key: cord-0005643-vac061r8
authors: nan
title: Physicians Abstracts: EBMT 2010
date: 2010-04-07
journal: Bone Marrow Transplant
DOI: 10.1038/bmt.2010.40
sha: 4afee7d50704ca86db033ea887dc7cb5745a7ee5
doc_id: 5643
cord_uid: vac061r8

nan

B cells have been demonstrated to present antigen to T cells in vivo. CD40-activation dramatically improves antigen presentation by normal and malignant B cells and has therefore been studied as an approach to generate autologous "non-artifi cial" antigen presenting cells for active immunotherapy. Furthermore, CD40-B cells have recently been shown to expand tumorantigen and viral specifi c CTL as well as regulatory T cells and are therefore of great interest for post-transplant immunotherapy. Human B cells when activated via CD40-L/IL-4 can be expanded from small amounts of peripheral blood in 12-14 days. CD40-activated B cells can prime naïve CD4 + and CD8 + T cells, expand memory T cells and express important surface homing molecules. Nevertheless, it remains unclear whether CD40-activated B cells migrate to secondary lymphoid organs (SLO) in vivo to attract and interact with T cells. To address this question we established a methodology to generate murine CD40-activated B cells. At day 14 of culture, these cells are >95% CD19 + and CD80/86/MHCI/MHCIIhi. Murine CD40-activated B cells present a 'homing phenotype'; migrate toward SLO chemokines such as CCL19, CCL21 and CXCL13; and induce T-cell chemotaxis in vitro. Upon CD40L activation, B cells up-regulate CCR7 while down-regulating CXCR5 expression which suggests direction of activated B cells toward the B-zone/T-zone boundary. We compared the homing of GFP + CD40-activated B cells to resting GFP + B cells and show for the fi rst time that CD40-activated B cells home to SLO significantly more effi ciently than resting B cells. Furthermore, CD40activated B cells localize in the B-cell areas, and a signifi cant fraction move to the B-T boundary close to the T-cell zone over time. To dissect T-cell-APC interactions on a single cell we analyzed three-dimensional migration in collagen matrix using time-lapse videomicroscopy. Interestingly, antigen-loaded CD40-activated B cells differ from immature and mature DC by displaying a rapid migratory pattern undergoing highly dynamic, short-lived (7.5 min) and sequential interactions with cognate T cells. Taken together, these data revealed that CD40-activated B cells can home to secondary lymphoid organs and interact dynamically with T cells thus underlining their potential as cellular adjuvant for cancer immunotherapy.

Long-term follow-up of upfront tandem autologous -RIC (reduced intensity conditioning) allogeneic transplantation versus autologous transplantation (NMAM2000) in multiple myeloma G. Gahrton, B. Björkstrand, S. Iacobelli, U. Hegenbart, A. Gruber, H. Greinix, L. Volin, F. Narni, P. Musto, M. Beksac, A. Bosi, G. Milone, P. Corradini, H. Goldschmidt, T. de Witte, C. 

Objectives: Treatment of multiple myeloma with allogeneic hematopoietic stem cell transplantation is controversial. The NMAM2000 study compares tandem autologous (ASCT)/ Reduced intensity conditioning (RIC) allogeneic transplantation (RICallo) to only ASCT or ASCT in tandem (ASCT2) in a prospective study based on genetic randomization. Patients and Methods: Out of 358 myeloma patients accrued during 2001-2005 from 26 EBMT centres 109 with an HLAidentical sibling donor were allocated to tandem ASCT/RICallo and 249 without to ASCT only. Previously untreated patients with at least stable disease after VAD (vincristine, doxorubicine, dexamethasone)-like induction treatment were included at the time of conditioning for ASCT. Single (n = 145) or tandem (n = 104) ASCT was optional in the ASCT arm. Conditioning for ASCT was melphalan 200 mg/m 2 , and for RICallo fl udarabine 30 mg/m 2 x 3 plus TBI 2 Gy. The two treatment groups were well matched for standard prognostic parameters and response status at ASCT. Results: On an intention to treat basis the cumulative 24 and 60 months non-relapse-mortality (NRM) was 12 % and 16% with ASCT/RICallo and 3% and 4% with ASCT respectively (p = 0.0001 (Gray test)). The CR rate within 60 months was 52% with ASCT/RICallo and 41% with ASCT (p = 0.0009: improved in time (Fine/Gray model)). At 60 months after ASCT Relapse/Progression rate was 49% and 78% (p = 0.0014: reduction in time Fine/Gray model)), PFS 35% and 18% (p = 0.0012 reduction of risk in time (Cox)) and OS 65% and 58% (p = 0.0048 reduction in time (Cox)) (at 84 months 60% and 22%) for ASCT/RICallo and ASCT respectively. A comparison between those patients who received a second allo (n = 91) versus a second auto (n = 104) Relapse/Progression rate at 60 months from second transplant was 43% and 78% and PFS 39% and 19% in ASCT/RICallo and Auto2 respectively. Information about the chromosome 13 deletion (del(13q14)) was present in 214 patients. In 92 patients with the deletion OS at 60 months was 69% and 55%, and PFS 31% and 11% for ASCT/RICallo and ASCT respectively. GVHD in RICallo was as expected (aGVHD grade 0-1: 80%; grade 2-4: 20%; cGVHD limited: 36%; extensive: 27%). Conclusion: The risk of myeloma relapse is lower with tandem ASCT/RICallo as compared to ASCT or tandem ASCT, both in an intention to treat analysis and in patients that received the correct treatment. NRM is higher, but considered acceptable in view of the improved PFS and OS with tandem ASCT/RICallo.

A fi ve biomarker panel predicts acute graft-versus-host disease S. Paczesny (1) , D. Bickley (1) , S. Choi (1) , J. Crawford (1) , T. Braun (1) , S. Pitteri (2) , J. Hogan (2) , P. Reddy (1) , S. Hanash (2) , J. Ferrara* (1) , J. Levine* (1) (1)University of Michigan (Ann Arbor, US); (2)Fred Hutchinson Cancer Research Center (Seattle, US) * have contributed equally to this work Current evidence suggests that a composite panel of four biomarker proteins (IL2Ra, TNFR1, IL8, HGF) is diagnostic and prognostic for acute graft-versus-host disease (GVHD). Elafi n is a biomarker that has also been found to be diagnostic for GVHD of the skin, as well as prognostic for overall survival (OS). We sought to evaluate whether these fi ve biomarkers are able to predict future occurrence of GVHD, non-relapse mortality (NRM) and OS. We measured by sequential ELISA, levels of these fi ve proteins in plasma collected prospectively from 485 allogeneic HCT patients randomly divided into training (149 GVHD-, 175 GVHD + ) and validation (74 GVHD-, 87 GVHD + ) sets. We obtained samples 3-14 days before onset of GVHD (median day 29) and at equivalent time points in patients without GVHD. There were no signifi cant differences between sets in age, conditioning intensity, donor source, HLA match or GVHD grade between training and validation sets. The median day of sample acquisition was day 20 and day 21, respectively. Logistic regression determined that a linear combination of the fi ve proteins levels predict the occurrence of acute GVHD in the training set. The Receiver Operating Characteristic curves of each of the fi ve individual biomarkers are shown in Figure 1 with an area under the curve (AUC) for the composite panel of 0.77 (95%CI: 0.72-0.82). When this model was applied to samples of the validation set, the corresponding AUC was 0.76 (95%CI: 0.69-0.84). This 5-biomarker panel therefore discriminated between patients who later developed GVHD and those who did not. Proportional odds logistic regression models determined that the panel gave prognostic information regarding the eventual maximum grade of GVHD (p<0.001 in training set, p = 0.05 in validation set). Given this correlation, we next divided the patients into high and low risk groups based on their predicted probability of developing GVHD (high risk ≥ 0.7 and low risk < 0.7) and analyzed these groups for differences in NRM, relapse mortality and OS. The differences in 1 year NRM and OS between groups were signifi cant in both the training and validation sets (Table 1 ). When adjusted for other known risk factors (age, conditioning intensity, donor source, and HLA match), the difference in OS remains signifi cant (p = 0.02) in both sets. In conclusion, a 5-biomarker panel can predict GVHD before any clinical manifestation and provides prognostic information including long term survival. S3 their interaction with dendritic cells. Recent clinical data has indicated that the content of plasmacytoid dendritic cells (pDC) in an allograft is associated with the risk of relapse following allogeneic bone marrow transplantation (BMT). We have previously shown that the addition of donor pDC to a graft comprised of purifi ed HSC and T-cells led to enhanced Th1 activation of donor T-cells with improved GvL activity without increasing GvHD in multiple murine BMT model systems (Li and Waller JI 2009 ). Here we studied the mechanism that donor pDC augment GvL without increasing GvHD, and present a novel model for regulation of post-transplant immunity. Methods: Donor hematopoietic stem cells (HSC), T-cells and pDC in the allograft were rigorously purifi ed using immuno-magnetic selection and high-speed fl uorescent-activated cytometry (FACS) in the H2b B6->H2k B10.BR transplant model. Donor cell subsets purifi ed from wild type (WT) or interferon gamma knock-out(KO),(IFN-gKO), interferon gamma receptor KO (IFN-gRKO), and indolamine 2,3 dioxygenase (IDO) KO (IDO-KO) were combined to determine the role of the IFN-g and IDO in the separation of GvHD from GvL in murine BMT. Results: In the B6->B10.BR BMT model, the addition of 50,000 donor pDC to a graft comprised of 3,000 HSC and 300,000 T-cells lead to Th1 immune polarization and enhanced proliferation of donor T-cells with higher levels of donor T-cell chimerism and improved GvL activity without increasing GvHD. The absence of allo-reactivity was dependent upon donor T-cell synthesis of IFN-g and the presence of IFN-gR on donor pDC. Co-culture of T-cells with syngenic pDC in one-way MLR lead to up-regulation of IDO in donor pDC. Increased GvHD after transplanting pDC from IDO KO donors in combination with WT T-cells and HSC demonstrated the central role of IDO in the initiation of counter-regulatory effects that limit GvHD. WT T-cells and pDC lead to the generation of donor-derived T-reg in BMT recipients that limited GvHD. pDC from IDO-KO donors had markedly reduced numbers of donor T-reg.

Conclusions: Donor pDC regulate the initial activation and GvL activity of donor T-cells and subsequently generate T-reg that limit GvHD. The dynamic regulation of the activation and alloreactivity of donor T-cells by donor pDC suggest novel clinical approaches to enhance GvL activity without GvHD.

J. Peccatori (1) , D. Clerici (1) , A. Forcina (1) , M. Bernardi (1) , R. Crocchiolo (1) , C. Messina (1) , M. Noviello (1) , S. Mastaglio (1) , F. Giglio (1) , S. Malato (1) , M.T. Lupo Stanghelllini (1), S. Marktel (1) , A. Assanelli (1) , M. Battaglia (1) , A. Ferraro (1) , S. Rossini (1) , M.E. Bernardo (2) , A. Bondanza (1) , M. G. Roncarolo (1) , C. Bonini (1) , F. Locatelli (1) , F. Ciceri (1) (

Background and aim: Rapamycin, in contrast to calcineurin inhibitors, allows T regulatory cell (T-regs) proliferation while inhibits effector T cell expansion. We investigated the safety of infusion of T-cell repleted unmanipulated PBSC from haploidentical donor with a combination of Rapamycin, Mycophenolate and ATG as GvHD prophylaxis, to preserve early T-regs function (TrRaMM study, Eudract 2007-5477-54) . Patients and methods: Since 2007, fi fty-nine patients (pts) underwent SCT for AML (39 pts), ALL (9), MDS (3), CML-BC (4), NHL (2) or HD (2) . Median age was 50 years (range 14-69). At transplantation 7 pts were in early phase, and 52 in advanced phase. Median comorbidity index (CI) score was 1 (0-5). The conditioning regimen included Treosulfan (14 g/m 2 for 3 days), Fludarabine (30 mg/m 2 for 5 days) and an invivo T and B-cell depletion by ATG-Fresenius (10 mg/kg for 3 times) and Rituximab (a single 500 mg dose). All pts received allogeneic unmanipulated PBSC from an HLA-haploidentical related donor. GvHD prophylaxis consisted of Rapamycin (target level 8-15 ng/ml, till day + 60) and MMF (15 mg/kg tid till day + 30). Results: All patients engrafted and all but eight were in disease remission at fi rst marrow evaluation on day + 30. Cumulative incidence of grade 2-4 and grade 3-4 aGvHD were 29 and 13% respectively. Only 12 patients developed cGvHD. Cumulative incidence of TRM and relapse incidence were 25% and 44% respectively. After a median follow-up of 8 months, projected OS at 1 year is 43%. Immunoreconstitution was fast and sustained with a median 221 circulating CD3 + cells/μL (range 43-1690) from day 30. The immune-reconstitution was polarized towards central memory (CD45RA-CD62L + cells 32.7% ± 4.8), IL-2 producing cells (IL-2 + cells 26.2% ± 5.3). We detected high levels of CD4 + CD25 + CD127-FOXP3 + T regulatory cells (up to 30% of circulating CD4 + T lymphocytes) starting from day 30.

These cells were able to suppress in vitro proliferation of autologous effector cells demonstrating to be regulatory T cells. Conclusions: Rapamycin-Mycophenolate-ATG are effective as GvHD prevention in T-cell replete unmanipulated haploidentical peripheral SCT and are associated with an early T-cell immunoreconstitution characterized by the in-vivo expansion of earlydifferentiated T cells and T-regs. Further studies are warranted to gain insight on the role of Rapamycin as platform for exploitation of T-regs in allogeneic HSCT from mismatched donors.

Direct interaction of human NK cells with Aspergillus fumigatus induces a Th1 immune response and provokes signifi cant fungal killing but not via their usual cytotoxic pathways M. Bouzani, M. Ok, O. Kurzai, H. Einsele, J. Loeffl er Wurzburg University (Wurzburg, DE) Objectives: Invasive aspergillosis (IA), caused mainly by Aspergillus fumigatus (AF), is a highly devastating disease for immunosuppressed subjects. Host's defence is principally confi ned to innate effector cells like alveolar macrophages, neutrophils and dendritic cells. In our study, we questioned the possible interaction of AF with another potent component of the innate immunity, the Natural Killer (NK) cells. Methods: Human NK cells were isolated after magnetic depletion of the peripheral blood of volunteers and they were used after 24h priming with 500 U/ml recombinant human interleukin 2, rhIL2. Interferon gamma (IFN-g) and Tumor Necrosis Factoralpha (TNF-a) regulation were assessed after NK-AF coculture. Fungal damage was investigated through plate killing assays. To investigate the infl uence of rhIL2 on NK cells, plate killing experiments were performed with resting and primed NK cells. Transwell permeable membranes, NK cell granule depletion (treatment with strontium chloride), surface expression of degranulation markers CD107a/b and neutralization of NK death ligands (TNF-related apoptosis-inducing ligand [TRAIL] and FasL) by blocking antibodies were used to evaluate the means of interaction. Results: Fungal germlings induced towards NK cells a Th1 immune response with upregulation of IFN-g and TNF-a (p<0.05). NK cells displayed strong fungicidal effect against germlings (p<0.05), but they were inactive against conidia. Priming with rhIL2 (p<0.05) and direct effector-pathogen contact (p<0.001) were required for their interaction. The cytotoxic effect was not attributed neither to the release of perforingranzyme, nor to the engagement of NK cell death ligands. Conclusion: Human NK cells are stimulated in vitro by AF, which triggers a Th1 immune response and causes important fungal killing. This interaction requires priming of NK cells with rhIL2 and conditions of direct contact between NK cells and fungus. Interestingly, NK cells do not mediate their cytotoxic effect via perforin -granzyme pathway, neither through the engagement of death ligands, suggesting that another pathway is involved in NK cell -AF interaction. Our study attributes to NK cells anti-Aspergillus properties, suggesting them as a future potential immunotherapeutic tool against IA.

Objective: To conduct a survey on indications, effi cacy and safety of growth hormone (GH) treatment in children and ado-lescents after haematopoietic stem cell transplantation (HSCT) within the EBMT. Methods: In this retrospective survey using a two step approach, we asked (1st questionnaire) for information on endocrine follow-up, the use of GH, reasons for not using GH, indications for GH treatment, number of patients treated with GH and diagnostic tests to diagnose GH defi ciency (GHD); and (2nd questionnaire, to follow) data on growth, diagnosis of GHD, interval between HSCT and GH treatment, dose schedules, other endocrine defi ciencies. Results: 1st questionnaire: 53 centres replied until 15.11.2009: Endocrine follow-up (n = 50) was performed by the HSCT-clinic and an endocrinologist in 34%, the HSCT-clinic in 12%, an endocrinologist in 38%, a specialist in another clinic in 10% and the HSCT-and another clinic in 6%. 63% (33/52) centres treated patients after HSCT with GH, whereas 37% (19/52) did not. Of the centres reporting, 3/10 centres from Italy used GH, whereas 7/10 did not. In contrast, 6/7 centres from France used GH whereas 1/7 did not. German centres reported the use in 3/6 centres. All centres from Finland (2), Great Britain (2), Spain (2) and Switzerland (2) reported the use of GH. Indications for GH treatment were growth failure in 6% (2/32), growth failure or GHD in 50% (16/32) and GHD only in 44% (14/32). Reasons for not using GH were: no indication in 74% (14/19), risk of side effects in 11% (2/19) and lack of fi nancial support in 5% (1/19) . The numbers of patients treated in centres was less than 5 patients in 28% (9/32), 5 to 10 patients in 31% (10/32), 10 to 20 patients in 22% (7/32) and more than 20 patients in 19% (6/32). Diagnostic tests for GHD used were the insulin-test (n = 17), arginin-test (n = 14), spontaneous GH-secretion (n = 10), clonidin-test (n = 10) and GHRH-test (n = 9). Conclusions: In the majority of centres the endocrine follow-up was performed by an endocrinologist. GH treatment was regularly used for the treatment of growth failure after HSCT. About 50% of the centres used GH treatment for growth failure due to GHD only. In centres not using GH the main reason was that they saw no indication for GH treatment. Only a minority did not use GH due to the risk of side effects. The insulin-test, arginintest and spontaneous GH-secretion were the major tests used for the diagnosis of GHD. A. Rovo, M.T. van Lint, M. Aljurf, N. Salooja, G. Sucak, A. Hunter, M 

Within nonmalignant complication after hematopoietic stem cell transplantation (HSCT) gonadal dysfunction with absence of sperm production leading to defi nitive infertility is a common long-term sequela. Young age at HSCT and longer interval time since HSCT has been associated as a favourable factor for spermatogenesis recovery. The total body irradiation (TBI) used as part of the conditioning regimen plays a central role in posttransplantation infertility. We assessed in a retrospective multicenter study on a large cohort of male survivors, the probability of sperm recovery after HSCT, and evaluated associated factors for fertility recovery. Male patients in which at least one seminal fl uid analysis was performed after HSCT being in complete remission and in whom the results were available were the target population. Ninety centers reporting to the EBMT were asked to participate, 23 accepted and so far 15 centers contributed with 217 patients. The median age at HSCT and at time of 1st sperm fl uid analysis (SFA) was 23 and 29 (5-64) years respectively. The median time interval between HSCT and SFA analysis was 54 months . 206 (93%) received an allogeneic HSCT, 169 (73%) had bone marrow as source of stem cells, 201 (94%) received a myeloablative conditioning and 149 (67%) received TBI (with a median doses of 12 Gy (7.5-14.4) as part of the conditioning. During the follow-up 130 recipients (63%) had any grade of acute GvHD, 128 (64%) any type of chronic GvHD (cGvHD); 23 (17%) from them had ongoing cGvHD at SFA time. Presence of at least one spermatozoa in SFA was considered as recovery of spermatogenesis. Spermatozoa were detected in the semen in 58/217 (27%) patients; 19 (9%) of them showed normozoospermia. In the univariate analysis following associated factors were signifi cant (Table 1) : younger age at HSCT, longer time interval between HSCT and SFA, conditioning without TBI, patients without ongoing cGvHD. In the multivariate analysis, absence of TBI (p<0.00001), longer time interval (p = 0.002), absence of ongoing cGvHD at SFA (p = 0.037) were signifi cant. In conclusion, this is the largest population of male survivors evaluating spermatogenesis after HSCT. In this study we confi rm the established factors which infl uence recovery of spermatogenesis such as age (univariate analysis), time interval between HSCT to SFA and TBI. Here for the fi rst time evidence for graft versus testis effect can be demonstrated.

Introduction: Development of leukemia or myelodysplasia derived from donor cells termed donor cell leukemia (DCL) is a rare but severe complication following allogeneic hematopoietic transplantation. The estimated incidence is low ranging from 124 cases of DCL per 100,000 transplants (mostly myeloablative conditioning, retrospective analysis) to 5,000 per 100,000 transplants (non-myeloablative, single institution experience). Although about 60 cases have been reported in the literature, very little is known about pathogenesis and clinical management of donor cell leukemia. Factors proposed to be involved in malignant transformation and development of DCL include immunoregulatory dysfunction, immunosuppression, host environment triggering malignant processes, replicative stress and epigenetic reprogramming as well as antigenic or viral stimulation. Due to the increasing number of case reports throughout the last few years the Late Effects Working Party decided to analyze and evaluate the experiences with donor cell leukemia within the EBMT centers. Methods: A fi rst, short questionnaire will be sent to all EBMT centers asking for observed, proven or suspected cases of DCL. Centers reporting one or several cases will be asked to complete a second, more detailed questionnaire. The aim of the study is to identify and analyze a suffi cient number of cases to answer the following questions: How is the incidence of DCL evolving? Is development of DCL associated with viral status before and viral reactivation/infection during as well as after transplantation? Is there any infl uence of graft source and donor type (bone marrow vs. peripheral blood stem cells vs. cord blood and related vs. unrelated)? Is the risk of malignant transformation of grafted cells increasing with donor age? Is there any association with the form and extent of pretreatment (radiotherapy and chemotherapy) implicating a role of damaged microenvironment in the development of DCL? And fi nally, is DCL also observed as a very late complication of hematopoietic stem cell transplantation? Conclusion: The answers to these questions will help to further characterize donor cell derived leukemia and provide new insights into leukemogenesis in general.

A. Crotta, A. Tichelli, A. Ruggeri, I. Ionescu, A.L. Herr, K. Boudjedir, E. Gluckman, V 

Unrelated cord blood transplant (UCBT) is associated with a reduced incidence of chronic GvHD when compared to other sources of stem cells, however risk factors analysis for incidence is scarce in the literature. We retrospectively analyzed 1257 patients (pts), 755 children (age≤18) and 502 adults, receiving fi rst single (n = 1080) or double UCBT (n = 177) in EBMT centers, between 1995 and 2009 , for malignant and non-malignant diseases, who survived at least 100 days from transplantation with neutrophils recovery and without relapse or autologous reconstitution. Median age was 12 years (0. , most common disease was acute leukemia (60%). 11% of units were HLA-identical (antigen level for HLA-A and B, allelic for DRB1), while 41% and 48% had 1 or 2-3 mismatches, respectively. Median TNC infused was 3.7x10 7 /kg. Conditioning regimen was myeloablative (MAC) in 75% of cases (TBI>6Gy in 30%), RIC in 25%. Busulphan-based conditioning was used in 41% and ATG was added in 75% of cases. CSA + steroid was the most common GvHD prophylaxis (50%); in children prednisone was used in 70% of pts, mycophenolate mofetil was used in combination in 52% of adults. Median follow-up was 28 months (3-157). Incidence of cGvHD at 2y was 29±1% and 45±2% in children and adults respectively (p<0.001). Due to statistical difference between children and adults, risk factors analyses were performed separately. GvHD was extensive in 42% of children and in 52% of adults. Out of 412 pts who developed cGvHD, 219 had previously aGvHD (126 out of 212 children and 93 out of 200 adults). In univariate analysis in children, factors associated with increased incidence of GvHD were: age (>5y), advanced status of disease at transplant (>CR2) and number of HLA disparities; cGvHD was 25±2% and 36±4% (p = 0.03) in 6-5/6 cord blood unit and 4-3/6 respectively. In multivariate analysis only HLA disparities (HR = 1.95, p<0.001) and status of disease at transplant (HR = 1.74, p = 0.006) were associated with increased incidence of GvHD. In adults, higher incidence of cGvHD was associated with male sex, advanced disease at UCBT and use of MAC regimen. In multivariate analysis only the status at transplant was independently associated with cGVHD (HR = 1.46, p = 0.03).

In conclusion, incidence of cGvHD was lower in children than in adults. Status of disease at transplant was associated with increased incidence of GvHD in children and adults. Number of HLA disparities was associated with increased cGVHD only in children.

In MS, progression free survival has been reported to range from 50 to 70% at 5 years after HSCT. The most frequent conditioning regimen adopted in Europe is the association of BEAM and ATG, an intermediate-intensity scheme able to completely suppress MRI activity for at least 2 years in most of patients. Mortality has dropped in the last years from a relevant 7,3% to 1.2%, due to a better transplant management and patients selection. In rapidly progressive SSc 5-years mortality is estimated to be 40-50% [11] . The most commonly used conditioning regimen in Europe is Cyclophosphamide (200 mg/Kg) and ATG. In these patients, progression free survival after HSCT has been showed to be above 50% at 5 years. Furthermore, the use of high intensity conditioning regimens does not seem to offer special advantages over lower intensity schedules for this disease. The distribution of transplants per year shows a drop of activity in SLE and especially in infl ammatory arthritis, following the introduction of new biological agents after 2000. The emergence of pharmacological resistance in long-term treatments may suggest a possible renewal of the activity also in this subset of patients. In chronic infl ammatory bowel diseases (IBD) an increasing activity has also been shown in the last 6 years. HSCT is able to provide a high rate of long-term, immunosuppression-free remissions in AD patients resistant to conventional therapies. Appropriate selection of patients is crucial for providing the best risk-benefi t equipoise. Evidence of effectiveness and careful assessment of toxicity is expected from ongoing and future prospective clinical trials.

Data reported to the EBMT Registry at December 2009 102 European activity in multiple sclerosis R. Martin (1) , G. Mancardi (2) 

Intensive immunosuppression followed by autologous transplantation of hematopoietic stem cells (AHSCT) has been investigated as a possible strategy for the treatment of severe autoimmune disorders, including multiple sclerosis (MS) . It has been demonstrated that AHSCT leads to the elimination of self reactive T cells and complete or almost complete renewal of the immune repertoire. Immune system renewal has been demonstrated at the level of the T cell receptor repertoire in CD4 + and CD8 + T cells and in subsets of NK-and T cell population. According to the database of the European Blood and Transplantation Group (EBMT) registry, more than 400 MS cases have been treated with AHSCT. The retrospective survey, carried out in 2006, of all the patients recorded in the registry, with a median follow up of more than 3 years, showed that improvement or stabilization of neurological conditions occurred in 63% of cases. Transplant related mortality was very high (7,3%) in the 1995-2000 period, but subsequently dropped to 1.3% in the 2001-2007 years. The results appear to be particularly impressive in rapidly evolving severe MS cases unresponsive to conventional therapies. In these cases, defi ned also as malignant forms of MS, numerous European groups convincingly demonstrated the capacity of AHSCT to suppress the progression of the disease, with an unexpected relevant improvement of the neurological conditions lasting for a long period of time. The establishment of AHSCT as a rescue therapy in highly active MS cases unresponsive to other conventional therapies represents a very important step in MS clinical management, and the above mechanistic studies strongly support the rationale of this approach. Despite these advances, the acceptance of AHSCT as a rational treatment escalation remains low among neurologists, and this is probably one of the main reasons why the phase 2 study (ASTIMS), comparing AHSCT vs. Mitoxantrone with a MRI primary endpoint, had to be terminated due to recruitment diffi culties. Based on this experience, a new phase IIb/III study is currently being designed. This study will compare AHSCT with best available and approved therapy in the group of highly active relapsing-remitting MS patients, who failed fi rst-line therapy and one escalation step. The main goals of this trial will be to establish effi cacy in a controlled trial and to address important mechanistic questions. The status of the current discussion toward the trial design as well as strategies to recruit interested centers with MS-specialized neurologists and experienced transplant physicians at one location and potential mechanisms of funding will be presented.

Systemic sclerosis is a potentially life-threatening chronic autoimmune disease characterized by skin thickening, vasculopathy, and visceral involvement, mainly of lungs, gut, heart and kidneys. The clinical manifestations are diverse and refl ect a spectrum of subsets, ranging from limited to diffuse disease. Diffuse cutaneous systemic sclerosis (DcSSc) generally runs a more aggressive disease course, requiring intensive immunosuppressive therapy. Haematopoietic stem cell transplantation (HSCT) has resulted in long-term improvements of skin thickening, stabilization of organ involvement, but at the expense of mainly cardiopulmonary toxicity which has resulted in death in 5-8% of cases according to a recent analysis by the EBMT Working Party Autoimmune Diseases. To assess whether the benefi ts outweigh the risks, a prospective randomized controlled trial is nearing completion comparing HSCT with iv pulse cyclophosphamide in which 156 DcSSc patients have been enrolled in 27 centres. The primary endpoint is event-free survival, defi ned as survival until death or development of major organ failure during 2 years follow-up. Formal analyses of safety and effi cacy will be done in 2011. Interim safety analyses have lead to changes in the transplant protocol and eligibility criteria to also target patients with very early disease. The rationale of the trial is that effective intervention in very early disease is needed to induce long-term remissions. All patients are followed-up for at least 7 years. In anticipation of the results from the ASTIS trial, a new smallscale clinical trial will be done to comprehensively investigate the mechanism of action of HSCT in SSc focusing on its effects on biomarkers of fi brosis, autoimmunity, infl ammation and vasculopathy. It is hoped that the results of the ASTIS trial and those of the North American SCOT trial will provide solid data to decide on the design of a subsequent randomized trial. If ASTIS and SCOT indeed prove that HSCT is superior over conventional chemotherapy, then one option would be to modify the trans-plant regimen to reduce the risk of cardiopulmonary toxicity and/or enhance the intensity of the control arm.

Crohn's disease (CD) is characterized by chronic infl ammation in segments of the digestive tract leading to tissue damages. Uncontrolled infl ammation is associated with excessive immune responses towards the microbiota. Progress has been recently made in the management of CD, with increased use of immunosuppression and biologic therapies. Early optimized use of these treatments in patients with poor prognosis may provide a satisfactory control of the disease in the majority of patients. However, a fraction of CD patients experiences severe disease refractory to medical management, which may result in progressive tissue damages, need for surgery and chronic disability. The fi rst evidence of effectiveness of hematopoietic stem cell transplantation (HSCT) was reported with the observation that patients with CD, who underwent allogenic or autologous HSCT for a haematological or solid malignancy, experienced long-lasting remission of their IBD. Although few cases regarding autologous HSCT in concomitant CD and malignant diseases have been reported, collected data have shown similar results to allogeneic transplantation in terms of remission. Beyond case reports, autologous HSCT as primary treatment for CD has been investigated in two single-centre phase II studies. Despite the low number of patients and their limited follow-up, autologous HSCT was shown to have the capacity to induce complete clinical and endoscopic remission, with impressive results. The ASTIC trial, designed to assess its effi cacy and tolerability, is ongoing. Included patients undergo mobilization of stem cells, and are then randomized to transplantation after one month or after one year. Autologous HSCT may represent a therapeutic option in CD patients with refractory disease, and could change the natural history of the disease, inducing in some cases long term remission.

Haematopoietic cell transplantation for autoimmune diseases: EBMT/CIBMTR collaboration M. Pasquini (Milwaukee, US), R. Saccardi (Florence, IT) High dose immunosuppressive therapy with autologous haematopoietic cell rescue has the potential to halt the autoimmune process and result in a medication-free period. Since 1996, more than 1000 patients in Europe and 400 patients in North and South America who underwent transplantation for autoimmune disease (AID) have been registered in the EBMT and CIBMTR databases. Despite this activity, questions related to the timing of transplantation and what AID benefi ts more from transplantation remain unanswered. The development of a collaborative strategy has the potential to answer some of these questions and assist the development of prospective studies that can advance the fi eld. Following this strategy the CIBMTR and EBMT autoimmune disease committees joined forces in 2007 to expand the study portfolio and establish an infrastructure to facilitate future studies. The most common AID indication for transplantation is multiple sclerosis (MS). Data collection from both registries was harmonized into a single disease-specifi c form developed by a group of investigators affi liated to both committees. Collaborative project to focus on long term outcomes after transplant for MS was developed, especially to determine factors associated with prolonged disease-free intervals. Two in person discussions on the topic leveraged the development of an international phase III clinical trial with the objective to compare transplant with best medical treatment in a population with rapidly progressive MS. The next collaborative initiative is to focus on the second most common indication, systemic sclerosis (SSC). We plan to start harmonizing disease-specifi c forms and a collaborative study focused in this disease. Lastly, as the majority of patients present to transplantation with advanced AID, studying patients with coexistent AID who present for allogeneic transplantation, represent an opportunity to understand the graft-versus-autoimmunity effect. Thus, the CIBMTR is planning to identify these patients at time of registration and subsequently enrolled them on a prospect cohort study to evaluate the impact of graft-versus-host disease on AID. Outcomes database on transplant for AID are currently underutilized due to the heteroneity of diseases, scarcity of comprehensive level disease-specifi c information and loose integration with non transplant specialist. Bringing the EBMT and CIBMTR together in the last 3 years, helped engage investigators, including neurologists and rheumatologist to develop collaborative studies.

An update on autologous stem cell transplantation in severe systemic lupus erythematosus and in early diabetes type 1 D. Farge, R. Cervera, D. Jayne, A. Voskuyl, J.M. van Laar, A. Doria, M. Mosca, D. Boumpas, R. Saccardi, E. Snarski, J.F 

Objectives: To present the experimental and clinical rational for the ongoing EBMT or NIH trials for AHSCT in severe SLE and in early diabetes type I. Background: Since 1996, Autologous Hematopoietic Stem Cell Transplantation (AHSCT) has been used successfully in severe systemic lupus erythematosus (SLE) and early Insulin type I dependent Diabetes (T1D) to induce durable remission. In SLE BILAG A, the NIH or Eurolupus cyclophosphamide (CY) protocols and mycophenolate Mofetil (MMF) for induction are associated with 20% failure, 50% relapse and 10-15 % death at 10 yrs. Among 85 SLE out of 900 EBMT pts with fi rst AHSCT in 2009, 3 years PFS was 54 % (95% CI: 42%-66%) with 87% (95% CI: 79%-95%) overall survival (Farge D Haemetologica 2009 ). Higher remission rates and also some relapses, were shown in the North American experience (Burt R, JAMA, 2006) with the possibility of resetting the autoimmune response and inducing tolerance in SLE after HSCT and then in new onset Type I diabetes (Voltarelli and Burt et al, JAMA, 2007) . In susceptible strains of mice, allogeneic HSCT prevents insulitis and development of type 1 DM. In 2008, follow-up of the 23 AHSCT for T1D with in Brasil, showed that 14/21 remained insulin free (and normal HbA1C) with longest insulin free follow-up beyond 4 yrs and no correlation with pre-HSCT C-peptide, but AHSCT must be performed within 3 months of diagnosis. Eight patients with T1D (<6 weeks from diagnosis, C-peptide positive, Anti GAD -antibodies positive) treated in Poland using 200mg/CY conditioning and AHSCT plus antithymocyte globulin (ATG Genzyme) had no major complications after 6 (2-15) mths median follow-up and 8/8 pts became independent from exogenous insulin within 24 ( + 6-+ 60) days after AHSCT. Acarbose may have positive effects on the duration of remission of T1D after AHCT. Planned Studies: These were the basis for 1) the multicenter, EBMT approved ASTIL phase II study, to analyse the effi cacy of AHSCT followed by MMF as maintenance for severe SLE 5 yrs since diagnosis with sustained or relapsed active BILAG A SLE after at least 6 months of best standard local therapy among 30 patients. After CY 4 g/m 2 mobilisation, unselected AHSCT will use CY (200 mg/kg body wt in 4 daily doses) plus rabbit ATG and maintenance by MMF (2 g/day). The primary effi cacy endpoint will be the proportion of patients alive who achieve clinical success at 12 months and maintenance of this status until 24 months after inclusion; 2) the NIH approved phase I/II study of AHSCT in 15 newly-diagnosed T1D pts within 3 months of diagnosis, and a positive antibody to an islet cell autoantigen and fasting C-peptide > 0.2 nmol/l. to compare CY 4 g/ m 2 mobilization, unselected AHSCT using CY (200 mg/kg body wt in 4 daily Introduction: Graft versus Leukemia (GvL) effect of Hematopoietic Stem Cell Transplantation (HSCT) is mostly based on donor T cell-mediated alloreactivity. This is particularly relevant in the context of HSCT from family haploidentical and matched unrelated donors (MUD), in which mismatched HLA molecules are potent targets for donor T cells. Still, upon in vivo selective pressure by donor T cells, leukemia can undergo genomic rearrangements which result in loss of the patient-specifi c HLA, a mechanism which our group recently demonstrated to be frequently responsible for leukemia relapse after haploidentical HSCT (Vago et al., NEJM, 2009 ). Methods: 113 consecutive patients who underwent a partially HLA-matched transplantation for myeloid malignancies from 2002 to present, were included in our analysis. For 76 patients the donor was family haploidentical, and for 37 was unrelated and mismatched for a median of 2/12 HLA alleles. All patients received donor T cells as part of the HSCT protocol. Post-transplantation follow-up comprised monthly bone marrow examination, with Short Tandem Repeat (STR) chimerism analysis and HLA typing performed in parallel. In cases of relapse with suspected loss of the mismatched HLA, STR chimerism and HLA typing were performed also on purifi ed leukemic blasts. Results: Disease relapse occurred in 31 and 9 patients after haploidentical and MUD transplantation, respectively. After haploidentical transplantation, 11/31 relapses (35.5%) were due to mutated leukemic blasts which had lost the patientspecifi c HLA haplotype. Interestingly, 1 of the 9 relapses after MUD transplantation occurred through the same mechanism, in a patient who had received two subsequent transplants from HLA-C-mismatched donors. Upon detection of the mutated leukemic blasts, 9 of these 12 patients were enrolled to receive a subsequent HSCT from a different donor, mismatched for the remaining HLA haplotype. Conclusions: Our data consolidate the clinical relevance of this escape mechanism from the GvL effect mediated by alloreactive donor T cells. Screening for these mutants should be encouraged in patients who relapse after partially HLA-mismatched HSCT, to guide therapeutic strategies, avoid predictably ineffi cacious donor T cell add-backs, and quickly enroll the patients to a salvage transplant. To improve this approach, novel diagnostic and therapeutic tools are warranted, to provide earlier molecular detection and specifi c targeting of these mutants.

Impact of allogeneic stem cell transplantation on prognosis of patients with high-risk AML: results of the AML SHG 295 and 01/99 trials K. Wagner, G. Heil, M. Zucknick, D. Hoelzer, O.G. Ottmann, H. Martin, M. Lübbert, J. Finke, W. Heit, W. Fiedler, L. Kanz, G. Schlimok, H. Kirchner, A. Raghavachar, W. Brugger, A. Ganser, J 

Objective: Patients with high risk acute myeloid leukemia (AML) have a dismal prognosis after chemotherapy. Allogeneic stem cell transplantation (alloSCT) in fi rst complete remission (CR) is widely recommended for these patients but its impact on outcome is uncertain. Methods: In the prospective multicenter trials AML 295 and 01/99, we treated 825 adult patients with AML (except APL) up to 60 years. All patients received intensive double induction and early consolidation chemotherapy. High risk patients were defi ned by either bad response to induction I (persistent bone marrow blasts on day 15 after start of therapy) or high risk karyotype (all karyotypes other than normal or CBF-leukemias). All high risk patients with a matched related donor (MRD) were scheduled for an alloSCT as late consolidation. Patients without a family donor in the 295 trial should receive an autologous stem cell transplantation, whereas in the 01/99 trial an alloSCT from matched unrelated donors (MUD) was recommended. Results: Median follow up of the patients is 78 months. 393 of the patients fulfi lled the high risk criteria. Median overall survival (OS) of these patients is 17 months and the six-year OS is 30%. In multivariate analysis, a complex or monosomal karyotype (HR 1.82; 95% CI 1.36-2.43) and age above the median of 47 years (HR 1.53; 95% CI 1.18-1.98) were identifi ed as adverse prognostic factors for OS. 234 of the 393 high risk patients (59.5%) achieved a CR. Median relapse-free survival (RFS) is 12.8 months and six-year RFS is 33%. Leukocytes above the median (HR 1.65; 95% CI 1.18-2.29) and a complex/ monosomal karyotype (HR 1.61; 95% CI 1.06-2.44) were identifi ed as independent prognostic factors for RFS. Of the 234 patients who achieved CR, 97 received an alloSCT from MRD (n = 65) or MUD (n = 32) in fi rst CR. Median OS of the transplanted patients is not reached and six-year OS is 57%. Median RFS after alloSCT is 100 months and six-year RFS is 53%. Outcome did not differ between MRD and MUD alloSCT. The impact of alloSCT on prognosis was analyzed in a multivariate Cox-model with alloSCT included as a time-dependent covariable. In this analysis, alloSCT in fi rst CR signifi cantly improved OS (HR 0.48; 95% CI 0.34-0.69) and RFS (HR 0.48; 95% CI 0.33-0.69). Conclusion: AlloSCT from matched related or unrelated donors in fi rst CR improves overall and relapse-free survival of patients with high risk AML.

