and 39 patients for cGVHD (median age 25 years, median follow-up for surviving patients 4 yrs [1 month-12 yrs]). Patients were treated with ECP on two consecutive days at weekly intervals for the fi rst month, every two weeks during the second and third months, and then at monthly intervals for a further three months. Briefl y, patients with aGvHD were ruled out from the ECP protocol if they had: a) completed their planned 22 procedures, b) had aGvHD progression under ECP, c) had GVHD response but interrupted their treatment early on given their high risk of relapse. Patients with cGvHD were ruled out from ECP therapy if they reached complete response (CR), partial response (PR), or minor response (MR) following the 22 planned ECP procedures or before if they had aGvHD progression under ECP. Results: The crude aGVHD response was 67% and the complete aGVHD Free Survival was 50% (95% CI, 25-70). The crude cGVHD response was 77% while the cGVHD Free Survival was 34% (95% CI 21-47). Among aGVHD patients, a better aGVHD Free Survival was associated to aGVHD lower grade ( For TRM and RI no factors were identifi ed as predictors. Discussion: No role of hematological value or aphaeretic cell count were found in our patient cohorts, but a trend for improved outcome for patients with higher aphaeretic yield was observed.

Larger studies are warranted to support the cell dose eff ect. Disclosure of Interest: None Declared. Most patients in these studies received T-cell replete transplants. As a consequence, and because antithymocyte globulin (ATG) reduces the risk of GVHD, the impact of FM in the specifi c setting of allo-SCT with ATG remains poorly defi ned. Materials (or patients) and Methods: To explore the impact of FM in the ATG setting, we undertook a retrospective analysis of male adult patients transplanted with rabbit ATG at our center between 01/01/2000 and 12/31/2012. Overall survival and disease-free survival (DFS) were calculated using the Kaplan-Meier estimate, with comparisons by the log-rank test. Cumulative incidences (CI) were used for NRM, relapse (REL), and GVHD in a competing risks setting. MUD were matched at the allele level for HLA-A, B, C, DRB1, DQB1.

Results: 212 male patients were identifi ed and included in the present study. The median age was 56 years (18-67). Diseases were AML (n=61), ALL (n=16), NHL (n=36), Hodgkin's disease (n=10), myeloma (n=37), MDS (n=26), aplastic anemia (n=7), CLL (n=11), CML (n=1), and MPS (n=7). Status at transplant were CR1 or PR1 or chronic phase (n=74), > CR1 or PR1 (n=83), refractory (n=34), or untreated (n=21). Conditioning regimens were RIC (n=195) or MAC (n=17). Donors were MRD (n=110) or MUD (n=102). Eightyeight patients were transplanted with a FD. Source of stem cells were PB (n=193), BM (n=18), missing data (n=1). Prophylaxis of GVHD consisted of CsA+metho for 84 patients. The median follow-up was 42 months (5-149). In univariate analysis, the 3-year OS and DFS in FD vs MD groups were 48% ± 6% vs 56% ± 5% (P=0.3) and 42% ± 5% vs 50% ± 5% (P=0.4), respectively. The 3-year NRM and REL in the same groups were 25% ± 5% vs 19.6% ± 4% (P=0.4) and 30% ± 5% vs 31% ± 4% (P=0.9), respectively. The CI of aGVHD II-IV, aGVHD III-IV, SR aGVHD, and extensive chronic GVHD in FD vs MD groups were 37.5% ± 5% vs 33% ± 4% (P=0.5), 22.7% ± 4% vs 17% ± 3% (P=0.3), 17.2% ± 4% vs 8.3% ± 3% (P=0.049), and 20% ± 4% vs 22.5% ± 4% (P=0.9), respectively. Other variables considered in univariate analysis for SR aGVHD were recipient age (P=0.9), recipient CMV status (P=0.7), disease status (P=0.9), RIC vs MAC (P=0.16), MRD vs MUD (P=0.09), PB vs BM graft (P=0.4), and GVHD prophylaxis (P=0.2). In multivariate analysis, the risk factors for SR aGVHD were FD (HR=3.1, 95%CI: 1.2-7.9, P=0.01) and MUD (HR=2.9, 95%CI: 1.1-7.1, P=0.02). The CI of SR aGVHD were 32.3% ± 9.7% in the FD + MUD group (n=29), 9.7% ± 3.8% in the MD + MUD group (n=73), 10.7% ± 4.1% in the FD + MRD group (n=59), and 6.5% ± 3.7% in the MD + MRD group (n=51); P=0.004. Discussion: We conclude that FD was an independent risk factor for SR aGVHD after allo-SCT with rabbit ATG in male adult patients mostly transplanted with PB after RIC. There was no impact of FD on aGVHD II-IV or III-IV, chronic GVHD, NRM, REL or survival. Finally, the absence of impact of FD on NRM and survival should be interpreted with caution given the retrospective design of the study and because we cannot exclude a possible limiting eff ect of an insuffi cient number of patients. Disclosure of Interest: None Declared. Introduction: Extracorporeal photopheresis (ECP) is a useful second-line therapy for patients with chronic and acute graft-versus-host disease (GVHD). While in the USA only the closedloop system is used, in other countries both closed-loop and two-step open-loop procedures are employed. This study is aimed to prospectively compare the efficacy of the two ECP procedures, with each patient being treated with both treatment modalities. Materials (or patients) and Methods: Patients were scheduled to undergo the open-loop procedure (modality A) for 2 consecutive days followed by the closed-loop procedure (modality B) to be performed for 2 consecutive days two weeks later. Data on levels of hematocrit (Hct), WBC, monocytes (Mon), lymphocytes (Lym), neutrophils (Neu), product bag volume and collection duration were planned to be analyzed. Additionally, early and late apoptosis of product cells was to be evaluated. Samples were collected from the product bag prior to adding 8-MOP and following UVA irradiation. Mononuclear cells (MNC) were separated using Lymphoprep (Axis-Shield), counted and suspended in cell culture flasks at a concentration of 4x10 6 /ml in the RPMI-1640 medium supplemented with 10% fetal bovine serum. The cells were incubated for 48 hours at 37 ° C. MNC subsets (Lym, Mon) were identified by flow cytometry. Early apoptosis was evaluated using Annexin V and late apoptosis was assessed using Propidium Iodid. 3 

Introduction: Chronic graft-versus-host disease (GVHD) is a major complication following allogeneic hematopoietic stem cell transplantation and is associated with a substantial morbidity and mortality. It is a systemic infl ammatory disorder that refl ects the lack of immune tolerance between donor-derived immune competent cells and host organs. Human chorionic gonadotropin hormone (hCG) is a natural occurring hormone during pregnancy secreted by syncytiotrophoblasts of the placenta. We had previously observed (Koldehoff et al; J Leukoc Biol 2011) that the rejection of transplanted skin was signifi cantly delayed by hCG in a mouse skin transplant model and had also demonstrated that tryptophan-catabolizing enzyme, indoleamine-2,3-dioxygenase (IDO), interleukin-10 (IL 10) and T-regulatory cells (Tregs) increased signifi cantly in females treated with hCG as preconditioning therapy for in-vitro-fertilization. Since all these factors are known to induce tolerance and given the low rate of adverse eff ects, we off -label used low dose of hCG to treat 20 patients as forth-or fi fth-line therapy with steroid-refractory or intolerant severe-grade chronic GVHD. Materials (or patients) and Methods: Because all of these factors are known to induce tolerance and given the low rate of adverse eff ects in preconditioning therapy, we off -label used low dose of hCG (187 IU) to treat 8 male and 12 female patients (median age 48, r. 28-68) with moderate or severe grade of chronic GVHD according to the NIH criteria; all patients had been informed of the experimental state of this treatment and provided written consent Results: The median number of sites of chronic GVHD involvement per patient was 3 (range, 1-6). hCG therapy was started as 4 or 5th line-therapy together with preexisting medication with prednisone and a calcineurin inhibitor. Twelve of 20 patients (60%) had an objective partial response during 8 weeks of hCG treatment with at least 30% improvement according to the TSS score. Responses included softened skin and subcutaneous tissue; decreased erythema and extent of sclerodermatous, hidebound skin; improved joint mobility and gait; gastrointestinal improvements; and resolution of neuropathy. Nine patients had stable disease (6 with minor responses). Only one patient with previous ATG treatment showed progression of her liver GVHD (histologically proven) and died from GHVD. All other patients were well and alive. Daily low-dose hCG was well tolerated. Adverse events that were possibly related to hCG included reversible and asymptomatic CTCAE grade 4 hypertriglyceridemia (n=1), grade 2 constitutional symptoms (fever, malaise, fatigue; fl ush, breast enlargement). IDO expression increased up to 7 times and IL10serum level up to 2 times after 3 weeks of hCG therapy (P<0.003 and P<0.04). T-regulatory cell expansion was documented in one patients. Discussion: This successful use of hCG in an immune disorder warrants further studies to assess its role as an immunosuppressant in GVHD and potentially other autoimmune disorders. Disclosure of Interest: None Declared. Introduction: Several reports have shown that the measurement of CSA and/or MMF blood concentrations does not suffi ciently refl ect the eff ects of the applied drugs on immune cells; in fact, despite having blood concentrations within therapeutic range, some patients still experience acute GVHD. Pharmacodynamic (PD) biomarker monitoring may refl ect more accurately individual response to immunosuppressive drugs and clinical outcome. Materials (or patients) and Methods: We selected a battery of biomarkers for prospective PD monitoring in 37 RIC alloSCT patients: T cell activation measured by intracellular ATP production in CD4 + -T lymphocytes as a refl ection of the overall immunosuppressive status achieved with CsA+MMF; and specifi c biomarkers based on the mechanism of action of these drugs: soluble production (ELISA) and intracellular expression (Flow cytometry) of IL-2, IFNγ, and TNFα in CD4 + and CD8 + -T cells such as eff ector T-cell response quantifi cation. We also evaluated the frequency of CD4 + CD25 high CD127 low FOXP3 + T-cell (Tregs). Results: ATP levels at 0.5 months after alloSCT were significantly higher in those patients with grade 0-I GVHD (260 ng/ml, 6.6-876) in comparison to grade II-IV GVHD (144 ng/ml, 21.2-691.7) (P=0.028); however, at 1 month after transplantation a trend for higher levels was observed in patients developing grade II-IV GVHD in comparison to grade 0-I GVHD (357.1 ng/ml vs. 233 ng/ml, respectively) (P=0.26). ATP measured at GVHD diagnoses identified patients with severe (grade III-IV) GVHD with a ROC-AUC of 0.76. Patients with grade II-IV GVHD showed higher %CD3 + CD69 + CD4 + and %CD3 + CD69 + CD8 + lymphocytes expressing IL-2, IFNg and TNFa than those with grade 0-I GVHD along the period of study. These differences were statistically significant for CD3 + CD69 + CD8 + IL-2 + levels at 0.5 and 1 month after transplantation (P=0.032 and P=0.016, respectively). No statistically significant differences were observed in soluble IFNγ and TNFα levels between grade 0-I and II-IV GVHD patients. The median %Tregs lymphocytes was higher in 0-I GVHD during the whole period of the study, particularly at 2 months after alloSCT (P<0.0001). The ratio %CD3 + CD69 + CD8 + IL2 + /Tregs was higher in patients with grade II-IV GVHD than in those with 0-I GVHD, being significant at 1 (1.3, 0-12.06 vs. 0.07, 0-1.84; P=0.025) and 2 (1.96, 0-81.3 vs. 0.09,0-6.75; P=0.002) months after transplantation. Discussion: In conclusion, ATP production in CD4 + lymphocytes, %CD3 + CD69 + CD8 + IL2 + and %CD3 + CD69 + CD8 + IL2 + /Tregs rate monitoring after transplantation may help physicians to identify individual response to CsA+MMF and patients at high risk of GVHD. Disclosure of Interest: None Declared. Introduction: The combination of CsA and MMF is widely used for prophylaxis of GVHD in the context of RIC alloSCT. CsA and MMF show large inter-and intraindividual pharmacokinetic variability and confl icting fi ndings have been reported when studying the association between their blood levels and the risk of GVHD. Materials (or patients) and Methods: Measures of through levels (C min ) (37 patients) and area under the curve (AUC [0-8h] ) (12 patients) of CsA and MPA were prospectively performed along the fi rst 2 months after RIC alloSCT. CsA (5 mg/kg/d) and MMF (15 mg/ kg/8h) were started IV and then switched to oral administration when tolerated by the patients.

Results: CsA C min at 0.5 months was signifi cantly higher than C min at day 0 (P=0.001) and C min at 1 month post-transplantation (P=0.039). Higher MPA C min levels were achieved after oral MMF administration in comparison to IV (P<0.005); however, only a small proportion of subjects (3% with IV and 24% with oral MMF) achieved through concentration targets >2 mg/ml. There was no correlation between MPA C min and AUC. Median AU C[0-8h] of MPA at d+1 during IV MMF administration was 52.9 mg*h/ml (range 16. 3-93.9) . MPA concentrations signifi cantly declined after initiation of oral MMF (MPA AUC [0-6h] 27.6 mg*h/ml, 13.4-67.8; P=0.01) with only 36.4% of patients achieving levels > 30 mg*h/ml. No differences in CsA C min on day 0 were observed between grades 0-I and II-IV GVHD patients. Patients with II-IV GVHD had lower C min of CsA at 0.5 months than those with 0-I GVHD (P=0.018); however, no correlation was observed between CSA AUC [0-8h] and GVHD. No diff erences were observed between 0-I and II-IV GVHD patients in MPA C min measured at day +1, 48-72h after MMF oral initiation, and 3-4 weeks after alloSCT. All patients with grade 0-I GVHD had MPA AUC [0-8h] on day +1 and 48-72h after MMF oral administration higher than 30 mg*h/ml; however, most patients with grade II-IV GVHD (63% on day +1 and 57% 48-72h after oral MMF, respectively) did not achieve this value (P=0.08) (Figure 1 ). Discussion: In conclusion, our results show that trough MPA concentrations post-transplant do not accurately describe MPA AUC [0-8h] nor GVHD risk; and that high MPA exposure is associated with more efficacy in GVHD prevention. Individualized MMF dosing according to AUC [0-8h] levels during the first month after RIC alloSCT may be important for GVHD prophylaxis after RIC alloSCT. Disclosure of Interest: None Declared. Introduction: Graft versus host disease (GvHD) is a common complication following haematopoietic stem cell transplantation (HSCT). 1 st line therapies are corticosteroids (appropriate dose/route for grade) and topical treatment if required [1] . Many cases resolve but some develop steroid-resistant (SR)-GvHD. 2 nd line therapy should commence if no response after 5 days, or after 3 days if symptoms progress [1] , including additional immunosuppressive agents, monoclonal antibodies, serotherapy, mesenchymal stromal cells and extracorporeal photopheresis. We compared and evaluated practice at our unit in the last 5 years to recently published guidelines providing 1 st and 2 nd line treatment of SR-GvHD. Materials (or patients) and Methods: A retrospective review of patient records and the CBMTU database between 31/01/08-20/12/12. Patient Cohort: Patients developing SR-GvHD following HSCT for primary immunodefi ciency. Data collected: Diagnosis, HSCT date, GvHD: prophylaxis, occurrence date, grade, treatment, treatment start date. Results: Of 152 patients, 18 (12%) developed SR-GvHD, 5 (28%) of whom died. 8 (44%) developed Grade II GvHD (6 skin, 2 mixed), 5 (28%) Grade III (1 gut, 4 mixed), 5 (28%) Grade IV (1 skin, 4 mixed).

Median day of GvHD occurrence: 24 (Range: 0-125). Some patients received a boost infusion for falling chimerism that appeared to induce more severe GvHD. All patients received 1 st line treatment in accordance with the guideline. Median time to 2 nd line treatment: 24.5 days. 2 nd line: 33% of patients received 1 monoclonal antibody, 11% received 2 monoclonal antibodies -none received serotherapy or other therapy on/by the 5 th day of onset of symptoms [1] . Discussion: 1 st line therapy was extremely well implemented but there was signifi cant variation in time between GvHD diagnosis and 2 nd line treatment for SR-GvHD patients. Reasons why 2 nd line therapy was not given according to the guideline were potentially: Fluctuating symptoms, varying timeline/severity and presentation of GvHD in many patients. Potential cautious prescribing by clinicians as there are worries due to possible serious side eff ects of 2 nd line treatments [1] . Outside referral was required for MSC therapy, which delayed treatment. Limitations of the study include: Progressive symptoms could not be assessed, as often records were unclear whether symptoms were progressive and of the exact timeframe of events. Guidelines are useful, but there is no clinical consensus for application of 2 nd line treatment as this is an ever developing fi eld [1] .Our cohort includes patients treated before these guidelines were published [1] . SR-GvHD is a life-threatening diagnosis with poor outcomes. Increasing access of patients to new treatment options and reducing the time taken to implement these therapies may help to improve outcomes. Disclosure of Interest: None Declared. conditioning. Underlying diseases were chronic myeloid leukemia (n=11), acute nonlymphoid leukemia (AML, n=107), acute lymphoid leukemia (ALL, n=52), myelodysplastic syndrome (n=35), severe aplastic anemia (n=4), non-Hodgkin's lymphoma (n=10), and others (n=2 Introduction: The prevention of severe acute Graft-versus-Host-Disease (GvHD) and yet preserving an adequate Graft-versus-Leukemia-Eff ect are still challenging tasks for the patient's outcome after allogeneic HSCT (alloHSCT). We retrospectively compared the standard GVHD-prophylaxis consisting of Cyclosporine A (CSA) and Methotrexate (MTX) with the more tolerable combination of CSA and Mycophenolate-Mofetil (MMF) in a ten year single-center study. Materials (or patients) and Methods: We retrospectively analysed the outcome of 201 patients who underwent alloHSCT between 2000 and 2010 at our institution. The median age of patients was 48 years (15-71 years). GVHD prophylaxis consisted of CSA/MTX (n = 125) or CSA/MMF (n = 76). In case of mismatched or unrelated donor grafts ATG (n = 110) was additionally given. The analysed groups did not diff er signifi cantly in terms of age, conditioning regimen, gender, HLA-matching and CMV-serostatus. The median follow up of survivors was 47 months (4-124 months). Results: The overall rate of acute GvHD was 53.7% and occurred in median 29 days (9-187) after alloHSCT. After CSA/MTX acute GvHD occurred in median 9.5 days earlier than after CSA/MMF (26 vs. 35.5 days, P = 0.041), whereas the overall severity of acute GvHD was comparable in both groups. There was a trend for a higher frequency of aGVHD °IV of the liver in patients receiving CSA/MMF when compared to patients receiving CSA/MTX (54.5% vs. 15%, P = 0.071). This reached statistical signifi cance in the subgroup of patients without ATG (0% vs. 55.6%, P = 0.023). In addition the incidence of steroid-refractory aGVHD was higher in the subgroup without ATG after prophylaxis with CSA/MMF when compared with CSA/MTX (46.2 vs. 20.7%, P= 0.083). Of note, the incidence of intestinal acute GvHD was higher in the CSA/MTX/ATG subgroup (54.1% vs. 12.5%, P = 0.006). 29.6% of patients experienced chronic GvHD after a median time of 192 days (33-1454) after alloHSCT. However, neither in rate nor in severity of chronic GvHD statistically signifi cant diff erences were found between the groups. The incidence of CMV infections was 29.6%, without signifi cant diff erences in the analysed groups. The estimated survival after 5 years for the CSA/MTX vs. CSA/MMF group was: overall survival 47.0 vs. 50.7% (P = 0.91), relapse-free-survival 31.3% vs. 29.4% (P = 0.588), event-free-survival 41.8 vs. 41.9 (P = 0.725) and the non-relapsemortality 38.9 % vs. 47.2% (P = 0.588), respectively. Cox-regression-method identifi ed involvement of intestinal and liver acute GvHD, severity of acute GvHD and steroid-refractory acute GvHD as negative infl uence on OS. Presence of chronic GvHD had a positive infl uence on OS. Discussion: GVHD prophylaxis with CSA/MMF is a feasible alternative to CSA/MTX with respect to frequency and severity of acute and chronic GvHD, incidence of CMV infection, as well as survival after alloHSCT. Disclosure of Interest: None Declared. Introduction: The GVHD pathogenesis involves migration of the donor naïve T-cells into the secondary lymphoid organs in the recipient, which is mainly steered by CD62L and CCR7. Therefore, we aimed to study whether the expression of CD62L and CCR7 in T cells is associated with GVHD, both acute and chronic. Materials (or patients) and Methods: This single center study included 52 donor-recipient pairs. Samples were collected prospectively from the apheresis product and phenotyped by fl ow cytometry, and CD62L and CCR7 expression in CD4+ and CD8+ T-cells were compared between patients who developed acute or chronic GVHD (n=26) and those who did not (n=21). Results: Interestingly, in the case of the acute form, the GVHD group received a significantly higher percentage of either CCR7+CD4+ (P=0.03) and CCR7+CD62L+CD4+ T-cells (P=0.042) compared to the no GVHD group. These results were confirmed when the patients with acute GVHD were divided in degrees according to the severity of the disease. Regarding chronic GVHD, significantly higher percentage of CCR7+ CD8+ (P=0.011), CD62L+ CD8+ T cells (P=0.015), CD62L+CCR7+ CD8+ T cells (P=0.001) was observed in those patients who developed chronic GVHD in comparison to those who did not. These data were also confirmed when patients were subdivided in degrees of the disease severity. The multivariable analysis confirmed that these T cell subpopulations had no association with the classical clinical GVHD predictors studied. Discussion: Our results indicate that CCR7+ and CD62L+ expressing T-cells are the alloreactive cells in the graft, so they would mediate the pathogenesis of GVHD and are potential therapeutical targets for the disease. Disclosure of Interest: None Declared. Introduction: The aim of our retrospective analysis was to evaluate the clinical eff ect of extracorporeal photopheresis (ECP) and its impact on immunosuppressive therapy in patients treated for acute graft-versus-host disease (aGVHD) after allogeneic hematopoietic stem cell transplantation. Materials (or patients) and Methods: Over a 6-year period (2007) (2008) (2009) (2010) (2011) (2012) (2013) , 167 procedures were performed in 20 patients (pts) with steroid refractory aGVHD (10 male/10 female, median age 51, range 22-71). ECP was performed as third or fourth line of therapy in 10/20 pts. 15/20pts were in grade III/IV aGVHD at the beginning of the treatment. ECP was performed using an off -line technique (Cobe Spectra cell separator and PIT System UV-A devices) with the following schedule: 2 procedures a week for 1 month then 2 procedures every other week for 1 month. Stopping ECP was decided on the clinical basis. Venous access was provided by a CVC in 12/20 pts. The response was assessed according the 1994 Consensus Conference criteria: complete remission (CR), partial remission (PR) no response or progression (NR). Results: ECP was started at a median of 18 days after aGVHD onset (range 5-34). Pts underwent 3 to 15 procedures (median 8) with a mean treatment period of 26d (range 14-127).No relevant treatment related complications were observed. At one month, a clinical response (CR+PR) was obtained in 12 of 20 (60%) pts: 5/5 with grade II and 6/15 (40%) with grade III/IV aGVHD.The highest percentage of response was observed in patients with muco-cutaneous involvement (87% RC+PR) while only 40% of pts with prevalent liver involvement were responsive to ECP. The treatment was continued in 8 /20 pts (1 in CR, 5 in PR, 2 in NR): the patient in CR stay in remission, 3 of 5 pts in PR obtained a CR and 1of 2 pt in NR gained a PR. Finally, at the end of treatment 9 pts were in CR, 3 pts were in PR e 8 pts in NR. In responders patients, the prednisone equivalent dose was 1 mg/kg/die at the onset of ECP (range 0,5-2,7) and 0,33 mg/Kg/die at the end of treatment (range 0,06-0,6). Only two of the responders pts with grade III aGVHD received new additional immunosuppressive therapy (1pts infl iximab, 1 pts etanercept) started during ECP. The median follow up in responders group was 140 days ( range 63-1022), 5 pts survived and 8 pts died (4pts for relapsing disease, 2pts for infections, 1pt for PT-PTT). In the NR group, the ECP was stopped after 3 procedures in 3/8 pts due to clinical reasons (post transplant PTT, hepatic grade IV aGVHD) and after 8 procedures in the remaining 5 pts. All the 8 non responders patients died (6pts for aGVHD e 2 pts for pulmonary infections). The Kaplan Mayer probability of overall survival was signifi cantly higher among responders pts compared with those with NR ( 72% vs 25% at three months and 46% vs 0% at six months) ( P=0.001). Discussion: In our experience ECP proved to be a safe and eff ective treatment in patients with corticosteroid-resistant aGVHD. The low reported toxicity of ECP allowed the procedure to be performed also in pts with poor performance status. In our analysis better response rates were obtained in patients with cutaneous involvement and the reported steroid-sparing eff ect of ECP could be confi rmed. Finally a signifi cant association between the clinical response and the OS could be shown, which is in line with previous published results. Disclosure of Interest: None Declared.

[PH-P388] Introduction: Allo-HSCT can be burdened by life-threatening complications, being GVHD the major cause of morbility and mortality. Clinical and physio-pathological evidences showed that vascular endothelium could be a target of GVHD in very early phase; therefore markers of endothelial damage are warranted as valuable support in GVHD diagnosis. We conducted a study with primary endpoint to identify and count circulating endothelial cells (CEC) in peripheral blood of patients undergoing alloBMT as a function of endothelial damage. Materials (or patients) and Methods: The CellSearch System® is used to capture and enumerate CEC. Enriched and stained cells are dispensed into a MagNest® cartridge that is scanned and individual images of cells are scored as CEC, based on CD146+, CD105+, DAPI+ and CD45-phenotype. Patients undergoing allo-HSCT were tested before (T1), after the conditioning regimen (T2), at engraftment (T3), at GVHD onset (T4) and at 1-2 weeks after steroids treatment (T5). Ten healthy subjects served as controls. Results: We enrolled 40 patients with hematologic neoplastic diseases ( Discussion: Circulating endothelial cells can represent a promising marker to monitor endothelial damage in patients undergoing allo-HSCT. The confi rmation of the clinical utility of CEC counts in a larger series of patients, together with the use of a semi-automatic, standardized and reproducible technology, will allow a valuable help in the diagnostic defi nition of GVHD in early phase, and moreover could be a valid complement in the prognostic stratifi cation of patients candidates to allo-HSCT. Disclosure of Interest: None Declared. Introduction: Steroid-refractory graft-versus-host disease (GVHD) is a major and often fatal complication after allogeneic stem cell transplantation (alloSCT). Although the pathophysiology of steroid-refractoriness is not fully understood, evidence is accumulating that endothelial cell stress is involved, and endothelial thrombomodulin (TM) plays a role in this process. Here we assess if single nucleotide polymorphisms (SNPs) within the TM gene (THBD) predict outcome after alloSCT. Materials (or patients) and Methods: Seven SNPs within the THBD gene were studied (rs1962.TC, rs1042579.TC, rs1042580.AG, rs3176123.TG, rs3176124GA, rs3176126.GA and rs3176134.CT) in a training cohort of 465 allografted patients. The relevant genotypes were then re-assessed in an independent validation cohort (n=386). Results: In the training cohort, an increased risk of non-relapse mortality (NRM) was associated with three SNPs out of seven tested: rs1962.CC, rs1042579.TT (455V, linked with rs3176123GG) and rs1042580.GG. When patients were divided into risk groups (one vs. no high-risk SNP), a strong correlation with NRM was observed (HR 2.32 95% CI 1.36-3.95; P=0.002). More specifi cally, NRM was predicted by THBD SNPs in patients who later developed GVHD (HR 3.03 95% CI 1.61-5.68, P=0.0006), but not in patients without GVHD. In contrast, THBD SNPs did not predict incidence of acute GVHD. Multivariate analyses adjusting for clinical variables confi rmed the independent eff ect of THBD SNPs on NRM. All fi ndings could be reproduced in the validation cohort. Discussion: THBD SNPs predict mortality of manifest GVHD but not the risk of acquiring GVHD, supporting the hypothesis that endothelial vulnerability contributes to GVHD refractoriness Disclosure of Interest: None Declared. Introduction: Graft-versus-host disease (GVHD) remains a major limiting factor for hematopoietic stem-cell transplantation (HSCT). Risk factors for GVHD include, but are not limited to, HLA mismatches, a female donor for a male recipient, and stem-cell source. Especially sex mismatching combined with HLA mismatching seems to lead to an increased risk of alloreactive T-cell responses. The pathophysiological mechanisms via which these factors aff ect the GVHD risk, are poorly understood. GVHD evidently involves T-cell recognition of non-self antigens expressed by recipient cells, as depletion of T cells from the graft minimizes the GVHD incidence. We earlier reported signifi cant diff erences in skin-infi ltrating T cells in male pediatric patients who developed GVHD after receiving a graft derived from an unrelated donor. This previous study suggested that CD4/CD8 ratios among skin-infi ltrating T cells were aff ected by donor sex. In the current study, we retrospectively analyzed if HLA and sex mismatching quantitatively aff ects the composition of GVHD-induced cutaneous T-cell infi ltrates in a pediatric HSCT cohort. Materials (or patients) and Methods: We analyzed tissue sections from steroid-naïve pediatric HSCT-patients (n=15 males and n=8 females) with histologically and clinically confi rmed acute GVHD. The absolute numbers of CD3+CD4+ and CD3+CD8+ T cells were quantifi ed in tissue sections prepared from archived formalinfi xed paraffi n embedded skin biopsies. These sections were subjected to a routine immunofl uorescent staining protocol using a validated set of CD3, CD4 and CD8-specifi c antibodies, where after the T-cell infi ltrates were quantifi ed by fl uorescent microscopy in a blinded manner. Results: Graft source (HLA-identical related donor (IRD, n=7), matched-unrelated donor (MUD, n=12) or cord blood (CB, n=4)) had no signifi cant impact on the numbers of skin-infi ltrating CD8+ T cells. HSCT with CB seemed to result in the highest number of CD4+ T cells in the skin (compared to MUD P=0.02). Patients receiving an HSCT from a sex-mismatched donor displayed an increased number of CD4+ T cells when compared to recipients of a graft from a sex-matched unrelated donor when the donor was unrelated (either MUD or CB) (P=0.01). This sexmismatching eff ect was not observed in IRD transplant pairs. Furthermore, unrelated patient-donor pairs in which the donor can recognize T-cell epitopes derived from the recipient mismatched HLA, so called predicted indirectly recognizable HLA epitopes (PIRCHES), showed an increased number of skin-infi ltrating CD8+ T cells when compared to patients who did not express PIRCHES. The combined expression of HLA class I-or II-presented PIRCHES with a sex mismatch resulted in the highest number of skin-infi ltrating T cells. Discussion: Our results provide the fi rst indication that an increasing number of major (i.e. HLA) and minor (i.e. HY) histocompatibility antigen-derived T cell epitopes expressed by recipient skin cells leads to increased accumulation of CD4+ and CD8+ T cells in the skin. These observations may explain the pronounced eff ect of sex mismatching in HLA-mismatched transplants as evidenced by registry-based epidemiological studies. Disclosure of Interest: None Declared.

Introduction: The best conditioning regimen before allogeneic stem cell transplantation (alloSCT) for relapsed B-cell lymphomas remains to be identifi ed. Several studies have shown that Rituximab (R) is implicated in the pathophysiology of acute and chronic graft-versus-host disease (GVHD). Materials (or patients) and Methods: We analyzed 153 adult patients (pts) who received alloSCT for chemosensitive relapsed/ refractory B-cell lymphomas. All pts received the same preparative conditioning regimen consisting of thiotepa/cyclophosphamide± fl udarabine, and GVHD prophylaxis consisting in cyclosporine and mini-methotrexate. ATG was added to pts allografted from class I antigen mismatched sibling or unrelated donors (MUD). Eightytwo pts (group A, study group) received high-dose R (500 mg/m 2 on day -6) and were enrolled in a prospective multicenter study, whereas seventy-one consecutive pts (group B, control group) were transplanted without R in the conditioning. The two groups were not signifi cantly diff erent in terms of diagnosis (group A: n=48 indolent, n=34 aggressive; group B: n=32 indolent, n=39 aggressive; P=0.10), donor types ( Discussion: Rituximab administration in the setting of alloSCT is associated with reduced GVHD-related deaths and increased infection-related deaths, likely due to delayed B-cell immunereconstitution. Thus, as compared to historical controls, pre-transplant Rituximab fails to improve the outcome of alloSCT. Disclosure of Interest: None Declared. Introduction: Ocular manifestations occur in 25-45% of patients with chronic graft versus host disease (cGvHD) following allogeneic BMT. Treatment options are limited and often ineff ective. We describe a series of multiply treated patients, with severe, debilitating, refractory disease. All patients responded to treatment with autologous serum eye (ASE) drops. In some cases there was a remarkable restoration of vision related function and a dramatic improvement in Q.O.L. Materials (or patients) and Methods: Five patients were diagnosed as having severe ocular cGvHD following ophthalmic review which included schirmer's testing and slit lamp examination after fl uorescein staining to examine the tear meniscus and assess the degree of corneal surface ulceration. The validated ocular surface disease index (OSDI) questionnaire was used to assess functional debility. After processing, an autologous unit of blood provided a 3 month supply of individual daily ASE for each eye. Results: Patient characteristics and treatments are shown in table 1. All patients had already 'failed' systemic immunosuppression (IS) and multiple lines of ophthamological treatments. OSDI scores were uniformly high at the start of treatment and improved for all patients. Two patients noted improvement and discontinued ASE to be maintained on acetylcysteine drops. Three patients had dramatic functional improvement that signifi cantly improved their quality of life and remain well on ASE. This subjective improvement was confi rmed by ophthalmology assessment with marked improvement in the appearance of the ocular surface. No adverse reactions were described. Discussion: Ocular cGvHD manifests as dry, gritty, painful eyes and if severe can lead to corneal ulceration, perforation and visual loss. ASE have been found to have epitheliotrophic and antimicrobial, as well as lubricating properties, due to the presence of growth factors, fi bronectin and vitamins. It is these properties that may explain the benefi ts seen in patients who have had no signifi cant response to systemic IS, lubricants and ophthamological procedures. In our experience few treatments have made such an immense diff erence to patients quality of life and functional capabilities. Disclosure of Interest: None Declared. . The conventional treatment of acute GHVD (aGVHD) includes calcineurin inhibitors (namely cyclosporine A) associated to high dose methyl-prednisolone (namely 2 mg/Kg/day). Unresponsive patients may be salvaged with second-line therapy including immunosuppressive drugs (e.g. mycophenolate), polyclonal antibodies, monoclonal antibodies and anti-cytokine antibodies. Extracorporeal photoapheresis (ECP) is an immunomodulatory treatment for steroid-refractory acute and chronic graft versus host disease (a/cGVHD) and it has been proven to be eff ective both in acute and chronic GVHD. Materials (or patients) and Methods: From March 2007 to March 2013, 138 adult patients were addressed to allo-SCT at our Institution for haematological malignancies. Forty-three out of 138 patients (31%) developed aGVHD grade ≥2 according to Glucksberg grading and were treated with conventional approach (6-methyl prednisolone 2 mg/Kg/day associated with calcineurin inhibitor). Nineteen out of 43 patients (44%) achieved a complete response and 24/43 (56%) did not, according to the standard criteria. This report focus on this 24 patients with steroid-refractory grade II -IV aGVHD, 15 (62%) of whom were treated with ECP and 9 (38%) with other second-line immunosuppressive agents. Results: The clinical and the transplant characteristics of these two groups of patients were comparable. Interestingly, 93% and 67% of the patients had skin involvement in the ECP and non ECP group, respectively. The liver and the gut were involved in 20%, 13% and 56%, 88% in the ECP and non ECP group, respectively.

The median time from aGVHD onset to ECP initiation was 20 days (range 11-41). Overall, 13/15 patients (87%) addressed to ECP and 6/9 (67%) of those treated without ECP showed a complete response according to the standard criteria. The median time to achieve a response with ECP was 51 days (range 3-117) and the median of ECP cycles was 20 (range 8-50) for a median number of 117 days (22-271) under ECP. At last follow up 7/15 (47%) patients treated with ECP are alive (two cases with limited cGVHD) and 2/9 (22%) patients not addressed to ECP are alive (both with extended cGVHD). The projected EFS at 3 years is 30% (95% CI 0-61) for patients who received ECP and 22% (95% CI 0-49) for patients who received treatments other than ECP (P=0.28). Discussion: Our preliminary results show that the response rate in patients treated with ECP is higher, although not signifi cantly, than that of patients not addressed to ECP (87% vs 67% Discussion: Our data underline the clinical signifi cance of danger signals for HSCT. Surprisingly, we found that low uric acid levels at the time of HSCT were associated with increased incidence of acute GVHD. This association has to be confi rmed in independent cohorts and mechanisms behind it remain speculative at this point. Currently, the EBMT GVHD working party is planning a retrospective survey on the association of uric acid and GVHD using the EBMT database. Hopefully this study will add more detailed information on the association between uric acid and GVHD. Disclosure of Interest: None Declared. Introduction: Intestinal GvHD as a severe complication in patients undergoing allogenic stem cell transplantation (ASCT) seems to be aff ected by gastrointestinal (GI) microbiota. Analysis of stool microbiomes in patients after ASCT revealed al loss of bacterial diversity with predominance of enterococcal fl ora, which was associated with prophylactic and therapeutic use of antibiotics as well as GvHD.

We now extended our pilot study on a larger number of patients and asked, if enterococcal positivity infl uences GvHD and overall survival of patients with ASCT. Materials (or patients) and Methods: In a prospective cohort study of 77 patients undergoing ASCT between 1/2011 and 6/2013, stool specimens were collected weekly starting before conditioning until 4 weeks after ASCT. Presence of Enterococcus (E) faecium and/or E faecalis was monitored by PCR-analysis. During this period, we switched our strategy of antibiotic prophylaxis due to increasing microbial resistance: the fi rst 32 patients received the standard prophylaxis with ciprofl oxacin and metronidazol, whereas in the subsequent 45 patients rifaximin, a synthetic derivate of rifamycin with <1% GI absorption, was used at a dose of 200 mg twice daily. A clinical analysis of intestinal GvHD and documented infections as well as overall survival was performed, and risk factors for enterococcal positivity and gastrointestinal GvHD were analyzed. Results: On day 28, 24 patients (31%) remained negative for both enterococci, whereas 39 (50%) and 14 (19%) patients, respectively, tested positive for either or both enterococci. Enterococcal positivity was associated with severe GI GvHD (stage III-IV): 27.3% of patients positive for E faecalis developed severe GvHD as compared to 5.5% of patients negative for E faecalis (P=0.007).

In those testing positive for both germs, E faecalis and E faecium, this relation was 35.7% to 4.2% (P=0.008). One-year TRM analyses showed a poor outcome for patients with positivity for E faecalis (P=0.019) and both germs (P=0.007), respectively. Enterococcal positivity seemed to be aff ected by the type of antibiotic prophylaxis. Patients on rifaximin showed less frequently a shift to positivity for both germs compared to systemic decontamination (11.1% to 28.1%, P=0.02). For E faecalis only, a minor insignifi cant eff ect was observed (22.2% to 37.5%, P=0.14). This association was confi rmed in multivariate analysis, as merely the type of antibiotic prophylaxis (p 0.034), but neither systemic use of antibiotics nor other clinical risk factors correlated with enterococcal positivity. Between rifaximin and the standard prophylaxis we found no signifi cant diff erences regarding the incidence of fever of unknown origin (25% to 18.2%, P=0.47), bacteremia (9.1% to 8.3%, P=0.91) or sepsis (21.2% to 16.7%, P=0.61). Finally, rifaximin (P=0.015) and early stage of underlying disease (P=0.04) seemed to exert a protective eff ect against severe GvHD according to a multivariate evaluation. Discussion: In conclusion, positivity for E. faecalis alone or both, E faecalis and E faecium in stool specimens increases the risk of severe GI GvHD and poor overall survival after ASCT. The choice of antibiotic prophylaxis might contribute to enterococcal overgrowth. Enterococcal positivity at day 28 might be an indicator of loss of bacterial diversity or even directly contribute to GvHD associated damage; larger prospective clinical and preclinical studies are needed to clarify the exact interactions. Disclosure of Interest: None Declared. Introduction: Reg3a is a GI GVHD specifi c biomarker that is not elevated in patients with non-GVHD enteritis developing after engraftment. Plasma levels of Reg3a at the onset of GvHD have been shown to be highly prognostic. However, the kinetics of Reg3a post-transplant, are not known. To answer this question we serially measured Reg3a levels after SCT and correlated the results with GI GVHD and neutropenic enterocolitis. Materials (or patients) and Methods: Serum was cryopreserved weekly from admission until d42 and thereafter according to outpatient intervals (up to d560). A total of 86 patients were included. Reg3a serum levels were assessed by ELISA. Admission and peak levels between day 0-14, 21-42 and 90-360 were correlated with the development of GI GVHD and/or neutropenic enterocolitis. Diagnosis of intestinal GvHD was confi rmed by histology in all symptomatic patients. Results: Mean Reg3a levels were lowest at admission (62.0 + 12.6 ng/ml) and did not diff er according to later GVHD or neutropenic colitis. The mean peak Reg3a level between d0-14 was not statistically diff erent between patients who developed severe GI GvHD (stage 2 or higher) or not (272.9 ± 101 vs 126.7 ± 31.3 ng/ml, ns). However, neutropenic enterocolitis may have confounded this result. Mean peak Reg3a levels between d0-14 were highest in patients who developed both GI GvHD and neutropenic colitis (329.9±73.9 ng/ml) and lowest in patients who developed neither (63.2 ± 4.8 ng/ml), this diff erence was highly statistically signifi cant (P=0.000). Reg3a levels were similar in patients who developed enterocolitis but remained GI GVHD free or who developed GI GVHD without prior enterocolitis (187.8 ± 32.6 vs 145.8 ± 20.5 ng/ ml, ns, but in both cases were signifi cantly higher than for patients with neither complication. Mean peak Reg3a levels between d21-42 were signifi cantly higher in patients who developed GI GVHD (546.9 + 121.7 vs 239.8 + 65 ng/ml, P<0.001). Likewise, mean peak Reg3a levels after day 90 (measured in 14 patients who received DLI or after taper of immunosuppression) who developed late acute GVHD compared to those who did not (239.8 + 65 vs 59.7 + 25.1 ng/ml, P=0.03). Discussion: Our study demonstrates Reg3a release is associated with severe GI epithelial damage from both GVHD but also from neutropenic enterocolitis. We were able to confi rm that Reg3a as a biomarker of early GI GVHD as well as GI GVHD that occurs late post transplant after DLI. Our new fi nding shows that in contrast to non-GVHD enteritis developing after engraftment, during the early posttransplant period, neutropenic enterocolits causes additional and independent Reg3a release. Disclosure of Interest: None Declared. 2 Hematology, LUMC, Leiden, Netherlands, 3 Medical Department I, University Hospital Carl Gustav Carus, Dresden, Germany Introduction: Outcome analyses of studies in the fi eld of SCT are mainly focused on time of relapse and death, measured from the moment of transplantation. This approach does not take into account the application of post-SCT interventions, such as DLI. To assess the dynamic impact of post-transplant interventions and clinical events on eventual failures or successes, the methodology of multi-state models has been developed. In our current study we demonstrate how a multi-state model helps to gain insight into the events after SCT in a cohort of AML and MDS patients who were transplanted using a strategy including scheduled DLI. Materials (or patients) and Methods: 78 patients who underwent T-cell depleted allogeneic SCT for AML (n=69, 83% (very) poor risk) or MDS (n=9) in CR (85% CR1, 15% CR2) after myeloablative conditioning at the Leiden University Medical Center between January 2002 and June 2011 were included in this analysis. Patients were eligible for DLI in the absence of GvHD from 3 months after the transplantation onward. We constructed a multi-state model to analyse (1) the impact of DLI and GvHD on the failure probabilities of relapse and non-relapse mortality over time, and (2) the probability of treatment success, which we defined as the absence of disease and of GvHD requiring immuno-suppressive treatment (IS), over time (see Figure 1 ; states 1, 3, 4 and 6). To assess the impact of IS and DLI on subsequent failure, a dynamic prediction approach was chosen, which can be considered as an extension of landmark modelling, in which both starting and end time, and starting and end state are variable. For accurate and objective measurement of duration of severe GvHD, date of start of IS for GvHD treatment until cessation of IS was taken. Results: From ca. 8 months after SCT, a substantial proportion of patients (ca. 30%) remained in the DLI state, meaning they had received DLI, were still alive without relapse and did not require IS for GvHD after DLI. In general, only few patients required IS for GvHD after DLI. Both relapse and NRM were considerable forces of failure of comparable magnitude. Yet while relapse is the dominant cause of failure from the SCT state, its impact is reduced after GvHD and especially after DLI. The long-term estimated probability of treatment success from start was around 40%; if a patient survived the first months after SCT without subsequent events, this substantially improved the expected long-term outcome. A comparison between these patients and those having received DLI suggests that the latter's probability of treatment success is larger: approximately 80% at 60 months after SCT for a patient in the DLI state at 9 months. Discussion: Our multi-state model helps to get insight into the dynamics of post-SCT interventions and clinical events and gives summary measures of treatment success. This model helps to extract more information from observational data. When data of diff erent centers are analyzed by our model, a comparison of treatment strategies is possible. Disclosure of Interest: None Declared.

Introduction: Acute graft-versus-host disease (aGvHD) still remains one of the major causes of procedure-related morbidity and mortality after allogeneic hematopoietic stem cell transplantation (HSCT). Information on the outcome of paediatric patients experiencing this complication is limited. We, therefore, conducted a retrospective registry-based analysis on 4193 children who developed grade III-IV aGVHD and were reported to the EBMT registry. Introduction: Extracorporeal photochemotherapy (ECP) has become a recognized treatment for graft versus host disease (GVHD), but the optimal frequency and duration of treatment are yet to be established. For either acute (aGVHD) or chronic (cGVHD) disease, in the absence of controlled trials the most frequently applied schedule is two ECP sessions per week until maximum response, followed by tapering, tailored to the individual patient. The aims of this study were (1) to compare MNC collections performed with the off -line procedure with historical data from inline procedure in patients with GVHD to generate products for ECP, and (2) to evaluate the effi cacy of a reduced-intensity ECP regimen in a pilot study respect to that adopted by most centers. Materials (or patients) and Methods: In all, 15 patients with refractory GVHD (12 cGVHD [83% severe disease], 3 aGVHD [66% grade 2, 33% grade 3]) were enrolled on this prospective study. The schedule was as follows: one procedure every week for 6 weeks, one procedure every 2 weeks for 1.5 months, and then one procedure every month for 6 months or more until maximal response was achieved. Written informed consent was obtained from all patients and our institution's ethics committee approved the study. Apheresis procedures were performed with either the COBE Spectra (Terumo BCT, Lakewood, CO, USA; version 7.0) or Spectra Optia v9 (Terumo BCT) by processing two times the patient´s blood volume. The product was transferred to a UVA-permeable bag (UVA

S293 PIT, Med Tech Solutions Gmbh, Fürt, Germany) , added 5 mL (0.1 mg) of 8-methoxypsoralen (8-MOP) aqueous solution (S.A.L.F., Cenate Sotto, Italy), exposed to UVA irradiation (UVA PIT System), and then reinfused. Results: We analyzed 216 consecutive procedures (Cobe Spectra 100, Spectra Optia 116), having a mean white blood cell (WBC) and MNC dose of 128.2 and 122.1 x 10 6 /Kg, respectively. Compared with historical data obtained from adults using the in-line devices (Denney H, et al. 38th Annual Meeting of the European Group for Blood and Marrow Transplantation, 2012) , in our hands one treatment collection resulted in twice the number of WBCs, and most important, four times the number of MNCs. The median duration of treatment was 241 days, and the median number of ECP cycles was 14. The response rates were recorded according to the involved organ. The overall response rate in cGVHD was 100% (25% complete response [CR] ; 75% partial response [PR] ). The mean response rate in patients with cutaneous, mucosal and joint involvement was of 100%, and that of other organs (gastrointestinal, lung, liver) was of 50%. We observed excellent response rates in the three patients with aGVHD with an overall response rate of 100% CR (skin, gut and liver involvement). Discussion: Prospective clinical trials with ECP will enable recommendations to be made with regard to the number of cycles and cells to be infused. In our hands, one treatment collection with the off -line procedure was at least as eff ective as two treatment collections with the in-line system, in terms of MNC dose. This reduced-intensity ECP schedule with the off -line procedure is feasible and eff ective in the management of GVHD, results in a decrease in at least 50% the number of treatments that are usually performed, allows for substantial cost reduction, and is more convenient for the patient. Introduction: Treatment options for steroid refractory Graft versus host disease (SR-GVHD) after allogeneic stem cell transplantation are unsatisfactory. Extracorporeal photopheresis (ECP) has immunomodulating eff ect mainly on T cells and has been used for the treatment of SR-GVHD with encouraging results. Rituximab (Rtx) targeting B cells showed effi cacy in SR-GVHD. We present an analysis of ECP and Rtx combination targeting two pathogenetic GVHD treads T and B cells used in our center to treat SR-GVHD. The aim of study was o determine the safety and clinical benefi t of ECP and Rtx combination in SR-GVHD in terms of clinical response rate (RR) and site using the National Institutes of Health consensus criteria. To evaluate lymphocyte subpopulations as prognostic markers of response. Materials (or patients) and Methods: Patients with acute or chronic SR-GVHD, received ECP plus Rtx (off label use) as addon to steroids ± Cyclosporin A. 12 ECP cycles (each consisting of 2 procedures on 2 consecutive days) were given by weekly schedule followed by monthly ECP for up to 16 cycles or more. 1 g of Rtx was given after 1st and optionally 3rd ECP cycles. Evaluations were performed after 6, 12, 14 and 16 ECP cycles. Results: ECP+Rtx treatment was initiated for 60 patients with SR-GVHD. 48 patients who survived at least for 6 weeks were further analysed for clinical response. 25 patients had grade III-IV acute GVHD (aGVHD). 13 patients with chronic GVHD (cGVHD) and 10 with overlap syndrome (oGVHD) (10 -moderate, 13 -severe)

were analysed together. The median observation time was 10 months Results: As a whole, 36/72 (50.7%) of the patients developed monoclonal gammopathy after transplantation. A decreased relapse incidence (19% vs 40% at median +436 days) and an increased GvHD development (54% vs 34% at median +436 days) was observed in patients with an appearance of monoclonal gammopathy; no diff erence was observed regarding overall survival and post-transplantation mortality (not statistical signifi cance was reached). Comparing patients with or without monoclonal gammopathy at +90 and +180 days post transplant respectively, we have not detected any statistical diff erence in overall survival, GvHD development, relapse incidence and post-transplantation mortality. Vice versa, after 360 days from the transplant, among patients with a monoclonal gammopathy, was observed an increased rate of GvHD (50% vs 32%, P = 0.07). In those patients GvHD frequently seemed aspects of cGVHD. Discussion: Monoclonal gammopathy, also in our experience, is common after allo-BMT. The few papers published in the past, found this laboratory alteration more frequently associated with GVHD but without any long term adverse eff ect. Our data would seem to confi rm these conclusions. In the past the explanation for the evidence of a monoclonal gammopathy was correlated to an aberrant immune reconstitution after allo-BMT. Recently many evidence showed that B cells are involved in the pathogenesis of chronic GVHD (cGVHD) and anti-B-cell therapy has been proposed for the therapy mainly of cGVHD. We speculate that the presence of a monoclonal gammopathy after allogeneic transplant is expression of the activation of the B-compartment. A prospective study with a larger population should be considered, in order to confi rm our results and assay post-transplantation monoclonal gammopathy as an early marker for GvHD development. Disclosure of Interest: None Declared. Introduction: Extracorporeal photopheresis is increasingly used to treat chronic GVHD (cGVHD). While its limited toxicity is well established and accepted, data on effi cacy are still rare und most reports so far included less than 50 patients. While the optimal schedule for ECP is still unknown, most centers use a back-to-back treatment, initially repeated weekly or every two weeks. Materials (or patients) and Methods: In this single center analysis we retrospectively evaluated treatment success in 61 adult patients who received ECP for steroid refractory or relapsed cGVHD at our center. Treatment success was defi ned as "patient being alive at 6 months after initiation of ECP, having achieved a partial (PR) or complete response (CR) of cGVHD according to NIH criteria and having no relapse of underlying disease". ECP was performed once weekly for the fi rst 4 weeks, thereafter once every two weeks for responding patients. 3 Hematology, UMCU, Utrecht, 4 Hematology, VUMC, Amsterdam, Netherlands Introduction: About 60% of all patients receiving an allo-SCT and surviving beyond day 100 develop cGvHD, assuring high morbidity within this patient group. For this reason, chronic GvHD is associated with a substantial impairment in quality of life and loss of employment in long-term allo-SCT survivors. For physicians, cGvHD poses a big challenge to treat adequately and satisfactory.

