key: cord-0005620-fs6vjplh
authors: nan
title: Physicians Oral Abstracts: EBMT 2012
date: 2012-04-04
journal: Bone Marrow Transplant
DOI: 10.1038/bmt.2012.36
sha: e7484b7a686329ecd72031d10422bd9e1b9b494e
doc_id: 5620
cord_uid: fs6vjplh

nan

T cell receptor (TCR) gene-transfer in T lymphocytes is a promising tool for adoptive immunotherapy of cancer patients for whom natural tumor-specifi c lymphocytes cannot be isolated. Nevertheless, TCR-transferred T lymphocytes differ from their natural counterpart in carrying two different TCRs and this limits their effi cacy and alters their toxicity profi le due to reduced expression of tumor-specifi c TCR and inappropriate pairing of TCR chains. To overcome these issues, we developed a novel approach aimed at the complete genetic editing of T cell specifi city, based on the ZFN-mediated disruption of the endogenous TCR chain genes coupled to the transfer of genes encoding for a tumor-specifi c TCR. We selected Wilms' tumor antigen 1 (WT1) as a model antigen. For the complete editing of T cell specifi city, we established a protocol to sequentially disrupt the endogenous TCR alpha and beta chain genes with high effi ciency (averages: 34%±11 and 16%±10), and to permanently transfer WT1-specifi c TCR alpha and beta chain genes by lentiviral vectors (average effi ciencies: 53%±25 and 21%±15). This procedure resulted in a population of TCRedited lymphocytes encoding only the tumor-specifi c TCR that, in the absence of competition from the endogenous receptor, was expressed at high levels. Accordingly, TCR-edited lymphocytes were superior to conventional TCR-transferred cells in promoting specifi c recognition of WT1-expressing targets, including primary leukemias, and most importantly, were devoid of residual endogenous TCR reactivity including alloreactivity. Finally, for a comprehensive assessment of safety and antitumor effi cacy, we treated immunodefi cient mice, infused with primary human leukemias, with matched TCR-transferred, TCRedited, and unmanipulated cells and we monitored mice for Graft-versus-host disease (GvHD) and leukemia appearance. Mice treated with TCR-edited cells showed higher event free survival than TCR-transferred (p<0.05) or unmanipulated lymphocytes (p<0.01). These data demonstrate that the successful genetic re-programming of T cell specifi city in primary lymphocytes results in a functionally superior target specifi c killing activity and thus has the potential to greatly improve the safety and therapeutic activity of cancer immunotherapy. (Provasi and Genovese: equal contribution). Objectives: Self-renewal is a key property of hematopoietic stem cells (HSCs) and the hallmark of stem cells in general. However, despite decades of research, it is still unknown which extrinsic factors are necessary to optimize HSCs self-renewal ex vivo. The present experiments were designed to identify potential stromal-derived factors that promote mouse HSC selfrenewal ex vivo. Methods: Highly purifi ed HSCs isolated from mouse adult bone marrow (i.e. EPCR+, CD150+, CD48-, CD45+ cells, 42% having longterm repopulating ability) were cultured for 7 days in vitro in serum-free medium (SFM) and various combinations of cytokines, stromal cells and factors they produce. HSCs numbers were determined by performing limiting dilution transplants in sub-lethally irradiated congenic W41/W41 mice assessed for repopulation activity of the transplanted cells 4-6 months post-transplant. Culture conditions that supported the greatest expansion of HSCs were used to assess the HSC output of single input cells and to design comparative gene expression analyses on 6-hour-stimulated cells to identify potential stromal factors involved. These were then tested for their potential to replicate the effects of stroma cells and stromal cell conditioned medium. Results: Of the various conditions initially tested, the addition of stromal cells or stromal cell-conditioned medium (CM) to Steel factor and IL-11 gave maximal and equivalent HSC outputs after 7 days. (5-to 11-fold expansion of transplantable HSC numbers). This indicated a stromal cell contact-independent mechanism. Visual tracking of single-cell cultures revealed that stromal cell-derived factors prevented the apoptosis of ~50% of the input cells and transplant assays of the CM-containg cultures showed that 90% of the cells that produced at least one daughter HSC did so asymmetrically and only 10% did so symmetrically. Gene expression analysis showed that pathways activated by collagen 1 (Col1) and nerve growth factor (NGF) were signifi cantly upregulated during the self-renewing process and a direct test of these two factors combined indicated they could replace the activity of CM in promoting HSC survival and expansion. Conclusion: NGF and Col1 are key stromal-derived factors that can positively regulate adult HSC expansion when these are stimulated with proliferative cytokines.

Introduction: In April 2006, the Spanish Myeloma Group (PETHEMA/GEM) activated a randomized phase III trial comparing induction with TD vs. VTD vs. VBMCP/VBAD/Bortezomib in patients 65 years-old or younger with newly diagnosed symptomatic MM and ASCT with MEL-200 followed by maintenance with thalidomide/bortezomib (TV) vs. thalidomide (T) vs. alfa-2b-interferon (IFN). Primary end points: The primary end-point was time to progression from the initiation of maintenance therapy. Patients and Methods: The maintenance program consisted of TV (thalidomide 100 mg daily plus one cycle of bortezomib-1.3 mg/m 2 on days 1, 4, 8 and 11 every 3 months) versus T (single agent thalidomide at a dose of 100 mg daily) versus IFN (subcutaneous IFN at a dose of 3 MU three times per week). The planned maintenance duration was three years or until disease progression or toxicity. From February 1, 2007 to January 27, 2011 266 patients were randomized to maintenance therapy (TV:90; T: 89, IFN: 87). Response and survival were evaluated on an intention-to-treat basis. Responses and progressions reported by the investigators were centrally reassessed. Results: The patient's characteristics at diagnosis such as age, ISS stage, cytogenetics and presence of extramedullary plasmacytomas as well as induction regimen and diagnosis-randomization interval were similarly distributed among the 3 arms. The response status at the time of randomization after ASCT was CR: 51%, VGPR: 23%, PR: 24% and SD: 2% and was well balanced in the three groups. The CR rate with maintenance was improved by 23% with TV, 11% with T and 19% with IFN (p=NS). After a median follow-up of 24 months, the PFS was signifi cantly longer with TV compared with T and IFN (PFS at 2 yrs: 78% vs. 63% vs. 49%, p=0.01). OS was not signifi cantly different among the 3 arms. Grade 3 and 4 hematological toxicity was similar (22.2% vs. 16% vs. 21.8%). No peripheral neuropathy (PN) was observed with IFN being its frequency similar with TV (12.2%) and T (10.1%). No grade IV PN was observed. Dose reductions for TV, T and IFN were required in 33.3%, 33.7% and 19.5% of the patients, respectively. The discontinuation rate due to toxicity was signifi cantly higher with thalidomide compared with TV (30.3% vs. 15.6%, p=0.08) and with IFN (30.3% vs. 18.3%, p=0.17). Conclusion: The addition of bortezomib to thalidomide maintenance resulted in a signifi cantly longer PFS when compared with thalidomide alone or with IFN with no increased toxicity.

In allogeneic hematopoietic stem cell transplantation (allo-HSCT) patient and donor are usually matched for the HLA class I molecules A/B/C as well as for the HLA class II molecules DRB1 and DQB1. In contrast, the HLA-DPB1 locus is still ignored in donor selection. Clinical studies have demonstrated that disparities at HLA-DQB1 and distinct HLA-DPB1 alleles do not adversely affect the outcome of allo-HSCT. It has also been shown that HLA class II is predominantly expressed on hematopoietic cells under non-infl ammatory conditions. Thus CD4+ donor T cells recognizing patient-derived HLA-DQB1 or permissive HLA-DPB1 mismatch alleles may primarily target leukemic and hematopoietic cells, while sparing non-hematopoietic tissues. We used PBMC of healthy donors to generate mature monocyte-derived dendritic cells (DC), which underwent transfection with in vitro transcribed RNA coding for single HLA-DQ/-DP mismatch alleles by electroporation. These allo-HLA expressing DC were used to stimulate autologous naive CD4+ CD45RA+ T cells in mixed lymphocyte reactions (MLR) in vitro. Rapidly expanding MLR cells showed specifi c recognition of allo-HLA-DQ/-DP molecules as demonstrated by lack of immune reactivity to non-transfected DC as well as by complete inhibition of alloreactivity using HLA allele-specifi c antibodies. The allo-HLA-DQ/-DP specifi c T cells were also analyzed for reactivity to a broad panel of primary acute myeloid leukemia (AML) blasts. Strong IFN-g secretion could be observed only for those AML targets that carried the HLA allele used for T cell priming, demonstrating the specifi city of this approach. We further investigated HLA class II expression on hematopoietic and non-hematopoietic cells by fl ow cytometry. HLA class II was not detected on primary fi broblasts, keratinocytes, and normal kidney cells (each n=10), but was expressed at signifi cant levels on primary AML blasts and B-cell lines (each n=10). Expression levels followed the hierarchy: DP>DR>DQ. Up-regulation of HLA class II expression was observed on all cell types after pre-incubation with IFN-g, but not after addition of TNF-a, IL-1b and IL-6. Our approach appears suitable for generating allo-HLA-DQ/-DP specifi c CD4+ T cell lines that recognize leukemia cells while presumably sparing non-hematopoietic cells under non-infl ammatory conditions. It may be of potential use in adoptive immunotherapy of allo-HSCT patients who express single HLA-DQ or permissive HLA-DP mismatch alleles. The ratio of the FLT3-ITD mutation to the wt-FLT3 allele has signifi cant prognostic importance of FLT3-ITD mutations in AML (Thiede C et al., Blood 2002) . There is still uncertainty about the role of allogeneic transplantation (allo-SCT) in the treatment of patients with FLT-ITD mutation. In order to proof the presence of an allogeneic effect we compared the survival after allo-SCT in fi rst remission in patients with a mutant rate <0.8 with those ≥0.8. For comparison, the results in patients cohorts treated with chemotherapy alone with the same mutations rates were analyzed. Patients and Methods: Patients diagnosed with AML, aged 18-60 years, and treated in the AML 2003 trial of the SAL were analyzed. According to the risk-adapted treatment strategy of the trial, cytogenetically intermediate-risk (IR) and adverserisk (AR) patients should receive an allo SCT as consolidation treatment if a HLA-matched-sibling donor (IR) or HLA-matched related or unrelated donor (AR) was available. Patients with no available donor received high-dose cytarabine based consolidation or autologous SCT. Survival analyses were performed by using the Kaplan-Meier method including log-rank tests for signifi cance testing. Results: Of 1182 patients enrolled in the trial, 257 were FLT3-ITD+ (22%). The ratio of the FLT3-ITD mutation to the wt-FLT3 allele was < 0.8 in 182 patients and ≥ 0.8 in 75 patients. 47 (26%) of the low mutation rate group and 30 (40%) of the high mutation rate group received an allogeneic transplantation. In the cohorts having received an allogeneic transplantation, the 3-year disease-free survival (DFS) in the mutant rate < 0.8 and ≥ 0.8 groups was 58% and 50%, respectively (p=0.53). The 3-year overall survival (OS) was 61% and 59%, respectively (p=0.47). In the cohorts having chemotherapy as consolidation, the 3-year DFS in the mutant rate < 0.8 and ≥0.8 groups was 36% and 10%, respectively (p<0.001). The 3-year OS was 43% and 11%, respectively (p<0.001). Overall, allo-SCT lead to an overall and event-free survival in patients with a high FLT3-ITD mutation rate comparable with those with a low mutation rate. Without allogeneic transplantation and chemotherapy alone, again signifi cant differences in DFS and OS between patients with a high vs. low mutation rate as in our report of the previous AML96 could be shown. These data point toward a strong allogeneic effect after transplantation thus eliminating the negatve impact of a high mutation rate.