Cord blood transplantation for adults with acute lymphoblastic leukaemia -a survey of outcomes conducted by Eurocord and the Acute Leukaemia Working Party of the EBMT D. Purtill (1) Cord blood transplantation (CBT) is considered an option for adults with acute lymphoblastic leukaemia (ALL) for whom stem cell transplantation is indicated. However, little information is available regarding outcomes of this procedure in this group of patients. We conducted a retrospective review of the Eurocord database in order to describe outcomes of adults treated with single or double CBT for ALL at EBMT centres from 2000 until 2008. Two-hundred and thirty six patients with a median age of 30 years (18-62 years) were included. Most were transplanted in fi rst complete remission (CR1) (n = 101) or CR2 (n = 75). Median white cell count at diagnosis was 13.9x10 9 /L (0.6-624) and 56% (n = 56) of patients in CR1 had t(9;22) or t (4;11). Double cord blood transplantation was performed for 73 patients (31%). Overall median total nucleated cell dose at freezing was 3.9x10 7 /kg (1.45-8.81 ) and it was 4.5 (2.1-8.8 ) and 3.4x10 7 /kg (1.5-7 .0) respectively for double and single transplants. Most patients received cord units with one (n = 65) or two (n = 133) HLA disparities. Overall cumulative incidence (CI) of neutrophil recovery was 74±3% and acute graft-vs-host disease was 33±3%. Transplant-related mortality (TRM) was 41±3% and 35±3% for patients in CR1 and CR2 respectively, and 63±3% for the 60 remaining patients with more advanced disease (CR3, relapsed or refractory disease). CI of relapse were 20±6%, 27±8% and 33±11% respectively for CR1, CR2 and advanced disease. At the median follow-up of 2 years, leukaemia-free survival (LFS) was 30±3% for the whole population and 39±9% for CR1, 38±6% for CR2 and 4±3% for advanced disease. LFS for patients in CR1 with t(9;22) or t(4;11) was 39±7% and for double CBT it was 38±7%. On multivariate analysis, being in CR1 or CR2 at transplant was the only factor associated with improved LFS (HR 0.25, 95%CI 0.08-0.42, p<0.001). When this group was analysed separately, the use of a reduced intensity conditioning (RIC) regimen was associated with improved LFS (57±8% vs. 32±5%, HR = 0.53, 95%CI 0.27-0.79, p = 0.01). In particular, the 27 patients who received the combination of cyclophosphamide, fl udarabine and TBI (2-6 Gy) had an LFS of 64±10%. Cord blood is an alternative source of stem cells for allogeneic transplantation for adults with high risk ALL. Results with RIC and double CBT are encouraging and should be further investigated in a larger series of patients.

Superior long-term survival with a high rate of allogeneic stem cell transplantation in AML (non-APL) patients below 60 years of age G. Juliusson (1) Introdution: Allogeneic stem cell transplantation (alloSCT) prevents relapse in AML, but the indication in intermediate risk is unclear and depends on the transplant-related mortality. The Swedish Acute Leukemia Registry (Blood 2009; 113:4179) includes 3878 patients (pts) with acute leukemia diagnosed 1997-2006 covering 98% of all cases in Sweden, with 9 million inhabitants. Using this population-based registry, we evaluated the SCT rates in AML, and the long-term outcome. Allo-SCT rates in different Swedish regions were also evaluated. Transplantation data was individually validated in 2008, and survival was updated through the national population registry. Patients and results: Of 797 adult AML (non-APL) pts <60 yrs (median 51 yrs), 304 (38%) had an alloSCT (23% with secondary AML) in CR1 (n = 206; 26%), or at later stages (n = 92; 12%), with decreasing rates by age ( Figure 1 ). Donors were unrelated in 52%. Median follow-up was 6.2 yrs, and 170 (56%) of allografted pts were alive. Treatment-related mortality (TRM, i.e., death in CR), and AML deaths according to age, donor, and status at alloSCT is shown in Figure 2 . Estimated 5-yr and 10-yr OS was 41% and 35%, respectively. These results could be compared to recently published large AML studies from ECOG, EORTC-GIMEMA, UK MRC and GAMLCG, with alloSCT rates of 5-15% and 5-yr OS of 18-38%.

The health care region SE, which had the highest alloSCT rate (61% of AML pts <60 yrs) had a 5-yr OS of 52% for the total AML population, as compared to the other Swedish regions with a mean alloSCT rate of 35% and a 5-yr OS of 40% (log rank, p = 0.04). The difference in alloSCT rate was mainly due to more allografts among older pts: in the age cohort 50-59 yrs, 42% of CR1 pts underwent alloSCT yrs in region SE as compared to 15% in the rest of Sweden, associated with a more pronounced survival benefi t (p = 0.007). Conclusions: This is a real world population-based study on alloSCT rates, treatment-related mortality and long-term survival in AML pts. Our results indicate that alloSCT is performed in Sweden with relatively low total non-relapse mortality, and with a moderate relapse rate. The association found between a high alloSCT rate in pts 50-59 yrs and a better survival is intriguing, is suggestive of a positive impact of a more frequent transplantation practice in this age cohort, and supports the need for prospective studies.

Role of the graft-versus-leukaemia effect after reducedintensity conditioning allogeneic stem cell transplantation as treatment for acute myeloid leukaemia: a survey from the Acute Leukaemia Working Party of the EBMT F. Baron, M. Labopin, M. Mohty, N. Basara, D. Niederwieser, N. Milpied, J.J. Cornelissen, C. Malm, L. Vindelov, D. Blaise, J.J.W.M. Janssen, E. Petersen, G. Socie, V 

Previous studies have observed a lower risk of relapse in patients (pts) who experienced chronic GVHD after RIC allo-SCT versus in those who did not. The objective of the current study was to further investigate the association between chronic GVHD and relapse in a large cohort of pts given RIC allo-SCT as treatment for AML. Data from 1188 AML pts in fi rst or second CR transplanted between 2000 and 2008 following a RIC regimen at EBMT affi liated centers were analyzed. Patients were given PBSC from HLA-identical sibling (MRD, n = 879), or from HLA-matched unrelated donors (MUD, n = 309). Median pt age at transplantation was 55 yrs in pts given grafts from MRD, versus 57 yrs in those given grafts from MUD. The impact of chronic GVHD on relapse risk, non-relapse mortality (NRM) and leukemia-free survival (LFS) was assessed by time-dependent multivariate Cox models and in a landmark analysis. Three-yr incidences of relapse, NRM and LFS were 35 ± 2%, 14 ± 2%, and 50 ± 2%, respectively, while 2-yr incidence of chronic GVHD was 49 ± 2%. In a landmark analysis at 18 months after allo-SCT, 5-year relapse rates were 10 ± 2% versus 19 ± 3% for patients with or without chronic GVHD (P = 0.04), respectively. In multivariate Cox models, CR2 versus CR1 (P = .003), pt age >55 yrs (P = .008), alemtuzumab use in the RIC (P = .048), TBI-based RIC (P = .006), high-risk cytogenetics (P = .001), and absence of chronic GVHD (P = .015) were each associated with higher risk of relapse. Factors associated with high NRM were MUD versus MRD (P = .003), grade II-IV acute GVHD (P<.001), and chronic GVHD (P = .002). Factors associated with lower LFS were CR2 versus CR1 (P = .003), pt age >55 yrs (P = .007), alemtuzumab use in the RIC (P = .012), and high-risk cytogenetics (P = .003). In conclusion, in this cohort of AML patients transplanted in remission, chronic GVHD was associated with a lower risk of relapse while profound in-vivo T cell depletion with alemtuzumab was associated with higher relapse rate suggesting that GVL effects play a role in preventing AML relapse in patients given RIC allo-SCT. In spite of the presence of GVL effect, chronic CHVD was associated with higher NRM and therefore there was no net impact on LFS. Strategies to decrease NRM related to chronic GVHD should be further investigated in order to keep the benefi ts from GVL effect and improve LFS. Finally, the impact of chronic GVHD duration and severity on LFS need to be studied.

Allogeneic stem cell transplantation in acute myeloid leukaemia with normal karyotype: a risk factor analysis in 247 patients, based on molecular markers and stage at transplantation T. Pfeiffer (1) , M. Schleuning (2) Cytogenetically normal acute myeloid leukaemia (CN-AML) represents a heterogenous disease, that can be subdivided by molecular analysis. Only limited data is available on SCT in different molecular subgroups, particularly in advanced stage. Therefore, we retrospectively analyzed data on 247 pts. with CN-AML, who uniformly had received the FLAMSA-RIC regimen for SCT in 14 European centres. Pts. suffered from de novo AML (76%), sAML/MDS (21%), and tAML (4%). Median age was 52y (19-71). Donors were matched or mm family, and matched or mm unrelated donors in 30%, 2%, 50% and 18%, respectively. SCT was performed in untreated disease (6%), primary induction failure (PIF,23%), 1st complete remission (CR1,14%), and >CR1 (57%). Median follow-up of survivors was 19 mo. Overall survival (OS) and leukaemia-free-survival (LFS) at 2y from SCT was 51% and 47%. The stage at SCT was the most important factor (p = .001 for OS, <.001 for LFS). Results were promising after SCT in CR1 (2y OS/LFS 76%), and PIF (2y OS/LFS 69%), but were inferior after SCT in untreated disease (2y OS/LFS 34%), or >CR1 (2y OS 42%, LFS 34%). Information on molecular markers was available in 183 pts. (74%). As suggested (Schlenk, NEJM 2008) , analysis was based on 2 subgroups: 22 pts. with isolated NPM1 mutation (NPM1mut), and 161 pts. with other genotypes (FLT3-ITD, n = 66; or wildtype FLT3/wildtype NMP1 [FLT3wt/NPM1wt], n = 95). Pts. with NPM1mut had a 4y OS/LFS of 75/63% with identical outcome, when transplanted in PIF, CR1, or >CR1. Pts. with other genotypes showed an OS/LFS of 51%/48% at 2y and of 40%/39% at 4y, without differences among pts. with FLT-ITD and FLT3wt/NPM1wt. However, in this subgroup, outcome was highly dependent on the stage at SCT, with excellent results after SCT in CR1 (2y OS/LFS: 76%) or PIF (2y OS/LFS: 75%/74%), but inferior outcome after SCT >CR1 (2y OS/LFS 38%/33%; p = .004 for OS, .001 for LFS). Conclusion: AlloSCT following FLAMSA-RIC produces excellent survival in pts. with CN-AML, particularly when performed in CR1. Encouraging results in PIF support an early SCT, regardless of molecular subgroup, when CR is not reached after double induction. In NPM1mut, SCT in advanced disease achieved identical results as in early stage, supporting the delay of SCT until relapse has occurred. In contrast, pts. with FLT3-ITD or FLT3wt/NPM1wt achieved signifi cantly worse results when transplanted >CR1, arguing in favour of SCT in CR1 for this molecular subgroup.

Comparison of outcomes after allogeneic SCT for adult patients with AML in remission using either I.V. busulfan plus cyclophosphamide or TBI plus cy in the myeloablative conditioning regimen: an-ALWP-EBMT survey A. Nagler (1) , M. Labopin (2) , A. Shimoni (1) TBI/CY and oral Bu/Cy are the most common myeloablative regimens for adults with AML with comparable survival and relapse probabilities. Intravenous (i.v.) Bu in contrast to oral Bu has more predictable pharmacokinetics and a more favorable toxicity profi le. Outcomes with i.v. Bu/Cy, thus, may be superior to those achieved with TBI/Cy. In order to address these issues, the ALWP of the EBMT performed a survey comparing i.v. Bu/Cy to TBI/Cy as conditioning regimen for adult pts. with AML undergoing alloSCT. Overall, 1479 alloSCT were analyzed: Bu/Cy -332, TBI/Cy -1147. Median age was 40 (range, 18-60) and 39 (range, 18-65) years and median transplant year was 2007 (2000 -2007) and 2004 (2000 -2007) for the Bu/Cy and TBI/Cy groups, respectively (P<0.0001). Disease status at alloSCT was CR1 -81% vs. 80% and CR2-19% vs. 20% for the Bu/Cy and TBI/Cy groups, respectively (NS). WBC count at diagnosis was 15x109/L in both groups. Cytogenetic: good -6% and 8%, intermediate -26% and 42% and poor risk -6% and 6%, respectively (NA -44%-62% of pts). 76% and 82% of both groups underwent alloSCT from sibling donors, while 24% and 22% from MUD, respectively. Follow up was 15(0.5-96) and 41 mo. for both groups, respectively. 75% and 47% of the Bu/Cy and TBI/Cy groups received PBSC, while 25% and 53% received BM grafts, respectively (p<0.0001). Engraftment was similar in the Bu/Cy vs. the TBI/Cy groups, 97% and 97%, respectively. Cumulative incidence of NRM at 2y was 16 + 2% and 18±1%, respectively (p = 0.46). Acute GVHD both >II-IV and III-IV were signifi cantly lower in the Bu/Cy vs. the TBI/ Cy groups: 21% and 32%and 8.5% and 19.5%, respectively (P<0.0001). While, VOD was signifi cantly higher in the Bu/Cy vs. the TBI/Cy groups, 7.8% vs. 1.7%, respectively (P<0.001). Two year relapse incidence was comparable, 25 + 3% and 21 + 1% for Bu/Cy vs. TBI/Cy, respectively (p = 0.73). LFS was comparable, as well, 59 + 3% and 61 + 2%, respectively (p = 0.73). In multivariate analysis poor prognostic factors for LFS were: poor cytogenetics (P≤10-4), disease status (CR2 vs. CR1) (P≤10-4), age >40y (P = 0.001), MUD vs. Sib. donor (P = 0.037) and PB vs. BM grafts (P = 0.042). In conclusion, in this retrospective EBMT survey, outcomes of alloSCT including engraftment, TRM, RR and LFS was comparable, while acute GVHD probability was signifi cantly lower and VOD probability signifi cantly higher in adult pts. with AML in CR using myeloablative i.v Bu/Cy vs. TBI/Cy conditioning regiments, respectively.

Higher incidence of relapse with higher doses of CD34 + cells from leukapheresis products infused in adults with acute myelocytic leukaemia autografted during the fi rst remission N.C. Gorin, M.M. Labopin, D. Blaise, F. Witz, T. de Witte, G. Meloni, M. Attal, D. Carreras, V. Rocha on behalf of the ALWP Purpose: The cell source for autologous stem cell transplantation has shifted from bone marrow (BM) to peripheral blood (PB). For patients with AML in CR1, we previously showed that the risk of relapse was greater with PB than BM and a poorer outcome was associated with a shorter interval from CR1 to PB transplantation (≤80 days) (JCO 2009 in press). Leukemic and normal progenitors bear the CD34 + antigen and can be mobilized together; we questioned whether there was a relation linking the doses of CD34 + cells infused to the outcome. Methods: Out of 1262 patients autografted with PB more than 80 days post CR1 and reported to the EBMT registry using MEDB form, the dose of CD34 + cells infused was available in 772. Results: The CD34 + cell doses were categorized by percentiles to divide the whole group into fi ve categories containing the same number of patients. We identifi ed the fi fth percentile (>7.16 × 10 6 /kg) as the cut off point for Relapse Incidence (RI) and Leukemia-Free survival (LFS). Patients receiving the highest dose had a higher RI (57 ± 4%, vs. 44 ± 2%; p = 0.008) and a lower LFS (34±4% vs. 49 ± 2%; p = 0.007). In a multivariate analysis adjusted for differences, RI was higher in patients receiving the highest CD34 + cell dose ((hazard ratio [HR], 1.48; 95% CI, 1.12-1.95; p = 0.005).and the LFS was worse (HR, 0.72; 95% CI, 0.55-0.93; p = 0.01). Conclusion: For patients with AML in CR1 autografted with PB, risk of relapse is greater and LFS is lower in those receiving the highest doses of CD34 + cells.

Oral Session 2: Early side effects O122 DNA transfer from donor to host cells as a mechanism for epithelial chimerism and genomic alterations after allogeneic haematopoietic cell transplantation M. Waterhouse (1) , M. Themeli (2) , H. Bertz (1) , N. Zoumbos (2) , J. Finke (1) , A. Spyridonidis (2) (1)Albert Ludwigs University of Freiburg (Freiburg, DE) ; (2)University Hospital Patras (Patras, GR) Research in the fi eld of allogeneic hematopoietic cell transplantation (allo-HCT) revealed hidden consequences of the co-existence of two genetically distinct populations in the transplant recipient. First, epithelial cells with donor-derived genotype emerge and second, epithelial tissues of the host acquire genomic alterations. We asked whether horizontal transfer of donor-DNA to host epithelium is operating in (and linking) both phenomena. 176 buccal samples were obtained from 71 allotransplanted patients and analyzed with microsatellite markers for the presence of donor-DNA and microsatellite instability (MSI). The presence of graft-derived hematopoietic cells in the samples was excluded by immunocytochemistry. The results were associated with clinical data. Genomic instability induction and DNA transfer in epithelial cells by allogeneic lymphocytes was assessed in vitro. We found a high contribution of donor-DNA (mean 26.6%) in buccal samples in 61 out of 68 evaluable patients. In addition, 32% of the samples were positive for MSI (MSI + ). The extent of donor-DNA was signifi cantly correlated with the occurrence of genomic alterations (p<0.05). The age of the patient and a female-to-male transplantation were signifi cantly correlated with MSI. There was a trend for increasing risk of MSI for patients who experienced severe graft-versushost disease. During follow up secondary malignancy was diagnosed in 5 patients (14%) who exhibited MSI but only in 1 (3%) with no MSI. By applying a time-dependent statistical analysis we found that the probability of secondary tumor development was signifi cant higher in the MSI + as compared to the MSIpatients (p = 0.024, hazard ratio 6.68, 95% . In an in vitro model, we demonstrated that apoptotic lymphocytes may not only induce genomic alterations but also transfer DNA through a phagocytotic process to co-cultured epithelial cells. The ingested lymphocytic DNA was also incorporated into the nucleus and integrated into the genome of the recipient epithelial cells, resulting in the creation of hybrid chromosomes. Our results indicate that continuous charge of the transplant recipient with apoptotic donor-DNA and its illegitimate integration into host epithelium may come in light as epithelial cells with donorderived genome or genomic instability events and may provide a new model for elucidating protean clinical manifestations after allo-HCT, such as secondary malignancy.

The haematopoietic cell transplantation-specifi c Co-morbidity Index and non-relapse mortality after reduced intensity conditioning: fi nal analysis from the ALWP of EBMT in AML patients in fi rst complete remission M. Mohty, M. Labopin, N. Basara, J.J. Cornelissen, R. Tabrizi, C. Malm, J. Perez-Simon, A. Nagler, N. Kröger, B. Rio, R. Martino, M. Eder, K. Bilger, D. Bunjes, G. Socié, D. Blaise, E. Polge, V 

The current study was designed to test the performance of the adapted Charlson comorbidities index (CI; so-called "Sorror index") and its association with outcomes among a cohort of 345 patients (i) aged >50 y.; (ii) diagnosed with a single disease entity, AML in CR1, and (iii) who underwent a RIC allo-SCT reported to the EBMT registry between 1999 and 2006. In this series, the median year of allo-SCT was 2004, and the median age was 58 y. (range, 50-76) . 32% of patients needed more than one induction course to achieve CR1. A fl udarabinebased RIC regimen was used in 64% of patients, while 31% of patients received low-dose TBI as part of their RIC, and 6% received other non-specifi ed RIC regimens. 76% of the patients received allo-SCT from an HLA-matched sibling donor. Based on score calculated with hazard ratios (HR) estimated on the population studied, 161 patients (47%) had a CI score of 0, 96 patients (29%) had a score of 1, and 49 (14%) had a CI score of 2. The remaining 39 patients (11%) had CI scores of 3 or more. In this cohort, 2 years overall and leukemia-free survival rates were 64±3% and 54±3% and the 2 years relapse and non-relapse mortality (NRM) cumulative incidences were 32±2%, and 15±2% respectively. The 2 years NRM incidences according to comorbidities score 0, 1, 2 and 3 + were 9±2%, 15±4%, 18±5% and 31±7%, respectively. In multivariate models (adjusted for recipient age, donor type, use of TBI or not, and cytogenetics risk group) comorbidities such as moderate active liver disease, obesity, prior history of renal dysfunction, and prior history of severe liver disease were associated with the highest HRs for 2 years NRM (varying from 2.11 to 2.76) and 2 years cumulative incidences of NRM varying from 22% to 44%, whereas previous solid tumor, diabetes, rheumatologic abnormalities, moderate pulmonary diseases, cardiac abnormalities (other than arrhythmia and valve disease) were associated with the lowest HRs (varying from 0.2 to 1.0) with 2 years cumulative incidences of NRM varying from 5 to 17%. Results from this large study performed in a single disease entity and homogeneous allo-SCT setting, suggest that the hematopoietic cell transplantation-specifi c CI is a simple, informative and useful tool for capturing pre-transplant comorbidities, and for predicting NRM after RIC allo-SCT for AML in CR1 in patients aged >50 y. Such index may be used for clinical trials and patient counselling before RIC allo-SCT.

Fibrin glue directly applied on damaged bladder mucosa during cystoscopy is highly effective to treat severe, refractory, haemorrhagic cystitis after allogeneic transplant M.C. Tirindelli (1) (1) Background: Hemorrhagic Cystitis (HC) occurring after hematopoietic stem cell transplant (HSCT) signifi cantly affects quality life of patients, prolongs hospitalization and in some cases can become a life-threatening complication. Its management has not been established. Fibrin Glue (FG) is a hemostatic agent derived from human plasma with proven effi cacy in repairing damaged tissues. Study design and methods: This study included patients who met the following criteria: grade ≥3 HC not responding to hyperhydration, bladder irrigation, antiviral treatments and transfusion support. FG was obtained using Vivostat system, an automatic method for processing and applying FG. During conventional cystoscopy and maintaining bladder distension by a CO2 insuffl ator, FG was accurately sprayed through a specifi c applicator on bleeding mucosa. FG polymerized on contact and set over several days. The response to the treatment was defi ned complete (CR) for disappearance of hematuria, partial (PR) for at least one grade regression of HC and no response (NR). Results: From Jan 06 to Oct 09, 626 patients undergoing an autologous (n = 428) or allogeneic (n = 198) HSCT were registered at the RTN. No autologous HSCT recipients developed HC of severe grade, whereas 18 of 198 patients (9%) undergoing an allogeneic HSCT met the criteria to enter the study. These 18 patients (6 M, 12 F) with a median age of 32.5 years (range, 18-53) had been submitted to a HSCT from HLA identical sib. (n = 4), unrelated CB (n = 4), MUD (n = 2) or related haploidentical donor (n = 8) for different hematological malignancies. All patients, deeply immunosuppressed with positive BKV viruria >7x10 6 copies/ml, developed a very severe HC, refractory to all current treatments including antiviral therapy. At time of FG application, HC persisted for a median of 16 days (range, 7-65) and was grade 3 and 4 in 14 and 4 patients, respectively. The number of FG applications was 1 in 15 patients, 2 in 2 and 3 in 1 patient for a median of 11 mL (range, 6.3-16.2) of glue. The treatment was successful in 16 out of 18 patients (89%). All 14 patients with grade 3 HC responded and the response was complete in 12 (86%) and partial in 2 (14%), while of the 4 patients with grade 4 HC: 1 achieved CR, 1 PR and 2 NR. No patient died of HC. Conclusions: FG therapy is a feasible, safe, easy repeatable, not invasive, small time consuming, lightly expensive and highly effective procedure in treating severe, refractory posttransplant HC.

Glutamine-enriched intravenous total parenteral nutrition doesn't improve mucositis and clinical outcome after stem cell transplantation for childhood malignancies. A prospective double-blind controlled study on behalf of AIEOP Group C. Uderzo (1) , E. Marrocco (1) , P. Rebora (1) , F. Cichello (1) , M. Bonetti (1) , S. Cesaro (2) , S. Varotto (2) , P. Verlato (2) Introduction: High dose chemo-radiotherapy followed by stem cell transplantation (SCT) constitutes the principal cause of post-transplant severe mucositis and related complications. Intravenous glutamine-enriched solution (GEs) has been advocated as one of the support treatment capable to improve marked body protein wasting, oxidative stress which result in a mucosal damage, often "primum movens" of infectious diseases beginning from G.I.tract damage. Patients and methods: 118 patients (79 males, median age at SCT 8.1 years) with haematological malignancies have been treated from June 2005 to June 2008 with high dose chemotherapy and/or TBI followed by allogeneic SCT (35 match related,65 match unrelated,18 mismatch.Patients were randomly and double-blinded assigned to undergo either for total parenteral nutrition (TPN) or for GEs TPN at the dose of 0,4 g/Kg/day from day of SCT until the end of TPN. The principal endpoint of the study was the mucositis assessment according to WHO criteria.The evaluation of the clinical,haematological and laboratory parameters served to perform the evaluation of secondary end-points.Statistical analysis was set up to demonstrate a 20% difference in terms of mucositis incidence and severity with a power of 80% (alfa = 0.05; two-sided tests). Results: Both study groups were comparable for age, gender, diagnosis and type of SCT. The GEs patients (60) were clinically similar to the controls at the entry; however, they didn't experienced a reduction of rate and grade of mucositis, with an odd ratio of 0.98(0.26-3.63) adjusted for the type of SCT. Neither the type of analgesic treatment nor the duration of opiod treatment, were statistically different. Days of both different TPN, length of hospital stay, TRM at day + 180,acute or chronic GVHD, incidence of severe infectious diseases, immunological recovery, progression of malignancies were similar in both groups. Nutritional status at the beginning and at the end of both types of TPN didn't show any difference at univariate analysis. No toxicity of GEs was observed. Conclusions: The current study is one of the few designed to demonstrate the utility of GEs on mucositis in the setting of children undergoing SCT for malignancies. GEs failed to infl uence incidence and severity of mucositis after SCT and didn't offer any advantage on clinical outcome, as claimed by other retrospective studies often performed in adult transplanted patients.

Clinical and genetic risk assessment for overall survival in haematopoietic stem cell transplantation A. Dickinson (1) Objectives: Non-HLA gene polymorphisms and clinical risk factors impact on outcome after HSCT. Age, stage of the disease, time from diagnosis to transplant, histocompatibility and donor and patient gender combination are key transplant risk factors for HSCT. We have assessed the impact of non HLA gene polymorphisms on the EBMT risk score for overall survival in a EUROBANK cohort of 915 HLA identical sibling and matched unrelated donor (MUD) transplants. Methods: HSCT patients with acute leukaemia (AML + ALL (AL)) (n = 463), CML (N = 187) plasma cell neoplasia (n = 120) or lymphoma (n = 145) and their donors from 8 transplant centres were genotyped for non-HLA polymorphisms within the tumour necrosis factor receptor II (TNFRSF1B), estrogen receptor (ESR1), vitamin D receptor (VDR), interleukin (IL) 6 (IL6), IL-1 receptor antagonist (ILRN), interferon gamma (IFNG) and IL 4 (IL4) genes. Genetic factors were assessed using the log rank test (p value < 0.2). Candidate factors were included in addition to the EBMT risk score in a stepwise Cox regression procedure to select fi nal genes. Predictive value of the model for overall survival was assessed through the concordance (c) index and prediction error curves. Results: In the whole cohort the presence of allele C in the donor IL-6 genotype (SNP IL6-17) was associated with lower survival time (especially in the CML sub group p = 0.025) and improved the prediction of the EBMT risk score. In the AL subgroup the absence of patient IL4 (any T) and presence of IL1RN (any C) (high risk group) were associated with lower survival time (compared to the remaining patients (low risk group)), and taken together also improved the predictive value of the EBMT risk score. Figures 1A and B show the lower probability of survival (and increased TRM) in the high risk versus low risk group. Conclusions: This study confi rms the importance of non-HLA genotyping for risk assessment in allogeneic HSCT. Improvement of fi t of the EBMT risk score presents a powerful novel tool to assess this impact of gene polymorphisms in a complex heterogeneous HSCT population. In a confi rmatory study the EBMT risk score was associated with survival for an AL patient group. Assessment of patient IL4 (any T) and IL1RN (any C) in this second cohort is in progress. Background: Cognitive impairments have been found among patients receiving chemotherapy and HSCT. However, studies with representative samples are rare, but needed to clarify the short and long term impact of different conditioning regimens and HSCT. In the present multicenter trial, we assessed the prevalence and course of cognitive dysfunctions in patients with hematological diseases undergoing allogeneic HSCT. Moreover, the role of neurotoxic intensity of conditioning regimens was investigated. Methods: 102 patients (61% male) with mixed hematological cancers (41% acute myeloid leukemia) at an average of 48 years of age were assessed before (t0), 100 days after (t1) and one year following (t2) allogeneic HSCT. 36% of the participants received intense neurotoxic conditioning regimens. A battery of neuropsychological tests using computer-and paperpencil-based methods was used covering the domains of attention, memory, executive and psycho-motor function. Measures assessing self-perceived cognitive impairments and psychological distress were additionally applied. Results: Compared to published test norms at each assessment time point, signifi cant impairments in several neuropsychological test parameters across all cognitive domains were found. At baseline, patients had impaired cognitive functions in 50% of the test parameters. No signifi cant change in the prevalence of cognitive impairments was observed over time except for a mild increase in psycho-motor dysfunction at t2. Patients who were classifi ed as having intense neurotoxic conditioning were more likely to show impairments in the domain attention at t1 (P<0.05, d = 0.53) and showed a different time course of performance in attentional tasks compared to patients with mild neurotoxic conditioning (P<0.01, eta² = 0.05). Conclusions: A remarkable amount of patients was classifi ed as having cognitive impairments prior to HSCT. Possible explanations include the impact of the hematological cancer disease, invasive treatments applied prior to conditioning and HSCT, and treatment related distress. A decline in cognitive functioning was primarily limited to psycho-motor functions. However, subgroups of patients and in particular those with intense neurotoxic conditioning regimens might be at greater risk for developing short term cognitive impairments particularly in the domain attention. Acknowledgement of Funding: This study was supported by a research grant from the German José Carreras Leukemia Foundation.

Clinical outcomes of second allogeneic haematopoietic stem cell transplantation for acute leukaemia and myelodysplastic syndrome relapsing after fi rst allogeneic transplantation T. Yamashita, T. Kikuchi, I. Kamiya, R. Hanajiri, Y. Nagata, S. Wakabayashi, K. Taoka, T. Kobayashi, K. Ohashi, H. Sakamaki, H. Akiyama Tokyo Metropolitan Komagome Hospital (Tokyo, JP) Background: Allogeneic hematopoietic stem cell transplantation (HCT) is an effective therapy in acute leukemia and myelodysplastic syndrome (MDS). But even after HCT, these diseases recur in some cases and prognosis of these patients is very poor. Second allogeneic HCT (HCT2) may be one of the most effective therapies in some patients who relapsed after fi rst allogeneic HCT (HCT1). Because of the heterogeneity of HCT2, it is often diffi cult to fi nd useful prognostic factors and appropriate strategies. In our single-center study, we retrospectively analyze the clinical data of 465 patients with acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL) and MDS who received HCT1 in our hospital to investigate the factors that affect clinical outcomes of HCT2. Patients and methods: We included patients with AML, ALL and MDS who received HCT1 between 1986 and 2008 in our hospital. Probability of overall survival (OAS) and event-free survival (EFS) were calculated using the Kaplan-Meier method. Cumulative incidence rates were calculated using standard methods for hematopoietic recovery, acute and chronic GVHD, relapse and non-relapse mortality (NRM). Multivariate analysis was performed with variables that can affect the clinical outcomes using Cox proportional-hazards regression models. Results: A total of 465 patients were eligible for this analysis. Five years OAS of HCT1 is 50.8%. 5 years cumulative incidence of relapse (RI) 30.5% and 138 patients relapsed after HCT1. Relapsed patients group was signifi cantly lower incidence of chronic GVHD (p = 0.008). In this patients group, HCT2 was performed for patients with AML (n = 19), ALL (n = 11), MDS (n = 4). Myeloablative conditioning was used in 28 cases. Five years OAS of HCT2 was 15.2%, and OAS and EFS of myeloabative HCT2 was 24.4% and 13.1%, respectively. Five years RI was 57,2% and TRM was 30.7%. Multivariate analysis showed that 1) relapse to HCT2 over 180 days, 2) age at HCT2 under 33 years, 3) CR at HCT2, 4) related donors are signifi cant favorable prognostic factors for OAS of HCT2. Conclusion: Our study shows that myeloablative HCT2 brought sustained remission in a proportion of patients with AML, ALL and MDS who relapsed after HCT1. Age and disease status at HCT2 are important factors that infl uence the clinical outcomes. This study also suggests that the duration from relapse after HCT1 and HCT2 have impact on regimen-related toxicity of HCT2.

Hepatic dysfunction is frequent and diverse before and after allo-RIC and has a major impact on the transplant outcomes P. Barba (1) Introduction: Hepatic dysfunction is one of the most frequent and less studied organ impairments in the setting of allogeneic stem cell transplantation (allo). To analyze the impact of pretransplant liver function on the outcome of allo-reduced intensity conditioning (allo-RIC) and to determine the incidence, characteristics and risk factors of hepatic injury after allo-RIC we conducted a retrospective study in two Spanish centers. Patients and methods: We analyzed 281 adult patients receiving an allo-RIC. The median follow-up for survivors was 5 years (rg 0.2-10). RIC consisted in fl udarabine 150 mg/m 2 plus melfalan 70-140 mg/m 2 or busulfan 8-10mg/kg. Pretransplant indicators of liver injury analyzed were aspartate aminotransferase (AST), alanine aminotransaminase (ALT), gammaglutamyl transpeptidase (GGT), alkaline phosphatase (AP), total Bilirubin (Bil), prothrombin time (PT) and the category of severe hepatic disease according to HCT-Comorbidity Index (HCT-CI_hep). Posttransplant severe hepatic injury (SHI) was defi ned as maximal total serum bilirubin (Bil) level > 4 mg/dl or encephalopathy of hepatic origin (according to Hogan, Blood, 2004) . Results: Pretransplant liver laboratory tests abnormalities were found in 51 (18%) patients. We observed high levels of AST, ALT, GGT, AP, Bil, PT and HCT-CI_hep in 5 (2%), 16 (6%), 26 (9%), 21 (7%), 14 (5%), 2 (1%) and 27 (9%) patients, respectively. Among the pretrasplant indicators of liver injury analyzed, HCT_CI hep showed to be the best predictor of transplant outcomes. In multivariate analysis (MVA), HCT-CI_hep high risk patients showed higher 100-days and 2-years NRM (HR 3.1 [95%CI 1.5-6.2] p = 0.002 and HR 1.9 [95%CI 1-3.7] p = 0.04, respectively), lower overall survival (OS) (37% vs. 53%, p = 0.04), higher grade 2-4 aGvHD p<0 .001]), higher grade 3-4 aGvHD and a trend to higher cGvHD (1.6 [95%CI 0.9-2.8] p = 0.06). After allo-RIC, a total of 182 patients (65%) developed grade III-IV liver toxicities. A total of 74 (26%), 111 (40%), 160 (57%) and 51 (19%) patients developed grade III-IV levels of AST, ALT, GGT, and Bil at a median time of 194 (rg 0-1629), 146 (rg 0-1932), 195 (rg 0-3126) and 74 days (rg 0-783), respectively. Sixty-seven patients (24%) developed SHI at 5 years with a cumulative incidence of 19% (95%CI 15-25). Main causes of SHI were GvHD (n = 36, 54%) and pharmacological toxicity (n = 7, 10%). In MVA, risk factors for SHI were unrelated donors (HR 3. A hallmark of acute graft-versus-host disease (aGVHD) is the selective attack of certain tissues, namely the gastro-intestinal tract, liver and skin but not others such as the pancreas and kidneys. Imaging techniques such as whole body in vivo bioluminescence imaging, dynamic multiphoton-laser-scanningmicroscopy and ultramicroscopy utilized by others and our own group help to elucidate the elusive mechanisms underlying alloreactive T cell traffi cking. Here we dissected the molecular mechanism that direct alloreactive T cells via vascular endothelial cells to target tissues in a murine MHC major mismatch hematopoietic cell transplantation mouse model. As a strategy we employed endothelial ligand blocking antibodies during the aGVHD effector phase after alloreactive T cells leave secondary lymphoid organs. Immunohistochemical analysis of target and non-target tissues revealed organ dependent upregulation of infl ammation induced endothelial ligands. The identifi ed surface molecules subsequently served as therapeutic targets. We learned that during the aGVHD effector phase multiple endothelial ligands need to be blocked simultaneously to avert aGVHD target manifestation. Blocking of single molecules such as MAdCAM-1, VCAM-1, E-Selectin, P-Selectin alone could not prevent alloreactive T cell traffi cking to the intestinal tract, liver and skin. When we treated transplanted mice with the combination of two or even three antibodies we observed a reduction of alloreactive T cells. However, only when we blocked four targets simultaneously, namely MAdCAM-1, VCAM-1, E-Selectin, P-Selectin in combination we could prevent GVDH target manifestation in the skin, in the liver and the small intestines. From these data we conclude that alloreactive T cell homing depends on the recruitment by infl ammation induced endothelial ligands of target tissues. After clonal expansion of alloreactive T cells simultaneous blocking of several endothelial ligands was required to effi ciently abrogate aGVHD in target tissues. Our data indicate that T cell homing remains highly attractive for therapeutic interventions. Our results also point out that several pathways should be targeted simultaneously in order to prevent aGVHD.

Distinct temporal and spatial roles for host conventional and Langerhans cells in the development of graft-versushost injury F. Fallah-Arani, H. Goold, B.R. Flutter, J. Sivakumaran, C.L Bennett, R. Chakraverty Royal Free and UCMS (London, UK) Background: Following bone marrow transplantation, donor T cell reactivity can be confi ned to the lympho-haematopoietic system or, in the presence of infl ammation, can additionally extend to peripheral tissues leading to graft-versus-host disease (GVHD). In this study, we have explored the role of individual host dendritic cell (DC) subsets in regulating these processes using models involving inducible depletion following donor T cell transfer to allogeneic chimeras. Methods: BALB/c Thy1.1 + T cells were transferred to established allogeneic chimeras generated following reconstitution of lethally irradiated B6 mice with a mix of BALB/c and B6.CD11c-DTR bone marrow. CD11c.DTR mice express a primate diptheria toxin receptor (DTR) under the control of the CD11c promoter. In the resulting chimeras, host CD11c + conventional DC (cDC) were depleted at the time of donor T cell transfer by injections of diphtheria Toxin (DT). In experiments to test the role of host Langerhans cells (LC), BALB/c Thy1.1 + splenocytes were transferred to BALB/c + B6 > B6.Langerin. DTR, where co-treatment with DT leads to depletion of host, epidermal LC. In each of the above settings, infl ammation was induced in some chimeras by systemic or local application of a Toll-like receptor (TLR) agonist at the time of T cell transfer. Results: Absence of host cDC at the time of delayed T cell transfer abrogated accumulation of donor CD8 cells in recipient spleens and reduced in vivo cytotoxicity against host target B cells. This was associated with a lack of conversion to full donor chimerism, demonstrating a requirement for host cDC in priming the lympho-haematopoietic graft-versus-host response in the steady state. We then explored the role of host cDC and LC in a model of cutaneous GVHD, where local GVHD is induced following application of a TLR agonist (imiquimod). Maximal imprinting of E-selectin ligand expression upon donor CD8 cells within the draining lymph node required the presence of host cDC. However, even in their absence, donor T cells were able to access the epithelium and cause injury. Selective depletion of host LC had no effect on imprinting or tissue infi ltration of donor T cells. However, the capacity of donor T cells to induce GVHD was severely impaired in this case. Conclusions: These data show that host cDC are required for GVH reactivity in the steady state. In contrast, in a model of skin GVHD, host cDC are dispensable whereas LC have a unique post-priming role.