The current treatment relies heavily on corticosteroids, however refractoriness is a big problem as well as the unwanted eff ects of longterm corticosteroid use. Therefore, new therapies are urgently needed. Earlier studies by others and us have demonstrated amelioration of chronic GvHD symptoms by B-cell depletion. In addition inhibition of the PDGF pathway by tyrosine kinase inhibitors also provided provocative data. However, complete resolution of cGVHD is usually not reported with single drug therapies and cGvHD with ulcera has so far been refractory to most experimental therapies. Materials (or patients) and Methods: Hypothesis: We aim to test whether the sequential therapy of the anti-CD20 antibody Rituximab followed by a 6 month period of treatment with the tyrosine kinase inhibitor Nilotinib can further improve response rates. Methods: 30 patients are treated with a combination of 4 weekly infusions of the anti-CD20 antibody Rituximab followed by a 6 month period of treatment with the tyrosine kinase inhibitor Nilotinib. Patients have been evaluated monthly. Results: 4 patients have completed the study period whilst 7 patients are still being treated. Three out of four patients who completed the study showed a partial response. Two patients showed complete resolution of ulcerative skin lesions. Response rates for the remaining patients will be available shortly. Discussion: The sequential therapy of B-cell depletion and tyrosine kinase inhibition might show for the fi rst time a complete resolution of ulcerative cGvHD and provides a new and interesting alternative treatment option for this diffi cult patient category. Disclosure of Interest: None Declared. Introduction: Graft-versus-host disease (GVHD) is a serious complication after donor lymphocytes infusions (DLI), as well as part of an immunological response fl owing parallel eff ect-graft-versus leukemia. The presence of acute or chronic GVHD is associated with a lower risk of relapse after allogeneic stem cell transplantation (allo-SCT).

Aim of this study was to investigate the frequency of GVHD depending on the scheme DLI, and the impact of GVHD on the results adoptive immunotherapy in patients with relapsed hematological malignancies after allo-SCT.

Materials (or patients) and Methods: The study comprised 44 patients (pts) with relapsed hematological malignancies after allo-SCT from HLA-matched related (n = 42) and unrelated (n = 2) donors. The median age was 30 years (16 -60 years), male -29, female -15. Relapse was diagnosed in 9 months (range from 2 to 51 months) after allo-BMT. DLI were performed in two ways. The fi rst was DLI alone (n = 15) and the second -DLI after chemotherapy (during aplasia or after reconstitution of hematopoiesis) (n = 29). The number of DLI range from 1 to 8 (median 3 DLI) per patient. Total amount of the CD3+ cells varied from 1 to 76x10 7 CD3+cells/kg (median 20,5 x10 7 CD3+cells/kg). All pts received 2 -6 MUE Interleukin-2 (IL-2) subsequently after DLI. Chimerism was monitored by PCR ana lysis (VTTR and STR), and by FISH -analysis for centromers of X and Y -chromosomes. Results: Remission with full donor chimerism was achieved in 25 (57%) of 44 pts. The median duration of remission with full donor chimerism was 8 months (2 to 162 months). Acute GVHD after DLI was diagnosed in 16 pts: I-II grade -5 (31%); III-IV grade -11 (69%). Chronic GVHD after DLI was diagnosed in 13 pts (30%): limited -8 pts (62%), extensive -5 pts (38%). Acute and chronic GVHD were associated with full donor chimerism (88% and 92%, respectively). Chronic GVHD after DLI signifi cantly infl uenced on the overall survival. The 5-year overall survival in pts after DLI with chronic GVHD was 45% and without chronic GVHD -17% (P= 0,01). The eff ectiveness of treatment is two times higher after DLI with chemotherapy (33% and 69%, respectively) (P = 0.03). The frequency of acute and chronic GVHD was 2 and in 1.5 times higher, respectively. Discussion: DLI after chemotherapy is eff ective in patients with relapsed hematological malignancies after allo-SCT. GVHD after DLI is associated with longer overall survival (P = 0.01). Disclosure of Interest: None Declared. Introduction: Immune reactions are common complications of allogeneic hematopoietic stem cell transplantation (HSCT) and include: graft-versus-host disease (GvHD, acute-50%, chronic 40-70%), graft rejection-12%. To explore the pathogenesis of immune complications after allogeneic HSCT, we carried out genome wide transcriptional profi ling of RNA extracted from peripheral blood mononuclear cells collected before and after transplantation followed by pathway enrichment analysis in an attempt to identify biological pathways that were preferentially associated with immune reactions induced by the procedure. Materials (or patients) and Methods: Study group composed of 44 patients (pts.) referred to ) years old, 31 boys and 13 girls. Children were diagnosed with neoplastic-32 (73%) and non-neoplastic-12 (27%) diseases. Blood was collected before transplantation and 7 months in average after transplantation. The second collection was performed in 27 pts. 2.8-19.5 (average 10.4) years old, 20 boys and 7 girls. All patients with two collections were treated with allogeneic transplantations according to EBMT protocols including: match unrelated donor-60%, match sibling donor-33% and match family donor-7% transplantations. The mRNA samples isolated from peripheral blood mononuclear cells were evaluated for gene expression with the use of GeneChip® Human Gene 1.0 ST microarrays. DAVID annotation tools (DAVID Bioinformatics Resources 6.7) were used to explore which predefi ned gene sets were signifi cantly enriched before HSCT compared to after HSCT. The KEGG (Kyoto Encyclopedia of Genes and Genomes; www.genome.jp/kegg) was selected for analysis. A list of 250 genes diff erentially expressed between before and after-HSCT samples with P-values adjusted for multiple testing below 0.01 and with fold change greater than 1.5 was used as input for pathway enrichment analysis in DAVID.

Results: There were no signs and symptoms of the primary disease in patients at the second collection. Graft versus host disease was diagnosed in 52% of pts.(acute-33%, chronic-15%, overlap syndrome-4%). Organ localizations were as follows: skin-50%, gut-21%, skin/gutt-7%, skin/liver-15%, skin-gut/liver-7%. No patients presented with graft rejection. Four genes pathways were detected: "Allograft rejection", "Graft-versus-host disease", "Autoimmune thyroid disease" and "Type I diabetes mellitus" all connected with immune reactions after transplantation. The relationships between genes in the pathways were signifi cant (P<0.05). Discussion: Allogeneic transplantations by defi nition induce immune reactions with varied intensity. As mononuclear cells are involved in the process, activation of gene pathways regulating immune response should be visible. Graft-versus-host disease is the leading cause of morbidity and mortality after allogeneic hematopoietic stem cell transplantation (HSCT). So far, there is no validated diagnostic or predictive blood biomarker for GvHD which could improve diagnosis and prognosis and help guide therapeutic interventions. Our study confi rm activation of genes responsible for aggression of donor/recipient immune system against donor/recipient cells. Further study will explain if whole genome expression profi le may serve as a biomarker for immune reactions observed after stem cell transplantation in children. Introduction: Infusion of mature donor T cells with or early after allogeneic stem cell transplantation (alloSCT) is associated with a high risk for the induction of graft versus host disease (GvHD) in diff erent organs, including the skin. Postponed application of donor T cells at 6 months after T cell depleted alloSCT signifi cantly reduced the risk for coinciding acute GvHD. In this study we investigated the eff ect of pre-transplant conditioning on the numbers and phenotype of antigen presenting cells (APCs) and T cells in the skin. Materials (or patients) and Methods: Skin biopsies were routinely taken after informed consent of the patients at 0, 3, 6, 12 and 26 weeks after alloSCT at the site of the bone marrow aspiration from patients who did not display signs of GvHD. Skin biopsies taken at similar time-points from patients who underwent an autologous transplantation (autoSCT) served as autologous controls and skin biopsies taken from healthy individuals served as normal controls. Diagnostic biopsies taken from aff ected skin (GvHD/dermatitis) were analyzed in parallel. Biopsies were cryopreserved and stored. Cryo-sections were stained with antibodies for HLA-class II, HLA-DP, HLA-DR, HLA-DQ, CD3, CD8, CD68, CD54, and CD40, counterstained with fl uorescently labeled secondary antibodies and analyzed by fl uorescent microscopy. In biopsies taken from sex-mismatched donor/patient pairs we assessed the origin of the specifi c cell types by combining the surface immunestaining with XY FISH. Results: At the moment of transplantation directly following the conditioning regimen no signifi cant changes were observed compared to normal skin. At 3 and 6 weeks after the transplantation we observed an increased number of HLA class II positive macrophage-like cells in the dermal region of the skin. In patients without GVHD, this was not associated with a signifi cant increase in the number of local T cells. No overt diff erences associated with the type of pre-transplant conditioning (myelo ablative (MA) versus reduced intensity (RIC)) were observed. Both the HLA class II positive cells and the local T cells were found to be of patient origin at these time points. We observed a remarkable further increase in the number of HLA class II positive cells in the biopsies taken at 3 months after transplantation, which was still pronounced at 6 months after transplantation in both the allo-SCT and autoSCT biopsies. This increase was most pronounced in patients who received MA conditioning, even in the absence of any sign of GvHD. The origin of the HLA-class II positive cells and the local T cells diff ered between patients. We observed complete patient and complete donor origin, as well as mixed origin.

In contrast to the HLA class II positive cells recruited in aff ected skin from patients with GvHD or dermatitis, the HLA class II positive cells in the routine biopsies taken after alloSCT and autoSCT did not express dendritic cell like markers, including adhesion and costimulatory molecules. Discussion: Pre-transplant conditioning results in overt recruitment of HLA class II expressing macrophage-like cells after transplantation. Despite their relatively unprofessional APC phenotype, these cells may amplify the GvHD response resulting in the local formation of a pro-infl ammatory milieu. The origin of the class II positive cells and the patrolling T cells will dictate the chance of local GvHD induction. Disclosure of Interest: None Declared. Introduction: Our group has shown that severe hypoalbuminemia (‹3.0 g/dL) at day +90 post allografting is an adverse prognostic factor for non-relapse mortality and overall survival in patients with AML and MDS. However, the chronological relationship between development of severe hypoalbuminemia and acute GVHD onset and grading is not clearly understood. Materials (or patients) and Methods: We analyzed 1066 patients who received an allogeneic HCT at the H. Lee Moffi tt Cancer Center between January 01, 2005 and March 31, 2013. Serum albumin levels were obtained for all subjects from day of stem cell infusion (day 0) until day +100 post-transplantation whenever available.

The association between occurrence of severe hypoalbuminemia (defi ned as a serum albumin level ‹ 3.0 g/dL) and onset of acute GVHD and its severity grading was assessed using time-to-event analysis where time to follow up was defi ned as time from day 0 to the fi rst serum albumin level below 3.0 g/dL. Diff erences in timeto-event according to the severity of acute GVHD were assessed using the Cox proportional hazard model. The signifi cance level was set at ≤ 0.05. All analyses were performed using Stata statistical analysis software version 13. Results: The median age for all subjects was 52 (18-76) years. The most common preparative regimen consisted of fl udarabine and pharmacokinetically-targeted intravenous busulfan (FLU-BU) in 72% of cases. Acute GVHD prophylaxis consisted of a calcineurin-inhibitor in combination with methotrexate (53%) or sirolimus (25%), or mycophenolate mofetil (22%). These and other patient characteristics are shown (Table 1) . Median time to ANC engraftment was 16 days. Time-to-event analysis demonstrated severe hypoalbuminemia by day +20 in 804 (75%) subjects.

Severe hypoalbuminemia was not associated with a diff erence in the cumulative incidence of acute GVHD (all grades) whether it occurred before day +20 [OR=0.93 (95%CI=0.77, 1.11 *includes pharmacokinetically-targeted IV busulfan with a median daily AUC ranging from 3500 to 9000 μmoles min/L. Discussion: We demonstrate a chronological association between occurrence of severe hypoalbuminemia before day +20 post-allogeneic HCT and time of acute GVHD onset. Moreover, patients who develop severe hypoalbuminemia have a higher likelihood of developing a higher acute GVHD grade. Disclosure of Interest: None Declared.

E. Buces 1, 2, * , 3 , 3 , V. Guillem 6 , J. Nieto 7 , M. González 8 , R. de la Cámara 9 , S. Brunet 10 , I. Espigado 11 , C. Vallejo 12 , A. Sampol 13 , D. Serrano 2, 3 , M. Kwon 2, 3 , J. Gayoso 2, 3 , P. Balsalobre 2, 3 , C. Solano 14 , D. Gallardo 16 , 3 , I. Buno 2, 3 and on behalf of the GvHD/Immunotherapy committee of the Spanish Group for Hematopoietic Transplantation (GETH) 1 Department of Hematology, HGU G.Marañón, 2 IIS G.Marañón, 3 HGU G.Marañón, Madrid, 4 ICO, H.J.Trueta, Girona, 5 H.Clinic, IDIBAPS, Barcelona, 6 H.Clínico, Valencia, 7 HU M.Meseguer, Murcia, 8 HU, Salamanca, 9 HU La Princesa, Madrid, 10 HSC Sant Pau, Barcelona, 11 HU V.Rocío, Sevilla, 12 HC Asturias, Oviedo, 13 HU S.Espases, Mallorca, 14 HCU, Valencia, 15 HRU, Málaga, 16 ICO, HJ Trueta, Girona, Spain Introduction: The interleukin 1 (IL1) cytokine family includes two major pro-infl ammatory molecules, IL1A and IL1B, which play a key regulatory role in initiating and modulating immune responses. Polymorphisms in IL1 genes have been shown to be associated with plasma levels of these cytokines and have been associated with the development of complications after SCT. However, reported data are inconsistent as a result of the heterogeneity of the transplant cohorts or the small number of subjects. The objective of the present study was to evaluate the impact of IL1 SNPs on the development of complications and survival in a large cohort of HLA-identical sibling allogeneic stem cell transplant patients. Materials (or patients) and Methods: The study includes 509 patients with hematological malignancies and their donors (1018 individuals) from the DNA Bank of the Spanish Group for Haematopoietic Stem Cell Transplantation (GETH; Table 1 ). Four SNPs were genotyped, one in the promoter of IL1A gene (rs763780), 2 in the promoter of IL1B (rs16944, rs1143627) and one in IL1B exon 5 (rs1143634), by mass spectrometry (MALDI-TOF, Sequenom MassARRAY; CEGEN). Results: The association between genotypes and complications after allo-SCT after univariate analysis (Table 2 ) are as follows. Acute GVHD: patients harbouring the minor allele C for rs1143627 showed a higher incidence of grade II-IV and III-IV aGVHD. Patients with genotype GG for the SNP rs16944 showed a lower incidence of grade II-IV and III-IV aGVHD. Chronic GVHD: Donors and recipients with genotype CC for rs1143634 or allele C for rs1800587 associate with a lower risk of cGVHD. In addition, donors with allele C for rs1143634 associate with a lower risk of extensive cGVHD. Relapse: A higher risk of relapse was observed in patients harbouring genotype CC for rs1800587 or genotype CC for rs1143634 as well as in patients transplanted from donors CC for rs1143627. On the other hand, donors with the major allele G for rs16944 associate with a lower incidence of relapse. All polymorphisms in the IL1 gene remained as independent risk factors in the multivariate analysis (Table 3) , except for the association of patients with allele C for rs1800587 and the development of chronic GVHD and that of donors with allele C for rs1143634 and the risk of extensive chronic GVHD. Discussion: This results further support the idea of a genetic predisposition to certain complications after allo-SCT. IL-1A and IL-1B SNP genotyping might be useful to anticipate complications after sibling HLA-identical allo-SCT and, therefore, to improve the clinical management of transplanted patients. Disclosure of Interest: None Declared.

Introduction: Vitamin D has emerged as a central player in the immune system, with defi ciency being implicated in the pathogenesis of several autoimmune diseases, including chronic graftversus-disease (cGvHD). This study aimed to evaluate possible associations of vitamin D defi ciency prior to allogeneic stem cell transplantation (ASCT) with complications after ASCT, with cGvHD of grade moderate to severe as the main endpoint. Materials (or patients) and Methods: This is a retrospective cohort analysis on 162 patients who have undergone ASCT at Karolinska University Hospital, Huddinge, between 2005 and 2011 , with stored serum samples available. Children under the age of 12 and cord blood transplantations were excluded. A total of 137 patients were eligible for evaluation regarding cGvHD after excluding patients with graft rejection or with a survival of less than 100 days after transplantation. Patient data were retrospectively collected from clinical patient fi les. For chronic GvHD diagnosis and scoring the NIH consensus criteria were used and both classical chronic GvHD and overlap syndrome were included, but not late onset acute GvHD. Rejection was defi ned as lack of engraftment, engraftment with recipient cells or later developing full (>95%) recipient chimerism in the absence of relapse of the underlying disease. Vitamin D was analyzed as 25-OH-cholecalciferol from cryopreserved serum samples using a chemiluminescence method (CLIA), performed by the laboratory for clinical chemistry at Karolinska University Hospital, Solna, Results: Median level of vitamin D before transplantation was 42 nmol/l (range 10-118), hence below the level of insuffi ciency (50 nmol/l).

For chronic GvHD grade moderate to severe, the cumulative incidence was 56% when vitamin D level prior to transplantation was below 25 nmol/l, 34% when between 25 and 50 nmol/l and 26% when vitamin D was above 50 nmol/l. This was confi rmed in a multivariate analysis including vitamin D level, age, conditioning, anti-T-treatment and sex mismatch (female donor to male recipient) as covariates with death as competing risk, identifying vitamin D level prior to transplant as a signifi cant independent risk factor for development of cGvHD (P=0.007). Incidence of acute GvHD grade II-IV showed no correlation to vitamin D level. There was a trend towards lower overall survival in the vitamin D-defi cient patients, which was signifi cant at the median vitamin D level in the cohort (42 nmol/l, P=0.03). Patients with graft rejection tended to have lower vitamin Dlevels. As 11 out of 12 cases of rejection were in patients with reduced intensity (RIC) conditioning, we chose this subgroup for a post-hoc analysis. This showed a median vitamin D level of 34 nmol/l in patients with rejection as compared to 44 in the others (P =0.017). Discussion: In this study, we confi rm the previous fi ndings by Glotzbecker et al (2012) of an association between low levels of vitamin D before transplantation and increased incidence of chronic GvHD. An unexpected fi nding was the indication of an association between low levels of vitamin D and graft rejection. These results must be interpreted with caution as rejection was not a predefi ned endpoint and the eff ect was in the RIC subgroup, but are interesting. These fi ndings could be highly relevant for the care of ASCT patients worldwide, and prospective, randomized studies on the eff ect of vitamin D supplementation are therefore needed. Disclosure of Interest: None Declared. 

Introduction: Previous studies have shown that germ-free or completely decontaminated rodents showed a decrease in aGVHD event in recipients of MHC mismatched blood marrow transplants. Oral metronidazole (MTZ) reduced the incidence of aGvHD in matched sibling donor in allo-HSCT. We analyze if the use of oral metronidazole with quinolones vs quinolones alone decreases the incidence of aGvHD in unrelated donor (UD) allo-HSCT. Materials (or patients) and Methods: We collected a total of 123 consecutive UD Allo-HSCT from 2001 to 2012 submitted to a HSCT in our center. Our study compared two diff erent regimens of intestinal bacterial decontamination in order to investigate the incidence of aGVHD and were classifi ed into two groups; Arm A: oral metronidazole (500 mg tid) and quinolones (ciprofl oxacin 500 mg td/levofl oxacin 500mg od) (n=65) and Arm B: quinolones alone (n = 58). Both groups were similar in age, sex, underlying disease and aGVHD prophylaxis. There were diff erences in stem cell source and conditioning regimen. In group A (with MTZ), 22 patients received cord blood (cb), 19 bone marrow (bm) and 24 peripheral blood (pb). In group B, 11 patients received cb, 29 bm and 18 pb (P = 0.044). The primary objective was to evaluate the incidence of aGVHD (grades 0-I vs II-IV) between groups. Secondly we analyze overall survival and transplant-related mortality at +100 days (TRM100). The data analysis was performed with SPSS 17.1 software, using the Kaplan-Meier tests, LonRank test, T-student and Chi-squared test. Results: the incidence of aGVHD was 51% (n=46) for group A and 49% (n=45) for group B with no statistical diff erences between groups. When we analyzed the incidence of aGVHD grades 0-I vs II-IV we didn´t fi nd statistical diff erences (A=51% Vs B=36% and A=49% vs B=64%) (P=0.2). TRM +100 was 14% in the global serie (n = 14), with no statistical diff erences between groups (P=0.1). The most frequent causes of death at +100 days were infection (n=7) and aGVHD (n=2). Discussion: In our experience the use of oral metronidazole do not decrease the incidence of aGVHD in unrelated donor Allo-SCT. Disclosure of Interest: None Declared.

Introduction: G-CSF stimulated peripheral blood (G-PB) as a donor source has a higher frequency of chronic GVHD (cGVHD). Recently the Canadian BMT Group (CBMTG) performed a prospective Phase III clinical trial entitled "A Randomized Multicentre Study Comparing G-CSF Mobilized Peripheral Blood and G-CSF Stimulated Bone Marrow in Patients Undergoing Matched Sibling Transplantation for Hematologic Malignancies (CBMTG 0601)". The donor grafts were evaluated for immune cell populations that may predict the onset of cGVHD. Materials (or patients) and Methods: This study evaluated donors enrolled on CBMTG0601 for graft lymphocyte populations. Proportions of specifi c lymphocytes in donor grafts were associated with cGVHD in the recipient using logistic regression. Associations were signifi cant if the corresponding P-value was less than 0.05. Recipients were considered cGVHD positive if diagnosed with extensive cGVHD and cGVHD negative if free of extensive cGVHD for a minimum of 12 months post transplant. Relapsed recipients and those who died without previous diagnosis of GVHD were excluded from analyses. Cell phenotypes and cytokine production of lymphocytes in the donor grafts were analyzed by fl ow cytometry in a set of panels that enabled identifi cation of distinct NK, T and B cell subsets. Results: Analyses of cGVHDvs. cGVHD + showed the following associations: higher proportions of IFNg-producing T helper cells and CD56 bright NK cells in donor grafts were associated with a lack of cGVHD (P<0.005 and P<0.05, respectively). On a multinomial logistic analysis, there was a higher graft proportion of immature B cells (CD19 + IgD -CD27 -), a lower proportion of memory B cells (CD19 + CD27 + ), and higher proportion of TLR9+ B cells (P <0.01) with development of de novo cGVHD. Evaluation for diff erences between G-BM and G-PB showed that there were signifi cant diff erence between both memory B cells (P = 0.002) and immature B cells (P <0.0001). Discussion: We have identifi ed two population, IFNg-producing CD4 + T cell and CD56 bright NK cells that are associated with tolerance after both G-BM and G-PB transplantation. We found three donor B cell populations, including immature and memory B cell populations associated with development of de novo cGVHD in donor grafts that impact the development of cGVHD. These fi ndings require confi rmation in a larger population but suggest that donor B cell populations rather than T cell populations may have the strongest impact on the high frequency of cGVHD after G-PB transplantation. Disclosure of Interest: None Declared. Introduction: Keratoconjunctivitis sicca syndrome (KCS) due to chronic GvHD (cGvHD) is responsible for major alteration in quality of life of patients undergoing allogeneic stem cell transplantation (allo-CST). The conjunctival fi brosis secondary to infl ammation is often slow and subtle and is responsible for impaired corneal and conjunctival epithelial surfaces potentiated by tear quantitative and qualitative defi ciency. Treatment of KCS remains disappointing; variable success in the correction of the conjunctival dryness has been obtained with variety of topical treatments with or without systemic immunosuppressive agents. These treatments, though, do not seem to have any eff ect on sicca symptoms and patients' quality of life. Sclero-corneal lenses bring a valid therapeutic option by creating a pre-corneal reservoir of tears allowing permanent lubrication of the epithelium, the protection of the corneal surface against the eyelid and ciliary mechanical friction and against environmental stresses and the creation of a uniform refractive surface to be taken into optimum optical load and stable visual acuity. Materials (or patients) and Methods: We describe the safety and effi cacy of Sclero-corneal lenses in a retrospective analysis of 7 patients with KCS due to cGvHD following allo-CST. Table 1 summarizes patients' and GvHD characteristics. All patients had superfi cial punctate keratitis refractory to standard treatments. Evaluation of patients was carried out by the same ophthalmologist and hematologist. cGVHD was recorded according standard criteria. Ocular surface disease index (OSDI) and Oxford score were used to evaluate ocular symptoms. The scale of "Monoyer" was used and converted into visual acuity "LOG MAR" for comparative purposes of the study. Results: All patients but one agreed to hold the lenses. The 6 remaining patients were equipped with lenses type ICD (LCS Company). With a median follow-up of 6 months (4-13) all patients have experienced an improvement in their quality of life with a clear improvement of dry-eye symptoms. This quality of life is also improved by decreasing the frequency of eye-drop instillation and the attenuation of the discomfort and post-instillation visual fl uctuation. We observed 100% improvement in OSDI score with an average improvement of 67.08 points, 100% improvement or stability of visual acuity with an average gain of 0.23 LOG MAR acuity and 100% improvement or stability of the Oxford score with a mean gain of 1,917 points. Discussion: Despite its limited size, this cohort of patients treated with sclero-corneal lenses is promising. Whenever possible, this approach should be considered in patients experiencing KCS due to cGVHD. Disclosure of Interest: None Declared. Introduction: Chronic graft-versus-host disease (cGvHD) as a major complication after allogeneic stem cell transplantation (SCT) is associated with late morbidity, mortality and inferior quality of life. In 2005. the National Institute of Health (NIH) proposed new criteria for the diagnosis and classifi cation of cGvHD. Overall data suggested that the proposed NIH classifi cation is prognostically useful, but the impact of both the syndrome and global severity grading on outcome of the SCT is still unknown. We evaluated the association of the proposed NIH classifi cation (syndrome classification and the global severity) with transplantation outcomes. Materials (or patients) and Methods: We retrospectively analyzed 174 adult patients (pts), average age 29 years, F/M ratio 68/106, with diff erent hematological malignancies who survived beyond 100 days following allo-SCT. All pts have HLA identical sibling donor. Peripheral blood was source of stem cells (SC) in 99 pts, main conditioning regimen was myeloablative and GvHD prophylaxis was combination of Cyclosporine A with Methotrexate. cGvHD was retrospectively classifi ed with NIH proposed classifi cation and was correlated with overall survival (OS) and non relapse mortality (NRM) after SCT. Results: From our cohort of pts, 114 (65%) have cGvHD with late onset in 12 (10,5%), classic in 46 (40,3%) and overlap syndrome in 56 (49,2%). According to onset, 16 (19,6%) pts have progressive, 28 (29,6%) de novo and 58 (59,8%) quiscent type of cGvHD. Median time of onset of cGvHD was 158 days (range 100-473). Skin, mouth, eyes, gastrointestinal tract and liver were aff ected mostly. Considering of global severity scoring system, cGvHD was mild in 27 (23,8%), moderate 49 (42,9%) or severe in 38 (33,3%) pts. Pts with progressive type of onset and with severe form of cGvHD had signifi cantly worst OS in comparison to other subgroups (log rank P < 0.0001, P= 0.009, respectively). Pts with classic cGvHD had signifi cantly better OS in comparison to pts with late acute and overlap syndrome (log rank P= 0.009). NRM was signifi cantly higher in pts with progressive and severe cGvHD (log rank P< 0.0001, P<0.0001, respectively).

Discussion: cGvHD have impact on global outcomes following SCT. NIH proposed classifi cation of cGvHD (syndrome classifi cation and the global severity) is useful and predictive, but further investigations in homologous subgroups of pts (according to type of the diseases, SC sources, conditioning regimen and donor) are needed. Disclosure of Interest: None Declared. Introduction: Recent studies demonstrate that polymorphisms in non-HLA genes, including cytokine genes, innate immunity genes, are significantly associated with the risk of graft versus host disease (GVHD). Now microRNAs have been demonstrated to play important roles in GVHD pathogenesis. So we hypothesize that microRNA gene polymorphisms may contribute to transplantation outcome by controlling different essential gene regulatory modules in posttransplantation immune reconstruction. Materials (or patients) and Methods: The cohort was consisted of 74 pairs of recipients and their unrelated donors who underwent allo-HSCT in our single center from January 2008 to December 2010. Recipient genomic DNA was extracted from bone marrow of recipients before transplantation. Donor genetic DNA was extracted from recipient bone marrow 100% donor chimerism by short tandem repeat(STR) after transplantation. Eleven Single-nucleotide polymorphism were detected by multiplex SNaPshot technology,including mir-101(rs2549855), mir-21(rs175 39020,rs9684042), mir-92a(rs2716830), mir-164a(rs2910164), mir-142(rs11934028), mir-125b(rs10503344), mir-142(rs11934028), mir-19b(rs67308836) . Microrna genotypes were analyzed by the Fisher exact test. The cumulative incidence of aGVHD, II-IVaGVHD, cGVHD, relapse, transplantation related mortality(TRM), and overall survival(OS) were estimated using the Kaplan-Meier method. The Cox proportional hazard model was applied to multivariate analysis of outcomes of transplantation. All probability values were two-sided and P values less than 0.05 were considered statistically signifi cant. P values between 0.05 and 0.1 were considered to be indicative of a trend. Results: Our results show that presence of rs11934028AA genotype in mir-142-3p of the recipient (AA genotype 100% vs. AC/ CC genotypes 55.3.P=0.012) and the donor (AA genotype 100% vs. AC/CC genotypes 55.3%, P=0.001) were associated with aGVHD than other genotypes. Besides, the presence of recipent rs11934028 AA genotype was associated with higher incidence of II-IV aGVHD(AA genotype 100%versus AC/CC genotypes 44.9% P=0.08). Interestingly, the presence of donor rs10503344TT genotype in mir-125b reduce the incidence of aGVHD and II-IV aGVHD statistically signifi cant in the study. Multivariate Cox regression analysis was carried out to sort out risk factors of aGVHD and II-IVaGVHD (Table.1 ). The results confi rmed that donor11934028 AA genotype was an independent risk factor for aGVHD and donor rs10503344 genotype was a protective risk factor (Table 1) . No association was observed between mirSNPs and cGVHD, TRM, relapse and OS. Discussion: The present study fi rst demonstrate that mirSNPs have an infl uence on the development of aGVHD and higher grades aGVHD. It suggests that mirSNPS are potential to be the novel biomarker for prevention and trentment of aGVHD. This is instructive and meaningful to the choose of donor and early prevention of aGVHD. Studies remains to be carried out in vitro and in vivo to further explore the molecular mechanism how mirSNPs infl uence the outcome in the posttransplantation immunity reconstruction. Disclosure of Interest: None Declared. Results: Donor T cell proliferation in response to Allo-DCs were suppressed by auto-NK cells. Donor NK cells killed proliferating T cells(CFSE low ) more efficiently than nonproliferating T cells (CFSE high ) at all E:T ratios tested. Donor NK cells degranulated in response to activated, but not resting T cells. CD56 dim degranulated more than CD56 bright NKcells, and NKG2A-degranulated more than NKG2A+ NK cells. Activated T cells express high levels of the NKG2D ligand s(MIC-A, MIC-B and ULBP-1) and C226 ligand PVR. Blocking of LFA-1,NKG2D or DNAM-1 led to significant reduction of NK cell cytotoxicity, whereas blocking of NKG2A and TIM-3 resulted in an increase in NK cytotoxicity. In the early period after HSCT, reconstituted NK cells were mainly CD56 bright and NKG2A + subgroup. NK cells in GVHD patients expressed lower level of NKG2D and CD226 when compared with NK cells in non-GVHD recipients or their corresponding donors. In GVHD patients, both degranulation and cytotoxicity of NK cells towards activated auto-T cells were depressed. NK cells in GVHD up-regulated the expression of CD226 and NKG2D when stimulated ex vivo with IL-15, but not IL-2. Both IL-2 and IL-15 enhanced the degranulation and cytotoxicity of NK cells towards activated T cells. Discussion: This study provides new insight into the role of NK cells in the regulation of GVHD, demonstrates for the fi rst time that unlike in mice models, the ability of donor NK cells to inhibit and lyse autologous activated T cells is impaired during human GVHD, possibly due to altered subgroup and down regulated activated receptors. When treated with IL-15 ex vivo, this phenotype and function impairment can be partly reversed, which potentially may provide an opportunity for therapeutic treatment of GVHD. Disclosure of Interest: None Declared. 1 1 Hematology, 2 Dental Medicine, 3 Neurology, 4 Dermatology, 5 Nutrition, 6 Pediatric Hematology, 7 Gynecology, 8 Rehabilitation Medicine, UHC Zagreb, 9 Infectious Disease, University Hospital for Infectious Diseases "Dr. Fran Mihaljević", 10 Transfusiology, UHC Zagreb, 11 Ruder Boskovic Institute, 12 Immunology, UHC Zagreb, Zagreb, Croatia, 13 Transplantation & Immunology Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, United States Introduction: Due to increasing safety of allogeneic hematopoietic stem cell transplantation (alloHSCT), the number of survivors is increasing and more patients are at risk of developing chronic Graft-versus-Host disease (cGVHD), a leading cause of non-relapse mortality and morbidity after alloHSCT. Approximately 50% of alloHSCT recipients develop cGVHD. CGVHD is a disorder aff ecting many organ systems in highly variable fashion and requires multidisciplinary medical management. Division of Hematology at the University Hospital Center Zagreb (UHC Zagreb), Croatia, has 30-year long experience with alloHSCT. Recognizing cGVHD as one of the most serious challenges after alloHSCT and wishing to achieve better clinical care for long-term survivors, a Center for cGVHD and long-term follow-up was recently established at UHC Zagreb. Multidisciplinary clinic infrastructure was set up using established cGVHD-related grading scales and measurements in collaboration with the National Cancer Institute, USA. Materials (or patients) and Methods: From 1983, a total of 753 patients received alloHSCT at UHC Zagreb. From 2009 to July 2013, 119 transplanted patients were analyzed, of which 102 patients are alive and 33 (32%) developed cGVHD. Since the establishment of a comprehensive cGVHD Center, patients are fi rst seen by a hematologist, with history and physical exam, focusing on prior acute GVHD (aGVHD) history, current symptoms and onset of cGVHD, performed as part of an organ-system oriented exam. Standard cGVHD evaluation and scoring forms are fi lled according to NIH Consensus recommendations. Selected nurse is engaged in visit organization. Comprehensive laboratory workup is done, and patients are seen and evaluated by sub-specialists (Dental,

Dermatology, Rehabilitation, Neurology, and other) with further workup as needed. Quality of life questionnaires (SF 36, EORTC QLQ-C30, Lee symptom scale) are fi lled during the visit. All data are stored in a database and weekly team meetings are organized. Results: Using comprehensive cGVHD approach, 19 patients (1 pediatric) were assessed, median age 36.5 years. Three patients were ineligible, as they did not meet criteria for cGVHD diagnosis. Twelve transplants were from matched related donor, 11 using myeloablative conditioning, and 11 using PBSC as stem source. Eleven patients had previous aGVHD. Seven patients had de novo cGVHD, 7 quiescent and 2 progressive onset. Fifteen were classifi ed as classic and 1 as overlap. Ten patients had score 3 according to NIH global scoring, and 6 had score 2. Most involved organs were skin, joint/fascia, eyes, and lungs. Challenges included organizing timely access to sub-specialists and liberating dedicated research time for team members. An internationally peer reviewed grant through a World Bank program (UKF Fund) awarded in 2013 will help with strengthening research infrastructure and staff support. Discussion: Establishment of multidisciplinary team for cGVHD and long-term follow-up after HSCT in a new environment proves feasible. It enhances the quality of medical documentation and management of these complex patients. Such comprehensive clinical and research infrastructure will allow further collaboration and integration with other European and international cGVHD centers. Disclosure of Interest: None Declared. Introduction: CD4+ Foxp3+ regulatory T cells (Tregs) have been shown to play important roles in the maintenance of tolerance after allogeneic hematopoietic stem cell transplantation (allo-HSCT) and patients with chronic graft versus host disease (cGVHD) have a relative defi ciency of Tregs partly due to the abnormalities in Treg homeostasis. IL-2 is a critical homeostatic cytokine for Tregs. Low-dose IL-2 therapy have been reported to restore Treg homeostasis in patients with cGVHD and have a potential therapeutic effi cacy on cGVHD in some patients (John Koreth et al, New England Journal of Medicine, 2011; 365:2055-66) . So we analyzed the relationship among serum IL-2 levels, Treg frequency and cGVHD severity after allo-HSCT. Materials (or patients) and Methods: Patients undergoing allo-HSCT in our center from January 2000 to July 2012 were selected. They were divided into 3 groups according to cGVHD criteria including non-cGVHD, limited cGVHD and extensive cGVHD group. 10ml peripheral blood was drawn from all the selected patients for Tregs analysis by fl ow cytometry. Serum cytokine levels of TNF-α, INF-μ and IL-2 were evaluated by ELISA. Data were processed and analyzed using SPSS 17.0. Statistical signifi cance in diff erent groups was assessed by two-sample t-test. Spearman's correlation was used to test the correlation between two continuous variables. Results: 22, 18 and 30 patients were selected in non-cGVHD, limited cGVHD and extensive cGVHD group respectively. As shown in Figure 1A , we found that serum IL-2 level was the highest in non-cGVHD group (2.67±2.02pg/ml), higher in limited cGVHD group (1.04±1.14pg/ml) while the lest in extensive cGVHD group (0.28±0.55pg/ml) (P<0.05). IL-2 can facilitate the homeostasis of Tregs. We further found the percentage of Tregs was signifi cantly increased in non-cGVHD group(2.58±0.81%) compared to limited cGVHD (1.51±0.83%) and extensive cGVHD group (1.04±0.55%) (P<0.05) ( Figure 1B ). Spearman's correlation analysis revealed that the increased level of IL-2 was positively associated with increased Tregs (r=0.856, P<0.01), ( Figure 1C ). Moreover, we demonstrated increased Tregs were associated with less severity of cGVHD including reduced serum levels of INF-γ (r=-0.744, P<0.01) and TNF-α(r=-0.744, P<0.01) ( Figure 1D ). Discussion: For the fi rst time our result implied that serum IL-2 levels regulated cGVHD severity via increased Tregs after allo-HSCT and provided novel pathogenesis of cGVHD. It has been reported that low dose IL-2 administration was associated with preferential, sustained Tregs expansion in vivo and amelioration of the manifestations of chronic GVHD in a substantial proportion of patients. So our results might provide a prognostic factor associated with the effi cacy of IL-2 therapy for cGVHD after allo-HSCT. Disclosure of Interest: None Declared.

Introduction: In chronic graft-versus-host disease (cGVHD) a pathogenic role for B cells has been suggested by the presence of anti-HY antibodies in female-into-male transplants, by elevated plasma levels of B cell activation factor and by the therapeutic effi cacy of B-depletion. In systemic autoimmune disorders, such as SLE, Sjogren's syndrome, and scleroderma, pathogenic autoantibodies can be found in serum, and immune complex deposits may be identifi ed in aff ected tissues, and have been useful as markers of disease activity, severity or organ specifi city. Materials (or patients) and Methods: The focus of this study was to determine the incidence and associations of autoantibodies with cGVHD severity, activity and organ damage in a large patient cohort. 20 antibodies were characterized. Results: 280 cGVHD patients (29% moderate, 70% severe), with median duration of cGVHD 2 years, who received a median of four therapy lines, were consecutively enrolled in a cross-sectional natural history study protocol(NCT00092235). Circulating autoantibodies were present in 62% of cGVHD patients (n=174), and multiple antibodies were detected in 35% of patients (n=61) ( Table) . Patients with circulating autoantibodies had signifi cantly higher levels of IgM (<.0001), IgG (<.0001) and IgA (.001), elevated uric acid (0.008) and total protein (.0004), and higher numbers of CD3+ (.002), CD4+ (.001), CD8+ (.02) T cells and CD19+ B cells (<.0001). Prior rituximab therapy (n=66) reduced the incidence of autoantibodies (48 vs. 66% P=.01). Patients with moderate cGVHD severity by NIH criteria had higher numbers of CD19+ B cells (542 vs. 248 x10 6 P=.008) than those with severe cGVHD. Patients with moderate cGVHD had a higher total incidence of autoantibodies (67.5%) and of multiple antibodies (27.5%) than those with severe cGVHD (60% and 19.4% respectively). Moderate cGVHD was associated with signifi cantly higher incidence of cardiolipin M (.007) and higher titers of both RF (.001) and cardiolipin M (.002). Patients with active cGVHD (n=169) were not signifi cantly diff erent from non-active, however the frequency of autoantibodies accounted for more than 75% of ds-DNA, smooth muscle, Smith, and centromere in seropositive patients. Among organ systems aff ected by cGVHD, only oral cGVHD was associated with circulating antibodies, (67% vs. 53% P=.02), high ANA (32% vs. 18%, P=.002 and ANA titer 0.7 vs. 1.3 P=.008). Discussion: Circulating autoantibodies are common in patients with advanced cGVHD, its presence is associated with better quantitative immunologic reconstitution and do not refl ect the disease activity, severity or organ specifi city. Antibody detection has limited role in clinical assessment of cGVHD, however, the biological signifi cance of these autoantibodies remains to be determined. Disclosure of Interest: None Declared. Introduction: acute graft versus host disease (aGVHD) is one of the major complications after allogenic hematopoietic stem cell transplant (HSCT). Diagnosis is made with histology. Bio markers as a diagnostic and possibly prognostic tools are under continuous development. Total serum gamma-glutamyltransferase (TGGT) activity can be fractionated by size exclusion chromatography in 4 fractions: "big", "medium", "small" and "free" GGT ( b-, m-, s-, and f-GGT, respectively). These fractions are involved in cardiovascular, metabolic, nephrologic and neurologic diseases. Aim: we investigated the role of TGGT and its fractions in patients undergoing allogeneic (AlloTx) and autologous transplantation (AutoTx). We addressed if TGGT and its fractions increase due to conditioning, to aGVHD, if the increase is due to a predominant fraction, and if there is a correlation with transplant related mortality (TRM). Materials (or patients) and Methods: we prospectively enrolled in the study 65 patients (pts): 51 underwent an AlloTx and 23/51 (45%) developed aGVHD (study group). Twenty-eight pts without aGVHD and 14 AutoTx were considered our control groups. Conditioning in the AlloTx group were myeloablative (MyC) in 36 pts and Reduced Intensity (RIC) in 15 pts. Acute GVHD involved skin (N=10), lower intestinal (N=8), stomach (N=1) and hepatic+skin (N=2), lower intestine+skin (N=2). TGGT and its fractions were assessed before conditioning, at day+7,+14,+30,+60,+90, +180, +360 and any time GVHD occurred. Introduction: The therapeutic effi cacy of allogeneic stem cell transplantation (alloSCT) for hematological malignancies relies largely on the graft versus leukemia (GvL) eff ect exerted by the donor CD3 cells, but an uncontrolled graft-versus-host-disease (GvHD) bears a risk of complications. On the other hand, T regs cells (CD4+CD25high Foxp3+) are believed to maintain tolerance and to inhibit GvHD after alloSCT; also, the Foxp3 gene encodes a transcription factor that is a key for thymic development, so T regs cells could preserve an optimal microenvironment for the reconstitution of functional immunity after alloSCT. Moreover, when looking at post allotransplant patients' outcomes, there is no evidence that donor graft CD3/T regs ratio may determine an eff ect in terms of OS, NRM and relapse free survival rates so far. In this study we analyzed the graft CD3+/Tregs ratio (gCD3/Tregs R) and determined its impact on acute GVHD (aGVHD) and survival rates (OS, NRM and Relapse) after myeloablative alloPBSCT. Materials (or patients) and Methods: We analyzed 102 consecutive patients (median age 39 yy) transplanted with unmanipulated peripheral blood stem cells from an HLA identical related donor (n=78) or an HLA identical unrelated donor (n=32); diagnoses were acute myeloid leukaemia (n=82), acute lymphoblastic leukaemia (n=20 Introduction: Chronic GVHD (cGVHD) is the major cause of long term morbidity after allogeneic HSC transplantation. No biomarkers are currently known that can consistently predict its occurrence. We have previously observed that patients with cGVHD have increased numbers of circulating activated monocytes (Arpinati Transplantation 85:1826; 2008) . Also, the serum concentration of infl ammatory chemokines has been associated with acute GVHD (aGVHD). We thus aimed to evaluate whether PB numbers of antigen-presenting cell (APC) subsets or serum chemokine concentrations are biomarkers of chronic GVHD occurrence. S305 unmanipulated hematopoietc stem cell transplantation (HSCT) from HLA-haploidentical donors with the aim of inducing T regulatory cells (Tregs) in vivo and prevent graft-versus-host disease (GVHD). In one of these protocols (TrRaMM, Eudract 2007-5477-54) , rapamycin administration to 113 pts with high-risk hematological malignancies associated with a low risk of grade III-IV acute GVHD (21%) and chronic extensive GVHD (35%). Importantly, three-years relapse incidence was 48%, indicating benefi t from substantial graft-versus-leukemia (GVL) eff ects. Aim: A long-standing question is whether Tregs-based strategies may interfere with the antitumor eff ects of HSCT. We therefore investigated if diff erential homing capacities of in vivo rapamycininduced Tregs might explain the dissociation of GVHD from GVL eff ects. Results: In pts receiving rapamycin post-transplant, we documented the accumulation of circulating CD4+/CD25+/FoxP3+/ IL-7Ralpha-Tregs (at day 30, median 6.5% range 0.2-37.2, P<0.01 compared with controls), which were suppressive ex vivo in CFSEdilution assays and were selectively demethylated at the FoxP3 locus. Tregs also expressed high levels of CD45RO, suggesting extensive post-thymic proliferation and acquisition of a memory phenotype. Interestingly, Tregs frequencies inversely correlated with the risk of developing subsequent acute GVHD (P<0.05). These eff ects were rapamycin-specifi c, since 5 pts receiving cyclosporine A post-transplant had low to undetectable Tregs levels and all developed grade IV GVHD. Surprisingly, the bone marrow (BM) of pts receiving rapamycin was depleted of Tregs (at day 30, BM Tregs frequencies: median 0.3% range 0.0-2.2, P<0.01 compared with PB Tregs), but enriched with CD45RA -/CD62Leff ector memory CD8+ T cells. Compared with eff ector T cells, circulating Tregs had signifi cantly lower levels of the BM addressin CXCR4 (P<005). Tregs generated in vitro after CD3/CD28-beads stimulation and rapamycin culture also showed CXCR4 downregulation and displayed lower migratory capacities across SDF-1 gradients. These eff ects were specifi c for rapamycin-induced Tregs, as naturally occurring Tregs readily migrated ex vivo and could be inhibited by the CXCR4 antagonist plerixafor. Discussion: Our results hint to a yet unrecognized compartmentalization eff ect of rapamycin on Tregs. Rapamycin-induced Tregs appear to re-circulate in the peripheral tissues and protect from GVHD, but are excluded from the BM and likely do not interfere with GVL eff ects locally. Disclosure of Interest: None Declared. 1 Hematology, Hemostasis, Oncology, and Stem Cell Transplantation, Hannover Medical School, 2 Mosaiques Diagnostics and Therapeutics AG, Hannover, 3 Hematology and Oncology, Städtisches Klinikum Braunschweig Medizinische Klinik III, Braunschweig, Germany Introduction: Severe graft-versus-host disease (GvHD) is one of the major complications after hematopoietic stem cell transplantation (HSCT). Early prediction and diagnosis of GvHD are important for an early intervention and better survival of the patients. We aim to predict GvHD without invasive techniques and at an early stage using urine and plasma biomarkers. Materials (or patients) and Methods: 100 patients transplanted at Hannover Medical School were included in this study. Most patients were transplanted for acute leukaemia (n=69) and from HLA-matched donors (n=82). Reduced intensity conditioning was used for 42 pts and 93 pts received T-cell depleting antibodies prior to HSCT. Patients were divided into a test set (n=69, 80 samples) and a blinded validation set (n=31, 263 samples). For the test set plasma sampling was done close to time of diagnosis of GvHD and controls without GvHD were chosen on time-matched time points after HSCT. For the validation set plasma and urine samples were collected after transplantation weekly up to day 35 and biweekly thereafter. Here we analysed β2microglobulin (β2m) and CD99 and other GvHD-related biomarkers, such as CD25, Elafi n, REG3α, ST2 and sTNF-RI as reported by Paczesny et al. in 2010 , and Vander Lugt et al. 2013 in all plasma samples collected after allogeneic HSCT via ELISA. In addition we analysed the simultaneously taken urine samples from validation set patients with CE-MS in order to compare both methods. Results: As a follow-up to the test set data presented last year we measured seven biomarkers (β2m, CD99, CD25, Elafi n, REG3α, ST2 and sTNF-RI) in plasma samples in a validation cohort for a timeperiod of up to one year after HSCT. After unblinding of the samples a receiver operating characteristic (ROC) analysis was performed revealing a mean of areas under the curve (AUC) of 0.57 ± 0.05 over all seven plasma biomarkers. Based on the cut-off values calculated within the test set results of the plasma biomarkers the validation cohort showed moderate conformity with aGvHD-diagnosis represented by sensitivity and specifi city values of 0.60 ± 0.38 and 0.41 ± 0.25, respectively. We observed fi ve cases of non-relapse mortality (NRM) within the validation patients. In four of these fi ve patients all biomarker concentrations were highly increased. Especially β2m was up to 100-fold increased compared to the aGvHD-cut-off . Analysis of the urine samples of the same patients showed sensitivity and specifi city for correct classifi cation of patients with aGvHD of 0.75 and 0.50, respectively. Patients diagnosed with aGvHD and with positive urine proteomic patterns showed pattern positivity before clinical diagnosis (median 25.5 days before, interquartile range of 49 days). Discussion: We confi rmed published data that plasma biomarkers can predict poor overall survival and higher NRM-rates in patients after HSCT by a massive increase of marker concentration in plasma of those patients. The newly found β2m could be integrated into the list of markers predicting NRM. However, plasma markers did not allow for diagnosis of GvHD in our small cohort. In contrast, urine analysis showed higher sensitivity values for aGvHD even before clinical manifestation. Early intervention and pre-emptive therapy with steroids upon proteomic urine pattern analysis are currently under investigation in a clinical trial. Results: The incidence of acute graft versus host disease in the related and unrelated donor groups was 22.6% and 54%, respectively. Children less than 5 years of age required a much higher dose of tacrolimus than older children. The drug could be safely administered to achieve therapeutic range sublingually in all our patients. The adverse events commonly associated with tacrolimus included hypomagnesemia (100 %), renal dysfunction (5%), hypertension (8%), seizures (14%) and hyperglycemia (5%). Discussion: An initial dose of 0.03 mg/kg/day either oral or sublingual was started on all children. To assess if target whole blood concentrations of tacrolimus in children undergoing haematopoietic stem cell transplantation can be achieved reproducibly with this dose or not was assessed by checking trough drug levels 3 days a week for the fi rst two weeks. We reviewed the tacrolimus blood levels and calculated clearances for 55 children (aged 6 months to 18 years, median 9 years) using tacrolimus after allogeneic marrow, blood stem cell or cord blood transplantation. In children less than 5 years old undergoing haematopoietic stem cell transplantation we found tacrolimus clearance faster than older children and adults. Hence, careful therapeutic monitoring is needed in the fi rst 2 weeks after transplantation to avoid prolonged subtherapeutic dosing for this age group. Tacrolimus was well tolerated and eff ective in graft versus host disease prophylaxis in paediatric patients undergoing allogeneic stem cell transplantation. The less than 5 year old children undergoing transplantation required increased dosing and careful monitoring of drug levels. Hypomagesaemia and hypertension were the commonest side eff ects seen with the use of this drug. Disclosure of Interest: None Declared. Inline method (all 3 steps in one system, UVARXTS™, CELLIX™), offl ine method (Leukocyte collection by Leukapheresissystem (AMICUS™, OPTIA™,COBEspectra™) and separate irradiation system (Macogenic™)), and MINI method (manual blood drawing, and all steps are manually performed under GMP conditions). Whether there is a diff erence between the methods no data are available. To elucidate the question, we performed a retrospective analysis of our patients treated for acute GVHD refractory standard immunosuppressive therapy (SIT). Materials (or patients) and Methods: From 2002 until 2013 23 patients with > grade 2 acute GVHD refractory to SIT were treated with ECP (median age 6.9y (1.5 -12) , median bw 31 kg (7 -60), 11 f, 12 m). Patients chart were analysed for GVHD, SIT, GVHD treatment, side eff ects, number of ECP and outcome of acute GVHD. The schedule was performed to our in-house standards. ECP started by a frequency of 2 to 3 reinfusions/week, with individualized tapering due to response of the GVHD and the ability to reduce the immunosuppressive therapy (IT). Results: In total 338 procedures in 23 patients were enroled (123 inline in 4 patients (bw median 40 kg, median 32ECP(10-49)/patient); 174 offl ine in 13 patients (bw median 24 kg, median 10 ECP(3-55)/ patient); 38 MINI in 6 patients (bw median 10 kg, median 7 ECP (3-9)/patient). No severe side eff ects were observed in either method. 21/23 patients improved (20 CR, 1 PR (gut), 1 SD (skin+liver), 1 PD (skin, liver, gut)). Log-rank-sum-test showed no statistically significant diff erence (P = 0.98) between the used methods for outcome and survival. The groups were not comparable to age and bodyweight, therefore a bias has to be claimed. In 22/23 IT could have been reduced, especially corticoid dosage could be tapered. Discussion: ECP is in our hands an eff ective second line therapy in acute GVHD. Neither the method used nor the individualized schedule applied seems to infl uence the outcome. Due to the small patient number, this report could be only a step forward to prospective randomized trials, bringing hopefully an answer to these questions. Disclosure of Interest: None Declared. (3). Three patients had a previously allo-SCT. All patients had extensive moderate/severe cGVHD with cutaneous involvement, mucosal involvement (7) and/or lung involvement (6). The median therapy lines for cGVHD before acitretin treatment were 2,5 (0-5). Seven patients were refractory to several IS lines that included systemic corticosteroids in all cases and extracorporeal photopheresis in 4 cases. In one patient acitretin was started for severe cutaneous and muscular morphea at the beginning of cGVHD. The initial dose of acitretin was 25-35 mg/day added to previous IS, and it was reduced based on clinical course and toxicity. In all patients, liver function tests and lipids were monitored. The median time from the onset of cGVHD to the beginning of acitretin was 23 months . Results: The median time of acitretin therapy was 10 months (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) . All patients have shown improvement while receiving acitretin: 4 had cutaneous complete response (CR) and 4 partial response (PR). Two patients with PR showed cutaneous progression (at 4 and 9 months respectively) after withdrawal of the drug due to side eff ects. The median time to response was 2 months (1-3) and the median of duration of response was 13 months . Six patients developed side eff ects being the most frequent xerosis (6) and hyperlipidemia/hypertriglyceridemia (2) .Two patients discontinued treatment due to severe xerosis and alopecia. In three cases, a 10 mg dose was used as maintenance treatment (1 for a good clinical outcome and 2 for mild xerosis). IS therapy could be diminished in 5 patients and withdrawal in 3. After acitretin discontinuation, there was no relapse or progression of SCL-cGVHD in those patients who achieved CR. Discussion: In our experience, acitretin is an eff ective and well tolerated treatment in refractory SCL-cGVHD. It produces a high rate of complete responses and allows the reduction of systemic immunosuppression in most cases. Disclosure of Interest: None Declared.