The ALWP objectives are: (i) to organize high level accredited educational activities pertinent to acute leukemia (latest symposiums: Nantes in 2008, Barcelona in 2009, Milan in 2010, and Warsaw in 2011); (ii) to design and support prospective clinical trials in the fi eld of acute leukemia across member centres (the pan-european elderly AML randomized trial is currently recruiting patients: ClinicalTrials.gov Identifi er: NCT00766779); (iii) to generate high quality retrospective studies addressing different issues related to acute leukemia management and therapy; (iv) to increase within the EBMT registry the quality of data pertinent to HSCT for acute leukemia; and (v) to generate guidelines pertinent to the management of acute leukemia. Currently, the ALWP activities are organized and structured within 6 subcommittees (SC) focused on specifi c fi elds of interest: autologous HSCT SC, Immunotherapy SC, Alternative donors SC, RIC SC, Molecular markers SC, and the Developing centers SC. The ALWP is currently chaired by M. Mohty (France) and the secretary is S. Gieble (Poland) and includes representatives/ members from most EBMT centres/countries, with expertise in both auto and allogeneic HSCT for AML and ALL. The ALWP meets twice a year to discuss ongoing studies and new study proposals and review manuscript preparation. All EBMT members are encouraged to submit study proposals (registry-based studies) to the ALWP. After a quick feasibility assessment performed by the ALWP offi ce, projects will be discussed during the ALWP winter (usually October or November) and spring (during the EBMT annual meeting) business meetings. Registry studies as well as randomized studies compared the two types of preparative regimens for allo-HSCT in patients with AML with some confl icting results concerning outcome and toxicity. The original goal of the Bu/Cy protocol was to reduce toxicity, improve outcome and provide an alternative for patients that received prior radiation and would thus not be a suitable candidates for TBI. However, the risk of veno-occlusive disease of the liver (VOD) was still a matter of concern. Indeed, VOD was shown to be relatively high in patients that were conditioned with high dose of oral Busulfan. In contrast, IV Bu has a more favorable toxicity profi le as compared to the oral formulation. It is thus conceivable that the risk of hepatic VOD as well as transplant-related mortality (TRM) would be reduced when using the IV formulation of Bu, leading to improved allo-HSCT outcome. In order to address this issue, the ALWP of the EBMT recently performed a series of different retrospective registry-based studies comparing IV Bu/Cy to Cy/TBI for conditioning prior to allo-HSCT for adult patients with AML in complete remission, fi rst relapse and resistant relapse. In all, the IV Bu/Cy conditioning regimen resulted in similar transplantation outcomes including engraftment, non-relapse mortality (NRM), relapse incidence (RI), overall survival (OS) and leukemia-free survival (LFS) in comparison to the TBI/CY preparative regimen. Acute GVHD was signifi cantly lower in AML patients in CR who received IV Bu/Cy vs. TBI/Cy conditioning, respectively. Moreover, in AML patients in fi rst relapse undergoing allo-HSCT, IV Bu/Cy conditioning can result in higher post-transplant remission rates which resulted in better LFS and OS in comparison to TBI/Cy conditioning. This advantage in favor of the IV Bu/Cy regimen is likely due to lower overall toxicity and improved capacity for salvage therapy. In summary, the available research evidence strongly suggests a favorable safety and effi cacy profi le in favor of the IV Bu/Cy regimen. Thus, the future role of TBI in allo-HSCT for AML needs to be revisited in well designed controlled studies. Allo-HSCT in early phase is considered the treatment of choice for younger patients with high risk AML, although this presumed benefi t of allo-HSCT has not been analyzed in molecularly defi ned AML subtypes. In this context, several studies within MMS-ALWP have been conducted to elucidate the role of allo-HSCT for specifi c AML entities. Main results of these studies, performed in patients allografted in CR1, will the summarized: -AML with FLT3-ITD: the effect of FLT3-ITD was analyzed in a cohort of 206 patients with normal cytogenetics AML. FLT3-ITD (n=120) was associated to a higher relapse incidence (RI, 5-yr RI: 30% vs. 16%) and decreased LFS (5-yr LFS: 58% vs. 71%).

-AML with MLL rearrangement (MLL-r, n=138): type of MLL rearrangement had a strong prognostic impact, allowing the distinction between favorable MLL-r entities, such as t(9;11) and t(11;19), and unfavorable MLL-r subtypes, namely t(6;11) and t(10;11). Patients from the favorable subgroup showed a signifi cantly better OS (2-yr OS: 70% vs. 29%) due to a markedly reduced RI (20-25% vs. 46-47%).

-AML with monosomal karyotype (MK-AML, n=189): MK-AML showed an independent impact after alloHSCT, with a 3-yr LFS and RI of 24% and 55%, respectively. Of note, outcome of patients with MRC-defi ned poor cytogenetics after excluding MK-AML and 7q abnormalities did not differ from that of patients with intermediate-risk cytogenetic AML (IR-AML).

-AML with t(6;9) AML: 73 patients allografted in CR1 were analyzed, with a 5-yr OS of 55±6%, similar to IR-AML patients. In summary, the outcome following allo-HSCT of patients with high risk AML varies signifi cantly depending on the specifi c entity considered. Identifi cation of molecularly-defi ned AML subtypes with a high relapse risk after allo-HSCT justifi es the design of strategies for a preemptive intervention after transplantation. In the early days of autologous bone marrow transplantation (ASCT) for acute leukemias, leukemia free survivals over 50% at 3 years have been reported for adult ALL both with marrow purged with cyclophosphamide derivatives and with unpurged marrow. A very interesting fi nding was that engraftment of neutrophils and platelets was rapid in sharp contrast with acute myelocytic leukemias where the kinetics of hematopoietic recovery were considerably delayed. Most teams have long term survivors (more than 10 years follow up), some autografted despite poor risk prognostic factors such as t(4;11) or t (9;22) . Moreover, small series have advocated in favor of the introduction of maintenance chemotherapy following ASCT. However, several prospective studies have failed S6 to demonstrate a clear advantage of ASCT over conventionaldose chemotherapy and the relapse rate following ASCT has been considered to be too high for further prospective studies. More recent analyses demonstrated that results of ASCT may depend strongly on the level of minimal residual disease (MRD) and in the era of routine MRD monitoring the interest in ASCT is re-growing. Here, we will update the current status of ASCT in adult ALL using the EBMT ALWP data base and make proposals for new studies with MRD monitoring.

Despite considerable progress in the treatment of acute myeloid leukemia in the past several decades, the prognosis of the majority of patients with this disease remains guarded. Advances in supportive care and better characterization of disease subsets through cytogenetics and molecular analysis have led to signifi cant success in treating specifi c subsets of patients such as those with acute promyelocytic leukemia (APL) and core binding factor leukemias (CBF), particularly among the younger patients who are able to better tolerate the effects of cytotoxic chemotherapy. However, overall, only about 40% of younger patients and less than 10% of older patients with this disease are alive at 5 years. Current research is focusing on the identifi cation of new cellular targets amenable to specifi c inhibitors, designing the best strategies for combining these novel agents with traditional chemotherapy regimens, and determining prognostic indicators that may allow us to better stratify therapy.

Allogeneic HSCT is generally advocated as the treatment of choice to consolidate remission in intermediate and poor risk AML. The last decade, new cytogenetic and molecular markers have been identifi ed that specifi cally relate to poor or very poor-risk AML. The latter categories include the so-called monosomal karyotype (MK) leukemia's and those with high expression of EVI-1. Recent studies suggested that the benefi cial effect of allo-HSCT also applies to those categories of AML. Despite being more effi cacious than consolidation chemotherapy, the risk of relapse in poor-risk AML is still considerable and may estimate between 30 and 50%. It suggests that, although clearly operational, the GVL effect has not been exploited fully. Further improvement may be pursued by avoiding delay and proceeding to allo-HSCT before full peripheral hematopoietic recovery. In addition, the continuous application of new agents such as 5-azacytidine after allo-HSCT, may offer a new approach to facilitate ongoing GVL. Reduced intensity conditioning regimen have been developed in order to reduce NRM. Several studies have indeed shown a lower NRM after RIC conditioning as compared to MA conditioning, especially as patients assigned to RIC regimens were most times older and had higher comorbidity scores. While several retrospective studies have suggested a somewhat higher relapse rate in recipients of RIC allo-HSCT, some very recent studies suggested that relapse may not differ in intensively pretreated AML CR1 patients, who subsequently proceeded to either RIC allo-HSCT or MA allo-HSCT. Collectively, these studies may suggest that the benefi cial effect of allo-HSCT largely depends on the immunotherapeutic GVL effect and less on the intensity of the conditioning regimen in AML patients, who attained fi rst CR upon intensive chemotherapy.