Failure to imprint donor CD8 memory and exhaustion may explain loss of GVT responses following donor leukocyte infusions B.R. Flutter (1) , F. Fallah-Arani (1), S. Sivakumaran (1) , C.L. Bennett (1) , S. Ghorashian (1) , G. Freeman (2) , M. Sykes (2) (1)University College London (London, UK); (2)Harvard Medical School (Boston, US) Background: Donor T cell alloreactivity can be co-opted to deliver graft-versus-tumour (GVT) responses following donor leukocyte infusions (DLI) given after bone marrow transplantation (BMT). However, the major reason for treatment failure following DLI is relapse, suggesting a long-term failure of antitumour immunity. A distinctive element to the GVT response is that the antigens (Ag) to which the response is directed are cleared from the haematopoietic system but continue to be expressed by non-haematopoietic cells. In this study, we have explored the infl uence of non-haematopoietic Ag upon the antihost CD8 T cell response in a model of delayed DLI. Methods: B6 CD45.1 T cells were given after 8 weeks to allogeneic chimeras where Ag was ubiquitous (B6 + BDF1 >BDF1) or restricted to the haematopoietic compartment (B6 + BDF1 >B6). In each setting, direct priming of the initial T cell response by BDF1 Ag-presenting cells resulted in the eradication of BDF1 haematopoietic cells. However, in B6 + BDF1 >BDF1 chimeras, Ag was still present in peripheral tissues, whereas in B6 + BDF1 >B6 chimeras, Ag was cleared completely. The donor T cell response was tracked at timed intervals following transfer in both sets of chimeras. Results: The continued presence of non-haematopoietic Ag led to a failure to sustain donor CD8 cytotoxic and effector cytokine responses by 60d following transfer and this was associated with a failure to establish a central memory (T-CM) population. In contrast, where Ag was absent in peripheral tissues, functional T-CM populations were preserved. The failure to generate donor T-CM in chimeras with ubiquitous Ag expression occurred at two distinct levels. Firstly, residual DLI-derived CD8 T cells demonstrated a PD-1 hi/CD127 lo/ IFN-g neg 'exhausted' signature. Consistent with this, CD8 functions were partially restored by in vivo antibody blockade of the PD-1 pathway. Secondly, during the initial phase (<14d) of the response, we observed a lack of memory precursor formation when Ag was ubiquitous. Thus, when Ag was ubiquitous, the capacity of activated, post-mitotic, CD8 T cells to establish recall immunity following transfer to Ag-free hosts was much reduced. Conclusions: These data demonstrate that alloantigen within non-haematopoietic tissues is not ignored, but rather blocks memory imprinting and drives eventual CD8 T cell exhaustion. While these effects may lessen the risk of GVHD, they might also impair the durability of the GVT response. Background: Adoptive transfer (AT) of TCR-gene transduced T cells has become a promising therapeutic tool, however, limited in vivo survival and the development of an unresponsive state often termed 'exhaustion' has hampered reproducible clinical success. Here we demonstrate that a) upregulation of PD-1 on donor-derived TCR-engineered T cells is associated with a loss of GVL effects after late AT and b) PD-L1 blockade after hematopoietic stem cell transplantation restores robust GVL-effects. Methods: We chose a MHC-mismatched allogeneic BMT model (B10.A mice (H-2a) into C57BL/6 (H-2b) mice) and a retrovirally encoded TCR (OT-1 anti-ovalbumin) that was introduced into B10.A-derived donor T cells for AT after MHC-mismatched HCT. After HCT, leukemia-bearing (C1498-OVA) mice (syngeneic HCT-recipients served as reference) received AT and were monitored for GVHD, GVL, and in vivo T cell persistence/ functionality. For PDL-1 blocking experiments PDL-1 antibodies (clone 10F.9G2) were used. Results: 1) Up to 1x10 9 TCR-transduced CD8 + T cells were generated in vitro within 7 days using a retroviral vector system linking the genes for the á-and â-chain via a 2A sequence. 40% of CD8 + T cells coexpressed the respective TCR chains Vá2 and Vâ5. 2) The introduced TCR mediated comparable specifi c cytotoxicity against C1498-OVA in vitro being functional on autologous and allogeneic T cells. 3) After early AT transduced T cells rescued up to 60% of mice in a dose dependent manner. (p = 0.001-0.01 versus mock-transduced controls). 4) After late adoptive transfer (day 56) autologous T cells peaked in numbers by day 3 after AT and provided strong GVL-effects. In contrast, numbers of allogeneic TCR-modifi ed T cells declined rapidly within the peripheral blood and increasingly expressed PD-1. Lower frequencies of allogeneic T cells in the periphery translated into nearly abolished GVL effi cacy upon leukemia challenge 3 days after AT (p = 0.004). 5) PD-L1 blockade in vivo after allogeneic HCT restored GVL-effi cacy of adoptively transferred T cells raising leukemia free survival from 0% to 60% (p = 0.001 compared to isotype controls) without increasing GVHD rates. Conclusion: AT with TCR-engineered T cells after allogeneic HCT can decrease the risk of GVHD and provide potent GVL effects. However, increased expression of PD-1 on adoptively transferred T cells is associated with decreased GVL effi cacy and can effectively be reverted by PD-L1 blockade. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative treatment for many hematologic malignancies or hematopoietic dysfunction syndromes in adult patients, but the application is still limited due to major complications, such as severe graft versus host disease (GvHD). Diagnosis of aGvHD is based on clinical features and biopsies. We have shown earlier that aGvHD can be diagnosed using a proteomic pattern established by screening with capillary electrophoresis (CE) and mass spectrometry (MS). The aGvHD-specifi c proteomic pattern has been evaluated prospectively and blinded on more than 960 samples collected from more than 141 patients undergoing allo-HSCT at Hannover Medical School (MHH) and 7 additional clinics. Since 2007 a new diagnostic proteomic pattern for aGvHD was developed, using a higher resolution MS. Thus, 192 patients from MHH and 133 patients from 4 collaborating clinics were subjected to double screening for development and prospective evaluation of the MS-17-aGvHD-specifi c pattern. The majority of the patients included was transplanted for hematological malignancies (n = 313), 12 for hematopoietic failure syndromes (2 PNH, 9 SAA) . Conditioning regimens included reduced intensity conditioning regimens (RIC) for about 60% of the patients of MHH, as well as standard conditioning regimens (mainly TBI + Cy or Busulfan + Cy). GvHD-prophylaxis was cyclosporine A (CSA) and mycophenolate (MMF) or CSA metothrexate (MTX), as appropriate. In addition, about 80% percent of the patients received ATG (antithymocyte globulin) prior to HSCT.

The new aGvHD-specifi c pattern, MS_17, consisting of 17 differentially excreted proteins and peptides was developed and prospectively evaluated in parallel for the last 2 years. Prospective and blinded evaluation of the patients included in this analysis for early recognition of patients at risk for aGvHD development revealed the correct classifi cation of patients developing aGvHD about 7 days (range: 2-21 days) prior to the development of clinical symptoms for aGVHD with a sensitivity 76% and specifi city of about 85%. Pre-emptive therapy ad ministered upon positivity of the protemic patterns reduced the incidence and severity of aGvHD (p = 0.04). Thus, a multicenter study has been initiated in Germany to test the effi cacy and safety of pre-emptive therapy.

HHV6 and acute GvHD C. Pichereau, K. Desseaux, A. Janin, R. Peffault de Latour, M. Robin, P. Ribaud, S. Chevret, G. Socié Hôpital Saint Louis (Paris, FR) Background: Previous studies have suggested that HHV6 infection is correlated with acute GVHD following allogeneic HSCT, but whether HHV6 triggers, is associated with, or is a differential diagnosis of GVHD is unknown. Methods: 414 patients received an allogeneic HSCT at the Saint Louis Hospital (Paris) between January 2004 and December 2007. Whenever acute GVHD was suspected, HHV6 RQ PCR and organ biopsy was performed. Cumulative incidence of HHV6 reactivation was estimated using competing risks approaches. Predictive factors of increased risk of reactivation or acute GVHD were assessed using cause-specifi c hazard Cox models. Effect of reactivation on further occurrence of acute GVHD was tested by introducing a time-dependent variable.

Results: 376 pts were tested for HHV6: 229 (61%) were male, median aged 40 years . 300 (80%) had malignancies and 76 suffered from non malignant disorders (aplastic anemia, sickle cell disease, thalassemia). 188 pts were grafted with related donors (50%). The conditioning regimen was myeloablative in 237 cases (63%). The source of HSCT was bone marrow in 161 cases, peripheral blood in 157 cases and cord blood in 58 cases. 240 pts (64%) developed acute GVHD. HHV6 was detected in 100 pts, 66 of whom developed GVHD. HHV6 reactivation was signifi cantly associated with cord blood graft (59% versus 21%, p<0.0001) and unrelated transplant (39% versus 14%, p = 0.007). GVHD was signifi cantly associated with previous HHV6 reactivation (hazard ratio, 1.63; 95% confi dence interval, 1.01-2.62; p = 0.04). Tissue biopsies were available for 64 of the 100 HHV6 infected pts (32 skin and 32 gut samples). There were no signifi cant difference between pts who had a biopsy and pts who hadn't, except for the incidence of acute GVHD (75% versus 50%, p = 0.01). 24 pts who had no or poor pathological evidence of GVH (grade 0 or 1), later on developed severe clinical acute GVHD (grade III and IV), suggesting the role of HHV6 as a trigger of severe GVHD. Beside this, 10 pts who didn't have clinical and histopathological GVHD (grade 0) showed a significant lymphoid infi ltrate on their biopsy. Immunohistochemical analysis are ongoing, that could provide the evidence of "pure" HHV6-related skin rash after HSCT. Conclusion: This study suggests the role of HHV6 as a trigger of acute GVHD, particularly in its severe manifestations. Beside this, it confi rms that HHV6 infection is a differential diagnosis of acute GVHD in a signifi cant proportion of patients.

Prochymal® improves response rates in patients with steroid-refractory acute graft-versus-host disease involving the liver and gut: results of a randomized, placebo-controlled, multicentre phase III trial in GvHD P.J. Martin (1) , J.P. Uberti (2) Background and methods: Steroid-refractory acute GVHD (SR-GVHD) remains a signifi cant and life-threatening complication of allogeneic hematopoietic cell transplantation (HCT). Prochy-mal® (Mesenchymal Stem Cells, MSC, derived from unrelated volunteer adult donors) was evaluated in addition to standard of care, including institutionally selected second line treatment, in a randomized (2:1) trial in patients with SR-GVHD (Protocol 280). Patients received 8 infusions of 2 x 10 6 MSC/kg over 4 weeks (or volume equivalent for placebo), with an additional 4 infusions administered weekly after day 28 in patients who had a partial response, defi ned as improvement in at least one organ without progression in others. The primary endpoint was durable complete response (DCR) for ≥ 28 days. Additional prospectively defi ned outcomes included responses in patients by organ involvement. Patients and results: 244 patients with SR-GVHD (skin involvement n = 144, gastrointestinal involvement n = 179, liver involvement n = 61) were enrolled and treated on a 2:1 basis: Prochymal (n = 163), placebo (n = 81). There were no signifi cant differences in age, pre-transplant conditioning, graft source, HLA-matching or second-line therapy between treatment arms. For the Prochymal and placebo arms, respectively, the grades of GVHD at entry were B (22% vs. 26%), C (51% vs. 58%), and D (27% vs. 16%). The respective DCR rates were 35% vs. 30% (p = 0.3) in the intent-to-treat population and 40% vs. 28% (p = 0.08) in the per protocol population. Results for secondary endpoints are shown in table 1. Patients with GvHD affecting all 3 organs had overall complete or partial responses rate of 63% vs. 0% (n = 22, p<0.05, Fisher's exact test) at day 28. Patients treated with Prochymal had less progression of liver GVHD at weeks 2 and 4 respectively (32% vs. 59%, p = 0.05; and 37% vs. 65 %, p = 0.05). The incidence of infections was not different between arms. Incidence rates were 9% vs. 8% for recurrent malignancy, 1.8% vs. 2.5% for infusional toxicity, and 0.6% vs. 4.6% for AE-related discontinuation in the Prochymal and Placebo arms, respectively. Conclusion: GVHD with liver or gut involvement is a life-threatening complication of HCT. These results suggest that the addition of Prochymal produced signifi cant improvement without additive toxicity in patients with SR-GVHD involving visceral organs. S18 O137 Treatment of steroid-refractory acute GvHD with mesenchymal stem cells improves outcomes in paediatric patients. Results of the paediatric subset in a phase III randomized, placebo-controlled study P. Szabolcs (1) , G. Visani (2) Background: Successful treatment of steroid-refractory acute graft-versus-host disease (SR-GVHD) following allogeneic hematopoietic cell transplantation remains a signifi cant challenge. Because of their immunomodulatory properties and safety profi le, adult mesenchymal stem cells (MSCs) have been proposed as a treatment for SR-GVHD. Intravenous allogeneic MSC therapy (Prochymal) for SR-GVHD was independently evaluated in the pediatric subset of a double-blind, placebocontrolled study. Methods: Pediatric patients (<18 years old) with grade B-D SR-GVHD were randomized to receive either Prochymal or placebo in addition to standard of care, including institutionally selected second line agent. Patients received 8 infusions of 2 × 10 6 cells/kg for 4 weeks (or volume equivalent for placebo), with 4 more infusions weekly in the case of a partial response (PR). The primary endpoint was durable complete response (CR 28 days); secondary endpoints included incidence of CR, PR and progression through 100 days, survival, and safety. Results: Twenty-eight children were randomized to Prochymal (50% male, 79% Caucasian) or placebo (71% male, 71% Caucasian), with a median age of 7 yrs (range 1-15) and 10 years (range 1-18), respectively. The dominant transplant graft source was cord blood (71% Prochymal, 57% placebo), with mostly unrelated donors (93% vs. 79%, respectively). The median duration of aGVHD prior to enrollment was 20 days for Prochymal and 8 days for placebo (p<0.05). At baseline, aGVHD grades B:C:D were 3:8:3 for both arms. For Prochymal, organ involvement was 64% skin, 43% GI, and 36% liver. For placebo patients, organ involvement was 57% skin, 79% GI, and 29% liver. Durable CR was 64% for Prochymal and 43% for placebo (p = 0.5). Prochymal improved rates of CR and OR (Table) . The median time to CR was 25 days vs. 63 days. The safety data showed no infusional toxicity and no evidence of Prochymal leading to ectopic tissue. There were no AEs leading to discontinuation of therapy. Conclusion: In a SR-GVHD population in which 79% of patients had grade C/D disease, the addition of Prochymal to standard of care resulted into a faster and better CR. In view of their increased response rates and a well-tolerated safety profi le, MSCs appear to be a safe and effective therapy in the treatment of pediatric patients with SR-GVHD. Aim: response rate after induction and after ASCT. Patients and methods: TD consisted of thalidomide 200 mg/d and dexamethasone 40 mg on days 1-4 and 9-12 at 4-week intervals for 6 cycles. VTD was identical to TD plus Velcade 1.3 mg/m² on days 1,4,8,11 of each cycle. Combination chemotherapy plus Velcade consisted of 4 cycles of VBMCP/VBAD followed by 2 cycles of Velcade. Results: As of December 31, 2008, 306 patients (median age: 57 yrs, M 156, F: 150; IgG 183, IgA 71, light chain 43, others 9) entered the study. 55 of 253 (22%) had high-risk cytogenetics (t(4;14), t(14;16) and/or 17p deletion). All 306 patients (TD: 104, VTD: 102 and VBMCP/VBAD/Velcade: 100) were evaluable for response and toxicity to induction therapy. The CR plus VGPR rate was signifi cantly higher with VTD than with VBMCP/VBAD/Velcade (59 vs. 37%, p = 0.003) and TD (59 vs. 28%, p<0.001). The progressive disease (PD) rate was higher with TD than with VTD (22 vs. 8%, p = 0.005). In patiens with high-risk cytogenetics VTD was associated with a higher CR rate than TD and VBMCP/VBAD/Velcade (42 vs. 0%, p = 0.003 and 23 vs. 0%, p = 0.04, respectively) as well as with a lower PD rate (0 vs. 37%, p = 0.003 and 0 vs. 23%, p = 0.04, respectively). The incidence of > grade 2 thrombotic events was higher with TD than with VTD (8% vs. 1%, p = 0.01) and grade >2 peripheral neuropathy was higher with VTD than with TD or VBMCP/ VBAD/Velcade (14 vs. 1% vs. 0%, p<0.001). 218 patients were evaluable for response after ASCT. The CR rate was higher with VTD (52%) than with VBMCP/VBAD/Velcade (49%) and TD (37%) although the differences did not reach statistical signifi cance. The estimated OS at 2 yrs. was 82% with no significant differences among the 3 arms. TTP and PFS were shorter with TD (p = 0.05 and p = 0.012, respectively). Summary: 1) induction with VTD produced a higher CR + VGPR rate, 2) in patients with high-risk cytogenetics, TD resulted in a signifi cantly lower CR and higher PD rate than bortezomibcontaining regimens, 3) the PFS was shorter with TD and 4) ASCT increased the CR rate in 23%, 21% and 26% in the TD, VTD and VBMCP/VBAD/Velcade arms, respectively.

Interim analysis of a phase III, prospective randomized trial of melphalan Stem cell source was PBSC in 93% pts. 15% pts received myeloablative allogeneic SCT as part of a planned tandem strategy, whereas 85% received a non myeloablative conditioning.

Overall response at allo SCT was 84% and included complete remission (CR 19%), very good partial remission (VGPR 9%), partial remission (PR 56%). Results: All pts except 3 had sustained donor engraftment. Among pts, 36%, 13% and 20% achieved CR, VGPR and PR respectively at day 100. Median follow-up from allo SCT was 37 months (range 1-122). Event-free survival (EFS) was 58% (49-66) at 3 years. Cumulative incidence of transplant related mortality (TRM) at 100 days was 10% and was mostly related to pneumonia. A single line of treatment before tandem auto/allo SCT was associated with improved 3-year-EFS (p = 0.02). 28% pts experienced grade I to II acute graft-versus-host-disease (GVHD) and 13% grade III to IV. 11% pts had limited and 18% extensive chronic GVHD. Whereas grade I to II acute GVHD was associated to better 3-year-EFS (p<0.0001), extensive chronic GVHD had no statistically signifi cant impact on EFS. Moreover the absence of deletion 13 was not associated with a better EFS. Conclusion: Our results suggest that the number of treatment lines received before tandem auto/allo SCT is an important issue, with an improved EFS for pts treated by a single line before tandem. Whereas better control of acute GVHD might further improve survival, impact of the deletion 13 on outcome seems limited. Purpose: The outcome of MM patients treated with low dose TBI and allogeneic related SCT following autologous SCT has been shown to be superior to that from two autologous SCT (Gahrton et al.; Bruno et al.) . Since related donors are available for approximately 30% of patients only, the outcome of allogeneic unrelated SCT in patients without a related donor was compared to that of patients with a related donor in an intention to treat analysis.

Patients and methods: MM patients (n = 44) with a median age of 51 (range 32-65) years were treated at the University of Leipzig with autologous SCT followed by allogeneic SCT. Autologous stem cells were mobilised with cyclophosphamide and transplanted after Melphalan 200 mg/m 2 on day -3. Allogeneic SCT was performed at a median of 5 (range 2-11) months after autologous SCT using Fludarabine (30 mg/m 2 /d from days -3 to -1) and TBI (2 Gy on day 0) followed by cyclosporine (6.25 mg/kg twice daily from day -1) and mycophenolate mofetil (15 mg/kg daily from day 0). Patients received an unrelated (n = 22) or related (n = 22) stem cell graft. Results: Patients with unrelated donors (n = 22) showed hematological toxicity comparable to those of related donors. All patients had stable hematopoietic engraftment with T-cell chimerism of 100% 3-6 months after SCT. With a median follow up of 41 (range 2-87) months overall survival (OS) at 5 years was 51±12% and 52±13% (p = 0,19) in patients with unrelated and related donors, respectively. Progression free survival (PFS) was 34±12% at 5 years for patients with an unrelated compared to 14±8% for patients with a related donor. This difference was due to a decreased relapse incidence (RI) in patients with unrelated donors (54±12%) compared to related donors (84±9%). Non relapse mortality was 22±9% in the whole population with no difference between the two groups. Of the 22 patients having received allogeneic unrelated SCT, 7 (32%) are currently in CR, 2 (10%) in VGPR, 3 (14%) in PR and 1 (5%) has stable/progressive disease, with no difference between related and unrelated SCT. Conclusion: Our data confi rm the feasibility of autologous SCT followed by low-dose TBI conditioning regimen and unrelated SCT. PFS seems to be higher and RI lower in unrelated compared to related SCT. These feasibility data provide the basis for a phase III study comparing auto-allo unrelated to auto-auto SCT. Background: POEMS syndrome is a rare paraneoplastic syndrome resulting from a clonal plasma cell proliferation producing a small monoclonal protein usually lambda type restricted. This syndrome is a multisystemic disease, its major clinical feature being a progressive peripheral neuropathy. Single osteosclerotic lesions should be treated with radiation therapy only. In widespread disease, high-dose therapy followed by autologous hematopoietic stem cell rescue (ASCT) has shown to be an effective therapeutic option in short series, but with significant morbidity. Patients and methods: Between December 1999 and September 2009, 19 patients with POEMS syndrome (12 female/7 male) were treated with melphalan-200 (16 patients) or melphalan-140 (3 patient) followed by ASCT at 9 Spanish institutions. Median age was 54 years (range: 26-67). All of them presented a M protein lambda type (IgA: 16; IgG 3), peripheral polyneuropathy (18), osteosclerotic lesions (12), organomegaly (16), endocrinopathy (7), skin lesions (18), extravascular volume overload (14), papilledema (6), polycythemia (2) and thrombocytosis (12). Four patients had pulmonary hypertension, two portal hypertension, and three Castleman disease. Four patients were previously untreated. The median number of prior therapies was 2 (range, 0-4). Median time from diagnosis to ASCT was 8 months (range, 2-95).

Results: No transplant-related-mortality (TRM) was observed. After a median follow-up of 45 months (range, 3-89), one patient has died of progression 90 months post-ASCT. Five patients presented an engraftment syndrome and two a primary graft failure resolved, one with a back-up infusion on day + 26 and the other presented a delayed engraftment. Sixteen patients were evaluable for response, 8 achieved a complete hematologic response (CR) (IF-), 7 a near-CR (negative electrophoresis but IF positive) and one died of progression 90 months post-ASCT. Clinical improvement was observed at 4-6 months post-transplantation. All patients had a signifi cant organic improvement. The four patients with severe pulmonary hypertension and the two with portal hypertension improved of both pressures. Conclusions: In this series, ASCT proved to be a highly effective therapy for patients with widespread POEMS syndrome. Despite no TRM was observed, these patients may have a delayed hematopoietic recovery and may develop an engraftment syndrome that must be diagnosed/treated promptly.

A salvage treatment containing novel agents consolidated by allogeneic stem cell transplantation with reducedintensity conditioning improves outcome of multiple myeloma patients failing autologous transplantation F. Patriarca (1), H. Einsele (2)

Objectives: Allogeneic stem cell transplantation (allo-SCT) employing reduced intensity conditioning (RIC) is a feasible procedure in selected patients with relapsed multiple myeloma (MM), but its effi cacy is still a matter of debate. We investigated the role of RIC allo-SCT in MM pts who relapsed after auto-SCT and were then treated with a salvage therapy based on novel agents. Our study was structured similarly to an intention to treat analysis and included only those pts undergoing a HLAtyping immediately after the failure of auto-SCT. The cohort of pts having a donor (donor group) was compared with the one not having a suitable donor (no donor group). Patients and methods: One hundred thirty-six consecutive pts were retrospectively evaluated. Fifty-seven found a donor and 50 (88%) underwent an allo-SCT: 16 identical sibling (32%), 34 MUD (68%). Conditioning regimens were fl udarabine, melphalan±thiotepa in 21 patients (42%), fl udarabine + 2 Gy TBI in 15 cases (30%), fl udarabine and cyclophosphamide in 8 patients (16%) and fl udarabine and treosulfan in the remaining 6 cases (12%). Seven pts having a donor did not receive allo-SCT for progressive disease or severe comorbidities. Median age of donor group was signifi cantly younger than no donor group (53 versus 60 years, p = 0,000045). The 2 groups were balanced with regard of time between auto-SCT and relapse, treatment of fi rst relapse, duration of salvage treatment, and quality of response after salvage treatment. Results: The median follow-up for all patients was 15 months (1-97) after relapse. The median time to progression-freesurvival (PFS) and overall survival (OS) for all patients were 16 and 25 months, respectively. Two-year PFS was 41% in the donor-group and 18% in the no-donor group (p = 0.0003). Twoyear OS was 57% in the donor-group and 49% in the no-donorgroup (p = 0.08). In multivariate analysis the availability of a donor was statistically signifi cant for PFS (p = 0.003); moreover a signifi cant difference in outcome was observed comparing patients who achieved CR + VGPR versus PR versus SD/PD after salvage treatment (p = 0.05, p = 0.002, p = 0.0001 for PFS, p = 0.01, p = 0.001, p = 0.0001 for OS).

Conclusions: This study comparing salvage treatment with novel agents consolidated by RIC alloSCT versus salvage treatment alone after a failed auto-SCT provides evidence for a signifi cant PFS benefi t and a trend for a prolonged OS in MM patients having a donor. Grouping patients based on d and non-d KIR2DS4 allelic status revealed individuals homozygous for KIR2SD4(d) had significantly shorter duration of fever, compared to those heterozygous or homozygous for KIR2DS4 (non-d), (P = 0.003). This held true on multivariate analysis. No signifi cant association was seen between KIR2DS4 genotype and neutrophil engraftment time or microbial isolates. Of note, there was a signifi cant association between KIR2DL3 positivity and gram-ve bacteremia (P = 0.017). These data support a role for KIR genotype and infectious complications post-ASCT. KIR genotyping may aid risk assessment of infectious complications post-ASCT and optimization of anti-infectious prophylaxis, surveillance and treatment in those deemed at risk of sepsis. These results suggest a functional role for the deleted KIR2SD4 variant, previously believed non-functional in the innate immune response. In the era of novel therapeutic agents, high-dose chemotherapy and autologous stem cell transplantation (ASCT) remains an integral part of treatment for multiple myeloma (MM). Therefore, the choice of new drug combinations for induction therapies must take into consideration the requirement to collect a suffi cient number of haematopoietic stem cells (HSCs) for one or more courses of ASCT. Lenalidomide and Melphalan have been shown to impair HSC mobilisation, but induction therapies containing either Thalidomide or Cyclophosphamide do not have a relevant impact on the HSC collection yield. We considered the possibility that the combination of Cyclophosphamide and Thalidomide could have an additive impact on the HSC compartment and carried out a retrospective analysis of the outcome of peripheral blood HSC mobilisations performed in 111 MM patients. Patients who had received induction therapy with CTD (oral Cyclophosphamide, Thalidomide, Dexamethasone; n = 55) were compared with a control group of patients (n = 56) who had received VAD (n = 30) or Z-Dex (Idarubicin, Dexamethasone; n = 26) during the same 4-year period. All mobilisations were performed with Cyclophosphamide and G-CSF. Our standard collection target was 4 × 10 6 CD34 + cells/ kg, with a minimal target of 2 × 10 6 being accepted if patients failed the standard target. The total number of CD34 + cells harvested was signifi cantly lower in the CTD group (5.2 vs. 9.7 × 10 6 /kg, p = 0.002). The number of CD34 + cells harvested on the fi rst day of apheresis and per apheresis procedure were also lower in the CTD group (2.8 vs. 7.3 × 10 6 /kg, p = 0.002; 2.6 vs. 6.7 × 10 6 /kg, p = 0.002). More patients in the CTD group failed to achieve both the standard (36.4% vs. 16.1%, p = 0.021) and minimal (18.2% vs. 5.4%, p = 0.036) stem cell harvest target, despite a higher number of patients in the CTD group undergoing two or more apheresis procedures (52.8% vs. 32.1%, p = 0.012). The failure rate on the fi rst day of apheresis was also higher in the CTD group both for the standard (56.3% vs. 28.6%, p = 0.003) and the minimal target (36.7% vs. 16.1%, p = 0.041). Age or number of induction therapy cycles did not have an impact on mobilisation failure in the entire cohort or the CTD group alone. These observations provide novel evidence that drugs with no previously demonstrated effect on HSC mobilisation can have a considerable negative impact when used in combination, which can result in a high rate of HSC collection failures.

A. Nagler (1) Background: Allogeneic transplantation of hematopoietic stem cells (allo-SCT) from an HLA-matched related (MRD) or unrelated donor (URD) is a curative option for patients (pts) with high-risk hematological disease (HRHD). In the absence of a MRD, pts have been offered investigational transplant strategies such as umbilical cord blood (UCB) or family haploidentical SCT (haplo-SCT). In our Institution, all patients with HRHD are typed at entry; if a suitable MRD donor is missing a URD search is promptly activated. Our policy is to offer an haplo-SCT to adult pts lacking an MRD or URD in order to adequately treat HRHD in the ideal appropriate time according to clinical indications to allo-SCT agreed in ongoing protocols for primary disease.

Methods: Here we report the retrospective intention-to-treat (ITT) analysis of alternative donor transplantation at our Institution. Data were obtained from Institutional database. Results: Between Jan-2004 and Nov-2009, 361pts (100% of the following ITT analysis; median age 48y) received indication to allo-SCT according to EBMT recommendations. Eightytwo pts (23%) received a transplant from a MRD; 170pts (47%) activated an URD search in the IBMDR registry; 78pts (22% of total pts, 46% of URD searching) received a URD transplant; 35pts (9.7%-20,5%) received an haplo-SCT due to lacking of a suitable URD in the appropriate timing according to disease status, or absence of suitable criteria to engage an URD donor; 9 pts (2,5%-5,2%) received a UCB because lacking a suitable haplo donor. Overall, 129pts received an haplo-SCT (36%): 35 after IBMDR research, 94 up-front. Nineteen pts died before receiving a transplant (5%), 44 (12%) are searching for a suitable donor.

If we consider only pts with acute leukemia (213pts, median age 47y, range 15-72; over 50y 92pts) 47pts (22%) received a transplant from a MRD; 66pts (40%) activated an URD search; 37pts received a URD transplant (17%), 6pts (3%) a UCB, 97pts an haplo-SCT (46%). Eight pts died before receiving a transplant (4%), 18 (8%) are searching for a suitable donor. Allo-SCT was performed in complete remission in 104pts (62% alive), in persistence of disease in 83pts (20% alive). Conclusion: In ITT analysis, 83% of overall pts candidate received an allo-SCT: 60% from an alternative donor, 23% from a MRD. The highly committed policy performed in the alternative-SCT setting and the implementation of an alternative donor option are essential prerequisite to obtain these results. There was a marked variation in total numbers of transplants performed between countries ranging from 38 to 2629. The median age of HSCT programs was 10 yrs (range 3-23). The total number of HSCT performed per year has continued to increase and is yet to plateau. A greater proportion of transplants was allogeneic HSCT (allo-HSCT) compared to autologous HSCT (ASCT) (77% vs. 23%). Acute leukemia constituted the main indication for allo-HSCT(37%). There was a relatively high rate of HSCT for bone marrow failure (n = 1001, 17%) and hemoglobinopathies (n = 885,15%) when compared to data reported to EBMT and IBMTR. CML continued to be an indication for 8.7% of allo-HSCT in 2007. The rate of unrelated donor HSCT remained low, with only 2 non-umbilical cord unrelated donor transplants over the surveyed period. The use of peripheral blood stem cells varied between countries though increasingly constituted the main source of stem cells in allo-HSCT. RIC was used in 13.4% of allo-HSCT in the the survey. ASCT rates continue to increase; while acute leukemia was the main indication for ASCT in the 1980s, overall the main indications for ASCT in the survey were lymphoma (45%) followed by myeloma (26%). Conclusion: We present the fi rst survey of HSCT activity in the EM region over 23 years which refl ects the unique health conditions of this region, accounting for notable differences in transplant practices from Europe and North America. Annual HSCT rates continue to rise. Economic, logistic and other factors are likely to be responsible for disparities in activity within the region. This survey may be valuable in providing a basis for healthcare planning in the fi eld of HSCT in the region. Background: Related haploidentical donors, as cord blood, can be alternative donor sources in stem cell transplantation (SCT). Severe graft-versus-host disease (GVHD), however, has interfered the progress of haploidentical stem cell transplantation (haploSCT). To deal with this strong GVHD, T cell depletion has usually been used in European countries. On the other hand, based on the difference of ethnicity prone to less severe GVHD in Japan, we have performed unmanipulated haploSCT using myeloablative or reduced intensity preconditioning regimen for ten years. In this meeting, we will summarize our experience of ten years. , and patients over 45 years old or with comorbidities or repetitive SCT (including second to fi fth SCT) underwent reduced intensity preconditioning regimen consisting of FLU/(CA)/BU/ATG or FLU/(CA)/MEL/ATG (haplomini, n = 146). High dose Ara-C (CA) was optional to reduce tumor burden. ATG (Fresenius) dose was 2 mg/kg/day for four days. GVHD prophylaxis consisted of taclolimus (TAC), methylprednisolone (mPSL) 2 mg/kg/day, short term MTX, and mycophenolate mofetil (MMF) 15 mg/kg/day in haplo-full, and TAC, mPSL 1 mg/kg/day in haplo-mini, respectively. For elderly patients over 50 years old in haplo-mini, MMF was added. Results: Hematopoietic engraftment in haploSCT was as rapid as that in HLA-identical SCT, except eight cases of graft rejection. Acute GVHD (grade II-IV) was observed in 30%. Overall survival in fi ve years is 30% in haplo-full and 40% in haplo-mini, respectively. If limited to CR cases, overall survival reached over 60% in haplo-mini. There is no difference in survival rate among patients' diseases. Discussion: Unmanipulated haploSCT is feasible and effective for refractory diseases. ATG dose used in haplo-mini is rather low compared with that of European cases reported so far. Although it should be too early to refer long term outcome, unmanipulated haploSCT could be considered as an option to fi ght against refractory diseases.

Background: While chemotherapy + G-CSF (G) is effective for autologous stem cell mobilization (SCM) it exposes patients (pts) to risks and side effects. Plerixafor (P) + G provides a safer alternative for mobilization of hematopoietic stem cells in pts with MM or NHL. To better understand the comparative effectiveness of standard and newer approaches for SCM, we conducted a study aimed to: (1) determine the clinical outcomes and cost of SCM with cyclophosphamide (CY) plus G (2) compare outcomes of CY + G to a clinical trial of P + G. Methods: A retrospective study was conducted in all pts undergoing fi rst SCM from 1/04 to 3/08 (n = 241) with CY (3 gm/m 2 ) and G (10 mcg/kg started 1 day after CY for 10 days with planned fi rst day of collection on day 11). Apheresis was initiated when peripheral blood had >15 CD34 + cells/μL. Positive clinical outcome (PCO) was defi ned as >2 × 10 6 CD34 + cells/kg collected on planned day of collection and within 1 or 2 apheresis without a prior negative event that led to additional clinic/inpatient evaluation. The cost of drugs, apheresis, product processing, and clinical events were reported based on Medicare part B physician, laboratory, and ancillary fee schedule. Data on pts undergoing SCM with P + G were obtained from published clinical trials in pts with myeloma (MM) or lymphoma (L) undergoing fi rst SCM using G (10 mcg/kg without dose escalation) and a maximum of 4 days of P. Results: Among CY + G pts, 141 (61%) were males; 121 (50%) had MM, 115 (48%) had L (5-other diagnoses); and 61 (25%) pts. had prior radiation. PCO was seen in 48 (20%) pts.; 23% of MM and 15.7% of L pts. Median total cost of CY + G SCM was $10,732 (range, 6988-30827). PCO was associated with a lower cost than non-PCO in the overall group, (mean, $10,371 vs. $12,870, p = 0.001), in MM pts. (mean, $10,511 vs. $12,152, P = 0.026), and in L pts (mean, $10,133 vs. $13,627, P = 0.006). Assuming a similar distribution of PCO in 100 pts. with MM and L, the projected per pt cost of SCM would be $11,774 and $13,067 (mean, $12,421) with CY + G. Projected costs of SCM using P + G for 100 pts. with MM and L were $12,852 and $8986 (mean, $10,919). Conclusion: SCM with CY + G was associated with a lower rate of PCO and higher cost than P + G, supporting front-line use of P + G for SCM. Future prospective studies should investigate whether SCM with P + G translates into decreased resource utilization and improved quality of life for pts undergoing SCM. .5x10 9 /L, platelet count (Pt) >150 × 10 9 /L, and hemoglobin level (Hb) >120 g/L], good partial response (GPR) [ANC >1 × 10 9 /L, Pt >50 × 10 9 /L, and a Hb level >100 g/L] and poor partial response (PPR) [ANC >0.5 × 10 9 /L, Pt >20 × 10 9 /L, Hb level >80 g/L and transfusion independency]. No response (NR) was defi ned as failing criteria for at least PPR. Non-responding patients received a second-line therapy (HSCT or androgens), a second course of IST with LG (15 mg/kg/day/ × 5 days) or TG (3.5 mg/kg/day/ × 5 days); or no treatment. Subgroup analyses were conducted and differences in response were tested using the chi-square statistic test. Immunoablation followed by immune reconstitution by transplantation of autologous peripheral blood stem cells is a promising approach to treat refractory or otherwise incurable S26 autoimmune diseases. It has recently been suggested that this treatment could lead to remission in some cases of the early diabetes type 1 if administered prior complete destruction of beta cells by autoimmune mechanisms. The objective of this study was to verify these fi ndings on an independent group of patients. Methods: Eight patients (age 19-32) with early diabetes type 1 (no more than 6 weeks from diagnosis, C-peptide positive, Anti GAD -antibodies positive) were qualifi ed for the treatment. The patients were subjected to 2-4 plasmaphereses to remove circulating immune complexes and then mobilized with cyclophosphamide and G-CSF. Leucaphereses were continued until >3,0 x 106 CD34 + cells/kg were collected. The conditioning consisted of cyclophosphamide (50 mg/kg/day on days -5, -4, -3, -2 prior to transplant) and antithymocyte globulin (ATG Genzyme -of 0.5 mg/kg/day given on day -5, 1.0 mg/kg/day given on days -5, -4, -3, -2 and -1). Results: Median follow up for the patients as of December 2009 is 10 months (range 6-19 months). No major complications were observed during the transplantation and in the posttransplantation period. All patients (8/8) became independent from exogenous insulin after the transplantation. Median day of insulin withdrawal was + 24 (range + 6 to + 60). One patient resumed insulin 7 months post transplant. Six out of 8 patients were given acarbose to reduce the potentially toxic infl uence of observed hyperglycemias. After the transplantation patients exhibited good control of glycemia -the average HbA1c concentrations were 12.3% at the diagnosis, and 5.59%; 6.23%; 5.9% -3, 6, and 12 months after the transplantation respectively. This was correlated with increased levels of C-peptide after the transplantation. Conclusion: This report independently confi rms that independence of exogenous insulin could be reached in diabetes type 1 patients following immunoablation and reconstitution of the immune system with autologous peripheral blood stem cells providing that the procedure is performed prior total destruction of beta cells in the pancreas by the autoimmune mechanisms. The use of acarbose may have potentially positive effect on the duration of remission of diabetes type 1 in the patients after the transplantation.

New perspectives on the use of autologous haematopoietic stem cell transplantation in multiple sclerosis: three strategies of high-dose immunosuppressive therapy with autologous haematopoietic stem cell transplantation During the last decade high-dose immunosuppressive therapy with autologous hematopoietic stem cell transplantation (HDIT + ASCT) has been used with increasing frequency as a therapeutic option for MS patients. We aimed to study clinical outcomes in multiple sclerosis (MS) patients after early, conventional, and salvage of HDIT + ASCT. 155 MS patients (secondary progressive-54, primary progressive-28, progressive-relapsing-5, relapsing-remitting-68) were included in this study (mean age-33.0; male/female -63/92). BEAM or mini-BEAM conditioning regimens were used. 136 patients did not have any supportive treatment after HDIT + ASCT; 29 patients with risk factors were administered mitoxantrone during a year after ASCT as a consolidation therapy. 68 patients underwent early transplantation (EDSS 1.5-3.0), 79 -conventional (EDSS 3.5-6.5), 8 -salvage transplantation (EDSS 7.0-8.5). Median EDSS at base-line -4.0. The mean follow-up duration -22 months (range 6-125). Neurological evaluation was performed at baseline, at discharge, at 3, 6, 9, 12 months, and every 6 months post transplant.

Transplantation procedure was well tolerated with no transplant-related deaths. The effi cacy analysis was performed in the group of patients who did not have consolidation treatment.

Out of 78 patients with the follow-up at least 9 months the following distribution of patients according to clinical response at 6 months post transplant was observed: 36 patients (46%) achieved an objective improvement of neurological symptoms; 41 patients (53%) had disease stabilization; 1 patients (1%) progression. In the patients who underwent early transplantation (n = 32) 15 (47%) patients stabilized and 17 (53%) improved. After conventional transplantation (n = 41) 22 (54%) patients stabilized, 18 (44%)-improved, and 1 (2%)-progressed. After salvage transplantation (n = 5) 4 patients were stable, and 1 improved. At long-term follow-up (median-26.5 months) out of 60 patients improvement was registered in 35 (59%) patients; stabilization -in 20 (33%) patients, and progression in 5 patients (1 patient after early; 4 -after conventional transplantation). No active, new or enlarging lesions were registered in patients without disease progression. Thus, HDIT + ASCT appears to be a safe and effective treatment for MS. Further studies should be done to establish the best timing for transplantation and to evaluate consolidation therapy in patients receiving early, conventional, and salvage transplantation. In the last decade, the so-called nonmyeloablative or reduced intensity conditioning (RIC) regimens for allogeneic stem cell transplantation (allo-SCT) have emerged as an attractive modality to decrease allo-SCT-related toxicities and nonrelapse mortality. Indeed, RIC allo-SCT represents an attempt to harness the immune graft-versus-tumor effect while attempting to control or overcome toxicity. The work of different pioneering groups rapidly proved that this approach is feasible in several disease settings or patients' categories, and had the added benefi t of expanding the transplant option to patients who are ineligible for standard myeloablative allo-SCT. Currently, the use of RIC allo-SCT has challenged the need for high-dose conditioning regimens. Unfortunately, and despite several thousands of patients receiving RIC allo-SCT, the true value of RIC allo-SCT in the management of hematological and non-hematological malignancies, especially AML is, as yet, diffi cult to delineate. Currently, there are only very few, if any, prospective, randomized, or controlled trials that addressed the specifi c role of RIC allo-SCT versus other treatment strategies in AML. Based on a large number of registry-based analyses, the RIC subcommittee of the ALWP has attempted over the last few years to better defi ne the role of RIC allo-SCT in AML patients through fi ne analysis of leukemia-free survival and overall survival balanced against treatment-related toxicity, complications, and death. A major advance was the identifi cation of the most important comorbidities that are likely to impact outcome after RIC allo-SCT for AML. Another major achievement is the launch in 2010 of the multicenter/multinational EBMT randomized phase 3 study evaluating the role of RIC allo-SCT in elderly patients with AML. At present, the use of RIC prior to allo-SCT appears to be on the cutting-edge. The RIC approach proved to be much more complex than originally thought. Since the fi rst reports published in the late 90s', the RIC allo-SCT literature has exponentially expanded. The complexity of RIC allo-SCT practice is progressively deciphered and the optimism to regard RIC allo-SCT as a potential and promising treatment modality for many AML patients remains very high among investigators, warranting continuous and renewed clinical and therapeutic research in this area.