S307 and transplantation. The release of soluble HLA-G molecules is known to be infl uenced by genetic factors within the HLA class I and II regions of chromosome 6. Materials (or patients) and Methods: Here, we studied the course of sHLA-G levels in 22 patients undergoing HLA-identical (N=13) or HLA-mismatched (N=9) allogeneic stem cell transplantation (alloSCT) due to acute myeloid leukemia (AML, N=17), secondary AML (sAML, N=4) and myelodysplastic syndrome (MDS, N=1). Plasma samples were serially procured from the patients before and 1, 2, 3, 4, 5, 6, 9 and 12 month(s) after transplantation. The samples were analyzed by sHLA-G specific ELISA and related to acute and chronic Graft-versus-host disease (GvHD). Results: In line with the expected tolerance-inducing function of HLA-G, the sHLA-G levels were overall signifi cantly increased in patients with low or no GvHD (N=13) compared to patients with acute GvHD grade 2 -4 (P=0.02, N=9; two-way ANOVA). The diff erence in the sHLA-G courses was more prominent when patients with no or limited chronic GvHD (N=11) were compared to patients suff ering from extended chronic GvHD (P=0.0008; N=11). Similar results were obtained for patients undergoing HLA-identical allo-SCT for acute (P=0.01) and chronic GvHD (P=0.0001), s. Figure 1 . However, in HLA-mismatched alloSCT the sHLA-G levels did not diff er in patients with and without severe acute or chronic GvHD. Discussion: This study clearly demonstrates that high levels of sHLA-G favor tolerance after HLA-identical alloSCT, whereas in HLA-mismatched alloSCT genetic factors must be taken into consideration responsible for the release of HLA-G molecules which in turn infl uence the transplantation outcome. Disclosure of Interest: None Declared. Introduction: CD4+CD25++ lymphocytes are recognized as cells representing T regulatory cells subpopulation. Indeed, more than 80% of CD4+CD25++ are equipped with the suppressor cells machinery being FoxP3+ Baecher-Allan J. Immunol., 2001 , Roncador Eur. J. Immunol., 2005 . Materials (or patients) and Methods: 99 patients (age: 7 to 65 years, median 42; 59 unrelated (MUD) and 40 sibling (SIB) transplantations: all except one MUD transplanted patients received anti-lymphocytes antibodies -ATG: 51 patients or Campath: 7 patients, in variance in the SIB group ATG or Campath received only those on reduced intensity conditioning -13 and 3 patients, respectively) transplanted due to hematological malignancies were followed for the proportions and numbers of CD4+CD25++ lymphocytes in blood. Observation started on the day of hematological reconstitution (granulocytes > 500/μl), and was continued in one week interval for four weeks. 1 III. Medizinische Klinik, Universitätsmedizin Mannheim, Mannheim, 2 Medizinische Klinik V, Universität Heidelberg, Heidelberg, 3 Medizinische Klinik II, Frankfurt, 4 Medizinische Klinik I, Universitätsklinikum Carl Gustav Carus, Dresden, 5 Zentrum für Knochenmarktransplantation, DKD, Wiesbaden, 6 5. Medizinische Klinik, Klinikum Nürnberg, Nürnberg, 7 2. Medizinische Klinik, Klinikum Augsburg, Augsburg, 8 Klinik für Hämatologie, Onkologie und Klinische Immunologie, Heinrich Heine Universität, Düsseldorf, Germany Introduction: Steroid refractory intestinal acute GvHD (aGvHD) is a severe complication with a poor prognosis. Pentostatin has shown effi cacy as salvage therapy in this situation. Here we report on a retrospective analysis on patients with histologically proven severe intestinal steroid-refractory aGvHD treated with pentostatin at 8 centres. Since pentostatin had been combined frequently with additional immunosuppressive therapies in the participating centres it was a specifi c aim of this study to elucidate the eff ect of such simultaneous immunosuppressive approaches. Materials (or patients) and Methods: 124 patients who had received pentostatin as salvage therapy due to III° (58) or IV° (66) intestinal steroid-refractory aGvHD between 2000 and 2011 were included. Steroid-refractory aGvHD was defi ned as progression or no improvement despite treatment with prednisolone (≥2mg/kg/d) for ≥ 3 days. Pentostatin was infused at a dose of 1mg/m 2 for 3 days. Patients received 1-4 cycles. Steroids and calcineurin inhibitors were continued. Response was classifi ed as complete (CR, no ongoing symptoms), very good partial (VGPR, residual symptoms) or partial (PR). 50 females and 74 males with a median age of 50 (range: 19-70) years were included. The underlying diseases were AML (71), ALL (15), MPN (6), lymphoma (11), MDS (11), multiple myeloma (9) and aplastic anaemia (1) . Patients had been transplanted from matched related (39), matched unrelated (54) Hematology department, 2 Hospital Universitario de Salamanca, Salamanca, Spain, 3 Hematology department, Hospital de San Antonio, Porto, Portugal, 4 Hematology, 5 Pharmacy department, Hospital Universitario de Salamanca, Salamanca, Spain Introduction: Acute GVHD is one of the major limiting factors in successful allogeneic hematopoietic stem cell transplantation (HCT), and early onset (hyperacute GVHD -haGVHD-) may be associated with lower response rate to fi rst-line therapy and a higher non-relapse mortality rate. Few studies have been reported in the last years focusing on this aGVHD type. Materials (or patients) and Methods: Between 1996 and 2013, 564 patients have received HCT in our center, 49 of them (8,7%) developed haGVHD defi ned as acute GVHD occurring before day +14, independent of neutrophil engraftment. Aims of this study were to analyse the outcome of these patients and to identify risk factor predicting survival. Results: Median age was 45 years (16-65). Diagnosis were: 34% acute leukemias, 14% MDS, 32% NHL, 4% MM, 14% other diagnosis. Status of the disease at HCT was in 57% complete response (CR), 10% partial response, 10% stable disase and 22% non-response. In 41% of patients, conditioning regimen was myeloablative (22% including total body irradiation). 59% received HCT from unre-lated donor (75% HLA 10/10). The cell source was peripheral blood in 81%. GVHD prophylaxis was based on cyclosporine in 50% (42% plus methotrexate and 8% plus MMF) or tacrolimus (45%; 10 % plus methotrexate and 35% plus rapamicine), 2 patients had received rapamicine-bortezomib; in 2% ATG was added. Median day of neutrophil engraftment (count > 0.5x10 9 /L) was day +15 (8-29) . Median day to development haGVHD was 9 (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) and the clinical characteristics were: 12% fever, 75% rash, 67% gastrointestinal symptoms, 26% hepatic dysfunction and 10% weigh gain. haGVHD grades were: 14% grade I, 57% grade II, 22% grade III and 6% grade IV. Biopsy was performed in 75% of patients, and the clinical diagnosis was proven in 26%. Overall response to fi rst line treatment was 85%, with 75% of CR. 63% of patients achieving CR, relapsed in a median time of 52 days from the fi rst episode; 38% with grade III-IV; 66% of them responded to the next line therapy. 66% of the patients at risk alive at day + 100 developed chronic GVHD. With a median follow-up of 43 months for patients alive (13-174), the estimated 6 months, 1 and 5 years OS was 67%, 53% and 20%. Overall transplant related mortality was 61% (23% at day +100). The main cause of death was GVHD (61%). In the univariate analysis, factors signifi cantly associated with a better OS an PFS were: reduce-intensity regimen (not achieved (NA) vs 6 months, P=0.002), GVHD grade I-II Vs III-IV (NA vs 4 months, P=0.013) and CR with fi rst line treatment (NA vs 3 months, P=0.02). Regarding to aGVHD reactivation, if it was >100 days after the fi rst episode (NA vs 12 months, P=0.04) or grade I-II-II vs IV (NA Vs 8 months) were associated with favourable outcome. Discussion: Although it is assumed that engraftment is a requisite for the appearance of acute GVHD, in almost 10% of our patients it appears before. Histology confi rms the diagnosis in only about a quarter of the patients. Although in our experience CR is high, it seems to be associated with a higher relapse and mortality. Much more biological knowledge is needed in these patients in order to identify them and improve prognosis. Disclosure of Interest: None Declared.

F. Alby-Laurent 1,* , J.-H. Dalle Robert Debré, Paris, France Introduction: Because of its effi cacy and for historical reasons, Cyclosporine A (CSA) still remains the basis of immunosuppression used in allogeneic hematopoietic stem cell transplantation (HSCT) as graft versus host disease (GVHD) prophylaxis at least in Europe. However, some published studies showed that Tacrolimus (FK) could be equivalent to, even better than CSA in this indication. Finally, some patients experimented evolutive acute GVHD despite CSA or CSA intolerance, then we may switch CSA for FK. We report here our experience about this switch. Materials (or patients) and Methods: We included all the children who received an HSCT and a switch from CSA to FK between 2007 and 2012 in our pediatric hemato-immunology department. Results: 302 allogeneic HSCT were performed during this period in 287 patients. All of them received CSA based GVHD prophylaxis. Among these 287 patients, 23 switched from CSA to FK. The median age of the patients who switched to FK was 7,2 years old and the sex ratio was 47.8%. 86.9% of them had malignant disorders including 65% with acute disorders in either CR1 (46%) or CR2 (54%) before the HSCT. 65.2% of the donors were unrelated and bone marrow was the most common source of stem cells (78.2%). The initial GVHD prophylaxis was CSA alone (26.1%), CSA + Methotrexate (39.1%) or CSA + Mycophenolate Mofetil (34.8%). 43.4% of the patients received antithymocyte globulin as pre-HSCT. The median follow-up was 556 days after the HSCT (range 31-2101). The OS and the DFS were 78.2%. The cumulative incidence of TRM was 8.6%. Only 1 patient died because of GVHD. 69,6% of the patients were alive and free from disease and GVHD. The switch between CSA and FK was made averagely on day 59 (range: 15-134) after the HSCT. The main cause of the switch was a CSA-refractory acute GVHD despite association with corticosteroids +/-other immunosuppressive drugs (95.7%). The average time for obtaining FK effi cient serum level (T0 ≥5ng/ml) was 7.5 days. The average minimal posology to have an effi cacy T0 of FK was 0.01 mg/kg/day IV and 0.12mg/kg/day p-o. Before the switch, 39.1% of patients had grade I-II acute GVHD, 47.8% had acute GVHD grade III-IV and 13% had chronic GVHD. The average grade of GVHD was 2.56 +/-0.90. After the switch we observed a signifi cant reduction of GVHD with an average grade of 1.57 +/-1.41 (P<0.01) when the T0 was obtained. The diff erence continues being signifi cative 3 months after the switch with an average grade of 0.038 +/-0.89 (P<0.01). 17.3% of the patients stopped tacrolimus because of side eff ects but 47.3% of the rest of the patients no longer needed the administration of other immunosuppressive therapies. Thrombotic microangiopathy episode in two patients -one of them had previously experienced it under CSA -, one seizure episode and 2 diabetes mellitus cases which could be also be explained by the concomitant administration of corticosteroids, were reported as side eff ects of FK. 43.4% of pts had presented an episode of viral replication. 3% needed antiviral treatment. 13% had documented bacterial events and no patients had fungal events. Discussion: This study suggests that FK may be helpful for some patients with CSA resistant GVHD and may be used safely. It would be interesting to do a prospective study comparing CSA and FK as GVHD's prophylaxis. Disclosure of Interest: None Declared. Hematology and Bone Marrow Transplantation, 2 CHRU Lille, Lille, France, CHRU Lille, Lille, France Introduction: Diarrhea is a frequent and common symptom after allogeneic stem cell transplantation. Etiological investigation can be challenging as treatment options are often opposed. The purpose of this study was to evaluate the impact of the global diagnostic approach on the outcome of patients suspected of having acute (a) gastrointestinal (GI) graft-versus-host disease (GVHD). Materials (or patients) and Methods: Patients who had diarrhea underwent extensive gastrointestinal examination and screening for infection. Small bowel exploration using video-capsule endoscopy (VCE) (associated or not with colonoscopy and/or upper endoscopy with systematic biopsy) was performed within 48 hours for each patient with atypical symptoms (without evidence of stage II or higher acute graft-versus-host disease (GVHD)related skin lesions). Results: A total of 37 VCE were performed on 30 consecutive patients who underwent allogeneic stem cell transplantation. The indications were isolated diarrhea (n=15), persisting or relapsing diarrhea (n=5), febrile or hemorrhagic diarrhea (n=17). The mean time between transplantation and the exploration was 48 days. The final diagnoses were GVHD (n=19), functional diarrhea without any other cause (n=9), viral infection (n=9), of which 3 were associated with GVHD. The evolution was favorable (n=24), treatment change (n=8), or death (n=5). The correlation between VCE results and the final diagnosis was perfect. For all diagnosis, the positive predictive value was 100% and the negative predictive value was 80%. Of the 20 patients with GVHD lesions observed with VCE, the final diagnosis was identical (2 cases with co-infections). Of the 10 patients with normal VCE, diarrhea was either toxic (n=8) or infectious (n=2). In 2 cases, VCE showed infectious lesions matching with the final diagnosis. Of note, VCE could not be interpreted in 3 cases because the capsule stayed in the stomach sack. In these 3 cases, the diagnoses were GVHD, motor diarrhea and GVHD with co-infection. In the 8 cases with uncertain or non-specific histological diagnosis, VCE results were normal (ruling out GVHD diagnosis) or showed GVHD lesions (leading to immunosuppressive treatment prescription). Discussion: The VCE, associated with infectious screening, has enhanced the authors' ability to establish the right diagnosis and adapt treatments in patients suff ering from diarrhea. This approach appears highly useful in patients with atypical GVHDlike symptoms, particularly with a view to avoiding unnecessary immunosuppressive treatment. Disclosure of Interest: None Declared.

Z. Al-Kadhimi 1,* , Z. Gul 2 , W. Chen 3 , D. Smith 3 , A. Mitchell 3 , M. Abidi 4 , L. Ayash 4 , L. Lum 4 , A. Deol 4 , V. Ratanatharathorn 4 , J. Uberti 4 1 Hematology/Bone Marrow Transplant, Emory University/ Winship Cancer Center, Atlanta, 2 Hematology/Bone Marrow Transplant, University of Kentucky/ Markey Cancer Center, Lexington, 3 Biostatistics, 4 Hematology/Bone Marrow Transplant, Wayne State University/ Karmanos Cancer Center, Detroit, United States Introduction: Chronic graft-versus-host disease (cGVHD) is a major factor determining long term outcome and quality of life following allogeneic hematopoietic cell transplantation (AHSCT). In fact, cGVHD is the leading cause of late non-relapse mortality (NRM) and morbidity after AHSCT. The rates of severe cGVHD using NIH consensus criteria reported in the literature are 15-38%. Eff ective regimens are needed to reduce the severity of cGVHD and improve NRM. We analyzed and updated the long term outcomes of our phase 2 trial evaluating the effi cacy of intermediate dose rabbit anti-thymocyte globulin (Thymo) in combination with tacrolimus and sirolimus for the prevention of acute and chronic GVHD after unrelated donor transplantation. Materials (or patients) and Methods: In a Phase II trial, 47 adult patients (pts) underwent AHSCT from unrelated donors using Thymo, tacrolimus, and sirolimus for GVHD prophylaxis. Thymo was given as follows: 0.5 mg/kg day -3, 1.5mg/kg day -2, and 2.5mg/kg day -1. Chronic GVHD was measured using NIH consensus criteria. Results: Twenty-two pts received 8/8 and 25 received 7/8 HLA matched unrelated grafts. Thirteen pts received a reduced intensity preparative regimen, while 34 pts received a full intensity regimen. The median follow-up duration was 45.2 months (95% CI 37. 7-48.8) , with minimum follow up of 30 months. At 4 years of follow up, the cumulative incidence of NIH severe cGVHD a, was 6.4 % (95% CI 1.6-15.9), and overall cumulative incidence of cGVHD was 48.9% (95% CI 33.6-62.6). Of 20 pts who are alive and disease free, only 4 pts continue to need systemic immune suppression at the last follow up. At 4 years of follow-up, the cumulative incidence of NRM and disease relapse were 27.7% (95% CI 15.7-41.0) and 30.0% (95% CI 17.5-43.6), respectively. There has been one non relapse death beyond 12 months and none after 18 months of follow up. Only one pt died from cGVHD and bronchiolitis obliterans, while two other pts had minimal symptoms from bronchiolitis obliterans. Progression free survival (PFS) and overall survival (OS) at 2 years were 50% (95% CI 35-64) and 56% (95% CI 42-70). Median OS is 33.9 months [95% CI (9.8 -*) *the upper limit of the CI was not calculated due to the pattern of censoring] All patients were censored after 40 months, so PFS and OS could not be calculated at 48 months. The median karnofsky performance status at 2 years was 90%. Discussion: At long term follow up, the combination of Thymoglobulin, Tacrolimus, and Sirolimus was associated with low incidence of severe cGVHD, low incidence of late NRM, and good performance status. Most of the acute complications associated with this regimen could be minimized with maintaining Tacrolimus and Sirolimus at therapeutic levels. Further validation of these results in randomized phase III trials is needed. Disclosure of Interest: None Declared.

Introduction: Chronic GVHD is the main cause of late non-relapse morbidity and mortality after allo-SCT. Pathophysiology of chronic GVHD remains poorly understood. Chronic GVHD presents clinical features that mimic autoimmune diseases. The identifi cation of proinfl ammatory Th17 cells which contribute to autoimmune diseases pathophysiology, raised the issue of the role of Th17 cells in human chronic GVHD. Indeed, the contribution of Th17 cells in chronic GVHD was assessed in GVHD mouse models. Furthermore, theses studies indicates that the expansion of Th1 and Th17 cells is favored by the progressive loss CD4+CD25+Foxp3+ regulatory T cells (Treg) leading to chronic GVHD. This report investigated the role of Th17 cells and Treg in liver biopsies taken from patients with or without chronic GVHD. Materials (or patients) and Methods: The cohort included 17 patients who underwent allo-SCT for diff erent hematological malignancies (cGVHD group) with a median age 54 (range, 27-71) y. The stem cell source was PBSCs in 9 cases (53%) and BM in 8 cases (47%). 13 patients received transplant from a MRD, and 4 patients from a MUD. A RIC regimen was used in 8 patients (47%) and a MAC regimen in 9 patients (53%). The control group included 8 patients without hematological malignancies who underwent liver biopsy for sleeve gastrectomy for morbid obesity (n=2) or adjacent tumor surgery (n=6). Immunohistochemistry was performed on deparaffi nized tissues sections. A quantitative evaluation of antigens expression was performed by counting the number of positive cells in the whole biopsy. Results: In the cGVHD group, based on standard pathology criteria, all 17 patients had a histologically proven liver chronic GVHD. Biopsies were taken at time of fi rst hepatic symptoms declaration or during their reappearance and prior to corticosteroid treatment initiation or resumption. In the control group, all patients present histologically normal liver biopsy. In order to identify the Th17 cell population, biopsies were tested for expression of CD161 CCR6 and RORgt, the key transcription factor for the diff erentiation of Th17 cells. Signifi cantly higher numbers of RORgt+, CD161+ and CCR6+ cells were counted in the liver of patients with chronic GVHD compared with the control group, mainly found in the portal space (P=0.0001, P=0.03 and P=0.03 respectively). We also assessed the presence of T cells expressing Tbet (the transcription factor characterizing Th1 cells) and Foxp3 (the master regulator gene of Treg cells) in the liver biopsies. There was no diff erence in the number of Th1 and Treg cells between the 2 groups (P=0.88 and P=0.12 respectively). Finally we look at the Th17/Th1 and Th17/Treg ratios, considering Th17 cells as RORgt positive cells as RORgt is the more specifi c hallmark of Th17 cells. Both Th17/Th1 and Th17/Treg ratios were signifi cantly increased in the liver of patients with liver cGVHD (P=0.005 and P=0.002 respectively). Discussion: The current study shed some light on the role of Th17 cells in the context of liver chronic GVHD. We show that Th17 cells infi ltrate liver biopsies from patients with chronic GVHD. In addition, Th17/Treg ratio was signifi cantly increased in the liver of patients with liver chronic GVHD, suggesting a regulatory defect in liver chronic GVHD. These data raise the prospect of future innovative approaches to optimize immunosuppression regimens for the treatment or prophylaxis of chronic GVHD by targeting the Th17 response. Disclosure of Interest: None Declared. Hematology, Hokkaido University Graduated School of Medicine, Sapporo, Japan Introduction: The pathogenic of chronic graft-versus-host disease (cGVHD) is still elusive. Prior acute GVHD (aGVHD) increase the risk of cGVHD and in vivo T cell depletion decrease it. These preclinical fi ndings indicate that cGVHD is strongly linked to donor T cell; not only reconstituted T cell (recon-T) but also graft derived T cell (graft-T). To understand the roles of distinct T cells, murine model that replicating clinical features including T-cell reconstitution in human is required. Here, we established clinically relevant murine cGVHD model and characterize each T cells from distinct origin during cGVHD. Materials (or patients) and Methods: C3H.Sw (H2b) recipients were received 9Gy TBI and transferred T cell depleted BM (TCD-BM) with or without spleen CD4 and CD8 T cells from B6 (H2b) donor. After 9 weeks, histological analysis was performed in cGVHD aff ected organ (lung, liver, skin and salivary gland) following NIH criteria. The kinetics and function of graft-and recon-T in aff ected organ and secondary lymphoid organ (SLO) by using congenic system. Interventions included use of administration of Thy1.2 monoclonal antibody (mAb) from day 21 to deplete distinct T cell selectively. Results: In this model, recipient mice develop multiorgan cGVHD ( Figure 1A ) with immunodefi ciency. Both graft-and recon-T involved in aff ected organ more extent to SLO. Notably, graft-T was long-lasting and predominated over recon-T even at 63 days posttransplant ( Figure 1B) . Moreover, not all graft-T were exhausted and retained higher proliferation potential as compared to recon-T ( Figure 1C ). To demonstrate the contribution of long-lasting graft-T or recon-T in cGVHD pathogenesis, we performed selective depletion in chronic phase. Graft-T depletion failed to block cGVHD development because of recon-T proliferated and activated compensatory just in aff ected lesion. On the other hand, recon-T depletion resulted in acute exacerbation and immediately death due to graft-T reactivation. These data showed that recon-T was pathogenic enough to induce cGVHD independently, and graft-T possessed high-cytotoxicity and induce acute exacerbation in chronic phase. These compensatory activations indicated that each T cells was regulated competitively, and the extent of its activation depended on T cell pool size in aff ected organ. Discussion: These fi ndings elucidate the distinct roles of each T cell in chronic phase by using clinically relevant murine cGVHD model, and also suggest that the extrinsic regulation determine the extent of pathogenicity. Disclosure of Interest: None Declared. Introduction: IMiDs such as lenalidomide (Lena) have immunostimulatory eff ects and therefore potential to reduce relapse after allogeneic (A) HCT by increasing GvT eff ects. However, early clinical experience using IMiDs after AHCT has been limited by increased GvHD. Although Lena has been shown to augment mitogen-stimulated T cell responses in various ways, the eff ects of Lena on T cell alloresponses that mediate both GvT and GvHD have not been well defi ned. We determined the eff ects of Lena on S311 functional human T cell alloresponses using a HLA-mismatched in vitro model. Materials (or patients) and Methods: We cocultured CFSE-labelled PBMC from healthy donors with irradiated allogeneic PBMC in the presence of 1μM Lena or vehicle control. Functional alloresponses were quantifi ed after 9 days of allo-coculture by fl ow cytometry. In addition, allo-coculture responders were fl ow-sorted into alloproliferative or non-proliferative fractions and extracted RNA used for gene expression profi ling. Data was interrogated by O-miner, R and Ingenuity Pathway Analysis. Results: Addition of Lena to allo-cocultures (n=12) led to a median 17% increase in the total number of responder T cells (P<0.001) due to an increase in proliferation of allospecific responder CD8 (alloCD8) T cells (P<0.001). In contrast, Lena had no effect on proliferation of CD4 cells. Proliferation kinetic analysis showed that Lena did not increase the number of times alloCD8 cells divided, but increased the precursor frequency of these cells within the responder cell pool (median 4% vs 11%, P=<0.001) consistent with a lowering of the activation threshold of alloCD8 cells. However addition of Lena to allo-cocultures did not increase the proportion of alloCD8 cells secreting TNFα or IFNγ or expressing CD107a. As expected, alloCD8 cells from untreated allo-cocultures demonstrated >2-fold altered expression of >500 genes mostly associated with DNA synthesis/cellular proliferation pathways when compared to non-proliferative CD8 cells. Lena treated alloCD8 cells when compared to treated nonproliferative CD8 cells showed further increases in expression of many of these genes, consistent with potentiation of pathways intrinsic to proliferative CD8 alloresponses. Importantly, Lena treated alloCD8 cells also demonstrated significant changes in expression of additional genes when compared to untreated alloCD8 cells. These included >8 fold increases in expression of multiple genes known to potentiate T cell immune responses in other settings including PFKFB4, PIR and SOCS2 (part of the E3 ubiquitin ligase complex known to associate with cereblon and Lena) and >5 fold decreases in expression of genes which suppress T cell activation and memory differentiation including FAIM3 and PMCH.

Discussion: We have shown for the fi rst time that Lena potentiates human alloresponses by selectively increasing proliferation of alloreactive CD8 T cells. Lena likely mediates this eff ect by increasing expression of genes common to the intrinsic CD8 alloproliferative response but also by modulating expression of additional genes important in control of T cell activation and diff erentiation. These fi ndings could be exploited to use Lena more eff ectively to potentiate GvT without increasing GvHD after AHSCT. Disclosure of Interest: None Declared. Introduction: MicroRNA (miR)-181a enhances T cell receptor signaling by repression of several downstream phosphatases. Thus, we hypothesized that manipulation of miR-181a expression in donor T cells may alter Graft-versus-Host Disease (GvHD) after allogeneic bone marrow transplantation (BMT). Materials (or patients) and Methods: Lentiviral gene transfer was used to over-express miR-181a and eGFP in murine T cells. To study miR-181a loss-of function, T cells from miR-181a/b-1 deficient (miR-181a/b-1 -/-) mice were used. miR-181a-over-expressing (miR-181a-high) and miR-181a/b-1 -/-T cells were tested in vitro and in vivo using a murine haploidentical BMT-model (C57BL/6 into BDF1). T cell proliferation, apoptosis, alloreactivity and traffi cking of donor T cells were assessed at diff erent time points after gene transfer and BMT, respectively. GvHD symptoms were monitored using a clinical GvHD scoring system and survival was analysed using Kaplan-Meier statistics. Results: miR-181a was over-expressed up to 200-fold upon lentiviral gene transfer into primary T cells with transduction rates [PH-P438] S312 between 50% and 70%. In in vitro cultures stimulated with anti-CD3 and anti-CD28 beads, proliferation of miR-181a-high T cells was reduced 5-fold as measured by 3 H-thymidine incorporation. Furthermore, competitive in vivo homing assays showed significantly reduced numbers of miR-181a-high T cells in GvHD target organs seven days after BMT as compared to controls as assessed by FACS-analysis of eGFP positive CD4 + and CD8 + cells. Further analysis of donor T cells revealed signifi cantly increased Fas ligand expression on miR-181a-high T cells as compared to control-T cells four days after BMT. Interestingly, recipient mice receiving miR-181a-high T cells showed no or fewer signs of GvHD and survived signifi cantly longer as compared to control-T cell recipients. In contrast, miR-181a/b-1 -/-T cells showed a trend towards more severe GvHD than wild-type T cells in our GvHD model. Discussion: Our gain-and loss-of function data indicate a role of miR-181a in T cell allo-responsiveness. Although the precise molecular mechanism is not yet known over-expression of Fas ligand and activation induced cell death of miR-181a-high T cells may contribute to their reduced presence in GvHD target organs and concomitant prolonged survival in our haploidentical GvHD model. Hence, manipulation of miR-181a expression in allogeneic T cells modulates T cell allo-responses after BMT and miR-181a may thereby serve as a potential immunotherapeutic target to prevent or ameliorate GvHD. (*MS, ME and CK contributed equally to this work). Disclosure of Interest: None Declared. Introduction: GvHD is a major obstacle to safe HSCT. Reliable markers facilitating the monitoring of this invalidating disease are warranted to improve its management. The long pentraxin (PTX) 3 increases in diff erent human infl ammatory pathologies and the correlation between PTX3 levels and disease severity suggests the potential usage of this molecule as GvHD marker. Materials (or patients) and Methods: We analyzed plasma samples from 70 pediatric patients with hemato-oncologic diseases, who received HSCT at S. Gerardo Hospital (Monza), collected before, at HSCT and weekly thereafter, until day +100. Blood samples were further collected at the day of GvHD onset, before the beginning of GvHD-specifi c drug therapy. PTX3 plasma levels were monitored by ELISA. 34 patients developed acute GvHD very early (i.e. within day +28), 20 between day +28 and day +100 and 16 did not show GvHD in the monitored time-frame (jointly called 'no early GvhD' group). Results: PTX3 plasma levels were signifi cantly higher at the onset of early GvHD compared to levels of 36 patients in the 'no early GvHD' group: the median was 33.1 ng/ml (range: 9.4-847.4) versus 14.2 ng/ml (range 4.1-200.6), respectively (P-value<0.0001). Among patients who developed GvHD anytime within day+100, 17 with grade I GVHD showed signifi cantly lower levels of PTX3 at disease onset (median 15.7 ng/ml, range: 7.3-38.3) as compared to levels of 36 patients with grade >I GVHD (median 37.1 ng/ml, range: 7. 1-847.4, P-value=0.0012 ). In addition, we investigated the potential of PTX3 measured at day +0, +7 and +14 since HSCT in the prediction of early GvHD occurrence. After normalization with respect to baseline levels, i.e. before conditioning regimen, individual change in PTX3 at each time-point was calculated for patients in the 'early GvHD' and 'no early GvHD' groups. Comparison of median change at day +0 in the 2 groups showed that the conditioning regimen induced an increase in PTX3 levels in both groups, but with no statistically signifi cant diff erence between them (P-value=0.62). However, at day +7 the median increase in PTX3 level was signifi cantly higher in patients experiencing early GVHD versus those who did not (2.5-fold versus 1.2-fold increase, P-value=0.014). Data at day+14 showed a similar although less marked trend (0.98-fold versus a 0.3-fold increase, P-value=0.045). Finally, we also investigated the role of PTX3 plasma levels in a murine model of acute GvHD. PTX3 increased in allogeneic transplanted mice upon GvHD occurrence, compared to syngeneic controls. Interestingly, PTX3 levels were higher in the allogeneic group also before GvHD onset. This fi nding is consistent with the evidence of our pediatric cohort and supports the predictive value of PTX3 for GvHD onset. Discussion: Our data showed that PTX3 plasma levels not only increased in patients experiencing GvHD, but they could also help in predicting patients at high risk for developing GvHD early after HSCT. Disclosure of Interest: None Declared. Introduction: The success of allogeneic stem cell transplantation, a standard therapy for hepatopoietic melignancies and inherited hematopoietic disorders, is limited by graft-versus-host disease (GVHD) morbidity and mortality. The impact of the microbiota on GVHD development is known to be signifi cant. Our previous studies revealed a highly signifi cant correlation of HPSE gene rs4693608 SNP with the risk of developing graft vs. host disease (GVHD). The discrepancy in this SNP between recipients and donors was found to be a more signifi cant factor for the risk of acute GVHD, than patient genotype. Moreover, heparanase is upregulated in response to pre-transplantation conditioning, followed by a gradual decrease thereafter. Expression of the HPSE gene correlated with the rs4693608 both before and after conditioning. Materials (or patients) and Methods: Analysis of HPSE gene expression before and after LPS treatment was performed in 128 umbilical cord blood (CB) samples and 104 peripheral blood (PB) samples from healthy individuals. Evaluation of heparanase expression in cell subsets was analyzed by RT-PCR and immunocytochemistry. Statistic analysis was performed using the NCSS software. Results: Heparanase (both mRNA and protein) is expressed primarily in PB and CB derived neutrophils, monocytes and macrophages. In addition, we found, for the fi rst time, that heparanase is expressed in NK cells. LPS treatment was found to up-regulate HPSE gene expression (P<10 -7 ) through TLR4. Our results indicated that monocytes are the key cells involved in the observed LPS, TLR4, HPSE cascade. Analysis of heparanase expression in resting MNCs did not reveal diff erences among individuals with various HPSE gene genotypes in both PB and CB samples Post-treatment heparanase expression correlated with rs4693608 SNP in both PB and CB MNCs. RQ (relative quantifi cation) of AA carriers after LPS treatment of PB MNCs was 16.3 (7.5-34.5) , while RQ of GG possessors was 6.4 (3.3-19.9 ), P=0.014. Analysis of HPSE gene expression in CB MNCs showed similar results (RQ of AA possessors was 24.9 (15.2-40.5), while RQ of GG carriers was 5.4 (3.1-13.0), P=0.0006). Ratio test revealed that MNCs with the AA genotype increased the level of heparanase to a higher extent than their counterparts with the GG genotype (P= 0.001 for PB MNCs and P=0.0006 for CB MNCs). Moreover, exposure of PB and CB MNCs with genotype AA and GG to increasing amounts of LPS revealed up-regulation of the HPSE gene in MNCs with the AA genotype, while MNCs with the GG genotype disclosed non-responsiveness to increasing amounts of LPS. Discussion: The present study indicated that the level of heparanase strongly correlates with the rs4693608 SNP and depend on cell type. The observed correlation between rs4693608 and heparanase expression in neutrophils and in LPS-treated MNCs suggests a yet undefi ned specifi c switch occurring only in activated cells. The present study helps to clarify how discrepancy in rs4693608 SNP between recipient and donor may elevate risk of acute GVHD. Disclosure of Interest: None Declared. Introduction: To add knowledge on the mechanisms contributing to acute GVHD, we fi rst established a chemotherapy-based minor mismatch mouse model. We then studied the proteins expression in the liver during GVHD using a mass spectrometry-based proteomic approach. Materials (or patients) and Methods: We used Bu/Cy conditioning in the LP/J → C57BL/6 model. At the time of maximum clinical manifestation of GVHD (day+23) the liver proteins were isolated. Quantifi cation data of labeled peptides were measured considering N-termini and lysine dimethylation on light (+28Da), medium (+32Da) or on heavy (+36Da) modifi cation per free primary amine. Protein and peptide quantitation information were extracted from MaxQuant 1.2.2.5. Protein interactions were studied with "String 9.05" (http://string-db.org/). Results: We compared the results of WT mice to syn-BMT recipients and to allo-BMT recipients. In total, 2238 proteins and 14489 peptides were quantifi ed by dimethylation labelling. 120 and 48 proteins were up (fc > 4) and down (fc >2.8) regulated in the liver of allo-BMT recipients as compared with syn-BMT recipients. Tap1 (28), ICAM1 (6), Cnx (4.6), H2-K1 (5.8) and H2-L (17) were up regulated; Iigp1 (13), Tgtp1 (38), Gbp2 (24) and Igtp (14) were remarkably overexpressed (Fig1, red circle). Another cluster of proteins, like Steap4 (10), Fmo5 (4) and Cytochrome P450 family proteins (around 4.5), were all signifi cantly overexpressed (Fig1, blue circle). Lrg1 and ICAM1 were signifi cantly up regulated in GVHD liver (Fig2). Discussion: During GVHD two clusters of proteins were over expressed, which contribute to activation of recipient antigenpresenting cells and donor T cells (Fig1, red circle), and are important to oxidation-reduction process and cellular metabolic process (Fig1, blue circle). Another cluster of proteins involved in cellular iron ion homeostasis was down regulated, possibly refl ecting a disturbed iron metabolism during GVHD. Interestingly ICAM1 and Lrg1, proteins that are involved in endothelial biology and angiogenesis, were signifi cantly up regulated. Further proteomics studies, with more time points/organs, may help us to elucidate the mechanisms, and ultimately may help to identify novel potential therapeutic targets for GVHD therapy. Disclosure of Interest: None Declared. Introduction: Chronic graft-versus-host-disease may aff ect a broad range of organs, including the skin, gut, liver, eyes and lung. Patients may exhibit single or multiple organ pathology. Factors which infl uence particular organ involvement are poorly understood. Migration molecules including integrins & chemokines have been illustrated to licence allogeneic T cell entry to target organs. We have previously reported associations between diff erential chemokine expression and cGVHD organ profi le (EBMT 2013, O325). We used a panel of chemokine receptor & tissue specifi c adhesion molecule expression profi les to determine whether organ involvement is refl ected in circulating eff ector T cell populations Materials (or patients) and Methods: Peripheral blood lymphocytes from 20 cGVHD adult patients were examined using fl ow cyto metry. Eff ector, memory & regulatory T cell populations were

S314 identifi ed. Expression of skin homing receptor, cutaneous lymphocyte antigen (CLA), gut tropic β7 integrin vascular addressin & a panel of chemokine receptors were examined & compared with normal controls (NC). 15/20 patients had active skin disease, with 6/15 exhibiting isolated cutaneous symptoms, 6/20 liver cGVHD, 5/20 current or recent gut GVHD. Results: The proportion of circulating CLA+ CD4+ eff ector T cells were signifi cantly elevated in patients with active skin disease (n=15), compared to patients without skin disease (n=5) (Median 13% vs 6%, P=<0.05). Proportion of CLA+ CD4+ eff ectors was highest in patients with isolated skin disease (n=6, median 18%). CD4+ eff ector T cells bearing chemokine receptor CCR5 were elevated in skin only disease patients compared with non-skin (Median 60% vs 20% (P=<0.05). In 5/6 skin patients with isolated skin disease a CLA+ CCR5+ CD25+ CCR7-CD4+ T cell population could be identifi ed which was absent in patients without skin disease. Proportions of CXCR3+, CCR5+ & dual CXCR3+ CCR5+, CD4+ eff ector T cells of a Th1 phenotype were all elevated compared to NC (P=<0.01), however no association with diff erent organ involvement could be determined. There was a trend toward higher proportions of CXCR6+ T cells in patients with liver or gut disease compared to those without (Median 3% vs 0.5%. P=0.10). Discussion: The skin homing receptor CLA and its interaction its ligand, E-selectin, represent an important axis in the ingress of T cells into the skin. The majority of CLA+ T cells are resident in the skin under resting conditions so whilst recruitment of additional T cells from circulation may be a feature of cutaneous infl ammatory conditions, it may not be prerequisite. However the presence of elevated proportions of CLA+ eff ector T cells in skin cGVHD patients, and in skin disease only patients, the existence of an activated phenotype CLA+ eff ector T cell population, which in lacking CCR7 is unlikely to recirculate to the lymph node, is supportive of a skin specifi c eff ector cell population which may be contributing to diff erential target organ pathology. The elevated eff ector CCR5 & CXCR3 T cell chemokine receptor expression is consistent with allogeneic expansion and cellular migration / tissue entry studies in GVHD (Palmer et al, 2009 , Croudace et al, 2012 . However, the lack of association of CXCR3+ CD4+ eff ector T cells with specifi c organ involvement may refl ect the diverse tissue production of CXCR3 ligand chemokines. Our fi ndings may contribute toward understanding diff erential organ phenotypes in cGVHD Disclosure of Interest: None Declared. Introduction: Graft versus Host Disease (GVHD) is a major complication of bone marrow transplantation therapy. Here we sought to investigate if BMSCs are ideal candidates for the treatment of human GVHD by utilizing our previously well-established mouse model of humanized x-GVHD. Materials (or patients) and Methods: Mice were allowed to develop lethal GVHD and monitored for >50% human Th1 cell engraftment. On suffi cient Th1 cell engraftment and disease development, murine recipients were treated with BMSC (2M per mouse) three times every four days. Control cohorts received irradiated allogeneic human monocytes (2M per mouse). At the end of treatment, mice were evaluated for weight loss, long-term survival and pathological changes in the GVHD target tissues. Results: We found that BMSCs were (1) effi cient at treating x-GVHD long term (2) diminished T cell numbers in the GVHD target organs and (3) maintained T cell immune competence ex vivo. Cohorts post BMSC treatment showed increased weight gain compared to control cohorts (P=0.04), diminished pathological damage of GVHD target organs (P=0.03), with signifi cant changes in the liver (P=0.0005), lung (P=0.03) and skin (P=0.05). Immunological parameters were measured in secondary lymphoid organs such as the spleen and BM. Interestingly, murine recipients that were treated with BMSCs showed signifi cant decrease in human cytokine production in the spleen (control vs. treated; IFNg γ ; P=0.04, TNF-α ; P=0.04). A similar phenotype was noted in the BM compartment reaching a trend towards signifi cance (control vs. treated; IFNg; P=0.07, TNF-α ; P=0.08). We next explored the immune-phenotype of human T cells in the GVHD target organs. In contrast to the secondary lymphoid organs, a signifi cant decrease in human Th1 cell number was noted in the liver of the treatment cohorts (P=0.01) with no diff erence in the cytokine profi le of the human Th1 cells. A similar trend towards decreased cell numbers were also noted in the lungs of the treated cohorts. Finally we sought to determine the mechanism by which BMSCs diminished GVHD in the treatment cohorts as compared to the controls. We found that the human T cells from the treatment cohorts exhibited diminished proliferation capacity as measured by Ki67 (P=0.06) and had a signifi cant decrease in Bcl 2 expression (P=0.01) as compared to the control cohorts. On ex-vivo stimulation, human Th1 cells from the treatment cohorts exhibited strong allo-responses as compared to the control cohorts (P=0.03). Discussion: These data taken together may suggest a novel mechanism by which BMSCs can inhibit human Th1 cell mediated x-GVHD. In summary, our study is the fi rst to assess the function of clinical grade BMSCs in inhibiting x-GVHD. The data presented here suggests that BMSC can effi ciently decrease infl ammation by decreasing T cell number while maintaining T cell immune competence. These results may suggest that BMSCs are ideal cellular therapeutic candidates for inhibiting Th1 mediated allo and autoimmunity. Disclosure of Interest: None Declared. Introduction: Defi ning the biodistribution and immunomodulatory mechanisms of human mesenchymal stromal cell (hMSCs) following allogeneic BMT (alloBMT) is critical to optimizing the clinical utility of these cells for the treatment and prevention of GvHD. We initially reported that hMSC can modulate murine T-cell (TC) alloreactivity in vitro and in vivo (ASH 2011). Importantly, hMSCs had no eff ect on CTL activity and potent GvL eff ects were preserved following hMSC administration. We now use novel, cryo-imaging technology to determine where hMSCs co-localize with donor T-cells to attenuate activation and proliferation. Materials (or patients) and Methods: Cryo-imaging provides highresolution anatomic imaging as well as single-cell detection and can facilitate unique quantitative analysis of fl uorescently-labeled cells within an organ of interest. In initial experiments, irradiated, B6D2F1 mice were IV injected with 1.0 x 10 6 fl uorescently labeled (Qtracker 625 red quantum dots) hMSCs on D1. Mice were sacrifi ced at several time points, embedded en bloc in optimal cutting temperature compound and snap frozen. In subsequent experiments, CFSE labeled donor (B6 or B6D2F1) splenic T cells were infused on D0 followed by the infusion of Q-dot labeled hMSCs on D1. Results: Six hours after infusion, fl uorescent cryo-imaging with single cell resolution demonstrated that the majority of labeled hMSCs can be found in the lung (83%) and liver (17%), with relatively few dispersed in other locations. At 24 hours (D2 after BMT) early re-distribution of cells to the liver and spleen was detected. Specifi cally, we found that hMSCs migrated to splenic marginal zones after BMT. Greater numbers of hMSCs were found in the spleens of alloBMT recipients (compared to syn) at all time points through 96hr (D4) after BMT (P <0.05). Interactive visualization enabled imaging hMSC co-localization with T-cells within the spleen at 48, 72 and 96 hours after BMT. For each mouse, white pulp volume was fi rst quantifi ed and then CFSE intensity was measured. Since CFSE dye intensity decreases with T-cell expansion, intensity histogram analysis of T-cell brightness served as an in vivo T-cell proliferation assay. Analysis revealed that T-cells in allo-BMT mice given hMSCs retained their CFSE fl uorescence intensity to a greater degree than allogeneic mice not given hMSCs. Correspondingly, whole spleen and white pulp volumes were also signifi cantly reduced compared to alloBMT controls. Finally, MLRs containing hMSCs demonstrated that decreased T-cell proliferation was associated with increased PGE 2 levels. Indomethicine and E-prostanoid 2 (EP2) receptor antagonism reversed, and EP2 agonism restored, hMSC-mediated T-cell suppression, confi rming a role for PGE 2 in vitro, and cyclo-oxygenase inhibition after alloBMT abrogated the protective eff ects of hMSCs in vivo. Discussion: In sum, novel, whole-mouse, cryo-imaging shows for the fi rst time that hMSCs migrate to the marginal zone of the spleen within 24 hours of administration. hMSCs subsequently accumulate, co-localize with donor T-cells in the splenic white pulp and attenuate alloreactive T-cell proliferation and expansion in a manner that is PGE-2 dependent and preserves allo-specifi c GvL responses. These results provide crucial insights connecting biodistribution with local regulatory eff ects of hMSCs that may serve to optimize the delivery of hMSCs to regulate clinical GVHD. Disclosure of Interest: None Declared. Introduction: Sickle cell disease (SCD) with approximately 300.000 newborns diagnosed each year is one of the most prevalent hematological diseases worldwide. Despite signifi cant advancement in the prevention and treatment of SCD-related complications, associated morbidity and mortality remains substantial. Allogeneic hematopoietic stem cell transplantation (SCT) is currently the only curative option. Until recently SCT was off ered to patients with advanced SCD only if an HLA identical sibling was available. As MSD are available only for a minority of patients haploidentical SCT is becoming an intriguing alternative. Materials (or patients) and Methods: Three adolescent and young adult patients with homozygous SCD, 1 male, two female (age 18, 17, 13 years) with a Turkish background and a high degree of disease burden were transplanted from a haploidentical donor.