Background: In the last years allogeneic stem cell transplants after reduced intensity conditioning (RIC) have become increasingly popular. As a result of minimized cytotoxic conditioning and therefore lower early treatment related mortality and morbidity, allogeneic transplantation can now be considered in patients who would otherwise not qualify for this treatment. As the success of the transplantation relies mainly on the graft-versustumour effect, which often goes in parallel with a more or less distinct graft-versus-host reaction, long term complications and problems in various aspects of quality of life (QoL) have to be expected as they are already known from conventional stem cell transplantation HSCT. Aims of the study: The aim of the study is to compare quality of life in both groups (reduced vs. conventional conditioning) in the longitudinal course over three years after transplantation. We want to analyse QoL after non-myeloablative HSCT over the long term outcome. Methods: Based on a prospective longitudinal design we assess QoL as primary endpoint with a validated questionnaire (EORTC QoL-C30, EORTC HDC26). Secondary end points are transplant related mortality, overall survival, event free survival and graft versus host disease. We are measuring QoL before transplantation (T0), 100 days after (T1) and each year over a prospected time frame of 3 year (T2 1 year, T3 2 years and T4 3 years). It is planned to reach a fi nal sample size of 300 patients in each group at T0. The study is carried out as a multicenter study recruiting patients in four centres (France, Germany, Italy and Switzerland). Results: As the study is still ongoing, the focus of this presentation is an interim analysis of the data collected until end of 2011. We present results on QoL and medical data from 218 patients (37.2% standard; 62.8% reduced conditioning), showing that also patients with RIC have substantial problems in various functions and activities of daily living over time. Discussion: Due to high drop out rates further centres will be included to improve patient recruitment and guarantee substantial subgroup samples over the longitudinal course of the study. Objectives: Genital chronic graft-versus-host disease (cGVHD) is a complication of allogeneic hematopoietic stem-cell transplantation (alloHSCT) responsible for signifi cant morbidity and impairment of quality-of-life, but remains under-diagnosed. We conducted a retrospective monocentric study to describe the main characteristics of women diagnosed with genital cGVHD and to evaluate the impact of a systematic gynecological follow-up after alloHSCT on genital cGVHD diagnosis, treatment and evolution. Methods: Women seen at the gynecology consultation of our alloHSCT center, between 2008 and 2010 were included if they had received alloHSCT after year 2000 and were diagnosed with genital lesions of cGVHD. The diagnosis could be clinical or histological in atypical cases. All patients received local Background and Methods: in women, genital chronic graftversus-host disease (gcGvHD) is well described, but only limited data with regard to male recipients after HSCT exist. In a prospective cross-sectional single center study, we aimed to address this issue in post HSCT male recipients performing a whole body skin examination, focusing on genital changes. Furthermore posttransplant sexual contentedness and sexual functioning were assessed by two self-assessment questionnaires: the 5-Item Version of the International Index of Erectile Function (IIEF-5) and the modifi ed Brief Sexual Symptom Checklist (mBSSC).

S8 Results: all 155 asked patients accepted to participate. The characteristics of the study population are listed in Table 1 . The median time between HSCT and genital examination was 50 months (range 21-72). Thirty-one out of 155 patients (19%) showed remarkable genital skin changes, 21/155 patients (13%) had gcGvHD-related infl ammatory genital lesions (12 had Zoon ' s-like balanoposthitis; 6 lichen sclerosus-like lesions; 5 phimosis; 2 patients had more than one feature). Patients with gcGvHD had signifi cantly higher coincidence of oral mucosal (p<0.0001), ocular mucosal (p<0.002) and non-mucosal cutaneous cGvHD (p<0.026) compared to patients without gcGvHD (Table 2) . Further posttransplant genital lesions included unspecifi c balanitis, hyperpigmentation and melanotic macules of the glans penis, vitiligo and penis deviation. The rate of questionnaire returning was 51% (79/155) for the mBSSC and 65% for the IIEF-5 (88/155); 64% out of the patients, who returned the mBSSC reported to be uncontented with sexual functioning. The major sexual problem reported was erectile dysfunction affecting 58/88 (65%) of those who returned the IIEF-5. Only 9 out of 21 patients with gcGvHD returned the questionnaires, reporting no relevant uncontentedness or sexual dysfunction. This small return rate does not allow further interpretation in this particular aspect. Conclusions: the high participation in this study refl ects the great interest of posttransplant male recipients regarding genital issues. gcGvHD was observed in 13% of the evaluated population mostly in correlation with mucosal and non-mucosal skin cGVHD. Erectile dysfunction was the most prevalent sexual problem reported. Transplantation centers should be aware of possible male genital changes and their implications in sexual life. A regular screening should be standard part of the post-transplant control. S9 centre. In addition we have investigated the frequency of use and the outcome of attempting parenthood using stored material. A postal questionnaire was sent to 434 patients surviving a minimum of 2 years. 221 patients responded including 112 male patients and 72 female patients under the age of 42. Among male patients, 42 of 112 banked sperm and 25 subsequently attempted parenthood. Of 72 women within a reproductive age group, 12 stored either embroys and-or eggs and-or ovarian tissue. Following transplant 5 of 8 women with cryopreserved embryos used them to attempt pregnancy, one in a surrogate. Not everyone was offered the opportunity to store tissue and the reasons for this will be discussed together with outcome data of using stored material. Methods: This retrospective observational study was based on 294'922 patients who underwent autologous or allogeneic HSCT from 1980 to 2009. The primary endpoints were the cumulative incidence of death from suicide and accident, the deaths rates by suicide and accident, as well as the standardized mortality ratio (SMR) and the absolute excess risk (AER) of these causes of death. To defi ne factors associated with suicide or death from accident, a case control analysis was performed. Findings: In total 116,149 (39.4%) died. From these deaths, 189 were due to suicide and 125 to an accident. The 10 year cumulative incidence of death by suicide and by accident was 101.8 and 55.6 per 100,000 patients, respectively. There were 20.7 suicidal deaths (males 27.4; females; 14.0) and 13.7 accidental death (18.0 in males; 8.2 in females) per 100,000 personyears. Based on the Eurostat statistics,, the expected suicidal and accidental death rates per 100,000 person-years were 9.2 (males 14.0; females 4.4) and 10 .5 (males 16.3; females 4.9), respectively. The SMR and the AER of suicide after HSCT were 2.12 (p<0.001), and 10.91 higher than in the European general population for 100,000 deaths, respectively. The SMR and AER of accident were 1.23 (p<0.05) and 2.54, respectively (Table 1 ). In a case-control study, relapses were signifi cantly more frequent among patients who commit suicide after autologous HSCT (37% vs. 18%; p<0.0001). Chronic graft-versushost disease (GVHD) was signifi cantly higher among patients who committed suicides after allogeneic HSCT, as compared to controls (64% vs. 37%; p=0.001) ( Table 2) . Interpretation: This study demonstrates that there is an excess of deaths due to suicide and to a lesser extend to accidents after HSCT as compared to the European general population. Relapse was associated with more deaths by suicide and accident after autologous HSCT and chronic GVHD with more deaths by suicide after allogeneic HSCT. Patient at risk for suicide or accident after HSCT should be recognized and preventative strategies initiated. 

There is now a 15 year experience with hematopoietic stem cell transplantation for treating severe autoimmune disease. In addition, a newly evolving fi eld of mesenchymal stem cell therapy of autoimmune disease is evolving. Phase I/II studies (1, 360 patients in the EBMT/EULAR data base and 130 patients in the CIBMTR data base) using autologous HSCT in several major autoimmune disease have shown a satisfactory benefi t risk ratio. Over one-third of patients achieved a durable relative drug free remission with a treatment-related mortality of around 8-10%. ranging from 12% for SLE and 2% for multiple sclerosis). Large phase III randomized controlled trials are advanced in systemic sclerosis (ASTIS -156 patients and SCOT -75 patients Trials) and Crohn's disease (ASTIC Trial -48 patients) as well as non-randomised phase II study in MS (HALT). Results should be become available in 2012, but so far no unexpected TRM has been observed. A small randomised study in SSc (ASSIST -19 patients ) was positive. In systemic sclerosis, data of the past 12 months suggest remodelling of collagen and normalization of microvasculature after hematopoietic stem cell transplantation, a new fi nding. In MS and SLE durable remission was observed in some patients despite full immune reconstitution.

[101] S11 The interpretation of the existing trials and planning of future studies needs to take into account an evolving situation regarding improving transplantation techniques and alternate less toxic treatment options for autoimmune disease e.g. biologics. Mesenchymal stem cells have shown promise in exerting an immediate anti-infl ammatory immunomodulatory role in some autoimmune disease with little evidence of acute toxicity. Large randomised prospective trials are needed to confi rm the promising in vitro, in vivo animal model and small anecdotal clinical data. Introduction: Graft-versus-host-disease (GVHD) is a complication frequently occurring after HSCT and a potentially life threatening condition. One strategy for the prevention of GVHD is the administration of anti-thymocyte globulins (ATG), a set of polyclonal antibodies which lead to immune cell depletion in the recipient. It is still an open question whether ATG administration is benefi cial for the patients due to the complications arising from the consecutive severe immunosuppression. So far, there are no systematic meta-analysis data about this issue. Methods: We searched CENTRAL (the Cochrane Central Register of Controlled Trials), MEDLINE, trials registries and conference proceedings. Only randomized controlled trials comparing the prophylactic use of ATG versus no ATG were included into the analysis. Results: Six RCTs involving 568 participants meeting the predefi ned selection criteria were identifi ed by our search strategy. The primary outcome overall survival was not signifi cantly changed by the addition of ATG for the prophylaxis of GVHD (HR 0.88; 95% CI 0.67 to 1.15, P=0.33). The incidence of treatment-requiring or severe acute GVHD (grade II to IV) was signifi cantly lower in patients who received ATG (RR 0.68; 95% CI 0.55 to 0.85, P=0.009; NNT: 8). Also the incidence of severe acute GVHD (grade III to IV) (HR 0.53; 95% CI 0.33 to 0.85, P=0.0005; NNT: 7) was signifi cantly reduced (HR 0.53; 95% CI 0.33 to 0.85, P=0.0005; NNT: 7), but comparable data were only available for rabbit ATG. However, pooled study results regarding the incidence of acute GVHD of all grades (I to IV) showed no signifi cant benefi t (RR 0.89; 95% CI 0.74 to 1.06, P=0.20). Due to a lack of suffi cient data, a meta-analysis concerning the incidence of chronic GVHD was not possible. Pooled data regarding the incidence of relapse were not differing signifi cantly (RR 1.13; 95% CI 0.75 to 1.68, P=0.56), as well as pooled data regarding non-relapse mortality (HR 0.82; 95% CI 0.55 to 1.24, P=0.35). Conclusions: From the currently available data no recommendation on a general use of ATG in allogeneic HSCT can be conducted. Therefore, a careful consideration of the use of ATG based on the patient´s condition and risk factors of the transplantation setting should be made.