Transfusion of donor lymphocytes (DLT) for treatment of leukaemia relapse after allogeneic stem cell transplantation (SCT) has been a milestone in the fi eld of immunotherapy against malignant disease. It can be regarded as the proof of principle for the Graft-versus-Leukemia effect. During recent years, the immunotherapy subcommittee of ALWP has initiated a variety of retrospective studies to evaluate in detail the role of unmanipulated DLT in acute leukaemia. In contrast to CML and early stages of MDS, DLT was less effective in the treatment of post SCT relapse in highly proliferative diseases as acute leukaemia. Nevertheless, when given as maintenance therapy after effective cytoreduction, DLT could induce long term remissions in selected patients in AML, both after standard and reduced intensity conditioning transplants. Strategies that included DLT were more effective than treatment without the use of donor cells. In contrast, no such effects could be shown in ALL. These data have prompted the strategy to use donor cells already before the occurrence of overt haematological relapse, i.e in minimal residual disease, mixed or falling donor chimerism, or even prophylactically. This strategy is now widely used across European transplant centres and has shown promising results. Future strategies include the combination of targeted therapies and donor cell based strategies, as well as the use of specifi c subsets of donor cells in order to augment the antileukemic effi cacy and control for side effects of DLT, such as GvHD. Allogeneic Stem Cell transplantation (AlloSCT) with reduced intensity conditioning (RIC) is being increasingly used for acute myelogenous leukemia (AML) patients with high comorbidities not eligible for standard myeloablative conditioning. New compounds and formulations including intravenous Busulfan, Treosulfan and recently Clofarabine have been introduced into the pre transplant conditioning regimens in an attempt to reduce transplant related toxicities while keeping or increasing the anti leukemic effect, overcoming the increased relapse rate usually associated with RIC AlloSCT. Similarly, replacing BM by peripheral blood stem cell (PBSC) grafts may increase the GVL, albeit with the price of increasing toxicities, GVHD and TRM. Thus, the optimal conditioning regimen for adults with AML is yet unknown. In order to address these issues, the ALWP of the EBMT recently performed several surveys analyzing the use of IV Bu for AlloSCT, comparing IV Bu/Cy to TBI/Cy, comparing PBSC and BM, and is about to compare Treosulfan to IV Bu as conditioning regimens for adult patients with AML. In the IV Bu/Cy vs. TBI/Cy survey that included 1479 patients, the fi ndings indicated that outcomes of AlloSCT including engraftment, TRM, RR and LFS were comparable while GVHD probability was signifi cantly lower and VOD probability signifi cantly higher in adults patients with AML using IV Bu/Cy vs. TBI/Cy conditioning, respectively. In a separate survey analyzing the use of IV Bu AlloSCT for AML the 1 year cumulative incidence of VOD was found to be 8.4 + 2%. The factors associated with VOD were mismatched unrelated donors and being not in remission. The PBSC vs. BM survey included 1537 patients. We observed signifi cantly higher incidence of GVHD and NRM and lower incidence of relapse rate with the use of PBSC, resulting in equivalent LFS. The Treosulfan vs. IV Bu survey is ongoing. In house comparison revealed similar LFS in AML patients in remission, while results were better using 4d IV Bu combinations in patients with active disease. For MDS the Flu/Treo conditioning regimen was the best. In conclusion, it is conceivable that introducing new compounds in the pre AlloSCT conditioning regimens in combination with pre-and post-AlloSCT targeted therapy will enable us to tailor the conditioning regimen to the specifi c disease category, reducing toxicity while improving the anti tumor activity. Allogeneic HSCT (alloHSCT) is generally considered the best consolidation option for younger patients diagnosed with an AML with high-risk cytogenetics. Nonetheless, the cytogenetic high-risk category comprises different AML subtypes, and the specifi c results of alloHSCT for most of these cytogenetic entities are mostly unknown. Moreover, several molecular lesions, such as mutations of FLT3, MLL or WT1 genes, identify subsets of patients with a poor-prognosis disease despite harboring a non-adverse-risk cytogenetic abnormality. Therefore, the precise role of alloHSCT for the management of high-risk AML should be determined after a comprehensive analysis of the outcome of transplant in all these biological AML varieties with an adverse prognosis. In this context, the Molecular Markers SC is conducting a project to analyze the results of alloHSCT for diverse high-risk AML subsets, such as AML with t(6;9), AML with 3q abnormalities, and normal karyotype AML with internal tandem duplication of FLT3 gene (FLT3-ITD Three-year LFS was 28 9% and 18.9% for patients in CR1 and PIF, respectively. Finally, the prognostic impact of FLT3-ITD on the outcome of alloHSCT has been analyzed in a cohort of 120 patients (median age: 41, 18-60) with a normal karyotype AML. Donor was an HLA-identical sibling in 75% of cases, whereas conditioning regimen was a myeloablative regimen in all transplants. All patients received transplant in CR1. Of note, 2-year LFS, RI, and NRM was 58.5%, 30.5%, and 13.3%, respectively, without a signifi cant difference between transplants from related and unrelated donors. In summary, the outcome of alloHSCT in patients with two high-risk AML subtypes such as AML with t(6;9) and FLT3-ITD AML showed a relatively favourable outcome, which compares favourably with reports from non-transplanted patients. On the contrary, the outcome of patients with AML and 3q26 rearrangement seems inferior, even in patients transplanted in an early phase of the disease. Therefore, in order to elucidate the current role of alloHSCT for high-risk AML, transplant analyses should be focused on specifi c biological AML entities, since the results might differ largely among AML subtypes. Finally, a careful collection of biological features at diagnosis is essential for addressing studies based on biological categories.

HSCT is a well-recognized option for the treatment of acute leukemia with the number of transplantations continuously growing over the last decades. However, the HSCT rates vary strongly between countries. In particular, the activity in Central and Eastern (C&E) Europe is markedly lower compared to the Western part of the continent, suggesting the role of various socio-economic and geographic factors. Using the Human Development Index (HDI) as a surrogate marker we demonstrated that the socio-economic status (SES) of a country strongly correlates with the total number of HSCT per population, as well as separately with sibling-HSCT, unrelated-HSCT and autologous-HSCT in Europe. We further speculated that in line with the transplant activity the results of HSCT may also vary between countries or regions. Direct comparison of HSCT performed from sibling donors for AML in CR1 showed superimposable results obtained in C&E and Western Europe, however, the transplant characteristics differed with younger recipient age, longer interval from diagnosis to HSCT and less frequent use of peripheral blood, TBI, T-depletion and reduced-intensity conditioning in C&E countries. Subsequent analysis demonstrated that the outcome of sibling-HSCT for acute leukemias in C&E Europe improved over time. In particular the cumulative incidence of NRM decreased from 22 2% for patients treated between 1990-2002 to 15 3% for HSCT performed between 2003-2006, despite increasing recipient age. In another analysis we demonsatrated the infl uence of the SES on outcome in Europe, which, however, could not be explained by regional differences. Best outcome (increased LFS and reduced RI) after myeloablative HSCT from either related or unrelated HLA-matched donor was observed in 8 countries with the highest HDI, while no differences could be demonstrated for the remaining 22 analyzed countries. Interestingly, transplantations in countries with the highest HDI were characterized by increased recipient age and shorter interval from diagnosis to HSCT. Finally, we speculated that the center experience may infl uence results of HSCT. Using transplants with reduced-intensity conditioning (RIC-HSCT) as a model we analyzed results according to the team experience as defi ned by: 1) time from the fi rst alloHSCT and from the fi rst RIC-HSCT, 2) the total number of alloHSCT and the number of RIC-HSCT performed in a study period (2001) (2002) (2003) (2004) (2005) (2006) (2007) . We found that results in centers performing few RIC-HSCT (<15 in 7 years) were inferior compared to the remaining centers in terms of all, LFS, RI and NRM, which has been confi rmed in a multivariate model. Altogether, in a series of retrospective analyses based on the EBMT ALWP registry we demonstrated that both rates and outcomes of HSCT for acute leukemia vary among countries and centers. Both socio-economic factors and team experience infl uence results. We previously reported the EBMT experience with salvage high-dose chemotherapy (HDCT) in pediatric patients with extragonadal germ-cell tumour (GCT) (De Giorgi U et al. -BJC 2005) . We analyzed a total of 23 children with extragonadal GCT, median age 12 years (range 1-20), treated with salvage HDCT with haematopoietic progenitor cell support. The GCT primary location was intracranial site in nine cases, sacrococcyx in eight, retroperitoneum in four, and mediastinum in two. Twenty-two patients had a nongerminomatous GCT and one germinoma. Nine patients received HDCT in fi rst-and 14 in second-or third-relapse situation. No toxic deaths occurred. Overall, 16 of 23 patients (70%) achieved a complete remission. With a median follow-up of 66 months (range 31-173 months), 10 (43%) have been continuously disease-free. Of six patients who had a disease recurrence after HDCT, one achieved a diseasefree status with surgical resection followed by chemotherapy and radiotherapy. In total, 11 patients (48%) were diseasefree at the time of analysis. Eight of 14 patients (57%) with extracranial primary and three of nine patients (33%) with intracranial primary GCT were disease-free at the time of analysis. HDCT induced impressive durable remissions as salvage treatment in children with extragonadal extracranial GCTs. After 7 years from this fi rst analysis (done in 2003 published in 2005), we decide to analyze the long-term results of this experience integrating these data with other from other cases from the EBMT registry previously not included, and with data from additional patients from Italian registries (GITMO and AIEOP).

High-dose chemotherapy and autologous stem cell transplantation in relapsed germ cell tumours: do we need a randomized study? P. Pedrazzoli (1) Nasopharyngeal carcinoma (NPC) is an Epstein-Barr virus (EBV)-related malignancy expressing a restricted set of viral antigens. The outcome of patients with NPC failing conventional radio-chemotherapy is poor, the median overall survival of patients receiving second line treatments being less than two years. Hence, the need for alternative therapies capable of improving disease-free survival and associated with reduced toxicity.

Since 2001, we have implemented a T-cell therapy program for patients with NPC failing conventional treatment. So far, we have treated 27 patients with disease resistant/relapsing after >2 lines of radio-chemotherapy in two sequential trials of cell therapy with autologous EBV-specifi c cytotoxic T-lymphocytes (CTL). In detail, EBV CTL (4 escalating doses to a maximum of 8 × 10 7 CTL/dose in the fi rst trial, or 2 doses of 3-4 × 10 8 CTL in the second trial) were administered in patients without or with preceding lymphoablative chemotherapy and recombinant human interleukin-2 (rhIL-2). The proportion of patients achieving response (partial, PR, complete, CR) or stable disease (SD) lasting at least 3 months (RECIST criteria), as well as immunologic correlates of effective treatment, were evaluated. Overall, the objective control of disease was 57%, with 6/27 patients showing complete (n = 1), partial (n = 4) and minimal (n = 1) responses, and 10/27 disease stabilization (median duration 7 months). No severe adverse events were observed after cell therapy; 4 patients showed an infl ammatory reaction at the tumor site. The use of preparative chemotherapy and increased CTL dose did not infl uence outcome. Importantly, patients showing response to cell therapy showed the emergence of EBV LMP2 antigen-specifi c T-cells in their peripheral blood. EBV-specifi c CTL therapy is safe and associated with clinical benefi t in patients with advanced NPC refractory to standard therapies. The use of CTL with higher specifi city for the EBV subdominant antigens expressed by the tumor, such as LMP2, at an earlier stage of disease, could further implement the strategy and ameliorate the outcome of patients with relapsing/refractory NPC. This study was supported by a grant from the Italian Association for Cancer Research (AIRC). , and CSA given orally (5 mg/kg/d). Patients randomized to receive G-CSF were given glycosylated G-CSF from day 8 to 120 (150 microg/m 2 /d, sc). OS at 6 years was 75%, it was 82% for patients with severe AA and 66% for patients with very severe AA (P = 0.001). Survival decreased with increasing age from 100% (<20 years) and 92% (20-40 years) to 71% (40-60 years) and 56% (>60 years) (P < 0.001). There was no difference, overall and when stratifi ed according to age and AA severity in OS (P = 0.64) and EFS, defi ned by death, need for transplantation, relapse and non response as events (P = 0.36) between the study arms at 6 years ( Table 1 ). The median neutrophil count was signifi cantly higher between day 30 and 240, in the G-CSF group; this difference did not persist to day 360, when G-CSF was stopped. There were fewer infections (36% no G-CSF; 24% with G-CSF; P = 0.006), and less days of hospitalization during the fi rst 90 days (P = 0.03) in the G-CSF group. 44 patients died, mostly from infection (55%).There was no difference in death rates, cause of death and response rates between both groups. Overall 73% of patients with and 62% without G-CSF did respond to IS (P = 0.49). 57 patients did not respond to fi rst-line therapy, 31 patients relapsed during the fi rst year of treatment (no difference between both groups).

In conclusion, G-CSF added to standard IS increases neutrophil counts, and decreases rate of infections and days of hospitalization but has no impact on OS, EFS, remission, death and relapse rates. This has to be weighed against possibly higher risks of MDS/AML, as suggested by previous studies. an HLA-identical sibling (65%), 157 pts received a myeloablative regimen (78%) and 135 bone marrow (64%) with GvHD prophylaxis consisted in CSA ± MTX in 154 pts (73%). During evolution, 14 pts did not engraft (7%). Acute GvHD of grade >I developed in 52 patients (29%). Chronic GvHD developed in 57 pts (26 extensive). After a median ±SE follow-up time of 61±6 months, 64 pts died (35 infections, 18 GvHD). The 5-year OS rate was 68 ± 3%. None of the common variables examined for association with OS were statistically signifi cant, except for SCT indications with an increasing risk from Rsh, SAA and Thr ( Figure 1 , p = 0.03). Based on the 24 pairs obtained through the matching process, the 5-year OS estimate after Thr ( Figure  2 ) was 45 ± 11% for pts with SCT and 87±8% for pts without SCT [HR 10.0 (95%CI 1.3 to 78.1); p = 0.01]. Of note, the OS among non-SCT pts was signifi cantly worse (p = 0.01) in the non-matched than in the matched group, raising the question of pts over selection due to the matching process. Concerning pts with SAA complication, after exclusion of pts with Thr and of non-SCT pts with a follow-up time <6 months after SAA, the 5-year OS was 96±2% for pts without SCT and 81±4% for pts with SCT, but the 2 groups differ signifi cantly for age at SAA and year of SAA.

Conclusions: Further matching processes are necessary to conclude on this large cohort of PNH pts in order to defi ne the exact place of SCT in PNH, especially in the era of the Eculizumab. Introduction: Currently peripheral blood (PB) is more commonly used as stem cell source than bone marrow (BM) in both autologous and allogeneic hematopoietic stem cell transplantation (HSCT) (Gratwohl, BMT, 2005) . However, PB is associated with an increased risk of chronic graft-versus-host disease (GVHD), which is a disadvantage in non-malignant diseases. A recent study of EBMT/CIBMTR showed that incidence of chronic GVHD and overall mortality were higher after HSCT with PB than after HSCT with BM in young patients with severe aplastic anemia (SAA) (Schrezenmeier H, Blood, 2007) .

The EBMT transplant activity survey shows that the number of PBSCT has increased rapidly since early 90's and exceeded the one of BMT in 1999 in family donor and in 2002 in unrelated donor (UD)-HSCTs. The number of PBSCT has also increased in SAA. We were therefore interested to review the current status of stem cell source selection in bone marrow failure ( In conclusion, PB has been increasingly used as a stem cell source in BMF despite of its higher risk of chronic GVHD. There were major differences in stem cell source distribution, regarding donor type, the global region as well as countries in regions. It may refl ect the differences in infrastructure in each center/country, donor and physicians preferences and policy of marrow donor program. Clearly, recommendation for stem cell source is warranted in BMF. We conclude that 5q-syndrome can be cured by allogeneic transplantation, but additional abnormalities reduce the curability. The role of lenalidomide given before transplant will also be analysed. patients. An interim toxicity analysis was performed when the fi rst hundred patients had been included. The safety committee agreed to resume the trial. We will report the feasibility and the toxicity in both arms. An interim analysis is currently ongoing.

Other preliminary studies demonstrate that VTD is a highly active and relatively well tolerated regimen. The combination is used in the relapse setting, as well as fi rst line, consolidation and maintenance. In this protocol, the starting doses of Velcade and Thalidomide are relatively high and the duration of treatment is long. We will assess the superiority of VTD over TD in the relapse setting as well as its toxicity. Objectives: Chronic lymphocytic leukemia (CLL) is susceptible to well characterized graft-versus-leukemia effects. The leukemic clone is eradicated at the molecular level in >50% of patients. Yet, up to 50% of patients have persisting disease or relapse after allogeneic hematopoietic cell transplantation (HCT). The application of donor lymphocyte infusions (DLI) appears to be attractive in this situation. The aim of this retrospective analysis is to study the long-term effects of DLI in patients with CLL. Methods: Data from patients with CLL who received allogeneic HCT between 1997 and 2008 and had received at least one DLI were included. The outcome of DLI was analysed in two situations. 1) "pre-emptive" DLI prior to frank relapse or disease progression 2) "therapeutic" DLI given after documented relapse. Baseline and follow-up data were downloaded from the EBMT database.

Results: 217 patients fulfi lled the inclusion criteria. Major characteristics were median age 52 years, a median of three prior chemotherapy regimens, 71% matched sibling donor HCT, 74% reduced intensity conditioning, 85% received peripheral blood stem cells. ATG or alemtuzumab was used in 37%, ex vivo T-cell depletion (TCD) in 12%, no TCD in 50% of the patients. The median follow-up after the fi rst DLI was 20 months. In the cohort of 109 patients who received DLI prior to relapse, 38 patients received more than one DLI, 21 patients more than two DLI, 11 patients more than three DLI. Grades II to IV acute GVHD was reported in 9 out of 35 patients (26%) with informative data. PR or CR was documented in 10 out of 20 patients (50%). Overall survival and progression-free survival 5 years after the fi rst DLI was 42% (95% CI, 28% to 56%) and 22% (95% CI, 12% to 22%) respectively in this group of patients. 108 patients received DLI as treatment of relapse. Among these, 42 patients received more than one DLI, 16 patients more than two DLI, 5 patients more than three DLI. The overall response rate was 38% (13 out of 34 informative patients). Nine out of 50 patients (18%) experienced acute GVHD grades II to IV. In this group the 5-year overall survival after the fi rst DLI was 21% (95% CI, 9% to 33%). Only 3 patients had a follow up of more than 5 years. Conclusion: In patients with CLL donor lymphocyte infusions appear to have only moderate long-lasting activity. Further investigation to delineate factors associated with improved outcome is warranted. , 3 centres 8%, 5 centres 7.5%, 2 centres 6%, 5 centres 5%. 3 centres using 10% DMSO washed cells prior to infusion although another 9 centres also washed cells in occasional patients mostly less than 10% of their patients. 43 centres added additional agents to the freezing mixture, mostly albumin (25), hydroxy ethyl starch (6), heparin (8) and tissue culture medium (5). The median amount of DMSO given per centre per patient was 20 g, although the upper limit set by the centre was often considerably higher. 75% of centres did not use any delay between bags of stem cells and the median duration of infusion was 22 minutes. Side effects were defi ned using NCI criteria (version 3.0) and initially results were analysed by each centre. Patients in centres who used washed cells or 5% DMSO experienced less nausea and vomiting and 'severe other' effects, although hypotension and hypertension did not seem to be affected (p = 0.026, 0.014, 0.624 and 0.208 respectively). On a centre basis, the use of an upper limit to the amount of DMSO which could be given (70 g or the amount given ≤20 g versus >20 g) did not result in any reduction of any of the groups of side effects. Similarly when the amount of DMSO given and the duration of DMSO infusion were compared on a centre-basis, no signifi cant differences were found. Further analysis will be undertaken using individual data where appropriate and the results presented. Introduction: CMV infection and disease still remain serious and frequent complications after HSCT. Both morbidity and mortality have been reduced by prophylactic and pre emptive strategies based on biological tests and on more or less effective anti-viral drugs. However, in absence of comparative studies, there is no consensus for diagnosis, prophylaxis and treatment. Then we conduct a survey to assess the current CMV management for patients less than 18 years in EBMT centres. Materials and methods: In December 2009, a 40-item questionnaire about diagnostic tests, monitoring schedules, prophylactic and pre emptive strategies and treatment modalities was send to centres. Defi nition of treatment failure, recourse to drug resistant CMV mutant research, practice of cell therapy and combined anti-CMV therapy were also tested. Results: By the 31st December '09, we have received 35 responses from 14 countries. A second shipment will send in order to increase the number of responding centres. As expected, whatever the considered aspect i.e. monitoring, prophylactic or pre-emptive applied strategy, used drug in front or second and more line therapy etc. none centre report same procedure than another. All results will be presented and discussed during PDWP meeting in Vienna. Then, some proposals will be made regarding studies comparing results in different centres with "opposite" procedures. The fi nal goal of this work is to collect enough data to build consensual procedure. Here we describe the results of the fi rst 87 patients. All cases were confi rmed by PCR in nasopharyngeal or bronchoalveolar lavage samples. The fi rst case was diagnosed on July-9. Patients' characteristics are shown in Table 1 . Eight cases were considered nosocomial infections. In SCT patients, the median time from transplant to infl uenza diagnosis was 588 days (7-6155). S-OIV characteristics are given in Table 2 . The most frequent symptoms were: fever (86%), cough (84%), rhinorrhea (51%), odynophagia (26%), myalgia (22%), headache (11%) and dyspnea (11%). Diarrhea was rare (2%). Of those who presented with upper respiratory tract infection alone, 9 (14%) had progression to pneumonia. 30% had pneumonia, with no difference between SCT patients (31%) and non-SCT patients (28%). Fifteen out 17 pneumonias in SCT patients occurred in patients during the fi rst postransplant year or later but under immunosuppressive treatment. Five patients were transferred to the intensive care unit (ICU). Four out of 5 of the most severe cases (admitted to ICU or fatal course) had lymphopenia (<500/ mm 3 ) compared with 26% of those with less severe forms (P 0.010). Compared with adults, children had more pneumonia (39% vs. 28%) although the difference was not statistically signifi cant (P 0.3). Only one of the vaccinated patients against seasonal infl uenza were diagnosed with pneumonia compared with 41% who were not vaccinated (P .005). All patients except one were treated with oseltamivir for a median of 5 days (5-27). Zanamavir was added (in combination or sequentially) in 4 cases. No major adverse events relating to anti-infl uenza treatment were reported. Summary: The Novel S-OIV is a serious disease in STC and oncohematological patients with a high incidence of pneumonia (26%) and signifi cant mortality (3%). Diarrhea was not a frequent symptom. Lymphopenia was linked to the most severe forms. Vaccination for seasonal infl uenza might protect against the development of pneumonia. Nevertheless, the severity of new pandemic infl uenza seems to be similar to seasonal infl uenza in this patient population.

Background: EBV-PTLD (Epstein-Barr Virus-related Post-Transplant Lymphoproliferative Disorder) is a rare but serious complication after HSCT and number of patients at risk is increasing over time. Available data do not refl ect general practice of diagnosis and treatment of this complication. Objective: Assessment of the current strategy of diagnosis and preemptive use of rituximab for EBV-PTLD in EBMT transplant centers (TC). Methods: In 2009 survey, data regarding EBV strategy from 74 participating TC were registered on specifi c forms and centrally analyzed. Responses from 10 centers were excluded due to lack of specifi c strategy. Results: Regular monitoring for EBV after HSCT is done by 47/64 (73%) TC, and in 7/64 (11%) TC is related to clinical situation. The monitoring is performed in all alloHSCT patients in 32/47 (68%) TC, while in remaining 22 TC in following subgroups: MUD (17/22), T-depletion in vitro (8/22), T-depletion in vivo (21/22), family mismatched (14/22), cord blood transplants (16/22), and in single centers in patients with SAA, EBV mismatch, or after autoHSCT for autoimmune disorders. Quantitative EBV-DNA by PCR is performed in 62/64 (97%) centers. The assay is performed in: whole blood -50/64 TC (78%), plasma -9/64 (14%), lymphocytes -4/64 (6%), serum -1 center. The test is repeated twice a week in 8/64 TC (12.5%), once a week in 39/64 (60.9%), once per 2 weeks in 8/64 (12.5%), once a month in 5/64 (7.8%) and adjusted to risk in 4/64 (6.2%) TC. The monitoring for EBV reactivation after HSCT is performed for a period of: less than 3 months in 2/64 (3%), 3 months in 22/64 (34%), 6 months in 19/64 (30%), 12 months in 8/64 (12%) and adjusted to risk in 13/64 (20%). Rituximab as a pre-emptive therapy for EBV-PTLD is routinely administered in 51/64 (80%). The number of EBV-DNA copies as an indicator for preemptive therapy with rituximab was given by 60 TC, but varied between the centers, and were based also on symptoms and signs (Table) . Conclusions: EBV-PTLD strategy exists in most of the responding centers. Differential approach regarding indications for preemptive therapy is seen between centers: rituximab is administered as preemptive therapy in 80% of participating transplant centers. Objectives: The aim of this study was to analyze the clinical risk factors, donor and recipient cytokine gene polymorphisms associated with cytomegalovirus (CMV) infection within 100 days after allogeneic hematopoietic stem cell transplantation (allo-HSCT). Methods: We studied in a total of 203 pairs of recipients and their donors, who underwent allo-HSCT at our center. We analyzed 12 single nucleotide polymorphisms (SNPs) in 6 pro-and anti-infl ammatory cytokines genes, tumor necrosis factor ( 92.6%of122 patients developed CMV positive antigenemia without disease, only 9 patients developed CMV disease (7 patients with pneumonia and 2 patients with enteritis). There was a higher incidence of early CMV infection in transplantation with unrelated donors (66.4% vs. 51.7%, P = 0.043) and in recipients who were CMV seropositive before transplantation(72.9% vs. 54.9%, P = 0.018). The recipient's TGF-beta1-509 and + 869 genotypes were signifi cantly associated with the incidence of early CMV infection in both the unrelated transplantation cohort and sibling transplantation cohort, but the infl uence of the donor's TGF-beta1-509 and + 869 genotypes was signifi cant only in the sibling transplantation cohort. Multivariate analysis identifi ed two risk factors for early CMV infection: CMV seropositive recipients (RR: 1.712, 95%CI: 1.177-2.490, P = 0.005), recipients with TGF-beta1 + 869 C allele (RR: 2.225, 95%CI: 1.401-3.536, P = 0.001). Donors with TGF-beta1-509T allele was found to be less signifi cant factor (P = 0.094). Conclusion: Although CMV disease has been reduced in the era of antiviral prophylaxis and preemptive therapy, our fi ndings suggest the incidence of CMV infection remains high and provide the fi rst report of relationship between genetic background of donor and recipient to the risk of CMV infection in a Chinese population.

HSCT. Toll-like receptors (TLR) are essential components of the innate immune system. C3H/HeJ mice that lack functional TLR4 did not develop cystitis after intravesical instillation of E. coli. Individuals with the Asp299Gly polymorphism of the TLR4 gene showed a diminished infl ammatory responsiveness to endotoxin. Because of the requirement of TLR4 in infl ammatory response we hypothesized that TLR4 Asp299Gly reduces the risk of BK virus-induced HC after HSCT. 166 children (median age, 13 years) who underwent allogeneic bone marrow (n = 105) or peripheral blood stem cell transplantation (n = 61) in a single center and their respective donors were genotyped of TLR4 for rs4986790 (A896G, Asp299Gly) using TaqMan real-time polymerase chain reaction. The donor was HLA-matched unrelated in 57% of transplants or HLA-identical related in 33% of transplants. Conditioning regimen was myeloablative in all cases. Two forms of post-transplant immunosuppression predominated, cyclosporine A and methotrexate in 64% of transplants and cyclosporine A alone in 25% of transplants. The Asp299Gly polymorphism was present in 21 of the 166 patients (12.6%) and in 24 of the 166 donors (14.4%). Interestingly, we found a signifi cantly reduced incidence of BK virus-induced HC in patients with the Asp299Gly polymorphism (0% vs. 22.8%, p = 0.009). In addition, we observed a significantly reduced incidence of BK virus-induced HC in patients who were transplanted from a donor with this specifi c polymorphism (4.2% vs. 22.5%; p = 0.05). In ten of the donor-patient pairs the polymorphism was present in both individuals and no HC was observed. The occurrence of the TLR4 Asp299Gly polymorphism, in either recipients or donors, had no signifi cant impact on acute and chromic GVHD, relapse rate, bacterial and fungal infectious complications, transplant related mortality, and overall survival. In conclusion, Asp299Gly polymorphism of the TLR4 gene in the recipient and/or the donor is associated with a signifi cant decrease of BK virus-induced HC after HSCT in childhood. This study provides the fi rst evidence that the innate immune system through TLR4 signaling pathway seems to play an important role in the pathogenesis of BK virus-associated HC after HSCT. Design and methods: One hundred-seventeen patients, median age 52 (20-67) years, with various haematological malignancies transplanted between 1999 and 2007 entered the study. Eighty-seven recipients negative for HBV surface antigen (HbsAg), anti-hepatitis B core antigen antibodies (anti-HBc) and HBV-DNA with HBsAg negative donors were defi ned as at no risk of HBV reactivation whereas all the remaining 30 patients were defi ned as at risk. In accordance with the Italian national guidelines for immunocompromised hosts, patients at risk transplanted after 2005 received lamivudine to prevent HBV reactivation from conditioning up to 12-18 months after discontinuation of all immunosuppressive drugs whereas before 2005 no prophylaxis was given.

Results: Patients at no risk did not experience HBV reactivation/ hepatitis. Among patients at risk, HBsAg negative recipients from HBsAg positive donors (3/3), HBsAg positive recipients from negative donors (2/2) and 11/25 anti-HBc positive were treated with lamivudine. None developed hepatitis B after a median follow-up of 33 months (7-55). Hepatitis developed in 3 anti-HBc positive untreated patients conditioned with a reduced intensity regimen up to 19 months after discontinuation of immunosuppression and in none of those on prophylaxis. Conclusions: We observed: no risk of hepatitis B in recipients serologically negative for HBV, transplanted from HBsAg neg donors; effi cacy of lamivudine in controlling HBV reactivation in both HBsAg positive recipients and negative recipients from S38 HBsAg positive donors; a signifi cant risk of HBV reactivation in HBsAg negative/antiHBc positive recipients and effi cacy of prophylactic lamivudine in this setting and the importance of prolonged prophylaxis even after the discontinuation of immunosuppressive drugs.

Republic -a single-centre experience P. Hubacek (1,3) , D. Boutolleau (2, 4) , C. Deback (2, 4) , P.

Keslova (3) Despite the improvement of infection monitoring and antiviral treatments, Cytomegalovirus (CMV) infections remain an important cause of morbidity and mortality in allogeneic hematopoietic stem cell transplant recipients (alloHSCT). Aim was to investigate CMV resistance in paediatric and adult alloHSCT recipients and to clarify the CMV load in whole peripheral blood and target tissue in situation of CMV disease. Between I/2002 and XII/2009, we tested 17504 whole blood samples from 427 adults (median 22/patient) and 211 children (median 32/patient) after alloHSCT. Additional 1348 biological non blood samples were tested too.Samples were tested for CMV and albumin gene quantity by RQ-PCR and results were normalized to 100000 human genome equivalents. Ganciclovir as fi rst-line therapy was initiated when CMV load exceeded 1000 normalized viral copies (NVCs) and switched to foscarnet or cidofovir in case of none response.Clinical resistance was suspected after 2 weeks of unsuccessful well-conducted treatment.Resistance was studied using UL97 and UL54 gene sequencing.In the patients deceased in consequence of CMV infection,DNA quantity in the tissue was tested too. Virostatic treatment was started in 142 adults (33%) and 63 (30%) children.Clinical resistance was observed in 58 adults and 25 children.Known genotypic CMV resistance was proved in 11 of them (5.4% of treated) detecting mutation C592G, A591V, A594V, L595S, L595F, G598S and del597-599 in UL97 and del981-982, N408K, V715M, P522S in UL54). Natural polymorphisms and unknown phenotype changing were also detected in both genes. Despite the virostatic treatment, symptomatic CMV disease was observed in 15 adults and 10 children. Ten patients deceased in consequence of CMV infection,mainly combined with GvHD. Typical pathological signs of CMV infection and antigen detection in the tissue were found in 1 patient only.Despite this median of CMV quantity in mainly affected lung tissue was 9534 NVCs (range 1029-384456) while median in the whole blood was only 1065 NVCs (range 183-89072).

Quantifying of CMV DNA is very useful in the treatment of CMV infection. Unfortunately our observation suggests the limits in case of CMV disease. Genotypic evidence of CMV resistance proved to be very useful in the improvement of therapeutic management of patients. Further studies are required to ascertain the true nature of the novel mutations described within UL97 and UL54 same as compartmentation of CMV infection. Supported by MZ0FNM2005, MSTM0021620813.

CMV reactivation often occurs in patients after allogeneic HSCT. Recently, PDGFR-alpha was suggested as the cellular receptor for human CMV (Soroceanu et al., Nature, 2008) . PDGFR-alpha binding and activation of a downstream PI3kinase signaling pathway was essential for CMV propagation. Many TKIs, such as imatinib, effi ciently inhibit PDGFR-alpha at concentrations readily achievable in patients. The aim of this study was to retrospectively assess a possible link between CMV reactivation and TKI therapy. . Also, at 2 years, the relapse incidence was not signifi cantly associated with the type of RIC regimen: 37±3%, 34±2% in the TBI-based RIC and chemobased RIC groups respectively (p = 0.17). Finally, NRM was also comparable between both groups (21±3% and 20±2% in the TBI-based RIC and chemo-based RIC groups respectively; p = 0.60). Despite its retrospective nature, results from this large study suggest that RIC allo-SCT is a valid option for AML patients not eligible for standard allo-SCT. The overall outcomes (LFS, NRM and relapse) appear not to be signifi cantly different between AML patients in CR1 receiving a low dose TBI-based RIC allo-SCT versus those receiving a chemotherapy-based RIC allo-SCT using an HLA-identical sibling donor. Patients with ALL who relapse after allogeneic HCT have a grim prognosis and the optimal salvage therapy remains unknown. Aim of this retrospective analysis was to report the outcome of relapsed ALL after allogeneic HCT, to identify factors which have a prognostic signifi cance and to evaluate whether any relapse treatment strategy was associated with better outcome. For this study the EBMT database was searched for: 1) adult patients with ALL aged >18 y at HCT; 2) transplanted in remission; 3) with hematological relapse after an allogeneic HCT; 4) HLA identical or MUD HCT performed from 1995 to 2006; 5) suffi cient MED-C data. 465 adult pts (median age 32 y, 18-66) met these eligibility criteria. Diagnosis was 76% B-lineage ALL, 21% T-ALL, 33% Ph + ALL and previous allograft (68% were HLA identical and 32% MUD) was performed in remission (CR1 65%, CR2 32%, CR3 3%). The median interval from transplant to relapse was 7 months (0.8-52). Post-relapse interventions were: no further treatment in 13% pts, 43% pts received chemotherapy at a median of 5 d after relapse , 23% pts received DLI at a median of 45 d post-relapse (4-344) and 20% pts underwent a second transplant at a median of 71d after relapse (7-286). 61% of Ph + ALL pts in addition received a TKI inhibitor. Overall, 40% pts reached a subsequent CR. After a median follow up of 56 months (3-141), the estimated 1-, 2-and 3-year overall survival (OS) was 32%, 19% and 10%. The Table below shows the statistically signifi cant parameters associated with 2-year OS. In the multivariate analysis, OS was associated with disease status at HCT (CR2 + vs. CR1 p = 0.005, HR: 0.57), time to relapse after HCT (>vs< median 7 months, p<0.0001, HR: 2.27 ) and number of peripheral blood blasts at relapse (>vs<10%, p<0.0001, HR:0.49), but it was not signifi cantly infl uenced by cytogenetics (Ph + vs. other), donor type (sibling vs. unrelated), stem cell source (BM vs. PB), type of conditioning (TBI vs. chemo) or type of treatment for relapse (DLI or HCT-based therapy vs. therapy without cell infusion). This study highlights the extremely poor outcome of ALL pts who relapse after allogeneic HCT. Better outcomes were found in patients transplanted in CR1, with late relapse after HCT (>7 months) and lower tumor burden at relapse (<10% of PB blasts). These factors should be taken into consideration when discussing further therapeutic options for these patients. Introduction: Prognosis and survival in AML patients is associated primarily with cytogenetic abnormalities but also with molecular markers. The internal tandem duplication (ITD) of the FLT3 gene is one such marker associated with poor prognosis. We report here an analysis of the impact of allogeneic SCT on clinical outcome of cytogenetically normal AML patients according to their FLT3-ITD status. Patients and methods: FLT3-ITD mutation was analyzed by capillary electrophoresis after PCR amplifi cation of material frozen at the time of diagnosis from a total of 116/277 patients with normal karyotype entered between 1997 and 2008 into two OSHO studies (AML 96 and AML 2002). SCT was performed after conditioning with cytoxan and 1200 cGy TBI followed by GvHD prophylaxis with cyclosporine and methotrexate. Related and unrelated donor search was initiated as soon as possible and patients allocated to groups with or without a donor. A total of 46 patients were allocated to the donor group and 70 patients to the no donor group. Results were analyzed as intention to treat. Results: Event free survival (EFS) of all patients was 38% after 5 years. In patients (n = 46) allocated to receive HCT from an allogeneic donor, EFS was 44% compared to 33% in those allocated to the non-transplant treatment (n = 70). As previously described, detection of a FLT3-ITD mutation had a negative impact on EFS at 5 years, which was 25% in FLT3-ITD positive and 46% in FLT3-ITD negative patients (p = 0.06). Subsequently, EFS was analyzed according to FLT3 status and post-remission treatment. In the FLT3-ITD positive group, EFS was inferior in patients without donor (19% at 5 years) compared to those with a donor (36% at 5 years). This difference was due to a decreased relapse incidence in the donor group (48%) compared to the no-donor group (80%). However, a difference in EFS was also observed in FLT3-ITD negative patients in the donor (50% at 5 years) compared to the nodonor group (43% at 5 years). This difference was also associated with a reduced relapse incidence in the donor group (26%) versus the no-donor group (55%; p = 0.003). Conclusion: In conclusion, allogeneic HCT improves EFS in AML patients with a normal karyotype by reducing relapse incidence irrespective of the FLT3-ITD mutation status. The median follow-up of living patients was 20 months (range, 1 to 53 months). At 3 years the overall survival (OS) was 38% and progression-free survival (PFS) was 36%, while the incidence of non-relapse mortality (NRM) and relapse were 29% and 34%. We performed multivariate Cox regression modelling for OS and PFS while competitive event statistics were applied for NRM. In order to adjust for a potential selection bias the HCT-specifi c comorbidity score, Karnofsky performance status, delay between diagnosis and HCT and age were forced into the multivariate model. We performed a complete case analysis based on 223 patients. The hazard ratios (HR) of a matched or a mismatched UD compared to a sibling donor were 1.1 (95% CI, 0.7 to 1.7) and 1.3 (95% CI, 0.6 to 2.0) for OS and 1.3 (95% CI, 0.8 to 2.0) and 1.3 (95% CI, 0.8 to 2.1) for PFS. In interaction analyses we did not identify signifi cantly differing effects according to age and disease stage. Of note, having a matched UD did not signifi cantly affect the risk of NRM (HR=0.8; 95% CI, 0.4 to 1.6).

Our results confi rm the current practice in high risk AML to deliver allogeneic HCT from matched UD and SIB. An updated analysis will be presented at the meeting. (14) (

This phase-III randomized EBMT-intergroup trial studied the impact of a consolidating haematopoietic stem cell transplantation (autoHSCT) vs. wait and watch for patients with CLL in Binet stage A progressive, B or C, in CR, nodular PR or VGPR after fi rst or second line therapy. The primary objective was to show that autoHSCT increase the 5-year progression-free survival (PFS) by 30%. Here we present a fi rst analysis based on 80% of expected follow-up forms. Results: Between November 2001 and July 2007, 223 patients were enrolled (SFGM-TC/FCLLG n = 99, MRC n = 63, GCLLSG n = 36, SAKK n = 10, other EBMT centers n = 15). There were 74% males and 26% females. Binet stages were progressive A 13%, B 67%, C 20%; 59% were in CR, and 41% in very good or nodular PR. Of note, SFGM-TC/FCLLG included only patients in CR. Eighty three percent of the patients were enrolled in 1st, and 17% in 2nd line treatment. Patients were randomized between group 1 (autoHSCT n = 112) and group 2 (observation n = 111) after an induction treatment which was left at the discretion of the investigators. Median PFS was 26.6 months (18.3-35) in the observation group and 50.1 months (40-60.1) in the autoHSCT group; the 5-year PFS was 29% and 42%, respectively (p<0.001). Accordingly, the 5-year relapse incidence was 54% vs. 70%; p<0.001. The Cox modeling for randomization arm, Binet stage, disease status, line of treatment, contributing group (country), and the interaction between randomization arm and contributing group confi rmed that autoHSCT signifi cantly improved PFS (HR 0.45 [0.30-0.66] p<0.001). The benefi cial effect of autoHSCT was stable over all contributing groups. At 5 years, the probability of OS was 83% and 82% for autoHSCT and observation, respectively; p = 0.81. Signifi cant differences in terms of non-relapse death were not observed. In conclusion, in patients with CLL in fi rst or second remission, consolidating autoHSCT reduces the risk of progression (PFS) by more than 50%, but has no effect on overall survival. Further analyses on variables affecting the outcome are underway and results from a quality of life study on both groups are awaited.