Results: Conditioning consisted of ATG, Fludarabin, Thiotepa and Treosulfan in two patients, who achieved a complete donor chimerism without developing any relevant SCT related complications. The other patient received Campath, Fludarabin, Thiotepa and Melphalan. The graft of his mother contained only 2.93 x 10 6 /kg CD3/CD19 depleted cells as a result of poor mobilization. Due to graft failure the patient received an ineff ective boost of 4 x 10 6 /kg CD 34+ cells on day +30. The patient was reconditioned with ATG (Fresenius), Fludarabine, Cyclophosphamide and 2 Gy TBI and was retransplanted from his father with a total of 8.4 x 10 6 /kg CD3/CD19 depleted cells. The patient achieved a complete chimerism on day +20 and maintains currently a stable mixed chimerism without any further complications. In all three patients immunosuppression consisted of CsA and MMF. Transplant related morbidity was mild. Two patients developed aGvHD of the skin that resolved immediately with prednisolone alone. Currently the patients are 410, 260 and 92 days post SCT in excellent condition with stable chimerism of 89%, 100% and 100% respectively. Discussion: Socioeconomically the costs of conventional medical care outweigh by far the costs of SCT. Haploidentical SCT is a safe and curative option for patients with SCD, even in advanced stage and can be a feasible option for SCD patients with poor donor availability in developing countries. Disclosure of Interest: None Declared. 3 

Introduction: The Mucopolysaccharidosis VII is an inherited disease characterized by the cellular accumulation of undegraded glycosaminoglycans due to the defi ciency of the lisosomal enzyme B-glucuronidase. The usefulness of HSCT is due to the ability of hematopoietic cells to distribute and diff erentiate as microglial cells in the brain, alveolar macrophages in the lungs and Kuppfer cells in the liver and to a cross-correction phenomenon by which the engrafted cells provide normal enzyme to the neighboring defi cient cells. At our knlowledge only one case of MPS VII submitted to HSCT is described in literature. Materials (or patients) and Methods: We describe a case report of a two years old female. Clinically she presented skeletal deformations, hepatomegaly, umbilical hernia and a history of frequent upper respiratory infections. She did not show neurocognitive neither cardiac alteration. The height and weight were at a lower centile for the age (3 rd centile ). The diagnosis of MPS VII was carried out by the demonstration of high levels of glycosaminoglycans in urine, a low level of B-glucuronidase in leucocytes and mutations in the B-glucuronidase gene. In order to ameliorate skeletal problems and to avoid a possible neurocognitive involvement the patient was submitted to an allogeneic bone marrow transplantation from an unrelated donor with HLA identity. A reduced intensity conditioning regimen based on fl udarabine, melphalane and alemtuzumab was used. The prophylaxis of GVHD was based on cyclosporine and mofetil mycophenolate. After 1 year, because of the loss of the engraftment, she received a second stem cell transplantation. The source of progenitors was blood cord from HLA-matched unrelated donor. A myeloablative conditioning regimen based on busulfan (adjusted doses), cyclophosphamide and rabbit ATG was used. Cyclosporine was given for the GVHD prophylaxis. Results: After the fi rst transplantation the patient achieved total chimerism on day +26. The laboratory tests confi rmed progressive normalization of B-glucuronidase level from day +55 until day +272 when started decreasing. The chimerism was progressively lost with fi nal rejection of engraft. On day +300 post HSCT the B-glucuronidase level was <3% of normal. In order to improve the engraftment six infusions of donor lymphocytes were carried out, nevertheless no improvement was observed. After the second transplant full chimerism and normal B-glucuronidase level were achieved on day +26 and +116 respectively and persisted during the 48 months after transplant. Glycosaminoglycans in urine presented a progressive decrease. Discussion: HSCT may be a valid therapeutic option in patients with mucopolysaccharidosis VII. In our report total engraftment and normalization of B-glucuronidase level were achieved after both transplantations nevertheless only with a myeloablative conditioning regimen stable engraftment was maintained allowing persistent metabolic correction. Clinically we observed the stabilization of skeletal abnormalities, improvement of phenotype and resolution of hepatomegaly. During the 48 months after the second transplantation the patient has been showing a normal neurocognitive development and physical growing, being able to perform normal daily activities. Disclosure of Interest: None Declared.

S. Bhat 1,* , S. Badiger 1 , N. K.S 2 , S. Damodar 2 1 Pediatric Hematology and Oncology, 2 Hematology, Mazumdar Shaw Medical Center, Narayana Health City Bangalore India, Bangalore, India Introduction: Hematopoietic stem cell transplantation (HSCT) is the defi nite treatment for patients with thalassemia major. Materials (or patients) and Methods: We retrospectively analyzed the outcomes in thalassemia major patients classifi ed as Class 2 and Class 3 by the Pesaro criteria, who underwent allogeneic HSCT between November 2004 and December 2012. Results: Forty one patients (Males n=27; females n=14) with a median age of 6 years (range 3 to 22 years) underwent HSCT during this time. Thirty fi ve patients were classifi ed as Class 2 and 6 as Class 3. The conditioning regimens used consisted of Bu-Cy-ATG (n=24), Bu-Cy (n=2), Flu-Bu-Cy-ATG (n=8) and Treo-Flu-Thio (n=7) with cyclosporine and methotrexate as GVHD prophylaxis. Stem cell source was marrow, peripheral stem cells and cord in 33, 7 and 1 patients respectively. The median TC cell dose was 6x10 8 /kg (range 3 to 16.3 x10 8 /kg) and median CD34 cell dose of 3.8 x10 6 / kg (range 1.0 to 11.0 x10 6 /kg). Hematological recovery was seen in all patients except three patients, who died before engraftment. Neutrophil and platelet engraftment occurred at a median of 13 days (range, 12-20 days) and 18 days (range, 16-35 days), respectively. Sixteen patients developed veno-occlusive disease, 4 patients developed Gr II-IV acute graft-versus-host disease (GVHD), and 4 patients had chronic GVHD. At day +28, twenty two patients showed more than 90% donor chimersim and rest had mixed chimerism. Two patients experienced secondary graft rejection. Treatment-related mortality (TRM) for the whole cohort was 14.6% at 100 days and 19.5% at 6 months post transplant.

TRM was 14.2% and 50% for Class 2 and class 3 patients respectively. At a median follow up of 1144 days, overall survival and thalassemia free survival are 80.4%, and 75.6% respectively. Discussion: Conclusion: Similar outcomes have been reported from developed countries. Outcome of Class 3 patients still continues to be poor and VOD rates are high in this patient group. Disclosure of Interest: None Declared. Introduction: HSCT is considered the only curable treatment option for patients with thalassemia. Here, we report our hematopoietic stem cell transplantation experience in the treatment of thalassemia major patients, using PBSCT versus BMT. Materials (or patients) and Methods: From 1991 to 2013, 574 patients underwent HSCT in our centre. 221 patients underwent HSCT from BMT and 353 patients received PBSCT. The median age in the BMT group was 7years (2-26years) and in the PBSCT group was 8 years (2-29 years) (P-Value=0.001). In the BMT group, 89(40.3%) patients were class III, whereas in the PBSCT group 121(34.3%) were class III (P-Value=0.196). Results: Acute graft versus host disease (GvHD) occurred in 141(63.80%) and 253(71.70%) of patients in the BMT and PBSCT groups, respectively (P-Value=0.048). Chronic GvHD was 19.30% in the BMT and 32.70% in the PBSCT (P-Value=0.001) survivors after 100 days. With a median follow-up of 50 months, the 5-year thalassemia-free survival rate (TFS) of BMT and PBSCT was 76.3% and 67.5%, respectively (P-Value=0.294). The 5-year overall survival rate (OS) in the BMT and PBSCT groups was 80.1% and 73.8%, respectively (P-Value=0.119). The rate of rejection was 16.3% and 6.5% in the BMT and PBSCT groups, respectively. The most common causes of death were GvHD and infections in both groups. TFS and OS results stratifi ed by disease classes showed better survival rates in lower classes (Table1). Discussion: According to results of the study, HSCT is considered for younger patients with lower class of thalassemia. Appropriate GvHD prophylaxis regimen should be used to decrease the mortality rate in these patients. Compared to BMT, PBSCT with short time hospitalization has a lower cost and is an easy procedure to perform for donor Engraftment was faster in the PBST group, compared to BMT group. Acute and chronic GvHD were more frequent in the PBSCT group, but most of them were limited and manageable and chronic GvHD was more mild but was alleviated over time. Disclosure of Interest: None Declared. Introduction: After the early successes in therapy of Gaucher disease (GD), allogeneic hematopoietic stem cell transplantation (allo-HSCT) was completely abandoned in developed countries in the 1990s in favor of life-long medical treatment, the most often in form of enzyme replacement therapy (ERT). The aim of this study was to describe a long-term outcome of Swedish patients who successfully underwent allogeneic bone marrow transplantation (allo-BMT) for GD at Karolinska University Hospital between 1982-1991. Materials (or patients) and Methods: Six patients underwent allo-BMT because of severe GD; 4 patients for Norrbottnian, neuronopathic type 3 Gaucher disease (N-GD3) and 2 patients for non-neuronopathic, type 1 GD (GD1). The median age of the patients at the time of transplantation was 2.5 years (range 2-9 years). All patients were splenectomized before transplantation (two of them underwent partial splenectomy). Five patients underwent BMT from related donors (4 HLA-identical siblings and 1 from a one HLA-B antigen locus-mismatched father) and one patient was grafted from an unrelated donor (HLA-A, -B and -DR matched, with one -DP antigen mismatched graft). The myeloablative conditioning in all cases was based on BuCy in 4 patients and CyTBI in 2 patients. Results: All patients engrafted; however, the parental graft was rejected (the patient is still alive on ERT). Four patients developed mild acute GVHD, although requiring prednisolone therapy. One patient developed chronic extensive GVHD requiring combined immunosuppressive therapy with prednisolone, cyclosporine A, and PUVA in long periods over many years, until his death from sepsis 7 years after allo-HSCT. Four patients (2 F, 2 M) are still alive with fully functioning graft (2 pts with GD1 and 2 pts with N-GD3) after a mean and median follow-up time of 27 years (range 22-31 years). Unfortunately, both N-GD3 patients (2 F), with a follow-up time of 31 and 27 years after allo-BMT, developed epilepsy which is a debilitating sign of brain involvement in GD3 (after 16 and 22 years after allo-BMT, respectively). All four successfully transplanted patients have normal complete blood counts, no signs of visceral disease as well as skeletal and neurological status fully comparable with other Swedish patients with GD1 and N-GD3 receiving ERT. Analysis of plasma chitotriosidase activity (control range: <40 nkat/L), a biomarker of GD, disclosed completely normal levels (10-14 nkat/L) in patients who underwent allo-HSCT as compared with patients treated with ERT (96-1398 nkat/L). One woman with N-GD3 gave birth to a healthy child at the age of 20 years (18 years after allo-BMT). Discussion: The literature on the topic of allo-HSCT in GD is limited, restricted to only case reports and small series of patients. Our long-time results indicate that Gaucher disease patients can benefi t from allo-HSCT. It seems reasonable that, under some circumstances, allo-HSCT could off er a valuable treatment option for Gaucher patients failing medical therapy, as well as for patients suff ering from GD3, before development of neurological damage. However, as long as transplant-related mortality of allo-HSCT will not be close to 0% and the risk of GVHD remains signifi cant, it would be questionable to implement allo-HSCT for GD in developed countries, where safe but extremely expensive medical therapy is available for GD patients. Disclosure of Interest: None Declared. CsA (Cyclosporine A); Hb (hemoglobin in g/dL); RBC Tx (red blood cell transfusion); *Donor chimerism assessed on whole peripheral blood; § This patient received 5 donor lymphocyte infusions, the last of which on day +372 at a dose of 10 7 T-cells/kg but did not seem to improve until CsA was discontinued.

However, there is no published evidence on effi cacy and toxicity of treosulfan as part of conditioning regimens in patients with sickle cell anemia. We present seven children with sickle cell disease receiving a treosulfan-based conditioning regimen before 9/10 or 10/10 HLAmatched sibling donor BMT. Materials (or patients) and Methods: Children with homozygous sickle cell anemia, at least one severe sickle cell anemia related event and a well HLA-matched sibling donor were eligible for allogeneic BMT. The conditioning regimen consisted of treosulfan (3 x 14 g/m2), thiotepa (2 x 5 mg/kg), fl udarabine (5 x 30 mg/m2) and anti-thymocyte globulin (ATG Genzyme®, 10 mg/kg). GvHD prophylaxis was maintained with ciclosporine A, methotrexate, metronidazole and IVIGs. All patients received G-CSF from day +5 until engraftment. Results: Seven patients were enrolled in our centre. The chemotherapy was well tolerated and all patients showed successful engraftment. The median leukocyte engraftment was observed on day +13 (range 11 -17). Some patients experienced mild complications as mucositis or fever of unknown origin. Three patients developed transient skin GvHD grade I/II. There was no evidence of VOD in any of our patients. After a median follow-up of 16 months all children were alive, disease-free and free of GvHD. On day +100 four children had a complete donor chimerism and three children had a mixed donor-chimerism. Two of these children had a donorchimerism of >80% and >90%, respectively. Only one child who also suff ered from a glucose-6-phosphate dehydrogenase deficiency, had a declining donor-chimerism and required analgetics for diff use pain during an febrile infection of the upper respiratory tract early after transplantation while in rehabilitation. It is unlear if this was due to the remaining sickle cells at this point, as he has not encountered complications during following infections. None of the patients required erythrocyte transfusions or had direct evidence for any SCD-related event.

Discussion: To our knowledge we present the fi rst data of treosulfan, fl udarabine, thiotepa and ATG-G as conditioning regimen for children with sickle cell disease. Chemotherapy as well as the post-transplantation course proceeded without severe adverse events. All children showed a successful engraftment. There was no case of VOD. From these pilot data, we conclude that treosulfan-based conditioning prior to MFD BMT seems to be safe and eff ective in patients with sickle cell anemia. Larger cohorts are necessary to substantiate these observations. Introduction: Osteopetrosis is an heterogeneous group of bone remodelling disorders characterized by an increase in bone den-sity due to a defect in osteoclastic bone resorption. We previously presented a retrospective study of 20 children, 11 males and 9 females; the disease is diagnosed at a mean age of 4 months (range 10 d to 9 m). We found 9 children with mutations of the ATP6i (TCIRG1) gene, 2 patient of the CLCN7 gene, 3 of OSTM1 gene and one of the RANKL gene. The mean number of infused CD34+ cells was 18x10 6 /Kg, the number of TCR g/d+cells 66x10 6 /Kg and the number of TCR a/b 21.85 x10 6 /Kg obtained from the "non target" fraction. Overall the procedure allowed us, in the fi rst 2 cases, to achieve engraftment 100% donor by day 12 with no evidence of acute GVHD with a mean follow up of 5 months. One case too early to be evaluated. Discussion: This seems a promising procedure to enhance engraftment due to an high CD34+ inolculum combined with a good number of γ/δ TCR +cells in a genetic disease diffi cult to be grafted such as osteopetrosis. We have added a controlled number of α/b+ TCR +cells that in our previous experience didn't provoke GVHD in the MUD setting. Disclosure of Interest: None Declared.

Chimerism for fi rst HSCT patient was 95% at 2 years post HSCT, 57% at day +27 post HSCT for the third patient and 0% for the second patient after the second re-transplant. The second transplant patient at 6 months post HSCT had a total of 15 units of irradiated red cell concentrates, developed iron over load (serum ferritin of 1800ng/ml) and currently on oral iron chelator. He also had a total of 32 irradiated platelet concentrate and current parameters show a platelet count of 50,000/ul, hemoglobin of 7g/dl, WBC of 4500/ul and ANC of 1800/μl. Discussion: Postulated reasons for rejection include a lower dose of transplanted total bone marrow stem cells (nucleated cells of 5.7X10 8 /kg) compared to 9.8X10 8 /kg for the fi rst and 8.2X10 8 /kg for the third. Also there was minor ABO incompatibility between patient/donor (A+/O+) and the inability of the centre to monitor serum cyclosporine levels which was done for the third patient. Reported overall survival and disease free report from our centre is comparable with other centres. Introduction: Sickle cell disease is the most prevalent monogenic hereditary disease in Brazil. In 2001 a national newborn screening program was started in several states of the country. In 2005 the ministry of health launched the fi rst offi cial program for the treatment of SCD in Brazil. This program regulated the use of antibiotic prophylaxis, vaccination and family education. In 2009 the use of hydroxyurea was also introduced in the public health system for defi ned population of patients. We estimate that yearly about 3500 children with sickle cell diseases (SCD) are born in Brazil and more than 30 000 patients with the diseases are alive nowadays. Most of them came from the northern states of the country and from low incoming families. Stem cell transplant is the only curative approach but has been performed mostly in children.

Since the newborn screening program fi nally covered the whole country only in 2012, many patients with SCD in Brazil are adult candidates for stem cell transplantation (SCT). We started a SCT program at our center and accepted patients without upper age limit. Materials (or patients) and Methods: We have transplanted 16 patients (9 females and 7 males) transplanted between the ages of 8 and 38 (median age 18 years) and compared the outcome of patients older and younger than 18 years. Indications for transplant were: recurrent priapism (2 patients), neurological alterations (4 patients, one with MoyaMoya), repeated vasoocclusive episodes unresponsive to hydroxyurea (9 patients) and recurrent and severe acute chest syndrome with pulmonary hypertension (one patient). Each patient received bone marrow from a healthy HLA identical sibling; in 4 cases the donor had a sickle cell trait. Fifteen patients (the 38 years old patient with liver hemosiderosis grade 3) received myeloablative conditioning regimen consisting of Fludarabine 120 mg/kg, Busulfan per os 12 mg/kg and ATG 4.5 mg/kg, one patient with liver hemosiderosis grade 3 was conditioned with Fludarabine and Cyclophosphamide. GVHD prophylaxis consisted of CsA and a short course of MTX.

Results: All patients engrafted (median day for neutrophils 18 and platelets 22). The median follow up is 2 years and 15 of 16 patients are alive. One patient died on day 416 of hemorrhagic stroke as a result of Moya Moya. Severe acute GVHD was not observed, the incidence of grade I and II was 25%. None of the patients has yet developed chronic GVHD. Actuarial overall survival is 94%. We have not seen any diff erence in terms of complications and mortality in patients younger or older than 18 years using the same protocol. The death occurred in a 16 years old girl with severe neurological complications already before transplant. Discussion: In Brazil the incidence of SCD is high and transplantation is the only curative approach; it should be off ered for every patient with a complicated course independently of his age. Careful evaluation and preparation for transplant are essential for the best outcome; iron chelation, reduction of hemoglobin S, treatment of ulcers and prolonged use of anticonvulsant are indispensible. Patients should be off ered before irreversible damage has occurred. Disclosure of Interest: None Declared. Introduction: SCA and TM are associated with short survival. Allogeneic sibling HSCT off ers high overall and disease-free survivals (OS & DFS) when done in childhood and young adolescents. Studies in adults report signifi cantly lower eligibility rate due to signifi cant co-morbidities accumulated with age leading to higher transplant related mortality (TRM) and rejection. Reducing the intensity of conditioning (RIC) increased rejection while myeloablative conditioning (MAC) may be associated with increased toxicity and mortality. We aimed to study the feasibility and outcome of delivering steroids and cytoreduction to reduce rejection in a pre-conditioning phase followed by reduced toxicity conditioning (RTC) with lower doses of ATG in adolescents and young adults with SCA and TM. Materials (or patients) and Methods: 9 patients with SCA (n=8) and thalassemia major (n=1) (5 F/4M) with a median age 18 y (14-30y) transplanted between 1/11 and 10/213 from identical siblings (6 bone marrow, 2 peripheral blood and 1 both BM+PB) with a median dose of CD34+ cells of 5.79 x10 6 /kg (range 2.88-8.58). Preconditioning anti-rejection phase included intense cyto-reduction with hydroxyurea (to achieve an acceptable level of leukopenia (TLC >3,000/μl, ANC >1500-<3000/μl, Platelet >150-<250 x10 3 /μl), prednisone 0.5 mg/kg every other day for at least 3 weeks, hyper-transfusion and intense chelation. In case of avascular necrosis prednisone is replaced by Cellcept. Reduced Toxicity Conditioning: Fludarabine 40mg/m 2 /d on d-9 to d-6, Busulfan 0.8mg/kg/dose Q 6h x 14 doses (total 11.2mg/kg), ATG (Thymoglobulin; Genzyme) 1.5mg/kg/d d-4 to d-2 (total dose 4.5mg/kg). (1 patient received 7mg/kg and 2 received 5.5 mg/ kg),GVHD prophylaxis included: Cyclosporine (CSA) starting d-1 and ending 9 months after SCT and short course of Methotrexate. Results: 9 patients were followed for a median of 11 months ( 2-32 mo). All patients tolerated Immunmodulation and cytoreduction before conditioning with only one episode of short (3 days) neurto penic fever. TRM was 0% and all patients engrafted and living (OS 100%) and disease free (DFS 100%) and Transfusionfree (TFS 100%). Median time to ANC engraftment for the eligible patients (n:8) was 21 days (16-24) and for platelets 17 d (9-36 d) .

Only 2 of 7 evaluable patients developed grade II-III a-GVHD that completely resolved on steroid therapy and one of the 5 evaluable patients (20%) developed cGVHD that resolved on systemic steroids. 7 patients were evaluable for chimerism studies and all (100%) were chimeric with 100 % myeloid in 7/7 (100%) and lymphoid chimerism was 100% in 1 (14%), high (>50-95% in 4 (57%) and low (<50%) but stable in 2 (29% Introduction: Bacterial infections are an important cause of morbidity and mortality in patients receiving hematopoietic stem cell transplantation (HSCT). This justifi es the wide use of antibacterial prophylaxis with fl uoroquinolones in the period of neutropenia. This retrospective cohort study was performed to compare prophylaxis with levofl oxacin (LF) 500 mg oral per day versus no bacterial prophylaxis (no-LF). LF was administered from hospitalization for HSCT to the recovery of neutropenia (>1x10 9 /L) or the emergence of fever (intravenous antibiotic treatment Table 1 summarizes the main data in the LF and no-LF groups. No differences were observed with regards to the baseline characteristics in the two cohorts, as well as according to the type of HSCT. The rate of MDI with or without bacteremia was similar in the two groups; however, gram-negative bacteria were isolated more frequently in the no-LF cohort. Significant differences were observed with respect to the empirical use of antibiotics (imipenem+amikacin and teicoplanin was more used in the no-LF group). The incidence of resistant-bacteria was significantly higher in the LF group. No significant differences were observed on comparing of the outcome of infection in the two groups. Discussion: In this study no signifi cant diff erences were observed in the incidence of microbiologically documented infections and in the outcome of infections in HSCT patients according to the use of antibiotic prophylaxis with levofl oxacin or not. Prophylaxis with levofl oxacin reduced the frequency of infections by gram-negative bacteria in patients receiving HSCT. Supported by grant RD 12/0036/0029 from RTICC, Instituto Carlos lll, Spain. Disclosure of Interest: None Declared.

[PH-P465]

HIV infection, hematological malignancies and transplant recipients. The most eff ective prophylactic therapy is TMP-SMZ but its administration is often associated with intolerance and toxicity. Other prevention strategies as dapsone or inhaled pentamidine have been associated with higher incidence of pneumonia. Introduction: Prophylaxis of infections using High dose IgG after allogeneic transplantation do not decrease infectious risk, furthermore it may be associated to increased liver toxicity. IgMenriched immunoglobulin are able to neutralize endotoxin and, thus, it may reduce toxicity and TRM after allogeneic HSC transplantation. Materials (or patients) and Methods: 114 patients aff ected by acute leukemia and treated with allogeneic transplantation were studied. All patients received as anti-infectious prophylaxis:-IgM-enriched Immunoglobulin (Pentaglobin) at dosage of 200 mg/weekly for 6 doses, -prophylactic systemic antibiotics, -Total GI tract decontamination. 54 patients received also prophylaxis with Defi brotide as a further anti-toxicity. 81 patients were aff ected by AML and 33 by ALL, 66% received BM and 33% PBSC, in 65% donor was a HLA identical sibling (SIB) while in 35% was an Unrelated volunteers, 88% received a Myeloablative conditioning. 56% percent of patients were in 1st CR while 44 % in more advanced phase. Early Infections were categorized as: "FUO intervening early post-HSC infusion", "FUO after day +2", "Gram positive bacteremia or CVC infections", "Gram negative pneumonia or focal pneumonia". Results: No patients died because of early infections and overall TRM at 100 days was 5%. "FUO after day +2" was registered in 39% of patients and it was not associated to Conditioning type, Diagnosis, Advanced disease at transplant or to HSC source. In univariate analysis, "FUO (after day +2)" was signifi cantly associated with low immunoglobulin level of at admission (P=0.01), "FUO (after day +2)" was more frequent also after MUD transplants (51% versus 32%, P=0.05) and in patients who previously had a prior transplant (P=0.01). In multivariable analysis, only low level of serum immunoglobulin at admission remained signifi cant for an increased risk for this type of infection (P=0.02). Gram negative bacteremia or focal pneumonia interested 4.5% of patients and these types of infections were signifi cantly more frequent in MUD transplant in respect to SIB (9.5% vs 1.5%, P=0.04). A low grade (< 38.5 °C) and transient fever presenting early after infusion of HSC ("FUO intervening early post-HSC infusion"), interested 22% of patients (4% in family sibling and 55% in MUD, P=0.0001) no association of this type of FUO with immunoglobulin level, phase of disease or other factors was found. "Gram positive bacteremia or CVC infections" interested 26% of patients and was not associated to any factors. Overall incidence of Liver Sinusoidal Obstructive Syndrome (SOS) was 4.9%, no SOS graded as "severe" was registered. Patients presenting an increase in bilirubin above 2 mg/dl were 20%. The only factor associated to liver toxicity was advanced phase disease (P=0.004). Patients treated by Defi brotide had a reduction of liver toxicity, in terms of SOS and Bilirubin >2mg/dl, that, however, did not reach signifi cance. (17), non-Hodgkin lymphoma (n=6), severe aplastic anemia (n=12), neuroblastoma (n=5), inborn metabolic disorders (n=4). Their conditioning regimens were myeloablative (35%), or non-myeloablative (65%) of cases. HSCT was performed from related (16%), matched unrelated (63%), or haploidentical donors (21%). The patients received either bone marrow, or peripheral blood stem cells (resp., 47% and 53% of the cases). Second transplants were excluded from the analysis. Anti-infective treatment included pre-emptive acyclovir therapy and antimicrobial treatment at a standard schedule. Conventional GvHD prophylaxis was performed with Cyclosporin A and methotrexate, while ATG was applied in most cases. DNAbased diagnostics of T. Gondii (a total of 2975 tests) was performed in blood leukocytes on a weekly basis, and, if available, in сerebrospinal fl uid (CSF) cells, broncho-alveolar lavage (BAL), and urine sediments, using commercial gene-specifi c PCR kits, at a cutoff value of 1000 copies/mL, as verifi ed by quantitative PCR kits.

Results: PCR tests for T. Gondii DNA were positive in 13% of blood samples, 9% of CSF specimens, 11% of BAL samples, and 5% of urine sediments. We could not fi nd any signifi cant correlations between Toxoplasma activation and type of donors or source of transplanted cells for individual cases. Initial incidence of the pathogen was similar for diff erent age groups (1…4, 5…9, 10…14, 15…19, 20> years). T. Gondii activation (>2 consecutive positives within 100 days post-HSCT) was found in 12% of the patients, pointing to a special subgroup of patients prone to T. Gondii activation. Comparing subgroups of patients with diff erent conditioning regimens, we have shown a sharply increased T. Gondii incidence in the patients of 10 to 14 years old who received myeloablative (MA) treatment, when compared with (47) or not naïve (249): no significant difference was observed in overall survival (54% vs 50%) and cumulative incidence of acute (19% vs 25%) and chronic GVHD (18% vs 22%). Within familiar donor/recipient pairs the incidence of antiHBc+ was significantly higher in recipients (14% vs 5%, P<0.001). Forty-six of 296 (15%) patients were at risk of HBV reactivation at transplant: 2 HBsAg+ recipients, 35 antiHBc+ recipients, 1 HBsAg-recipient who received transplant from HBsAg+ donor, 9 HBsAg-/antiHBc-recipients transplanted from anti HBc+ and antiHBs-donors. Five (11%) of 46 patients at risk had HBV reactivation. The cumulative incidence of HBV reactivation was 21% at 8 years. The incidence of HBV reactivation was not influenced by donor type (identical related, identical unrelated, familiar haploidentical), MAC or RIC conditioning, ATG use, bone marrow or peripheral stem cells source, recipient age, recipient sex. HBV reactivation was associated with serum HBsAg+, HBV DNA more than 8 log10 copies per ml and an ALT increase 1.5-to 36-fold (median 4). HBV reactivation was observed in 3 patients after immunosuppressive therapy withdrawal and in 2 patients under immunosuppressive therapy for chronic GVHD. All five patients had genotype D of HBV. Three patients showed acute icteric hepatitis B that evolved in acute liver failure in one case. Shortly after the diagnosis of HBV reactivation, one patient underwent a liver biopsy which evidenced both mild HBV-related inflammation and minimal colangiopathy suggesting chronic GVHD. At a median follow-up of 30 months after HBV reactivation, three patients are undergoing continuous antiviral therapy with tenofovir or entecavir. Regarding the virological status, at last follow-up one of them is HBV chronic hepatitis active carrier and two are inactive carriers. One patient died of AML relapse at 27 months from reactivation, as active hepatitis B carrier. Another patient while undergoing lamivudine prophylaxis, developed fatal antiviral resistance unresponsive also to therapy with entecavir and died of acute liver failure after 7 months from reactivation. Introduction: Ebstein-Barr Virus (EBV)-associated post-transplant lymphoproliferative disease (PTLD) is a life-threatening complication after allogenic stem-cell transplantation (SCT). EBV monitoring was initiated in 2000 in our department as a standard procedure. In a previous study, we found a high frequency of EBVassociated diseases including PTLD, which resulted in reduction in the ATG dosage in conditioning. This study aimed to evaluate the current practice of screening for EBV in plasma with PCR among high risk patients the fi rst 100 days post-SCT and early pre-emptiv treatment with rituximab to prevent PTLD after the reduction of ATG dosage. Materials (or patients) and Methods: Patients who underwent allogeneic SCT with an unrelated donor, a mismatched related donor and patients who received a T-cell depleted graft were regarded as high risk for EBV-associated disease and were followed weekly with EBV-PCR in plasma for at least 100 days post SCT. All other patients were sampled at clinical suspicion of EBVassociated disease. All 222 patients who underwent allogeneic SCT between 2007 and 2012 were retrospectively evaluated regarding EBV-viral loads, clinical signs of EBV-disease and treatment with rituximab. Results: EBV-reactivation (defi ned as > 1000 genome equivalents (gEq)/ml in at least one test) was found in 39/222 (18%) of all patients. All patients with EBV-viremia had received ATG-containing conditioning (39/149, 26%), 26/101 myeloablative conditioning and 13/48 reduced intensity conditioning regimens. No patient without ATG as part of the conditioning was diagnosed with EBV-viremia (0/73). Of 39 patients with EBV-viremia, 26 patients received treatment with rituximab (median 3 doses, range 1-6). In 23 cases (88%) complete response was acquired. Of the 26 rituximab treated patients, 23 patients had clinical signs of EBV-associated disease as lymph node enlargement, fever, meningo-encephalitis or hemolysis, and in three cases biopsy proven PTLD were documented. Two patients with proven PTLD received additive treatment with chemotherapy, still one of them died. The 13 patients who did not receive treatment all resolved their EBV-viremia spontaneously. Discussion: After the reduction of ATG dosage, the biopsy proven cases of PTLD have been less prevalent (in this study 3/149 cases of PTLD in the high risk group compared to 3/15 before reduction of ATG). However, the incidence of EBV-associated diseases remains high (26%) for patients who received ATG-containing conditioning. Weekly monitoring in the high-risk group of EBV-PCR and early pre-emptive therapy with rituximab seems to be feasible. Disclosure of Interest: None Declared. cell (HSCT) or organ transplantation (SOT). Early or pre-emptive treatment interventions in high risk patients require both (i) improvement in manufacturing of virus-specifi c T cells without long-term ex vivo stimulation (e.g. by using IFN-γ cytokine capture system, CCS) maintaining antiviral CD4+ and CD8+ cells as well as (ii) quick recruitment of the respective T-cell donor including at least 3/6 HLA-matching. Materials (or patients) and Methods: Frequency assessments of antiviral T cells in HLA-typed healthy as well as in HSCT donors were performed over 4 years at Hannover Medical School using so far 17 diff erent overlapping peptide pools to detect CMV-, EBV-, ADV-, HHV6-and BK-specifi c T cells. Potential T-cell donors are fi rst identifi ed by IFN-g ELISPOT assay followed by a detailed phenotypical and functional analysis with multimer staining and cytokine secretion assay (CSA). Using CSA a donor was defi ned as eligible for donation of CMV-specifi c T cells (CMV-CTLs) if the number of CD3 + IFN-g + T cells was more than 0.03% of the CD3 + cells after 4 hour incubation with the respective peptide pool. Additionally, the potential enrichment of antiviral T cells was tested to predict eff ective enrichment of T cells of interest (TOIs) and should result in a purity of >60% of CD3 + IFN-g + T cells.

In order to obtain a manufacturing license according to the German Medicines Act (AMG), purifi cation of clinical grade CMV-CTLs from 3 healthy CMV-seropositive donors (covering the broad spectrum from low to high CMV-CTL frequency) was performed aseptically under GMP conditions using the CCS and the CMVpp65 overlapping peptide pool. Results: Frequency assessments of antiviral T cells in >400 HLAtyped healthy donors as well as in HSCT donors were performed over a period of 4 years at Hannover Medical School. Further quality control of three validation runs for CMV-CTLs starting with 0.05-1.7% IFN-g secreting CD3 + T cells resulted in 19.2%>81.2% of CD3 + IFN-g + T cells at the end of the manufacturing process. We were able to enrich a total of 0.3-1.8x10 6 CD3 + CD56 -CD45 + T cells. The total amount of CD3 + IFN-g + T cells in the fi nal product ranged from 0.54-14.2x10 5 cells with a purity between 19.2%>81.2% of CD3 + IFN-g + T cells. Among CD3 + T cells we found 11.5-68.0% CD8 + IFN-g + and 4.9-53.2% CD4 + IFN-g + cells, respectively. Despite this low purity of IFN-g secreting CD3 + T cells the amount of contaminating IFN-g -T cells was acceptably low (2.3-6.7x10 5 T cells). In all preparations we found contaminating B cells (4-5%), granulocytes (5-27%), monocytes (16-30%), and NK cells (12-21%) . Discussion: The manufacturing of antiviral T cells in the clinical scale using the CCS was validated at Hannover Medical School. To allow for a timely donor evaluation and also for a future registry the acceptance criterion for donor eligibility was a 3/6 HLA-match or better. So far the results obtained in the donors' pre-testing indicate, that a starting frequency of ≥0.03% might be suffi cient for a successful purifi cation of TOIs. Unfortunately the results obtained in pre-testing are not predictive for the effi ciency of antiviral T-cell isolation by CCS: It is likely, that a high proportion of antiviral CD4+ T cells negatively infl uence the purity and therefore the effi cacy of the antiviral cellular preparation. Disclosure of Interest: None Declared. (6), cord blood (3). The viral load was determined by quantitative PCR (Nanogen Advanced Diagnostic S.r.L) in cell-free body fl uids such as plasma, bronchoalveolar lavage (BAL), cerebrospinal fl uid (CSF), bone marrow (BM) aspirates or in gastrointestinal biopsies. Results: Median time from alloSCT to HHV-6 reactivation was 34 days (range: 0-705). Thirty-one patients presented HHV-6 positive in plasma, 9/54 in BM, 33/54 in gut biopsies or BAL, 7/54 in CSF. At the time of viral positivity all pts were receiving acyclovir as viral prophylaxis except fi ve. Twenty-nine patients had acute graft versus host disease (GvHD). Twenty-two out of these twenty-nine patients experienced a grade III-IV acute GvHD, requiring high dose steroids in twenty-six cases. A concomitant CMV positivity was detected in 15/54 patients. The median absolute count of CD3+ lymphocytes was 262 cells/mcl. In 52/54 cases we reported HHV-6 clinical manifestations: fever (43), skin rash (22), hepatitis (19), diarrhoea (24), encephalitis (10), BM suppression (18), delayed engraftment (11). HHV-6 positivity led to antiviral pharmacological treatment in 37/54 cases, using as fi rst choice therapy foscarnet. Amongst the total fi fty-four patients with documented HHV-6 positivity thirty-one solved the clinical event. However the mortality rate was relatively high in this population (only 30% of patients were alive), mainly related to severe infections or GvHD. A better overall survival (OS) is signifi cantly associated with CD3+ cells higher than 200/mcl (P-value 0.011) and time after allo-SCT more than 2 months (P-value 0.035). The overall survival is not signifi cantly modifi ed when patients with a better immune reconstitution (defi ned as CD3+ cells higher than 200/mcl) were separated in groups according to the time after transplant. In this analysis the overall survival was not signifi cantly infl uenced by steroids administration, presence of acute GvHD, plasma viral load and organ involvement. Discussion: This retrospective study confi rms a correlation of HHV-6 with high morbidity and mortality rates after alloSCT, underlying the importance of a regular HHV-6 monitoring in allo-SCT recipients. Despite HHV-6 detection typically occurred in the fi rst month after alloSCT, a better immune reconstitution has the potential to improve clinical outcome. Disclosure of Interest: None Declared. (14%) . Together, the two methods identifi ed a total of 586 microorganisms in 509 (26%) episodes: Gram-positive bacterial species (80%), Gram-negative bacterial species (17%), and fungal species (3%). Concordance between the two methods was 77%, with most of the discordant samples that tested negative by culture but positive using the molecular approach (50% of the total positive samples). The cases positive by SF alone were mostly samples from patients already receiving antimicrobial therapy, or, importantly, sample positive for fungal pathogens such as Aspergillus fumigatus, which is hard to detect by the traditional approaches.

Discussion: This analysis demonstrates a signifi cant correlation between the molecular test and the standard BC in haematological patients with febrile neutropenia. A molecular test such as SF, in combination with traditional assays, can play an important role in the diagnosis of sepsis, particularly in persistent fever despite antibacterial therapy, when a non-responding bacterial infection or an invasive fungal infection is suspected, therefore leading to a rapid diagnosis and an earlier targeted antimicrobial therapy. Disclosure of Interest: None Declared. Introduction: Long-term management of CMV infection after allogeneic stem cell transplantation (SCT) is based on the reconstitution of the specifi c T-cell immunity against the virus. In the clinical setting there are no well-established methods for the measurement of the CMV-specifi c immunity. Recently, a standardized quantitated assay was developed, which measures responses to CMV peptide antigens targeting mainly CD8+ cells HLA class I haplotypes, thus covering >98% of human population (Quantiferon-CMV, Cellestis). The aim of the study is to evaluate the levels and time of CMV-specifi c immunity reconstitution after SCT, in order to establish criteria for the management of CMV infection. Materials (or patients) and Methods: 26 adults (age 18-61) who underwent SCT from identical sibling (n=6), haploidentical sibling (n=1), unrelated donor (n=13) and dual cord blood (n=6) were enrolled in the study. Each recipient was examined before transplant and then monthly for the fi rst trimester post transplant. All patients were monitored for CMV replication with RealTime PCR once or twice per week. The pre-transplant CMV serostatus was as follows: D+/R+ (n=8), D+/R-(n=3), D-/R+ (n=11) and D-/R-(n=4). The latter group was excluded from the study.

Results: All seropositive recipients tested positive prior to transplantation with the Quantiferon CMV, with a median value of >10 IU/mL. Nine recipients with related (n=4) or unrelated donor (n=5) reconstituted their cell-mediated immunity against CMV early in the post transplant period (<3 months) and never experienced CMV infection. In some cases, low transient viral loads were detected that did not demand specifi c antiviral treatment. On the contrary, 8 out of 19 seropositive recipients or recipients with seropositive donor who did not reach levels of CMV Quantiferon >1 IU/mL during the follow-up period had multiple episodes of CMV reactivation with high viral loads, and even CMV-disease in one patient. Among the cord blood recipients, none reconstituted CMV-specifi c immunity during follow-up period, but one who developed limited immunity (0.67 IU/mL) after a year. Discussion: Evaluation of the level of CMV-specifi c immunity with the Quantiferon-CMV method seems to off er the ability to discern patients at high risk for developing CMV disease after SCT, as well as those who are capable to manage CMV reactivation on their own without specifi c antiviral treatment. Additionally, the use of the assay in everyday practice may contribute to the decision making for the follow-up schedule of the patients, thus probably resulting in more cost-eff ective strategies. Disclosure of Interest: None Declared. Introduction: CMV reactivation due to delayed specifi c immunological reconstitution causes signifi cant morbidity and mortality after HSCT. The use of adoptive CMV-specifi c T cell line (CMV-TCL) therapy is able to control CMV reactivation. The inclusion of adoptive therapy into routine clinical care requires effi cient methods for TCL generation. We investigated the effi cacy of CMV-TCL for the treatment of refractory high risk pts. Materials (or patients) and Methods: We performed a single center retrospective analysis of consecutive pts treated with allogeneic HSCT between February 2010 and October 2013. CMV reactivation was monitored using biweekly quantitative blood PCR starting from day +15 until day +100 and then weekly until day +180. A preemptive therapy with gancyclovir or foscarnet was started for number of ≥ 3,000 copies/mL. Pts receiving a haplo or cord blood unit (CBU) HSCT with at least 1 CMV reactivation, or any HSCT recipient experiencing at least 2 CMV reactivations, were considered at high risk. CMV-TCL were produced from the HSCT or from third-party family donors, according to a GMP protocol, by stimulation with a CMV peptide pool and culture for 24 days. Pts were treated on a compassionate use basis and an informed consent was obtained. Results: 118 allogeneic transplantations were performed: 47 from HLA-identical donor, 65 from haploidentical donor and 6 from CBUs. 76 pts were considered at high risk for CMV reactivation (71 because of donor type and 5 because of more than 2 reactivations). 35 out of 71 (49%) pts did not reactivate CMV; 23 pts (30%) had one reactivation, 18 (24%) more than one and 9 (12%) more than 2. TCL expansion was attempted in 24 donors (59%); in a third donor was CMV-seronegative. Target CMV-TCL dose was obtained from 22 donors (92%), in 2 cases TCL expansion failed to reach the target. Eight pts (20%) were infused with CMV-TCL. In those pts, 30 CMV reactivations were diagnosed. The median day of fi rst CMV reactivation and of fi rst infusion were respectively +40 (33-62) and +141 (77-511). 5 out of 8 pts were experiencing GVHD. Mean number of reactivations before receiving TCL was 3 (1-6). The mean total dose of infused cells was 3 x 10 5 /kg body weight. At the time of fi rst infusion, mean absolute lymphocyte count was 1.23 x10 3 /mL and PCR CMV was negative in all but one pt. After TCL infusion, 6 out of 8 pts had no more CMV reactivations, and 1 pt had 1 reactivation and 1 pt 2 other reactivations. After TCL infusion, one pt developed cGVHD that did not require treatment, and 2 pts experienced a late acute grade II GVHD at the time of discontinuation of immunosuppressive therapy. No pts died of CMV infection, at a median follow up of 288 days. Discussion: Adoptive T cell therapy with CMV-specifi c T cells is safe and eff ective in the management of refractory CMV infection, in adult patients after HSCT. However, only a small percentage of patients candidate for immunotherapy can be eff ectively treated. Disclosure of Interest: None Declared. Introduction: Human herpersvirus type 6 (HHV6) is a member of the β herpesvirus subfamily. More than 90% of the general population has been infected by HHV6. In immunocompromised patients, viral reactivation has been associated with fever, cutaneous rash, interstitial pneumonitis, encephalitis, and myelosuppression. The aim of this study was to evaluate the incidence of HHV6 reactivation in patients receiving ASCT with delayed hematopoietic reconstitution Materials (or patients) and Methods: We performed a single center retrospective analysis of 117 consecutive hematological patients treated with ASCT between February 2011 and October 2013. Patients with a delayed engraftment (defi ned as ANC<500/μL at day +14 after HSCT) were tested for quantitative blood PCR HHV6, CMV, and EBV. Conditioning regimen used was carmustine, etoposide, cytarabine, and melphalan (BEAM) in 45% and high-dose melphalan in 55% of cases. Results: 117 ASCT were performed and 62 of them presented a delay in engraftment. Among this group, 34 (55%) were tested and 7 out of 34 (21%) turned out to be HHV6+. Characteristics of the patients with engraftment delay are presented in Table. In 48% of cases, pre-transplant disease status was complete remission. In multivariate analysis, patients with POEMS syndrome reactivated HHV6 more than other diseases (50% vs. 21%, OR=5.00 (95% CI: 0.56-44.34, P=0.15); same trends were observed also with higher age (P=0.14) and lower CD34+ cells infused (P=0.13). Interestingly, no patient with Hodgkin's lymphoma reactivated the virus. Median time of HHV6 reactivation was day +20 (14-27) and

100% of patients experienced fever but no one had clinical HHV6related disease. 4 patients were treated with foscarnet and 3 with ganciclovir. Virus DNA integration was searched only in 3 cases and all resulted negative. Neutrophil engraftment (ANC>500/μL) occurred in 100% of HHV6+ patients after a median of 29 days and in 87.5% (95% CI: 74.5-100, P=0.73) of HHV6-patients after a median of 24 days; platelet engraftment (PLT>20,000/μL) occurred in 57% (95% CI: 20-94) of HHV6+ vs. 72.5% (95% CI: 53.5-90. Introduction: In one of the largest mono-institutional series of HIV-1 positive (HIV+) relapse/refractoring lymphoma patients (pts) submitted to Autologous Stem Cell Transplantation (ASCT), EBV and KSHV-DNA loads were evaluated in order to assess the frequency of γ-herpesviruses reactivation and their prognostic and predictive value during the ASCT procedure and follow-up. Materials (or patients) and Methods: 22 relapse/refractory HIV+ lymphoma pts who underwent ASCT at the National Cancer Institute (Aviano, Italy) and with at least 2 months (mo) follow-up after ASCT were included in this retrospective immunovirological study. EBV-and KSHV-DNA were measured by RT TaqMan PCR using the ABI PRISM 7900 HT SDS (Applied Biosystems). The HIV RNA level was quantifi ed by using the Versant HIV-1 RNA 3.0 assay kit (bDNA; Bayer Diagnostics). CD4+ T, CD8+ T, CD56+ and CD19+ lymphocyte subsets were evaluated by fl ow cytometry (EPICS XL -Beckman -Coulter). Immunovirological parameters were evaluated before and after debulking chemotherapy (DCT), and at mo 0.5, 1, 3, 6, 12 after ASCT. The median follow-up of the pts after ASCT was 49.75 mo (range 2-108.91 mo). Results: Pre-treatment plasma and PBMCs EBV-DNA were detectable in 12 (median 12135 copies/mL) and 18 pts (median 417 copies/10 6 PBMCs). After DCT, EBV-DNA viremia in both compartments predicted overall survival (HR, 5.88, 95% CI, 1.40-24.66, P=0.02; HR, 5.52, 95% CI, 1.08-28.16, P=0.04, respectively). Moreover, response <75% to DCT was associated to persistent EBV-DNA detection in plasma (P=0.03). A positive signifi cant correlation between EBV-DNA in PBMCs and residual post-DCT B cell counts and percentage was found (r=0.69, P=0.002; r=0.6 P=0.009) and a negative correlation between plasma EBV-DNA and CD4+ T cell percentage and CD4/CD8 T cell ratio (r=-0.55, P=0.01; r=-0.45 P=0.04) was also observed. After ASCT, plasma EBV-DNA was detectable in 5/10 pts who died very early after transplantation. No occurrence of lymphoproliferations or of life threatening clinical complications was diagnosed for the remaining 12 pts during year 1 after transplantation and afterward. In these pts, EBV-DNA in plasma was always negative, except for 2 sporadic increase, and EBV-DNA in PBMCS was detected with increasing frequency from mo 0.5 to 12 after ASCT (4/12 pts at mo 0.5, 11/12 pts at mo 12, P=0.01). At this time point, median EBV-DNA in PBMCs was 316 copies/10 6 PBMCs (range: undetectable-2083 copies/10 6 PBMCs) and no correlation to B cell counts was found (r=0.07, P=0.83); the inverse correlation with CD4/CD8 T cell ratio was kept like after DCT (r=-0.74 P=0.006). KSHV-DNA load was detected in plasma samples of 3/16 pts before DCT (18.8%) and 2 of them showed detectable KSHV-DNA also in PBMCs. After DCT, a decrease in plasma KSHV load, ranging from 83 to 100%, was observed in all 3 pts, while KSHV-DNA in PBMCs became undetectable. After ASCT, KSHV-DNA was positive only in 1 patient who died for evolution of the initially diagnosed plasmablastic NHL to Primary Eff usion Lymphoma. Discussion: The presented fi ndings underline that γ-Herpesvirus load is associated with deep immune defi ciency and with precocious mortality after ASCT in HIV+ pts. This suggests the utility of EBV-and KSHV-DNA monitoring as a simple complementary tool for the management of HIV+ lymphoma pts submitted to ASCT. Disclosure of Interest: None Declared. Introduction: Invasive aspergillosis (IA) remains a major concern among patients with hematological malignancies who receive cytotoxic chemotherapy or allogeneic stem cell transplantation (allo-SCT). With a view toward better understanding the diff erences in terms of clinical signifi cance and outcomes of IA between allo-SCT recipients and patients treated for leukemia, we report a single-center study of 739 unselected consecutive patients treated in our unit between August 2000 and February 2004. IA episodes were classifi ed according to the 2008 EORTC/MSG criteria. Materials (or patients) and Methods: Patients were hospitalized either for allo-SCT (n=135), or for the treatment of acute leukemia (n=378), chronic lymphocytic leukemia (n=116), or myelodysplasic syndrome (n=105). Probable or confi rmed IA were observed in 29 patients, among whom 7 were undergoing allo-SCT (5.2%), 20 were treated for acute leukemia (5.3%), 1 for chronic lymphocytic leukemia (0.8%), and 1 for myelodysplastic syndrome (0.95%).