Fibroblast growth factor-7 sustains thymic expression of tissue-restricted antigens during experimental acute graft-versus-host disease S. Dertschnig (1), G. Nusspaumer (1), G. Holländer (2), W. Krenger (1) (1)University Basel (Basel, CH); (2)University Oxford (Oxford, UK)

Outcomes of allogeneic hematopoietic stem cell transplantation (alloHSCT) depend on the successful reconstruction of T-cell adaptive immunity. This process requires in turn the de novo production of naïve T cells in a competent thymus. Unfortunately, acute graft-versus-host disease (aGVHD) compromises intrathymic T-cell maturation. Impaired thymopoiesis was previously shown in murine models to be due to anti-host alloimmunity directed against components of the thymus stroma. In particular, thymic epithelial cells (TECs) -which deliver under physiological conditions the extrinsic signals required for T-cell development -are targets of disease. The systemic administration of human fi broblast growth factor 7 (Fgf7; palifermin) can, however, correct the disarray of the thymic epithelial network via its mitogenic effect on TECs and thus maintain normal T-cell development during experimental aGVHD. Fgf7's molecular and cellular mechanisms of action on thymic stromal cells are presently incompletely understood. Here we used a haploidentical murine transplantation model to study the effects of Fgf7 on medullary TEC (mTEC). A subset of these mTEC normally expresses and presents to developing T cells an array of ectopic tissue-restricted antigens (TRAs). TRAs are central to the process of thymic negative selection and are partly under control of the transcription factor Aire ("autoimmune regulator"). As detected by fl ow cytometry and immunohistochemistry, mature mTEC cell numbers were found to be progressively diminished to almost undetectable levels in the course of aGVHD. Oligonucleotide microarray analysis revealed a contraction in the diversity of medullary TRA expression. However, Fgf7 sustained a stable population of mTEC, including the subset expressing Aire, in allogeneically transplanted mice, even if total mTEC numbers remained lower than in age-matched controls. On a whole mTEC population level, Fgf7 therapy maintained in the long-term a much more diverse array of TRA expression than in untreated mice with aGVHD. The cellular expansion of the thymic epithelial compartment in response to Fgf7 therefore conserved a crucial function of mTEC despite the continued existence of injurious donor T cells in the host thymus. Hence, a therapeutic strategy using Fgf7 may promote regeneration of a functionally competent T-cell adaptive immune system following alloHSCT. Background: Cyclosporine (CSP) or tacrolimus (TAC) in combination with methotrexate (MTX) are the most common regimens for the prophylaxis of acute graft-versus-host disease (aGvHD). To evaluate the current status of prophylaxis for aGvHD in Japan, we retrospectively analyzed data from the transplant outcome database of the Japan Society of Hematopoietic Cell Transplantation. The effect of Busulfan/Cyclophosphamide (BuCy) vs. Cy(VP16)TBI on aGvHD has been evaluated in patients receiving BMT but not in patients receiving PBSCT (1). Data from 96 consecutive AML patients undergoing non-T-depleted HLAidentical PBSCT after myeloablative conditioning from Jan 2001 to Jan 2011 were included in the study. 54 Patients received 12Gy TBI (+Cy (n=26); +Cy/VP-16 (n=28)) and 42 patients received non-targeted full dose i.v. busulfan (BuCy(n=7) or CyBu (n=35; see (2) for details). GvHD prophylaxis was with MTX/CyA. TBI and noTBI groups were comparable for all relevant pretransplant risk factors (Table 1) except for year of transplant (median 2004 vs. 2008; p<0.001) and age at transplant (41.6 yrs vs. 50.0 yrs; p<0.001) which refl ects the gradual introduction of Bu based regimens and increased age limits at our institution. The cumulative incidence of aGvHD at day 100 was 67.3% in the TBI and 38.6% in the noTBI (p<0.001; severity and organ involvement is summarized in table 1. The median onset of aGvHD was 12(6-45) days in the TBI and 17 in the noTBI group (p=0.04). The cumulative incidence of TRM and the relapse incidence at one year and two years respectively were 22.7 vs. 14.8% (p=0.96) and 24.2% vs. 29.6% (p=0.7). Overall survival at two years post transplant was similar (56.8% vs. 53.7%, p=0.78.). The cox regression analysis after correction for age, remission status, year of transplant and CMV status confi rmed a signifi cant aGvHD risk reduction (RR=0.4 (0.18-0.90; p=0.028)) for the noTBI group without infl uence on TRM and OS. In the TBI group, the median duration of clinical mucositis was 12 days (0-30 days), whereas in the noTBI group it was 5 days (0-19 days; p<0.001). The cumulative median dose of morphine used for mucositis related pain was 269 mg (0-3055 mg) in the TBI group vs. 36 mg (0-483 mg) in the noTBI group (p<0.001). Reduced morbidity resulted in a signifi cant reduction of inhospital stay from a median of 27 days (mean 34.9 (29.2-42.3 95% CI) in the TBI group to a median of 21 days in the noTBI group (mean 24.5 (21.6-27.4 95% CI) p=0.014). The use of non-targeted i.v. busulfan regimen demonstrated a lower incidence of aGvHD, less toxicity and morbidity and shortened inhospital stay, making thus an attractive alternative for AML patients Toll-like receptors (TLRs) are evolutionary conserved receptors activated by exogenous and endogenous stimuli, having been implicated in the development of infectious and autoimmune diseases. Graft vs Host Disease (GvHD) is characterized by damage of epithelial surfaces in target organs caused by alloactivated T-cells recognizing host-tissue antigens. Intestinal bacterial breakdown products and conditioning-induced loss of gastrointestinal tract integrity result in LPS-TLR4 interaction contributing to acute GvHD whereas TLR4,9 polymorphisms have been associated with GvHD occurrence. We studied TLR2,4 expression by qRT-PCR and FCM, in peripheral blood mononuclear cells (PBMCs), lymphocyte subpopulations and target tissues as well as the activation of TLR2,4 signaling pathway by specifi c agonists (HKLM, LPS respectively) in PBMCs from acute and chronic GvHD patients (n=18, n=42, respectively). As control groups served healthy volunteers (normal group) and allo-transplanted patients who either never developed or were cured from GvHD (no GvHD group). There was a signifi cant upregulation of TLR2,4 expression in PBMCs of aGvHD patients, both at a transcript (p<0,001, p<0,001) and a protein level (p<0,001, p<0,001), whereas cGvHD patients overexpressed mainly TLR2. In purifi ed by immunomagnetic separation, T-and B-cell subpopulations, a similar TLR2 (CD3+/TLR2+: p<0.001, CD19+/TLR2+: p=0.005) and TLR4 (CD3+/TLR4+: p<0.001,CD19+/TLR4+: p<0.05) upregulation was observed in aGvHD patients whereas there was only a trend to signifi cance for B-cell TLR4 overexpression in cGvHD patients. TLR2,4 immunohistochemistry in skin, oral mucosa and salivary gland biopsies, showed a distinct expression pattern in aGvHD and cGvHD patients as compared to no GvHD group. In response to TLR ligand activation, a signifi cant TLR2,4 hyporesponsiveness of PBMCs was observed by qRT-PCR and ELISA in aGvHD and cGvHD as compared to control groups: the mean fold change of TNFa and IFNb in presence vs absence of agonist per GvHD patient was signifi cantly lower as compared to its control group counterpart [TLR2/TNFa: p<0.01, TLR4/TNFa: p<0.05, TLR4/ IFNb: p<0.05]. This study describes for fi rst time in an acute and chronic GvHD setting, the PBMC and target tissue expression of specifi c TLRs as well as a TLR hyporesponsiveness to agonists. Induction of TLR tolerance by repeated exposure to TLR agonists before alloHSCT might represent a novel anti-GvHD prophylaxis, worthwhile to be investigated. We performed an international multicenter retrospective study of steroid refractory (SR) (progression after 3 days (d) or no response after 7d) and steroid dependent (SD) (recurrence during taper) acute graft-versus-host disease (aGVHD) comparing extracorporeal photopheresis (ECP) (Austria (AT), US, UK) to systemic immunosuppression (France) as second-line therapy (SLT) in patients (pts) undergoing transplant after January 2005. Response was assessed at end of ECP (8 weeks after ECP in AT cohort) or at 4 weeks after systemic SLT (inolimomab (anti-IL-2R)-21 pts; anti-TNF-alpha-20 pts). Table 1 outlines the characteristics of the 127 pts (ECP: AT-35, US-29, UK-22; non-ECP: 41). SLT was initiated at a median of 50 (range, 8-99) d after transplant/DLI (10 pts). All pts received steroids prior to SLT (median of 16 d). Median number of ECP treatments was 12 (range, 2-45) over a median duration of 50 d. aGVHD response was higher in the ECP group (73% vs. 32% P <0.001) and after an ablative regimen (68% vs. 47% P=0.024). Lower grade (gr) aGVHD ≤ 2 (87% vs. 7% P<0.001)) and gut (GI) aGVHD ≤ stage 2 (68% vs. 43% P=0.008) at SLT-onset were associated with a higher response rate to SLT. aGVHD gr (≤ vs. >2) (adjusted for type of SLT, regimen intensity) was an independent predictor of response (OR=72. In pts with SR/SD aGVHD ≤ gr2 at onset of SLT, non-ECP SLT (adjusted for regimen intensity) was associated with an inferior survival (HR=2.53, 95% CI 1.04-6.11 P=0.039). Although limited by its retrospective nature and patient selection bias, this international multicenter study suggests that ECP is an effective SLT therapy for SD or SR aGVHD and maybe superior to non-ECP intervention. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the only curative treatment of hematologic malignancies and hematopoietic failure syndromes in adult patients, but compromised by acute graft-versus-host disease (aGvHD). aGvHD-specifi c urinary peptide biomarkers were combined in a classifi er enabling examiner-independent, unbiased diagnosis of aGvHD as reported earlier, based on urine from 141 patients transplanted in 4 transplant centers. Here, additional 315 patients transplanted between 2006 and 2010 were prospectively monitored with this aGvHD-specifi c proteomic classifi er. The median age at transplantation was 48 years, ranging from 17 to 71. The majority of the patients were transplanted for acute leukemias (n=169) followed by chronic diseases (MDS, MPS, CML, n=115), 21 lymphomas and 10 non-malignant disease. Complete remission (CR) at transplantation was achieve in 144 patients, while 146 were not in CR. Unrelated donors were found for 203 patients. Reduced intensity conditioning was [O121] S14 applied to 224 patients. GvHD-prophylaxis was a combination of CSA and MTX or MMF in 279 patients, 22 had other GvHDprophylaxis, 3 had none, about 80% received additional antithymocyte globulin prior to transplant. 10 recipients were T-cell depleted. Applying the previously established aGvHD classifi er to the blinded, prospectively collected urine samples allowed correct classifi cation of patients with aGvHD about 10 days (1 to 21 days) prior to clinical diagnosis with a sensitivity of 80.26% [CI 95%:74.6-85.28] and 75.5% specifi city [CI 95%: 72.2-78.5]. aGvHD grade III and IV development could be predicted 21 days prior to clinical diagnosis with accuracy greater 85%, distinguishing patients with aGvHD grade I and II from those with higher grade complications. We identifi ed peptides from collagen, albumin, beta2 microglobulin and CD99 as biomarkers, indicating signifi cant disturbance in collagen metabolism and T-cell activation. Safety and feasibility of a proteomic-classifi er-driven pre-emptive steroid treatment was assessed: 49 patients received pre-emptive treatment within 72h upon proteomic results positive for aGvHD. Pre-emptive therapy did not lead to increased relapse rates or infections. The results indicate that proteomic screening of urine after allo-HSCT predicts aGvHD development accurately at least 10 days prior to clinical diagnosis and allows prediction of developing aGvHD grades III and IV 3 weeks prior to clinical manifestation. Single (s) or double (d) cord blood transplantation (CBT) are valid treatment for hematologic diseases. One limitation of using dCBT is the cost of the 2 units, but comparison of the total cost of the 2 procedures has not yet been investigated. We analyzed costs-effectiveness of dCBT compared to sCBT in 134 pts with acute leukemia. CBT was performed in France from 2001-2009. All hospital costs were estimated from donor search to 1-year after CBT, according to the French public health system. A Markov decision analysis model was used to calculate the QALY (quality-adjusted life years) and cost-effectiveness ratio (ICER). For cost-effectiveness analysis, reduced intensity conditioning (RIC) and myeloablative conditioning (MAC) were analyzed separately. Forty pts were transplanted for ALL and 94 for AML, in CR1. Sixty one pts received a sCBT and 73 a dCBT. Seventy nine pts had a RIC and 55 a MAC. Neutrophil recovery was achieved in 115 pts in a median time of 23 days. Neutrophil recovery was not different after sCBT or dCBT. dCBT was associated with higher acute GVHD grade II-IV: 56% vs 30% for sCBT, p=0.003. At day-100, 53% of pts had CMV reactivation (37% after sCBT and 71% after dCBT p=0.01), 45% had viral infection other than CMV and 49% had bacterial infection. Fifteen pts (11%) had a 2nd transplant, 6 for graft failure (4 in sCBT group and 2in dCBT group) and 9 for relapse (6 in sCBT group and 3 in dCBT group). Cumulative incidence of relapse at 2-year was lower after dCBT: 29% vs to 42% after sCBT (p=0.04). No difference was observed for NRM and chronic GVHD. Survival at 2-year as 40% vs 58% after sCBT and dCBT (p=0.04). Leukemia-free survival at 2-year was 30% in sCBT vs 49% in dCBT (p=0.09). The mean cost for donor identifi cation and UCB acquisition was 28.164€ for sCBT and 48.929€ for dCBT. The estimated costs within 1 year after RIC-sCBT was 133.790€ and it was 211.735€ after MAC-sCBT. The estimated cost was 180.549€ after RIC-dCBT and 205.375€, after MAC-dCBT. Table 1 summarizes costs by type of graft and conditioning. In the MAC group, dCBT was associated with lower cost (minus 13.554€) and better effectiveness (plus 0,53 QALY). The cost per QALY obtained after RIC-dCBT compared with sCBT was 91.199€. In France, dCBT is associated with higher incidence of acute GVHD, lower relapse and better survival in adults transplanted for acute leukemia. With MAC, dCBT is the best option and the cost per QALY obtained for dCBT when using RIC is acceptable. During pregnancy maternal B and T cells may become primed against fetal HLA and minor Histocompatibility Antigens (mHA) inherited paternal antigens (IPA). Anti-IPA immunity accounts for the lower relapse rates in maternal haplo-identical hemato-