Improved long-term outcome after highly purifi ed peripheral blood CD34 + cell transplantation from

Here, we report about the long-term follow-up results of a prospective phase II study in patients after transplantation of highly purifi ed peripheral blood CD34 + stem cells from HLA-identical sibling donors for CML in 1. CP. A total of 58 pts with a median pretransplant EBMT risk score of 2 (range 1-4) have been included in this study. All patients received a myeloablative conditioning regimen with TBI, Cyclophosphamid with/or without thiotepa and ATG, but without any prophylactic immunosuppression post transplantation. One patient received an unmanipulated graft due to poor CD34 + donor cell mobilization, while two patients were successfully retransplanted with an unmanipulated graft from the primary donor after secondary graft failure.

Of the 58 pts, 56 were eligible for the application of DLI, but 6 pts did not receive DLI due to sustained molecular remission and complete chimerism. Thirty-two pts (64%) received DLI because of increasing BCR-ABL transcript levels or hematologic relapse, and 18 pts (36%) as programmed T-cell add-back. The median starting dose was 0.33 (0.01 -10) x 10 6 CD3 + cells per kg with a median maximum dose 3.3 (0.17 -100) × 10 6 CD3 + cells per kg. DLI alone induced a lasting reduction of median BCR-ABL transcript levels (BCR-ABL/GAPDH ratio) of more than 3 log10 and the estimate of being in a complete molecular remission at 8 years is 76% ± 5%. 12 pts. (24%) did not respond to DLI alone, but 8 of these pts. attained a cytogenetic and molecular response by imatinib (n = 9), dasatinib (n = 1) and/or interferon treatment. Four patients were retransplanted with an alternative donor. The cumulative risk of grades II-IV acute GvHD is 15% ± 5% for all study pts, and the risk of chronic GvHD is 25% ± 6%, respectively. After a median follow-up period of 88 months (range 6 -125) for all patients, the cumulative 10-year survival estimate is 91,3% ± 4%. Including all therapies for molecular relapse after transplant, 50 of the 58 included patients (86%) were in molecular remission at the last time point of observation. Causes of death (n = 5) were disease progression, secondary malignancy, liver failure, septicemia, and systemic capillary leak syndrome in one patient each.

In conclusion, the concept of highly purifi ed PB CD34 + cell transplantation in conjunction with adoptive DLI is associated with a particularly low risk of non-relapse mortality and allows induction of lasting molecular disease control in the majority of 1. CP CML.

Allogeneic stem cell transplantation (ASCT) after reducedintensity conditioning (RIC) has become a reasonable treatment option for patients with advanced myelofi brosis. The role of characteristic molecular genetic abnormalities such as JAK2V617F on outcome of ASCT is not yet elucidated. In 139 out of 162 myelofi brosis patients with known JAK2V617F mutation status who received ASCT after RIC the impact of JAK2 genotype, JAK2V617F allele burden and clearance of mutation after ASCT was evaluated. A non-signifi cant higher treatment-related mortality (31% vs. 19%, p = 0.1) and relapse incidence (30% vs. 21%, p = 0.2) was noted in patients harbouring JAK2 wild-type, resulting in a decreased overall survival in a multivariate analysis (HR 2.23, p = 0.007). No signifi cant infl uence on outcome was noted for the mutated allele burden analyzed either as continous variable or after dividing in quartiles. Achievement of JAK2V617F negativity after ASCT was signifi cantly associated with a decreased incidence of relapse (HR 0.22, p = 0.04). In a Landmark analysis, patients who cleared JAK2 mutation level in peripheral blood 6 months after ASCT had a signifi cant lower risk of relapse (5% vs. 35%, p = 0.03). We conclude that JAK2V617F mutated status but not allele frequency resulted in an improved survival and rapid clearance after allografting reduced the risk of relapse.

Early autologous stem cell transplantation in poor-risk chronic lymphocytic leukaemia. Long-term results from the GCLLSG CLL3 trial focusing on incidence and type of secondary malignancies F. McClanahan (1) (2), P. Dreger (1) (

Introduction: As previously reported, early autoSCT as conducted in the CLL3 protocol is a feasible therapy option for younger patients (pts). Purpose of the present analysis was to study the impact of FISH karyotype and IGHV mutational status on long-term PFS and OS and the incidence and type of secondary malignancies occurring in this trial. Trial design and patients: The design of the protocol has been extensively described previously. From 1996 to 2002, 216 pts were registered. As 47 cases had to be excluded due to screening failure (n = 21), withdrawn consent (n = 19) or other reasons (n = 7), 169 pts were eligible for the current analysis. Male to female ratio was 5:1 and the median age at diagnosis was 51 yrs (range 27-60). Results: SCT was performed in 131 pts (78%) at a median time of 17 months (range 4-159) after initial diagnosis, whereas 38 pts did not proceed to SCT (mobilization failure n = 14, disease progression n = 4, early death n = 3, pts preference n = 6, unknown reasons n = 11). At a median follow-up of 99 months (range 4-137), median OS of all 169 pts was 10.5 yrs (10.5 yrs after SCT, 6.1 yrs without SCT, HR 0.26, 95% CI 0.13-0.54, p<.0001). Median PFS was 6.3 yrs (6.8 yrs with SCT and 4.8 yrs without, HR 0.39, 95% CI 0.23-0.67, p = 0.0007). Diagnostic samples for assessment of IGHV mutational status were available for 143 (85%). An unfavorable IGHV rearrangement was present in 104 pts (73%), and was associated with signifi cantly worse PFS (p = .0001) and OS (p = .017). FISH was possible in 160 pts. Whereas PFS (p <.0001) and OS (p <.0001) were considerably reduced in pts with del 17p-, there were no signifi cant differences between the other subsets. Altogether, 20 secondary malignancies were observed, with 6 cases of t-MDS/ t-AML, translating into a 10-year incidence of 20% (95% CI 11-30%), with no signifi cant difference among pts treated with and without SCT (p = 0.68). However, all cases of t-MDS/ t-AML occurred after SCT, yielding a 10-year incidence rate of 9% (1-18%). Overall survival after secondary neoplasms was 22 months , with no difference between t-MDS/AML and other malignancies. Conclusions: Unmutated IGHV remains an adverse prognostic factor. While del 17p-is associated with a very poor outcome, autoSCT may eliminate the unfavorable impact of del 11q-seen with conventional therapy. Secondary neoplasms are a serious problem after early autoSCT, but do not appear to occur more frequently than reported after SCT for other diseases. Reduced intensity conditioning (RIC) has reduced non-relapse mortality (NRM) associated with allogeneic HSCT in Hodgkin Lymphoma and relapse is now the commonest cause for treatment failure. However, there is little published evidence to guide management of relapse. We performed 81 RIC allografts for HL (44 matched related, 37 unrelated donor) incorporating T cell depletion with alemtuzumab. 35 (43%) were primary or salvage-refractory, 47 (58%) had failed a prior autograft, and the median number of prior treatment lines was 5. Patients were monitored by PET-CT and relapse defi ned by recurrence or progression of FDG-avid lesions in sites of prior disease. If FDG-avidity occurred only in new sites, relapse was confi rmed by biopsy if accessible (n = 7), otherwise an interval scan was performed at 6-8 weeks. Following cyclosporine withdrawal and in the absence of GvHD, patients received dose-escalating DLI for mixed chimerism (MC) or progression. The 3 year CI of NRM was 17% and of relapse was 45%. 44 patients received a total of 87 DLIs (median 2, range 1-5). Treatment for MC alone (n = 22) resulted in conversion to full donor in 15/20 evaluable patients, continuing improvement in 2, and 3 remained stable. Only 1 of these relapsed (CI 7% at 3 years from initial DLI), receiving further DLIs following chemotherapy. Relapse was treated in 23 patients with DLI either alone (n = 13) or following debulking chemotherapy (n = 10). Median maximal doses were 1 × 10 7 in unrelated and 1 × 10 8 CD3 + T cells/kg in related donor transplants. Complete responses were seen in 12 and partial responses in 5 of 22 evaluable patients (response rate 77%), and did not differ signifi cantly according to donor source (10/13 MRD vs. 7/9 UD). The majority of responders developed GvHD (5 Gd III-IV acute, 4 extensive chronic). 7 CR are maintained at a median of 4.6 years from last DLI (5 without prior chemotherapy), 3 died of GvHD-related complications, and 2 progressed (at 1.6 and 2.3 years). 3/5 with PR progressed. The projected 3 year OS and PFS from relapse are 64% and 54% in this group. The OS and current PFS for the entire cohort of 81 patients are 61% and 53% at 4 years. In conclusion, the data demonstrate favourable long term survival in this heavily pretreated HL cohort, a strikingly low relapse incidence in those receiving DLI for MC, and the ability of DLI guided by PET-CT to induce high response rates and durable salvage in the setting of relapse post T cell depleted transplantation.

A recent update of 3 consecutive prospective trials with high-dose therapy and autograft, without or with rituximab, as primary treatment for advanced-stage follicular lymphoma shows a sizeable group of patients surviving in continuous complete remission C. Tarella, M. Ladetto, F. Benedetti, U. Vitolo, A. Pulsoni, C. Patti, V. Callea, A. Rambaldi, A. Piccin, L. Devizzi, M. Musso, E. Iannitto, P. Spedini, A.M. Liberati, F. Ciceri, A. Gallamini, F. Rodeghiero, G. Gini, A. De Crescenzo, F. Di Raimondo, A. Levis, T. Chisesi, T. Perrone, D. Rota Scalabrini, G. Rossi, A.M. Carella, G. Parvis, I. Majolino, R. Passera, M. Ruella, A. Pileri, A.M ) or in the proportion of patients with systemic symptoms or with elevated serum lactate dehydrogenase. The median time from the diagnosis to ASCT was also comparable as well as the number of previous treatment lines and the use of rituximab before ASCT (49% vs. 53%). The elderly patients were less often in fi rst complete remission at the time of ASCT (42% vs. 53%, p = 0.003). The most common conditioning regimen used was BEAM (62% in the elderly vs. 58%) followed by cyclophosphamide plus total body irradiation (14% vs. 14%). Non-relapse mortality (NRM) was comparable at 3 months (3.7% in the elderly vs. 2.1% in younger patients) and at 1 year (4.2% vs. 2.9%) but was higher in the elderly patients at 5 years (8.6% vs. 3.2%, p = 0.04). The most common causes of NRM were infections (53% vs. 44%). With a median follow-up of 21 months for the elderly patients and 25 months for the younger patients, the risk of relapse was 57% and 54% (n.s), respectively. Progression-free survival was also comparable (38% vs. 40% at 5 years) without any plateau.

Overall survival was worse in elderly patients (60% vs. 69% at 5 years, p = 0.01). Autologous stem cell transplantation is feasible in selected elderly patients with MCL. Early NRM is low and comparable to younger patients. Risk of relapse and progression-free survival are also comparable but overall survival is worse in the elderly patients. Continuously increasing relapse risk indicates the need for improvements in pre-and posttransplant strategies in order to improve long-term outcome in this lymphoma type.

Phase II study of radioimmunotherapy with yttrium- (1) Allogeneic hematopoietic cell transplantation (HCT) using reduced intensity conditioning (RIC) offers a potential curative therapy to patients with advanced indolent NHL. Combined use of radioimmunotherapy (RIT) with RIC may increase anti-lymphoma activity of RIC while avoiding additional toxicities. Forty patients have been enrolled in a multicenter phase II study of allogeneic HCT using RIT with yttrium-90-ibritumomab tiuxetan (Y90-CD20, Zevalin) with 0.4 mCi (15 MBq)/kg on day -14 combined with RIC using fl udarabine (30 mg/m² day -4 to -2) and 2 Gy TBI (day 0) followed by allogeneic HCT from matched related or unrelated donors. GVHD-prophylaxis consisted of cyclosporine and mycophenolate mofetil. Diagnoses were follicular lymphoma (FL, n = 17), chronic lymphocytic leukemia (CLL, n = 13), mantle cell lymphoma (MCL, n = 8), marginal zone lymphoma (n = 1) and immunocytoma (n = 1). Median age was 56 (range, 35-68) years. PBSC grafts were either from matched related (n = 13) or matched unrelated donors (n = 27). All patients were "high risk" with refractory disease or relapse after preceding autologous HCT. Engraftment was rapid and sustained with no graft rejections. Median time to >500 granulocytes/μL was 13 (range, 0-69) days, and to >20000 platelets/μL 4 (range, 0-69) days. In 13 patients platelets were never <20000/μL and in 6 patients granulocytes never <500/μL, illustrating the nonmyeloablative intensity of the conditioning regimen. TRM in the fi rst 100 days was 10% (n = 4) and overall 40% (n = 16). No additional toxicity due to RIT compared to our previous experience with the same RIC as single modality was observed. Incidence of grade II-IV GVHD was 45% (II = 3, III = 12, IV = 3). To date, chronic GVHD occurred in 19 patients (48%, limited = 8, extensive = 11). Deaths occurred due to infections = 8, GVHD = 7, relapse = 2 and cerebral bleeding = 1. 22/40 (55%) of all patients are alive with a median follow up of 672 (range, 251-1086) days, resulting in a Kaplan-Meier estimate 2 year survival of 51% for all, and 67% in FL, 49% in CLL and 38% for MCL patients. Risk factors for decreased overall survival and TRM in multivariate cox regression analysis were non-FL histology (p = 0.032 and p = 0.024) and aGVHD >grade 3 (p = 0.001).

In conclusion, a combination of RIT with RIC is feasible with no additional toxicity due to RIT and with stable engraftment in all patients. Disease response and overall survival seems promising even in this elderly and heavily pretreated cohort of patients.

Allogeneic stem cell transplantation after autologous haematopoietic stem cell transplantation relapse in aggressive lymphoma patients: fi nal report of a retrospective GITMO study L. Rigacci, A. Bosi, B. Puccini, P. Corradini, L. Castagna, N. Cascavilla, G. Milone, A. Bacigalupo, R. Scimé, G. Specchia, A. Rambaldi, P. Leoni, F. Ciceri, A. Levis, S. Guidi, B. Bruno, R. Oneto, R. Fanin on behalf of GITMO Autologous hematopoietic stem cell transplantation (AHSCT) has been shown to be an effective therapy for patients (pts) with aggressive lymphoma. Pts who relapse after an AHSCT have a very poor prognosis. Allogenic hematopoietic stem cell trasplantation (AlloHSCT) has shown to be effective in the rescue of indolent lymphoma pts relapsed after conventional therapies. According to this data we have retrospectively analized all pts with diagnosis of aggressive lymphoma in the GITMO data-base who have performed an AlloHSCT after an AHSCT relapse. From 1995 to 2008, 181 pts were selected from the GITMO data-base. One of the principal objective in this fi nal report was the completeness of the data. Data missing were: acute GVH was evaluable in 90% of pts, chronic GVH in 70% of pts, response after allogeneic transplantation in 66%, condition at the moment of allotransplant in 94%, therapy pre-Allo-HSCT in 77% of pts. The other characteristics were evaluable in all patients. One-hundred and one were male (56%), 160 presented a diagnosis of DLBCL. The stem cell donor was related in 115 pts, the stem cell source was peripheral blood in 151 pts and bone marrow in 30. The conditioning regimen was conventional in 55 pts and reduced intensity (RIC) in 126 pts. The median time between AHSCT and allo-HSCT was 12 months. Ninethy-seven pts (54%) obtained at least a partial remission before allo-HSCT, 74 (41%) were treated with active disease and in 10 cases the data was not available. After allo 78 pts (43%) obtained a CR and 10 a PR with an ORR of 49%, 69 pts (38%) showed a rapid progression of disease. Ninethy-seven pts died, 46 due to disease and 51 to treatment related mortality (TRM). Acute graft versus host disease was recorded in 61 pts and a chronic one in 42 pts. With a median follow-up period of 16 months (4-28) from allo the OS was 56% and after a median period of 12 months (6-17) from allo the PFS was 49%. In multivariate analysis the only factors which affected PFS were the status at allo and a CR after allo transplant. This conclusive retrospective analysis confi rms that the only one parameter affecting either OS or PFS was the response status at the moment of all-HSCT and does not confi rm that RIC could permit to obtain better results in aggressive lymphoma. Probably a myeloablative conditioning should be reconsidered in pts with aggressive disease because of the slow-acting graft versus lymphoma effect is overriden by the rapid growth of the tumor. patients (5%), II in 100 (28%), III in 80 (23%), and IV in 155 (44%). Forty-nine percent of the patients had B symptoms, 7% bulky disease, and 54% extranodal involvement. Karnofsky performance status < 80 was reported in 72 patients. Treatment following relapse consisted on conventional chemotherapy and/ or radiotherapy in 294 patients (64%), second ASCT in 35 (8%), and allogeneic stem cell transplantation in 133 (29%). A significant higher proportion of patients, age < 50 years, received a second transplantation in comparison to older patients (39% vs. 16%, p =.006). After a median follow-up of 49 months (range 1-150), the OS from relapse after ASCT was of 55% at 2 years and 32% at 5 years. In multivariate analysis, independent risk factors for OS were early relapse (< 6 months), stage IV, bulky disease, poor performance status, and age > 50 at relapse. In conclusion, according to the EBMT database, most HL patients relapsing after ASCT were treated with chemo-radiotherapy approaches and some of them with a second transplantation. Those patients in a good performance status presenting with a localized late relapse seem to be the ones showing the best prognosis. Hematopoietic cell transplantation (HCT) is the therapy of choice for a variety of malignancies. HCT provides disease benefi t through both the high-dose conditioning regimen and an allogeneic graft versus tumor effect (GVT). However graft-versus-host disease (GVHD) still remains a major obstacle. It has been proposed that host conditioning may not only be crucial in the activation of alloreactive T cells but also determine acute GVHD organ manifestation. We wanted to investigate how the host-conditioning regimen affects the host target tissues in terms of infl ammatory cytokines and their role in donor T cell recruitment. Utilizing a non invasive bioluminescence imaging (BLI), and fl ow cytometry in a murine allogeneic hematopoietic cell transplantation (allo-HCT) model, we compared lethally irradiated (8Gy) vs. non-irradiated Balb/c wild type, and Balb/c Rag-/-cGC-/-(DKO) (H-2d) mice that received allogeneic luciferase + FVB/N T cells (H-2q). We observed that the proliferation (BLI, CFSE), acquisition of activation markers (CD25, CD44, CD69) and homing receptors (a4b7, aEb7, P-selectin ligand, E-selectin ligand) by alloreactive T cells occurred independently whether allogeneic recipients were conditioned or not conditioned before allo-HCT. As T cell recruitment may have occurred as a result of alterations of the milieu infl ammatory cytokines in GVHD target tissues, we compared the cytokine profi le in conditioned vs. non-conditioned hosts. At days 3 and 6 after allo-HCT, host tissues were analyzed for a TH1/TH2/Th17a cytokines from the target tissues; liver, large bowel, small bowel, peripheral blood and a non target tissue: kidney. At day 3, high levels of INF-g, TNF and IL-6 were detected in the Balb/c WT conditioned compared to the non-conditioned host in all target tissues and most markedly in blood and the large bowel. More importantly both the Balb/c Rag-/-cGC-/-(DKO) conditioned and non-conditioned host displayed very high levels of the infl ammatory cytokines, with Balb/c WT and Balb/c Rag-/-cGC-/-(DKO) conditioned hosts displaying higher levels. Similar results with a reduced expression of the cytokines were observed at day 6 indicating that the cytokine storm peak was maybe at day 3. In summary host conditioning is not a requirement for alloreactive T cell activation rather induced infl ammatory cytokines such as TNF and INF-g are the determinant factors for effector T cell recruitment to GVHD target tissues. 

Acute Graft-versus-Host Disease (GVHD) after gender mismatched stem cell transplantation may be caused by female donor T cells recognizing HY minor Histocompatibility antigens expressed by male recipients. HY-specifi c CD8 positive cytotoxic T lymphocytes (CTL) have been detected in peripheral blood samples collected at the onset of acute GVHD. However, it is still unclear whether these CTL reach acute GVHDaffected tissues. We validated the sensitivity and specifi city of fl uorescently labeled multimeric HLA-A2 molecules containing HY peptide, i.e. HLA-A2/HY dextramers, to stain cryosections prepared from skin biopsies obtained from healthy male HLA-A2 positive volunteer donors. These skin explant tissues were cultured with female HY-specifi c CTL or control HA-1 specifi c CTL before cryopreservation. Unlike conventional HLA-A2/HY tetramers, HLA-A2/HY dextramers stained HY-specifi c CTL, but not HA-1 specifi c CTL, when applied to already cryopreserved skin explant tissue. Control staining of serial sections with HLA-A2 dextramers containing infl uenza peptide showed no positive signal. Next, the presence of HLA-A2/HY dextramer positive cells was analyzed in cryopreserved skin biopsies derived from 7 male HLA-A2 positive pediatric patients who developed acute GVHD of the skin after receiving a hematopoietic stem cell graft from a HLA matched unrelated donor. While a few CD8 positive HLA-A2/HY dextramer negative cells were detected in the skin of 4 boys who received male hematopoietic stem cells, significantly higher numbers of CD8 positive cells were detected in the skin of 3 boys who received a female graft; 50-75% of these CD8 positive cells were specifi c for HY. Skin-infi ltrating HYspecifi c T cells were visualized as early as 2 weeks after SCT when total peripheral blood lymphocyte counts were still low.

Our results underline the usefulness of this multimeric staining reagent for in situ characterization of donor-derived T cells that infi ltrate acute GvHD affected tissues. (1) Granulocyte-colony stimulating factor (G-CSF) is increasingly described as an immuno-modulatory agent for a diverse range of diseases and although the cytokine is usually associated with the regulation of immune responses, clear evidence exists that it can also exacerbate pathology in some settings. The clinical shift toward utilizing G-CSF mobilized stem cell grafts has resulted in enhanced hematopoietic reconstitution, reduced relapse rates in advanced disease, similar levels of acute GVHD but a striking increase in chronic GVHD. The mechanisms by which stem cell transplantation (SCT) promotes chronic GVHD are unclear. Comparison of cytokine expression following TCR activation of splenocytes from naïve and G-CSF treated B6 or BALB/c donors (low and high responders respectively) showed little effect of G-CSF on Th1 (IFN-gamma and TNF) or Th2 (IL-4, IL-5 and IL-10) cytokine production. In contrast, IL-17 production was dramatically enhanced in response to G-CSF treatment in both strains but was highest in BALB/c donors.

The amplifi cation of IL-17 production by G-CSF occurred in both CD4 and CD8 conventional T cells and by using relevant knock-out mice or blocking reagents we demonstrated that this was independent of IL-6, TGF-beta or IL-23 signalling. However, the induction of IL-17 by G-CSF was completely lost in the absence of IL-21 signalling. G-CSF induced the production of IL-21 in CD4 T cells, and this occurred independently of IL-17 production itself. We next utilized multiple models of GVHD using G-CSF mobilized B6 or BALB/c wild-type or IL-17 deficient donors, in both MHC matched and mismatched transplant settings. Surprisingly, IL-17 was critical for the induction of sclerodermatous chronic GVHD occurring late after transplant using either donor strain. Importantly, IL-17 controlled the dramatic sequestration of macrophages into skin that coincided with the fi brogenic response. This study provides a logical explanation for the propensity of allogeneic stem cell transplantation to invoke sclerodermatous GVHD and suggests a therapeutic strategy for intervention. Non-HLA gene polymorphisms contribute to the immune response leading to Graft-versus-host Disease (GVHD). We applied a systematic approach using microsattelite (ms) marker typing for a large number of immune response genes on pooled DNA of Japanese donors and recipients of haematopoietic stem cell transplants (HSCT) to identify recipient and donor risk loci for GVHD. Ms, due to their multiple alleles, are more informative than single nucleotide polymorphisms (SNP). We selected 4,231 ms markers, tagging 3,093 target genes (representing the 'immunogenome') at close proximity (<100kb). We selected 922 unrelated HSCT donor/recipient pairs from the Japan Marrow Donor Programme (JMDP) registry, based on clinical homogeneity (acute leukaemia, age 4-40 years, myeloablative conditioning, bone marrow source). 42.61% of pairs had a 10/10 HLA match. The population was split into discovery and confi rmation cohorts with 460/462 pairs each. Eight DNA pools, four for each of the two independent screening steps were created using a highly accurate DNA pooling method. While 4,321 ms were typed on the four pools of the 1st screening step, only markers positive here were typed on the 2nd screening pools. Fisher's exact test for 2x2 (each ms allele) and 2xm ChiSquare tests were performed, comparing allele frequencies of recipients with GVHD grade 0-1 with GVHD grade 2-4 (donors accordingly). Markers positive after both independent screening steps (p-value <0.05, same associated allele, consistent odd's ratio (OR)) were genotyped for confi rmation on individual samples of all 922 pairs. The independent, 2-step pooled DNA screening process has effectively reduced false-positive associations. In the fi nal analysis, 39 (recipient) and 58 (donor) ms loci remain associated with risk or protection from GVHD. (1), G. Afram (2) 

We have evaluated the impact of NIH score system in the outcome of 820 patients and studied additional prognostic factors among patients who develop cGVHD at 3 European centers. Furthermore, we tried to determine the prognostic impact of the different organs involved in order to defi ne which ones must always be evaluated and which ones could be avoided in a routine examination. Patients' characteristics: Patients transplanted from 2000 to 2006 at 3 different institutions were analyzed; 77% received stem cells from a related donor and 23% from alternative donors. Results: Median follow up was 1087 days (range: 7 to 2944). 53% and 31% developed overall and extensive cGVHD, respectively. According to NIH classifi cation 29%, 24% and 17% of the patients were categorized as mild, moderate and severe cGVHD, respectively. Type of onset was de novo in 22%, quiescent in 27% and progressive in 12%. Cumulative incidence of delayed acute GVHD was 10% while the respective value for overlapping syndrome was 17%. Among patients with cGVHD, the extent of cGVHD infl uenced on survival (55% vs. 71% for extensive vs. limited cGVHD, p<0.001), as well as the development of overlapping syndrome (OS 57% vs. 71% for patients who did or did not have overlapping syndrome, p = 0.01), of severe cGVHD (68%, 68% vs. 48% for patients with mild, moderate vs. severe cGVHD, p<0.001), and the type of onset (70% vs. 57% vs. 51% for de novo, quiescent and progressive cGVHD, p<0.001), while delayed acute GVHD did not infl uence on outcome. In multivariate analysis patients with severe cGVHD displayed a dismal outcome [HR = 0.31, 95%CI (0.17-0.58); p = 0,001]. The type of onset allowed to identify prognostic subgroups even among patients with severe cGVHD, so that the combination of both variables discriminate different patients subpopulations in terms of outcome [HR = 0.18, 95%CI (0.06-0.51); p = 0,001]. Among targeted organs performance score at the time of cGVHD had the most signifi cant effect on survival [HR = 4.73, 95%CI (2.39-9.36), p<0.001]. Also liver, gut and lung involvement adversely infl uenced on survival while eyes or mouth involvement had no infl uence on outcome. Conclusion: NIH scoring system plus cGVHD type of onset allows to defi ne patients subgroups in terms of survival. Patients with overlapping syndrome have a poor outcome. Among targeted organs, performance score was an independent prognostic factor while eyes and mouth involvement did not infl uence on outcome. Chronic GVHD (cGVHD) remains the leading cause for late morbidity and mortality after allogeneic hematopoietic stem cell transplantation (HSCT). The consensus conference summarized the current evidence on treatment of cGVHD and developed guidance for clinical practice. The consensus process included HSCT centers within Germany, Austria, and Switzerland participating in four meetings accompanied by two surveys on treatment of cGVHD send to all centers within the German and Austrian working party on Bone Marrow and Blood Stem Cell Transplantation and the Basel transplant center (Switzerland). Evidence was graded into: I (based on randomized trials), II (based on case controlled or well designed phase II trials), III-1 (several case series), III-2 (one case series), III-3 (case reports) according to all available publications. Recommendations were based on effi cacy and safety profi le: A (should always be applied), B (should generally be applied), C-1 (may be applied 1st-line), C-2 (may be applied 2nd-line), C3 (may be applied advanced line), C-4 (may only be applied in special circumstances). For 1st line treatment of cGVHD the following agents have been proposed: prednisone A I, calcineurin inhibitors (CNI) C-1 II, MMF C-1 III. For 2nd-line treatment the following treatment options have been proposed: prednisone B III-1, CNI C-1 III-1, mTOR inhibitors C-1 III-1, MMF C-1 III-1, photopheresis B II. A pulse of steroids C-2 III-2 and rituximab C-2 III-1 should be generally applied after 2nd-line treatment but may be used earlier in special circumstances. Advanced line treatment options being proposed are: MTX C-2 III-1, hydroxychloroquine C-2 III-2, clofazimine C-2 III-2, pentostatine C-2 II, thoracoabdominal irradiation C-2 III-2, and imatinib C-2 III-1. Additional agents in advanced line treatment are retinoids C-3 III-2, azathioprine C-3 III-1, and thalidomide C-3 II. Other treatment options have been rated as experimental and may only be applied in clinical trials or special circumstances: alemtuzumab C-4 III-3, etanercept C-4 III-3, alefacept C-4 III-3. The low level of evidence of the proposed treatment options resulting in a low level of recommendation indicates the urgent need for further evaluation in clinical trials. Moreover, the lack of indicators for response to certain agents requires a "trial and error" approach in choosing a treatment option and clinical trials for evaluation of biomarker for response are urgently needed. Recently, we reported an elevation of immature CD19 + CD21-B lymphocytes in patients with active cGVHD. Here, we investigated B lymphocyte subsets in cGVHD cohorts with abnormal serum immunoglobulin levels. Patients and methods: Seventy-fi ve patients were enrolled into our study a median of 42 months after HCT. They consisted of 3 groups: 25 with cGVHD and high serum IgG (>1600 mg/dl), 22 with cGVHD and low IgG (<700 mg/dl) and 28 patients with resolved cGVHD and normal IgG (control group). Severe cGVHD was present in 64% and 45% with more than 2 organs involved in 44% and 36% in the high IgG and low IgG group. Signifi cantly more patients in the high IgG group than in the low one had autoantibodies (72% vs. 18%, p = 0.02). Peripheral blood cells were analyzed by multiparameter fl ow cytometry after staining for CD19, CD21, CD27, CD10, CD38, CD95, and IgD. Results: While the high IgG group had equal CD19 + B cell numbers, the low IgG group had signifi cantly diminished ones compared to the control (400 vs. 160 vs. 400 × 10 6 /L p<0.0001). Numbers of CD10high most immature B cells in the high IgG, low IgG and control group were 26, 7, and 7 × 10 6 /L. Immature CD19 + CD21-B cell proportions were 9.2%, 16.5%, and 7.5% in the high, low IgG and control group (p = 0.01). In the high IgG group transitional B cells were signifi cantly increased (28 vs. 14 × 10 6 /L, p = 0.02) compared to the control. In the high IgG group naive and class-switched memory B cells were equal to the control whereas in the low IgG group both naïve (13 vs. 35 × 10 6 /L p = 0.02), class-switched (16 vs. 22 × 10 6 /L, p = 0.03) and non-class-switched B cells (4 vs. 10 × 10 6 /L, p<0.0001) were signifi cantly lower compared to the control. CD10-CD21low CD95high B cell proportions were the same in the high IgG and control group, while they were signifi cantly elevated (23.4% vs. 9.6% p<0.0001) in the low IgG group. Conclusions: cGVHD patients with hypergammaglobulinemia have normal B cell numbers with elevated immature and transitional B cells but no signifi cant impairment of B cell maturation. In contrast, patients with hypogammaglobulinemia have both a signifi cant B cell defi ciency and a distortion of B cell homeostasis in the circulation. Our data indicate different pathogenetic mechanisms of cGVHD leading to different clinical presentations.

Preliminary results of a phase II trial of montelukast for the treatment of bronchiolitis obliterans syndrome after HSCT K. Williams (1) Bronchiolitis obliterans syndrome (BOS) after allogeneic HSCT is a serious manifestation of cGVHD. Current treatments yield poor and transient responses. Although the pathogenesis of BOS after HSCT is unknown, a similar disease, BOS after lung transplant is associated with elevated leukotriene levels. We present preliminary results from an IRB-approved prospective, open label, phase II trial to test the effi cacy of montelukast, a leukotriene inhibitor, for the treatment of BOS after HSCT. BOS diagnostic criteria included: FEV1<75%, FEV1/VC < 0.7 or air trapping on CT and RV>120% or RV/TLC>120% in the absence of infection and presence of another cGVHD manifestation. Subjects had stable or declining FEV1 on stable or decreasing immunosuppression. Eleven patients have enrolled. One withdrew prior to medication and 9/10 patients have reached the primary endpoint (6 months) and have continued on study medication (10 mg montelukast po nightly). Study participants ranged from 15-64 years, 7/11 female, with baseline FEV1 from 33 to 71% predicted, and 3/11 patients required supplemental oxygen. FEV1 increased 6-10% predicted in 3 patients, remained stable in 4, and declined less than 15% in 2. Comparison of patient pre-study FEV1 decline to on-study FEV1 values was generated using the slope of FEV1 volume vs. days post-transplant. The difference in pre-and primary endpoint slope revealed: 7/9 improvement and 2/9 decline. Three patients reduced immunosuppression on study with complete cessation of tacrolimus in 1, cessation of steroids in 1, and decreased tacrolimus in 1 (including 2 with stable FEV1 and 1 with increase in FEV1); 1 patient had a steroid increase less than 1mg/kg/day. Two patients had worsening of other cGVHD manifestations, including skin fl are that resolved with local therapy (1) and gastrointestinal cGVHD fl are that improved with increased steroids and local therapy (1) . Notably, after 6 months, 1 patient demonstrated FEV1 increase of 14% predicted (from baseline) on tacrolimus taper after steroid cessation and 1 patient no longer required oxygen supplementation for exercise and sleep. Montelukast was well-tolerated with only one grade II attributable adverse event (insomnia) during the six-month collection period. Using NIH consensus criteria, improvements were also noted in 4/4 with buccal mucosa cGVHD and in 2/5 with evidence of liver disease. These fi ndings suggest that montelukast is a promising therapy for BOS after allogeneic HSCT.

Oral Session 11: Paediatric issues 1 O256 Objective: Allogeneic hematopoietic stem cell transplantation (HSCT) remains the main treatment option for children with advanced primary or secondary myelodysplastic syndrome (MDS). Relapse rate following HSCT for advanced MDS ranges between 20%-50%. This analysis was performed to asses the outcome of patients treated with a second HSCT (HSCT2) for disease recurrence after HSCT1. Patients and transplant: Within studies EWOG-MDS 98 and 2006, 73 patients with advanced MDS relapsed after fi rst HSCT (HSCT1); a second allograft was performed in 32 of these patients. For the 32 patients with HSCT2, diagnosis prior to HSCT1 was primary advanced MDS (n = 18), therapy-related MDS (n = 9) and secondary MDS after bone marrow failure disorder (n = 5). Preparative regimen of HSCT1 consisted of busulfan, cyclophosphamide and melphalan in 22 / 32 patients. Median time to relapse was 13.8 months (0.9-77.3) after HSCT1, time from relapse to HSCT2 3.7 months (0.4 -14.8), and median age at HSCT2 13.0 years (5.4 -25.3). Twelve patients were transplanted from a matched sibling donor, 16 from a matched unrelated and 4 from an alternative family donor. In 18 cases HSCT2 was performed using the same donor than in HSCT1. Stem cell source was peripheral blood (n = 22) or bone marrow (n = 10). Conditioning regimen and GvHD prophylaxis varied widely according to the centers preferences, a TBI-based regimen was given to 12/32 patients. Outcome: Median follow up was 3.0 years (0.3-5.4) after HSCT2. All but 2 patients engrafted. Acute GvHD grade II -IV was seen in 10 patients, chronic GvHD in 8 of 22 patients at risk. Transplant related mortality occurred in 9 patients; 4 of these died from GvHD, 1 from sudden cardiac arrest 6.5 years post HSCT2. A second relapse was noted in 12 patients, all of whom succumbed to their disease. There was no signifi cant difference in relapse rate according to the choice of donor (identical or different donor than in HSCT1) or the occurrence of acute/ chronic GvHD. For the whole cohort of 32 patients, the probability of event free survival at 5 years was 0.35 (0.17-0.53). Conclusion: We conclude that HSCT2 after MDS relapse is feasible and may rescue a considerable proportion of patients.

Reduced-intensity conditioning for children with refractory cytopenia: results of the EWOG-MDS study B. Strahm (1) , P. Bader (2) Objective: Refractory cytopenia (RC) is the most common subtype of childhood myelodysplastic syndrome (MDS). Hematopoietic stem cell transplantation (HSCT) in RC following a myeloablative preparative regimen has resulted in an eventfree survival of 80% with a very low probability of relapse and a probability of therapy related mortality of 15%. This analysis was performed to asses the outcome of children transplanted for RC following a reduced intensity conditioning. Patients and Transplant Procedure: Fifty-six patients (32 male/24 female) were diagnosed with RC at a median age of 11.2 years (1.8-17.9). None of the patients had an abnormal karyotype. The median time to HSCT was 0.71 years (0.12 -9.1). Patients were grafted from a matched sibling donor (MSD) (n = 15), an alternative family donor (n = 1) or an unrelated donor (UD) (n = 40). UD were matched 10/10 or 9/10 based on HLA-A, -B, -C, -DRB1 and -DQB1 molecular typing. Stem cell source was bone marrow (n = 48) or peripheral blood (n = 8). All patients were prepared with thiotepa (3 × 5 mg/kg) and fl udarabine (4 × 40 mg/m 2 ). Prophylaxis for graft-versus-host-disease (GVHD) was CsA ± MTX, ± MMF for MSD, and CsA, MTX and anti-thymocyte globuline for patients transplanted from an UD. Results: Two patients each experienced primary or secondary graft failure. Fifty-four patients reached neutrophil engraftment at a median time of 24 days (11-105). Platelet engraftment was demonstrated in 48 patients at a median time of 28 days (1 -201) ; in 2 patients an additional stem cell boost was necessary. Out of the four patients with graft failure three patients are alive after second HSCT. The incidence of acute GVHD grade II-IV and grade III-IV was 23% and 5%, respectively. Ten of 52 (19%) patients at risk developed chronic GVHD which was extensive in 5. One patient died of from veno-occlusive-disease with multi-organ-failure. Viral infections were the most prevalent complication. Fifty-four patients are alive with a median follow-up of 2.0 (0.1 -7.0) years resulting in a probability of overall survival of 0.95 and 1.0 for patients transplanted from MUD and MSD, respectively. Conclusion: In summary the conditioning regimen with thiotepa and fl udarabine offered an excellent survival for patients RC.

Chronic GVHD and viral infections were the main complications. Long term follow is needed to assess the expected reduction in long term sequelae. Allogeneic stem cell transplantation (SCT) is indicated in approximately 20-30% of children with acute lymphoblastic leukemia (ALL). Unrelated umbilical cord blood (UCB) is an established stem cell source for SCT. We retrospectively analyzed 532 children with ALL in complete remission (CR)1 (n = 186), CR2 (n = 238) and CR3 or advanced disease (n = 108) who received UCBT as fi rst transplant. Patients were transplanted in EBMT centers from 2000-2008. Median age was 6.8 years (y) (29 patients less than 1 y). The most common immunophenotype was B-cell precursor ALL and 17 patients had biphenotypic ALL. Of 186 patients transplanted in CR1, 45% had poor risk cytogenetics (t4;11 or t9;22). Grafts consisted of one (n = 504) or two (n = 28) units; 62% had 0-1 HLA mismatches with recipients while 38% had 2-3 mismatches (antigen level for HLA-A and B, allelic level for DRB1). Median TNC cell dose at freezing and infusion was 5.9 × 10 7 /kg and 4 × 10 7 /kg, respectively. Conditioning regimen was myeloablative in 96% of cases and TBI >6Gy was used in 52%. Other regimens included busulphan with cyclophosphamide ± thiotepa or melphalan. ATG was added in 88% of cases and GVHD prophylaxis was CSA ± steroids in 75%. Median follow-up was 18.5 months (3-109). Cumulative incidence (CI) of neutrophil recovery, acute GVHD and TRM were 82 ± 2%, 27 ± 3% and 21 ± 3% respectively. In multivariate analysis TNC infused >4 × 10 7 /kg (p = 0.001) and remission status at UCBT (p = 0.01) were associated with improved neutrophil recovery. CI of 2y relapse was 37±3% (31% for CR1, 34% for CR2, 50% for advanced). Disease status at UCBT (HR = 0.36, p 0.001) and use of TBI>6Gy (HR = 0.58, p = 0.01) were independently associated with lower CI of relapse. 2 y probability of leukemia-free-survival (LFS) was 38±2% (49% for CR1, 42% for CR2, 10% for advanced; p = 0.001). In multivariate analysis, disease status at UCBT was the only factor associated with improved LFS (HR = 0.32, p = 0.001). Causes of death were infections or other transplantrelated events (n = 172) or disease progression (n = 95).

In conclusion, in the absence of an HLA identical donor, UCBT remains a valuable alternative option for children with high risk ALL. Disease status at transplant and cell dose are the most important risk factors for outcomes. Use of TBI in the conditioning regimen was associated with decreased incidence of relapse but was not associated with improved LFS. The role of minimal residual disease as a predictor of outcomes of UCBT is under investigation.