Most of the clinical signs were mild and nonspecifi c, especially in allo-SCT recipients. Results: Comparison of patients undergoing allo-SCT to the others revealed signifi cant diff erences. In allo-SCT recipients, the onset of IA occurred later than in the other patients, after the neutropenic Introduction: Despite anti-infective prophylaxis, bacterial, fungal and viral infections remain an important complication after allogeneic stem cell transplantation (SCT). Recovery from infections depends on the complete integration and balance of innate and adaptive immune responses, which may aff ect survival. In this complex interplay, Toll-like receptors (TLRs) play a key role and recognize pathogen-associated molecular patterns (PAMPs), such as common protein, carbohydrate or DNA/RNA pattern motifs. Extracellular PAMPs, especially of bacteria and fungi, are recognized by surface TLRs (TLR-1,TLR-2,TLR-4,TLR-5, and TLR-6). Intracellular TLRs (TLR-3,TLR-7,TLR-8 and TLR-9) bind mainly to foreign nucleic acids. To our knowledge, no studies deal with expression and function of all human TLRs together in relation to infectious complications in the setting of allogeneic SCT. In this study we analyse 9 TLRs on T-lymphocytes and monocytes of transplanted patients in relation to bacterial, fungal, viral infections and outcome. Materials (or patients) and Methods: Bacterial infections included all blood stream infections with or without organ localization. Fungal infections were evaluated and defi ned according to the revised criteria of EORTC/MSG Consensus Group. CMV, EBV and HHV6 reactivations/infections were monitored weekly by quantitative real-time PCR until the third month after SCT. The expression of TLRs on T-lymphocytes and monocytes was analysed in 35 patients by fl ow cytometry as mean fl uorescence intensity at day +30. Functional data were obtained by ELISA assay after TLRs activation. The cell supernatants were collected and assayed for TNFalpha, IL-4, IFN-gamma, and MCP-1. Lymphocyte subsets were evaluated by fl ow cytometry. Results: Clinical/transplant characteristics and infections before day +30 did not infl uence levels of TLRs (multifactor analysis of variance). In multivariate Cox regression analysis, levels of TLR-9 expression on T-lymphocytes (HR 0.97; P=0.01) and values of NK cells (HR 0.95; P=0.01) correlated negatively with bacterial infections after day +30. We observed a trend for negative correlation between TLR-7 levels on T-lymphocytes and fungal infections after day +30 (HR 0.36; P=0.07). Values of monocytes were negatively associated with CMV reactivation (HR 0.99; P=0.03), whereas levels of TLR-5 on T lymphocytes were positive predictors (HR 1,2; P=0.01). Age and bacterial infections negatively infl uenced overall survival (HR 1.06 P=0.03; HR 10 P=0.006). Monocyte values were positive predictors of survival (HR 0.96 P=0.003). Discussion: Bacterial, fungal and CMV infections were associated with a diff erent expression of some TLRs. The protective role of TLR-7 and TLR-9 against fungal and bacterial infections, respectively, seems to be dominant in comparison to other known TLRs, which recognize these pathogens. The atypical involvement of TLR-5 in the response against CMV may suggest more complex and pleiotropic functions of TLR-5 in the immune reactions. A specifi c TLR profi le and an adequate recovery of monocytes could positively infl uence the outcome of allogeneic SCT by promoting an eff ective control of infections and infl ammatory responses. Disclosure of Interest: None Declared. Introduction: Several recent studies have shown that HHV-6 DNA is frequently found in blood (HHV-6 DNAemia) of hematological stem cell transplantation (SCT) recipients. Reactivations of HHV-6 are associated with the development of acute Graft versus Host Disease (GvHD), CMV reactivations, poor engraftment and higher overall mortality. Ongoing discussions revolve around the nature of the association, i.e. whether HHV-6 triggers the other complications, whether HHV-6 DNAemia is caused by the above mentioned complications, or whether HHV-6 DNAemia has no signifi cance at all. In this observational study we investigate not only the presence of HHV-6 DNA in blood of SCT recipients, but also the duration and the height of the HHV-6 DNAemia, and how these factors relate to the occurrence of GvHD. Materials (or patients) and Methods: Frozen whole blood samples from 50 consecutive SCT patients were tested for HHV-6 by quantitative PCR. All samples taken until one year after transplantation were analyzed retrospectively (total 961 samples). Patient data about the underlying hematological disease, type of transplantation (myelo-ablative, reduced intensity, cord blood, sibling, or matched unrelated donations) were recorded from existing medical fi les, as well as the occurrence of acute or chronic GvHD and overall outcome. Results: Forty-six of 50 patients had at least one blood sample which tested positive for HHV-6 DNA. 25 patients had occasional samples with low level HHV-6 DNAemia (<4 consecutive positive samples with HHV-6 DNA <1000 copies/ml). 21 patients had consistent HHV-6 DNAemia (≥4 consecutive HHV-6-DNA positive samples). Eighteen out of these 21 patients had DNA levels >1000 copies/ml for at least 4 consecutive blood samples. In the group of patients with consistent HHV-6 DNAemia, 13 out of 21 (62%) were diagnosed with acute GvHD, and 8 of these patients had clinical presentations compatible with viral illness or GvHD at the time of the fi rst blood sample with detectable HHV-6 DNA. In the group of patients with occasional HHV-6 DNAemia, 6 out of 25 (24%) were diagnosed with acute GvHD (P<0.05). Discussion: While HHV-6 DNAemia is a frequent fi nding following hematological SCT, the pattern of HHV-6 DNAemia (either occasional or consistent) may be the most important predictor of complications. Routine monitoring of HHV-6 DNAemia is necessary to distinguish patients with occasional HHV-6 DNAemia form those with consistent HHV-6 DNAemia. Disclosure of Interest: None Declared. Introduction: Thanks to recent advances, a platform of T-cell replete haploidentical hematopoietic stem cell transplantation (HSCT) using post-transplant cyclophosphamide (as a part of graft-versus-host disease prophylaxis) is now available, and it is characterized by a high reproducibility and an acceptable safety profi le. However, detailed data on the timing and etiology of infections occurring after such a transplant approach are still limited. Materials (or patients) and Methods: Present analysis reported data on 40 consecutive haplo-HSCT and compared them with a cohort of 72 consecutive HSCTs from a HLA-identical donor treated at the same center after a reduced-intensity or nonmyeloablative conditioning, with the aim of reporting etiology and timing of infections occurring after HSCT in both groups. For haplo-HSCT, antimicrobial prophylaxis consisted of acyclovir, levofl oxacine, caspofungin followed by itraconazole, and cotrimoxazole until day -2. In HLA-identical group acyclovir, levofl oxacin, fl uconazole and cotrimoxazole were administered.

Results: In haplo cohort, a total of 38 patients presented at least one infectious episode: 22 had bacterial (55%), 28 had viral (70%) and 5 (12.5%) had at least one fungal infection. In particular, we found a 72% prevalence of viral infections/reactivations between day+30 and +100 with a median of viral events per patient of 2 (range [1] [2] [3] [4] . Twenty episodes of CMV reactivation occurred in 17 patients (42%), BK-virus-associated cystitis was observed in six patients (15%). Interestingly, beyond day +365 only fi ve bacterial events were observed out of 24 patients at risk (21%) and no fungal infections were detected. Incidence of viral infections or reactivations was higher in haplo group compared with HLAidentical one, while late bacterial events occurred less frequently (prevalence 21% vs. 55% beyond day +365), probably due at least in part to low rate of chronic GVHD among haplo-HSCTs (10% vs. 51%, P<0.0001). Two-year NRM was 10% (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) and 25% in HLA-id and haplo cohort respectively (P=0.02). Discussion: Despite a quite high rate of viral infections in the early period, present data suggest a satisfactory immunological recovery after T-cell replete haplo-HSCT using post-transplant CTX, in particular with few late bacterial events and no fungal infections beyond one year after HSCT. Disclosure of Interest: None Declared. Introduction: Posaconazole (posa) is used for prophylaxis against invasive fungal infections in BMT. There is large interpatient variability in posaconazole kinetics as well as an established concentration-eff ect relationship, which makes it an ideal candidate for therapeutic drug monitoring (TDM). It has been recommended to maintain trough concentration >0.7 μg/ml for prophylaxis against invasive mycosis. We report our experience.

Materials (or patients) and Methods: All patients who received posa prophylaxis during peritransplant period between September 2013 and December 2013 were included. Posa prophylaxis was started at least 3 days prior to start of conditioning regimen in all patients. All patients received standard oral dose of 200 mg TDS. Posa levels were measured on day 7 of treatment by HPLC assay. Subsequent levels were done at physician's discretion or a week after dose modifi cation, if any. If Cmin was <0.7 μg/ml, frequency of dosing was increased as per standard therapeutic guidelines for posa. Posa was continued for at least 100 days for all patients undergoing transplant and longer if patients were on steroids due to engraftment fever or GVHD. Breakthrough fungal infection was documented as defi nite, probable or possible as per standard guidelines. Infl uence of covariates included age, sex, height, weight, body mass index (BMI), diarrhea, liver and renal function parameters and concomitant drugs were analyzed by linear regression with backward elimination. Results: Twenty-nine patients were monitored during this period The median age was 33 (10-57) years. The median number of TDM was 3 (1-8) per patient. High interpatient variability was observed (CV=97%). Seven patients (24.13%) had trough levels < 0.7 μg/mL after the fi rst monitoring. Six patients had their dose modifi ed after fi rst monitoring due to low levels. All of them achieved target trough concentration after dose modifi cation. Six patients who had adequate levels at fi rst TDM had fall in Cmin below 0.7 μg/mL during subsequent TDMs. Introduction of steroids signifi cantly reduced posa trough levels [N=4; (presteroid median Cmin-2.035 μg/mL (1.12-2.78 ); post steroid median-0.94μg/mL (0.5-1.59); P=0.06 -Wilcoxon sign rank test)]. None of the demographic or laboratory parameters infl uenced posa levels. Breakthrough fungal infections were seen in 2 patients (defi nite-1, possible-1). One of them had undetectable levels and the other had adequate but decreasing trend in previous Cmin levels. Discussion: About one in fi ve patients required dose modifi cation after fi rst TDM. The incidence of breakthrough fungal infections was approximately 7% which is comparable with other centres in the West. Introduction of steroid may lead to fall in Cmin levels. Such patients would require a repeat TDM and appropriate dose correction, if the levels are low. Interestingly, none of the baseline characteristics and laboratory parameters infl uenced posa level.

Other studies have shown diarrhea and concomitant medication (PPIs) to have an impact on posa levels. This needs to be confi rmed in our setting in larger sample size. Pharmacoeconomic analysis showed that as little as Rs. 2,500 (approximately 30 euros) spent on TDM would benefi t one patient. Thus TDM could be adopted even in resource limited settings to reduce breakthrough fungal infection in BMT patients on posa prophylaxis. Disclosure of Interest: None Declared. Introduction: Most of the therapeutic drug monitoring (TDM) studies of voriconazole applied the single blood concentration of each patient. The aim of this study is to elucidate the dynamics of blood concentrations of voriconazole by serial monitoring in hematopoietic stem cell transplantation (HSCT) recipients. Materials (or patients) and Methods: Thirty-eight patients who received voriconazole (400 mg/day) orally after allogeneic HSCT for hematological diseases with weekly monitoring of plasma voriconazole concentration were identifi ed by using the institutional database and enrolled into the analysis. Data, including patient demographics and voriconazole concentration, were retrospectively collected from the medical records. Plasma trough concentrations of VCZ were measured by high-performance liquid chromatography weekly starting 5-10 days after initiating voriconazole administration.

Results: Median age and body weight of the patients were 44.5 years (range: 22-66) and 51.7 kg (range; 43.1-71.5), respectively. The plasma voriconazole concentration initially measured 5-10 days after starting voriconazole was 1.69+1.10 μg/ml, which was less than 1.0 μg/ml, defi ned as a low concentration, in 12 patients (32%, Group A), and 1.0-5.0 μg/ml, defi ned as an optimal concentration, in 26 patients (68%, Group B). In Group A, 6 patients subsequently achieved an optimal concentration 2-4 weeks after initiating voriconazole without an adjustment of the voriconazole dose. No patients showed voriconazole concentration greater than 5 μg/ml at any time points in Group A. In Group B, voriconazole concentration decreased to a low concentration in 4 patients at the second week and exceeded 5.0 μg/ml in 2 patients 4-5 weeks after initiating voriconazole. In 6 patients whose voriconazole concentration was monitored longer than 8 weeks, the concentration was at a steady state and no accumulation was observed. Discussion: These results suggest that the voriconazole concentration measured 5-10 days after initiating voriconazole could predict its subsequent dynamics. However, since a proportion of patients attain a steady state at later than 4-5 weeks, a serial monitoring of voriconazole concentration is required to elucidate the optimal dose in each patient to prevent toxicities and maximize its effi cacy. Disclosure of Interest: None Declared. Introduction: Though end-organ disease has radically reduced, CMV DNAemia continues to be a frequent event after Allo-SCT with potential deleterious eff ects. Experimental evidence suggests that polyfunctional CD8 + T cells are critical in the control of some chronic viral infections. Aims: 1. To determine the functional profi le of CMV-specifi c CD8 + T cells that associate with protection from and control of CMV DNAemia after Allo-SCT, 2. To investigate whether NKG2C + NK cells contribute to aff ording protection. Materials (or patients) and Methods: We prospectively analysed 108 non-consecutive pts from Apr-2010 to May-2012. We enumerated pp65 and IE-1-specifi c CD8+ T cells expressing IFN-γ, TNF-α and CD107a, and NKG2C+ NK cells, at days +30 and +60 after transplant. CMV DNAemia was quantitated once or twice weekly by CMV RT-PCR, (Abbott Molecular or LightCycler CMV Quant Kit, Roche). Quantitation of monofunctional (IFN-γ, TNF-α, and CD107a), bifunctional (IFN-γ/ TNF-α, IFN-γ/CD107a, and TNFα/CD107a) and trifunctional (IFN-γ/TNF-α/CD107a) CMV-specifi c CD8+ T cells was performed by ICS using whole blood. Specifi c responses were considered as those that were >0.1% for IFN-γ, TNF-α, and CD107a-expressing CD8+ T-cell populations. Polyfunctional CD8+ T cells those positive for 2-3 markers. NK cell immunophenotyping was performed in EDTA WB samples in a FACScant II fl ow cytometer (BD Biosciences). Patients with active CMV infection were pre-emptively treated with valganciclovir (900 mg/12 h), ganciclovir (5 mg/kg/12 h) or foscarnet (90 mg/kg/12 h) upon detection of >500 CMV DNA copies/mL. Antiviral therapy was discontinued after two consecutive negative PCR results. Results: Fifty-nine out of 108 pts (54.6%) did develop CMV DNAemia within the study period. CMV-specifi c CD8+ T-cell responses (of any functional type) were more likely to be detected in patients not developing CMV DNAemia (P=0.04). Qualitatively, no major diff erences in the functional signature of CMV-specifi c CD8+ T cells were noted between patents who had or had not CMV DNAemia. However, the level of polyfunctional CD8+ T cells at day +30 in patients who had a subsequent episode of CMV DNAemia was lower than in patients who did not, though not achieving statistical signifi cance (P=0.32). The presence of bifunctional and trifunctional CD8+ T cells was associated with lower levels of CMV replication, and higher frequency of self-resolved episodes. The levels of NKG2C+ NK cells were comparable in patients with or without CMV DNAemia (P>0.1), and in those with or without reconstituted polyfunctional CD8+ T-cell responses (P=0.4). Discussion: Our data indicate that the enumeration of IFN-γproducing CD8+ T cells permits an estimation of the polyfunctionality of the CD8+ T cell population, and should therefore be considered a reliable surrogate marker for predicting protection against the occurrence of CMV DNAemia. Our results do not confi rm a direct role of NKG2C+ NK cells in CMV control or in promoting the reconstitution of CMV-specifi c polyfunctional CD8+ T-cell responses. The data reported further clarify the immune mechanisms involved in preventing the occurrence and in control of CMV DNAemia in Allo-SCT recipients, and may be helpful in the design of therapeutic strategies aimed at optimizing the management of CMV infection in the clinical setting Disclosure of Interest: None Declared. Introduction: Hematopoietic stem cell transplantation (HSCT) is a potentially curative procedure to many hematological diseases. Human herpesviruses may cause severe complications after HSCT such as interstitial pneumonia, encephalitis and post-transplant lymphoproliferative disease (PTLD). A prospective survey on the incidence of primary infection or reactivation and clinical features of herpesvirus infections after HSCT has not yet been performed in Brazilian patients. Additionally, the impact of most of these infections on the HSCT outcome is still unclear. This study aimed to develop a test to screen and quantify all known human herpesviruses (HSV1, HSV2, VZV, EBV, CMV, HHV6, HHV7 and HHV8) in plasma samples from patients undergoing a HSCT and evaluate the impact of these infections on hematopoietic transplantations outcomes. Materials (or patients) and Methods: Between August 2010 and December 2012, peripheral blood samples from 98 allogeneic HSCT recipients were collected weekly after transplant until day +100, from four transplant centers in Sao Paulo (Brazil), totalizing 1045 samples.Median age was 16 years (range: 2-73), 61% were male, and acute leukemias were the most frequent diagnosis (55%). Stem cell sources were bone marrow in 54%, umbilical cord blood in 25% and mobilized peripheral blood in 21%. 40% percent of donors were related identical. In a semi-automated workfl ow, the DNA was extracted from plasma in the QIAcube robot. A test based on quantitative real-time PCR (Taqman®) was optimized to screen and quantify all known human herpesviruses. Infected cell cultures and plasma specimens with a known viral load/amplicon copy number have been used as controls. For all the viruses the lower limit of detection (LOD) was around 5 copies of target per reaction, representing 250 copies/ml of plasma. No cross-reaction or false positive results were detected and within-run and between-run precision estimates are equal or higher than 95%. Results: The frequencies of herpesviruses reactivation or primoinfection were: CMV=43%, HHV6=18%, HHV8=6%, EBV=3%, HSV1=3%, VZV=3%, HHV7=2%, and HSV2=1%. CMV reactivation was signifi cantly more frequent in adults (71% vs. 26% for children, P<0.0001). HHV6 reactivation was signifi cantly more frequent after umbilical cord blood transplant than after transplant from other sources (41% vs. 6%, respectively, P<0.0001). CMV reactivation was associated with a higher risk of acute GVHD, with a cumulative incidence at 100 days of 35% vs. 16% (P=0.04), but had no impact on the other outcomes. HHV6 reactivation also had no signifi cant impact on outcomes. HHV8 reactivation was associated with an increased risk of chronic GVHD (83% vs. 48%, P=0.001). Discussion: HHV6 primo-infection or reactivation is more frequent after transplant from umbilical cord blood than from adult donor grafts. CMV and HHV6 reactivation are frequent after HSCT, but had no signifi cantly impact on the transplant outcomes, possibly due to monitoring and preemptive measures. Monitoring these viruses constitute an essential measure to improve outcomes. Additional prospective studies are required to confi rm these fi ndings and determine whether therapy infl uences survival. Disclosure of Interest: None Declared. Introduction: EBV-associated lympoproliferative disorders (EBV-LPD) in the setting of allogeneic haematopoetic transplantation (alloHCT) occur with an increasing incidence due to innovations in transplant techniques. Several risk factors have been correlated with a higher probability of EBV-LPD. Serial monitoring and pre-emptive treatment with monoclonal anti-CD20 antibody in patients with elevated EBV viral load has proven eff ective in reducing the risk of EBV-PTLD in transplanted patients. Although a donor sero-positivity has been recognized as a risk factor for EBV-LPD in sero-negative recipients, the role of EBV DNA in whole peripheral blood (PB) of donors has not yet been elucidated and measurement of EBV DNA in donors is not a standard part of donor evaluation. EBV DNA in whole blood represents both the EBV genomes from mononuclear cells (mainly B cells refl ecting latency) and EBV DNA released in plasma due to infection. Thus, a positive result from donor's blood could harbor the risk of transmitting active EBV infection. Materials (or patients) and Methods: In our study, we retrospectively analyzed the role of EBV DNA in PB of donors at the time of transplant in developing EBV-LPD. Forty-two patients (29 male/13 female) with a median age of 41 (13-61) who underwent alloHCT for hematologic malignancies and aplastic anemia were evaluated. 

In our population other consistent risk factors (EBV seronegativity, aGvHD, RIC) were not signifi cantly correlated with EBV reactivation. Moreover, none of these donors developed an EBV overt infection or any other EBV-related disorder. Discussion: Our results indicate that even in a high risk population measurement of EBV DNA in whole blood of donors does not predict EBV-LPD and should not be done routinely. Since there is still no evidence to suggest that a healthy donor with high copy number is of greater risk than a healthy donor with an undetectable copy number, these donors should not be excluded as haematopoietic stem cell donors. Disclosure of Interest: None Declared. Here, we present our experience on the eff ectiveness and safety of ciprofl oxacin for the prevention of severe BKHC in high-risk cord blood plus third-party donor (CB-TP) dual transplant recipients. Materials (or patients) and Methods: Nineteen consecutive alloHCT candidates who in the absence of compatible related or unrelated volunteer donors received a myeloablative (fl udarabine, cyclophosphamide, busulphan, ATG) CB-TP dual transplant in our center from feb/2009 to sept/2013, were included in this study (11 AML/MDS, 3 ALL, 3 CLPD, 2 CML; 11 men; median age 44 [20-64]; 10 refractory to ≥1 chemotherapy lines, 4 relapsed after previous BMT, 5 with active disease at time of transplantation; 02/2009-09/2013). In this series, patients received no formal antibacterial prophylaxis. Eight patients receiving no BKHC ciprofl oxacin prophylaxis were compared to 11 patients receiving 60 days of post-transplant ciprofl oxacin for BKHC prophylaxis (500mg po bid or 400 mg IV bid) regardless of IV broad-spectrum antibiotics. GVHD prophylaxis (cyclosporine A and corticosteroids), antiviral (acyclovir), anti pneumocystis jirovecii (pentamidine or trimethoprim-sulfamethoxazole) and antifungal (posaconazole) prophylaxis strategies and supportive therapy were identical in both groups. Results: Three patients in the non-ciprofl oxacin group developed severe BKHC (one grade III and two grade IV) after neutrophil engraftment (25-35 days after alloSCT) compared with no cases in the ciprofl oxacin group (37.5% vs 0, P=0.027). BKHC episodes increased the hospital stay a median of 11 weeks (8-12) during which patients required intense transfusion support (median 30 packed red blood cells, median 41 [22-56] pool platelets) and up to four lines of treatment; two patients achieved complete response and one, died in refractoriness 8 weeks after BKHC diagnosis. All eleven patients in the ciprofl oxacin group completed the prophylaxis as initially planned and no ciprofl oxacin-related severe adverse events were reported. Time to neutrophil engraftment (mean 12.5 +/-2.9 vs 12.8 +/-4.16, P=n.s.) and 100-day survival after CB-TP dual transplantation (87.5% vs 82%, P= n.s.) were comparable between ciprofl oxacin and non-ciprofl oxacin groups. Discussion: Our results suggest that ciprofl oxacin prophylaxis may reduce the incidence of severe BKHC in high-risk cord blood transplant recipients, and the morbidity and resource use associated with it. They also confi rm that the maintenance of the prophylaxis during 60 days after transplantation regardless of IV broad-spectrum antibiotics is safe and well tolerated. Further studies with larger series of patients are warranted. Disclosure of Interest: None Declared. had recent resolution of GvHD. Only 5 patients with active GvHD were on steroids at dose of 1-2 mg/kg/day and 10/20 allogeneic HCT were on steroids but at a dose lower than 1-2 mg/kg/day. 11/24 had pulmonary involvement. Fungemia was notable with S. prolifi cans (3/5) and 4/4 of Coccidioides immitis had pulmonary involvement (2 had disseminated infection). 17/24 received antifungal prophylaxis in the 2 weeks preceding the diagnosis of IMI. Response to antifungal therapy at 4 weeks: none of the patients had CR, 6 had PR, 4 had stable disease and 14 died. At 90 days after diagnosis 14/24 had died (58%). No diff erence in mortality noted in those diagnosed early after HCT (within 90 days of HCT) versus those with late onset disease (6/10 vs. 8/14). 10/16 (62.5%) who had active GvHD or recent resolution of GvHD died vs. 4/8 (50%) with no GvHD. Mortality by fungal pathogen at 90 days; 3/4 with Fusarium died (75%), 7/11 with Scedosporium spp died (63%), 2/2 with Paecilomyces spp died, 1 each with Chaetomium spp and Alternaria spp died, while none of the patients with Coccidioides immitis died. Discussion: Although NA/NM IMI is infrequent after HCT, the response to antifungal therapy is sub-optimal and the mortality associated with these infections is signifi cantly high. Disclosure of Interest: None Declared. Introduction: Adenovirus (HAdV) is a signifi cant cause of morbidity and mortality in pediatric hematopoietic stem cell transplant recipients. As yet, HAdV DNA detection and quantifi cation in plasma is the only objective parameter for identifying HAdVinfected patients at high risk for disseminated disease. However, viral monitoring has not been widely adopted in clinical practice, owing mainly to the relatively low frequency of HAdV disease and to the lack of safe and eff ective treatment. Pediatric HSCT recipients could benefi t from HAdV surveillance, provided that cost-eff ectiveness of the procedure be enhanced by a centerfocused determination of the subgroups of patients at high risk of developing HAdV infection. Aim of this study was to assess risk factors for HAdV infection in a single-center cohort of pediatric HSCT recipients monitored prospectively, and evaluate effi cacy of a viro-immunological surveillance program and preemptive treatment approach on infection and patient outcome Materials (or patients) and Methods: 104 recipients of HSCT from matched related (n=18), matched-unrelated (n=41), or haploidentical family donors, transplanted at our center between 03/2010 and 04/2013 for malignant (n=70) or non-malignant (n=34) disease, were monitored weekly for HAdV DNA by real time polymerase chain reaction in blood, and for immune reconstitution. In case of persistently positive HAdV DNA >10.000 copies/ml and/or HAdV disease, patients received cidofovir treatment. Donor HAdVspecifi c T cells were employed as rescue therapy in unresponsive patients.

Results: HAdV viremia occurred in 22/104 HSCT recipients (cumulative incidence, CI, 21%) at a median time of 27 days from transplant (range 6-124 days). Peak HAdV DNA was 1900 copies/ml (range, 100 -4.123.100). Fourteen of the 22 HSCT recipients experienced self-limiting viremia without signs of viral disease, while in 8 patients, an increase in HAdV DNA from a median value of 6700 copies/ml up to a median peak of 1.65 x 105/ml was observed. Factors signifi cantly associated with HAdV viremia were a diagnosis of malignant disease, peripheral blood as stem cell source, development of grade II-IV acute GVHD, and failure to reconstitute CD4+ T cell subset and HAdV-specifi c T cell frequency. Although the rate of HAdV infection was double in recipients of haplo-HSCT compared with MUD HSCT, this diff erence was not statistically signifi cant. Patients with HAdV DNA >10.000 copies/ml were treated with cidofovir; in 4 of the 7 evaluable cases, viraemia further increased, and HAdV disease developed, that required a rescue with HAdVspecifi c T cell therapy. Viremia and clinical symptoms cleared in all patients, and no death due to HAdV disease progression was observed. Despite the good HAdV infection outcome, cumulative incidence of transplant-related mortality in the HAdV group was 30%, compared to 10% in the patients that remained HAdV DNAnegative (P<0.01). Discussion: There seems to be a benefi t for prospective monitoring and preemptive therapy in pediatric HSCT recipients with HAdV viral load >10,000 copies/ml in terms of HAdV disease prevention. However, HAdV-associated morbidity exerts a toll on TRM, that may possibly be prevented by a prompter intervention with HAdV-specifi c T cells in patients with aGVHD and delayed immune reconstitution. Disclosure of Interest: None Declared. based guidelines for placement and management of central venous catheters in patients undergoing hematopoietic stem cell transplantation. The central venous catheter infection prevention bundle consists of recommendations regarding hand hygiene, full barrier precautions, cleaning the insertion site with chlorhexidine, avoiding femoral sites for insertion, and removing unnecessary catheters. The aim of this study was to survey recommendations included in standard operating procedures (SOP) and current clinical practice regarding CLABSI prevention, infection monitoring and education on CVCs at hematologic and oncologic EBMT centers performing HSCT. Materials (or patients) and Methods: Setting and sample This cross sectional study was performed during January to June 2013 among European centers performing HSCT and being a member of EBMT. Among the 545 EBMT transplant centers worldwide 103 (19%) participated. The questionnaire A questionnaire concerning recommendations included in standard operating procedures (SOP) of the center for management of CVCs to prevent CLABSI (Part A) and current practice regarding management of CVCs (Part B) was specifi cally developed for this study by the authors. Results: CLABSI prevention bundle The recommended CLABSI prevention bundle consists of hand hygiene, full barrier precautions, cleaning the insertion site with chlorhexidine, avoiding femoral sites for insertion, and removing unnecessary catheters. In 33% of the centers all parts of the bundle were included in SOP. This corresponded to 27% of centers that fully incorporated recommended CLABSI prevention bundle into current clinical practice. The most common unfulfi lled parts of CLABSI prevention bandle were lack of complete full barrier precautions and lack of use of chlorhexidine in clinical practice. Full barrier precautions All fi ve parameters of full barrier precaution (use of cap, mask, sterile gloves, sterile gown, full size body drape: 60 x 60 cm) were fulfi lled in 57% of the centers SOP and 36% reported these procedures being used in clinical practice. The most common unfulfi lled criteria for full barrier precautions was lack of use of full size body drapes during CVC insertion. In clinical practice 35% of the centers used a full size body drape, 48% used a larger drape and 17% used a smaller drape than 60 x 60 cm. Discussion: This is the fi rst report on the current practice on CLABSI prevention among the EBMT centers. Among the surveyed centers only minority report implementation of the full recommended practices. The most common recommendations missing in clinical practice are use of all full barrier precautions and use of chlorhexidine containing solutions. Also considerable variation exists regarding preferred site of the central line insertion and every day care. The results of the study show that prevention of CLABSI is still a challenge for many centers and there is a need for improvement. The further study is needed to assess if observed diff erences between the centers could infl uence HSCT outcome at the centers. Disclosure of Interest: None Declared. Introduction: We evaluate the incidence and outcome of blood stream infections (BSI) caused by multi-drug resistant (MDR) bacteria in patients (pts) with hematologic diseases and in hematopoietic stem cell transplantation (HSCT) recipients. Materials (or patients) and Methods: 256 BSI have been recorded between october 2008 and october 2013 in 187 pts. The characteristics and outcome of MDR-BSIs have been evaluated. MDR-BSIs and multi-drug-sensible bacteria-BSI (MDS-BSI) have been compared. Moreover the results were evaluated according to the status of underling hematologic disease and the HSCT. Results: 316 isolates have been documented in 256 BSIs. Among these, 55 (17%) bacteria were MDR and 261 (83%) were MDS. MDR isolates were: MDR Pseudomonas aureginosa, 28/55 (51%); Escherichia coli ESBL, 16/55 (29%); Stenotrophomonas maltophilia, 6/55 (11%); Klebsiella pneumoniae-KPC, 4/55 (7%); Enterococcus sp-VRE, 1/55 (2%). A progressive increase of MDR-BSI was documented in the last two years (with 1,55 episodes/month in 2013).Septic shock and infection related mortality were significantly higher in the MDR-BSI group compared to MDS-BSI (36% vs 12% and 35% vs 11%, respectively -P less than 0,001). Occurrence of septic shock was particularly elevated in KPC-BSI (75%) and in MDR-P. aureginosa BSI (50%). Pts with refractory/relapsed hematologic disease had higher infection mortality rate (70%). In the MDR-BSI group, a higher mortality rate was observed in HSCT pts, (52% in HSCT vs 20% in No-HSCT, P less than 0,05). Discussion: MDR-BSI (particularly MDR-Pseudomonas aureginosa and Klebsiellae-KPC) are an emerging and serious problem in oncohematologic pts. The onset of this MDR-BSI is often associated with septic shock and with high mortality rate, despite a target and combined antibiotic therapy. Mortality MDR-BSI related is particularly elevated in HSCT recipients (52%) and in pts with refractory/relapsed hematologic disease (70%). These results underline the urgent need of guidelines for surveillance and treatment of hematologic and HSCT pts with MDR bacteria colonization or/and active infection. Disclosure of Interest: None Declared. Introduction: Cytomegalovirus (CMV) infects and establishes persistent lifelong infections in 50-85% of adults. Reactivation of the virus is a frequently occurring complication of immunosuppression following transplantation and can signifi cantly contribute to morbidity and mortality in such patients. Reconstitution of CMV-specific T cell immunity after allogeneic hematopoietic cell transplantation (alloHCT) has previously been quantified using CMV tetramers, and shown to be a valuable aid in predicting patients at risk of developing CMV reactivation. Materials (or patients) and Methods: We have developed an assay for quantifying CMV-specifi c CD8+ T cells using CMV-specifi c Dextramers. Dextramers are MHC multimer reagents that are used in fl ow cytometry to detect antigen-specifi c T cells in the blood. Dextramers have much higher resolution than conventional MHC multimers like Tetramers, and thus provide a more reliable means for identifi cation of antigen-specifi c T cells. Results: We here show that the CMV Dextramer assay including 7 alleles (HLA-A*01,-A*02, A*03, A*24, B*07, B*08 and B*35) has high specifi city and sensitivity and accurately enumerates CMVspecifi c T cells in both healthy donors and alloHCT patients, with a lower detection limit of 0.08 cells/ul. The assay is highly reproducible with low intra and inter assay variation. Using the CMV Dextramer assay we were able to quantify reconstitution of CMV T cell immunity post transplantation at day 30, 100, and 365 in 89 patients. Furthermore, in some recipients receiving transplants from HLA-mismatched donors we could measure CMV-specifi c T cells restricted by donors HLA and not recipients HLA, indicating that adoptive transfer of CMV-specifi c T cells can occur with alloHCT. Discussion: This study demonstrates that CMV Dextramers are reliable reagents that can be used to monitor reconstitution of CMV immunity post-alloHCT, and shows that adoptive transfer of anti-CMV immunity can be quantifi ed. Disclosure of Interest: None Declared. Introduction: Non-myeloablative allografts (NMA) remain contentious in the treatment of multiple myeloma (MM). We report our experience in 21 patients from 2005-2013 using a fl udarabine and low-dose TBI regimen for NMA within 6 months of autologous stem cell transplantation (the Seattle regimen, see Rotta et al, Blood, 2010) . Materials (or patients) and Methods: We reviewed our transplant database medical fi les of all patients who had undergone NMA for MM since 2005. Results: The median age was 56 (range 40-62). Males accounted for 15 out of 21 patients. Five patients had sibling donors. All patients had a creatinine clearance of 45ml/min or greater at time of NMA. 13 patients were treated in fi rst clonal remission, 7 in second remission and one in third remission. 16 had measured cytogenetic abnormalities, and the majority of patients had either poor risk cytogenetics and/or were ISS Stage III. 15 patients had their NMA in partial remission, three in very good partial remission and three in complete remission. Median patient follow-up from NMA was 13.5 months (range 1-49). Estimated median overall survival was 28 months. Five patients have died since NMA. Transplant-related mortality at 2 years was 10%. No patients died as a consequence of the Seattle conditioning and the majority died from progressive disease. The patient who underwent NMA in third remission died within two months of progressive disease. Median estimated progression-free survival was 21 months. Five patients had evidence of clonal relapse within a year of NMA. Five patients received DLI and six patients received lenalidomidebased salvage treatment, with good clonal responses observed in the majority. Acute graft-versus-host disease (GVHD) was observed in 7 patients: 4 skin, 2 oral and 2 gut, Grades 1-3. Chronic GVHD was observed in 12 patients: 6 oral, 5 skin, 2 gut and 2 liver. 3 mild, 5 moderate and 4 severe. Discussion: The role of NMA in the treatment for MM remains undefi ned. The Seattle regimen is easy to administer and well tolerated, even in poor risk and heavily pre-treated patients. GVHD was common but there was a trend that this was associated with improved survival. Clonal relapses were not uncommon in this particularly poor-risk population but salvageable in the majority. Larger numbers and further clinical trials are still required to determine the optimal place for NMA in the treatment of myeloma, and the optimum conditioning regimen. Disclosure of Interest: None Declared. Introduction: Renal failure (RF) is a common complication in multiple myeloma (MM). It has been associated with inferior survival in patients (pts) treated with conventional regimens. The impact and 2011 at Nantes. The cohort included 23 males (64%) and 13 females (36%) with a median age at time of allo-SCT of 55 (range, 37-65) years. Patients received a median of 2 lines of treatment (ranges, 1-4) before allo-SCT. 14 patients (39%) with high-risk MM (del13 by FISH + high beta2-microglobuline at diagnosis) received allo as part of frontline therapy, in a tandem auto-RIC allo program (protocol IFM99-03, Moreau et al, Blood 2008; 112:3914-3915) and 22 patients (61%) received allo as part of salvage therapy and all but 1 had chemosensitive disease at the time of allo. PBSC were used as stem cell source in 35 patients (97%), while one patient received BM. A MRD was used in 24 cases (67%) and an UD in 12 cases (33%). The conditioning regimen was based on the combination of fl udarabine and busulfan and high-dose ATG in patients who received frontline RIC allo (14 cases), while the others received either fl udarabine and 2 Gy TBI (2 patients) or fl udarabine, melphalan and bortezomib (20 patients) as part of conditioning prior to RIC allo. Results: The median time between MM diagnosis and allo-SCT was 66 months. The median follow-up among surviving patients was 48 months (range, 25-132). Overall, the cumulative incidence of acute GVHD grade III-IV was 25% and the cumulative incidence of extensive chronic GVHD at 1 year and 5 years were 30% and 30%, respectively. At 3 months, 43% of patients were in CR. Relapse or disease progression occurred in 22 patients at a median of 11 months (ranges, 2-46) after allo-SCT, the cumulative incidence of relapse/progression was 33% at 1 year and 63% at 5 years. The cumulative incidence of NRM was 19% at 1 year and 34% at 5 years. For the whole group of 36 patients, the KM estimates of OS and PFS at 5 years after allo-SCT were 46% and 12%, respectively. When comparing the group of 14 high-risk patients treated with the tandem auto/RIC allo program as part of front treatment versus the group of 22 patients treated by RIC for chemosensitive relapse later in the course of the disease, the 5-years PFS and OS estimates were 21.4% vs 6.1% (P = .02), and 64.3% vs 32.7% (P =0.12). Discussion: A frontline tandem allo-RIC program may induce durable responses (and cure?) in one fifth of the patients. PFS of patients treated with RIC as salvage therapy is dramatically poor, indicating the low probability for cure. Nevertheless, OS in both groups of patients is high, indicating that patients can be salvaged by novel-agent based therapies following relapses after RIC-allo. Disclosure of Interest: None Declared.

M. Veeraputhiran 1,* , T. Jain 2 , A. Deol 1, 2, 3 , S. Kim 4 , G. Dyson 4 , J. Uberti 1, 2, 3 , M. Abidi 1, 2, 3 1 Hematology/Oncology, Karmanos Cancer Institute, 2 Internal Medicine, Wayne State University, 3 Blood and Marrow Stem Cell Transplant Program, 4 Biostatistics Core, Karmanos Cancer Institute, Detroit, United States Introduction: Salvage ASCT is increasingly utilized for eligible MM patients (pts). Despite concerns for resistance, high dose melphalan (HDM) is predominantly used as conditioning regimen for 2 nd /or salvage ASCT. BEAM (Carmustine, etoposide, Ara-C and Melphalan) may have superior outcomes compared to HDM during 1 st ASCT for MM pts as shown in a single institution study. We conducted a retrospective analysis at our institution to evaluate toxicity and response with BEAM conditioning used as salvage therapy after HDM ASCT. Materials (or patients) and Methods: Thirty eight pts who received 2 nd ASCT for MM (2008) (2009) (2010) (2011) (2012) (2013) were identifi ed. Nineteen pts received HDM (140-200mg/m 2 ; Group A) and an equal number received BEAM (carmustine 300 mg/m 2 , etoposide 100 mg/m 2 , cytarabine 100 mg/m 2 , and melphalan 140 mg/m 2 ; Group B). Patient characteristics, toxicity and response at day 100, progression free survival (PFS) were collected and compared between the 2 groups. Fischer's exact, Cochran-Mantel-Haenszel tests and log rank tests were used to compare responses, toxicities and PFS. Results: The median age of the patients was 59 yrs (range: 39-72) in Group A and 58 yrs (range: 44-68) in Group B. Group A had more female pts (53% vs 32%) and a decreased baseline renal function (median creatinine clearance 84 vs 108 ml/min) compared to Group B. All patients in both groups received salvage therapy prior to 2 nd ASCT and received stem cells which were collected/ cryopreserved prior to the 1 st ASCT. The median time between 1 st and 2 nd ASCTs were 47 (range: 17-68) vs 34 (range: 18-59) months in group A and B respectively. There was no signifi cant diff erence in disease status pre-2 nd ASCT between the 2 groups. In group A, eight pts (50%) received reduced dose melphalan (140mg/m 2 ) due to renal insuffi ciency while the remaining eight pts received HDM (Median 200mg/m 2 , range 180-240mg/m 2 ). At day +100 post 2 nd ASCT nine patients in each group had CR ( Table 1 ). The median time of follow-up after 2nd ASCT was 9 and 5 months for groups A and B respectively. The median PFS were 12.9 vs 7.7 months (P=0.62) for Group A and B respectively (Figure-1) . Group B had a higher incidence of febrile neutropenia (19 vs 13 pts, P=0.02) and longer hospitalization (22 vs 15 days, P <0.0001). Other toxicities were not signifi cantly diff erent between these groups. Discussion: BEAM seems to be a reasonable alternative conditioning regimen for 2 nd ASCT with comparable safety and efficacy to HDM, however requires prolonged hospitalization and has increased incidence of febrile neutropenia. Longer follow-up

S348 is needed to determine whether there is any signifi cant diff erence in PFS and or OS between the two groups. Disclosure of Interest: None Declared. Introduction: We analyzed main modalities and clinical outcomes of the early discharge outpatient model-autologous stem cell transplantation (EDOM-ASCT) for multiple myeloma in Italy. Materials (or patients) and Methods: This retrospective study has been conducted through the GITMO trial offi ce, which promotes independent clinical research studies in the setting of both autologous and allogeneic transplantation in Italy. A fi rst questionnaire was mailed to 75 GITMO Centers accredited for autologous transplant to evaluate how many had been involved in outpatient transplant models for MM patients between 1998 and 2012. If a given Center was involved, further specifi c queries included patient selection criteria, infectious prophylaxis, supportive care, criteria for hospital re-admission, management of febrile neutropenia, and clinical outcomes. Overall, 55/75 (73.3%) answered the fi rst questionnaire, and among these 6 had been involved in outpatient transplant programs according to "EDOM". Results: EDOM-ASCT was applied in 382 patients, for a total of 522 procedures, between 1998 and 2012. Our study showed high homogeneity among Centers in terms of inclusion criteria and supportive care, and in-hospital re-admission criteria. Overall, re-admissions during the aplastic phase occurred in 98 of 522 transplants (18.8%). Major extra-hematological complication was neutropenic fever in 161 cases (30.8%) that required re-admission in 76 cases. Incidence of severe WHO 3-4 mucositis was 9.6%. By univariate analysis, fever, mucositis, renal function at diagnosis, second transplant, transplant performed late in the course of the disease were signifi cantly correlated with re-admission, whereas fever, mucositis, renal function and timing of transplant were the only independent predictors by multivariate analysis. Overall, transplant related mortality was 1.0%. No center eff ect was observed in this study (P=0.36). Discussion: The low rate of re-admission and the safety of this EDOM-ASCT in myeloma indicates that this strategy could be extended to other transplant centers if a stringent policy for patient selection and management is applied. Disclosure of Interest: None Declared. Introduction: Immunoglobulin light chain amyloidosis (AL) is a plasma cell proliferative disorder characterized by deposition of insoluble fi brils composed of immunoglobulin light chains, causing progressive organ dysfunction. In 50% of cases of AL, there is documented renal involvement, which, if left untreated, progresses to end-stage renal disease and is often associated with signifi cant morbidity. Materials (or patients) and Methods: We performed a retrospective analysis in 72 patients (pts) with AL who underwent highdose chemotherapy and autologous hematopoietic stem cell transplantation (auto-HCT) at our institution between 1999 and 2011. Fifty-fi ve of these AL pts had renal involvement, as defi ned by the International Consensus Criteria. Primary objectives were to assess hematologic and organ response, progression free survival (PFS), overall survival (OS) and correlation between hematologic and organ response. Results: Median age at auto-HCT was 59 years (range, 41-74). Median time from diagnosis to auto-HCT was 6.6 months (range, 2.2-121.9). Forty-one pts (75%) had lambda light chain disease. Additional visceral organs involved were: heart in 5 pts (9%), liver in 5 pts (9%), GI tract in 3 pts (5%), and peripheral nerves in 1 pt (2%). Median baseline proteinuria was 5.055 grams (range, 0.174-27.854 grams) in a 24 hour urine specimen. Median base-line serum creatinine was 1.20 mg/dL (range, 0.53-9.28). All pts received melphalan or melphalan-based combinations as their preparative regimen. Median time to neutrophil engraftment was 10 days (range, 7-15). Median follow up from auto-HCT was 37 months (range, . Six pts died of non-relapse causes with a treatment-related mortality (TRM) at both 100 days and 1 year of 11% (95% CI 5-23). Fifty pts were evaluable for a hematologic response, while 5 pts were inevaluable due to early death. Eight (16%) pts achieved complete remission (CR), 10 (20%) achieved a very good partial remission (VGPR) and 21 (42%) achieved a partial remission (PR), with an overall response rate of 78%. Organ response was evaluated at 6, 12, and 24 months, and was defi ned as a >50% decrease in total proteinuria over 24 hours without an increase of >25% of serum creatinine. Organ response was demonstrated in 8 (16%) of 50 evaluable pts at 6 months, in 13 (31%) of 42 evaluable pts at 12 months, and in 17 (49%) of 35 evaluable pts at 24 months, with a median decrease in proteinuria of 1.21 grams, 1.24 grams, and 2.86 grams at 6, 12, and 24 months, respectively. Organ responses at 2-years post auto-HCT were seen in 15 of 28 (54%) pts with > PR and 2 of 7 pts (29%) with <PR (P=0.2). Kaplan-Meier estimates of 3-year PFS and OS were 54% and 73%, respectively. In a multivariate model, induction prior to auto-HCT and a low absolute diff erence between involved and uninvolved serum free light chains (FLC diff ≤ 26 mg/dL) were associated with longer PFS and OS. At the time of last follow-up, 24 pts (44%) were alive and in remission. Discussion: High-dose melphalan and auto-HCT is associated with durable hematologic and organ response and prolonged survival in patients with AL and renal involvement. Induction therapy prior to auto-HCT and absolute FLC diff ≤ 26 mg/dL predicted longer PFS and OS. Disclosure of Interest: None Declared.

Grade II-IV. 4 patients received DLI; 3 for disease progression all of whom died within the following 12 months. 1 received DLI for low T-cell chimerism. This was followed by the onset of GVHD 1 month later and the patient remains alive 6 years later. 100 day NRM was 5% (n=2). Cumulative 1 year NRM was 11% with 1 patient having a GVHD-related death. 3 year NRM mortality remained at 11%. Best responses post allograft were 72% CR and 28% PR from all evaluable cases (17% data unavailable). Median overall survival (OS) from RIC allo-SCT for all patients was 5.2 years. Median progression free survival (PFS) was 2.4 years. ISS and age at time of transplant had no eff ect on PFS or OS. Achieving a CR vs PR preallo SCT had no signifi cant eff ect on OS, but resulted in a significant diff erence in median PFS (63% vs 23% at 60 months; P=0.03). A signifi cant diff erence in PFS was seen in patients who received ≤3 vs 4 lines of treatment prior to allo-SCT (median PFS 2.15yrs vs 0.54yrs; P=0.0003), and demonstrated a trend towards an eff ect on median OS (5.6yrs vs 3.2yrs; P= 0.06). Discussion: RIC-allo-SCT in myeloma using fl udarabine/ cyclophosphamide conditioning is an eff ective treatment strategy with an acceptable NRM. PFS is signifi cantly increased if a CR is achieved prior to allo-SCT, and if the patient has received ≤3 lines of treatment. Disclosure of Interest: None Declared. S352 the new studies have testifi ed that the CD137 ligand (CD137L) is expressed by various carcinoma cells too. However, the expression of CD137L and its function on multiple myeloma cells is unknown. Materials (or patients) and Methods: We identifi ed the constitutive expression of CD137L by fl ow cytometry on U266, RPMI 8226, LP1, MY5 and KMS-11 of Multiple myeloma (MM) cell lines as high as 96%, 97.5%, 89%, 93% and 94%.But, CD137 expressed on the cell surface was low as 4%, 5%, 1%, 2%, 5% respectively. We investigated CD137L expression of MM cells from 85 BM samples of patients diagnosed with active multiple MM, CD137L protein was expressed by a select group of CD45-CD38++CD138+cells as higher than 95%, the same, CD38 and CD138 are expressed more than 90% of the cells of CD45-CD137L+.There were 22 samples from 11 cases collected before and after treatment and this was further evidence that CD137L molecule was consistently expressed on the MM cell surface. However, CD137L expression was not or hardly detectable on normal plasma cells confi rmed by CD45+CD38++CD138+ CD56-CD19+, indicating that CD137L was ectopically expressed by MM cells and probably a specifi c marker of MM cells. Results: We selected U266-a MM cell line to explore the biological effect of CD137L reverse signaling and its underlying mechanism. The proliferation and survival of U266 was enhanced by stimulating-CD137L mAb (1F1) than those induced by control mouse IgG by cell counting (4.2 X10 5 /ml VS 3.3 X10 5 /ml, the average of 6-well plates), WST-8(1.15 VS 0.81, the average of 6-well plates absorbance) and CFSE assay (930 VS 991, fluorescence measured by flow cytometer) at incubation for 48h. In addition, the cell cycle analysis showed that CD137L induces proliferation and increases the number of cells in the S phase from 36.1% to 42.5% after 72h incubation. The percentage of apoptosis cells was 19.6% VS 21.2% with no statistical significance. Intracellular cytokine staining showed that treatment of cells with 1F1 increased the production of IL-6 from 3.8% to 63.9% by Flow cytometry. When neutralizing anti-IL-6 mAb (5 μg/ml) was added to the culture medium, the cells (2X10 5 /ml) were cultured for 48 h in pure medium or plus 10 ng/ml Fc or CD137-Fc protein and the cell proliferation measured by WST-8 was 0.79 VS 0.80 VS 0.72, the average of 6-well plates absorbance.1F1-induced cell proliferation was eff ectively inhibited. IL-6 can promote cell proliferation and survival of MM. An increase of these cytokines might explain why CD137L expression could stimulate the proliferation of U266. Finally, the U266 cells were treated with bortezomib and the growth of cells was analyzed by WST-8 assay. It demonstrated that bortezomib could inhibit the function of 1F1 and the inhibition ratio of bortezomib was 22%, 51% and 58% at 24h, 48h and 72h. Discussion: In our study, CD137L is not only a novel ectopic constitutive marker of MM, but also a promoting proliferation factor. This suggests the possibility that its expression on MM cells may be directly target for immunomodulatory therapy for MM. Disclosure of Interest: None Declared. Introduction: Post transplant Cyclophosphamide (PT-Cy) after nonmyeloablative conditioning regimen (NMAC) after T-cell replete haploidentical transplantation reduces graft-versus-host disease (GVHD) and graft rejection. This major HLA incompatibility can be supposed to harness an important anti-tumor eff ect, in several haematological malignancies. In this retrospective study, we evaluated the safety and effi cacy of haplo and PT-Cy in refractory multiple myeloma (MM) patients. Materials (or patients) and Methods: Between April 2009 and November 2013, 124 patients with hematological malignancies were treated in two diff erent centers. 9 refractory multiple myeloma patients were retrieved. Conditioning regimen consisted of Cy 14.5mg/kg/day i.v. on days −6 and −5, fl udarabine 30 mg/m 2 / day i.v. on days −6 to −2, and 200cGy of TBI on day −1. Cy was administered. Prophylaxis of GVHD consisted of PT-Cy (day +3 and +4), 50 mg/kg tacrolimus or cyclosporine and mycophenolate mofetil (MMF). Unmanipulated haplo marrow or peripheral blood stem cells was infused on day 0. Results: Disease status before transplant was PD for 4 patients (44%) and DRI was HR in 5 and INT in 4. All patients relapsed after 1 or 2 HDC and autologous stem cell support. HCT-CI was ≥ 2 in most patients (67%). Median quantity of CD34+ cells infused was 4.7x10 6 /kg. Only one patient (11%) developed a cutaneous grade II aGVHD; no cGVHD was recorded except one occurred after DLI infusion. The median times to neutrophil (>500/μL) and platelet recovery (>20,000/μL) were 20 and 24 days, respectively. No graft failure occurred. Bacterial, viral and fungal infections were diagnosed in 4, 3 and 2 patients, respectively. With a median follow-up of 7 months (1-35), the cumulative incidence of NRM was 11%. Current overall survival (OS) is 67% (see Table) . Discussion: Despite short follow-up time, our multicentric experience suggests that in advanced multiple myeloma patients, NMAC haploidentical transplant with PT-Cy is well tolerated without an excessive GVHD and NRM, and it seems to aff ord an interesting anti-myeloma eff ect. More patients and longer follow-up are needed. Disclosure of Interest: None Declared.