S16 poietic stem cell (HSC) transplants compared to paternal ones (Stern, et al. Blood 2008) . We hypothesized that CB graft recipients who share the same antigen(s) as their CB donor IPAs might have a lower incidence of relapse due to the presence of primed maternal cells than recipients who have no antigens shared with their CB IPAs. Methods: We carried out a retrospective, observational study in 1,155 patients with hematological malignancies who received single unit CB grafts from NYBC between 1993-2006, including 453 with ALL, 392 with AML and 310 with CML, MDS and lymphomas. Primary study endpoints were relapse and grade III-IV acute GVHD; secondary endpoints were engraftment, mortality and treatment failure relapse or death). Maternal HLA typing was available, permitting inferred identifi cation of CB IPAs at HLA-A, -B (intermediate resolution) and DRB1 (high resolution). Patients were classifi ed by whether they shared the same antigen(s) as the CB IPAs or not, as well as by the number of HLA mismatches, mismatch direction and match to non-inherited maternal HLA (NIMA. Results: Among 1,094 patients given HLA mismatched CB units, 1,030 shared antigens with one or more CB IPA targets and 64 (6%) had no shared IPA target. Transplants with a shared IPA target had a signifi cantly lower incidence of relapse during the fi rst 3 years post-transplantation compared to those that had no shared IPA target. This effect was detected in both ALL and AML (combined hazard ratio=0.4 in a multivariate analysis with other risk factors, P<0.001. Figure 1A) but not in patients with other malignancies. The lower relapse rate was strongest in patients with 1 HLA mismatch (HR=0.1, P<0.001) and was refl ected in a lower treatment failure rate in this group (HR=0.5, P=0.012). Acute GVHD in patients with a shared IPA target was increased but not signifi cantly so ( Figure 1B ). Maternal HLA match to the recipient (a control for possible effects of unprimed maternal cells) was not associated with relapse risk. We conclude that microchimeric IPA-primed maternal cells in CB can after HSCT reduce relapse in patients with ALL and AML and might play a role in cancer surveillance in general. Table 1 . Results: We used the EBMT risk score based on 5 pre-transplant variables, age (<> 30 yy), disease phase (early vs advanced), interval time from 1st to 2nd transplant (<> 1 year), donor type (SIB vs ALT) and donor-recipient sex mismatched (F->M vs others) In multivariate Cox analysis negative predictive factors on overall survival (OS) were: interval time from 1st to 2nd transplant < 1 year (RR 0.6, 95% CI 0.4-0.9, p<0.01), age >30 yy (RR 1.4, 95% CI 1-1.9, p<0.04), advanced disease phase (RR 1.4, 95% CI 1-2, p<0.03) and alternative donor (RR 1.7, 95% CI 1.2-2.3, p<0.002). Disease phase and donor type were favorable predictive factors on non-relapse mortality (NRM) in multivariate Cox analysis. We also analyzed the impact on the outcome of a different donor type for second allogeneic HSCT and we observed no effect both on OS and NRM in multivariate analysis. We therefore identifi ed three groups of patients: group A score 0-2 (n=117), group B score 3 (n=81), group C score 4-5 (n=51). The actuarial 5yy OS according to group A (Low risk), group B (Intermediate) and group C (High risk) was 26% vs 19% vs 14% respectively, p<0.001 ( Figure 1 ). The cumulative incidence of [O126] S17 NRM according to Low, Intermediate and High risk group was 19% vs 26% vs 35% respectively, p<0.01. Conclusions: This study confi rms that the EBMT score is a useful predictor of outcome for patients with acute leukemia receiving a second allogeneic transplant because of a leukemia relapse. The 5 year actuarial ranges from 26% to 14%, and it may thus be possible to select patients who will most benefi t from a second transplant. In our series changing donor had no impact on the outcome. Objectives: There are no systematic data on the prognostic impact of the recipient's and donor's cytomegalovirus (CMV) serostatus focused on the era of preemptive therapy of CMV infection in patients after allogeneic stem cell transplantation (allo-SCT). Methods: We analyzed 16 628 adult de novo acute leukemia patients documented in the registry of the European Bone Marrow Transplantation (EBMT) group allografted between 1998 and 2009 (cohort I). Additionally, 97 acute leukemia patients who underwent allo-SCT (University Medicine Berlin; Germany) were analyzed in more detail including kinetics of the hematopoietic chimerism and CMV infection (cohort II). Results: CMV seronegative recipients (R-CMV-) allografted from a CMV seronegative donor (D-CMV-) had a signifi cant (p<0.001) better median overall survival (OS) (56 months) and leukemia-free survival (LFS) (49 months) than R-CMV-/ D-CMV+ (49 and 44 months), R-CMV+/D-CMV-(49 and 43 months) and R-CMV+/D-CMV+ (51 and 45 months) cohort I cases, respectively. The recipient/donor CMV serostatus remained as a signifi cant prognostic factor in a multivariate analysis. The negative impact of the recipient's and/or the donor's CMV seropositivity was due to both an increased treatment-related and relapse-related mortality and strongest in patients with a matched unrelated donor. Conversely, the CMV serostatus had no signifi cant impact on the occurrence of graft-versus-host disease or graft failure. The CMV infection risk in comparison to R-CMV-/D-CMV+ cases was 3-fold for R-CMV+/D-CMV+, 4-fold for R-CMV+/ D-CMV-and 9-fold for R-CMV-/D-CMV+ cohort II patients. Recipient's CMV seropositivity was also in this cohort associated with an inferior OS (66% vs. 35% 3 years after allo-SCT; p=0.03) and, even more interestingly, this impact remained if analyzing only patients who did never develop a CMV reactivation until the last follow-up. Finally, recipients seropositivity was associated with a lower rate of complete day +30 chimerism (68% vs. 92%, p=0.02) without infl uencing time to engraftment or immune reconstitution kinetics of CD4+ T cells, CD8+ T cells and NK cells. Conclusion: The recipient/donor CMV serostatus remains an important prognostic factor despite the increasing use of sophisticated techniques to diagnose and treat CMV infections in acute leukemia patients after allo-SCT. Interestingly, this observation seems to be mainly explained by indirect virus effects and not a CMV reactivation itself. Objectives: In patients with AML, bone marrow (BM) cytomorphology allows monitoring the remission status post-transplant, but the impact of other morphologic features such as dysplasia or cellularity in this specifi c setting remains unclear. Methods: We analysed the frequency and the prognostic impact of these parameters in stem cell recipients with AML. 112 patients (60 males, 52 females; 17-72 yrs) received unrelated/related allogeneic HSCT following reduced/myeloablative conditioning resulting in blast clearance ≤5%. BM cytomorphology was performed at days +30 (d30) and/or +100 (d100) posttransplant. Following the criteria of Goasguen et al. (1992) , thresholds of 10%, 20%, and 50% of dysplastic cells were used to defi ne any hematopoietic lineage as "dysplastic". Overall BM cellularity and cellularity in granulopoiesis (GP), erythropoiesis (EP), and megakaryopoiesis (MP) were analysed performing age-related adjustment. Results: Dysplasia in ≥10% of cells was frequent in all hematopoietic lineages on d30 (MP: 21/44; 47.7% pts evaluable in this lineage; EP: 18/52; 34.6%; GP: 15/60; 25.0%) and also on d100. On d30, normal overall BM cellularity was seen in only 34/75 pts (45.3%) with available morphology, while cellularity was increased in 13 (17.3%) pts and reduced in 28 (37.3%). Also on d100, aberrant cellularity was frequent (increased: n=4, 6.5%; reduced: n=24, 38.7%). Relapses were less frequent in pts with normal BM cellularity on d30 (7/34; 20.6%) than in those with reduced (9/28; 32.1%) or increased cellularity (10/13; 76.9%; p=0.001). Reduced overall cellularity on d30 signifi cantly correlated with inferior 2-year OS rates as compared to increased or normal cellularity (31.4% vs. 44.0% vs. 59%; p=0.009). Reduced cellularity separately in GP, EP, and MP on d30 also correlated with inferior OS. Competing risk factor analysis revealed that pts with increased cellularity on d30 had a cumulative relapse incidence of 62% (95%CI 35%-89%) with a HR of 6.68 (p=0.00014). Pts with reduced cellularity had a NRM of 36% (95%CI 18%-54%) with a HR of 2.6 (p=0.006). In contrast, dysplasia in at least 10%, 20%, or 50% of cells on days +30 and +100 and cellularity on d100 did not correlate with survival/relapses. Conclusion: Aberrant cellularity is frequent in the post-transplant period and seems to represent an additional prognostic parameter for stem cell recipients with AML. In contrast, dysplasia seems to be an unspecifi c phenomenon post-transplant. In this study we test the predictive value of WT1 expression on leukemia relapse in 131 consecutive patients with acute myeloid leukaemia (AML) patients undergoing an allogeneic hemopoietic stem cell transplant (HSCT). WT1 expression was assessed before and after HSCT, and expressed as copy numbers/ 10 4 abl; 72 patients were in fi rst complete remission (CR1), 28 in second or subsequent tremission and 27 had active disease at the time of transplant; the median age was S19 44 years (15-69); 87 patients received a myeloabltive conditioning, and 34 a reduced intensity regimen. End points of the study were cumulative incidence of transplant related mortality (TRM), relapse related death (RRD) and actuarial overall survival (OS). Pre-transplant WT1. The median pre-transplant WT1 level in marrow cells, was 100 copies: TRM was 12% vs 13% for patients with pre-transplant WT1 levels </>100 (p=0.8); RRD was 23% vs 47% (p=0.004) and OS 65% vs 40% (p=0.004). Post-transplant WT1. On day+30 post transplant patients were again re-evaluated: TRM was 12% vs 10% for post-transplant WT1 levels </>100 (ns); RRD was 29% vs 53% (p=0.01) and OS was 58% vs 37% (p=0.03). Pre+Post Transplant WT1. When combining WT1 expression pre and post transplant we could identify 2 groups: group A (patients with WT1 <100 either pre and/or post transplant), and group B (patients with WT1 >100 pre and post transplant). The RRD is 26% in group A and 70% in group B (p<0.00001). We then asked the question: can we identify patients in group A who will relapse. To answer this question we looked at WT1 expression on day +60, and looked at the difference between WT1 expression on day +60 and day+30. We found that when WT1 expression was higher on day+60 than on day +30 (n=36) patients) then the risk of RRD was 39% compared to 21% if WT1 expression on day +60 was not higher as compared to day+30 (n=72) (p=0.04). In conclusion WT1 expression pre and post-HSCT identifi es 3 groups of patients: low risk -patients with low WT1 expression pre or post HSCT day+30 and no increase in WT1 on day+60 (n=72; RRD 21%), the intermediate risk with WT1 low expression but increase on day +60 (n=36; RRD 39%) and the high risk patients with high WT1 pre and post trasnsplant (n=26; RRD 70%). High risk and especially intermediate risk patients could be considered for early pre-emptive therapy. In haploidentical HSCT, in vitro T-cell depletion of the graft is an effective method to prevent GvHD. We investigated a new T-cell depletion method (using the Clini-MACS system) which removes alpha/beta T cells and B cells, while retaining gamma/ delta T-cells, NK and other cells in the graft. So far, 32 pts have been treated in Tübingen and Rome. PBSC manipulation resulted in 4.5 (range 3.8-5) and 4.3 (3.7-5) log-depletion of alpha/ beta T cells in Tübingen and Rome, respectively. The median number of CD34+ cells in the 2 centers was 12x10 6 /kg (5-38) and 11x10 6 /kg (8-40), respectively. Pts were given 107x10 6 /kg (35-192) and 83x10 6 /kg (34-242) CD56+ NK cells, respectively. The median number of gamma/delta T cells was 11.3x10 6 /kg (5-30) and 7.5x10 6 /kg (1.4-25), respectively. No further post-transplant GvHD prophylaxis was given. The 15 Tübingen pts had advanced/refractory leukemias (8 ALL; 7 AML/MDS/JMML; 9 active disease, 6 in CR2-6). For these poor-prognosis pts, a RIC regimen (L-PAM, Thiotepa, fl udarabine or clofarabine and OKT-3 or ATG) was used. All pts engrafted, the median time to PMN and PLT recovery being 10 (8-12) and 11 days (6-28) respectively. Three and 1 pts had grade II and III acute GvHD, respectively. Three pts experienced chronic GvHD (1 extensive). Three pts relapsed after HSCT and died; 1 pt died of multi-organ failure. 8 pts are in remission (median FU: 7 months, 5-15). The Rome cohort comprised 17 pts with ALL (13), AML (3) and NHL (1). All children but 1 had relapsed/refractory disease. In particular, 10 pts were transplanted in CR2 and 6 with more advanced disease. Myeloablative conditioning included fractionated TBI, Thiotepa, fl udarabine/L-PAM and ATG. All pts but 1 engrafted, the median time to PMN and PLT recovery being 12 (10-18) and 13 (8-15) days, respectively. Only 1 pt had grade I acute GvHD. No pt of the 12 at risk experienced chronic GvHD. At a median FU of 7 months (1-13), 14 pts are alive and disease-free; 3 pts relapsed (1 died) and 1 had fatal lung aspergillosis. In both cohorts, gamma/ delta T cells started to expand faster than alpha/beta T cells in the early post-HSCT period, whereas at day +100 alpha/beta T cells were predominant. These data indicate that transplantation of TCR alpha/beta/CD19 depleted cells from a haploidentical donor results in sustained engraftment, rapid immune recovery and low incidence of GvHD. The anti-leukemic effi cacy of this approach needs to be evaluated with a longer follow-up. Engraftment rate and speed of a single CBU in adults remains unsatisfactory. Tx of 2 CBU may overcome this problem with an S20 increased rate of GVHd. Until now, attempts at using ex-vivo expanded CBU have been unsuccessful to promote long term engraftment. We report the results achieved in the fi rst 8 Pts included in a PCT of transplantation of a single ex-vivo expanded allogeneic CBU. Eudract 2008-006665-81, Clinicaltrials. gov NCT 01034449. Methods: Adults pts with an indication for SCT and unable to tolerate MAC (age>45, co-morbidities, previous HD therapy) were included if no Id sibling, no MUD 9 to 10/10 HLA matches and no CBU fulfi lling the HLA matching (≥4/6) and richness ≥3 to 4 x 10 7 Nuc C/kg before thawing) criteria. RIC consisted of Flu (40 mg/m 2 :d x 5d), Cyclophosphamide (50 mg/kg x 1d) ICT 2 Gy. GVHd prevention consisted of MMF (d-3 to d28) and CSA from d-3. Graft engineering: 1 CBU with > 2 and < 3 TNC/kg and 4 to 6 HLA compatibilities was thawed, CD34+ cells were selected through magnetic device (Miltenyi) and ex-vivo expanded in SF medium ( HPO1-Mako) supplemented with SCF, Flt3l, G-CSF and TPO during 12 days (Ivanovic, Cell Transplant 2011) CD34-cells were cryopreserved. On d0, expanded cells were washed and resuspended in HSA 4% and injected to the pt. Cd34-cells were thawed and injected 3 h later. Results: From 03/2010 to 06./2011 8 pts have been included, med age 55y.o. (26-64) with AL: 3, Hodgkin's: 2, MDS: 3. Pts had received 1 to 3 lines of Tx (med :2). In 1 case the expanded product was contaminated. The pt then received a back-up CBU, engrafted correctly and is AW at 19 m. For the 7 other cases, the med fold expansion of CD34+ cells and TNC was 39 (29-75) and 390 ( 127-526) respectively, leading to a graft containing 1.3 to 13x10 6 CD34+ cells/kg (med: 2). The CD34counterpart contained 3x10 6 CD3+/kg (1-5) and 0,9x10 6 CD19+ cells/kg (0,3-1,5). At d42, 6/7 pts who received the expanded graft had engrafted with ≥ 99% donors cells. The time to reach 500, 1000 PMN's and 20 000 plts was 7d (6-19), 8d (6-21) and 24d (0-39) respectively. 5 pts experienced AGVHd (1 grade IV). With a med FU of 15m (6-21m) 6 pts are alive. For the 6 pts who engrafted with expanded product the chimerism remains full donor up to 1,5 year after Tx. Conclusion: Ex-vivo expansion of CBU is feasible and reproducible, produces rapid, complete and sustained engraftment in adults after RIC. Background: A subset of 161 patients from the 278 subjects randomized to receive BM grafts in the BMT CTN 0201 study had aliquots of the BM allografts analyzed at a central laboratory for the content of CD34+ progenitors and immune cell subsets by fl ow cytometry. Demographics, disease characteristics, conditioning regimens (all myeloablative) and GvHD immuneprophylaxis of the 161 patients were similar to the entire BMT CTN0201 population.