Treosulfan-based conditioning in children: retrospective analysis of the German and Austrian experience R. Beier (1) Background: Treosulfan (TREO) is an alkylating agent with good myeloablative activity and a favourable toxicity profi le. It has been widely used in hematopoietic stem cell transplantations (HCT) of high-risk adult patients. Here we report on our retrospective analysis of children treated in Germany and Austria. Patients and methods: All 25 pediatric transplant centers in Germany and Austria were asked to submit data on TREO conditioning. Nine centers reported a total of 92 transplantations, 3 autologous and 89 allogeneic. Results: The median age at transplantation was 9,9 (range 0,1-20,4) years (n = 74). Underlying diseases were malignancies (n = 44: 12 AML, 12 ALL, 6 MDS, 5 Ewing sarcoma, 2 neuroblastoma, 1 rhabdomyosarcoma, 1 B-ALL, 1 LGL, 1 biphenotypic leukemia, 1 nasopharynx carcinoma, 1 ALCL, 1 nephroblastoma), immunodefi ciencies (n = 26: 8 SCID, 4 CGD, 3 unclassifi ed, 3 Hyper-IgM, 2 interleukin-10 receptor defi ciency, 2 HLH, 1 WAS, 2 CVID, 1 MHC class II defi ciency), hematologic diseases (n = 20: 8 thalassemia, 3 CAMT, 3 osteopetrosis, 2 SDS, 1 sickle cell disease, 1 SAA, 1 DKC, 1 SCN), and metabolic disorders (n = 2: 1 MLD, 1 mannosidosis). Donors were matched sibling or matched related (n = 21), matched unrelated (n = 50), autologous (n = 3), mismatched unrelated (n = 2) or haploidentical family donors (n = 12). Two patients received a combination of a cord and a haploidentical graft. Treosulfan was given at a total dose of 30 g/m 2 (n = 8), 36 g/m 2 (n = 31), and 42 g/m 2 (n = 53). Additional conditioning drugs included fl udarabine (n = 82), thiotepa (n = 43), melphalan (n = 22), cyclophophamide (n = 4), or TBI (n = 1, 4Gy). ATG (n = 37), Alemtuzumab (n = 23), OKT-3 (n = 11), and a combination of ATG and OKT3 (n = 3) were used together with CSA ± MTX or MMF as GVHD-prophylaxis. Eight patients experienced non relapse mortality, six of them had heavily pretreated malignancies. One engraftment failure (thalassemiahaplo) and three rejections (1 immunodefi ciency, 1 thalassemia, 1 SCN) were reported. All other patients engrafted rapidly. Toxicity and GVHD grades in these patients were mostly I and II. Event free survival after 3 years (data available for 65 patients) was 92% for non-malignant and 37% for malignant diseases. Conclusion: TREO based conditioning regimens in children were effective in terms of engraftment and survival. Optimal TREO dosing for children is still not well defi ned. Pharmakokinetic studies in children are needed. Objectives: Hepatic veno-occlusive disease (VOD) is a severe complication following hematopoietic stem cell transplantation (HSCT), with a mortality rate of 20%-50%.

The diagnosis of VOD is based on clinical criteria. However, the late onset of clinical signs may delay the due treatment. Studies in adults have addressed to plasminogen activator inhibitor-1 (PAI-1) a possible role as marker of VOD.

Our aim was to prospectively evaluate the role of fi brinolytic parameters to discriminate VOD from other liver disorders occurring after HSCT in a pediatric population. Methods: We studied 161 children (males 96, females 65, mean age 7,91 ± 5,17 years) who underwent 195 HSCT performed at the Pediatric Hematology Oncology of Padua University. The prevalence of VOD was recorded.

In 105 HSCT the levels of alanine amino-transferase (ALT), total bilirubin, PAI-1 antigen (PAI-1:Ag) and activity (PAI-1:act), t-PA antigen (t-PA:Ag), D-dimer, PT, aPTT, antithrombin, fi brinogen and platelet count were assayed before and weekly for 1 month after HSCT. Results: The prevalence of VOD was 5.6% (11/195 HSCT), and it was signifi cantly higher in patients with pre-existing risk factors for VOD as compared with those without (9.7% vs. 1.9%; p<0.05). All but one patient who developed VOD showed an early increase in PAI-1:Ag, PAI-1:act, t-PA:Ag and D-dimer levels, even before the clinical diagnosis of VOD. The increase in fi brinolytic parameters, and in particular PAI-1:Ag, was statistically signifi cant in comparison with that observed both in patients without complications (PAI-1:Ag p<0,0001) and in those with non-VOD liver disorders (PAI-1:Ag p<0.0001). Similar fi ndings were also seen when VOD patients were compared with subgroups of children with infections and/or different types of hepatic diseases complicating HSCT. Conclusions: Our study demonstrates a role of fi brinolytic tests in the diagnosis of VOD after HSCT in the pediatric population.

In particular, the early rise in PAI-1 antigen levels may suggest an incoming VOD, even in the absence of clinical signs. In addition, PAI-1 antigen increase may help to discriminate VOD from other hepatic complications after HSCT. Late onset non infectious pulmonary complications (LONIPC) after allogeneic hematopoietic stem cell transplantation (HSCT) such as chronic graft-versus-host disease (cGvHD) of the lung, bronchiolitis obliterans (BO), bronchiolitis obliterans organizing pneumonia (BOOP), sinu-bronchial syndrome and restrictive lung disease are well characterized. Immunodefi ciency, with or without cGVHD, and recurrent respiratory tract infections may be a vicious circle which requires optimal supportive therapy. Mucociliary dysfunction is described in chronic pediatric lung diseases like cystic fi brosis (CF) and primary ciliary diskinesia (PCD). Treatment with dornase alpha as a mucolytic drug with anti-infl ammatory effect improves morbidity and mortality rate in children with various chronic lung diseases. We hypothesized that outcome of our patients with LONIPC measured by mortality, lung function, inhalative therapy and systemic immunosuppressive treatment could be infl uenced by dornase alpha inhalations 2500 I.E. once daily. We analyzed 23 children with LONIPC who have received an allogeneic HSCT between 1994 to 2009 for malignant (n = 10) and non malignant (n = 13) diseases. We compared 12 children (study group) obtaining dornase alpha inhalations with 11 children without dornase alpha. Characteristics of both groups are shown in Table 1 . Additional supportive care was similar in both groups. Dornase alpha inhalations were well tolerated with no side effects and high compliance. Children treated with dornase alpha had a lower mortality rate (17%) than in the control group (36%) and showed in the long-term follow-up an improved lung function, predominantly of the peripheral airways (mean expiratory fl ow (MEF) after 25%, 50% and 75% of forced ventilation capacity), in the forced expiratory value after 1 second (FEV1), in resistance (R eff) and in the intrathoracic gas volume (ITGV) compared to the control group. Treatment with dornase alpha reduced signifi cantly the inhalation frequency of salbutamol, fl uticason or tiotropium after 6, 12 and 24 months (4 ± 2 times per day vs. 2 ± 2, p < 0.05; vs. 2 ± 2, p < 0.05; vs. 1 ± 1, p < 0.05) compared to the control group. Immunosuppressive medication was similar in both groups. Our data showed that inhalation with dornase alpha in children with LONIPC is a well tolerated treatment option which may improve mortality rate and lung function. Additional inhalative therapy but not systemic immunosuppression could be reduced. In order to fi nd hallmark of protection we studied the functional composition and magnitude of CMV-specifi c T cell response in CMV-seropositive (CMVpos) and CMV-seronegative (CMVneg) healthy donors and compared it to children after HCT. Polychromatic fl ow cytometry was used for detection of CMVspecifi c immune responses. The ability of CD4 + and CD8 + Tcells to produce cytokines: interferon-ã (IFNg) and interleukin-2 (IL-2) and to express activation marker CD40L and/or to mobilize the degranulation marker CD107a in response to CMV antigens was evaluated by intracellular cytokine staining method after in vitro stimulation. Peripheral blood samples were collected from 61 patients who underwent allogeneic SCT and 58 healthy donors. Monitoring of viral load was performed weekly from day + 7 using quantitative RT-PCR. Monitoring of CMVspecifi c T-cells begun on day + 28 and was performed weekly during the time of hospitalization and on day + 56, + 90, + 180 and + 365 after SCT.

We have compared functional signatures in patients controlling their PCR documented reactivations (controllers) and patients non-controlling reactivations (non-controllers). As a reference we have included healthy CMVpos and CMVneg donors. Among CD8 + T-cells we found three functional signatures (IFNg + , IFNg + /IL-2 + and IFNg + /IL-2 + /CD40L + ) that signifi cantly differed between controllers and non-controllers. Whereas single IFNg + CD8 + T-cells were the most frequent of CMV-specifi c T-cells, they were not restricted to CMV controlling individuals. We found signifi cant correlation between number of weeks with low viral load reactivations and frequency of IFNg + CD8 + Tcells (r = 0.4, p<0.0001). Dual IFNg + /IL2 + CD8 + T-cells were present in majority of controllers, but they were virtually absent in non-controllers. Triple IFNg + /IL-2 + /CD40L + positive CD8 + T-cells represented a rare subset of novel CD8 + helper phenotype. No CD4 + T-cell functional signature resolved controllers and non-controllers.

We (1) (

Aberrant DNA methylation contributes to the malignant phenotype in cancer including myeloproliferative neoplasms and myeloid leukemia. We studied aberrant DNA methylation in 14 candidate loci in blood and bone marrow samples of 87 children with juvenile myelomonocytic leukemia (JMML) and asked whether it is associated with clinical, hematologic or prognostic features of the disease. The pattern of hypermethylation allowed the categorization of JMML cases into several groups. High methylation was strongly associated with higher age and increased hemoglobin F level at diagnosis. Importantly, hypermethylation at diagnosis characterized cases with signifi cantly increased probability of leukemia relapse after HSCT: the 5year relapse incidence was 0.22 (0.11-0.45) in the no-methylation group but 0.69 (0.49 -0.96) in high-methylation cases (p<0.01). The predictive power of high methylation for outcome following HSCT was also refl ected in a multivariate Cox model which included age at diagnosis, sex, platelet count and mutational category (RAS, PTPN11, CBL mutation or clinical diagnosis of NF1); here methylation was the only signifi cant prognostic factor (p = 0.013). Future guidelines for intensity of graft-versushost prophylaxis in JMML will have to take the methylation status at diagnosis into account. ECFCs have recently been described as vascular progenitor cells with robust proliferative potential and vessel-forming capacity. Vascular integrity depends on endothelial and pericyte functions. This study was performed to establish ideal conditions for the generation of stable vessels by ECFC/MSCcotransplantation.

MSCs were propagated as previously described. ECFCs were isolated with a novel recovery strategy and propagated under animal protein-free culture conditions with pooled human platelet lysate (pHPL) replacing fetal bovine serum (FBS). ECFC long-term proliferation potential was monitored and phenotype was analyzed by fl ow-cytometry and immune-cytochemistry. Genomic stability was assayed with chromosome G-banding and array-comparative genomic hybridization (array-CGH). Additionally we compared telomere-length and telomerase-activity of ECFCs at different time points during culture with fl ow-fl uorescence in situ hybridization (Flow-FISH) and telomere repeat amplifi cation protocol-assay (TRAP). Functionality was studied during vascular network assembly in vitro and in human vessel formation in immune-defi cient mice in vivo. A mean of four ECFC colonies/mL of peripheral blood could be recovered. The progeny of these cultures could be expanded to mean 1.5 ± 0.5 × 10 8 ECFCs within 11-25 days. Consecutive analysis confi rmed ECFC purity, immune phenotype and sustained proliferation potential for >30 population doublings with preserved progenitor hierarchy. Genomic stability was confi rmed by karyotyping and array-CGH. Telomere-length analysis revealed longer telomeres in cord blood compared to peripheral blood-derived ECFCs and a constant shorting of chromosomal ends through passaging, which could be further confi rmed by a lack of telomerase activity. Large-scale expanded ECFCs functioned to form complex vascular networks in vitro and assembled stable CD31 + /Vimentin + /von Willebrand factor + human vessels when transplanted together with MSC in vivo under defi ned conditions. These human vessels were connected to the mouse circulation as indicated by a rich content of Ter119 + murine erythrocytes. This demonstrates that functional and stable human vessels can be generated in vivo as result of ECFC/MSC-cotransplantation. This procedure represents a promising tool to develop innovative experimental, diagnostic and therapeutic strategies. Hematopoietic progenitor cells (HPCs) which circulate in the blood are increasingly recognized as co-regulators of the development of tumor solid tumors through their ability to support neovessel formation and the transduction of the HPCs with suicide genes offers a new modality targeted antitumor cellular therapy. Further to clarify mechanisms of their tumor homing, we investigated the role of the small GTPases Rac1 and 2 expressed in hematopoietic cells during tumor development in mice. Mice with an inducible deletion of the Rac1 gene in hematopoetic cells and/or constitutive deletion in Rac2 were used. Rac1 deletion was induced by polyIC treatment in Rac-1fl ox/fl ox mice harbouring a hematopoietic-specifi c inducible Cre recombinase. Mice had previously been inoculated with Lewis Lung Carcinoma by s.c. injection. We found that in mice with hematopoietic specifi c deletion Rac genes, LLC tumors grew more slowly in the absence of Rac genes 1 and 2 compared to wild type (wt) controls. At the same time, numbers of mononuclear cells and HPCs increased in the blood. Mobilization was further increased in tumor bearing mice. In contrast, the content of CD45 + cells in tumors of mice with an induced defi ciency of Rac1 and 2 in HPCs was greatly diminished as shown by immunohistochemistry. We then performed competitive homing experiments by co-injecting green and red fl uorescence-stained wt and Rac1/2-/-HPCs in tumor bearing mice. This revealed a 60-70% decrease in numbers of Rac1/2-/-HPCs homing to tumors compared with controls. Next, we studied in vitro adhesion in parallel plate fl ow chambers to delineate specifi c adhesion defi ciencies of the Rac deleted HPCs. We found decreased arrest and persitstent adhesion of Rac1/2-/-HPCs on endothelial cells on SDF-1a precoated endothelial cells. This correlated with a decreased ability of these cells to migrate through gradients fo SDF-1a in transwells. Moreover, analysis of fi rm adhesion on both, Vascular Cell Adhesion Molecule-1 and Intercellular Cell Adhesion Molecule-1, either alone or in the presence of SDF-1a revealed a critical role of Rac1 and 2 genes in adhesion strengthening of HPCs.

In conclusion, Rac1 and 2 negatively regulate mobilization of HPCs in tumor bearing mice, but are critically involved in entry of HPC into tumor tissue in a process involving SDF-1a. Modulation of Rac activation in progenitor cells may thus be a tool to modify tumor growth and to target suicide gene delivery to tumor sites.

The However, HOXA9, HOXB7, HOXC10 and HOXD8 were defi ned as good candidate markers to discriminate CB-MSC and BM-MSC from USSC. Thus, our data suggest that the "biological fi ngerprint" based on the HOX code can be used to distinguish functionally distinct stem cell populations from bone marrow and cord blood. This work was supported by a grant from the DFG KO2119/ 6-1.

Combined action of endothelial and mesenchymal niche cells to amplify haematopoietic progenitor expansion in a humanized system D. Thaler (1) The hematopoietic stem/progenitor cell (HSPC) niche is an anatomically confi ned space governing HSPC proliferation, differentiation and self renewal. Recent research identifi ed distinct compartments described as stromal and vascular niches. This study was initiated to directly compare mesenchymal stromal cell (MSC) and endothelial colony-forming progenitor cell (ECFC) contribution to niche functions in a humanized co-culture system. MSCs and ECFCs were propagated under animal protein-free conditions. Autologous MSC-ECFC pairs were established to avoid donor variation. Purifi ed CD34 + HSPCs were used as responders in a cytokine-driven (IL-3, 6 ng/mL; Flt-3-L, 50 ng/mL; SCF, 20 ng/mL) niche cell-regulated co-culture system. Different standard media were employed to compare serum-free conditions with humanized cultures supplemented with pooled human platelet lysate (pHPL). Primary expansion culture of HSPCs with and without niche cell support was followed by colony forming cell (CFC) assays to determine maintenance of clonogenicity. Liquid cultures supplemented with 10% pHPL were more efficient than serum-free cultures resulting in a mean 16-84-fold increase in nucleated cell progeny within 11 days. MSCs or ECFCs further amplifi ed HSPC proliferation. Interestingly, the supportive effect of MSCs or ECFCs was constantly more pronounced in humanized pHPL-supplemented compared to serum-free cultures, indicating an important role of the natural platelet-derived growth factors. The combination of MSCs + ECFCs resulted in HSPC proliferation with up to 567-fold NC increase. The HSPC progeny was mainly comprised of differentiating myeloid cells. CFC assays revealed that pHPL-supplemented liquid cultures were more effi cient than serum-free cultures resulting in a 1.2-4.5 fold CFC expansion. Both MSC and ECFC initiated a more than 6-fold increase of CFCs, whereas combined action of MSCs + ECFCs resulted in a maximum of 23.5-fold CFC increase in pHPL-supplemented cultures. The impact of human ECFC as compared to MSC cotransplantation on human HSPC engraftment and reconstitution in immune defi cient mice is currently under investigation.

Our data indicate that MSCs and ECFCs are effi cient in supporting HSPC proliferation and CFC amplifi cation in a humanized co-culture system. The combination of both niche cell types had no further additive effect. The humanized co-culture thus provides a novel model system for analyses of HSPC-niche cell interactions.

A. Reinisch (1), N.A. Hofmann (1) , A. Ortner (1) , N. Etchart (1), C. , C. Dullin (2) , C. Diwoky (2) Osteosarcoma (OS) and Ewing's Family Tumors (EFT) are the most frequent bone sarcomas in young adults. Unresectable or metastatic presentations are currently characterized by an extremely severe prognosis, with a consequent great need for new therapeutic approaches. We conducted a preclinical study to investigate the potential effi cacy of Cytokine-Induced Killer (CIK) cells as adoptive immunotherapy for bone sarcomas. CIK cells are a heterogeneous subset of ex-vivo expanded T lymphocytes presenting a mixed T-NK phenotype and endowed with a MHC-unrestricted antitumor activity. CIK cells can be used both as autologous adoptive immunotherapy or even infused from a donor after allogeneic HCT. We successfully generated CIK cells from 6 patients with metastatic bone sarcomas (5 OS; 1 EFT). CIKs were generated from PBMC, cultured for 3-4 weeks with the timed addition of IFN-a; Ab anti-CD3 and IL2. At the end of the culture we obtained a heterogeneous T cell population with a median of CD3 + CD56 + cells of 40% (15-45), 52% (51-55) were CD8 + and presented a high expression (>95%) of NKG2D, the receptor mediating most of CIKs' antitumor activity. The median ex-vivo expansion, calculated on the CD3 + CD56 + fraction, was 50 fold (22-400). Cytotoxicity experiments (n = 4) demonstrated that CIK cells effi ciently killed 2 different OS cell lines (86%, 85%, 81% and 68% of specifi c killing at a 40:1, 20:1, 10:1 and 5:1 effector/target ratio respectively). More striking, in selected experiments (n = 2), we could confi rm the potent killing activity of CIK cells against the autologous OS tumor (Figure 1 ). The test was performed by staining target cells with CFSE and co-culturing them with CIKs at different ratios; after 5 hours data were analyzed by fl owcytometry. We confi rmed that both cell lines and autologous OS cells presented a high expression of MIC A/B, main ligands for the NKG2D receptor of CIK cells. The tumor-specifi c cytotoxicity of CIK cells was confi rmed by the absence of any signifi cant killing against non-tumoral MHC-mismatched targets.

Our data are the fi rst report of antitumor activity of CIK cells against autologous OS cells. Our fi ndings are encouraging and support the designing of adoptive immunotherapy clinical trials with autologous CIK cells for OS patients. The observed safety across major HLA-barriers suggests that also donor-derived CIK cells might be considered in peculiar experimental clinical settings exploring allogeneic HCT for solid tumors.

M. Nonn, J. Knapstein, A. Brunk, D. Tomsitz, S.A. Khan, E. Distler, M. Theobald, W. Herr, U.F. Hartwig Johannes Gutenberg University (Mainz, DE) Donor lymphocyte graft engineering to separate graft-versushost (GVH) and graft-versus-leukemia (GVL) immunity is of central interest in allogeneic hematopoietic stem cell transplantation. Therefore, we established a xenograft-transplantation model using NOD/SCID/IL2R gamma chain null (NSG) mice to evaluate engraftment, residual GVH-reactivity and GVL-immunity of human leukemia reactive CD8 + T cell lines in vivo. Previous studies demonstrated engraftment of resting or polyclonally stimulated human CD3 + T lymphocytes to high levels in spleen and bone marrow following adoptive transfer into NSG mice and induction of xenogeneic GVHD (x-GVHD) within 3-5 weeks depending on the applied T cell dose. The engrafted human T cells contained both CD4 and CD8 subsets in the same ratio as analyzed prior to transfer. Further transfer experiments using isolated human CD4 + and CD8 + T cell subpopulations revealed long-term engraftment of CD4 + T cells and induction of xenoreactivity. CD4 + T cells isolated ex vivo from spleen from these mice demonstrated an activated and memory phenotype suggesting a xenoantigen driven response in vivo. In contrast, adoptively transferred CD8 + T cells did neither engraft nor induce x-GVHD in NSG recipients. Similar results were obtained upon transfer of several leukemia-reactive CD8 + T cell lines generated by short-term mixed leukemia lymphocyte coculture of CD8 + donor lymphocytes and acute myeloid leukemia (AML) blasts. Investigating different CD8 + T cell growth and differentiation promoting cytokines we found that administration of human interleukin (IL)-2 or IL-7 failed to improve CD8 + T cell survival and expansion in vivo, whereas co-transfer of 10% autologous, naïve CD4 + T helper cells or repetitive injections of human IL-15 resulted in robust CD8 + T cell engraftment in NSG recipients. In addition to these data, fi rst results of studies obtained in a therapy model to investigate GVL-responses of HLA-mismatched AML-reactive CD8 + cytotoxic T cells upon transfer into primary AML engrafted NSG mice in the presence of autologous T helper cells or IL-15 will be reported. In summary, we have shown that CD8 + T cell engraftment and homeostatic proliferation in NSG mice is dependent on CD4 + T cell-mediated or, alternatively, IL-15-driven signals. We conclude that our NSG transplantation model provided with these signals may be a valuable tool for evaluating GVH-and GVLreactivity of ex vivo modifi ed human donor T cell grafts in vivo.

Oral Session 13: Reduced-intensity conditioning O334 RIC allo-HSCT for haematological malignancies up to 70 years is feasible without extratoxicity: a retrospective study of 619 patients on behalf of the SFGM-TC P. Chevallier, D. Blaise, N. Milpied, M. Michallet, J.P. Vernant, N. Fegueux, M. Renaud, B. Rio, N. Gratecos, J.Y. Cahn, G. Socie, I. Yakoub-Agha, A. Huynh, S. Francois, J.O. Bay, C. Cordonnier, A. Buzyn, N. Contentin, E. Deconinck, M 

The aim of this report was to assess the outcome of 619 patients aged ≥60 years who received a RIC allo-SCT, with a special emphasis on the comparison of the outcome of patients aged 60-65 years (y.) and patients aged >65 y. Between 1998 and 2008, 619 patients aged ≥60 years with different haematological diseases were treated with a RIC allo-SCT, and reported to the SFGM-TC registry. Patients twice allografted were excluded from this study. This series included 408 males (66%) and 211 females (34%). The median age for the whole cohort was 62 (range, 60-71) y. 369 patients (60%) were diagnosed with a myeloid malignancy. 468 patients (76%) had high risk disease features. 380 patients (61%) received a RIC allo-SCT from an HLA-matched related donor. PBSC were used in 82% of the patients (n = 511). The conditioning regimen consisted of Fludarabine and Busulfan in 307 cases (50%), Fludarabine and low dose TBI in 148 cases (24%). The remaining 164 patients (26%) received other socalled RIC protocols. ATG was used in half of cases. With a median follow-up of 23 (range, 1-125) months after allo-SCT, engraftment was 96%, grade II-IV and grade III-IV acute GVHD occurred at a median of 31 days after allo-SCT in 29% (n = 180) and 12% (n = 75) of patients, respectively. Chronic GVHD was observed in 142 out of 528 evaluable patients (27%). NRM was 31% (n = 194). 2-year OS was 48% (95%CI, 44-53%). In order to assess the applicability of RIC allo-SCT to the older age group, we next compared the outcome of patients aged from 60 to 65 y. (n = 509; median age 62 y.) and those aged >65 y. (n = 110; median 66 y.). Except for age, in univariate analysis, these 2 groups were not statistically different in terms of demographic, disease or transplant characteristics. Overall, the median time to ANC>500/μL was 18 (range, 1-99) days in each group. The grade II-IV acute GVHD (29% vs. 31%), the overall NRM (31% vs. 33%), the relapse (32% vs. 24%) and the deaths (54% vs. 54%) incidences, as well as the 2-year OS (48.5% (95%CI, 44-53%) vs. 48% (95%CI, 38-57%)) were comparable between the younger population vs. the older one (p = NS). In a Cox multivariate analysis accounting for all relevant factors, age >65 y. was not found to be a statistically signifi cant factor associated with worsened survival. Despite its retrospective nature and the inherent selection biases, this data support the use of RIC-allo-SCT in patients up to 70 years.

Final results of UK multi-centre, phase II study of Campath-1h dose de-escalation prior to non-myeloablative hla-identical sibling stem cell transplantation G. Orti, K. Peggs, M. Collin, R. Clark, D. Milligan, H. Roddie, E. Liakopoulou, C. Crawley, A. Clark, R. Pettengel, N. Chowdhry, M. Roughton, J. Snowden, E. Morris, K. Thomson, P. Kottaridis, A. Fielding, S. Mackinnon, R. Chakraverty on behalf of the UK Collaborative Group

Background: Inclusion of 100mg CAMPATH-1H (CAM) as part of a Fludarabine/Melphalan conditioning regimen is effective at preventing GVHD and reducing NRM following allogeneic SCT. However, these benefi ts are offset by high rates of infection and potentially a loss of graft-versus-tumour effects. We reasoned that reductions in the dose of CAM would permit improved immune reconstitution post-transplant. Methods: We performed a multi-centre trial in which the total dose of CAM was reduced step-wise in cohorts from 60 mg to 20 mg prior to HLA-identical sibling transplantation (n = 101). Eligibility criteria were patients with haematological malignancies aged 18-65, life expectancy >3 months and unsuitable for myeloablative conditioning. Primary endpoints were NRM, incidence of GVHD or infection. The study received IRB approval and all patients gave informed consent. The CAM dose was reduced over 4 cohorts: 60 mg (group A, n = 26); 40 mg (group B, n = 24); 30 mg (group C, n = 27) and 20 mg (group D, n = 24). Results: Median follow up is 2.6 years. Median age was 50 yrs (range 17-64). 3-yr OS was 71% with no differences observed in between the groups. 3-yr NRM was 16% overall (23.1% in group A, 12.5% in B, 7.4% in C and 20.8% in D, with no significant differences between the groups). Cumulative incidences of grade II-IV acute GVHD was 3.9% in group A, 8.3% in B, 3.7% in C and 12.5% in D. 3-yr cumulative incidences of extensive chronic GVHD were 7.7%, 4.2%, 0% and 16.7% in groups A, B, C and D respectively. The incidence of infection and kinetics of CD4, CD8, total lymphocyte recovery did not differ between the groups. The lowest dose cohort (20 mg, group D) compared unfavourably to the other 3 cohorts combined (30-60mg, groups A-C), in terms of cumulative incidence of severe grade III-IV acute or chronic, extensive GVHD (HR 4.2, 95% CI 1.1-15.6) and a trend for greater number of non-relapse deaths due to infection (HR 3.1, 95% CI 0.9-10). 2 patients died of grade III-IV acute GVHD in group D with no deaths related to this complication in the other groups. Conclusions: Signifi cant dose de-escalation of CAM prior to HLA-identical sibling transplantation is feasible without increasing NRM, although reductions below 30mg are associated with a higher risk of severe acute and extensive chronic GVHD. In future studies, we will determine whether use of CAM at a dose of 30mg for HLA-identical sibling transplantation will translate into improvements in progression-free survival.

Comparison of Bu-Cy-based standard myeloablative conditioning vs. fl udarabine and busulfan (Flu-Bu)-based reduced-intensity conditioning M. Mohty, M. Labopin, G. Socié, N. Milpied, M. Attal, D. Blaise, O. Ringden, P. Brown, B. Lorentz, M. Brune, R.M. Hamladji, R.F. Duarte, A. Nagler, V 

The aim of this analysis was to compare outcomes (LFS, NRM, and relapse incidence) between patients receiving the classical Bu-Cy myeloablative conditioning (MAC) regimen vs. patients receiving a fl udarabine and busulfan (Flu-Bu) reduced-intensity conditioning (RIC) regimen prior to allo-SCT. Since the use of RIC in young patients is still limited, the analysis was restricted to patients aged >40 years, who were transplanted using an HLA identical sibling donor for AML in CR1. In summary, in this specifi c setting of AML in CR1, LFS is not statistically different when using MAC or RIC allo-SCT for patients older than 50 years. However, for younger patients (40-50 y. age group), the use of RIC is associated with a higher relapse rate. Prospective trials addressing the use of RIC in patients younger than 50 years are needed. Indeed, reducing toxicity without compromising disease control could be of signifi cant benefi t to many patients, but "more intensive" regimens, despite the hazard of increased toxicity, may be necessary in others. Thus, the trade-off between dose intensity, toxicity, and disease control will remain to be assessed for each individual patient.

Long-term results of reduced-intensity conditioning with busulfan and fl udarabine ± ATG prior to allogeneic HCT: 10-year follow-up K. Sockel, M. Bornhäuser, N. Kröger, D. Beelen, A. Fauser, R. Schwerdtfeger, W. Siegert, L. Kraut, A. Cook, T. Zabelina, C. Theuser, G. Ehninger, J 

Reduced intensity conditioning (RIC) prior to allogeneic stem cell transplantation (HCT) is being used for 10 years now. The long-term outcome of RIC conditioning with respect to late relapse and late non-relapse mortality (NRM) is still subject to ongoing research. Therefore, we sought to determine factors predictive for long-term NRM and overall survival (OS). Methods: We retrospectively analyzed data from patients who received RIC based on busulfan and fl udarabine prior to allogeneic HCT between March 1998 and July 2001 in six major german transplant centers. Overall survival was analysed using multivariate Cox regression models, while competitive event statistics were applied for NRM. Results: We identifi ed 196 patients with a median age of 52 years (range, 12 to 67). The underlying hematologic diseases were: AML/MDS (47%), CML (20%), CLL (16%), lymphoma (10%) or other hematologic malignancies (7%). Donors were matched siblings in 42%, other matched relatives in 5%, matched unrelated in 40% and unrelated with single mismatches in 12% of the patients. The incidence of acute GVHD grades II-IV was 53%. The cumulative incidence of any episode of extensive chronic GVHD at 5 years after HCT was 46%. 66 patients were alive with a median follow up of 9 years (range, 2 to 11). The 10-year OS and relapse-free survival was 32% (95% CI, 21% to 35%) and 28% (95% CI, 21% to 35%). Non-relapse mortality at 10 years was 30% (95% CI, 23% to 37% that, these 2 regimens produce similar 1 year OS (primary endpoint). However, FBA is associated with better early PFS and EFS and socially acceptable cost-effectiveness ratio but worse early QOL. FBA is also associated with better long term disease control, whereas FTBI tends to produce lower TRM and higher rejection rates. Clinical data might help designing individual and optimal strategies for each candidate patient, while economical data may help hospitals to tailor their transplant program, depending on the patient population that they care for.

Allogeneic-related stem cell transplantation with reduced-intensity conditioning versus best standard of care in older patients with acute myeloid leukaemia in fi rst complete remission. Interim analysis of a prospective multi center phase III trial T.L. Kiss (1) Introduction: Allogeneic transplants after reduced intensity conditioning (RICT) are increasingly used in patients (pts) with acute myeloid leukemia (AML). There is however a lack of prospective data supporting this practice. We present an interim analysis focusing on safety of an ongoing phase III multi center trial aiming at determining whether RICT from an HLA matched sibling donor (MSD) leads to improved overall survival (OS) compared to standard of care in elderly pts with AML in fi rst complete remission (CR1). Patients and methods: Elderly pts with de novo or secondary AML in CR1 with non favourable risk cytogenetics were enrolled if they had a potential sibling donor and assigned to the RICT group or control group depending whether a MSD was identifi ed. Both groups were followed prospectively. In the RICT group 85% of pts were conditioned with fl udarabine (150 mg/m 2 and busulphan 6.4 mg/kg IV or 8 mg/kg PO). Graft versus host disease (GVHD) prophylaxis consisted of ciclosporine with mycophenolate mofetil or methotrexate. The study is supported by the Canadian Blood and Marrow Transplant Group. Results: Between 2003 and June 2009 101 pts (53 males, 48 females), median age 62 (53-74) yrs were enrolled and allocated to the RICT (n = 61) or control (n = 40) group. Median follow up for the RICT and control group was 17 and 23 months, respectively. After a median of 2.3 months (1) (2) (3) (4) (5) (6) (7) (8) 49 pts in the RICT group (80%) underwent allografting. 12 pts did not reach transplant due to non-relapse death (n = 3); relapse-related death (n = 6) or comorbidities (n = 3; infections, cardiac disease). Six pts (12%) died after transplant without previous relapse, due to GvHD (n = 4), infection (n = 1) and VOD (n = 1). Relapse rate for patients assigned to the RICT group was 34% (n = 21), for patients actually transplanted 31% (n = 15). In the control group there were 2 (5%) non-relapse deaths and 18 (45%) relapses. Kaplan-Meier estimates of 3 year OS and PFS in the whole study group was 50% and 45% respectively, with no signifi cant difference between RICT and control arms. Conclusion: This early analysis indicates that RICT can be performed with relatively low transplant related mortality and with a moderate relapse rate. The full potential of RICT can only be assessed after inclusion of more patients, longer follow-up, and by reducing the rate of pts not reaching transplant due to early relapse. In this interim analysis, no conclusions can be drawn on the putative positive effect of RICT in elderly patients with AML.

T-cell depleted reduced-intensity transplantation followed by donor leukocyte infusions to promote graft-versuslymphoma activity results in excellent long-term survival in patients with multiply relapsed follicular lymphoma K. Thomson (1) Follicular lymphoma is an indolent disorder, which is treatable but considered incurable with chemotherapy alone. The curative potential of allogeneic transplantation using conventional myeloablative conditioning has been demonstrated, but this approach is precluded in the majority of patients with FL, because of excessive toxicity. Reduced intensity conditioning regimens are therefore being explored. This study reports the outcome of 85 consecutive patients with FL transplanted using fl udarabine, melphalan and alemtuzumab. Patients were heavily pretreated, having received a median of 4 lines of prior therapy, and 27% had failed previous autologous transplantation. Median patient age was 45 years and 53% received stem cells from unrelated donors. With a median follow-up of 4 years, the non-relapse mortality was 15% at 4 years (8% for sibling, 21% for unrelated donor transplants), acute GVHD grade II-III occurred in 14%, and the incidence of extensive chronic GVHD was only 18%. Relapse risk was 26%, and this was signifi cantly reduced where mixed chimerism had been converted to full donor chimerism by the use of DLI (10% vs. 36%; p = 0.03). In addition, 10/13 (77%) given DLI for relapse post-transplant remitted, with 9 of these responses sustained and one responding to further DLI. In these patients, CR is currently ongoing at a median of 44 months (12-74) following last DLI. Current PFS at 4 years was 76% for the whole cohort: 90% for those with sibling donors and 63% for those with unrelated donors. The excellent long-term survival with associated low rates of GVHD and frequency and durability of DLI responses makes this an extremely encouraging strategy for the treatment and potential cure of FL. We analyzed the incidence and the risk of secondary solid tumors for 17,997 adult recipients (5,598 related bone marrow S58 (BM) recipients, 3,747 related peripheral blood (PB) recipients, 6,530 unrelated BM recipients, 2,093 unrelated cord blood (CB) recipients Objectives: Allogeneic hematopoietic stem cell transplantation (HSCT) can severely compromise patients' health related quality of life (HRQoL), but there is lack of evidencebased data in this area. This is the first report to provide a longitudinal assessment of HRQoL in transplanted thalassemia children and their families also using a parent proxy-report evaluation. HRQoL scores in children and parents were prospectively evaluated as well as parent healthcare satisfaction. Methods: Fifty-eight thalassemia patients from Middle Eastern countries were evaluated at baseline. Twenty-eight children (median age 9.5 years) were transplanted from an HLAmatched donor (25 sibling, 2 unrelated and 1 haploidentical) in 2 Italian HSCT Centers. The PedsQL 4.0 Generic Core Scales were administered to patients and their parents both before and after (3, 6, 18 months) transplantation. Parent healthcare satisfaction was assessed using the PedsQL Healthcare Satisfaction Generic Module. These questionnaires were also administered to 30 conventionally treated thalassemia patients and their parents, all coming from the same geographical area. Change from baseline at 3, 6 and 18 months was assessed using the Wilcox Signed Rank Test. Results: Two-year overall survival, thalassemia-free survival, mortality and rejection were 89%, 79%, 11% and 14%, respectively. The cumulative incidence of acute and chronic GVHD was 35% and 18%, respectively. Eighteen months after transplantation, total PedsQoL scores were signifi cantly higher than baseline ratings, both in child ( + 8.3, p = 0.05) and parent reports ( + 13.4, p<0.01) (Figure 1 ). While physical, emotional and social functioning scores generally tended to decrease after 3 months in comparison with baseline values, they all returned to higher levels at 18 months after transplantation. Final PedsQoL scores were also signifi cantly higher than those obtained for conventionally treated patients (p<0.05).

Conclusion: This is the fi rst longitudinal study showing a trend towards a better HRQoL in thalassemia children at 18 months after transplantation compared to pretreatment levels and conventionally treated patients. Also parent satisfaction with cure and doctor-family communication was high. These data could help to relieve some of the uncertainty surrounding the diffi cult choice of transplantation in a chronic disease like thalassemia. Allogeneic stem cell transplantation (SCT) is potentially curative therapy for patients (pts) with acute leukemia and MDS. SCT is associated with substantial mortality during the fi rst 2 years after SCT due to relapse and non-relapse mortality (NRM) whereas after 2 years survival curves often reach a plateau. The pattern of late events was reported in myeloablative conditioning (MAC) but is not well defi ned in the reduced intensity (RIC) setting. To explore late outcomes we analyzed SCT results in a cohort of 401 pts with AML/ MDS and ALL. Pts meeting standard eligibility criteria were given MAC (BuCy or Cy/TBI)). Pts considered at excessive risk for NRM were given reduced toxicity conditioning (RTC) such as fl udarabine (F) with high-dose busulfan (Bu) or treosulfan or a RIC regimen of F and reduced Bu or melphalan.

The 5-year overall survival (OS) was 35% (95CI, 30-41) and was similar with the various regimens. We identifi ed 115 pts with AML/MDS (n = 96) or ALL (n = 19) who were alive and leukemia-free 2 years after SCT from related (n = 70) or unrelated donors (n = 45), using RIC (n = 34), RTC (n = 37) or MAC (n = 44). Median age was 51 (17-72). At the 2-year time-point, 41 pts had a history of acute GVHD and 55 had active chronic GVHD which still required immune suppressive therapy (IST) in 48. The probability of remaining alive for the next 5 years was 80% (95CI, 70-90). It was 76%, 85%, and 83%, after RIC, RTC and MAC, respectively (p = NS). There were 16 deaths beyond 2 years; 9 due to relapse and 7 due to NRM. NRM included 3 deaths due to solid cancers. There were 4 additional secondary cancers in pts currently alive. Two pts died of myocardial infarction and two of chronic GVHD. In all, the cumulative incidence of late NRM and relapse mortality was 9% (4-20) and 11% (6-22), respectively. Advanced age (>50) was the most signifi cant predicting factor for shortened OS; 64% and 96% in the older and younger groups, respectively (p = 0.005 Results: Recipients with MBL levels <1000 ng/ml were at increased risk to suffer from one or more major infections (p = 0.002) during entire follow-up. Infectious susceptibility was increased after neutrophil recovery, particularly until 24 months (Hazard Ratio (HR) 3.4) with sustained effects afterwards (HR 2.9) (Figure 1 ). In multivariate analysis MBL serum concentrations <1000 ng/ml were independently associated with major infections after neutrophil recovery (p = 0.009) when corrected for recipient age at transplantation, sex, diagnosis, history of total body irradiation, history of splenic irradiation or splenectomy, acute or chronic graftversus-host disease, and duration of immunosuppression. In subgroup analysis occurrence of severe herpes virus infections in particular was associated with signifi cantly lower MBL levels (p = 0.02). Conclusion: Our fi ndings indicate that low MBL levels may predict markedly increased susceptibility to severe infections with sustained effects even late after allogeneic HSCT. Determinations of MBL status should therefore be included into pretransplantation risk assessment. Introduction: Aim of this study was to evaluate frequency and risk factors of both secondary myelodysplastic syndromes/acute leukemias (sMDS/AL) and solid tumors in lymphoma patients treated with the high-dose (hd) squential (HDS) chemotherapy with peripheral blood progenitor cell (PBPC) autograft. Patients and methods: Data have been collected on 1,347 lymphoma patients treated at 11 Centers associated to GITIL ( Figure 1 ). Patients received either the original or modifi ed HDS regimens, as shown in Figure 2 . PBPC were collected after hdcyclophosphamide (CY), and, in a subgroup, after a 2nd round of mobilization with hd-Ara-C. To reduce the incidence of acute and chronic graft versus host disease (GVHD) anti-T-cell globulins (ATG) have been incorporated into the preparative regimen for allogeneic stem cell transplantation (SCT) from alternate donors by many centers. Different ATG preparations are available and little is known about the optimal dosing. Therefore we conducted this retrospective registry study utilizing specifi c questionnaires to participating centers. Chronic myelogenous leukemia (CML) in chronic phase has been selected as underlying disease in order to have a rather homogenous patient population. A total of 1359 patients (pts) have been analyzed. 534 pts had received no ATG, 288 ATG-Fresenius, 122 Thymoglobuline® (Genzyme), 261 other in vivo T-cell depletion, mainly Campath, and 154 had received in-and ex-vivo T-cell depletion, utilizing Thymoglobuline® for in-vivo depletion. A cumulative dose of less than 40 mg/kg ATG-Fresenius or less than 10 mg/kg Thymoglobuline® has been defi ned as low-dose. The median follow-up for surviving pts is 62 months (range: with no statistically difference in the different pts groups. Only the use of ATG-Fresenius and Thymoglobuline® proved to be an independent positive prognostic factor for overall survival in multivariate analysis incorporating the EBMT risk score.