Introduction: Multiple myeloma remains incurable, but new drugs off ers a signifi cant prolongation of survival in clinical trials. There is need to look for real-life eff ectiveness of new drugs and ASCT role in daily practise. Materials (or patients) and Methods: The observational study was aimed to evaluate infl uence of ASCT consolidation on DOR and OS of 295 relapsed or refractory, bortezomib-naïve MM patients, responding to bortezomib-based salvage regimens, suitable for ASCT (≤65 yrs, Karnofsky index >60%). Bortezomib was given iv. alone, or combined with steroids or anthracyclines, Thal and alkylators. The drugs combination and ASCT were case-adjusted in order to balance antitumor potential and toxicity at physician discretion. The survival analyses were carried out using the Kaplan-Meier method and log-rank test. The multivariate Cox proportional hazard regression analysis was done for categorised variables (resistant/relapsed MM, sex, Ig isotype, D-S and ISS stage, number of previous treatment lines, usage of Thal before, undergone ASCT, response to therapy, B2microglobuline>2,5mg/ dl; WBC<4,0G/L) and continous (age, time from dgn to bortezomib therapy, protein M, Hgb, PLT, creatinine, albumine and calcium). Factors of prognostic value p≤0.05 were included into fi nal model. Results: The therapy result in ORR 67.9% for refractory MM and 69.9% for relapsed MM. The median DOR and OS were 16 and 56 months resp. Bortezomib salvage+ASCT vs bortezomib salvage only resulted in better DOR (median 23 vs 8 months; P<0.000001) and OS (median not reached vs 8 months; P<0.000001). The significant predictors for DOR were: relapsed MM (HR=2.14 P<0.000001), ISS stage (HR=1.29 P=0.033), partial response to bortezomib (HR=1.19, P=0.046), age (HR=1.03 P<0.01) and female (HR=0.70 P=0.016), but for OS were: relapsed MM (HR= 1,85 P= 0.022), ISS stage (HR= 1,78 P= 0.004), partial response to bortezomib (HR= 1,44 P= 0.025), PLT prior bortezomib (HR= 1,02 P= 0.004), HGB prior bortezomib (HR= 0,79 P= 0.00014) and female (HR= 0,54 P= 0.023). No toxicity to bortezomib therapy was reported for 38.6% of patients. Severe toxic events were reported for 35.9% of patients: mainly neurotoxicity (16.0%), neutropenia (6.1%), thrombocytopenia (4.4%) and infections (5.4%) but there were no infl uence of toxicity on DOR and OS by multivariate analysis. Discussion: The patients benefi t strongly from ASCT consolidation after bortezomib-based salvage and there are identifi able survival predictors. Toxicity after bortezomib-based treatment was predictable and manageable. Bortezomib-based regimens are highly eff ective salvage MM and should be supported by ASCT consolidation in suitable patients. Disclosure of Interest: None Declared. Introduction: Maintenance therapy with immunomodulatory drugs has shown to improve responses and delay relapse and progression in Multiple Myeloma (MM) patients after autologous stem cell transplantation (ASCT). Although maintenance therapy with Thalidomide (T) after ASCT increases progression free survival (PFS) in MM patients, it is associated with dose-limiting multiple toxicities. Lenalidomide (revlimid, R) has been reported to be associated with lower rates of toxicities than thalidomide. Materials (or patients) and Methods: We evaluated effi cacy, safety and the eff ect on circulating Natural Killer (NK) cells of continuous maintenance therapy with alternate-day low dose R (LD-R, 10 mg/day), after high-dose melphalan (HD-MEL, 200 mg/mq) and ASCT, in 8 MM patients (6 male and 2 female) with median age of 68 years (range 60-73), receiving pre-ASCT induction treatment with 4 cycles of conventional bortezomib, Thalidomide and dexamethasone (VTD) regimen. Of these 8 MM patients, 2 and 6 patients were in complete remission (CR) and in very good partial remission (VgPR) after ASCT, respectively. The eff ects of lenalidomide maintenance therapy on NK cells were evaluated the numbers of circulating NK cells by fl ow cytometry assessing the expression of T-cell antigens (CD3, CD4, CD8), NK-cell antigen (CD56) and NK-cell-activation antigens (CD2, HLA DR). All antibodies were obtained from Beckman Coulter. Phenotypic analysis of all these T-and NK-cell antigens were performed before starting maintenance, after the 1 st month, and then every 3 months during R therapy. Results: After a median follow-up of 20 months (range 9-48) from the initiation of LD-R maintenance, patients in CR maintained their CR, and all patients in VgPR improved the depth of response except one who showed disease progression. In this LD-R group, PFS and overall survival (OS) at 24 months were 83% and 100%, respectively. Noteworthy, no signifi cant specifi c R-related toxicity was encountered. All MM patients treated with continuous alternate-day lenalidomide showed progressive increase in the percentage of circulating CD56+ CD3-Natural Killer cell. The median percentage of NK cells was 4% before R maintenance versus 9%, 22%, 26% and 31% at +3, +6, +12 and +18 months, respectively. Discussion: Our preliminary results provide evidence that continuous therapy with alternate-day LD-R is a feasible and eff ective maintenance treatment after ASCT for MM patients, enabling a long-lasting maintenance therapy, and that prolonged lowdose lenalidomide treatment increases circulating NK cells further supporting that this drug may mediate its anti-MM eff ect, at least in part, by modulating NK cell number and function. These results require further validation in prospective larger studies. Disclosure of Interest: None Declared. Introduction: In the last 10 years, the overall survival (OS) of patients (pts) with multiple myeloma (MM) has improved considerably. Transplant eligible pts have a 5-year survival rate of more than 70% with modern therapy. Autologous hematopoietic stem cell transplantation (auto-HSCT) is standard therapy for initial or treatment of relapse for MM pts younger than 65 years or even older, but fi t pts in good clinical condition and it is connected with better OS and progression free survival (PFS) in comparison to conventional therapy. We retrospectively analyzed pts with MM treated with auto-HSCT in our center from 1998. till 2012. Materials (or patients) and Methods: A total of 155 procedures of auto-HSCT in 120 de novo MM pts were performed after one or two lines of initial treatment. Male/ female ratio in our group was 72/28, average age was 52 years (37-64). Majority of pts, 72 (60%) were in III clinical stage. Pts have received 3-6 cycles of induction therapy (VAD, PAD, CTD, TAD or TD). Prior to auto-HSCT 43 pts (35,8%) were in complete remission (CR) + very good partial remission (VGPR), 35 pts (35.8%) in partial remission (PR), 7 pts (5.5%) were in less than PR and 4 (3.4%) had progression of disease (PD). Median time from diagnosis till auto-HSCT was 13,9 months ( Introduction: The role of allogeneic stem cell transplantation (allo-HSCT) in the treatment algorithms for patients with multiple myeloma remains controversial although it is the only potentially curative approach currently available. Here we present the retrospective analysis of 88 allo-HSCT performed between 1994 and 2011 at Ulm University Hospital. We focused on the impact of cytogenetics, graft-versus-host disease (GvHD) and intensity of conditioning on overall survival (OS), progression-free survival (PFS), relapse and non-relapse mortality (NRM). Materials (or patients) and Methods: Median age at initial diagnosis was 49 years (range 25-64), median age at time of allo-HSCT was 51 years (range 26-65). Median time from initial diagnosis to allo-HSCT was 14 months (range 3-106). Indications for allo-HSCT were 1) primary allo-HSCT after induction therapy (10 pts), 2) planned tandem auto-allo-HSCT (40 pts), 3) relapse after single allo-HSCT (24 pts), 4) relapse after tandem-auto-HSCT (14 pts). The conditioning regimen in 60 pts was a reduced intensity conditioning (RIC), 28 pts received myeloablative conditioning (MAC). In 58 pts cytogenetic data were available: 14 pts were stratifi ed into standard-, 31 pts into the intermediate-and 13 pts into the high-risk group according to the mSMART recommendations. Results: Median follow-up was 71 months (95% CI, 59, 8) . The estimate 1-, 2-and 5-year OS was 63,6 %, 58.7 % and 42,8 % with a median OS of 36 months (95 % CI, 19.9 -52.1). For RIC versus MAC median OS was 35 months (95% CI, 21.4-48.6) versus 89 months (95% CI 0-193.2) but this diff erence was not statistically signifi cant. The cumulative incidence of TRM was not diff erent for RIC and MAC but there was a highly signifi cant diff erence in relapse (P=0.011). With respect to the indication for allo-HSCT outcomes were as follows: Median OS was 147 months for primary allo-HSCT, 47 months for tandem auto-allo-HSCT, 20 months for relapse after auto-HSCT and 15 months for relapse after double auto-HSCT. OS did not diff er signifi cantly. Median OS in the standard risk group was 20 months (95% CI, 5. 3-34.7) , in the intermediate group 41 months (95% CI, 21. 2-60.8) and in the high-risk group 7 months (95% CI, 0-14), showing no statistical signifi cance. 1-year OS was 64,3% vs 67.7% vs 38.5%, and 2-year OS was 40.8% vs 67.7% vs 38.5% (standard vs intermediate vs high risk). The median PFS was 12 months (95% CI 7.8-16.2). 1-year PFS was 49.9%, 2-year PFS 34.6% and 5-year PFS was 21.6%. PFS according to the cytogenetic aberration showed a median PFS of 12 vs 14 vs 5 months, 1-year PFS was 50% vs 54.5 % vs 30.8 %, 2-year PFS with 33.3% vs 37.5% vs 7.7% (standard vs intermediate vs high risk). These diff erences are not statistically signifi cant. Considering the impact of acute GvHD, OS signifi cantly diff ered between the groups with no aGvHD or aGvHD grade I and aGvHD grade II-IV with inferior survival for patients suff ering from aGvHD grade II-IV. Chronic GvHD had no impact on outcome. Discussion: Our data of a 17 year experience in treatment of patients with advanced multiple myeloma with allo-HSCT showed an eff ective treatment option with a curative potential even for patients after intensive pretreatment including autologous stem cell transplantation. Patients who received MAC had a signifi cantly lower cumulative incidence of relapse compared to RIC without increasing TRM in our study. Disclosure of Interest: None Declared.

N. Kröger 1,* , S. Iacobelli 2 , A. van Biezen 3 , L. Garderet 4 and CMWP 1 University Hospital, Hamburg, Germany, 2 Universita tor Vergata, Rome, Italy, 3 University Medical Center, Leiden, Netherlands, 4 Hospital Saint Antoine, Paris, France Introduction: To investigate the use and outcome of thiotepa in combination with fl udarabine or other drugs as conditioning regimen followed by allogeneic stem cell transplantation for multiple myeloma we screened EBMT database. Materials (or patients) and Methods: We found 45 patients with a median age of 53 years (range 37-69) who received thiotepa-based regimen. Allogeneic stem cell transplantation was performed either as fi rst (n=21) or second transplant (n=24) Male/female distribution was 28/17. Most patients had advanced stage III (73%) or II (23%) myeloma, which was common type (IgG or IgA) in 58%, light chain in 33% or non-secretory MM in 9% of the patients. At transplantation remission status was evaluable in 40 pts: CR: 18%, PR: 37%, SD/MR: 15% and Relapse/PD: 30%. In-vivo (29%) and ex-vivo (11%) or combined (9%) T-cell depletion was used in 49% of the patients. Donorwere: HLA-identical sibling (73%), matched unrelated (16%), and mismatched related or unrelated (7%) donor or other relatives (4%). Conditioning regimen was classifi ed as reduced intensity (62%) or myeloablative (38%). Bone marrow or peripheral blood was used in 16% and 84%, respectively. Results: The median time to leukocyte engraftment was 14 days. Acute GvHD grade II-IV was seen in 15% of the patients, severe grade III in only 1 patient (2%). The cumulative incidence of nonrelapse mortality at 1 year was 24% (95% CI: 11-37%) and of relapse at 3 years was 43% (95% CI: 27-60%). The 3-year progression-free and overall survival were 33% (95% CI: 17-49%) and 60%, respectively. The NRM was similar in patients who received single agent thiotepa/ fl udarabine only in comparison to those who received Introduction: Renal impairment is frequently encountered in Multiple Myeloma and AL Amyloidosis patients, with 20% of patients presenting in renal failure and 5% requiring dialysis. This is often perceived as a contra indication to eff ective high dose therapy which may have the potential of not only getting the patient into a deeper remission but also reversing their renal dysfunction. Autologous stem cell transplants can be performed safely in end stage renal impairment in Multiple Myeloma and AL Amyloidosis patients. We present our single centre experience. Materials (or patients) and Methods: Between the periods of 2008 and 2012, there have been seven successful autologous stem cell transplantations in patients with dialysis dependent renal failure. Five patients had Multiple Myeloma and two AL Amyloidosis. There were fi ve males and two females with a median age of 60 years (52-65). All patients underwent an autologous stem cell transplant in fi rst remission. Six patients were on haemodialysis and one patient was on peritoneal dialysis. Only one patient was admitted to intensive care during their inpatient stay at time of stem cell transplant. All patients were transplanted at Manchester Royal Infi rmary, Manchester, UK, using Melphalan 140mg/m 2 . They all received inpatient dialysis under the care of a renal physician (two to three times a week according to individual pre transplant schedule). One patient was on peritoneal dialysis which was performed successfully with no infections during the neutropenic phase. There was no increase in dialysis requirements during the inpatient stay. Results: The median follow up was 3.6 years (1-5.3 ). Four of the seven patients became dialysis independent, two of whom received renal transplants. One patient resumed dialysis following decline in renal function after a 2 year period. However, during the period off dialysis he reported enhanced quality of life. One patient died following neutropenic sepsis and multi-organ failure within 100 days of transplantation. A second patient died 15 months post transplant from severe mitral regurgitation (a complication seen in long-term dialysis patients) having relapsed with myeloma at 12 months post stem cell transplant. All fi ve patients alive at this time point continue to be in complete remission (two AL Amyloidosis, three Multiple Myeloma). Discussion: In the correct setting, autologous stem cell transplantation is feasible in patients with dialysis dependent renal failure using Melphalan (140mg/m 2 ) conditioning in conjunction with dialysis. As demonstrated in our cohort, transplant related complications do not diff er from that of the non-dialysis patients. There is potential for cessation of dialysis, although in some cases this may only be temporary. Furthermore, autologous stem cell transplantation can induce complete remissions which subsequently may enable renal transplantation. Disclosure of Interest: None Declared. 1 Hematology Department, Salamanca, 3 Hospital Universitario Virgen del Rocio, Sevilla, Spain, 4 Hematology Department, Hospital Universitario Virgen del Rocio, Sevilla, Spain Introduction: Allogeneic hematopoietic stem cell transplantation (HCT) is a potentially curative option in MM patients. However, the introduction of new drugs together with the transplant-related mortality/morbidity and the high relapse risk of this procedure make the use of HCT controversial. Materials (or patients) and Methods: A total of 48 consecutive MM patients who underwent HCT in our center between 2000 and 2013 were retrospectively evaluated. Feasibility and effi cacy of HCT was analyzed in terms of transplant-related mortality (TRM), progression free survival (PFS) and overall survival (OS). Results: Median age was 54 years (30-65). 92% had received previous auto-transplant. Disease status at transplant was complete remission (CR) in 8 (16%), partial response or very good partial response (PR/VGPR) in 27 (56%), stable disease (SD) in 7 (10%) and progressive disease (PD) or non-responding disease (NRD) in 8 (17%). Twenty patients had previous extramedullary disease, 8 of them (17%) at transplant. Median time to HCT was 30 months (8-256) from diagnosis and 21 (3-111) from auto-transplant. Twelve patients (27%) had high-risk cytogenetic abnormalities. All the patients but one received reduced intensity regimen (fl udarabine-melphalan, plus bortezomib in 11 patients); Donor was unrelated in 13 patients (27%), 90% HLA 10/10, and median CD34-positive cells was 5,2 (2,2 -13) x10 6 /kg (from peripheral blood in 96%). Graft-versus-host disease (GVHD) prophylaxis consisted of calcineurin inhibitor-regimens plus MMF or metotrexate in 39 (81%) and tacrolimus and rapamicine in 8 (17%). All patients engrafted. Cumulative incidence of acute GVHD (aGVHD) and grades 3-4 aGVHD was 54% and 18% respectively, (median onset:31 days, 6-172). Cumulative incidence of chronic GVHD (cGVHD) and severe cGVHD was 37% and 14% respectively (median onset: 197 days, 105-778) posttransplant. Regarding response to transplant at day +100, 21/38 (55%) patients improved pre-HCT response: 5/8 of the patients who were in PD/ NRD, 2/5 of the patients with SD, 12/22 of the patients with PR and 2/3 of the patients with VGPR. Among the 6 patients who were in CR, 5 patients maintained this response. Four patients were not evaluable. 38 patients subsequently relapsed or progressed, 20 of them with extramedullary involvement. Median time to relapse or progression was 11.8 months . With a median follow up of 30 months (2-138) among patients alive, median EFS was 315 days. In multivariate analysis to develop cGVHD (mild (HR 0.31; 95% CI 0.1-0.9; P=0.04) moderate (HR=0.1(0.03-0.6) P=0.01) and severe (HR=0.1(0.05-0.6) P=0.01)) favourably infl uenced EFS, while time from auto-transplant >1 year to HCT adversely infl uenced outcome (HR 2.2(1.3-3.6) P=0.02). The estimated 5 and 10 years OS was 41% and 20%. Overall TRM was 14% (6% at day +100). The main cause of death was progression disease (n=24). In multivariate analysis to develop grade 3-4 aGVHD adversely aff ected OS .5) P=0.02), while chronic GVHD favourable infl uenced outcome (mild (HR 0.2(0.05-0.83) P=0.02) moderate (HR=0.2(0.04-0.88) P=0.03) and severe (HR=0.1(0.02-0.6) P=0.01).

Discussion: According to our results, HCT is an alternative therapy in high risk MM patients with low TRM, however risk of relapse is very high. Eff orts should be focused on new strategies to control the minimal residual disease as well as the severe GVHD after HCT, maintaining the GVM eff ect. Disclosure of Interest: None Declared. Introduction: Infl uenza infection in multiple myeloma (MM) patients is often characterized by severe complications. Infl uenza vaccination is therefore generally recommended, however, immune response to vaccination is frequently insuffi cient. The aim of our diagnostic study was to determine the immune response after one and two doses of a novel trivalent infl uenza vaccine (Optafl u®) in MM patients. Materials (or patients) and Methods: During the season 2012/13, vaccination with a trivalent (A(H1N1)pdm09, A(H3N2), B/ Yamagata) infl uenza vaccine (Optafl u®, Novartis) was administered to 49 patients with confi rmed diagnosis of MM at our outpatient clinic. Fourty-eight patients received one dose and 24 of these patients received a second dose of vaccine. Blood samples were taken prior to immunization and four weeks after each vaccination for evaluation of humoral response employing hemagglutination inhibition assays. Results: Fourty-eight MM patients were vaccinated, none had protective immunity at baseline against all three viruses. After the fi rst vaccination, seroprotection against all three antigens was achieved in 14% (7/49) of the patients. Of 42 patients without protective immunity, 24 received a second dose resulting in 33.3% (8/24) seroprotection. Patients who failed to seroconvert were more likely to receive an immunosuppressive therapy. Discussion: We demonstrated in this pilot study that a single dose of infl uenza vaccine resulted in 14% seroprotection of MM vaccines. The frequency of protective titers could be more than doubled to 33.3% by a vaccine boost. In patients with MM the rate of seroconversion after a fi rst dose of a trivalent infl uenza vaccine was poor, but increased after a second dose. In conclusion, twodose vaccination with Optafl u® vaccine resulted in a signifi cant, yet limited serological response. We therefore recommend an infl uenza vaccine boost for myeloma patients in general. A prospective, randomized study would be highly appreciated to confi rm this recommendation. Disclosure of Interest: None Declared. Introduction: Achievement of complete response (CR) has been considered a new goal of therapy for multiple myeloma (MM). The depth of the response may also be important. We have used a sensitive real-time quantitative polymerase chain reaction (qASO-PCR) to assess the level of minimal residual disease (MRD) in bone marrow of myeloma patients who had achieved CR or near-to CR (nCR) after upfront autologous stem cell transplantation (ASCT). Low or negative MRD was earlier shown to predict the ASCT

G. Bos 1,* , S. Sarkar 2 , L. Wieten 2 , A. Martens 3 , W. Noort 3 , R. Groen 3 1 Internal Medicine, Hematology, 2 Maastricht University Medical Centre, Maastricht, 3 University Utrecht, Utrecht, Netherlands Introduction: Multiple Myeloma (MM) is an incurable plasma cell malignancy residing in the bone marrow (BM). Cell based immunotherapy with allogeneic natural killer (NK) cells could be a new treatment option for MM because NK cells are known to mediate potent anti-tumor immunity. Here, we studied graft versus myeloma (GvM) responses and the development of graft versus host disease (GvHD) after immunotherapy with human allogeneic, KIR-HLA mismatched NK cells in in vitro and in vivo models. Materials (or patients) and Methods: Freshly isolated NK cells killed KIR-HLA mismatched human MM cell lines in a fl ow cytometry based cytotoxicity assay in vitro. Killing effi ciency varied between the cell lines and could partially be explained by diff erences in HLA expression. One of the MM cell lines (U266), was injected in immunodefi cient RAG2 -/gc -/mice. At day 21, 23 and 25 after tumor injection, allogeneic PBMC or purifi ed NK cells were intravenously injected in MM bearing mice and MM outgrowth was monitored real time by bioluminescent imaging.

Results: Infusion of unactivated NK cells decreased MM severity in one out of four treated mice. Analysis of the BM revealed that the infused NK cells had migrated to BM. Infused PBMCs effi ciently eradicated established MM in all treated mice, and in this group T cells were found in the BM. However, in PBMC receiving mice, GvM eff ects coincided with the development of GvHD which will severely hamper clinical application of PBMC infusion. On the other hand, NK cells could be infused at high numbers without causing GvHD, emphasizing the therapeutic opportunities of NK cell therapy. We additionally show that the GvM response could be enhanced by activation of the NK cells with IL-2 and by conditioning of the mice with low dose chemo/radiotherapy (50 mg/kg cyclophosphamide and 2x2gy TBI). Discussion: Our study illustrates the safety and therapeutic potential of NK cell therapy in MM. Thereby, our study can be a fi rst step towards cell based immunotherapy for MM patients without the induction of GvHD. Disclosure of Interest: None Declared. Introduction: Current experience with TCRαβ-depleted HSCT is limited to a few studies mainly in patients with malignant diseases transplanted from haploidentical donors. The key objects for this approach are reduction of GVHD risk, fast immune reconstitution and preservation of GVL-eff ect. However, infections, GVHD and GVHD-associated complications still remain signifi cant problems in patients with non-malignant diseases transplanted from haploidentical or match unrelated donors (MUD). Aim of our study was to examine the eff ect TCRαβ-graft depletion for HSCT from MUD or haploidentical donors in patients with primary immunodefi ciency diseases (PID). Materials (or patients) and Methods: In 2012-2013 twelve patients (male:female = 9:3; median age 1,8 years [0,2-6,2 years]) with different PIDs were transplanted with TCRαβ/CD19-depletion of the grafts. Indications for HSCT were CGD (n=1), Hyper-IgM (n=1), WAS (n=4), HLH (n=1), SCID (n=3) and unidentifi ed PID (n=2). Stem cell source was peripheral blood from 10/10 HLA-MUD (n=8) or haploidentical donors (n=4). All patient received Treosulfan-based reduced intensity condition regimens with top-up of Campath-1H or ATG (horse) as serotherapy. The median number of TCRαβ+ cells was 7 x 10 3 /kg (0.8 x 10 3 /kg -3.9 x 10 4 /kg) and median number of СD34+ cells was 15 x 10 6 /kg (10.3 x 10 6 /kg -61.0 x 10 6 /kg). GVHD prophylaxis included Tacrolimus (n=6) or Tacrolimus with short course of MTX (n=6). Results: Engraftment occurred in all patients; median time to neutrophil recovery was +15 day ( Day +13 -+21), median time to platelet engraftment was +12 day (Day +9 -+20). GVHD grade II occurred in 30% of patients and we had no patients with GVHD grade III-IV. There were two late graft rejections (patients with CGD, [n=1] and Hyper-IgM, [n=1]) 3 and 6 months after primary HSCT; second HSCT were performed in both cases with good preliminary results. Other serious complications after HSCT were CMV-retinitis (n = 1) and secondary hemophagocytic syndrome associated with BCGitis (n = 1). Introduction: Treosulfan-based conditioning for haematopoietic stem cell transplantation (SCT) shows limited extramedullar toxicity and thus appears favorable especially in patients with nonmalignant diseases. Materials (or patients) and Methods: Retrospective analysis of clinically signifi cant mixed chimerism (MC; defi ned as >10% recipient cells) in 13 patients (median age 2 years) with non-malignant diseases after treosulfan-based conditioning from 2009 to 2013. Underlying diseases were Wiskott-Aldrich syndrome (WAS) (n=2), mucopolysaccharidosis type I (MPS I) (n=1), immunodefi ciency syndrome (ID) (n=5), hemophagocytic lymphohistiocytosis (HLH) (n=1) and beta-thalassemia major (TM) (n=4). Donors were 10/10 HLA-matched unrelated (n=5), sibling (n=7) and one haploidentical donor. All patients received treosulfan and fl udarabine, while thiotepa was added in seven cases. Pre-transplantat serotherapy (n=12) consisted of anti-thymocyte globulin (n=9) or alemtuzumab (n=3). Graft-versus-host disease (GvHD) prophylaxis was mainly based on cyclosporine (CSA) (n=12) with addition of methotrexate and/or mycophenolate mofetil in several cases. Chimerism was analyzed by short tandem repeats or XY-FISH. Results: Conditioning was well tolerated, however, two patients with ID died due to pre-existing severe infections prior to engraftment. One patient suff ered graft rejection after haploidentical transplantation which was successfully rescued after second haploidentical transplantation. Ten patients achieved primary engraftment with median neutrophil engraftment on day (d) 19.5 (range d 15-36). Mixed chimerism occurred in four of the engrafted patients (40%) (MPS I, WAS, TM, ID). In one patient (ID) MC spontaneously dropped from 24.3% to 2.1% during fi rst month and turned into full donor chimerism after day 100 (table1). MC persisted in three cases for a median follow up of 484 days (range 364-629 d) and rose to a median of 80% (range 60-90%). MC was detectable in T-and B-cell, myeloid and monocyte compartments in two patients (WAS, TM) and in T-cell and myeloid compartments in one patient (MPS I). CSA was thus tapered and two patients additionally received donor lymphocyte infusions (DLI). While no GvHD occurred, MC increased further despite these interventions (table1). Notably however, only one patient (WAS) presented with recurrence of disease symptoms, i.e. low platelets requiring occasional platelet transfusions. Due to stable MC in the T-cell compartment (40-60%) retransplantation could be avoided and platelet count stabilized after splenectomy. Discussion: Children with non-malignant diseases transplanted with treosulfan-based low toxicity conditioning appear to be at signifi cant risk to develop MC. Interventions to reverse MC (i.e. tapering of CSA and DLI) seem of limited value. Since MC is not uniformly associated with symptoms of disease recurrence any intervention requires careful risk-benefi t assessment. Long-term outcome with MC after treosulfan-based conditioning remains unclear and urgently warrants further analyses in larger prospective studies. Disclosure of Interest: None Declared.

[PH-P538] 3) x 10 5 /kg; 93% were 0/1 HLA mismatched. Outcome: The overall survival (OS) for NM diseases at 2 years was 76%, with HLH patients doing the worst (49%). The median day to neutrophil (>0.5) and platelet (>20) recovery was 19 and 32 days respectively. The incidence of aGvHD (II-IV) and cGvHD was 20% and 8% respectively. The OS for the M group was 52% at 3 years, with a signifi cantly worse outcome for lymphoproliferative disease (25%, P=0.01) and trends to better outcome with a TNC dose infused of >5 x 10 7 /kg (63 v 46%; P=0.06), and 6/6 matched cords (69% (6/6) v 50% (5/6) v 41% (4/6); P=0.08). Median days to neutrophil (>0.5) and platelet (>20) recovery were 24 and 42 days respectively, and the incidence of aGvHD (II-IV) and cGvHD was 26% and 9% respectively. TRM was 15% at 100 days and 21% at 1 year and relapse 19% and 27% at 1 and 3 years respectively. There was no infl uence of year of CBT before/after 2009, the use of reduced intensity conditioning or the use of serotherapy. Focusing on children with acute leukaemia (n=130), the 3 year OS for AML was 58% and for ALL 48%; there was a trend to reduced survival with advancing disease status in ALL: CR1 64%, ≥ CR2 38%, and advanced disease 22% (P=0.08); surprisingly there was no infl uence of disease status in AML: CR1 68%, ≥ CR2 51%, and advanced disease 62% (P=0.74), although the numbers are fairly small.

Discussion: Despite the relative reluctance to use CBT in the UK the outcomes are good and very comparable to adult unrelated donor transplantation. HLH patients appear to do less well following CBT. The surprisingly good outcome (>50%) in advanced AML suggests the potential for a strong graft-versus-leukaemia eff ect but requires confi rmation in a larger study. Disclosure of Interest: None Declared. Introduction: Human Adenovirus (HAdV) viremia is a frequent and potentially fatal complication after pediatric allogeneic hematopoietic stem cell transplantation (HSCT). HAdV infections are generally treated with cidofovir, but studies on the eff ectiveness of cidofovir treatment are inconsistent and often lack data on the immunological status of cidofovir treated patients. Materials (or patients) and Methods: In 36 children treated with cidofovir for HAdV viremia after HSCT, the change of plasma HAdV DNA load was measured 14 days after treatment initiation and related to concomitant lymphocyte reconstitution. Acute and chronic nephrotoxicity of cidofovir were evaluated by monitoring of glomerular and tubular kidney function. Results: During cidofovir treatment, the viral load increased ≥1 log in 7 cases (19%), viral load stabilization occurred in 17 cases (57%) and ≥1 log viral load reduction was measured in 12 cases (33%). Viral load reduction was always accompanied by lymphocyte reconstitution and in only 5 cases (14%), viral load stabilization was reached in the absence of lymphocyte reconstitution. Eventually, HAdV viremia was cleared in 27/36 cases (75%). Glomerular and tubular nephrotoxicity occurred in 23% and 31% of cases, respectively. At latest follow up, 2/21 long-term survivors (10%) had cidofovir related chronic kidney disease. Discussion: HAdV viremia progressed despite of cidofovir treatment in a substantial number of cases. A subgroup of patients possibly benefi ted from cidofovir through stabilization of the viral load but treatment was associated with considerable nephrotoxicity. Lymphocyte reconstitution was essential for reduction of the viral load during cidofovir treatment. Disclosure of Interest: None Declared. Introduction: The FLAMSA sequential treatment with chemotherapy followed by reduced-intensity conditioning (RIC) for allogeneic stem cell transplantation (allo-SCT) has been introduced few years ago for adult refractory acute myeloid leukemia (AML) showing high activity and relatively good survivals in this particular setting (Schmid, Blood, 2006) . There is no study at our knowledge reporting the results of the sequential approach in pediatric refractory AML patients. Here we report our own experience in 6 children using a debulking chemotherapy combining clofarabine and Ara-C followed by RIC before allo-SCT. Materials (or patients) and Methods: These preliminary results included 3 males and 3 females with a median age of 8,5 years (range: [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] . All cases had received a sequential regimen before allo-SCT at the CHU of Nantes (n=5) or at the CHU of Montpellier (n=1) for primary refractory AML (n=1), refractory relapsed AML (n=3), slow responder relapsed AML (n=1) and blastic JMML (n=1). Sequential regimen consisted of 1) clofarabine 30 mg/m2/d days-13 to -9, Ara-C 1g/m2/d days-13 to-9 followed by RIC combining cyclophosphamide 60 mg/kg/d day-5, iv Busulfan 3.2 mg/Kg/d days -4 to -3 and ATG 2.5 mg/Kg/d days -3 to -2 in 5 patients or 2) clofarabine 30 mg/m2/d days-13 to -9, Ara-C 1g/m2/d days-13 to-9 followed by RIC total body irradiation 4 grays day-5, cyclophosphamide 40 mg/Kg/d days -4 to -3, and ATG 2.5 mg/Kg/d days -3 to -2 in 1 patient. One patient received a graft from a sibling donor while the fi ve other patients received a graft from an unrelated donor (10/10 n=3; 9/10 n=2). GvHD prophylaxis consist in Ciclosporin and Mycophenolate Mofetil (MMF) (n=5), or Ciclosporin and Methotrexate (n=1). Results: Engraftment was observed in 4 patients (67%) who achieved complete remission (CR) after transplant. The 2 patients which had an autologous reconstitution, relapsed and died rapidly. Considering the 4 patients achieving full engraftment and CR, two relapsed at day+60 and at 13 months post-allograft. The fi rst one died. The two others patients are alive in CR at +35 and +51 months post-transplant. Discussion: Patients with primary refractory and relapsed AML are rarely cured without undergoing allo-SCT. The treatment's management is currently debated. Using re-induction chemotherapy with the aim of obtaining a second CR before allo-SCT after myeloablative conditioning may leads to added organ severe toxicity, and low poor CR rate. Newer strategies using "sequential allo-SCT" have been developed and increased OS and LFS in patient with high risk diseases. Our pilot study shows encouraging results in survival, consistent with adults series. In our experience, cytoreductive chemotherapy and conditioning were well tolerated, as in the adult's study. To our knowledge, this is the fi rst report of a sequential allo-SCT approach for refractory pediatric AML patients. Although the number of patients is limited in our cohort, the results showed here are very encouraging as 2 out of the 6 patients are alive in CR with full engraftment. These results have to be confi rmed prospectively. Disclosure of Interest: None Declared. , age, body weight, height, renal function and use of diuretics). Paediatric dose selection was aimed to target a similar exposure in children as obtained in adults after intravenous administration of the clinically eff ective dose of 14 g/m 2 /day treosulfan for 3 consecutive days for conditioning prior HSCT. Results: When doses were calculated in g/m 2 , the exposure to treosulfan (area under the curve, AUC) was signifi cantly increased in children, compared to adults. The PopPK model for treosulfan consisted of two compartments with a fi rst order distribution and elimination processes. A covariate analysis revealed that BSA was the only relevant covariate for clearance and volumes of distribution. The PopPK model provided an adequate fi t to the data and model diagnostics revealed no signifi cant bias. Based on the estimated PopPK model, the functional relationship between clearance and BSA could be derived (see picture). For feasibility purposes, a simplifi ed scheme of treosulfan dosing was deduced thereof to be applied in the planned paediatric studies:

BSA Daily treosulfan Dose ≤ 0.5 m 2 10 g/m 2 > 0.5 -1 m 2 12 g/m 2 > 1.0 m 2 14 g/m 2

Discussion: The PopPK model for treosulfan is robust and accurately predicted exposure in children older than 1 year of age. For infants a best estimate could be given only as the dataset merely included limited data of children below 1 year of age. The planned paediatric transplant trials will start with the defl ected dosing table and will update the PopPK model accordingly to further investigate its validity in children with malignant or non-malignant diseases qualifying for HSCT. Disclosure of Interest: P. Van Den Berg Confl ict with: medac for paid consultancy services, M. Ruppert Confl ict with: medac Introduction: Immune-mediated post hematopoietic stem cell transplant (HSCT) cytopenias comprise a rare and severe group of diseases. They are more frequent following allogeneic HSCT, although they are also described in the autologous setting. Immune cytopenias respond poorly to standard treatment and are associated with a high rate of morbidity and mortality. The most common presentation is haemolyitic anemia, followed by Evans Syndrome (immune thrombocytopenia plus haemolyitic anemia). Materials (or patients) and Methods: we have searched for haemolytic anemia episodes among pediatric patients who underwent HSCT in our center between years 2004 and 2011. Of each patient, data concerning primary disease, type of transplant, stem cell source, conditioning treatment, graft versus host disease (prophylaxis regimen, and clinical features and treatment when applicable), were collected; and laboratory parameters, treatment administered and outcome of each haemolytic anemia episode were also recorded. Results: Out of 102 allogeneic HSCT performed in our center between 2004 and 2011, 9 cases of haemolytic anemia were found. Ages at the time of transplant ranged from 6 months and 12 years old. Of the 9 patients, 6 had a hematologic malignancy as primary disease (three acute lymphoblastic leukemias, two acute myeloblastic leukemias and one chronic myeloid leukemia); the 3 remaining patients had an immune defi ciency. Three patients had an HLA-identical sibling donor, another two an HLA-identical non related donor, three a mismatched unrelated donor, and one patient received a haploidentical transplant. Peripheral blood apheresis was the stem cell source in 4 patients, umbilical cord blood in 3, and bone marrow in 2. Only two patients received a reduced intensity conditioning. Graft manipulation (T cell depletion) was performed in two cases. All patients engrafted, with full chimerism achieved in 8 patients. GVHD prophylaxis was performed with cyclosporin in 7 patients; 2 received also methotrexate, 2 mycophenolate mofetil, and 1 prednisone. Five patients had acute GVHD. Haemolytic anemia was developed in the early postransplant period, with the exception of a patient who had an episode 3 years after HSCT. All of them had increased bilirubin and LDH, decreased haptoglobin leves, and a positive Coombs test, IgG positive in 8 patients (the remaining case was passenger lymphocyte mediated). All patients were treated initially with steroids, adding immunoglobulin in fi ve of them. Three patients received rituximab. Of the nine patients, six required intensive care unit admission. Three patients died due to direct complications of the process; the remaining six patients had a full recovery of the anemia. Discussion: our experience in post-HSCT haemolytic anemia shows that it is a severe disease. 66% of our patients required intensive care and mortality was of 33%. More than half of our patients required add-on treatment over steroids. All our patients had received transplant-related aggresive therapy, with myeloablative conditioning in 77% patients and immunosuppressive treatment in all of them; these measures often result in the depletion of the T-cell compartment, with a secondary dysregulation of

Disclosure of Interest: None Declared. Discussion: HBV reactivation in highly immunosuppressed population may be a not infrequent and severe late complication. Active prophylaxis with nucleoside analogues are essential, but not suffi cient to control HBV reactivation

In addition, the median C-reactive protein level was lower in allo-SCT recipients (53 mg/L, 13-146) compared to the other patients (143.5 mg/L, 22.5-285), P=0.023. It is noteworthy that the 7 cases of IA in allo-SCT recipients occurred only in patients treated with corticosteroids for graft-versus

Complete remission of IA on CT was observed in 12 cases, stable lesions in 8 cases and progression in 9 cases. Mortality directly related to IA was 24% in our study and occurred early after the onset of infection. The 100-day, 1-year, 2-year and 10-year overall survival (OS)

5%) pts renal function was normal. In this group 76 (57.6%) pts had myeloma G, 27 (20,4%) -myeloma A, 23 (17.5) -light-chain myeloma, 3 (2.25%) -non-secretory myeloma, 1 (0.75%) -myeloma D and 1 (0.75%) -plasma-cell leukemia. Pts received VAD chemotherapy and/or bortezomib + dexamethasone +/-doxoadministration of G-CSF, pts with end-stage RF received G-CSF only. HDT/ASCT entailed conditioning with melphalan 140-200 mg/m 2 . Response was defi ned according to IMWG criteria. Results: After the induction the overall response (CR+VGPR+PR) was 90,1% with 34,8% of CR and 31,8% of VGPR in the group of pts with normal renal function. Among pts with RF overall response was achieved in 93,7% pts with 53,1% of CR and 21,9% of VGPR. Renal response was developed in 24/32 (75%) pts after induction therapy. An improvement in glomerular fi ltration rate to >15 ml/min/1,73 m 2 was noted in 2 more cases post HDT/ASCT. 1 pt became dialysis-independent after induction, another 2 pts -after ASCT. The treatment-related toxicity was slightly higher in pts with end-stage RF after ASCT, all of them required i.v. antibiotics due to neutropenic fever, while 30% pts with normal renal function do not undergo any infectious complications. The median time of follow-up after ASCT was 30 months (2,5-127). In all 164 pts the median of OS was not reached, the probability of PFS at 5-y was 45,3%. Transplant-related mortality was 0,6%. 138 pts (84,1%) are still alive. There was no signifi cant diff erence in PFS and OS in pts with or without RF at the time of diagnosis and ASCT. Discussion: In pts with myeloma and RF

Cavo 2 , R. Fanin 1 1 Division of Hematology

Positron emission tomography integrated with computed tomography (PET/CT) has been reported to be useful for screening of myelomatous lesions in multiple myeloma (MM) patients at diagnosis and for monitoring response to therapy and prognosis after autologous stem cell transplantation (auto-SCT). Aim of the study was to prospectively evaluate the prognostic signifi cance of PET/CT in MM patients who received allogeneic stem cell transplantation

On univariate analysis, persistence of at least 1 FL, SUV > 4.2 and EMD before allo-SCT adversely aff ected PFS , TTP,and OS. In a Cox regression analysis of prognostic factors before allo-SCT, persistence of EMD at transplant was an independent predictor of worst TTP and PFS. OS was negatively infl uenced by unrelated donor and SUV higher than 4.2. Six months after allo-SCT, PET-CT was negative in 27/59 pts (46%), whose 2-year rate of OS was superior to that of PET-positive pts (80% vs 63%, P=0.01). Multivariate analysis of pre transplant and post treatment variables showed that persistence of EMD and failure to obtain CR or VGPR after allo-SCT were strongly associated with shorter PFS, TTP and OS. Moreover, allo-SCT at relapse was an additional independent predictor of worse OS. Discussion: Our study shows that PET-CT adds information of prognostic signifi cance to conventional evaluation of clinical and biochemical response before and after allo-SCT in MM. Persistence of EMD by PET-CT before and after allo-SCT was an independent prognostic factor for poor PFS

Grade 3-4 non-hematologic toxicity was seen in 40 (70%) patients, with no signifi cant diff erences between the 4 dose levels. Two patients died of nonrelapse causes (viral infection 1, cardiac failure 1) with a TRM of 3%. Median time to both neutrophil and platelet engraftment was 11 days. At day 90, 8 (14%) patients achieved a CR, 25 (44%) achieved CR or very good partial response (VGPR), and 42 (74%) achieved CR, VGPR or partial response (PR), with no signifi cant diff erences in response rates among the 4 LEN dose levels. By day 180 and post auto-HCT, 12 patients (21%) had achieved a CR. Thirty-six (63%) patients received post auto-HCT maintenance therapy (lenalidomide 32, bortezomib 2, and pomalidomide 2). With a median follow up of 12.3 months (range 0.5-41), median progression-free (PFS) was 24 months and median overall survival (OS) has not yet been reached. Two-year PFS and OS were 48% and 72%, respectively. These compare favorably with our previous experience of 2-year PFS and OS of 24% and 59%, respectively, after salvage auto-HCT with melphalan alone. There was no signifi cant diff erence in median PFS (13 months) and OS (not reached yet) in patients undergoing a 2 nd auto-HCT, compared to patients undergoing fi rst auto-HCT (P=0.47 and 0.85, respectively). No second primary malignancy (SPM) has been seen so far. Discussion: LEN up to 100 mg daily x 7 days can be safely combined with high-dose melphalan for relapsed/refractory myeloma. The regimen is associated with a high overall response rate, 2 year PFS and OS of 48% and 72% in relapsed and heavily pretreated patients, and no SPM after a follow up extending to 41 months

Differences between the COX-2-positive and COX-2-negative patients in gender, hemoglobin, β2-microglobulin (β2M), creatinine, albumin, and disease stage were not statistically significant. Discussion: COX-2 expression neither had a role in prognosis nor signifi cantly aff ected OS and PFS. We conclude that stem cell transplantation might eliminate the detrimental eff ects of COX-2 positivity. Larger series of patients are needed to investigate this observation. Disclosure of Interest: None Declared. Introduction: Despite recent advances in treatment of multiple myeloma (MM), the disease is still incurable and prognosis for high-risk patients is dubious. Long-term remissions have been achieved after allogeneic transplantation (alloSCT). However, relapse rate and treatment-related mortality (TRM) can be substantial

Methods: Eligible for this retrospective analysis were patients with advanced MM who were considered to have poor prognostic factors and were allografted after TBI8/ CY between 09/09 and 06/13. Patients received 2x2 Gy TBI days -5 and -4, cyclophosphamide 40 mg/kg BW days -3 and -2, and ATG in case of unrelated donors (10 mg/kg BW in matched unrelated donors, MUD, and 20 mg/kg BW in mismatch donors, MMD) days -3 to -1. Ciclosporine, tacrolimus or sirolimus (in case of renal insuffi ciency) combined with mycophenolat mofetil (MMF) were used for GVHD prophylaxis. Results: Twenty-four patients referred for allografting met the criteria

n=2; relapse after 1 st alloSCT, n=1) and were thus conditioned with TBI8/CY. Median age and follow-up were 53 years [range 38 -65] and 24.1 months, respectively. Remission status prior to alloSCT was at least VGPR in 8 patients

Cy (120mg/kgBW) and Mel (140 mg/m 2 ). MNC were collected mostly with one "large volume" apheresis and median number was 9.85x10 8 /kg BW (range 3.8-28x10 8 /kg BW). OS and duration of response were showed with Kaplan Meier method. Peripheral blood progenitor cells were the preferred source of stem cells. The defi nition of response is estimated by International Myeloma Working Group in 2006. Results: In 112 pts engraftment was observed on +12 day after auto-HSCT (range 8-36, respectively), in 3 pts there were no reconstitution, in 3 pts it was delayed and 2 pts have died during the procedure. Overall response rate (ORR) is achieved in 72 (61%) pts. After auto-HSCT, 72 pts (61%) have achieved CR+VGPR, while 35 pts (29,7%) were in PR. 21 pts (17,5%) have died and 91 (75,8%) are still alive. 8 pts (6,7%) were lost from following . Median OS in 115 49 months. Discussion: Achievement of CR+VGPR before and after auto-HSCT is important predictive factor for PFS and OS. Better OS and PFS

Disclosure of Interest: None Declared. Introduction: Autologous haemopoetic stem cell transplantation (AHCT) and novel agent-based (NAB) therapies have improved multiple myeloma (MM) outcome. However, prognostic factors for post-AHCT survival in the era of novel agents are not clearly defi ned and data regarding the role of post-AHCT consolidation/ maintenance (C/M) therapy are inconclusive. The aim of the study was to record "real world data" on clinical features, prognosis and post-AHCT survival of a large cohort of consecutive MM patients treated with AHCT in Greece

Results: At MM diagnosis 189 patients had ISS 1, 141 had ISS 2 and 90 had ISS 3 (no data: 69 patients)

6-4.9 g/dL), 1.17 mg/ L (1.0-2.2mg/L) and 0.87 mg/dL (0.59-2.52mg/dL), respectively; the median % of bone marrow plasma cells before AHCT was 3% (range: 0-40%); 378 patients were mobilized with cyclophosphamide + GCSF, 79 patients with GCSF and 32 patients with multiagent chemotherapy. The median number of CD34+ cells infused was 5.9 x10 6 /kg (range: 2.6-7.43 x10 6 /kg). Median days for neutrophils >500/μL were 11.5 (range: 8-18); median days for PLT>

266 patients received post-AHCT C/M treatment (consolidation: 80, maintenance: 186) and 223 did not

At the last follow-up, 337 patients were alive (68.9%) and 149 patients (30.5%) died (progression: 88%, AHCT-related: 2%, secondary malignancy: 2%). With a median post-AHCT follow-up of 56 months (95% CI: 51.2-60.7) the median post-AHCT time to progression and post-AHCT survival were 46 months (95% CI: 38-54) and 101 months (95% CI: 79-122), respectively

Discussion: In the era of novel agents, patients treated with AHCT enjoy prolonged survival outcomes. Pre-AHCT NAB treatment and C/M therapy independent of the type of agents used predict positively for post-AHCT survival, indicating the importance of continuous treatment in MM. Despite the use of novel strategies, high risk features is a strong negative predictor for post AHCT survival