Methods: 46 progenitor and immune cell subsets were selected for study based upon the absence of a strong correlation with another graft subset (Pearson or Spearman correlation >0.8) and a priori interest. Graft characteristics were described separately for survivors and those who died and compared using a nonparametric Mann-Whitney Wilcoxon test. P-values were not adjusted for multiple comparisons, but only covariates for which the false discovery rate (q value) <0.2 are presented. Univariate and multivariable analysis (adjusting for factors known to be associated with transplant outcomes) were used to estimate the association of graft characteristics with transplant outcomes. Results: Numbers of nucleated cells/kg, CD34+ cells/kg, CD8+ T-cells/kg, and CD4+ T-cells/kg were not signifi cantly associated with OS. Two cell subsets in the BM graft were signifi cantly associated with transplant outcomes with univariate p-values <0.01 and q values of <0.2: pre-plasmacytoid dendritic cells (pre-pDC) (median number 0.3x10E6/kg; p=0.007), and naïve CD8+ T-cells (CD8N; median number of 1.3x10E6/kg; p=0.009). Multivariable analysis including the content of either pre-pDC or CD8N showed better overall survival and decreased transplant-related mortality among patients receiving larger numbers of each donor cell subset. Grade ¾ acute GvHD was lower among recipients of more CD8N. Estimated 3 year OS for patients receiving >0.3 x 10E6 pre-pDC/kg was 55% versus 35% for recipients of <0.3 x 10E6 pre-pDC/kg (p=0.025) and with 58% estimated 3 year OS for patients receiving >1.3x10E6 CD8N/kg versus 35% for recipients of <1.3x10E6 CD8N/kg (p=0.01). Recipients of more than the median number of CD8N w signifi cantly less grade III-IV acute GvHD 0.34 (0.12-0.96), p=0.041. Relapse rates were similar among recipients of larger versus lower numbers of donor pre-pDC and CD8N. Conclusions: Donor pre-DC and naïve T-cells have important effects on clinical outcomes after allogeneic BMT from unrelated donors. The standard of care for mobilising blood stem cells for autografting is chemotherapy (often cyclophosphamide) with G-CSF. Plerixafor is licensed in myeloma and lymphoma where chemotherapy + G-CSF have failed, but there are almost no data on the use of plerixafor as a fi rst line mobilising agent.