This was due to decreased treatment related mortality. However, any of the analyzed T-cell depletion strategies increased the risk of relapse, which did not translate into overall survival, since relapse after allo SCT is manageable in CML pts. When also analyzing the dosing of ATG, the use of high dose ATG-Fresenius was associated with the best long-term overall survival of about 70%. When comparing high dose Fresenius versus all others the use of high dose Fresenius had the same impact on overall survival as the EBMT risk score, indicating that the use of high dose Fresenius is an independent positive prognostic factor. Similar effects were not seen with high-dose Thymoglobuline®. Interestingly, the positive effects of ATG only became obvious after 4 months after transplant suggesting no protection against acute GVHD but protection against mortality from chronic GVHD. Although unrelated allogeneic SCT in chronic phase CML is nowadays a rather rare indication these data nevertheless prove benefi cial effects of in vivo T-cell depletion and also emphasize, that the different preparations are not interchangeable and that the dosing is of great importance.

Skewed T-cell receptor repertoires in very long-term survivors after allogeneic haematopoietic stem cell transplants A. Rovó (1), C. Arber (1), P. Fisch (2), U. Siegler (1) Disturbed immune reconstitution and late auto-immune like phenomena have been observed after HSCT and described as "late altered immunity" in very long-term survivors. In a previous cross-sectional study on 44 long-term survivors after allogeneic HSCT (>10 years) and their respective donor, we demonstrated that recipients with cGvHD and a female donor had signifi cant telomere shortening of the hematopoietic cells, and subtle signs of altered late organ function as compared to their respective donor (Baerlocher et al., Blood, 2009 ). We were therefore interested to learn more about these differences and analyzed in the same cohort the T Cell Receptor beta (TCRb) repertoire diversity by spectratyping of peripheral blood T lymphocytes. We compared by visual analysis the Gaussian like distribution of the peaks from 23 Vb subfamilies. Pattern of each Vb subfamily was classifi ed as normal or skewed. Normal refers to >19 Vb subfamilies with a Gaussian like distribution, abnormal to >5 Vb subfamilies with a non Gaussian pattern. The overall complexity of the TCRb repertoire (total number of peaks of all 23 Vb subfamilies) from the recipients was compared with the respective donor.

The median age at HSCT and at time of the study was 22.5 (6-50) and 44 (27-62) years for the recipients; 23.5 (11-45) and 45 (31-61) for the donors. The median time interval between HSCT and TCRb repertoire analysis was 20 years (12-25). During follow-up 4/10 analyzed recipients had cGvHD. The median number of skewed Vb subfamilies was 8.5 (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) for recipients, and 2 (0-8) for donors (P = 0.008). The median overall complexity was 149 (101-170) for recipients and 159 (138-183) for donors (P = 0.041). 8/10 analyzed recipients showed an abnormal pattern of the TCRb repertoire. All but two donors had a normal repertoire (P = 0.001); one of them suffered of 2 neoplasms after donation One patient with an abnormal TCRb repertoire ( Figure 1 , UPN498) had still extensive, severe cGvHD and showed the shortest telomere on the leukocytes and T-lymphocytes: 3.7 and 3.6 kb respectively, versus median 6.5 (5.4-8.1) and 5.7kb (4.8-7.4) respectively.

These data indicate that T cell repertoire diversity in long term survivors after allogeneic HSCT is abnormal and differs from the repertoire of their respective donors. The cause of this skewed T cell repertoire remains open but gives a formal basis for the clinically observed late altered immunity. The clinical consequences thereof need to be evaluated in the future.

A second allogeneic SCT was given to 119 patients (median time from SCT1 11 mo, from relapse 1.8 mo), using SIC (43 %) or RIC (57% Background: Center specifi c outcome data are being discussed as useful tools for patients, the insurance industry, and politicians to assess quality of specifi c centers and to guide decision making processes. Hematopoietic stem cell transplantation (HSCT) requires signifi cant infrastructure and is costly. HSCT therefore qualifi es as a target for such an exercise. Methods: We made use of the comprehensive database of the Swiss Blood Stem Cell Transplant Group (SBST) to evaluate center specifi c mortality rates. Nine centers reported a total of 4717 HSCT, 1427 allogeneic (30.3%), 3290 autologous (69.7%) in 3808 patients between 1997 and 2008 to treat acute or chronic leukemia (29%), lymphoid neoplasia (53%), non malignant disorders (3%) or solid tumors (15%). Because of mandatory reporting by legal requirement these include data on all HSCT performed in Switzerland over this period. Data were analyzed separately for recipients of allogeneic and autologous HSCT for survival and transplant related mortality (TRM) at day 100 and at 10 years. Results: Overall survival at 10 years was 49 + 4% after allogeneic and 40 + 3% after autologous transplantation. There were signifi cant differences among centers in unadjusted analyses of survival and TRM. These differences became absent or marginal, when results were adjusted for disease, year of transplant and the EBMT risk score (incorporating patient age, disease stage, time interval between diagnosis and transplantation, and for allogeneic transplants donor type, and donor recipient gender combination) in multivariate analysis.

Conclusions: These data indicate comparable quality among centers in Switzerland. They furthermore demonstrate that comparison of crude center specifi c outcome data without adjustment for patient risk factors may be highly misleading. Mandatory data collection and outcome analysis including all cases within a comprehensive quality management system may serve as a model for other cost intensive therapies to ascertain quality. . Contrary, the more relevant negative aspect is the increase of workload (38%). The major modifi cation produced by the QMS is to recognize the critical situations (51%) and to report in proper forms the NCA is useful also to resolve them (36%). Also, 67% think that internal audits are helpful to check the activities. The last question aimed to understand the relevance of QMS: 58% agree that QMS is part of daily activity and only for 2% "there is too much to do". The collected data show that the quality culture is, today, part of the work background and this result is recognized by 70% of lab staff, 47% of nurses and 62% of physicians. In general, the perception of QMS is good but there are still areas to be improved, particularly regarding the management and resolution of critical situations. It is clear that to study the perception of QMS is a tool to verify his effi cacy and continuously improve it, to make clear the diffi culties and fi nd appropriate corrective actions. So, QT plans to repeat the survey at least once at year in order to build a complete evaluation's system for work environment' quality and staff satisfaction. Background: The liver is the current site for pancreatic islet transplantation (Tx) but has many drawbacks due to immunological and non-immunological factors as well as important technical limitations. We asked whether pancreatic islets could be engrafted in the bone marrow (BM), an easily accessible and widely distributed transplant site that may lack the limitations seen in the liver. Methods: Pancreatic islets were implanted into the BM of diabetic C57BL/6 mice. Islet survival, function and morphology were evaluated in comparison with the liver site. Moreover a phase I/II open label pilot study to assess the feasibility and safety of BM as alternative site for islet Tx was started in patients with type 1 diabetes mellitus. Results: Syngeneic islets engrafted effi ciently in the BM of C57BL/6 mice rendered diabetic by streptozocin (STZ) treatment. For over a year post-Tx these animals showed parameters of glucose metabolism that were similar to those of non-diabetic mice. Islets in BM had a higher probability to reach euglycemia than islets in liver (2.4 fold increase, p = 0.02), showed a compact morphology with a conserved ratio between alpha and beta cells, and affected bone structure only very marginally. Islets in BM did not compromise hematopoietic activity, even when it was strongly induced in response to a BM aplasia-inducing infection with lymphocytic choriomeningitis virus. In humans, 3 diabetic patients received an intra BM islet infusion at the level of iliac crest without any adverse effects. In all recipients islets engrafted as demonstrated by the presence insulin producing cell In addition, preferential pairing facilitated by introduction of an extra disulfi de bond in the constant regions of the TCR chains can increase cell surface expression of the transferred TCR and decrease formation of mixed TCR dimers. Another strategy based on the fact that TCRs differ in their capacity to compete for cell surface expression, is to select recipient T cells with weak competitor phenotypes. Both weak and strong competitor phenotype virus-specifi c T cells were used to assess improvement of HA-1-TCR cell surface expression using the different strategies. The most marked improvement in HA-1-TCR expression and functionality was observed after TCR transfer of a cysteine modifi ed and codon optimized HA-1-TCR resulting in 70% and 35% of HA-1 tetramer positive weak and strong competitor T cells, respectively. This resulted in effi cient recognition of primary leukemic cells that endogenously process and present HA-1, independent of whether the recipient T cells were strong or weak competitor T cells. Furthermore, results demonstrate that next to increased HA-1-specifi c reactivity, neoreactivity after HA-1-TCR gene transfer was dramatically reduced by cysteine modifi cation of the HA1-TCR. Based on these results, cysteine modifi ed and codon optimized HA-1-TCRs will be used for the planned phase I/II clinical trial to treat patients with relapsed hematological malignancies. Reactivation of latent cytomegalovirus (CMV) infection is a frequent complication in CMV-seropositive patients after allogeneic hematopoietic stem cell transplantation (HSCT). The S66 CMV-related morbidity and mortality are particularly high, if the donor is CMV-seronegative and thus, no CMV-specifi c memory T cells are being transferred from donor to recipient during transplantation. Grafting non-reactive T cells of CMVseronegative donors by virus-antigen specifi c T cell receptors (TCR) may be an effi cient means to transfer CMV-specifi c T cell function into allogeneic HSCT recipients. In this study, we have reprogrammed T cells of CMV-seronegative donors with human TCR recognizing the immunodominant HLA-A*0201-binding CMVpp65 peptide epitope 495-503. To overcome the limitations of retroviral TCR gene transduction that hamper clinical translation, we used in vitro transcribed RNA encoding CMV-specifi c TCR for electroporation of non-reactive T cells. This procedure resulted in transient surface expression of the introduced TCR for at least 3 days as demonstrated on both CD4 + and CD8 + T cells by specifi c TCRVb chain and HLA tetramer staining analyses. The CMVpp65 TCR RNA-transfected T cells showed HLA-A*0201-restricted IFN-g secretion and cytolysis against CMVpp65 495-503 peptide-pulsed target cells and against human fi broblasts upon CMV infection. We also observed that TCR RNA transfection of CD4 + T cells turned them into potent CMVpp65/HLA-A*0201-specifi c T helper cells. This was demonstrated by co-incubating them with immature dendritic cells (DC), which resulted in maturation of DC only in the presence of the CMVpp65 epitope. Furthermore, we transfected pure naïve and memory CD8 + T cell subsets isolated from peripheral blood of CMV-seronegative donors. Although 90% of naïve CD8 + T cells were CMVpp65/HLA-A*0201 tetramer positive after electroporation, they mediated only marginal lysis toward CMV-infected fi broblasts. In contrast, memory CD8 + T cells showed strong cytotoxicity against CMV-infected fi broblasts for at least 3 days upon TCR RNA transfection. In summary, our data demonstrate that non-reactive human T cells can be easily redirected with CMVpp65 TCR RNA, thereby gaining CMV-specifi c T cell effector function for a considerable time period. We therefore believe that CMVpp65 TCR RNA has the potential to be further developed as a therapeutic "off-the-shelf" reagent for CMV-seropositive patients who undergo allogeneic HSCT from CMV-seronegative donors.

Functional cytotoxic T lymphocytes emerge in patients with Ph + acute lymphoblastic leukemia under Imatinib therapy and may be exploited for adoptive T-cell therapy C. Quadrelli (1) An effective specifi c immune response against adult ALL has so far been reported only in the allogeneic stem cell transplant setting. We have recently demonstrated that BM-resident, IFNg-producing, p190BCR-ABL specifi c autologous T cells may occur in Ph + ALL patients upon treatment with high dose IM, and that the presence of these cells correlates with lower minimal residual disease (MRD) values and better clinical outcome.

On the basis of these preliminary data, we proceeded to assess for the presence of cytotoxic T-cells among the BM resident, BCR-ABL-specifi c T-cells, and characterized the cytotoxic T-cell populations identifi ed, with the aim of optimizing a protocol for the ex-vivo expansion of p190BCR-ABL-specifi c cytotoxic T lymphocytes (CTLs) to be employed in protocols of T-cell therapy to control minimal residual disease after chemotherapy or HSCT. p190BCR-ABL-specifi c CTLs were identifi ed by co-colturing BM mononuclear cells (BMMC) of Ph + LLA patients with BCR-ABL peptide-pulsed autologous dendritic cells; the peptides employed derived from the complete spanning of p190BCR-ABL fusion region. The p190BCR-ABL-stimulated T cell lines generated with this protocol were mainly CD8 + /CD3 + or CD4 + /CD3 + cells with effector memory phenotype, that exerted a specifi c lytic activity against BCR-ABL fusion protein. In detail, we observed a p190BCR-ABL specifi c lytic activity >100 LU10/106 in 7 of the 10 patients tested (median lysis against p190-derived peptidepulsed targets: 1355 LU10/106). In 3 of the patients, for whom autologous leukaemia blasts were available, we could demonstrate a strong leukemia-directed lytic activity (median lysis: 10000 LU10/106). The lysis directed towards autologous blasts was mainly mediated by CD8 + T-cells, and was HLA class Irestricted. Lytic activity towards autologous, mock-pulsed PHA blasts, or allogenic Ph-blasts was low/absent. Our fi ndings demonstrated that p190BCR-ABL-specifi c T cell responses may emerge in patients with Ph + ALL under IM treatment, and may be stimulated and expanded ex-vivo. Whether these p190BCR-ABL-specifi c T cells may have a possible therapeutic role in Ph + ALL patients, and may be expanded from healthy donors to be employed in protocols of adoptive cell therapy after HSCT remains to be addressed in future studies. Background: A positive selection of leukemia associated antigen (LAA)-specifi c T cells would be highly desirable to amplify the GVL effect and to decrease the risk of GVHD. Wilms Tumor gene 1 (WT1) and the receptor for hyaluronic acid mediated motility (RHAMM) are LAAs recognized by CD8 + T lymphocytes. Streptamers constitute a novel multimer technology to select specifi c CD8 + T cells, available at good manufacturing production (GMP) level. Material and methods: Both tetramers and streptamers were used to detect the frequency of HLA-A2 restricted CD8 + T cells in the naïve peripheral blood (PB) from both healthy donors (HDs) and AML patients. RHAMM-and WT1-specifi c CD8 + T cells were further characterized for the expression of CD27, CD28, CD45RA and CCR7. In the next step, LAA-specifi c cells were positive selected by MACS columns after labeling with streptamers and thereafter immunophenotyped. Moreover, mixed lymphocyte peptide cultures (MLPCs) were performed to enrich WT1 specifi c T cells derived from the PB of HDs. RHAMM and WT1 specifi c cells were subjected to carboxylfl uorescein succimidyl ester (CFSE) based proliferation and cytotoxicity assays. Results: 21 of 40 HDs showed naïve LAA specifi c T cell frequencies of 0.5 to 1.6% of all CD8 + T cells. In AML patients in complete remission signifi cantly higher frequencies of LAAspecifi c T cells than in patients at diagnosis could be detected, up to 6.0% of all CD8 + T cells. These cells revealed to be CD8 + CD27-CD28 + CD45RA + CCR7-effector T cells in fl ow cytometry. After positive selection by MACS columns a purity of 20-87% could be achieved for LAA-specifi c T cells. After a maximum of three rounds of MLPC, only a frequency of 2-5% could be achieved, thus demonstrating the power of the streptamer technology. Both tetramer-and streptamer-based selections yielded similar amounts of LAA-specifi c T cells. After positive selection, these T cells preserved an effector T cell phenotype and showed active proliferation in CFSE staining. Streptamerselected T cells showed a trend towards higher CD28 expression thus indicating a more activated T cell status. Conclusion: In summary, the streptamer technology allows to select highly pure fractions of RHAMM-and WT1-specifi c effector T cells with proliferative and cytotoxic properties. In analogy to DLIs specifi c for viral antigens such as CMVpp65, production of leukemia specifi c DLIs is feasible on a GMP level. Further leukemia antigens are currently evaluated by our group.

Delicate balance between regulation and stimulation at the antigen presenting cell site determines the likelihood of successful priming and survival of antigen-specifi c T-cells from the naïve donor repertoire I. Jedema, T.S. Lam, M. van de Meent, C. Hoogstraten, J.H.F. Falkenburg Leiden University Medical Center (Leiden, NL) Although the in-vitro induction of tumor-and pathogen-specifi c T cells from the naïve donor repertoire has been shown to be feasible, the robustness of the procedure remains limited, hampering large scale clinical application. In this study, we investigated the role of individual parameters like the frequency of antigen-specifi c precursor T cells (Tprec) and regulatory T cells (Treg), the number of antigen presenting cells (APC) used as stimulator cells, and the number of targeted antigens (Ag) on the ability to prime, enrich and expand Ag-specifi c T cells from primary immune responses in-vitro. Therefore, we developed an in-vitro model system allowing the monitoring of Ag-specifi c activation and proliferation of individual naïve donor T cells in the fi rst 14 days of the immune response using PKH labeling, CD137 counterstaining and quantitative fl owcytometric analysis. In this model we exposed naïve Tprec to allogeneic APC in different responder/stimulator (R/S) ratios in the presence of different numbers of innocent bystander cells. Optimal T cell activation was seen at specifi c Tprec/S ratios between 1/1 and 1/5, irrespective of the number of innocent bystander cells.

Lowering the number of stimulator cells per Tprec resulted in incomplete activation and proliferation, but more importantly, exposure to an excess of stimulator cells resulted in induction of activation-induced cell death (AICD) of the antigen-specifi c Tprec. Next, we investigated the role of Treg in the induction phase of the immune response. Treg were like Tprec attracted to the site of the APC and their activation further increased their inhibitory potential. Especially when they were at a numeric advantage, Treg were capable of impairing antigen-specifi c Tprec priming. Increasing the number of APC in the induction phase of the immune response could overcome this negative regulation since the number of Treg per APC is diminished. However, at low Tprec frequencies, in case of single peptide responses, this will lead to AICD of the antigen-specifi c Tprec. This may be prevented by increasing the number of antigenspecifi c Tprec by simultaneous targeting of multiple antigens.

In conclusion, the in-vitro generation of antigen-specifi c primary immune responses can only be successfully and reproducibly performed by creating an optimal balance at the priming site of the immune response (e.g. the APC) between the number of negative regulators (Treg) and responding cells (Tprec). Fistulas are an invalidating, often diffi cult to treat, complication of patients with Crohn's disease (CD), a disabling, chronic, relapsing infl ammatory enteropathy caused by dysregulation of the immune tolerance towards intestinal bacteria in genetically susceptible individuals. Mesenchymal stromal cells (MSCs) represent a promising tool in approaches of regenerative medicine. We investigated the feasibility, safety and effi cacy of local injection of autologous bone marrow (BM)-derived MSCs for refractory CD fi stulas. MSCs were isolated and expanded ex vivo in the presence of platelet lysate from BM of 12 patients (7 males, median age 33 yrs, range 16-59). Patients received intrafi stular injection of MSCs scheduled every 4 weeks (median 4 infusions) and were monitored at time of each injection, and 1, 3, 6, 12 months after the last treatment. The cytokine profi le of MSCs and their ability to infl uence apoptosis of mucosal T cells obtained from involved and uninvolved colonic areas were also analyzed. MSC expansion was successful in all patients and no adverse event was recorded during and up to 12 months after treatment. Intrafi stular injection of MSCs was effective in inducing sustained closure of fi stulas, with the appearance of regenerative tissue along the tracks. In particular, 7 patients (70%) benefi ted from complete and sustained healing of fi stula tracks, while three had partial response. All patients showed a signifi cant reduction of both CD Activity Index (pre-and posttreatment median values: 294 SD 49 and 99 SD 32 at 6 months after the last infusion; p < 0.001) and Perianal Disease Activity Index (pre-and post-treatment median values: 13.0 SD 2.2 and 4.5 SD 2.4 at 6 months after the last infusion; p < 0.001) reaching disease remission usually after the second procedure. The immunephenotype of circulating T lymphocytes showed progressive increase of the number of CD4 + CD25bright FoxP3 + cells, which became signifi cant (p<0.01) after the second procedure and remained stable up to 6 months after the last infusion. No modifi cation of serum cytokines was observed at any time point. MSCs caused a sort of block of the rates of both apoptotic and living cells when incubated with T lymphocytes from diseased mucosa, whilst critically increased the apoptotic rate when incubated with T lymphocytes from apparently healthy mucosa. Local injection of autologous BM-derived MSCs appeared feasible, safe and successful in treating fi stulas associated with CD. We recently showed that NK cells require two signals to initiate lysis; the fi rst primes (S1) and a second triggers (S2). We found NK resistant cell lines which express S1 but lack S2 and that resting NK cells stimulated with these cells retain the primed state even after cryopreservation. These are termed Tumour Activated NK cells (TaNK). We are conducting a clinical trial in patients with AML in CR or PR with less than 25% blasts who have exhausted conventional treatment options. TaNK are generated from a haploidentical family donor by overnight incubation of NK cells with lysed CTV-1 cells. TaNK are then purifi ed from the lysate and released at a single dose of 10 6 NK/Kg with a T cell contamination of <10 4 /kg and cryopreserved. Pre-infusion conditioning -Fludarabine (25mg/m 2 /day) for 3 days plus 2Gy TBI on day 4. Eleven patients have been enrolled to date, 6 of whom have been treated. Patient Characteristics: Pt01 -57 yr female; previous autologous transplant, treated in CR3. Pt02 -71 yr male; in PR after 5-azacytidine. Pt03 -50 yr male; previous allogeneic transplant, treated in CR3. Pt04 -71 yr male; treated during relapsing disease and circulating blasts. Pt05 -73 yr female treated in CR1 following one course of induction chemotherapy. Pt06 -67 yr male; treated in CR2. Results: No infusional toxicity. All patients suffered a degree of bone marrow suppression needing in-patient supportive care. Aplasia ranged from 3 weeks up to 45 days. Pt01 remains in CR at month + 16; Pt02 achieved and remained in CR until month + 11, relapsed and treated with second TaNK infusion; Pt03 prolonged and severe pancytopaenia requiring CD34selected stem cell rescue, achieved and remained in CR until month + 11.5; currently undergoing re-induction chemotherapy. Pt04 cleared peripheral blasts for 2 months; now with progressive disease. Pt05 in CR, 3 months post infusion; Pt06 is non-evaluable at present. Extremely prolonged NK chimerism (up to + 6 months) has been seen in all patients in the absence Most cord blood transplants (CBT) are performed using cord blood units mismatched with the recipient at one or two HLA loci. It is not known whether the locus at which mismatches occur infl uences outcomes. We examined the effect of locusspecifi c mismatches on outcomes of 1305 HLA-mismatched (one mismatch, n = 626, two mismatches, n = 679) single CBT performed for haematological malignancies at EBMT centres from 1994 to 2008. Two digit typing for HLA-A and HLA-B and 4 digit typing for HLA-DRB1 was used. Patients were transplanted for ALL (n = 563), AML (n = 411), MDS (n = 166), chronic leukaemia (n = 87), lymphoma (n = 22) or myeloma (n = 10). Among the 626 patients who received a cord with one HLA mismatch, the mismatch occurred at HLA-A (n = 191), HLA-B (n = 247) or HLA-DRB1(n = 188). Transplant-related mortality (TRM) for this group at 2 years was 37±2% and was 35±4%, 35±3% and 42±4% for HLA-A, -B and -DRB1 mismatches respectively. On multivariate analysis the mismatched locus had no infl uence on TRM when adjusted for age, year of transplant, stage of disease, CMV status and cell dose. There was a trend for less relapse for DRB1 mismatches (21±3% vs. 29±3%, p = 0.051) but this did not lead to a difference in disease-free survival (DFS) between groups (38±4% vs. 39±3%, p = NS). Neutrophil engraftment and acute graft-versus-host disease (GVHD) were unaffected by locus of HLA mismatch. For the 679 patients who received a cord with two HLA mismatches, mismatch combinations included A + B (n = 250), B + DRB1 (n = 220), A + DRB1 (n = 122) or two mismatches each at either A (n = 20), B (n = 34) or DRB1 (n = 33). TRM for this group was 42±2% at 2 years and was 41±3% (A + B), 41±4% (B + DRB1), 39±5% (A + DRB1), 55±12% (A + A), 42±9% (B + B) and 58±9% (DRB1 + DRB1) by mismatch group. The loci at which HLA mismatches occurred did not infl uence TRM on multivariate analysis, nor when grouped according to MHC class (class I only vs. class II±I). However, double DRB1 mismatch was associated with more grade II-IV acute GVHD (50±8% vs. 32±2%, RR = 1.77, p = 0.03). In addition, on multivariate analysis A + B mismatches were associated with improved probability of DFS (34±3% vs. 30±2%, HR = 0.80, p = 0.03). A cell dose cut-point could not be defi ned in either population. In summary, for CBT with one HLA mismatch, the locus at which mismatch occurs does not infl uence outcomes. However, for CBT with two HLA mismatches, it seems that two mismatches at class I (HLA-A and -B) are preferable to other mismatch combinations.

Matched unrelated donor is associated with lower relapse than matched related donor in reduced-intensity conditioning transplantation: implications for graft-versus-malignancy effect V. Ho, H. Kim, J. Aldridge, G. Kao, D. Liney, J. Koreth, P. Armand, C. Cutler, J. Antin, R. Soiffer, E. Alyea Dana-Farber Cancer Institute (Boston, US) As success of reduced intensity conditioning (RIC) stem cell transplantation (SCT) relies on graft-vs-malignancy (GvM) effect, the extent of minor HLA antigen disparity in matched related donors (MRD) vs. matched unrelated donors (MUD) could impact clinical outcomes. While convention dictates that MRD is preferred over MUD, does this wisdom hold in RIC SCT?

We conducted a retrospective review of 477 (279 MUD, 198 MRD) RIC SCT performed at our institution from 9/01 through 12/08. All received uniform RIC conditioning with Busulfex (3.2-6.4 mg/kg) and Fludarabine (120 mg/m 2 ). GVHD prophylaxis was mostly tacrolimus/mini-methotrexate ± sirolimus (98%). SC source was PBSC (97%), marrow(3%). All donors were 10/10 allele matched at HLA A, B, C, DRB1, DQB1. The MUD and MRD cohorts were balanced in demographics, disease (MDS/AML 44% vs. 45%), high disease risk (81% vs. 82%), prior SCT (30% vs. 33%), and GVHD prophylaxis. Median CD34 + dose was higher in MUD (8.7 vs. 7.5 × 10 6 CD34 + /kg, p = 0.003). Cumulative incidence of grade II-IV acute GVHD and 2-yr chronic GVHD were similar, 21% vs. 16% (p = 0.21), and 50% vs. 47% (p = 0.17) in MUD and MRD, respectively. There was no difference in transplant related mortality (TRM), time to neutrophil or platelet engraftment, or % achieving over 90% donor chimerism at days + 30 and + 100. Cumulative incidence of relapse was 52% for MUD, 67% for MRD, p = 0.003. With a median follow-up of 25 and 28 months, 2-yr progression free survival (PFS) was 41% for MUD and 29% for MRD, p = 0.004 ( Figure 1 ). 2-yr overall survival (OS) was 57% for MUD, 49% for MRD (p = 0.35). Improvement in PFS did not translate to OS benefi t because many relapses were salvaged with second SCT or other therapy. In competing risks regression, MRD was associated with increased risk for relapse compared to MUD (HR 1.45, p = 0.004), but no difference for TRM. In Cox regression, MRD, prior SCT, high disease risk, and MDS/ AML diagnosis were associated with inferior PFS (Table 1) . Donor age, CD34 + dose, and year of SCT were not associated with worse survival. In RIC SCT, MUD is associated with signifi cant improvement in relapse and PFS compared to MRD, without increased GVHD or TRM. This improvement is not related to younger donor age or higher CD34 + count, suggesting that HLA minor antigen disparity in MUD SCT may mediate stronger GvM. While further investigation is warranted, these results could have great implications for the future selection of donors for RIC SCT. Tolerance of G-CSF-administration was different in the two groups: 12/171 (7,01%) SIB donors stated, that they only poorly had tolerated G-CSF administration, 4/106 (3,77%) MUD donors said that their tolerance to G-CSF had been poor. 3/171 (1,75%) SIBs indicated, that the infl uence of PBSC donation to their health status had been negative, while 2/106 (1,88%) unrelated donors said that there had been a negative infl uence of PBSC-donation on their health status. 5/171 (2,92%) SIBs described their current health status as "bad", while none of the 106 MUDs did so. The most serious adverse effects that occurred after donation were one case of Hodgkin's disease in one of the related donors and one case of acute lymphatic leukaemia in an unrelated donor. Furthermore, one case of squamous cell lung carcinoma occurred in a related donor and one case of mamma carcinoma in an unrelated donor. In addition to these malignancies the following health problems could be noted in the donors: Subdural haematoma a few days after donation: 1 SIB, 0 MUD. Sarcoidosis: 1 SIB, 0 MUD. Cardiac disorder: 2 SIBS; 0 MUD. Essential hypertension: 4 SIBs, 4 MUDs. Discus hernia: 0 SIB, 4 MUDs. Hypothyreosis: 0 SIB, 1 MUD. Depression: 1 SIB, 1 MUD. Tinnitus: 0 SIB, 2 MUDs. The presented data show that serious health problems after donation may occur and therefore global collection of these data is necessary in order to calculate the actual risks of PBSC donation in advance.

Evaluation, consent and care of related donors and recipients at haematopoietic stem cell transplant centres in the United States: practice patterns and potential for confl ict of interest P. Anderlini (1), T. Pedersen (2), D. Confer (2) Background: A real or perceived confl ict-of-interest may arise in the situation where the same physician is responsible for the care of two individuals whose care is interdependent. In hematopoietic stem cell transplantation (HSCT) from unrelated donors facilitated by the National Marrow Donor Program, donors and recipients are under the care of separate physicians in order to provide unbiased care and eliminate the potential for a confl ictof-interest. However, the practice patterns of evaluation and care for related donors and recipients at centers performing HSCT are unknown. Material and methods: A practice pattern survey of transplant centers in the United States reporting to the CIBMTR was conducted by the Donor Health and Safety Committee between December 2007 and July 2008. The survey was administered as an online survey tool sent to the center medical directors. The survey focused primarily on determining the type of provider involved in medical clearance, informed consent and medical management of the donor and the providers relationship to the recipient. Results: Of 222 centers surveyed, 98 evaluable responses were received (40%). The median number of related donor transplants per year at responding institutions was 15 (range: 1-400); the median total number of transplants per year was 70 (5-600). As shown in the Table, transplant physicians in greater than 70% of centers were involved in overlapping care of the donor and the recipient during the donor evaluation, clearance and collection phases. These patterns were similar between transplant teams caring for adult or pediatric donors and recipients. Conclusion: Among responding centers, it appears that medical management of recipients and their related donors by the same provider is common, and may not be viewed as a potential confl ict-of-interest. Whether this potential confl ict-of-interest affects donor care is unclear, and deserves further investigation.

Long-term follow-up and risk factors for outcomes after related HLA-identical cord blood transplantation for patients with malignancies. An analysis on behalf of Eurocord-EBMT A.L. Herr (1) , N. Kabbara (1), P. Teira (1), C. Bonfi rm (2) Outcomes after related HLA identical bone marrow or cord blood transplantation (CBT) in children have been reported to be comparable (NEJM 2000) . In order to analyze risk factors for outcomes, we studied 149 patients with malignancies who had a fi rst single unit related HLA identical unmanipulated CBT and were reported to Eurocord-EBMT. CBT were performed since 1990, in 25 countries and 68 centers. Median FU was 6.7 years (range 7 months to 17 years). Nearly all patients were children, median age being 5 years. Acute leukemia (AL) was the main diagnosis (n = 109) followed by myelodysplasia (n = 17), chronic myeloid leukemia (n = 12), solid tumors (n = 6), lymphomas (n = 3) and hemophagocytic lymphohistiocytosis (n = 2). According to IBMTR criteria, 39 patients had low, 71 intermediate and 39 high risk diseases. All CB donors were siblings except 2: one 2nd degree relative and one child. CB units contained a median of 5.1 × 10 7 /kg total nucleated cells (TNC) at collection and 4.1 × 10 7 /kg TNC after thawing. The conditioning regimen was myeloablative for 144 patients, 50% including TBI. The most frequent GVHD prophylaxis regimen was cyclosporine alone. Methotrexate (MTX) was used for 22 patients. The cumulative incidence (CI) of neutrophil recovery was 89% at d + 60 with 135 patients achieving a neutrophil count of 0.5 × 10 9 /l after a median time of 24 days. Infused TNC ≥ 4.1 × 10 7 /kg (p = 0.002) and the lack of MTX in GVHD prophylaxis (p = 0.002) were independently associated with improved neutrophil recovery. CI of acute and chronic GVHD was 11% at d + 100 and 10% at 2 years, respectively. CI of non-relapse mortality was 9% and of relapse was 46% at 5 years. Remission status (p = 0.04) and the use of TBI-containing regimens (p = 0.03) were independently associated with a lower relapse rate in AL patients. For CBT performed in year <2000 vs. ≥2000, disease-free survival at 5 years was 34% vs. 55% (p = 0.009), and overall survival (OS) at 5 years was 44% vs. 66% (p = 0.01). Transplant year ≥2000 (p = 0.005), low or intermediate risk disease status at CBT (p = 0.008) and infused TNC ≥ 4.1 × 10 7 /kg (p = 0.04) were independently associated with improved OS. Five-year OS in AL patients was associated with remission status (p = 0.001): 77% for patients in CR1, 50% in CR2, 32% in CR3 and 21% in advanced phase of the disease. Improving results over time and the low toxicity of related HLA-identical CBT in malignancies support banking and use of family CB units with suffi cient cell dose. In the past, the outcome of adolescents with acute lymphoblastic leukaemia (ALL) in 2nd complete remission (CR) given HSCT S72 prophylaxis was carried out with ATG or OKT3. Primary engraftment occurred in 85%. After reconditioning, fi nal engraftment was achieved in 98%. GvHD grade 0-1 occurred in 86%. 12% had grade II, 4% had grade III. Chronic GvHD occurred in 4 patients. No GvHD related mortality was observed. Median follow up was 5.9 years. EFS at 1 year was 56% (CR1), 51% (CR2) and 50% (≥CR3); EFS at 5 years was 49% (CR1), 46% (CR2) and 27% (≥CR3). Median survival of patients with active disease was 0.2 years. Relapse rates at 1 year were 0.32 (CR1), 0.47 (CR2) and 0.46 (≥CR3); relapse rates at 5 years were 0.32 (CR1), 0.47 (CR2) and 0.64 (≥CR3). TRM at 1 year was 14%, causes of death remained viral and fungal infections, no EBV LPD occurred. Conclusions: positive selection of progenitor cells or depletion of T and B cells can minimize acute and chronic GVHD in both matched unrelated and mismatched related transplantations and may prevent GvHD related mortality. Lethal infections occurred in 14% of the patients probably due to the delayed recovery of T cells. Despite profound depletion of donor T cells, relapse rates were acceptable and remained on a stable level after the fi rst year in patients with complete remission. Thus, absence of GvHD may not necessarily be associated with high relapse rates in childhood ALL. Apart from T cells, other effector cells like NK cells are likely to exert GvL effects and may contribute to the favorable EFS of our patients. Fanconi anemia (FA) is a rare disorder characterized by progressive bone marrow failure, congenital abnormalities and a predisposition to cancer. BMT is the only treatment able to restore normal hematopoiesis in this disease. Here, we update the Brazilian experience using only Cyclophosphamide 60 mg/ kg (CY60) as a preparative regimen for the treatment of pts with FA. Objective: Analyze the outcome of 70 pts with FA submitted to a related BMT using CY60 mg/kg. Material and methods: Period: 07/1999-07/2009; Gender: 32F/38M Age: 5-34 ys (M: 10,5). Type of donors: Matched siblings donors (MSD): 60 pts; Other related donors (ORD): 10 pts. 68 pts were in aplastic phase while 2 had MDS. All pts received a preparative regimen with CY60 mg/kg and GVHD prophylaxis with cyclosporine and methotrexate. Statistical analysis was performed using SPSS. Overall survival (OS) was estimated using the Kaplan-Meier method. Multivariate analysis was performed by Cox method. Results: 60 pts are alive between 135-3530 days (M: 1663) after BMT. OS: 85,5% at 5 ys. Pts transplanted below the age of 10 (37pts) had an OS of 97,1% at 5 ys. There was no statistical difference between transplants performed in Curitiba(56pts) and other BMT centers (86,1% × 83,2%). 2 pts died before D + 28 and were not evaluable for engraftment. Only one pt had primary graft failure and he is alive and well 5 ys after the 3rd transplant. Late rejection occurred in 6 pts (2 pts with MDS, 2 pts with ORD, 2 pts with MSD) at a median of 234 days after BMT. 2/6 pts are alive and well after a 2nd BMT. Cumulative incidence of rejection at 1 y was 10%. Pts receiving > 25 transfusions had a lower survival when compared to those less transfused (91% × 62% p = 0,03). Mucositis grade II-III occurred in the majority of pts;13/67 pts developed Acute-GVHD grade II-IV has been reported to be worse than that of children. In order to evaluate whether in recent years the outcome of adolescents has become more similar to that of children, we studied paediatric patients (age < 18 years at transplantation) reported to the AIEOP-HSCT Group Registry who received allogeneic HSCT for ALL in 2nd CR in a paediatric institution between 01/1990 and 12/2007. A total of 400 patients (64% males and 36% females) were anallyzed. 339 patients were children (<14 years old), whereas the remaining 61 patients where adolescents (14-18 years old). Median age at HSCT was 8.3 and 16.1 years, for children and adolescents, respectively. A MFD was employed in 50% of children and in 58% of adolescents; the remaining patients having received a MUD HSCT. Conditioning regimen included TBI in 90% of patients, in most cases associated with thiotepa and cyclophosphamide. Cyclosporine (CsA) alone was used for GvHD prophylaxis in 88% of patients transplanted from a MFD, whereas the combination of CsA, MTX and anti-thymocyte globulin (ATG, 3.75 mg/kg from day -4 to day -2) was employed in 77% of patients given MUD HSCT. No difference between children and adolescents was observed for patient-and transplant-related variables, with the exception of the infused cell dose which was greater in children.

As of October 2009, the probability of developing grade II-IV, grade III-IV acute GVHD or chronic GvHD was 54%, 19% and 32% for children, and 49%, 14% and 43% for adolescents (P = NS). Eight-year probability of EFS,TRM and RI were 53%, 22% and 25% for children and 55%, 23% and 22% for adolescents, respectively (P = NS). Univariate analysis showed that S3 + S4 BFM classes at relapse (P<0.0001), recipient male gender (P = 0.01), MUD HSCT (P = 0.02), grade 0 or IV acute GvHD (P<0.0001) were poor-risk factors for EFS. Only BFM class at relapse and acute GvHD (grade 0 or IV) maintained their prognostic signifi cance in multivariate analysis. Adolescents and children <14 years of age, treated in paediatric institutions, had the same outcome. Multivariate analysis confi rmed the GvL favourable effect in patients with grade I-III acute GvHD.Results obtained with MUD HSCT were comparable to those achieved with MFD HSCT. Adolescents should be referred for transplantation to treatment teams that have experience in the management of childhood ALL.