From a total of 43 patients who were randomly selected for molecular analysis, 25 had reached low/negative MRD status, and 18 had high MRD. The groups were otherwise well balanced but in the MRDhigh group the proportion of nCR patients was higher, 8 out of 18, while the respective fi gure in the MRDlow/neg group was two out of 25

Discussion: Negative or low MRD is associated with a prolonged PFS but seems not to lead to a statistically signifi cant benefi t in OS when compared to high MRD in patients with a good response (CR or nCR) after upfront ASCT for MM. The most apparent cause for non-benefi t in OS is probably the use of novel drugs, viz. bortezomib and lenalidomide

Discussion: TCR αβ/CD19 depletion -promising alternative technology for clinical application in haploidentical or matched unrelated transplantation. Low incidence of GVHD and rate of immune reconstitution are suffi ciently factors for decreasing of TRM and increasing the chance to long-term survival

Danby 2,3,4 , A

Sheffi eld Children's Hospital, Sheffi eld, 6 Royal Victoria Infi rmary

Bristol Royal Hospital for Children, Bristol, 10 Royal Hospital for Sick Children Glasgow, Glasgow, 11 Royal Marsden Hospital

Spaans Confl ict with: medac for paid consultancy services, M. Lioznov: None Declared, U. Pichlmeier: None Declared, B. Wullf: None Declared, J. Baumgart: None Declared. PH-P545 BLINATUMOMAB IN PEDIATRIC PATIENTS WITH RELAPSED/REFRACTORY (R/R) B-CELL PRECURSOR ACUTE LYMPHOBLASTIC LEUKAEMIA

Children's Hospital of Philadelphia, Philadelphia, United States, 10 Division for Stem Cell Transplantation and Immunology, Hospital for Children and Adolescents III

Blinatumomab is an investigational bispecifi c T-cell engager (BiTE®) antibody that has been shown to produce complete remissions (CR) in CD19-positive BCP-ALL, thus providing a bridge to hematopoietic stem cell transplantation (HSCT), in an exploratory study in adults with r/r ALL. Cytokine-release syndrome (CRS) and CNS toxicities are medically important adverse events (AEs) associated with blinatumomab in adult patients. We report results from the phase I portion of an ongoing phase I/II multicenter study that evaluates the optimal dose, toxicities

Results: Forty-one patients have received a total of 73 cycles. Nine (22%) patients had refractory disease and 7 (17%) had experienced ≥2 bone marrow relapses. Twenty-fi ve (61%) patients had relapsed after allogeneic HSCT; none had prior autologous HSCT. CRS was dose-limiting. MTD was established at 15 μg/m2/ day based on dose-limiting toxicity assessment and data safety monitoring board recommendation. To reduce the risk of CRS, a stepwise dose of 5-15 μg/m2/day was recommended for the phase II portion of the study. The phase II recommended dose was further evaluated in 18 patients across two age groups (2-6 and 7-17 years) in the phase I expansion part with PK analysis. One (6%) of those 18 patients developed CRS (grade 3)

Using body surface-area dosing, blinatumomab showed linear pharmacokinetics in paediatric patients. Transient elevations of serum cytokines (particularly interleukin-6, interferon-gamma, and interleukin-10) were observed in most patients, primarily within the fi rst 2 days initiation of infusion. Discussion: In the phase I portion of this study in paediatric patients with r/r BCP-ALL, 15 μg/m2/day was established as MTD. CRS was dose-limiting, but administration of a stepwise dose, 5-15 μg/m2/day, has been successful in ameliorating CRS. Thirtyseven percent of patients achieved CR, and more than half of those responders were able to receive allogeneic HSCT, thus confi rming a potential bridge-to-transplantation role of blinatumomab

None Declared, A. von Stackelberg Confl ict with

We recorded response, transplantation-related deaths, and other adverse events after infusion of mesenchymal stem cells. Results: The median age of the patients was 7 years (min-max range: 2-18 patients received one dose, two received two doses, and one received three doses. No patients had side-eff ects during or immediately after infusions of mesenchymal stem cells. Seven patients achieved complete response, two partial response after mesenchymal stem cells infusion. The total eff ective rate was 60% (9/15)

/38 (29%) were unconditioned infusions (UCC). 24/38 (63.2%) received immunosuppression with CSA alone or in combination with methylprenisolone or mycophenolate mofetil. Mean CD34+ cell dose in CC was 7.66x10 6 /kg (0.48x10 6 /kg for cord blood transplant) and 8.25x10 6 /kg for UCC. 28/38 (74%) patients are alive and well. 10/38 (26%) patients died (8 in immediate post transplant period, 2 died 2.5 years post transplant), 8/27 (30%) CC patients died (2 due to direct regimen related toxicity, 6 due to fl aring of pre-existing infection) and 2/11 (18%) UCC patients died. 13 patients developed grade I-II GVHD mostly skin, 1 grade IV liver GVHD requiring liver transplant. Immune reconstitution and function was analysed in 30/38 patients (28 alive till date, 2 late deaths at 2.5 years). All have 100% donor T cell chimerism, 26/30 (87%) have naive T cells >200

Of 9 patients who received Bu8Cy conditioning and TCD marrow, 3 developed myeloid chimerism, all of whom had Campath 1-M TCD, compared to 6 who had cellselected TCD, none of whom had myeloid chimerism. Discussion: Although small numbers, these data suggest that CgC/ JAK3 SCID have good T cell immunoreconstitution and function irrespective of whether they receive conditioning chemotherapy. B cell immune reconstitution and function are more likely when patient receives conditioning chemotherapy. However, for those receiving TCD marrow, more chemotherapy may be required to ensure myeloid engraftment

with an estimated NRM@1 year 13 +/-6% and relapse@1 yr 18 +/-6 %. Other endpoints: neutrophil engraftment 100% @d60, trombocyte engraftment 76+/-6% @d180, aGvHD 2-4@d180 was 18 +/-6% (gr 3-4: 6 +/-4%), extensive cGvHD@1 year 8 +/-6% and no VOD was noted. One graft failure was noted in a MDS patients grafted with a MUD. He was succesfully regrafted with a cord blood graft. Discussion: This cohort of children with ALL, AML and MDS was conditioned with a chemotherapeutic regimen that showed acceptable short term safety and effi cacy

Hematopoietic Stem Cell Transplantation Unit, Dept. of Pediatric Hematology and Oncology

Introduction: Reduced intensity conditioning (RIC) regimens have improved survival of patients with congenital hemophagocytic lymphohistiocytosis (HLH) undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT). We report the experience of HSCT-Unit of Gaslini Children Hospital (GCH) in allo-HSCT

Results: Engraftment was achieved in all pts but one, who received a second HSCT because of primary graft failure. Acute-GvHD was observed in 10 cases (62,5 %) (50% grade 1-2, 12,5% grade 3); chronic-GvHD in 7 (46,6%) (13,3% limited, 33,3% extensive). A mixed chimerism was observed in one patient, without disease relapse. CNS toxicity (posterior reversible encephalopathy syndrome [PRES]) was observed in 2 pts (13,3%), who had CNS involvement at diagnosis. Neurological evaluation at last follow up showed CNS residual damage in 6 pts, represented by temporomesial sclerosis with epilepsy in 2, brain atrophy with psychomotor delay in 3 and ataxic-spastic syndrome in 1. With a median follow up after HSCT of 3.71 years (1 month -17.17 years), 12 (80%) patients are alive and disease-free and 3 died (20%) because of transplant related events (1 VOD, 1 sepsis, 1 hemorrhagic complication). Cumulative DFS probability at 3-years was 77,1% (CI 43,2 -92,2), with trends in favour of better survival for patients conditioned with RIC (MAC 71

Three patients survived after the second HSCT with good engraftment, 2 patients died. Direct toxicity of conditioning regimen was minimal. No signs of acute GvHD were observed in 33 patients (62,2%), grade I-II aGVHD -in 17 patients (32,0%), grade III-IV aGVHD -in 2 patients (3,7%)

Twelve pts received no GvHD prophylaxis after HCT, tacrolimus (tacro) and short methotrexate (Mtx) were used for 43 pts, tacro 3 pts, Mtx 6 pts. Patients were divided in 3 groups according to remission status: CR1(31pts), CR>1(25pts), active disease (AD) (8pts). Results: Primary engraftment was achieved in 63 of 64 pts, the median time to neutrophil and platelet recovery was 16 and 15 days, respectively. Cumulative incidence of acute GvHD grade > II was 43% (95% CI: 32-58), grade III -16% (95%CI: 9-32), no case of grade IV aGvHD was observed. No correlation between graft composition and aGVHD was noted. Early mortality was low with a 100-day pTRM -1.6% (95%CI: 0,2-10). Cumulative incidence of relapse at 1 year was 33% (95%CI: 21-50), 1-year pTRM -15% (95%CI: 6-35). After a median follow-up of 6.8 months (0.8-15.4), 1-year pEFS was 52%

To prevent EBV-PTLD, a single dose of Rituximab 375mg/m 2 was given on day +1. Eight patients received additional short course of MMF (until day +28). Seven patients received lowdose donor lymphocyte infusion (DLI), either due to mixed chimerism or to boost T-cell recovery. Results: Graft rejection occurred in one patient (6%), who was successfully rescued by subsequent haploSCT with another donor

None of the patients transplanted in CR has relapsed so far, 3 died due to TRM (AdV hepatitis: 1, post-DLI, GvHD: 2). EFS for patients transplanted in CR was 0.7 (0.45-1). Cumulative incidence of TRM reached 0.2. Discussion: Haploidentical stem cell transplantation is a feasible treatment option for children with acute leukemia who achieve remission prior to transplantation. Even low-dose DLI after haplo-SCT may result in severe life threatening GvHD. Disclosure of Interest: None Declared. Introduction: Therapy-related myelodysplastic syndrome (tMDS) is a serious late event for cancer survivors and carries a poor prognosis. Allogeneic hematopoietic stem cell transplantation (HSCT) is the only curative treatment

Chromosomal analysis revealed a normal karyotype (n=9), abnormalities of chromosome 7 +/-additional aberrations (n=14), a structurally complex karyotype (n=7), other (n=12) or no result (n=6). The highest WHO type before HSCT was refractory cytopenia of childhood (RCC) in 4 patients and advanced MDS in 44 patients. Donors were matched siblings (n=22), a matched family donor (n=1) or unrelated donors (n=25). Stem cell source was bone marrow (n=28) or peripheral blood (n=20). Results: Neutrophil engraftment was achieved

Relapse of tMDS and the primary malignancy after HSCT occurred in 15 patients and one patient, respectively. With a median follow-up time for survivors of 4.9 years (0.9-13 years) after HSCT, the overall and event-free survivals at 5 years were 51% (35-67%) and 42% (27-57%), respectively, with no diff erence between matched siblings and other donors. The cumulative incidence of relapse and transplant related mortality (TRM) was 32% (21-49%) and 23% (14-39%) at 5 years, respectively. Causes of death (n=23) were relapse of tMDS (n=11), TRM (n=11) and relapse of the primary malignancy (n=1)

Novichkova 3 , A. Maschan 1 1 Hematopoietic stem cell transplantation, Federal Center for pediatric hematology, 2 Hematopoietic stem cell transplantation

All patients got uniform conditioning with total doses of treosulfan 42g/m 2 , horse ATG (ATGAM) 50 mg/kg, melphalan 140 mg/m 2 , fl udarabine 150 mg/m 2 . Fifteen patients were transplanted from matched unrelated donors, 10 -from haploidentical donors, 6 -from matched sibling donors. Decitabine courses were planned to start at 10mg/m 2 after day +30 as soon as WBC reached >1x10 3 /l, ANC >0,5x10 3 /l, PLT >30x10 3 /l, Hb >95 g/l. Courses were to be administered at 30-45 day intervals with the same starting criteria. Mild aGVHD and basic GvHD prophylaxis did not preclude therapy. Results: Median time from HSCT to fi rst course was 47 (31-127) days. Median number of decitabine courses was 4 (2-6). Hematologic toxicity included grade IV neutropenia in 35%, grade III in 33%, grade II in 19%; grade IV thrombocytopenia in 7,6%, grade III in 7%. Only 9 platelet and 13 RBC transfusions were required over all courses. Hepatic toxicity included grade II ALT/AST elevation in 16,8%. Renal toxicity -transient grade I azotemia in 13%. Infections complicated therapy in 22 patients. Overall grade I infections developed during 12,2% courses, grade II in 12,2%, grade III (requiring inpatient care) -in 7.6%. Microbiologically documented infections included one case of P.aeruginosae blepharitis and two cases of gram-positive bactremia. Ten of the 32 patients experienced disease relapse, cumulative incidence of relapse being 40%

PH-P453 HEMATOPOIETIC STEM CELL TRANSPLANTATION FOR BETA-THALASSEMIA MAJOR: A SINGLE CENTER EXPERIENCE M. Ertem 1 , T. Ileri 1,* , E. Unal Ince 1 , Z. Uysal 1 1 

Introduction: Allogeneic hematopoietic stem cell transplantation (HSCT) is currently the only therapy to cure patients with thalassemia major (TM) who have a human leukocyte antigen identical donor. Materials (or patients) and Methods: All the pediatric patients who received HSCT from full-matched related donor for thalassemia major between March 1997 and September 2013 at the Pediatric Hematopoietic Stem Cell Transplantation Unit of Ankara University School of Medicine were evaluated retrospectively. Results: Median age of the patients was 8.8 y (1.9 -18.5 y) . According to Pesaro risk criteria 28 of them were low risk (Class I: 6and Class II: 22) and 46 of them were high risk (Class III) patients. All donors were HLA identical, 75 were siblings and 3 were family member. Stem cell sources were bone marrow (BM) in 55, peripheral blood (PB) in 18, cord blood (CB) in 2, CB plus PB in 2 and BM plus CB in one patient. Preparative regimen consisting of bus ulfan+cyclophosphamide (BU+CY) was used in low risk patients, BU+CY+thiotepa was used in CB transplanted patients. All class III patients except one and second transplanted patients received Pesaro protocol 26 plus ATG and one class III patient received Pesaro protocol 26. The patients whose stem cell source was CB received only cyclosporine A (CsA), all other patients received CsA+methotrexate for GVHD prophylaxis. Engraftment was achieved in all transplants except two (97%). Autologous recovery was seen in 5 patients and three of them underwent second HSCT. Aplastic bone marrow was observed in one patient and she underwent second HSCT. All patients achieved engraftment after second HSCT. In our series, overall, 73 patients (93%) were alive at a median follow-up time of 54 months (range: 3-201 months) and thalassemia-free survival rate is found as 91%. During follow-up period fi ve patients died (pneumonia: one, acute GVHD: one and very late sepsis: three patients). The total mortality rate was found as 6.5% and mortality was found signifi cantly increased patients with high risk criteria (0% vs 11%). On the other hand, two of three patients who died because of late sepsis were also splenectomized patients. Discussion: In conclusion, this evaluation demonstrated that developing countries are be able to perform HSCT successfully. Unfortunately, during the long term follow-up period risk of infection increases signifi cantly because of additive eff ects of immunosuppression to the unfavorable eff ect of splenectomy and it is not possible to achieve early intervention after hospital discharge in a developing country. Disclosure of Interest: None Declared.

A. Svyatoslavovna Borovkova 1,* Hematology and Transplantation, First Pavlov State Medical University of St.Petersburg, St.Petersburg, Russian Federation Introduction: Mucopolysaccharidosis type I (MPS I) is a lysosomal storage disorder with progressive, multiorgan involvement caused by defi ciency of the lysosomal enzyme α-L-iduronidase (AIDU), leading to accumulation of glycosaminoglycans. Patients with most severe form of MPS I (Hurler syndrome) (HS) have an early-onset, rapidly progressive disease with CNS involvement, resulting in death by 8-10 years of age. Allogeneic HSCT (alloHSCT) is still the treatment of choice for children with HS. Traditionally myeloablative regimens (MAC) were used for conditioning of MPS I patients (pts). But this group of patients has a multisystem involvement and high risk of complications. MAC is associated with a lot of treatment related toxicity that limits its usage in pts with severe disease's complications. So reduced intensity conditioning regimens (RIC) may have some advantages because they are associated with lower rates of severe toxicity. The aim of study: to evaluate effi cacy of allo-HSCT with RIC in children with HS. Materials (or patients) and Methods: From 2009 to 2013 alloHSCT were performed in 7 pts with HS. Median age at the time of diagnosis was 14 months (3-24 months) , at transplantation -29 months (19 months-40 months). All patients received enzyme replacement therapy before HSCT for 10,7 months on average (5-19 months). Matched unrelated (n=5) or partially matched (9/10) (n=2) unrelated HSCT were performed. Peripheral blood stem cells (PBSC) (n=6) and bone marrow (n=1) were used as source of hematopoietic stem cells. The conditioning regimen contained Fludarabine 150mg/m 2 +Melphalan 140 mg/m 2 +ATG 60 mg/kg. CsA+short course of methotrexate or MMF were used for GvHD prophylaxis. In 3 cases we used CD34-selection or CD3/ СD19-depletion of PBSC (CliniMACS) for prevention of GVHD. Results: 4-years overall survival is 72%. 5 pts are alive with median follow-up of 17,8 months (1.5-44 months) . During RIC alloHSCT none patients developed severe toxicity (III-IV stage according to WHO classifi cation). The median time to neutrophil recovery >0.5x10 9 /L was 19 days (range12-23 days). All patients were engrafted with achievement of full donor chimerism and normal activity of AIDU on D+30. Median range of AIDU activity in leucocytes was 61.3 nM/mg/18h on D+30 and 77.6 nM/mg/18h on D+100 (normal AIDU activity range-61.0-175.5 nM/mg/18h). Complications: grade II aGvHD of skin and gut -in 2 pts, 1 pt had grade III gut GvHD, which was successfully treated by steroids and monoclonal antibody therapy. 1pt died on D+65 from grade IV aGvHD of gut. 1 pt died from TRALI-syndrome on day+45. Discussion: AlloHSCT is an eff ective treatment option for HS, but effi cacy of usage of diff erent conditioning regimens is controversial. In general, the outcome of children undergoing HSCT is varied and depends on degree of clinical involvement and the child's age at the time of transplantation (Lorne A Clarke, Jonathan Heppner et al., 2011) , HS pts usually have multiorgan involvement, that lead to an increased risk of complications post-HSCT, especially with MAC. RIC signifi cantly decreases development of diff erent complications. But some previous studies (JJ Boelens et al., 2009 showed, that use of RIC was the risk factor for graft failure in HS patients. In our study we demonstrate that RIC is eff ective in children with HS, has low toxicity and is the promising alternative to MAC. Disclosure of Interest: None Declared.

B. Hartz 1,* , J. Schrum 1 , R. Grosse 1 , U. Kordes 1 , M. Bleeke 1 , I. Müller 1 1 Pediatric Hematology and Oncology, Hamburg, Germany Introduction: Allogeneic bone marrow transplantation (BMT) currently is the only curative therapy for patients with sickle cell anemia (SCD) 1 . These patients are at risk for developing vaso-occlusive disease (VOD) 2 . Treosulfan has been used in several conditioning regimens showing a low risk for VOD and may thus represent an interesting alternative to classic regimens 3 . Patients with thalassaemia major undergoing allogeneic BMT after treosulfan-based conditioning regimens have a high rate of successful engraftment, a low rate of complications and no case of VOD 4 .Introduction: For the last 30 years, allogeneic hematopoietic stem cell transplantation (HSCT) has been used to cure many inherited non-malignant disorders. The transplanted stem cells diff erenti- ates, migrates and replace the defi cient enzyme, abnormally shaped, non-functional or missing cells. Materials (or patients) and Methods: We report 30 years of experience in 137 patients. The disorders include 40 immunodefi ciencies, 24 benign hematological disorders, 31 histiocytic disorders,15 mucopolysaccharoidoses, 14 leukodystrophies, 6 Gaucher's disease, 1 Sandhoff 's disease, 1 erythropoietic protoporphyria, 1 Shwachman's syndrome, 1 amyotrophic lateral sclerosis, 1 osteopetrosis and 2 aspartylglucosaminuria. Their median age was 3 (0-63) years. The donors were 50 HLA-matched related (MRD), 54 matched unrelated (MUD) and 33 HLA-A, -B or -DR mismatches (MM). The stem cell sources were 108 bone marrow (BM), 14 peripheral blood stem cells (PBSC) and 15 cord blood (CB) . 85 patients received a myeloablative conditioning regimen (MAC) and 52 patients received a reduced intensity conditioning regimen (RIC). Results: Among recipients of HLA-identical sibling grafts, 10% developed acute GVHD grades II-IV as opposed to 22% in the group receiving MUD-grafts and 29% in the MM group. The overall cumulative incidence of acute GVHD grades III-IV was 4%. The overall cumulative incidence of chronic GVHD was 15%. The rejection rates were low in the MRD group (6%), in the MAC group (9%) and the group receiving PBSC (7%) and high in the MM group (24%), the RIC group (29%) and the group of patients receiving a cord blood graft (27%).The 5-year survival rates were 96%, 84%, and 55% in recipients of grafts from HLA-identical siblings, MUD and MM, respectively. The overall 10-year survival rate was 76%. The patients transplanted in between 1984-2004 had an overall 5-year survival rate of 77% (n=66) compared to 86% for the patients transplanted in between 2005-2013 (n=71) (P=0.18). The surviving patients with immunodefi ciencies and hematological disorders are all well. The patients with mucopolysaccharoidoses are all well except for their skeletal problems. All patients with Gaucher's disease are alive. One rejected the graft and is well with enzyme replacement therapy, the others are well with donor enzyme levels. Among the leukodystrophies, the treatment failed to halt the disease in the infantile cases, but arrested the progression in the juvenile and adult cases. Discussion: In many patients with inborn errors of metabolism, HSCT remains the only curable treatment. HSCT however often remains controversial. The effi ciency and safety is diffi cult to prove in studies because of the small number of patients representing each disease. In this study, we looked at the whole group of patients transplanted for non-malignant disorders, mainly to evaluate the safety of the treatment. The overall 5-year survival for the group has improved in more recent years, conformably with the overall development of HSCT. In a vast majority of cases, the HSCT managed to either permanently halt or even to reverse the disease. In conclusion, HSCT provides a high overall survival in patients with inborn errors of metabolism, with an increased survival rate in the last years. The best survival rates are seen using PBSC from HLA-matched donors and the most benefi cial eff ects on the disease are seen in those who are transplanted early after time of diagnosis. Disclosure of Interest: None Declared. However morbidity and mortality of this procedure are high in these patients in particular with donor other than geno-identical sibling. These poor results are related to high rate of severe graft-versus-host disease (GVHD) and serious viral infectious com-plications often present before the HSCT procedure. Mice studies suggest that GVHD is mostly mediated by naïve T cells while memory T cells have anti-infectious properties with absent or strongly reduced capacity to induce GVHD (Anderson BE et al., 2003) . Thus, removing CD45RA cells from donor lymphocytes could clear opportunistic and viral infections without induction of GVHD. This procedure was evaluated in 5 patients. Materials (or patients) and Methods: 5 CID patients of age between 6 and 24 months (1 ORAI1 defi ciency, 2 MHC class II defi ciency, 2 CD25 defi ciency) with >1 viral infections (n=5) and autoimmunity (n=3) were transplanted with an immunoselected CD34+ graft followed by infusion of CD45RA negative T cell fraction. They were all transplanted with a 9/10 HLA donor (intrafamilial n=1, unrelated n=4). Conditioning regimen was myeloablative, based on busulfan, fl udarabine (and thiothepa in 4 cases). Serotherapy was administrated in 4 patients (upfront in 3 cases) CD34+ cell selection and CD45RA cell depletion procedures were performed using the Clini Macs system (Miltenyi Biotec). Grafts contain 2.9 to 12.3 x10 6 CD34+ cells/kg (median 8.8 x10 6 CD34+ cells/kg) with less than 5000 T lymphocytes/kg and 0.9 and 9.2x10 6 /kg CD45RO+ T cells (median 2.52x10 6 /kg). GVH prophylaxis was based on ciclospirine (D-1 to 6 months post HSCT) and mycophenolate mofetyl (D0 to D+45). Results: Four patients engrafted, one presented a mixt chimerim requiring unmanipulated donor lymphocytes infusion (DLI) on D+35. One patient rejected and died of infection in the absence of autologous reconstitution. Grade Materials (or patients) and Methods: The fi rst successful HSCT was performed in Nigeria in September 2011 for a seven years old sickle cell disease (SCD) patient with cerebrovascular accident (CVA). The second HSCT was a 12-years old SCD patient with recurrent vaso-oclusive crises (VOC) in August 2012 and a re-transplant in May 2013. The third HSCT was for a 15years old SCD patient with recurrent VOC in July 2013. All had Allogeneic HSCT from identical siblings and conditioning regimen was with reduced intensity conditioning (RIC) (FLU/BU) Fludarabin 160mg/m 2 days -6 to -2, oral Busulphan 16mg/kg days -5 to -2. GVHD prophylaxis and immunosuppression was with ATG(ATGAM) 73mg/kg total dose days -6 to -4, cyclosporine A and MMF. There was primary rejection for the second HSCT and had a second HSCT with CY/BU (cyclophosphamide 100mg/kg before and 100mg/kg day +3 and +4).

Results: There was no recorded acute or chronic GVHD in the patients. First patient had complications of infection with pseudomonas organism at the catheter site sensitive to ofl oxacine.Introduction: Mycamine® (micafungin sodium) is an echinocandin indicated in adults (>=16 years old) and children (<16 year old, including neonates) for the treatment of invasive candidiasis, and for prophylaxis of Candida infection in patients undergoing allogeneic haematopoietic stem cell transplantation or patients who are expected to have neutropenia for 10 or more days. Mycamine® is also indicated in adult for the treatment of esophageal candidiasis. MYRIADE was an observational, multicenter, national, prospective, longitudinal study conducted between January 2010 and December 2012, in patients treated with micafungin, among 117 physicians pre-selected for their potential involvement in treating invasive fungal infections. The main objective was to describe the use of micafungin in daily practice, in terms of reasons for prescriptions, therapeutic scheme, patient and physicians profi les and to assess prophylactic or curative patient response to treatment and potential adverse events as reported by physicians.Here we report the results from haematology patients. Materials (or patients) and Methods: Treatment and all management procedures/exams were chosen by physicians according to their normal practice. Inclusion criteria: informed patients who were receiving a treatment with micafungin. Exclusion criteria: patient who were participating in another micafungin clinical trial or who couldn't be followed up. Data collection was performed at the initiation of micafungin (baseline) and at the end of treatment (withdrawal or not Introduction: Immunization with live-attenuated vaccines (measles, rubella, mumps, and varicella) has been recommended after 2 years following allogeneic hematopoietic stem cell transplantation (allo-HSCT) in patients not receiving immunosuppressive drugs. There were several reports on eff ectiveness of vaccination for pediatric recipients, but few for adult recipients. Materials (or patients) and Methods: We prospectively analyzed 45 adult allo-HSCT recipients who come to outpatient clinic for posttransplant immunization in Toranomon Hospital from Jun 2012 to October 2013. All patients fulfi lled below criteria for immunization with live-attenuated vaccines; (1) two years or more after allo-HSCT, (2) one year or more after cessation of immunosuppressants, (3) no or under good control of chronic graft-versus-host disease (GvHD). Measles-rubella (MR), mumps, and varicella vaccines were subcutaneously injected once, respectively. Antibody titers for measles, rubella, mumps, and varicella were examined before and later than 2 months after immunization by enzymeimmunoassay methods. Results: The median age was 52 years old (range, 23 -72), 25 patients (55.6%) were male, and 21 (46.7%) underwent cord blood transplantation (CBT). The median counts of CD4-positive and CD19-positive cells were 540/uL (range, 202.8 -1202.8) and 520/uL (122. 2 -2014.8 ), respectively. The median duration from allo-HCT to vaccination was 5 years (range, 2 -12) . The rates of sero-positivities before and after immunization were 11.1% and 37.7% for measles, 31.1% and 88.9% for rubella, 5.6% and 11.1% for mumps, and 42.2% and 57.8% for varicella, respectively. The rates of no seroconversion after vaccination were 57.9% for measles, 6.6% for rubella, 88.9% for mumps, and 42.2% for varicella, respectively. In the group of CBT, there were tendency of higher CD19-positive cells counts and higher incidence of seroconverersion after immunization than the others, although the diff erences were not signifi cantly signifi cant. Discussion: There was not a few patient without seroconversion after immunization, and only once immunization of live-attenuated vaccines was considered not enough for eff ective immunization. Diff erent approach of immunization should be taken according to the type of donor sources. We are currently planning to vaccinate live-attenuated vaccines twice, and examine the relationship of immune recovery and eff ective immunization in each donor source. Disclosure of Interest: None Declared. S329 characteristic memory B cell response is defi ned as a mobilization of plasma blasts (PBs) 7 days after booster vaccinations into the peripheral blood (Odendahl et al., Blood 105: 1614 -1621 , 2005 . Materials (or patients) and Methods: In preparation of this study we analysed control patients (n=5) between day 180 to 240 after alloSCT in the course of routinely performed re-vaccinations with PENTAVAC. At the time of vaccination the patients received no immunosuppressive agents and showed no signs of active chronic GvHD. Before and 7 days after vaccination the number of CD19 + / CD38 high /CD27 high plasma blasts was analysed by fl ow cytometry and by ELISPOT assays after enrichment for total B cells. Data from alloSCT patients were compared with results from a control group of healthy donors (n=10), which also received a booster vaccination with the same vaccine. Results: After vaccination alloSCT patients showed no increase of circulating plasma blasts in the peripheral blood (mean 4.5% PB of B cells before vaccination and mean 4.6% PB of B cells d+7 after vaccination). In contrast, in the peripheral blood of healthy donors a 9.5-fold increase of plasma blasts was observed on 7 days after vaccination (mean 1.6% PB before and 15.5% PB d+7 after vaccination). In the ELISPOT assays, no increase of antigenspecifi c-IgG-secreting PBs was detectable in the alloSCT patients. In contrast, in the peripheral blood of healthy donors a profound increase of antigen-specifi c-IgG-secreting PBs against all vaccineantigens tested could be observed 7 days after booster vaccination, as expected. Introduction: We have recently reported that the preemptive use of the serum galactomannan assay (GM; Platelia, Bio-Rad) provides no diagnostic benefi t to manage high-risk hematology patients on posaconazole prophylaxis, in whom the pre-test incidence of invasive aspergillosis (IA) is very low. In 262 such highrisk episodes, with a 1.9% incidence of IA, all evaluable serum GM tests performed as a surveillance tool in asymptomatic patients were always either negative or false positive. However, GM tests remained useful as part of the diagnostic strategy for symptomatic patients, both for their contribution to the diagnosis of breakthrough IA, as well as for their high negative predictive value [Sánchez-Ortega, et al; P950 EBMT 2013 and AAA 2014] . Here, we now present a prospective validation of our new strategy, which aims to use more effi ciently serum GM tests in high-risk hematology patients on antifungal prophylaxis with posaconazole. Materials (or patients) and Methods: From June 2013, our antifungal management strategy discontinued the preemptive use of serum GM tests in asymptomatic patients, and focused it, in combination with CT scans, microbiological samples and cultures from urine, blood, BAL or other possible infected sites, on a diagnostic strategy for persistently febrile symptomatic patients only.

We present here prospective data until November 2013, on 49 consecutive high-risk episodes on posaconazole primary prophylaxis (31 AML-chemotherapy, 12 alloHCT and 6 GVHD) from 29 hematology patients from our centre. Introduction: High-dose chemotherapy followed by auto-HCT is the standard of care for pts with MM. Auto-HCT can be performed in the outpt setting in appropriately selected pts and may be as safe and more cost eff ective than in the inpt setting. We compared the characteristics and outcomes of pts with MM who underwent auto-HCT in both inpt and outpt settings between January 2008 and December 2012 at our institution. Materials (or patients) and Methods: We identifi ed a total of 1046 pts, with 669 transplanted as inpts and 377 as outpts. Primary endpoints were time to neutrophil and platelet engraftment, 100day treatment-related mortality (TRM), grade 2-4 adverse events (AEs), defi ned by the National Cancer Institute's Common Terminology Criteria, version 4, incidence of unscheduled admissions, and average cost of auto-HCT in both inpt and outpt groups. Secondary endpoints were overall response (OR) rates, progression-free survival (PFS), and overall survival (OS Introduction: Allogeneic transplantation in myeloma remains a controversial area. Following unacceptable non-relapse mortality (NRM) with myeloablative conditioning, the advent of reducedintensity conditioning allogeneic stem cell transplantation (RICallo-SCT) has reintroduced this treatment modality, although its role remains unclear. This study aims to analyse outcomes using this approach and identify which patients may benefi t from it most. Materials (or patients) and Methods: A retrospective analysis of 35 RIC-allo-SCTs performed in patients with myeloma between 1999 and 2011 was carried out. 71% had a sibling donor and the remainder had matched unrelated donors. Fludarabine and cyclophosphamide conditioning was used in 34 cases with HLA matched donors and 1 case where the donor was a 1X mismatch received Alemtuzumab in addition. Peripheral blood was the stem cell source in all cases. Results: 82% patients were male and the median age at the time of transplant was 52 years. 11, 6 and 10 patients had an ISS of 1, 2 and 3 respectively at diagnosis (data was unavailable on the remainder). 2 patients failed to harvest stem cells and had a RIC allo-SCT in fi rst response. 10 patients had a tandem autologousallogeneic approach having achieved a CR following autograft. 11 patients received maintenance or further chemotherapy following their initial autograft to improve response, prior to proceeding with RIC-allo-SCT. 4 patients had their RIC-allo-SCT following relapse and a second autograft and 2 had it following relapse but failure to harvest cells for a second autograft. There was no data on the remainder. Methotrexate and cyclosoporin were used for GVHD prophylaxis, apart from in 2 cases where cyclosporin was not tolerated. There were no cases of engraftment failure. GVHD occurred in 81% of cases (20% acute, 80% chronic Introduction: Infections are still a leading cause of morbidity and mortality after allogeneic hematopoietic stem cell transplantation (aHSCT), particularly during the initial hospital stay. Many measures are taken to prevent those life threatening conditions, including strict isolation, but few guidelines exist. Protected environment can have important consequences, especially on children's quality of life. We compared here the incidence of severe bacteraemia and invasive fungal infections (IFI) in a large pediatric cohort undergoing aHSCT in two units, including major changes in term of human-to-human interaction and room isolation. Materials (or patients) and Methods: We analysed the occurrence of severe infections from the beginning of the conditioning regimen to the end of the initial hospital stay in 286 patients undergoing aHSCT from 2002 to 2012 in our 2 centers. From 2002 to 2007, 144 procedures were performed in center 1 in laminar air fl ow rooms, wearing sterile gowns and gloves, hat and mask, eating sterile food. From 2008 to 2012, 142 procedures were performed in center 2 in positive air pressure rooms with HEPA fi lters, wearing clean gowns and mask, eating protected food. Furthermore, there was the possibility for one parent to sleep in his child's room in center 2. Children received the same oral decontamination and antibiotics/antifungals used for febrile neutropenia remained unchanged during the two periods. We recorded the occurrence of bacteraemia, defi ned by the association of fever with a positive blood culture (except for Coagulase negative staphylococci (CoNS) bacteraemia, where two distinct positive blood cultures were necessary), and proven IFI (according to the EORTC criteria). Results: We reported 85 episodes of severe infections (29.7%): 44 episodes (30.6%) occurred in center 1 while 41 (28.9%) occurred in center 2 (p: 0.53). Infections were mostly bacteraemia with 80 episodes (94.1%) including 6 polymicrobial events: 41 episodes (28.5%) occurred in center 1 and 39 episodes (27.5%) in center 2 (p: 0.71). Gram Positive bacteria (n=71, 81.6%) were more prevalent than Gram Negative bacteria (GN) (n=16, 18,4%) with a large predominance of CoNS (n=57, 65.5%) in both centers. We observed a trend towards more GN bacteria in center 2 (n=12, 26.1%) compared with center 1 (n=4, 9.8%), not signifi cant. Only the use of an unrelated donor was signifi cantly associated with bacteraemia (p: 0.003). We recorded 5 IFI episodes: 3 candidemia and 2 mold infections. Three occurred in center 1 and 2 in center 2 (p: 0.66). Four patients died of infection (1.4%): 3 died from an IFI and 1 from bacteraemia. Discussion: This is the fi rst study comparing the occurrence of infection between 2 centers with diff erent protected environments, but same protocols concerning antibiotics and prophylaxis. We showed similar infection rate in both centers. We noticed small changes: an increase in the proportion of GN as well as a later bacteraemia onset, but they never reached statistical signifi cance. Graft source was independently associated with more bacteraemia, as described in the literature. These results support our decision to make life better for children undergoing HSCT, while keeping them safe in a protected environment. Disclosure of Interest: None Declared. Introduction: Several studies have been published on late effects after haematopoietic stem cell transplantation (HSCT) in childhood, but few have focused on children transplanted at a very young age. The aim of this study is to assess the prevalence of and risk factors for late effects in these young children and compare with older children undergoing HSCT. Also, risk factors for late effects in these young children will be analysed. Materials (or patients) and Methods: This is a multicentre crosssectional study. Patients transplanted for a haematological malignancy under the age of three years and alive at least 5 years after HSCT were identifi ed from the EBMT database. Paediatric HSCT centres in Europe were sent a questionnaire for each eligible patient regarding the presence and severity of selected late eff ects: endocrine function (growth, thyroid and gonadal function), CNS problems and secondary malignancy. After return of the completed questionnaires, data were entered in a database and supplemented with patient data in the EBMT database for analysis. The following risk factors for late eff ects will be analysed: sex, age, conditioning regimen, donor type and stem cell source, acute GVHD, and chronic GVHD. Results: From the EBMT database 650 potentially eligible patients were identifi ed. 295 (45%) questionnaires were returned and valid for data collection. 171 of 295 (58%) patients are male. Median age of the study group at HSCT is 1.8 years (range 0-3 yr). The majority of patients received a myeloablative conditioning regimen (197 of 209 (94%) patients in whom conditioning was registered). In 66 of 295 (22%) patients this included total body irradiation with a dose of at least 6 Gy) and 2 patients received total lymphoid irradiation <6 Gy. Median follow-up of the study group is 10.5 years (range: 4.8 to 31.6 yrs). The frequency of late eff ects is as follows: 26% (n=77) of patients had only one of the selected late eff ects, 22% (n=65) of patients had 2 or 3 late eff ects and 2% (n=6) had 4 or 5 late eff ects. Severe or disabling late eff ects were present in 8 of 295 (3%) patients. The most prevalent late eff ect was thyroid dysfunction in 30% (88/289 with thyroid function known, hypothyroidism in 79). Growth disturbance was present in 27% (n=78) and gonadal dysfunction was present in 27 of 153 (18%) patients aged 12 years or more at time of study. CNS problems were present in 14% (39/288), cognitive problems (n=39, 14%) and concentration impairment (n=32, 12%) being the most prevalent CNS late eff ects. Nine patients developed a secondary malignancy at a median of 11.1 years (range: 0.94 to 22.75yrs) post HSCT. Risk factors for these late eff ects will be analysed and presented. Discussion: The preliminary results of this study indicate that 50% of the patients had at least one endocrine or CNS late eff ect and/or a secondary malignancy at a median follow-up of 10.5 years after HSCT performed before age 3 years. The analysed late eff ects have been described as the most common after HSCT in any paediatric age group and the prevalence of these problems does not appear to be increased in young children. Interestingly, at this young age 23% received a conditioning regimen including irradiation, a known risk factor for late eff ects. Discussion: In our center, viral infection prevalence is higher than previously reported, although mortality is lower. A probable explanation could be our pre-emptive screening test policy. CMV and HHV6 were the most common etiologies. Viral infection prevalence was higher in allogeneic HSCT, especially when cord blood was the stem cell´s source. ALL and AHSCT were the only factors associated with a higher risk of viral infections in our patients. Disclosure of Interest: None Declared. Introduction: Allogeneic hematopoietic stem cell transplantation (allo-SCT) may provide donor cytotoxic T cell/NK cell-mediated disease control in patients with Ewing Sarcoma (ES). However, little is known about the prevalence of graft-versus-ES eff ects and only a few case experiences have been reported. We evaluate in a prospective national wide protocol the feasibility and effi cacy of a reduced intensity conditioning regimen (RIC) followed by allo-SCT from related (RD) or unrelated HLA donor (UD) in advanced Ewing Sarcoma. Materials (or patients) and Methods: From 2009 and 2012, 14 pts, aged 5-22 years, aff ected by resistant or relapsed ES, were enrolled and submitted to an allo-SCT after a RIC consisting of Thiotepa 15 mg/kg and Melphalan 140 mg/sqm. The donor was a RD in 9 cases and an UD in 5. At time of transplant 5 pts were in CR, 1 in VGPR, 5 in PR and 3 with active disease. Seven pts received allo-SCT as fi rst line graft. Graft versus host disease (GVHD) prophylaxis consisting of Cyclosporin A in related and Cyclosporin A + Antilymphocytic serum and short term methotrexate in UD setting. SC sources were bone marrow.

Results: The reconstitution of bone marrow function was obtained in all the pts. Acute GVHD of grade II occurred in 5 pts and a complete marrow donor chimerism was observed after a median time of 40 and 60 days in sibling and UD setting respectively. After a median follow-up of 14 (5-42) months, 7 pts relapsed, 5 dead 9 are alive and well. The 2 years probability of OS (SE) and EFS (SE) were respectively of 0.51 (15.4) and 0.43 (12.2) respectively. The 100 days probability of TRM was 0.0. Discussion: The use of allo-SCT in patients with advanced ES is currently experimental but in a subset of patients it may constitute a valuable approach for consolidating CR. Related Donor, CR and fi rst-line allo-SCT assure the best outcome and may be purposed in prospective trials. Disclosure of Interest: None Declared. Introduction: After allogeneic stem cell transplantation (ASCT), children are isolated in hospital to prevent neutropenic infections. Fifteen years ago, we challenged this routine by allowing patients to be treated at home during the neutropenic phase. Initially, mainly adult patients were included in home-care. A matched-pair analysis showed that patients treated at home had several advantages compared to isolation in the hospital, such as early discharge to the outpatient clinic, fewer days of total parenteral nutrition, fewer episodes of acute graft-versus-host disease (GVHD), lower transplantrelated mortality (TRM), and improved survival (Svahn et al, Blood 2002) . Subsequently home-care in this setting was used on a routine basis. Children have a lower probability of and a lower severity of acute GVHD than adults. Thus, we do not know whether homecare would reduce the risk of acute GVHD in children in the same way as found in adults. The main reason for the present study was to determine whether it is safe to treat children at home. Materials (or patients) and Methods: Patients living within two hours' drive from the hospital were given the option of treatment at home after ASCT. Daily visits by an experienced nurse and phone calls from a physician from the unit were included in the protocol. We compared 29 children and adolescents treated at home with 58 matched hospital controls. Results: The children spent a median time of 13 days at home (range 2-24 days). Before discharge to the outpatient clinic after ASCT, they spent a median of 6 (0-5) days in hospital. Time to absolute neutrophil count >0.5×10 9 /L and to reach platelets >30×10 9 /L was not signifi cantly diff erent in the home-care and hospital-care children. Need for platelet transfusion and erythrocyte transfusion was similar in the two groups. The cumulative incidence of acute graft-versus-host disease (GVHD) grades II-IV was 21% in the home-care children and 39% in the controls (P=0.1). Chronic GVHD and probability of relapse were similar in the two groups. Transplant-related mortality at fi ve years was 11% in the home-care patients and 18% in the controls. Overall survival at three years was 77% and 62%, respectively (P=0.33). None of the patients died at home and no adverse events occurred. Median costs were 38,748 Euro in the home-care patients and 49,282 Euro in those treated in the hospital (P=0.2). Discussion: Initially, our aim was to improve the quality of life of our patients by treating them at home instead of isolating them in the hospital. Children in particular appreciate much more being at home than being locked up in an isolation room for most of the day. The result must be treated with caution because of the limited numbers. However, there tended to be less acute GVHD, lower TRM and a better survival in children treated at home, to the same extent as we have seen in many more adult patients (Svahn et al, Blood 2002) . Inclusion of many more children in the homecare program will be necessary to establish whether or not these observations have true statistical signifi cance. Our overall experience of home-care in adults involves more than 200 patients, and to date, none of them have died at home. We conclude that it is safe for children and adolescents to be treated at home during the pancytopenic phase after ASCT. Introduction: Treosulfan is an alkylating agent applied in regimens prior to allo-HSCT in children. It has strong myeloablative and immunosuppressive activity and a mild toxicity profi le in comparison to busulfan. In contrast to busulfan, pharmacokinetic (PK) data to optimize treosulfan dosing are still scarce in pediatric patients. We developed a population pharmacokinetic model and limited sampling strategy (LSS). Furthermore, we describe the PK profi le of 22 pediatric patients and an interim analysis on the correlation with clinical outcome. Materials (or patients) and Methods: Patients received treosulfan based conditioning prior to their HSCT for various malignant and non-malignant indications. Intravenous treosulfan was combined with fl udarabine and thiotepa; the treosulfan dose was 42 g/m 2 (or 30 g/m 2 if BSA<0.5 m 2 (n=1)), divided over 3 days. Multiple (n=6) sampling scheme was used in the fi rst 20 patients to develop a PK model and LSS which was applied in 2 patients. Samples were collected after the fi rst dose of treosulfan and measured with a reversed phase high pressure liquid chromatography method using UV detection. The PK model en LSS were developed with non-linear mixed-eff ects modelling. This model and LSS were applied to determine the PK profi le of each individual patient. Results: A total of 22 patients were included into the study with a median age of 4.4 years (0.1-16.8 years) and a median follow-up of 1.1 year (0.3-2.4 years). The cohort included patients with hemoglobinopathy (n=14), hematologic malignancy (n=6) or immune defi ciency (n=6). A one-compartment model was used and clearance and volume of distribution were allometrically scaled using body weight. The scaling exponent for clearance was fi xed at 0.75 and for volume of distribution at 1.0. The population estimates for clearance and volume of distribution were 6.9 L/h and 13.2 L for a patient of 20kg, respectively. Treosulfan AUC could be adequately determined based upon two serum samples, at 4 and 7 hours. The average AUC was 1625±251 mg*hr/L and interpatient variability was 15%. The toxicity profi le was in general mild; elevated liver enzymes, mucositis and skin toxicity were the most common toxicities. Neutrophil engraftment was 83% in patients receiving a fi rst SCT and 75% patients receiving a second SCT. In this limited cohort we did not fi nd a relation between treosulfan exposure and SCT outcome parameters e.g. engraftment, chimerism, toxicity and survival.