Here we present an interim report on the fi rst 60 harvested patients that entered an ongoing trial of fi rst line plerixafor with G-CSF in myeloma or lymphoma patients requiring an autograft (acronym PHANTASTIC). All entrants were aged ≥18 years and had no prior attempt at harvesting. G-CSF at 5 ug/kg was given on days 1-4, and plerixafor at 2200 hours on day 4, at a dose of 240 ug/kg if creatinine clearance ≥50 mls/min (59 cases) or 160 ug/kg if creatinine clearance 30-50 mls/min (1 case). Stem cell harvesting was carried out on day 5 and if necessary on days 6, 7 and 8 with continuation of G-CSF and plerixafor, until the target yield of 4 x 10 6 CD34+ cells /kg recipient weight was achieved. The median age was 54 years (range 20-68), and 26 patients had underlying myeloma, 26 NHL and 8 Hodgkins disease. In the 30 days following G-CSF commencement, no SAE were seen though 5 patients subsequently died of disease progression before receiving their transplant. 118 AE were seen, including 10 patients with mild transient gastrointestinal symptoms, insomnia and headaches which may have been plerixafor-related. The rest were attributable to known short term effects of leukapheresis (especially citrate toxicity). 48 of the 60 patients achieved the primary composite endpoint of BOTH an adequate stem cell harvest (≥4x10 6 CD34+ cells/ kg in ≤ 2 aphereses) AND no evidence of a neutrophil count <1x10 9 /L over the 3 weeks after starting mobilisation. The median number of CD34+ cells collected was 5.49x10 6 /kg (range 1.35-21.11), and this was completed in 1 apheresis in 26 patients, in 2 for 25 patients, in 3 for 5 patients and in 4 for 4 patients. 46 of 60 successfully harvested patients are currently assessable for engraftment. The median time to neutrophil engraftment of 0.5 x 10 9 /L was 12 days, and for platelet engraftment to 50 x 10 9 /L was 21 days. Of 21 cases assessable at 12 months post transplant, none have died though 4 have relapsed; the remaining 17 are alive and well. We conclude that plerixafor is safe and effective as a fi rst line mobilisation agent and the resultant stem cells can safely support a clinically effective subsequent autograft. Background: Peripheral blood stem cells (PBSCs) are increasingly used as an alternative to bone marrow in autologous transplantations. In adult patients, the peripheral blood CD34+ cell count is a good predictor of CD34+ cell yield in apheresis. However, the determinants of stem cell yield in the pediatric population have not been well established. Methods: We retrospectively studied 396 apheresis procedures in 301 pediatric patients. Receiver operating characteristic (ROC) curves based on pre-apheresis peripheral blood CD34+ cell count were generated to facilitate prediction of the optimal timing of PBSC collection. The associations between CD34+ cell yield and age and mobilization regimen were analyzed. Results: Signifi cant differences in CD34+ cell yield among different age groups were observed. Furthermore, higher CD34+ cell yields were obtained in patients receiving chemotherapy as a part of the mobilization regimen than those without chemotherapy. A correlation was noted between the CD34+ cell yield and blood surrogate markers, including white blood cell count, absolute neutrophil count, and pre-apheresis peripheral blood CD34+ cell count. A cutoff value of >35 CD34+ cells/microliter in patients <15 years old and >45 CD34+ cells/microliter in patients ≥ 15 years old were strong predictors of an adequate PBSC collection in one apheresis session. For clinical use, ROC curves and tables were generated to assist advance planning for PBSC collection.

[O134]