Long-term survival and relapse rate after transplantation of highly T and B cell-depleted stem cells from alternative donors in paediatric patients with acute lymphatic leukaemia P. Lang (1) We present long term results in 68 children with ALL who received highly T and B cell depleted stem cells from matched unrelated (n = 17) or full haplotype mismatched related donors (n = 51) at our institution. Our aim was to minimize GvHD and to avoid EBV LPD in both matched and mismatched transplantations. Remission status was: CR1, n = 18; CR2, n = 23; ≥CR3, n = 12; non remission or second transplant, n = 15. Graft manipulation was carried out with microbeads and the CliniMACS TM device (indirect depletion of T and B cells with CD34 + positive selection (n = 50); or direct depletion with anti-CD3/anti-CD19 coated microbeads (n = 18)). T and B cell counts were reduced for 4-5 log with median numbers of residual T cells of 15 000/kg bw (CD34 + selection) and 49 000/kg bw (CD3/19 depletion). No pharmacological immune suppression was given after positive selection, whereas patients with CD3/19 depletion received MMF. The conditioning regimens were either TBI or Bu based (n = 50) or a toxicity reduced protocol (Flud, TT, Mel) was used (n = 18). Rejection

more than 500 candidate donors were evaluated at our Institution (median age 43 y, r 16-79 y; 185 donors ≥50 y; male 283). 460 donors were evaluated with marrow smear and cytogenetic test, 328/460 were eligible for donation (4 bone marrow harvest, 324 peripheral blood GCSF mobilization), 10 proceeded to stem cell donation without marrow evaluation. 393 out of 460 marrow were morphologically normal (85,43%), 58 (12,63% -median age 54 y, 31 donors >50 y) presented abnormalities, 9 (1,96%) were not evaluable. We registered 19 cases of plasma cells hyperplasia (10 not eligible to donation)

7 (1,52%, 5 donors > 50 y) presented abnormalities, 1 was not evaluable. We registered 5 cases of chromosome Y deletion (4 not eligible to donation), 1 of pericentric inversion of chromosome 3 (p22;q29

Additional follow-up will give us more important information on safety of HSC donation procedure. O373 Impact of KIR-ligand mismatches and KIR genotypes on the outcome of patients receiving unrelated donor stem cell transplantation for lymphoid malignancies M. Morelli (1), A. Bermema (1)

All pts received rabbit anti-thymocyte globulin for in-vivo T cell depletion. Nineteen donor/recipient pairs were matched at molecular level for the HLA-A, B, C, DRB1, DQB1 loci and 30 were mismatched (27 at class I, 6 at class II). The majority of pts had chemosensitive disease (n = 38, 78%). The median follow-up was 24 months (range 3-65). We studied KIR genotypes using the KIR-typing kit (Miltenyi Biotec) and HLA-Cw KIR ligand using high-resolution molecular techniques. Results: Twenty-one pts were C1/C2 heterozygous, 16 were C1/C1 and 12 were C2/C2 homozygous. In the combination patient C1 absent/donor KIR2DL2/KIR2DL3 present (n = 12, 25%), we observed no TRM and only 3 cases of acute GVHD (aGVHD) as compared to other combinations (n = 37) in which 3 cases of TRM and 11 cases of aGVHD were observed. This combination was associated with a trend of better cumulative incidence of TRM, OS, and PFS

OS: 61% vs

67%) or in the recipient (n = 36, 74%) showed no signifi cant association with a better OS or PFS. sibling donors (n = 8), haploidentical CD3/CD19 depleted grafts (n = 3), CD34 + selected graft with OKT3 (1)

HLH (5), leukocyte adhesion defi ciency (4), chronic granulomatous disease (3), severe immune dysregulation (3), other PID (15)

MFD (13)

HLA matched 7-10/10). 16 of 45 were cords (7-10/10 HLA matched). Follow up is 2-59 months (median 16 months)

13 children died: 6/40 in the treosulfan/fl udarabine group, 7/30 in the treosulfan/cyclophosphamide group -HLH disease day-1, graft rejection and CMV, infection (4), GVHD, GVHD +infection, GVHD + cerebral infarcts, cerebral haemorrhage, pneumonitis, VOD and MDS/ AML. Skin toxicity was common. VOD occurred in 2 children in combination with cyclophosphamide and both had enterovirus in the gut. 2 patients rejected (MHC II defi ciency successfully retransplanted

Autoimmunity after BMT in primary immunodefi ciency diseases: single-centre report of 184 children

had AML (10 in CR1 at high risk, 10 in ≥CR2 and 2 in relapse), 5 had ALL (4 in CR1; 1 in relapse) and 1 had high grade NHL in relapse. Conditioning was: 8Gy single fraction TBI, thiotepa (4 mg/kg × 2), fl udarabine (40 mg/m² × 5), cyclophosphamide (35 mg/kg × 2)

No GvHD developed in 24/26 patients, 2 developed ≥ grade II GvHD. Ten patients died (3 VOD, 2 fungal pneumonia, 1 bacterial sepsis, 1 CNS aspergillosis, 1 systemic toxoplasmosis, 1 adenoviral infection, 1 MOF). CD4 and CD8 counts reached, respectively, 50/μL medianly on days 34 (range 19-63 days) and 24 (range 15-87)of CMV reactivation were signifi cantly fewer than after our "standard haplo" transplants. In KIR ligand-mismatched transplants, speed of NK cell reconstitution/maturation and size of donor vs. recipient alloreactive NK cell repertoires were preserved. In conclusion, in the setting of haploidentical transplantation infusion of Tregs makes administration of a high dose of T cells feasible for the fi rst time

Donor NKT cells and outcome following HLA-identical peripheral blood stem cell transplantation D. Nachbaur

Oral Session 15: Myelodysplasia / Regulatory issues and Quality management O350 Monosomal karyotype, higher blast count and ex vivo T-cell depletion predict poor outcome after allogeneic stem cell transplantation for MDS/SAML patients with chromosome 7 abnormalities M. van Gelder, J. Schetelig, L. Volin, J. Maertens, G. Socié, E. Petersen, H. Thomssen, A. Biezen, R. Brand, T. de Witte, N 

Background: High-risk MDS/sAML patients with a chromosome 7 abnormality have a poor prognosis with conventional therapies. These patients usually proceed to allogeneic SCT (allo-SCT). Data on the effect of this approach are scarce. Objective. Description of outcome and identifi cation of risk factors of alloSCT in MDS/sAML patients with a chromosome 7 abnormality. Methods: From the EBMT database 277 patients with MDS/sAML having any chromosome 7 abnormalities (median age 45 years) who underwent fi rst alloSCT between 1981 and November 2006 were identifi ed. Stem cells from related donors were transplanted in 191 patients (177 HLA-identical) while 85 received unrelated grafts. Bone marrow was used as stem cell source in 148 patients. Standard conditioning was applied in 222 patients. Sixty-three patients fulfi lled the criterion for complex cytogenetic abnormalities (CA). A monosomal karyotype (MK), defi ned as at least one autosomal monosomy and at least one other chromosomal abnormality (Breems DA et al., JCO 2008; 26:4791) , was present in 68 patients, of whom 24 did not have CA. Results: Median follow-up (FU) of patients alive at last FU was 5 years (range 0-18 years). Estimate 5-year overall (OS), and relapse-free (RFS) survival was 28 ± 6% and 26 ± 6% respectively. The relapse rate at 3, 12 and 60 months was 9 ± 3%, 29 ± 5% and 39 ± 6% resp. Non-relapse mortality at 3, 12 and 60 months was 21 ± 5%, 36 ± 6% and 40 ± 6% resp. In multivariate models including age, MK, CA, % blasts at alloSCT, donor type, sex match, conditioning regimen, T-cell depletion and year of alloSCT three factors remained statistically signifi cant predictors for poor outcome: MK, ≥5% blasts at alloSCT and ex vivo TCD (adjusted HR for OS resp. 1.65 [1.15 -2. 35 95%CI], 1.67 [1.14 -2. 46 95%CI] and 1.83 [1.21 -2.76 95% CI] resp.). When patients with ex vivo TCD were excluded, 5-year OS in the remaining 97 MDS/sAML patients with any chromosome 7 abnormality and <5% blasts at alloSCT with or without MK was 8 ± 14% and 53 ± 12% resp. and for the 91 patients with ≥5% blasts 5-year OS was 9 ± 12% and 29 ± 11% resp (see Figure) . Conclusion: This large study on outcome of alloSCT for patients with MDS/sAML and any chromosome 7 abnormality shows that long-term survival can be achieved in patients without the poor risk factors MK, ≥5% blasts at alloSCT and ex vivo TCD. For patients with MK new approaches that tackle the high and early relapse rate are warranted.

Bone marrow fi brosis on outcomes of patients with MDS/sAML undergoing allogeneic stem cell transplantation N. Kröger, T. Zabelina, A. van Biezen, R. Brand, T. 

Bone marrow fi brosis has an important impact on the prognosis of patients with MDS. We evaluate the impact of bone marrow fi brosis in 721 patients who underwent allogeneic hematopoietic stem cell transplantation (HSCT) for MDS/AML and were reported to the EBMT. No fi brosis was noted in 483 pts, mild or moderate fi brosis in 199 pts and severe fi brosis in 39 pts. Diagnosis in the none, mild/moderate and severe fi brosis group were RA/RARS (36%, 39% and 30%), RAEB (43%,43% and 48%) and RAEB-t (21%, 18% and 20%). Stem cell source were from related (63%, 63% and 62%) or unrelated (36%, 36% and 38%) donors. Leukocyte engraftment (>1.0 × 10 6 /L) was observed after a median of 16, 17 and 19,5 days for the days for the none, mild/moderate and severe fi brosis group, respectively (p = 0.002). There was a trend for more graft failure in severe fi brosis group (10%, p = 0.07). In a multivariate analysis bone marrow fi brosis did not infl uence non-relapse mortality, but severe fi brosis signifi cantly infl uenced risk of relapse (HR 5.7, p = 0.02), resulting in a reduced disease-free (HR 5.9, p = 0.01) and overall survival (HR 7.5, p = 0.006). Other factors for reduced survival were advanced disease at transplant (HR 14.9, p = 0.005), non-CR before transplant (15.8, p = 0.001) and HLA-mismatch (HR 5.3, p = 0.02). In summary, apart of the disease status, no complete remission at transplant, and HLA mismatch transplantation, severe bone marrow fi brosis is an independent prognostic factor for disease-free and overall survival for MDS patients undergoing allogeneic stem cell transplantation.

Management and outcome of myelodysplastic syndrome (MDS) and secondary acute myeloid leukaemia relapsing after allogeneic stem cell transplantation -A retrospective analysis on 888 patients by the MDS Subcommittee of the EBMT Chronic Leukaemia Working Pa C. Schmid, A. van Biezen, R. Brand, J. Finke, L. Volin, A. Vitek, D. Selleslag, D. Heim, P. van dem Borne, A. Ganser, M. Mohty, M. Boogaerts, T. de Witte, N 

The aim of this study was to obtain reliable data on management and outcome of relapse after allogeneic SCT for MDS and sAML in adults. Median age of 888 patients was 51 years. Patients (pts.) had been transplanted for RA/RARS (8%), RAEB (19%), RAEB-T (13%) and sAML (60%), either previously untreated (20%), in CR (44%), relapse/progression (13%) or primarily refractory (24%). Standard (SIC) and reduced intensity (RIC) transplants were performed in 59% and 41%, donors were id siblings (58%), MUD or mm relatives (42%), and 5 syngeneic twins. Median time from SCT to relapse was 6 mo . Median follow up from relapse of 188 survivors was 6.7 mo. Overall survival (OS) at 1, 2 and 4 years was 24%, 16% and 10%. In a cox regression model, the MDS subtype at time of transplant (RA/RARS vs. RAEB [HR 1, 524, CI 1, 390, CI 0, 071 vs. sAML HR = 2, 209, CI 1, 114, p<0.001) , and the interval from SCT to relapse (1st vs. 2nd [HR = 0, 770, CI 0, 969] vs. 3rd [HR = 0, 609, CI 0, 768] vs. 4th quartile [HR = 0,404, CI: 0,318-0,513], p<0.001), were associated with survival. In 342 pts., data on the management of relapse were available. 223 pts had received DLI. Among them, median time from SCT to relapse was 6.5 mo, median time from relapse to DLI1 was 1 mo. 149, 42 and 32 patients received 1, 2, and 3 DLI. Median OS from relapse was 7.6 mo. Again, remission duration after SCT (p<.001) and less advanced disease at time of SCT (p = .038) were associated with better OS.

Constitutional variability in genes involved in innate immunity (IRF-3, HAMP, PTX3) and in cell proliferation (ATBF1 and NFAT5) infl uences disease free survival after allogeneic stem cell transplantation B. Martín-Antonio, I. Alvarez, F. Márquez-Malaver, P. Trujillo, M. Carmona, J. Falantes, I. Espigado, Á. Urbano-Ispizua Hospital Universitario Virgen del Rocío (Seville, ES) Genes involved in innate immunity and in regulation of cell proliferation may play an important role in infections and modulating the intensity of the infl ammatory response after allogeneic stem cell transplantation (allo-SCT). Thus, genetic variability in donor and recipient in these genes might be an important factor infl uencing the outcome of allo-SCT. We have studied the potential infl uence of 15 single nucleotide polymorphisms (SNPs) in donor and recipient in genes of innate immunity (IRF-3, HAMP, PTX3, HBD2) and regulation of cell proliferation (ATBF1, NFAT5, AKT2, NM1, CD151, TCIRG1, SH3KBP1) on clinical outcomes after allo-SCT, specifi cally on the incidence of acute GvHD, transplant related mortality (TRM), relapse, and disease free survival (DFS). Study population consisted of 106 donor-patient pairs undergoing HLA identical sibling allo-SCT in a single institution. Patient median age was 38 years (range, 5-66). 42% of the patients were in advanced phase of disease. Cumulative incidence for acute GvHD, TRM, relapse and DFS was computed with the cmprsk package and with Kaplan-Meier. Patient IRF3 rs2304205 AA, donor ATBF1 rs719327 AA and patient AKT2 rs12460555 CC dominant genotypes, were associated with a higher incidence of relapse (p = 0.02, p = 0.04 and p = 0.002, respectively) and lower DFS (p = 0.02, p = 0.04 and p = 0.009, respectively). All of them retained signifi cance at multivariate analysis. Variant rs719327 AA in ATBF1 showed the same prognostic values when present in donor or in patient (relapse: 55% vs. 34% and DFS: 26% vs. 46%). When it was present both in donor and patient, the differences were more prominent (relapse: 69% vs. 33%, p = 0.003 and DFS: 15% vs. 45%, p = 0.01). Donor HAMP rs7251342 AG genotype and donor NFAT5 rs6499244 AA dominant genotype were associated with a lower and a higher incidence of TRM (p = 0.007 and p = 0.02, respectively) infl uencing in a higher and lower DFS (p = 0.04, p = 0.02, respectively). They retained its signifi cance at multivariate analysis. NFAT5 is necessary for optimal T cell development and rs6499244 AA variant showed the highest mRNA expression. Interestingly, none of the 25 patients with a donor carrying PTX3 rs18040680 AA recessive genotype had TRM but showed a higher tendency in the incidence of relapse (61%). In conclusion, genetic variability in innate immunity and in cell proliferation has a strong infl uence on the clinical outcome of allo-SCT, which might be important when choosing allo-SCT protocols.

The factors to predict adverse events occurred within 30 days of post-peripheral blood stem cell donation at family donors Y. Kodera, S. Kim, K. Nagafuji, M. Hino, K. Miyamura, R. Suzuki, A. Tamakoshi, M . Fukushima on behalf of The Japan Society for Hematopoietic Cell TransplantationBackground: It has become obvious that certain adverse events might occurre at peripheral blood stem cell (PBSC) donors during or after the donation process. It is therefor crucial that certain factors which can predict the occurrence of such adverse events at normal donor are cralifi ed to prevent the adverse events. The JSHCT has continued the nation-wide consecutive follow up of PBSC family donors and in this project, the Day 30 Report of post donation was included. This time, we present the outcome of the Day 30 Report analysis. Materials and methods: Among 3,264 PBSC family donors who were consecutively pre-registered to JSHCT Donor Center between April 2000 and March 2005, 2,873 Day 30 Reports were obtained and subjected for analysis. The relationship between family donors' 1) gender, 2) age, 3) body-weight, 4) daily and 5) total dose of granulocyte-colony stimulating factor (G-CSF), 6) current and 7) past health conditions and the occurrence of 1) thrombocytopenia, 2) prolongation of hospitalized period, 3) any adverse events (bone pain, fatigue, head ache, insomunia, anorexia, nausea, vomiting, splenomegary) as well as the mobilized CD34 + cell numbers were statistically examined. Results: Risk factors for thrombocytopenia were higher total dose of G-CSF and older age. Risk factors for prolongation of hospitalized period were higher total dose of G-CSF, any present illness, older age and low body-weight. The risk factor for bone pain was higher body-weight. The risk factors for fatigue were female and lower body-weight. The risk factors for insomnia were older age and female and the risk factors for anorexia were female and low body-weight. Predictive factors for lower CD34 + cell mobilization/donor's body-weight were older age and higher total G-CSF administration, and predictive factors for higher CD34 + cell mobilization were male and younger age. Discussion: It was revealed that certain adverse events which occurred within 30 days of post-donation and CD34 + cell numbers to be mobilized could be predicted by utilizing the basic information of donors and of PBSC mobilization/harvest process. To predict them, to notify them to donors and to prepare for possible events will contribute to keep PBSC donor's safety and trust. Background and aim: Allogeneic hematopoietic stem cell transplantation (HSCT) has become an established therapy and the numbers of such procedures increase year by year. The risk for while 17/66 pts developed Chronic-GVHD (extensive: 11 pts). In the multivariate analysis only acute-GVHD grade II-IV and extensive chronic-GVHD were associated with a lower OS. 3 pts developed oral squamous cell carcinoma (SCC) between 3-5 ys after BMT (all had C-GVHD) and 1pt died. 10pts died between 12-2034 days after BMT. Major causes of death were related to rejection or GVHD. TRM at 100 days: 4% and at 1 y: 8%. Conclusions: This regimen was well tolerated and had an excellent survival especially for pts below the age of 10. Long term follow up is still needed to determine the incidence of cancer in this group of pts.

Survey of outcomes of unrelated cord blood transplant in patients with haemoglobinopathies: a retrospective study on behalf of CIBMTR, NYCB and EUROCORD A. Ruggeri (1) Allogeneic hematopoietic cell transplantation (HCT) from HLA-identical donors is curative in patients with hemoglobinopathies. In order to extend the availability of HCT, unrelated cord blood transplantation (UCBT) has been investigated as an alternative. Between 1996 and 2009, 51 patients receiving a single UCBT for hemoglobinopathy were reported to the Eurocord and CIBMTR registries. Thirty-four had thalassemia major (Thal) and 17 had sickle cell disease (SCD). All Thal patients were transfusion-dependent with a median time from diagnosis to UCBT of 26 months and 12 of 17 patients with SCD had a history of stroke. Median age at UCBT was 5 years and median follow-up was 2 years. Cord units were matched at 6 of 6 (15%), 5 of 6 (34%) and 4 of 6 (51%) HLA loci (antigen-level for class1, allele-level for class 2). Median infused cell dose was 4.6 × 10 7 /kg (1.5-13). Thirty-nine patients received a myeloablative conditioning regimen with busulfan (BU) and cyclophosphamide (n = 33) or BU with other agents (n = 6). Reduced intensity conditioning regimens (RIC) (n = 12) were fl udarabine-based with BU < 8 mg/kg or melphalan < 150 mg/m². Cumulative incidence (CI) of neutrophil recovery at 60d was 72±6%, with 35 of 51 patients reaching neutrophil recovery at a median time of 22 days. Higher cell dose (>4.6 × 10 7 /kg) (80% vs. 54%, p = 0.004) and UCBT performed after 2004 (75% vs. 61%, p = 0.001) were associated with improved neutrophil recovery. Day-100 chimerism analysis was available for 47 patients: 19 patients had complete donor chimerism, 4 mixed chimerism and 24 autologous recovery. Among 16 patients who did not achieve neutrophil recovery, 5 had available data on subsequent treatment. Two had an autologous back-up and one received a second UCBT which engrafted. CI of grade II-IV acute-graft versus-host disease (GVHD) was 23% and 9 patients had chronic GVHD. The 2-year probability of overall survival was 77%. Thirty-eight patients (22 Thal, 16 SCD) are alive, 16 with full donor chimerism (8 Thal, 8 SCD). Of the 13 deaths, 5 occurred prior to day-100, mainly due to infections. No deaths due to GVHD were reported. Ten of 12 patients who received RIC are alive, 4 with donor engraftment. Despite the small number of subjects, these results show a particularly high risk of graft failure after UCBT for hemoglobinopathy. TNC dose plays a key role in engraftment. Novel approaches modifying the conditioning regimen and/or increasing cell dose in prospective clinical trials are needed.

Graft rejection and second haematopoietic cell transplantation in patients with Thalassaemia major S. Santarone, E. Di Bartolomeo, P. Bavaro, P. Di Carlo, P. Olioso, G. Papalinetti, P. Di Bartolomeo BMT Center (Pescara, IT) Graft rejection for patients with thalassemia major (TM) receiving haematopoietic cell transplantation (HCT) includes early graft failure (GF) (no evidence of engraftment), late GF (lost of engraftment), and recurrence of TM. We examined our series of 126 consecutive patients aged 1 to 29 years (median 10 years) with TM who were transplanted either from HLA-identical related (n = 123) or unrelated (n = 3) donor. Conditioning regimen consisted of busulfan (BU) and cyclophosphamide (CY). Graftversus-host disease (GvHD) prophylaxis included cyclosporine (CSA) and methotrexate (MTX). All patients received bone marrow (BM) cells. The 10-years overall cumulative incidence of graft rejection was 9% + 0.07. The impact of pre-transplant patient risk factors (age, gender, number of transfusions, ferritin, splenectomy, AST, ALT, hepatomegaly, HBV and HCV infection, liver fi brosis, CMV serology), donor characteristics (age, heterozygous state) and HCT regimens (dose of BU, GvHD prophylaxis, BM dose) was studied in univariate analysis. No factor was univariately associated with signifi cantly increased probability of graft rejection. Eleven events occurred: 4 early GF, 2 late GF, and 5 recurrences of TM. Patients with either early or late GF underwent an urgent second HCT, whereas 3 of 5 patients with recurrence of TM choose to be submitted to second HCT. At all, 9 patients received a second HCT from the same donor. Their median age was 14 years (4-19). Conditioning regimen for second HCT was immunosuppressive for 5 patients (ATG in addition to fl udarabine (FLU) or CY) and myeloablative for 4 patients (BUCY in 1 and thiotepa (TH), treosulphan (TREO) and FLU in 3). GvHD prophylaxis consisted of CSA alone for 5 patients and CSA /MTX for 4 patients. Six patients were given BM cells and 3 received peripheral blood stem cells. Results: 2 patients died from HCT related causes (heart failure at day 29 and acute GvHD at day 83, respectively). Seven patients (78%) are living. One patient had no sign of engraftment and is now living with recurrence of TM following a third HCT with double unrelated cord blood cells. TM recurred in an other patient at 3 months post-second HCT. This patient is now living with transfusion therapy. Five patients (56%), including the 3 patients transplanted with TH-TREO-FLU regimen, are cured with a median follow up of 3 years (0,2-21). This study shows that the incidence of graft rejection following HCT for patients with TM is low and second transplant is successful in an high proportion of patients. Treosulfan is a bi-functional alkylating agent which causes less VOD than busulphan and does not require anticonvulsant prophylaxis and drug monitoring. We report the use of treosulfan in 70 children undergoing HCT for primary immunodeficiency (PID) at two supraregional transplant centres in the UK. Children received 42 g/m² or 36 g/m² treosulfan in combination with either cyclophosphamide 200 mg/kg (30) or fl udarabine Autoimmunity is a complication after haematopoietic stem cell transplantation in patients with primary immunodefi ciencies (PID) that sometime occurs when BMT is performed from alternative donors with extensive stem cell manipulation. The fact that only purifi ed stem cells (CD34 positive cells) are injected could lead to a slower and incomplete post-transplant immunological reconstitution. Since the majority of children treated in our unit received haploidentical CD34 positively selected grafts, we retrospectively analysed the population of PID patients grafted in our institution between 1990 and 2008. Of the 114 affected by severe combined immunodefi ciency (SCID), 70 received an haploidentical transplant, while the others received a matched unrelated or a HLA-identical transplant. 94 patients were affected by inborn errors; 15 received an haploidentical transplant, the others received a matched unrelated or a HLA-identical transplant.We analysed the incidence of autoimmunity in these two groups of patients. Out of 184 children, 17 experienced autoimmune symptoms after HSCT. Among this group, 11 had a diagnosis of SCID, 2 of leaky SCID, 2 of Omenn's syndrome, one of Chediak-Higashi and one of Wiskott-Aldrich syndrome. Autoimmune symptoms included autoimmune haemolytic anaemia (8), autoimmune hypotiroidism (6 patients), cutaneous autoimmunity (2), vasculitis (1) and chronic bronchoreactivity. All patients except for 2 received a myeloablative conditioning therapy. 10 of them received an haploidentical BMT and 7 a MUD or a HLA-identical BMT. No correlation was found analysing onset of autoimmunity and split chimerism in all groups of patients nor we could fi nd a correlation with immunsuppressive therapy. In conclusion we found a low incidence of autoimmunity despite the fact that the majority of our patients received a CD34 positive selected graft with an even distribution of the complication independently from the type of donor or stem cell manipulation procedure.

Introduction: There is still a discussion about the best BM harvesting method to obtain enough progenitor cells during the bone marrow (BM) harvesting procedure. The technique used in most BMT centers is the multiple aspiration of a small (2 ml) amount of BM from the iliac crest, sternum or tibia. This is proposed to minimize the dilution with peripheral blood. Others are using few large amount aspirations. This procedure was not evaluated in children so far. To fi nd out possible differences in graft composition between this two methods and to evaluate the feasibility in children we initiated the following prospective study. Patients and methods: 20 BM harvestings were done in 18 patients (median age 8,78, (2,48 -16,6 y), F 5, M 13). All patients were transplanted in our BM transplantation unit. There were 7 BM harvests for autologous BMT and 13 for allogeneic BMT. The autologous donors were median age 6,93 (2,5-17 y), F 1, M 5), the allogeneic donors were median age 19,75 (6,45-50 y), F 8, M 5). The amount of harvested BM was median 900 ml (440-2380 ml) and median 37 ml/kg bw of the recipient (20-55 ml/kg). The method A was defi ned as an aspiration of 2 ml bone marrow at most before the position of the needle was changed. The method B was defi ned as an aspiration of 100 ml at least before the location of the needle was changed. The BM was collected and analyzed among others for MNC, CD 34 positive progenitor cells, CFU, and T-cells. Two operators carried out the BM harvest. One began with method A and changed in the second part to method B. The second operateur carried out the procedure in the opposite way. Both operators collected the same amounts for each harvesting method so long that the amounts of BM harvested in one method was identically with the second. Results: We found no statistically signifi cant difference between method A and B regarding the content of MNC, CD3 + T-cells, and CFU (MNC/ml 824572 A versus 725000 B Wilcoxon Test p = 0.7; MNC/kg 3.1 10E07 A versus 2.9 10E07, p = 0.3; CD3/ml 162500 A versus 300000, p = 0.3; CFU/10E05 MNC 1678 A versus 1315 B, p = 0.09), but a clinically not relevant but statistically signifi cant difference between CD34 positive cells (CD34/kg 2.62 A versus 2.09 B, p = 0.045). All transplanted patients showed a regular engraftment. Conclusion: BM harvest by large amount few punctures method (B) is as suffi cient as the common used small amount frequent punctures method (A), and could be therefore used equally. Haploidentical transplantation, with extensive T cell depletion to prevent GvHD, is associated with a high incidence of infectionrelated deaths. The key challenge is to improve immune recovery with allogeneic donor T cells without triggering GvHD. As T regulatory cells (Tregs) controlled GvHD in preclinical studies, Background: Natural killer T (NKT) cells represent a small but signifi cant lymphocyte population which have been demonstrated to play a regulatory role in autoimmune disease as well as in GVHD/GVL effects. Methods: The number of CD3 + CD56 + NKT cells in the graft was retrospectively correlated with clinical outcome of fi fty-eight patients receiving a fi rst allogeneic peripheral blood stem cell transplant (myeloablative conditioning, n = 31; reduced-intensity conditioning, n = 27) from their HLA-identical sibling donors between Dec 2004 and Jul 2009. Results: A median number of 0.15 (range, 0.01-0.57) × 10 8 /kg CD3 + CD56 + NKT cells was transplanted with the graft. NKT cells within the graft signifi cantly correlated with the number of CD3 + T cells, CD56 + NK cells, and MNC (p<0.05). Overall survival for the entire cohort at two years was 55% (41%-69%, 95% CI) with no difference between pts. receiving a high (>0.15 × 10 8 /kg) or low (<0.15 × 10 8 /kg) NKT cells. The cumulative relapse incidence for the entire cohort was 40%, with a trend for a lower relapse incidence in pts. receiving a low vs. high NKT cell dose (32% vs. 47%). The nonrelapse mortality was 19% showing a trend for a lower NRM in patients receiving a high NKT cell dose (12% vs. 27%). The cumulative incidence of aGVHD II-IV was 38% for the entire cohort. The incidence of aGVHD II-IV was lower in pts. receiving a high NKT cell dose (30% vs. 45%). The cumulative incidence of cGVHD in pts. at risk was 48%, with no difference between pts. receiving low or high NKT cell numbers. By multivariate Cox analysis including CD3, NK, NKT, TREG, CD34 cell number, recipient age, and time interval between diagnosis and SCT as numerical variables, and risk category by the underlying disease (standard risk vs. high risk), sex mismatch (male recipient/female donor vs. other combinations), and myeloablative vs. reducedintensity conditioning as categorical variables the most powerful predictive parameters for survival were standard vs. high-risk disease (p = 0.01), a high NK (p = 0.02) and a low NKT cell dose (p = 0.05). The most powerful predictive parameters for relapse were high risk disease (p = 0.01), a low NK (p = 0.02) and a high NKT cell dose (p = 0.08). Discussion: These data indicate that in the setting of peripheral blood stem cell transplantation from HLA-identical sibling donors a higher number of NKT cells in the graft might increase the risk of relapse by lowering the incidence of acute GVHD. Objectives: Adoptive transfer of NK cells with or without previous allogeneic progenitor cell transplantation may represent a promising therapeutic option in patients with hematological or oncological diseases. Different NK cell isolation strategies have been pursued. Here, two different methods were compared with respect to recovery, immunophenotype and cytotoxic capacity of purifi ed NK cells. Methods: NK cells of healthy donors were isolated from a LRS chamber of a platelet apheresis. Cells were enriched by (I) CD56 positive selection of T cell (CD3) depleted lymphocyte fractions (method I, MI) or (II) depletion of non-NK cells (untouched protocol, method II, MII) by magnet-activated cell sorting (MACS). Recovery of NK cells obtained by M I or II were compared. NK cells derived by MI were activated by stimulation with interleukin (IL)-2 (100U/ml) or IL-15 (10ng/ml) overnight and analyzed for cytotoxic activity. For expansion of NK cells CD2/CD335 antibody coated MACSiBeads (Miltenyi) were used as artifi cial stimulators in IL-2 supplemented media (500U/ml) in a two weeks culture. Finally, immunophenotype of effector cells and killing of target cells were assessed. Results: Median NK cell recovery was 35.9% (n = 4) and independent of strategy of enrichment (MI versus MII). Overnight activation of NK cells led to a 2-fold enhanced killing of K562 cells. NK cell expansion led to a 100-fold increase in numbers of functional active NK cells. Expanded NK cells showed characteristics of activated NK cells as indicated by the induced expression of NKp44 (CD336). Surface expression of TAC-TILE (CD96), a marker for certain activated immune cells, was found with CD96 antibody TH-111 elevated on bead-expanded NK cells in comparison to NK cells cultured only in IL-2 conditioned media, which may enhance reactivity of expanded NK cells against the ligand of CD96, namely polio virus receptor expressing tumours. Conclusion: T cell depletion and CD56 positive selection led to an acceptable recovery of enriched NK cells that could be successfully activated by IL-2 or IL-15. The described NK cell isolation and activation protocol can easily been transferred into a GMP laboratory. For clinical application, further expansion of isolated NK cells under GMP conditions may be preferable. NK cells enriched, expanded and activated by these means may be an adoptive cellular therapeutic option that in addition may be improved by combination with tumour antigen specifi c antibodies.

S76 O386 A randomized trial comparing CD34 enriched grafts versus CD3/CD19 depleted grafts in partial T-cell depleted allogeneic stem cell transplantation N. Schaap, D. Eissens, F. Preijers, A. van der Meer, A. Schattenberg, H. Dolstra, W. van der Velden, T. de Witte, M. Smits, N.M.A. Blijlevens Radboud University Medical Centre (Nijmegen, NL) We initiated a randomized trial in AML, ALL and MDS patients (pts) treated with partial T cell depleted allogeneic SCT using immunomagnetic CD34 selection versus CD3/CD19 depletion. A benefi cial effect was hypothesized on long term engraftment and DFS using CD3/CD19 depletion. Residual cell populations in the graft after CD3/CD19 depletion may play a protective role during early phase after SCT and generate an increased GVL effect. From May 2006 until May 2009 pts with AML, ALL and MDS were stratifi ed for diagnosis and randomized. Median age of the pts was 46 and 45 yrs, respectively. Stem cells were obtained after G-CSF (10 ug/kg). T cell add back to a fi xed dose of 5 × 10e5 CD3 + T cells/kg BW was standard in all pts. Pts without significant (> grade 2) acute and/or chronic GVHD were eligible for prophylactic DLI. In this interim analysis we determined immune reconstitution of T cells, B cells and NK cells using fl ow-cytometry and functional essays and evaluated clinical outcome. Reconstitution: NK cells are completely maintained in CD3/ CD19 depleted transplants. B cell depletion is equally effective. CD34 + cells are signifi cantly higher in the CD3/CD19 depleted grafts (median 4.3 versus 6.5 × 10e6/kg, p<0.05). All pts engrafted and the time to engraftment was not different. In the fi rst 3 months after SCT, the immune reconstitution of lymphocytes, especially CD4 T cells, NK and NK-T cells was faster in the CD3/CD19 group. Fourteen days after SCT the cytotoxic NK and regulatory NK subsets were already present in the CD3/CD19 group whereas the CD34 group needed 2 months to full recovery. The cytolytic response against a leukemia target was stronger in the CD3/CD19 group (p = 0.02). NK receptor expression of NKG2A (inhibitory) was strongly decreased whereas expression of NKG2C (activating) was enhanced. Clinical outcome; Acute GVHD > grade 2 was 17% in both groups. Extensive chronic GVHD was 13% (CD34) vs. 18% (CD3/CD19. One year TRM, RR, LFS and survival using CD34 selection and CD3.CD19 depletion were 5% vs. 39% (p = 0.066), 36% vs. 39% (p = 0.385), 64% vs. 35% (p = 0.063), 74 vs. 41% (p = 0.016), respectively. Conclusion: Compared to CD34 selection, CD3/CD19 depletion resulted in a faster reconstitution of CD4 cells,NK-T and NK cells. However survival was signifi cantly better in pts transplanted with CD34 selected grafts. Due to a loss of power in favor of our hypothesis (15% increase in overall survival) using statistical simulation models, this trial was stopped. (1) Introduction: In haploidentical Hematopoietic Stem Cell Transplantation (HSCT), the infusion of donor lymphocytes transduced to express the Herpes Simplex Virus Thymidine kinase (HSV-Tk) suicide gene allows to control GvHD, to mediate GvL, and to rapidly provide an effective and polyclonal anti-infective T cell repertoire (Ciceri and Bonini et al., Lancet Oncology, 2009 ). Even though their engraftment is necessary to achieve these effects, HSV-Tk + cells represent the minority of lymphocytes circulating in treated patients. Therefore, we investigated the putative role of HSV-Tk + cells in promoting thymic activity and T cell development from graft progenitors. Methods: Thymic function was assessed in adult patient who underwent haploidentical HSCT and infusion of suicide genemodifi ed donor T cells for hematologic malignancies, after validating the methods in healthy pediatric and adult controls. Single joint T cell Receptor Excision Circles (sjTREC) were quantifi ed by qPCR in peripheral blood mononuclear cells (PBMCs) and purifi ed T cells, and the proportion of CD31 + recent thymic emigrants (RTEs) in CD4 + naïve T cells was measured with immunophenotype analysis in PBMCs. Thymic output was correlated with thymic volume, assessed by Computed Tomography (CT) scans. T Cell Receptor repertoire was assessed by Vbeta spectratyping. The relative contribution of HSV-Tk + and HSV-Tk-donor T cells to post-transplantation anti-host alloreactivity was studied by mixed lymphocyte cultures. Results: At the moment of T cell immune reconstitution (defi ned as CD3 + cells > 100/μl peripheral blood), the CD4 + naïve T cell subset was almost entirely comprised by CD31 + RTEs, and this percentage remained high, as compared to age, also in subsequent months. In informative patients, RTE frequency before HSCT and before HSV-Tk + cell infusion was in line with agerelated healthy controls, suggesting a direct role of the infused cells in mediating the effect. Accordingly, in treated patients CT scans documented an increase in thymic volume following HSV-Tk + cell add-backs. Finally, low absolute sjTREC counts could be detected, in line with the documentation of extensive peripheral proliferation and low RTE absolute numbers. Conclusions: These data show that after the infusion of suicide gene-modifi ed T cells a renewal of thymic activity takes place, which contributes to the recovery of the peripheral T cell repertoire.

Elevated pretransplant serum ferritin levels have been associated with an increased susceptibility for opportunistic infections and increased incidence of morbidity and mortality after allogeneic HCT. We studied in 81 patients who underwent myeloablative allogeneic HCT for acute myeloid leukemia pre-and posttransplant serum ferritin levels and correlated the serum ferritin levels with the TLR9 expression and the cellular immune reconstitution 3 and 12 months post transplant. Further, we studied in vitro-experiments in Kasumi 1 cells the TLR1, TLR2, TLR3, TLR5, TLR7, TLR9 and TLR10 expression after overwhelming iron exposure. The average pretransplant serum ferritin level was 1245 μg/ml (mean) in all AML-patients (mean 1100 μg/l for patients with AML in 1.CR and mean 1820μg/l for patients with AML > 1.CR). Post transplant serum ferritin level increased up to 2080 μg/ml (mean) for all AML patients (mean 1290 mg/l for AML in 1.CR and mean 2350 μg/l for patients with AML > 1.CR). The application of 300 ng iron to acute leukaemia cell lines SD1, and Kasumi-1 cells increased signifi cantly TLR1,TLR2, TLR3, TLR5, TLR7 and TLR9 expression in relation to the housekeeping gene abl measured by real-time RT-PCR. In Kasumi-1 cells TLR1 increased up to 50,6% (p = 0.014) TLR2 35.5% (p = 0.046), TLR3 57,8% (p = 0.006), TLR5 62.9% (p = 0.005), TLR7 46.2% (p = 0.02), TLR9 44.2% (p = 0.026) and TLR10 54,7% (p = 0.07) compared to untreated Kasumi 1 cells. Further, patients with elevated post transplant ferritin level > 2000 μg/l had an increased TLR9 expression in mononuclear cells (TLR9/ABL quotient 6485 versus 4543; p<0.05) 3 months post transplant. The number of cytotoxic T cells CD4 + CD8 + in patients with elevated serum ferritin level was signifi cantly elevated after transplant (mean 189 versus 95 cells/μl 12 months post transplant, p = 0.034), whereas no differences were found in the number of CD3 + CD4 + T helper cells, B19 + cells, Nk cells.

These results indicate that elevated ferritin levels might activate the innate immune system by increasing TLR expression. This might be of importance since we recently showed that increased TLR9 expression was associated with adverse impact on nonrelapse mortality in the transplant setting. Further exaggerated TLR9 expression has been discussed to induce overwhelming immune responses as SIRS or ARDS. More studies are definitely necessary to evaluate the role of elevated iron overload on the innate immune system.

Clinical scale generation of functional human natural killer cells from umbilical cord blood CD34-positive cells for immunotherapy J. Spanholtz, M. Tordoir, C. Trilsbeek, J. Paardekooper, T. de Witte, N. Schaap, F. Preijers, H. Dolstra UMC St.Radboud (Nijmegen, NL) Immunotherapy based on natural killer (NK) cell infusions is a potential adjuvant treatment for many cancers. We established an extremely effi cient cytokine-based culture system for ex vivo expansion of NK cells from hematopoietic stem and progenitor cells from umbilical cord blood (UCB). Systematic refi nement of this two-step system using a novel clinical grade medium resulted in a therapeutically applicable cell culture protocol. The use of GBGM culture media in a clinical applicable protocol allows the ex vivo expansion and differentiation of CD34 + cells to more than 4-logs into CD56 + CD3-NK cells with very high purity. These ex vivo-generated CD56 + cell products contain NK cell subsets expressing CD94/NKG2A and KIR as well express high levels of activating NKG2D and NCRs. Functional analysis showed that CD56 + cell products containing alloreactive NK cells effi ciently kill myeloid leukemia and melanoma tumor cells as well as primary acute myeloid leukemia (AML) cells. We have currently translated the protocol to clinical scale production using a closed cell culture bioprocess ( Figure 1 ). CD34 + selection from cryopreserved UCB samples (n = 9) using the CliniMACS device was optimized and the selection resulted in a CD34 + enrichment with a mean number of 2,2 ± 1,6 × 10 6 cells and a purity of 71 ± 14 % CD34 + cells. Validation experiments using these cells showed, that the generation of suffi cient numbers of NK cells without contaminating T-cells or Bcells under a closed system process is feasible. The cell cultures using bags show a mean expansion of 1273 ± 506 fold (range 759-1770 fold, which generated 8,6 × 10 8 -1,9 × 10 9 NK cells from cord blood-derived CD34 + cells within maximal 6 weeks of culture. The mean purity of the NK cell product was 71 ± 9 % (range 63-80%) of the total cells in the bag cultures. In order to improve the purity of the product, we include bioreactors during the differentiation culture process. Using this modifi cation we were able to synchronize the differentiation process in small (plate) and large scale (bag) experiments, which allows the generation of large scale NK scale products with a purity of more than 95% CD56 + cells devoid of T-and B cells (Figure 2A and B) . These NK cell products will be used for immunotherapy in elderly AML patients. This study is a phase I dose escalation study in a series of 12 AML patients who have successfully achieved CR (<5% blasts in the bone marrow) after standard intensive chemotherapy.