Discussion: In this study a bio-analytical method, PK model and LSS for treosulfan were developed and validated. Furthermore, PK parameters of 22 pediatric patients were analysed. This interim analysis did not demonstrate a relation between treosulfan exposure and SCT outcome. However, there is need for further analysis in a larger group of patients and in disease specifi c subgroups. There are limitations to the Optia system for smaller patients as a minimum inlet fl ow rate (IFR) of 10ml/min is required compared to 2-3ml/min for Spectra. The following strategy was devised to overcome this limitation by using a citrate/heparin combination and increasing the inlet:anticoagulant (AC) ratio. Materials (or patients) and Methods: For patients with a total blood volume (TBV) of < 700ml, heparin is continuously administered during the procedure. This is achieved by adding heparin (1:1000 iu/ml preservative free) directly to the anti-coagulant (ACD-A). 5000 iu heparin is added to 500ml ACD-A giving a concentration of 10iu heparin / ml ACD-A. The procedure may then be run at Inlet:AC of 30:1. For every 30ml whole blood processed patient will receive 1 ml of heparin/citrate combination. Our default AC infusion rate is set at 0.9mls/min but this may safely be increased to 1.2mls/min if patient has high platelet count or evidence of clumping in the circuit. For patients with a TBV of 700 to 1000ml, the required IFR can be achieved by altering the inlet: AC ratio only.Discussion: The above strategy has been devised to overcome the limitations imposed by the minimum fl ow rate required by Introduction: Because of a signifi cant improvement in the survival of children and adolescents with cancer, fertility preservation has to be a major concern for paediatric oncologists. The aim of our study was to report all our ovarian tissue cryopreservation's (OTC) cases before stem cell transplantation (SCT) in order to specify the interest and indications of this method and to study the clinical and hormonal outcome in girls. (1), relapsed osteosarcoma (1), retroperitoneal primitive neuroectodermal tumor (PNET) with tumor spillage (1), and supratentorial PNET (1). Two patients (1 supratentorial PNET in second partial response, 1 refractory intracranial GCT) underwent double (tandem) ASCT. Regarding ASCT, the median age at transplantation was 5 years (3.5-8.5). One patient dead of septic shock after a second conditioning regimen (tandem transplant) for supratentorial PNET. One patient aff ected of intracranial GCT, and heavily pretreated with platinum compounds developed tubulopathy and severe ototoxicity after transplantation. No other major complications were registered. With a median followup post ASCT of 36 months (14.5-53.7) the median progression free survival (PFS) and median overall survival (OS) were 11 and 33 months respectively. The overall estimated 3-year PFS and OS were 43% and 47%, respectively. The median PFS was signifi cantly shorter in patients who did not achieve a CR before ASCT ( In these, early chimerism showed initial donor myeloid >50% in all 4 patients. Absent donor myeloid chimerism was present in 3 patients; all had poor T cell immunoreconstitution and are immunoglobulin dependent. Seven patients (38%) died posttransplant, 5 during the fi rst 6 months. 6/7 had engrafted (1 suffered graft loss); donor myeloid chimerism was 100% in 3 of these, partial in 1 and 0 in 2 patients. Mortality was higher among OS than SCID, TCD than TCR, and Bu-based than Treo-based patients. Main cause of death among OS was disseminated viral infection (4 patients), of whom 3 had grade 2/3 aGVHD and were immunosuppressed. SCID deaths were engraftment pneumonitis (n=1) and aGVHD grade 2 with EBV-driven cerebral lymphoproliferative disease (n=1). Discussion: SCID had better IR and survival than OS. Main cause of death among OS was infections. TCR transplant carry a better outcome than TCD. 100% donor myeloid chimerism seemed associated with poor outcome. The conditioning data suggested that some conditioning is required to get early myeloid donor chimerism, which enables B cell function and thymopoiesis with naive T cells. Disclosure of Interest: None Declared. Introduction: To prevent graft versus host disease (GvHD) and rejection in unrelated hematopoietic cell transplantation (HCT), children receive anti-thymocyte globulin (ATG) in the conditioning regimen. The therapeutic window is critical as over-exposure results in delayed immune reconstitution (IR) and under-exposure GvHD: both causing signifi cant morbidity and mortality. As the optimal exposure is unknown we studied the association between active ATG (Thymoglobulin) exposure and outcomes as a step towards an evidence based dosing regimen of Thymoglobulin. Materials (or patients) and Methods: All pediatric patients transplanted between 2004 -2012 receiving Thymoglobulin in the two study centers were included. Only first HCTs were included, patients mounting IgG anti-ATG antibodies were excluded. Different exposure related variables such as AUC (area under the curve) were calculated using a validated PK model and were tested for their association with the endpoints: event-free survival (EFS), overall survival (OS), acute GvHD and T-cell reconstitution. T-cell reconstitution was defined as a CD3+ T-cell count > 100 x10 6 /L in two consecutive occasions within 100 days after HCT. aGvHD was classified using the Glucksberg criteria. Cox proportional hazard and regression models were used. Results: Pharmacokinetic and -dynamic data were available from 196 pediatric HCT's; 28 were excluded due to 2 nd /3 rd HCT (13) or antibodies (15). OS was 67% with a median follow-up of 130 weeks . Post-HCT exposures correlate best with the endpoints. Survival was higher in patients with a low (<20 AU*days/L) post-HCT AUC (80.4% vs 60.7%, P=0.026). Multivariate predictors for lower OS were high (>20 AU*days/L) post-HCT AUC (P=0.024), mismatch donor (P=0.01) and malignancy (P=0.007). In addition, low post-HCT AUC of Thymoglobulin was associated successful T-cell reconstitution (P<0.001) while incidence of aGvHD was comparable irrespective of post-HCT AUC (P=0.48) (fig 1) . Discussion: High (>20 AU*days/L) post-HCT exposure of Thymoglobulin is associated with a lower OS and a lower probability of successful IR. These results may contribute to individualized dosing guidelines of Thymoglobulin in children, aiming to improved survival. S370 chemo-conditioning in children with ALL, the EBMT PD WP initiated a retrospective study on this topic. Materials (or patients) and Methods: We identifi ed 1728 children and adolescents with ALL who received a fi rst HSCT between 2000 and 2012 and who were registered in the EBMT data base. Results: The majority of patients were transplanted in fi rst or subsequent remission, 11% were not in complete remission at time of HSCT. The most common donor type were unrelated donors (52,9%), followed by matched siblings (32.8%), other relatives (13,9%) and some others like twins. The stem cell source was bone marrow in 43.6% and peripheral blood stem cells in 37.8%. 18.1% received cord blood and the remaining diff erent combinations of grafts. For 90.8% of patients the centres intended a myeloablative conditioning, and for 9.8% a reduced toxicity or a reduced intensity conditioning regimen was chosen. The preferred non-TBI conditioning regimen was a combination of busulfan and cyclophosphamide (48%), followed by a triple-drug regimen consisting of busulfan, cyclophosphamide and etoposide; the remaining patients received diff erent combinations like fl udarabine/ thiotepa/melphalan or treosulphan. Acute GVHD occurred in 52% (overall grade III/IV: 14,2%), chronic GVHD was reported for 15.2% (7.1% limited, 7.1% extensive). At time of analysis, 51.4% of patients were alive and 45,9% had died. Causes of death were relapse (49.8%) or transplant related complications (42.7%). There was a signifi cant improvement over time, as patients transplanted after 2008 had an overall survival of 60,9% with comparable relapse incidence but lower incidence of non-relapse mortality. Compared to the whole cohort, children who were transplanted below the age of 4 years had a lower relapse incidence (28.6%) and an overall survival of 56.3%. 336 children were under 2 years of age at time of HSCT and the overall survival was 56.3 % despite the high relapse risk of predominantly infant leukaemia phenotypes. Discussion: The preliminary results of this study indicate that more than 50% of children with ALL who received no TBI as conditioning regimen for allogeneic HSCT from diff erent donors in diff erent remission status are alive. This observation justifi es and requests a prospective evaluation whether TBI is still superior compared to conditioning regimen with chemotherapy only in comparable cohorts of patients. As most paediatric ALL-patients are cured with contemporary chemotherapy protocols, only a small proportion requires an allogeneic HSCT. International cooperation is essential to answer the question of the most appropriate conditioning regimen for these patients. Disclosure of Interest: None Declared. Introduction: Busulfan (Bu) as myeloablative agent is used in conditioning regimens prior to hematopoietic stem cell transplantation (HCT). In the recent past we reported on the relation between Bu exposure and outcome and defi ned an optimal exposure to minimize rejection, toxicity and relapse. From there we replaced classic cyclophsphamide by fl udarabine (fl u) to further reduce toxicity. This has also recently been published. In vitro a synergistic antileukemic eff ect of clofarabin (clo) to the Bu/Flu combination has been shown. We thus reasoned that the Clo Flu Bu combination might be optimal as an antileukemic conditioning regimen. Furthermore, recent interest to replace total body irradiation (TBI) by a chemotherapeutic approach to reduce late eff ects (endocrine disorders, second malignancies) accelareted our interest in chemo-based conditioning regimens. Within the framework of the Dutch Childhood Oncology Group (DCOG) it was decided to explore the Clo Flu Bu regimen for safety and effi cacy for lymphoid and myeloid malignancies. Materials (or patients) and Methods: Our conditioning regimen consisted of clofarabine 30 mg/m 2 over 1 hour, followed by fl udarabin 10 mg/m 2 in 1 hour, followed by a 3-hour infusion of once-daily busulfan (weight-based dosing) for 4 consecutive days. The cumulative target area under the curve (AUC) for Bu was 90 mg*h/l. Serotherapy in unrelated donors (expect AML patients receiving cord blood) and greaft-versus-host-disease (GvHD) prophylaxis were according to standard protocols. Primary endpoint: Leukemia-free-survival (LFS). Other endpoints: acute and chronic GvHD, veno-occlusive disease (VOD), non-infectious lunginjury, neutrophile (@day 60) and thrombocyte engraftment (@day 180). A risk factor analysis was performed using univariable and multivariable COX regression. Results: 38 patients were included (18 AML CR2, 11 ALL CR1 (of which 2 infants), 6 ALL CR2/3, 3 MDS (RAEB-t). 6/12 ALL patients with available MRD status prior to HCT were positive: 2>10 -3 and 4 <10 -3 . Donors used: 14 unrelated cord blood, 9 (matched) family donors (1 haplo id), and 14 matched unrelated marrow donors. . 20 pts were transplanted with and 14 pts without conditioning. OS was 85% and 79% respectively. Two pts died due to infectious complications (BCG, EBV), in four pts with respiratory failure no infectious agent was identifi ed; no pt died due to GvHD. T-cell engraftment was found in all pts except one who failed to engraft paternal cells without conditioning but was successfully retransplanted with a maternal graft. Successful retransplantations with conditioning were performed in two pts 6 and 12 years after the fi rst attempt without conditioning because of recurrent infections (in spite of Ig-substitution) and autoimmune hepatitis respectively. Stable B-cell function without the need for Ig-substitution was noted in 14/17 long term survivors (82%) after conditioning and in merely 3/11 pts (27%) treated without conditioning. Discussion: For pts with X-SCID without an available HLA-compatible donor, haploidentical transplantation with and without conditioning is a therapeutic option with excellent survival. Introduction: Historical data suggest similar overall survival for pediatric patients undergoing allogeneic hematopoietic cell transplant (HCT) for acute myeloid leukemia (AML) using busulfan (BU)-based myeloablative (MA) regimens as compared to total body irradiation (TBI)-based MA regimens. There has been a shift to increased use of BU MA regimens as compared to TBI MA regimens in pediatric AML patients. Therefore, understanding the short-term toxicities and early non-relapse mortality from BU MA regimens in these patients is crucial.

Materials (or patients) and Methods: We used data from an ongoing, prospective multi-institutional study (PBMTC ONC1001/CIB-MTR 09-MRD) to assess the incidence of toxicity and day 100 mortality by MA regimens. A total of 43 subjects enrolled in 09- MRD (2011 MRD ( -2013 had BU MA conditioning. All patients were in morphologic remission, had not had prior HCT, and had organ function suitable for MA transplant. Median age was 7 years , with 49% males and 51% females. Eighty-eight percent had Karnofsky/Lansky ≥90%, with 74% of patients having a comorbidity index of 0 or 1. Disease status was CR1 in 72% and CR2 in 28% and 91% were in cytogenetic remission. Donor types were: HLA-identical siblings in 19%, other relatives 5%, 7-8/8 unrelated donor 42%, 4-6/6 single cord blood in 28%, and 5-6/6 double cord bloods in 7%. Only 5% of patients overall received PBSC. Results: The following were the cumulative incidences (CI) of toxicities at day 100, based on CIBMTR data form 2100: infection 75% (95% CI 61-87%), idiopathic pneumonitis 5% (0-13%), liver toxicity 21% (10-34%), other organ toxicity 24% (12-37%) . In terms of graft-versus-host disease (GVHD), incidence of grade 2-4 was 21% (10-35%) and grade 3-4 0%. The CI of relapse was 19% (9-31%) and transplant-related mortality (TRM) 0%, for a disease-free-survival (DFS) of 81% (69-91%) at day 100 and an overall survival (OS) of 95% (87-100%). Discussion: While this is not a prospective treatment study, these data suggest that signifi cant day 100 toxicities in BU-based MA HCT are acceptable and the incidence of GVHD is low, despite 81% of patients receiving an alternative-donor graft. Moreover, a day 100 TRM of 0 suggests BU-based MA regimens are very well-tolerated in children with AML. These excellent results may be due to pharmacokinetic-based dosing which are standard now, although these data were not collected. Disclosure of Interest: None Declared.

A. 

Median hepcidin and ferritin levels did not diff er between patients treated for malignant (36,3 ng/ml and 1024 ug/l respectively) and non-malignant diseases (25,3 ng/ml and 117 ug/lrespectively) . Median values of sTfR, as well as median iron levels in children before HSCT (1,3 ug/ml and 13,2 umol/l respectively) did not diff er compared to controls (1,0 ug/ml and 17,9 umol/l respectively) .In 19/39 (43,6%) patients organ complications > 2 grade were diagnosed, in 20/39 children no organ complications or complications of 1 and 2 grade occurred. No diff erence was found in median ferritin, hepcidin sTfR and and iron levels between these groups of patients.In 17/39 patients (43,6%) high pre-transplant ferritin concentration (>1000 ug/l) was diagnosed (median 1902 ug/l). In this group of patients median hepcidin level was higher (62,8 ng/ml) than in low ferritin (< 1000 ug/l) group (median ferritin 117 ug/ ml; median hepcidin 21 ng/ml) [P<0,05] . No diff erence was found in the incidence of transplant related organ toxicities and infections in low and high ferritin/hepcidin group. Discussion: According to our analysis children undergoing stem cell transplantation are at risk of iron overload, which can be defi ned as well as high hepcidin-25 or ferritin serum concentration.We have not found a signifi cant association between the pretransplant serum hepcidin/ferritin levels and the cumulative incidence of organ toxicities and infections. Further studies are necessary to determine the correlation between iron overload and the incidence of complications in HSCT patients. Disclosure of Interest: None Declared. Introduction: Pediatric patients with hemato-oncological malignancies and neutropenia resulting from chemotherapy have a high risk of acquiring invasive fungal infections. Oral antifungal prophylaxis with azoles, e. g. fl uconazole or itraconazole, is preferentially used in pediatric patients after chemotherapy, even though only a few studies have been published. We also administered posaconazole based on data from studies in adult patients. Retrospectively, we compared the safety, feasibility and initial data on effi cacy of posaconazole vs. fl uconazole and itraconazole in pediatric patients and adolescents during neutropenia after chemotherapy. Materials (or patients) and Methods: In this single center study, 93 pediatric patients with a median age of 12 years (range 2 -17years) with neutropenia (absolute neutrophil count of 500 cells) from a minimum of 5 days after intravenously chemotherapy, who received fl uconazole, itraconazole or posaconazole as antifungal prophylaxis were analyzed. 32 of the 93 pediatric patients received antifungal prophylaxis with fl uconazole, while 31 were given antifungal prophylaxis with itraconazole. 30 other pediatric patients received antifungal prophylaxis with posaconazole. The observation period was defi ned as the day before start with oral antifungal prophylaxis with one of the three azoles, to the end of intensive chemotherapy and recovery from neutropenia or until occurrence of a proven or probable invasive fungal infection.

Results: The duration of the periods of prolonged neutropenia in the group of children treated with fl uconazole had a mean of 8.9 ± 3.2 days. In the group of pediatric patients treated with itraconazole the mean was 12.6 ± 6.1 days, while the mean for the group of children treated with posaconazole was 17.4 ± 11.4. Proven invasive fungal infections were reported in 2 (6.3%) pediatric patients in the fl uconazole or itraconazole group and in 1 (3.3%) pediatric patient in the posaconazole group. Possible fungal infections occurred in 3 (9,5%) pediatric patients in the fl uconazole or itraconazole group and in none (0%) of the pediatric patients in the posconazole group. The percentage of patients with adverse events potentially related to clinical drugs were 15.6% in the itraconazole group, 12.9% in the fl uconazole group and 10% in the posaconazole group. Posaconazole vs fl uconazole or itraconazole are comparably eff ective in preventing invasive fungal infections in pediatric patients with posaconazole being slightly more eff ective in pediatric patients with neutropenia after chemotherapy. Discussion: Conclusion: Posaconazole vs fl uconazole or itraconazole are comparably eff ective in preventing invasive fungal infections in pediatric patients with posaconazole being slightly more eff ective in pediatric patients with neutropenia after chemotherapy. Disclosure of Interest: None Declared. Introduction: Intravenous busulfan (iv-Bu) is considered to have a much more predictable pharmacokinetic (PK) profi le than oral-Bu. However there remains strong variation. For cases in which real-time therapeutic drug monitoring (TDM) is not applicable, new approaches using PK analyses of test doses have been proposed.

In the present study, we tested whether the administration of a test dose of iv-Bu before HSCT can be used to predict the PK of the fi rst dose of iv-Bu in children and investigated the association between the clinical outcomes and PK results. Materials (or patients) and Methods: iv-Bu was administered as a two-hour infusion every six hours for four days. Five dose levels (0.8-1.2 mg/kg) were evaluated according to body-weight strata. Following the administration of a single dose of iv-Bu as a test dose and the fi rst dose of Bu during the conditioning regimen, blood samples were collected from each patient at seven points. Dose adjustment was not allowed between the administration of the test and treatment doses. The Bu concentrations in plasma were measured using HPLC. The area under the curve (AUC) for both the test dose and fi rst dose was calculated using the linear trapezoidal rule. Results: Twenty-three children from 8 months to 17 years of age were enrolled in this study. The underlying diseases included AML (n = 10), advanced solid tumors (n = 6), and others (n = 7). The conditioning regimens included Bu/LPAM (n = 13), Bu/CY (n = 6) and Bu/FLU/ATG (n = 4 

The AUC for the fi rst dose was compared with that for the test dose. A linear regression analysis showed a statistical correlation between these values (r = 0.65; P < 0.05). Discussion: The body-weight strata dosing schedule used in the present study was associated with a low rate of achievement of the target AUC. The risk of graft failure among patients undergoing allogeneic HSCT should be minimized, keeping the AUC above 900 μM × min. When the maximum antitumor eff ect is required, then maintaining a high Bu exposure within the target AUC may be necessary. Three patients developed lethal Bu lung disease, not correlated with the AUC. Although a statistical correlation was observed between the fi rst dose and test dose AUC values, it is very diffi cult to predict the fi rst dose AUC precisely based on the test dose AUC. Developing a test dose strategy for assessing the effi cacy of iv-BU treatment would be a useful means of excluding extremely poor or extensive metabolizers. Disclosure of Interest: None Declared. Introduction: Treatment outcome of pediatric cancer has improved substantially during the past two decades, but limited progress has been achieved in improving the survival rate of high risk neuroblastoma. High-dose chemotherapy (HDCT) has been employed to improve survival, but many patients still suff er from relapse after HDCT. The intent of this research was to investigate the clinical eff ectiveness of tandem HDCT combined with autologous stem cell rescue therapy and establish a regimen that maintains a low TRM and relapse rate. Materials (or patients) and Methods: We retrospectively analyzed 60 children with high risk neuroblastoma who received single or tandem HDCT during the period between October 1997 and December 2010. The event-free survival (EFS), overall survival (OS), and toxicities were assessed. Results: Sixty patients diagnosed with high-risk neuroblastoma received single or tandem HDCT and autologous stem cell rescue. Thirty two patients received single HDCT and 28 patients received tandem HDCT. Patients ≥ 18 months consisted 78.1% of the single group and 82.1% of the tandem group. Patients in the single HDCT group received a conditioning regimen that consisted of melphalan, etoposide, and carboplatin (MEC) or topotecan, melphalan, carboplatin (CTM). The tandem HDCT group received topotecan, thiotepa, and carboplatin (TTC) with MEC. In June of 2007, a modifi cation of carboplatin dose was implemented as patients were observed to experience renal toxicity. The probability of 5-year event-free survival (EFS) and overall survival (OS) was higher in patients who received tandem HDCT compared to the patients who received single HDCT. The EFS of the single and tandem group was 48.8% and 63.9%, respectively (P=0.23). The OS of the single and tandem group was 77.6% and 85.7%, respectively (P=0.54). Treatment related mortality (TRM) was observed in 2 of the 32 patients that received single HDCT, and 2 of 28 patients that received tandem HDCT. A decrease in TRM attributed to renal toxicity was observed in patients who received lower doses of carboplatin. Discussion: Tandem HDCT could improve survival in high-risk neuroblastoma patients. Conditioning regimen related to renal toxicity may play a factor in increasing TRM, and further investigation of an optimal conditioning regimen is warranted. Disclosure of Interest: None Declared. Introduction: Cytokine-induced killer (CIK) cells are expanded in vitro with IFN-γ, αCD3 monoclonal antibodies (mAb) and IL-2. In 20 10 Bonanno presented a protocol with polyclonal thymoglobuline (TG) substituting mAb showing a successful diff erentiation of CIK cells. Due to high relapse rate adoptive CIK cell immunotherapy aims at controlling minimal-residual disease by the immune system, leading to eff ective tumour cell killing in posttransplant setting. Materials (or patients) and Methods: Peripheral blood mononuclear cells from 5 healthy potentially haploidentical donors were activated with IFN-γ on day 0 and fed with 50 ng/mL of mAb (standard protocol) or 500 ng/mL of TG (modified protocol) with IL-2 on day 1. Culture medium exchange with further IL-2 stimulation was performed every 2-3 days for 21 days. Aliquots of CIK cells were harvested weekly for cellularity, viability, cytotoxicity (against NK-sensitive K562 cells) and immunophenotype (standard lymphocyte subpopulations, regulatory T cells, TCRαβ/γδ). In one case cytotoxicity against fresh isolated MDS/sAML patient blasts was tested. In this case after local ethical board acceptance and obtaining a signed informed consent the treatment with CIK cells stimulated with TG was applied in a paediatric patient in progression of sAML (status post 4 alloHSCT, incl.1 haplo from KIR-mismatched father and 3 DLIs with no features of GvHD). 1 st TG-derived CIK cells were infused 4 weeks after the last DLI in the previously applied cell dose (1x10 6 CD3+cells/kg) with hydrocortisone, clemastine and metamizol prophylaxis without any cytoreductive chemotherapy. 2 nd infusion was performed 4 weeks after with increased dose of 2.7x10 6 CD3+cells/kg with clemastine and metamizol prophylaxis. 3 rd infusion was done 3 weeks after with increased cell dose of 5x10 6 CD3+cells/kg with clemastine. Two last doses were applied after chemocytoreduction of blasts. The patient was monitored for early and late immunological reactions. Results: The expansion of CIK cells stimulated with TG was more vigorous than mAb-driven CIK cells. This culture showed more effi cient K562 and fresh isolated blasts target cells lysis when compared with standard protocol. The frequency of Treg cells as well as TCRαβ cells was lower for the studied protocol in comparison with standard setting. Infusion of TG-derived CIK cells did not cause any side eff ects in the recipient. No features of GvHD was observed. After the 1 st infusion circulating peripheral blasts were reduced from 66% to 19% within 24 hours as well as transient autologous signal drop was noted. The patient died of progression 2,5 months after the 3 rd infusion and 4 months from the 1 st one. Discussion: TG allows suffi cient clinical grade functional CIK cell expansion without concomitant generation of Treg cells (a subset implicated in the down-regulation of anti-tumor immunity) and with reduction of TCRαβ cells (a population correlated with GvHD). Infusion of TG-derived CIK cells obtained from previous haematopoietic stem cell donor seems to be safe in allogeneic stem cell recipients who failed DLI and progressed without any features of GvHD. Further studies on safety and effi cacy of TGderived CIK cell therapy are required. Supported by NCN grant NN407686940. Disclosure of Interest: None Declared. Introduction: Sickle cell disease has a high risk of morbidity and mortality and bone marrow transplantation currently off ers the only chance of cure. However, less than 25% of children have a matched familial donor free of the condition and the chances of fi nding an unrelated donor are low. Materials (or patients) and Methods: Between July and September 2013 we performed three haplointetical BMT for patients with homozygous HbSS disease and severe cerebrovascular disease at St Mary's hospital. All children had suff ered at least one cerebrovascular event and had evidence of disease progression despite adequate transfusion therapy. The source of stem cells was G-CSF primed maternal bone marrow. Two patients had the same blood group as the donor and one had a major ABO incompatibility. Two patients were CMV+/+ and one -/+. Results: Endogenous haemopoiesis was suppressed for three months pre-transplantation with the use of hypertransfusions, hydroxycarbamide 30 mg/kg and azathioprine 3 mg/kg. The conditioning regimen comprised ATG (Thymoglobulin) 0.5 mg/kg day -9 and 2 mg/kg days -8 to -7 (total 4.5 mg/kg), thiotepa 10 mg/kg day -7, fludarabine 30 mg/m 2 days -6 to -2 (total 150 mg/ m 2 ), cyclophosphamide 14.5 mg/kg days -5 and -4 (total 29 mg/ kg) and TBI 2 Gy on day -1. GvHD prophylaxis was provided with cyclophosphamide 5 mg/kg day +3 and +4, MMF and sirolimus (target 10-15 ng/mL). MMF was weaned after day +35 over two weeks after molecular evidence of donor haemopoiesis. The cell dose was 5.13 x 10 8 TNC/kg and 1.123 x 10 6 CD34+/kg, 11.82 x 10 8 TNC/kg and 5.158 x 10 6 CD34+/kg, and 10.117 x 10 8 TNC/kg and 2.617 x 10 6 CD34+/kg, respectively. Neutrophil engraftment occurred on day +16, +17 and +21, respectively. All patients showed chimerism studies >95% donor in whole blood and T cell fraction on days +28, +60 and +90 and patient 1 is off all immunosuppression since day +119. Patient 1 developed a possible IFI of the lungs and CMV pneumonitis, reactivated HSV type 1 and type 2 and had haemorrhagic cystitis. Patient 2 had a possible IFI of mastoid. Patient 3 has a possible IFI of the lung. All infectious complications had complete resolution following fi rst line treatment and none of the patients had organ failure. Patient 1 developed acute GvHD of the skin stage 1 with complete response to topical steroids and patient 3 acute GvHD skin stage 3 with complete response to the use of MSC. None of the patients developed VOD. Discussion: In summary, these constitute the first three haploidentical transplants reported for sickle cell disease with the use of a post-infusion cyclophosphamide and reduced intensity conditioning in the paediatric population. This approach has led to prompt engraftment enabling the cure of patients with no related or unrelated donors. Although infectious complications are common all patients responded promptly and had full resolution. Disclosure of Interest: None Declared. Introduction: Patients with DNA-dsb repair defects often present with recurrent infection, bone marrow failure and predisposition to lymphoid malignancy. The systemic nature of the repair defect may cause patients not to tolerate myeloablative conditioning. We report multi-centre outcomes of HSCT in patients with Nijmegen Breakage Syndrome (NBS), DNA ligase 4 defi ciency (LIG4), Cernunnos defi ciency (XLF), and Ataxia Telangiectasia (AT). Materials (or patients) and Methods: Centres were contacted through EBMT IEWP and CIBMTR. Retrospective data were collected via a questionnaire, including conditioning regimen, outcome post-HSCT, incidence of GvHD. Conditioning regimens were classifi ed as myeloablative (MAC) (busulfan or treosulphan) or reduced intensity (RIC) (typically a combination of fl udarabine and melphalan or very low dose cyclophosphamide, including combinations of targeted antibodies and alemtuzumab). Results: 23 patients identifi ed, from 17 transplant centres in 11 countries, including NBS (8), LIG4 (7), XLF (4), AT (4) . Median age at HSCT was 52 months (range 22-240), median follow-up 13.5 month (range 2-102). 8 received MFD, 11 MUD, 4 mis-matched donors, 3 UCBT, 12 BM, 8 PBSC. 16/23 (70%) received RIC, of whom 14 (88%) are alive, 7 underwent MAC, of whom 2 (29%) are alive. Acute GvHD (grade 1-3) was common occurring in 14/23 (61%) cases, 12 ≥ grade 2 skin+/-gut/liver, 5 died. Acute GvHD occurred in 4/4 XLF, 4/8 NBS, 2/7 LIG4 and 3/4 AT patients. 5 developed chronic GvHD, 2 died. Discussion: HSCT can be successful for patients with DNA-dsb repair defects. A good short-term outcome can be achieved, particularly if RIC is used. Irradiation is not necessary as part of conditioning, and should be avoided. GvHD is common, and not well tolerated. Long-term follow up will be required to determine whether these patients are at greater risk of developing secondary malignancies. Disclosure of Interest: None Declared. Introduction: Children undergoing allogeneic stem cell transplantation (Allo-SCT) are at high risk of acquiring invasive fungal disease (IFD). However, in the last years the use of antifungal prophylaxis with anti-mold agents (especially azoles) has significantly reduced the incidence of IFD in this kind of pediatric patients. In this study we have prospectively assessed the safety and efficacy of voriconazole as primary antifungal prophylaxis during neutropenic phase in children undergoing Allo-SCT.

Materials (or patients) and Methods: From Oct-2004 to Nov-2013 a total of 77 Allo-SCT performed in 72 children and adolescents (<18 years) were included in this prospective study; median age was 9 years (range: 1-17). Acute lymphoblastic leukemia was the most frequent underlying disease (n=46). In all except 6 patients myeloablative conditioning regimen was administered while 39 children received antithymoglobulin. Stem cell source was bone marrow in 40 (51.9%) patients, peripheral blood in 15 (19.5%) patients and cord blood in the other 22 (28.6%). All pediatric patients received oral or intravenous voriconazole at a dose of 5 mg/kg/12 hours (n=23) or 7 mg/kg/12 hours (n=54) until a top dose of 200 mg/12 hours from day -1 to day +75 or later in patients with acute graft versus host disease (aGVHD). In this study we have just analyzed neutropenic phase, from stem cell infusion day to one week after myeloid engraftment (>500/mm 3 neutrophils). IFD (proven or probable) was diagnosed according to EORTC/MSG criteria. Voriconazole plasma levels were not monitored. AGA was measured twice a week from jun-2006 in 63 children.Results: Median engraftment was 16 days (r: 10-67) and no child died during neutropenic phase. Only two (2.6%) patients developed IFD (probable aspergillosis) within prophylaxis with voriconazole during neutropenic phase but none of them died due to IFD and both are alive today (28 and 22 months after Allo-SCT, respectively). In this large series, voriconazole prophylaxis failed in 11 (14.3%) children while a total of 66 (85.7%) patients successfully completed the prophylaxis without adverse events, no empirical or preemptive antifungal therapy neither IFD. Six (7.8%) receptors needed an empirical (n=4) or preemptive (n=2) antifungal therapy, voriconazole was stopped and replaced with amphotericin B in these six patients. In 3 (3.9%) children were detected grade III/IV adverse events due to voriconazole prophylaxis and needed a definitive withdrawal. These 3 receptors presented persistent hepatic damage with cytolysis enzymes elevation but it was readily and quickly solved in all cases once stopped voriconazole; no other adverse event was detected in this high risk phase. In our study 6 children <2 years old were included, voriconazole was stopped in one of them and empirical treatment was started for persistent fever; the rest successfully completed voriconazole prophylaxis. Discussion: According to the data of our series, voriconazole as primary antifungal prophylaxis during neutropenic phase in children undergoing Allo-SCT is effective in reducing the incidence of IFD (2.6%). Voriconazole is also well tolerated with a low percentage of adverse events easily solved once the drug was stopped without life-threatening complications, moreover only 7.8% of children needed empirical or preemptive antifungal treatment. Disclosure of Interest: None Declared. Introduction: Bronchoscopy performed with the use of modern equipment and adequate anaesthesia can be safely used for pediatric patients suff ering from oncological and hematological diseases.Objectives of our study were the comparison of diff erent anesthetics in children with oncological and hematological diseases during bronchoscopy, assessment of complications, and selection of preferable method of anaesthesia. Materials (or patients) and Methods: We analyzed retrospectively 165 episodes of anesthesia provided to patients aged 8 months -13 years undergoing bronchoscopy in 112 (68%) patients after hematopoietic stem cell transplantation (HSCT) and 53 (32%) patients after high dose chemotherapy. All the procedures were performed in the operating room with constant monitoring of heart and breathing rate, pulsoximetria, ECG, blood pressure (every 3 min). All the patients had spontaneous ventilation. We used the following methods of anesthesia-halothane 19% (n=31), sevofluran 53% (n=87), propofol IV 28% (n=47), combined with local anaesthetic (Lidocain 2%) for high-resolution CT guided bronchoscopy with endoscopic video system Olympus with external diameter of the distal end ranging from 3.6 mm. to 4.9 mm. At the end of the procedure all the patients were transported to the post anesthetic room where they were monitored until full recovery from anesthesia. Patients from intensive care unit (ICU) were not included in this research.Results: There were no serious complications and no one required further ventilation or transportation to the ICU. All complications after anesthesia were transient, required correction or resolved spontaneously. The complications associated with halothane included agitation 19.3% (n=3), laringospasm 6.45% (n=2), dysrhythmia 6.45% (n=2), hypotension 9.6% (n=3), transient hypoxia during emergence 9.6% (n=3) and other 2/2% (n=1). Propofol has some advantages such as quick induction and recovery providing better control, however the complications included hypotension 6.3% (n=3), dysrhythmia 8.5% (n=4), apnea 0.4%, (n=1). Only 2.3% complications were noted when sevofl uran was given (2 patients had agitation during induction). Discussion: At the moment of bronchoscopy sevofl uran is optimal for anesthesia in indicated group of pediatric patients due to quick induction, considerably fast recovery, low rate of complications during anesthesia and emergence. It also provides extra comfort to bronchologist by suppressing cough, swallowing refl ex and bronchodilator eff ect. Disclosure of Interest: None Declared. Introduction: T-cell depletion is an eff ective method to prevent Graft-versus-Host Disease (GvHD) in haploidentical stem cell transplantation. Materials (or patients) and Methods: In order to increase T-cell depletion effi cacy while maintaining anti-tumor and anti-infec-tious activity of the graft, we have evaluated a new method which removes αβ+ T-cells via a biotinylated anti-TcRαβ antibody followed by an anti-biotin antibody conjugated to magnetic microbeads while retaining γδ+ T-cells, Natural killer (NK) cells and others. CD19+ B-cells were concomitantly depleted to prevent EBV-LPD. The CliniMACS® system was used for manipulation of PBSCs from full haplotype mismatched family donors. Results: The overall depletion of αβ+ T-cells was highly eff ective with 4.6 log. Patients received a median number of only 14 x 10 3 / kg residual αβ+ T-cells. Recovery of CD34+ stem cells was 72%, and the median number of infused CD34+ stem cells was 12x10 6 /kg. Additionally, the patients received potential antileukemic eff ector cells: 107x10 6 /kg CD56+ NK cells and 11x10 6 /kg γδ+ T-cells. Diagnoses: ALL (n=20), AML/MDS/JMML (n=9), nonmalignant diseases (n=4), solid tumors (n=2); disease status: CR2-CR6 (n=17), active disease (n=11, 45%). 23 patients received a 2 nd /3 rd SCT (65%). The conditioning regimen comprised fl udarabin 40mg/m2 or clofarabin 50mg/m2 (day -8 to d -5), thiotepa 10mg/kg (d -4), melphalan 70mg/m2 (d -3 and d -2) . OKT3 was used as rejection prophylaxis from day -8 to day -1 in the fi rst 7 patients and was substituted since 2011 by a reduced ATG-F dose (15mg/kg) given at start of the regimen in order not to impair NK and γδ+ T-cells of the grafts (1 mg/kg d -12, 4 mg/kg d -11, 5 mg/kg d -10 and -9; n=28 patients). Short course MMF (until day +30) was given in 25 patients. Graft rejection occurred in 14% of the patients. However, after reconditioning and second stem cell donation, fi nal engraftment was achieved in all patients. Median time to reach neutrophil and platelet recovery in patients with primary engraftment was 10 and 11 days respectively. All patients showed a rapid immune reconstitution with 250 (OKT3 conditioning) and 273 (ATG conditioning) CD3+ T-cells/μl, 30 (OKT3) and 47 (ATG) CD3+4+/μl and 300 (OKT3) and 382 (ATG) CD56+ NK-cells/μl at day +30 posttransplant. γδ+ T-cells started to expand faster than αβ+ T-cells in the early post-transplant period (156 vs 82 cells/μl at day +30) whereas at day +90, αβ+ T-cells were predominant (170 vs 134 cells/μl). Acute GvHD grade 0-I occurred in 25 patients (71%); 6 patients had GvHD II (17%), 3 patients had GvHD III (9%) and one patients experienced GvHD grade IV (3%). 3 patients experienced chronic GvHD (8%). Incidence of acute GvHD was not infl uenced by the number of residual T cells or by type of serotherapy. 2 year EFS was: 80% (nonmalignant diseases); 37% and 9% (acute leukemias, any CR and active disease). Patients with any CR and 1 st SCT showed better results than patients with subsequent SCT (100% vs 30%). Discussion: These data indicate that transplantation of TcR αβ+/ CD19 depleted cells from a haploidentical donor results in fast immune reconstitution and low incidence of both acute and chronic GvHD. OKT3 could be substituted by ATG without negative eff ects. The anti-leukemic effi cacy of this approach in comparison to other methods of T-cell depletion needs to be evaluated with a longer patient follow-up. Disclosure of Interest: None Declared. Introduction: Allogeneic bone marrow transplantation (BMT) has become the treatment of choice for severe aplastic anemia (SAA) over the past decades and the outcomes of BMT have improved even from unrelated donors. Inclusion of rabbit antithymocyte globulin (rATG; ThymoglobulinR) as a part of conditioning regimen is known to reduce the incidence and severity of acute and chronic graft-versus-host disease (GVHD). However, the optimal dose of rATG is not still established in BMT for childhood SAA. We retrospectively compared the results of two doses of rATG in allogeneic BMT. Materials (or patients) and Methods: A total of 230 pediatric patients aged 1 to 15 years who received allogeneic BMT for SAA with the conditioning regimen including rATG at a dose of 5 mg/ kg (ATG-5 group; 81 patients) or 10 mg/kg (ATG-10 group; 149 patients) were enrolled in this study.

Results: There were no statistical diff erences between two groups in age and the male to female ratio. Introduction: There isn't specifi c antiviral drug with proven efficacy against BK virus (BKV) replication and for BKV hemorrhagic cystitis (HC). Lefl unomide is an immune suppressive drug with antiviral activity. To the best of our knowledge, these are the fi rst reported cases of a pediatric BKV-HC patients, who received lefl unomide therapy Materials (or patients) and Methods: Case 1. A 14 years-old male patient with relapsed ALL, underwent haploidentical HSCT from his 49 years old mother. Conditioning consisted of total body irradiation (TBI), thiotepa, melphalan and rabbit ATG. 4.5 x 10 6 /kg CD34+ stem cells were transplanted. There were no BKV HC, but he underwent 2th haploidentical HSCT from his 23 years old sister on the 50th day, because of primary graft failure. Cyclophosphamide, rabbit ATG and fl udarabine given for conditioning. Full engraftment was achieved. Neutrophil, platelet and erythrocyte recovery were 21 days, 30 days and 35 days respectively after the 2th HSCT. On the 24th day of the 2th HSCT (74th day of the fi rst HSCT), urine BKV loads increased to 9 × 10 9 copies/ml with HC. There were no improvement in BKV HC despite bladder irrigation, intravesical and i.v cidofovir, eptacog alpha, hyperbaric oxygen respectively; also despite nephrostomy and cystostomy because of pelviectasis. Oral lefl unomide therapy was started with a dosage of 40 mg/day, on the 105th day of HSCT, after the increase of viral BKV load to 4×10 5 copies/ml in blood also. After three days of this dosage, it is reduced to 20 mg/day. On the 10th day of lefl unomide therapy, his symptoms, especially pain and sypmtoms resolved, viral BKV was negative in blood and 10 5 copies/ml in urine (Figure) . Case 2. A 15 years-old male patient with CML underwent matched unrelated donor HSCT after his second remission. Conditioning consisted of TBI, rabbit ATG and cyclophosphamide. 5.8 x 10 6 /kg CD34+ stem cells were transplanted. Full engraftment was achieved with neutrophil, platelet and erythrocyte recovery being 16 days, 17 days and 23 days after the HSCT respectively. On the 15th day of the transplantation, urine BKV loads increased to 9 × 10 9 copies/ml with HC. He received the same therapies of case 1, but there were no improvement. Oral leflunomide therapy was started with the same dosage on the 43th day of HSCT but discontinued after two weeks, because of no improvement in BKV HC. Bladder irrigation went on and he was discharged on the 94th of HSCT, after the resolution of his signs and symptoms. Viral BKV load was negative in blood throughout the HSCT and decreased to 5 × 10 8 copies/ml in urine at his discharge. Results: No advanced GVHD occurred when lefl unomide was administered and no signifi cant side eff ects were observed during lefl unomide treatment in both of two cases. Discussion: Based on the satisfactory effects of leflunomide in treating BKV-associated nephropathy after renal transplantation, it is inferred that it might be effective for treating BKV-HC. Prospective, randomized control studies are needed to confirm its efficacy, but it may worth trying leflunomide in pediatric HSCT recipients, especially who are refractory to first line treatments. Disclosure of Interest: None Declared.

S380 Introduction: Hematopoietic stem cell transplant (HSCT) can benefi t with various non-malignant diseases but is limited by regimen-related toxicity especially in high risk patients, graft-versus-host disease (GVHD), donor availability, and graft rejection. To overcome some of these barriers, we adopted a novel conditioning strategy for these patients. Materials (or patients) and Methods: Between 2004 and 2013, 37 patients with high-risk non-malignant hematological disorders underwent allogeneic HSCT using a standard institutional protocol. All patients received preparative regimen consisting of oral busulfan 2 mg/kg given every 12 h for two days from days -8 to -7, fludarabine 35 mg/m 2 infused over 1 h daily for 5 days from days -6 to -2, horse antithymocyte globulin (ATG) 30mg/kg infused over 12 h daily for 5 days from days -6 to -2 was used in the first 10 patients, while thymoglobulin at 2.5 mg/kg once daily for 3 consecutive days was used in the rest, and total lymphoid irradiation administrated as a single fraction of 500 cGy on day zero prior to stem cell infusion for all patients. Cyclosporine and mycophenolate mofetil were used for GVHD prophylaxis. All donors were fully HLAmatched, 33 (89%) were HLA identical siblings and 4 (11%) were other family members (3 parents, and one cousin). Mobilized peripheral blood stem cells through 5 days of subcutaneous injections of GCSF at 5μ/kg/dose twice daily was used in all donors Results: Thirty-seven consecutive patients with a median age of 13 year (range, 0.3-25) were treated. Twenty-nine patients had high risk class 3 thalassemia major, 3 congenital pure red cell aplasia, 3 sickle cell diseases, one infantile osteopetrosis, and one autoimmune lymphoproliferative syndrome. All patients had significant morbidity before transplantation, including 10 patients with hepatitis C infection. All patients received peripheral blood stem cells, with median CD34 count of 4.9 x10 6 /kg recipient weight (3.5-8.2 Introduction: Treosulfan is a bifunctional alkylating drug with a structure similar to Busulfan. It is increasingly used in children prior to allogeneic haematopoietic cell transplantation (HCT). Pharmacokinetic (PK) data on Treosulfan in children is limited: we describe the PK profi le of Treosulfan in 22 children undergoing HCT in the UK. Materials (or patients) and Methods: Twenty-two children who underwent HCT in Newcastle and London between 2012 and 2013 were analysed as part of a pilot Treosulfan PK study. Treosulfan was administered iv over 2 hours at a dose of 10 (age < 3 months), 12 (age 3 months-1 year) or 14 (age > 1 year) g/m 2 / day for 3 consecutive days. Treosulfan concentration in plasma was determined at diff erent time points by a validated RP-HPLC method with refractometric detection. The non-compartmental AUC (0-∞) was calculated for each subject, and the data was modelled with the population PK software NONMEM. Results: Median age at HCT was 18 months (range 1-170). Indication to HCT was: primary immunodefi ciency (15), infl ammatory bowel disease (4), relapsed JMML (2) and osteopetrosis (1). Median weight was 10 kg (range 3.76-55.5). Conditioning was Treosulfan and Fludarabine 150 mg/m 2 (22) + Thiotepa 10 mg/m 2 (2) . Alemtuzumab was given to 18 children. Donors were: matched unrelated (8), 1-antigen mismatched unrelated (7) (1), VOD (1), cyclosporine-induced neurotoxicity (2) and severe mucositis (1). The 2-compartmental model provided best fi t with a clearance of 12.6 L/h/70kg and a Vss of 34. One patient died before engraftment. Median time to absolute neutrophile counts above 500/μl was 10 (9-32) days. Independence from platelet substitution was reached at a median time of 10 (8-22) days. 11% developed GvHD °III, 11% developed GvHD °II, 28% of all patients developed GvHD °I and 50% had no signs of GvHD. Severe organ toxicity was observed in 5 patients (bronchiolitis obliterans n=1, hemorrhagic cystitis n=2, leukencephalopathy n=2). Event free survival (EFS) of all patients at 2 years was 63%. 2 year EFS of patients with leukemias in complete remission was 75%. Patients with non-malignant diseases had 5 year EFS of 80%. Transplant related mortality at one year was 11%. Causes of death were: multi organ failure (n=2) and 4 patients with acute leukemias died of relapse. None of the patients rejected the second graft. Discussion: Thus, retransplantation from haploidentical donors with T cell depleted grafts after 7 Gy TNI based reconditioning is a realistic option to rescue patients with graft failure within a short time span and for whom a second stem cell donation of the original donor is not available. Introduction: The treatment of relapsed high risk neuroblastoma in children is a challenge. Haploidentical stem cell transplantation provides a therapeutical platform for the transfer of expanded allogeneic NK cells. The combination with GD2 monoclonal antibody (CH14.18/CHO) treatment is likely to exceed antitumor activity. To investigate the alloreactive potential of NK cell clones with and without KIR receptor-ligand mismatch, we sorted and expanded defi ned NK subsets. Materials (or patients) and Methods: 4 defi ned NK phenotypes from 3 healthy donors were sorted via FACSAria cell sorter. NK single cell clones were high effi ciently expanded (6.0-6.6 logs) using K562mb15 4-1BBL feeder cells and IL-2. After 21-28 days of expansion, phenotype of NK cell clones was confi rmed by FACS analysis. Cellular cytotoxicity as well as ADCC-mediating GD2mAB was measured in a 2 hour BATDA release assay against LAN-1 and LS. NK clones were matched by K562 lysis. NK phenotyping included CD56, CD16, CD158a, CD158b and CD158e. CD3 positive clones were excluded. LAN-1 (bw4 + , cw3 + , cw4 -) and LS (bw4 + , cw3 + , cw4 + ) were well characterized by fl ow cytometry. Results: Phenotype of NK cell clones was reliably defi ned by sorting procedure. LS showed signifi cantly higher expression of NKG2D ligands and GD2. LAN-1 expressed signifi cantly lower levels of HLA I per cell. CD158a + (R/L-mismatched) NK clones of all 3 donors lysed LAN-1 signifi cantly higher without (P=0.0047) and with GD2mAB (P=0.0001) than CD158b + (no R/L-mismatched) clones whereas there was no signifi cant diff erence in the lysis of LS by CD158a+ clones vs CD158b+ clones in all 3 donors without and with GD2mAB. Median specifi c lysis for LS 42.1% without and with GD2mAB 56% (E:T, 5:1) was signifi cantly higher than median specifi c lysis of LAN-1 13.7% and with GD2mAB 35.8% (n=198; for both conditions P<0.0001). GD2 antibody enhanced specifi c lysis signifi cantly in LAN-1 and LS (LAN-1 P<0.0001; LS P<0.0001) independent from phenotype in CD16+ NK cell clones. Increase of GD2mAB-mediated ADCC was signifi cantly higher in LAN-1 than LS (P<0.0001). Discussion: Here we could show that NK clones with R/L-mismatch exert a signifi cant better lysis than clones without KIR-mismatch. This model provides evidence for NK mediated alloreactivity not only in leukemias but also in neuroblastoma cell lines. In general NK alloreactivity in neuroblastoma may be signifi cantly infl uenced by KIR R/L-mismatch and should therefore be taken into account for donor selection strategies. Ex vivo expanded, highly activated alloreactive haploidentical NK cells could be used for NK cell transfer posttransplant in combination with GD2mAB to augment antitumor activity. NK cell clones can be expanded high effi ciently with predefi ned KIR-receptor phenotype via FACS sorting. NKG2D expression seems to be superior for the induction of NK cell mediated lysis in this certain range of HLA I quantity, which would explain higher specifi c lysis for LS than LAN-1. The signifi cantly higher increase of ADCC-mediated lysis may be due to signifi cantly higher expression of GD2 on LAN-1 compared to LS tumor cells. An effi cient NK cell clone expansion protocol has been established for further evaluation of NK alloreactivity in diff erent tumor entities. Disclosure of Interest: None Declared. Introduction: Allogeneic stem cell transplantation is still the curative treatment option of choice for the majority of patients with severe aplastic anemia and refractory cytopenias. However, a HLAmatched donor is not available for all patients. Alternative donor transplantation has been an experimental treatment option, limited by high rejection rates and transplant related mortality.

Materials (or patients) and Methods: We performed a prospective clinical trial to evaluate the safety and feasibility of haploidentical stem cell transplantation, since haploidentical family members are always available donors. We investigated a cohort of 10 pediatric patients with severe aplastic anemia or myelodysplatic syndrome (refractory cytopenia) transplanted with T-cell depleted grafts between 2004 and 2011. Results: 7 patients had myelodyslastic syndrome with refractory cytopenia (MDS-RC), 3 had severe aplastic anemias (SAA) refractory to immunosuppressive treatment. 3 patients received a 2nd SCT after rejecting the graft from matched donors. Median age was 11.4 years. Standard conditioning regimen consisted of Fludarabin 3-4x40 mg/m 2 , Thiotepa 1-3x5 mg/kg, Melphalan 2x70 mg/m 2 (n=8) and serotherapy using OKT3 (n=5) and ATG (n=5). 8 patients received additional total lymphoid irradiation (TLI 7 Gy) to prevent graft rejection. In vitro graft manipulation was carried out by direct depletion using antiCD3/19 magnetic microbeads. A median number of 10.1x10 6 CD34 + progenitor cells and 27x10 3 T-cells/kg body weight (BW) were transfused. Pharmacological GvHD prophylaxis (graft vs. host disease) was carried out with Mycophenolate until day 60, if residual T-cells in the graft exceeded 25000/kg BW. Primary engraftment occurred in all patients Median time to reach 500/μl neutrophiles was 9 days (9-11). Independence from platelet substitution was reached after 13 days (8-16). Three patients rejected the graft later on. 6/10 patients had no signs of acute GvHD or GvHD grade I, 2 patients had GvHD grade II. TRM at day +100 and after 1 year was 0% and 20%, respectively. After TLI no rejection was observed. Event free survival (EFS) at 3 years was 80%. Discussion: Haploidentical SCT with T-cell depleted grafts is a therapeutic option for refractory cytopenias and severe aplastic anemia after nonresponse to immunosuppressive treatment if no HLA-matched donor is available. Recovery of neutrophiles and platelets were fast and TRM was low, even if retransplantation was necessary. Since spontaneous outcome of these conditions are poor, alternative donor SCT is a realistic option for these patients. Disclosure of Interest: None Declared. Introduction: GMP-grade NK cell expansion for clinical purpose has been demonstrated feasible and safe. In our project we com-pared conventional fl ask-based cell culture NK expansion with the NK expansion using an innovative fl ow-through bioreactor (Zell-werk®) from healthy volunteer donors and produced and applied 7 NK products in 2 patients post allogeneic hematopoietic stem cell transplantation. Materials (or patients) and Methods: NK cells were expanded using 3 healthy donors' untouched isolated PBMCs and CD3depleted PBMCs (100ml PB) under static and dynamic expansion conditions. Isolated PBMCs or CD3-depleted PBMCs were pooled with 100Gy irradiated K562mb15 4-1BBL feeder cells 1:10 and divided equally into two portions. One portion was seeded in cell culture fl ask (175cm 2 ) at a density of 1.1E6/ml and one portion was seeded in the bioreactor chamber containing an absolute volume of 50 ml. Harvest of cells was performed at day 15-17. Isolated PBMCs and the expanded NK products were well-characterized by FC. 2h BATDA release was performed against various targets including K562, LAN-1 wo/w GD2mAB. Additionally two patients received static expanded NK cell products without and with GD2mAB. Results: NK cells were expanded 68-688 fold in 14-17 days. Conventional flask-based expansion revealed significant higher numbers of NK cells (350-688 fold) compared to bioreactor expanded NK cells (68-290 fold; P=0.0045) and untouched PBMC NK expansion reached significant higher NK cell count compared to CD3-depleted NK expansion of the same donor (P=0.008). Static and bioreactor expanded NKs showed high direct cellular cytotoxicity against all cell lines tested and showed excellent ADCC activity using GD2mAB (LAN-1; LS) and CD19mAB (MHH-4; Raji). There was no significant difference in viability of cells and NK phenotype but for functional properties static culture showed significant higher lysis than bioreactor expanded NKs (n=35; P=0.0178). The autologous NK expansions (n=7) post stem cell transplantation for NK transfer in two patients reached 64-5700fold expansion. The NK cell transfer of high numbers of NK cells (150-505E6 NK cells/kgBW) induced transient coughing and elevated temperature. Post transfer isolated PBMCs showed significantly increased direct and ADCC mediated cytotoxicity. CD69+ cells could be traced for several days after NK transfer. Patients' expanded NK cells showed the same phenotype and functional properties like NK cells expanded from healthy donors. Discussion: The expansion of NK cells under static rather than dynamic culturing conditions seems to be superior in terms of total NK cell count and functional properties. Nevertheless static culturing requires higher eff ort of maintenance and therefore needs an every-day engagement, whereas the bioreactor facilitates to program feeding rate of expanding cells for several days ahead. Thus we will further optimize bioreactor NK expansion to reduce handcraft during expansion period. NK transfer of autologous expansion NK cells post allogeneic SCT was tolerated alone as well as combined with GD2mAB (single dose 20 mg/m 2 ) and is likely to enhance NK cell antitumor activity as it has been shown for various targets in vitro. Disclosure of Interest: None Declared.