Conclusions: The pre-apheresis peripheral blood CD34+ cell count is most useful in predicting PBSC yield. Our new cutoff values have better operating characteristics in children than the conventional value of 20 CD34+ cells/μL used in adults. Autologous stem cell transplantation (ASCT) is potentially curative treatment of lymphoma. However peripheral blood stem cell (PBSC) mobilization may fail in some patients. New agents were recently developed which may improve the yield of PBSC. To optimize their use within current mobilization protocols we conducted a retrospective analysis on a large series of lymphoma candidates to ASCT, to identify factors infl uencing SC mobilization outcome, including potential "ongoing" markers, able to predict the risk of failure during mobilization attempts. Demographic and clinicopathological features at mobilization, as well as the number of circulating CD34+ cells and CD34+/WBC ratio calculated on the fi rst day scheduled for collection, were analysed in 415 consecutive mobilization procedures in 388 patients with lymphoma. Chemotherapy + G-CSF was used in 99% of mobilization procedures. Mobilization failed (<2x10 6 CD34+ cells/kg collected) in 14,2 % of attempts. At multivariable analysis, the use of cyclophosphamide as mobilizing chemotherapy, the number of circulating CD34+ cells and the CD34+/WBC ratio independently predicted mobilization failure. Using these three parameters a <1% to 89% risk of mobilization failure can be predicted on the fi rst day scheduled for collection, allowing to rationally add novel agents during mobilization attempts with chemotherapy and G-CSF at risk of failure. Objectives: Non-viable CD34+ cells are commonly assessed by standard fl ow cytometry using the nuclear stain 7-Aminoactynomicin D (7AAD). 7AAD only detects necrotic/late apoptotic cells not earlier apoptosis, which engraft poorly in animal models of cord blood (CB) transplantation. The objective was to assess the ability of Annexin V staining (AnnV) to allow for early apoptosis and thus better predict stem cell function. This would then provide a rapid and robust method of predicting graft potency in thawed CB units. Methods: AnnV labeling, in conjunction with 7AAD, was used to detect apoptotic events in CD34+ cells in fresh and thawed CB units by fl ow cytometry. Similarly, the effect of the presence of the cryo-protectant Dimethyl sulphoxide (DMSO) was observed. A parallel experiment was run, concurrently, whereby CD34+ cell numbers were quantifi ed by standard ISHAGE based cell enumeration, and stem cell function assessed by colony forming unit (CFU) assay. To confi rm the relationship, CB units and CB mononuclear cells were thawed and regularly reassessed for viability and function. Results: Signifi cant loss of CFU dose was observed following thawing of cryopreserved units (50% loss, n=10) with little difference found in the viability by 7AAD (<4%). However, signifi cant numbers of CD34+annV+ events were found within the 7AADgated population (60% increase AnnV+). The AnnV-dose correlated with the observed CFU in 78 fresh and 28 thawed samples (r2=0.53). In fresh samples (n=8), low dose DMSO (5-10%) was tolerated but high concentrations of DMSO (20-40%) led to a rapid loss of CFU dose (40% DMSO; 100% loss in 1 hour). This was only refl ected by viability by AnnV (40% DMSO; 100% AnnV+, 40% 7AAD+). Finally, using the relationship between CFU and viable cell dose, it was observed that the AnnV-dose predicted the observed CFU better than using 7AAD alone. Conclusion: These data indicate that measuring the level of apoptosis is a better predictor of functional dose than the standard method of enumeration following strong cellular insult such as cryopreservation. In the case of DMSO the adverse effects are confi ned to apoptosis and consequently missed by the standard assessment. This fl ow cytometric technique, therefore, has the potential to be a rapid test of potency of CB units that should be simple enough to allow standardization in a multi-centre scenario, and helps defi ne best practice for the storage and use of CB units in transplantation. Results: During the study period, 65 alloHSCT (100%) and 57 autoHSCT (44,5%) were searched at least once, and 672 samples were processed for RV detection. The median number of samples/pt was 3,46 (range: 0-23; alloHSCT: 7,76, and autoHSCT: 1,32). A total of 191 RVI episodes (147 in alloH-SCT, 43 in autoHSCT) were diagnosed in 85 pts. The median number of RVI episodes/pt was 2,24 (range: 1-10). The viral distribution was: 86 RV (45%), 39 RSV (20,5%), 36 PIV (19%), 30 BoV (15,7%), 28 IV (14,7%), 6 MPV (3,1%), 1 CoV (0,5%), and 1 AdV (0,5%). The types of IV found were: 17 pandemic 2009 H1N1, 3 seasonal IV-A, and 7 IV-B. There were 35 mixed RVI episodes. The median presentation was at day + 74 post-HSCT. 88% episodes were classifi ed as URTI, and 12% as LRTI (7,3% bronchitis and 4,7% pneumonia; the majority of them in allo-HSCT). The RVs responsible for the LRTIs were: RSV (11), H1N1 (5), RV (3), IV-B (2), PIV (1), and MPV (1). Risk factors for progression to LRTI were: 1) CMV seropositivity (83% vs 16%, p=0,037); 2) lower IgG (0,46 vs 0,61 g/L, p=0,019); 3) lymphopenia (295 vs 855/mm 3 , p=0,035), 4) use of corticosteroids (56% vs 14%, p=0,01), and 5) use of > 1 immunosuppressant (26,7% vs 5,3%, p=0,047). The overall mortality and the attributed mortality within 1 month from RVI diagnosis was 5,7% and 3,1% (6 cases: 3 RSV, and 3 H1N1). Conclusions: 1) RVI episodes are frequently discovered, particularly among the alloHSCT recipients, when an intense screening activity is implemented; 2) In our series, RV and BoV were the most common, and rarely associated with LRTI; 3) RSV and IV (specially H1N1) can lead to serious disease with high mortality; 4) lymphopenia, low IgG, corticosteroids, and CMV seropositivity seem to be associated with progression to LRTI. Alemtuzumab is a monoclonal anti-CD52-antibody frequently used to deplete T cells (TC) in the context of allogeneic hemopoietic stem cell transplantation (HSCT). We recently showed long-term persistence of CD52-negative T cells reconstituting in patients after Alemtuzumab mediated T-cell depletion (TCD). The lack of CD52 on the cell surface was due to missing glycosyl-phosphatidyl-inositol (GPI)-anchors. We found that the persisting GPI-anchor negative T cells exhibit reduced CMV-specifi c T-cell function. Since EBV-reactivation and EBVassociated diseases (for example post transplantation lymphoproliferative disorder PTLD) are life-threatening complications after HSCT, we also investigated on EBV-specifi c T-cell functions of GPI-anchor negative T cells to estimate their impact on the clinical course of our patients with EBV-viremia (n=6) and EBV-associated PTLD (n=4). PBMC were frozen at different time points in the course of EBV-reactivation and PTLD-therapy. We identifi ed EBV-specifi c TC by EBV-peptide/HLA-A2 Tetramer staining and analyzed peptide-specifi c cytokine production by intracellular FACS and ELISPOT-assay. Antigen-specifi c proliferation was tested in CFSE-assays and cytotoxicity was determined in chromium release assays. EBV-peptide loaded dendritic cells of our patients, peptide loaded K562-cells, as well as autologous EBVblasts (LCL) were used as stimulators in the functional assays. GPI-anchor negative TC showed a reduced EBV-specifi c TC-response. Patients with EBV-PTLD had no EBV-specifi c TC prior to the occurrence of PTLD. In patients with progressive or persistent disease, EBV-specifi c TC were fi rst detected after clinical recovery. Whenever EBV-reactive TC were found, they were mainly present in the GI-anchor positive TC-compartment. In conclusion, we confi rmed that GPI-anchor negative T cells reconstituting after Alemutzumab mediated TC-depleted HSCT show an impaired antiviral function. Our data support the hypothesis that these functionally impaired GPI-anchor negative TC, persisting years after HSCT, are partly responsible for some of the known viral complications in patients undergoing Alemtuzumab based conditioning regimen. We have demonstrated earlier that application of donor lymphocyte infusions (DLI) replenishes the GPI-anchor positive TC-compartment. Therefore, We hypothesize that DLI also improve immunologic control of virus associated diseases associated with CMV and EBV-reactivations. Background: Therapy with cidofovir (CDV) is often used in patients with BK-virus-associated hemorrhagic cystitis (BKV-HC) after allogeneic hematopoietic stem cell transplantation (HSCT) but the modifi cation of BKV load on blood during antiviral therapy has been scarcely investigated in previous studies. The aim of this study is to retrospectively assess the relationship between CDV therapy, clinical response and modifi cation of BKV viremia. Patients and Methods: Over 2 years, 32 cases of HC with documented BKV viremia and treated with i.v. CDV were collected among 13 centres. Results: The patients had a median age of 24 y, range 3.4-62 (12 of them < 18 year-old at HSCT), mainly affected by acute leukemia (22 patients), and in hematological CR at HSCT (21 patients). Donor origin was a sibling in 10 patients (31%) and an unrelated/mismatched family donor in 22 patients (69%). The source of stem cells was BM, PB, or CB, in 13, 11 and 8 patients, respectively. Conditioning regimen was mainly based on busulfan (21 patients, 66%) or on TBI (7 patients, 22%). ATG was used as prophylaxis of GVHD in 17 patients (53%). HC occurred at a median time of 29 days, range 6-28, from HSCT and was scored as grade I, II, III, and IV in 4, 4, 18, and 6 patients, respectively. Concomitant viral infection was recorded in 12 patients for CMV, and in 3 patients for adenovirus and EBV, respectively. In 31 of 32 patients evaluable, therapy with CDV was associated with complete response in 19 patients (60%), partial response in 8 patients (25%) and no improvement in 4 patients. Renal toxicity was reported in 9 patients (grade I, 4, grade II, 3, grade III, 2) whereas 1 patient had both grade III renal and cardiac toxicity. 12 of 31 patients had BKV viremia assessment before every CDV administration. In 10 of 12 patients, a virological response was observed, as follows: reduction of 1 log in 4 patients, reduction of 3 log in 4 patients, and reduction > 3 log in 2 patients. All these 10 patients showed a clear improvement of HC during CDV therapy: 9 had a complete response and 1 had a partial response. In the 2 patients without virological response no clinical benefi t was observed from CDV therapy with CDV. After a median follow-up of 259 days (range 62-581), the OS of the whole cohort was 66%. Conclusion. 1-3 log-reduction of BKV viremia is associated with clinical recovery or improvement from HC. Prospective studies are needed to confi rm this data. were serially monitored for cytomegalovirus (CMV). 645 pts (60,8%), had a myeloablative transplant (MAT); 415 (39,1%) having a reduced-intensity conditioning (RIC). Matched sibling/ family was 573 (54%) and unrelated donors was 482 (45,4%). Graft source was peripheral blood stem cells (PBSC) in 717 (67,6%), bone marrow (BM) in 266 (25%), and cord blood (CB) in 77 (7,2%). Anti-thymocyte globulin was administered in 541 (51%). Donor (D)/Recipient (R) CMV serostatus was analized as follows: D-/R-, D+/R-, D+/R+, D-/R+. Serial weekly monitoring for CMV reactivation was performed using either antigenemia or viremia according to Center policy. CMV reactivation was documented in 466 patients (43,9%) at a median of 80 days post HSCT (range 1-365). CMV reactivation was observed with incidence of 56%, 46,7%, 24%, 15% rispectively in D-/R+, D+/R+, D+/R-, and D-/R-, according to pre-transplant CMV serostatus. On univariate analysis, the signifi cant risk-factors identifi ed for CMV reactivation included D-/R+ and D+/R+ CMV serostatus (P<0.0001); matched unrelated versus sibling donor (P<0.0001), and acute graft versus host disease grade 2-4 versus 0-1 (P<0.0001). Hematological diagnosis and disease status at transplant, graft source (BM versus PBSC versus CB), and the type of conditioning regimen (MAT versus RIC) did not infl uence signifi cantly the risk of CMV reactivation. 725 patients (68,3%) are alive and well, at a median follow-up of 7 months (range 0,3-12). CMV reactivation had no signifi cant impact on overall survival. Regarding pretransplant CMV serostatus, survival was signifi cantly lower in D-/R+ patients compared with the other groups (P<0.005) (74% D-/R-; 65% D+/R-; 66% D+/R+; 54% D-/R+). Furthermore survival was highly affected by timing of reactivation: 41% in patients reactivating before day +30 and 69% in those with delayed reactivation (P<0.0001). These data, obtained in a large multicenter series, confi rm prospectively the impact of pre-transplant serostatus on CMV reactivation. Interestingly, SC source has no impact in this cohort, while type of donor (familiar or not) is a signifi cant factor. The timing of reactivation proved to infl uence survival signifi cantly. In patients receiving allogeneic HSCT, outcome of HCMV infection results from balance between viral load/replication and its control by the reconstituting T-cell response. Using a cut-off of 30.000 HCMV DNA copies/ml blood for initiation of pre-emptive therapy and cut-offs of 1 and 3 virus-specifi c CD4+ and CD8+ T-cells/uL blood for defi ning T-cell protection, we conducted a prospective study aimed at verifying whether patients attaining such immunological cut-offs were protected from HCMV reactivation. A total of 131 young patients (median age, 8 years; range, 1-23 years, 42M/89F), receiving allogeneic HSCT from January 2007 through January 2010, were studied. Patients had either malignant (n=98) or non-malignant disorders (n=33). Antiviral treatment was stopped following two consecutive negative blood controls. The frequency of HCMV-specifi c CD4+ and CD8+ T-cells producing IFN-gamma and IL-2 was determined by fl ow cytometry, following in vitro stimulation with autologous monocyte-derived, HCMV-infected, dendritic cells. In the fi rst 3 months after transplantation, 55/89 (62%) HCMV-seropositive patients had infection and 36/55 (65%) were treated preemptively, whereas only 7/42 (17%) HCMV-seronegative patients developed HCMV infection and 3/7 (43%) were treated.

Overall, only 1 patient developed HCMV disease (gastritis), which resolved under anti-viral treatment. After 12 months, 76 HCMV-seropositive and 9 HCMV-seronegative patients (cumulative incidence: 90% and 21%, respectively, p<0.01) displayed protective levels of HCMV-specifi c T cells. Overall, 80 of these 85 (95%) patients showed spontaneous control of HCMV infection, not requiring any additional antiviral treatment. Five patients after reaching protective T-cell levels needed pre-emptive therapy, because they developed GvHD. IFN-gamma+/IL-2+ CD4+ and CD8+ T cells were found to emerge later in patient PB than IFN-gamma+ T cells. In summary, HCMV infection was much more frequent in HCMV-seropositive patients than in seronegative patients. As a consequence, the virus-specifi c T-cell response was more frequent in HCMV-seropositive patients; however, protective activity of the T-cell response against HCMV was detected in both seropositive and seronegative patients. Protection was stable and long-lasting, unless steroid therapy for GvHD was administered, this leading in some cases to episodes of HCMV reactivation. Monitoring of HCMV infection may be tailored on reconstitution of specifi c immunity. Objectives: The emergence of cytomegalovirus (CMV) drug resistance, favored by long-term exposure to antiviral drugs and by profound immunosuppression, is a growing therapeutic challenge in allogeneic hematopoietic stem cell transplantation (allo-HSCT). The majority of ganciclovir (GCV) resistance is associated with mutations in the CMV phosphotransferase (UL97) gene. The aim of our study was to evaluate the prevalence of persistent CMV reactivation and CMV UL97 mutations at a single center in China, endemic for asymptomatic CMV infection. Methods: One hundred and thirty-eight pairs of patients and their unrelated donors (URDs) and 102 pairs of patients and their HLA-identical sibling donors, who underwent HSCT at our center from 2001 to 2009 were included in this study. Patients were monitored for detection of CMV infection for ≥2 years. CMV UL97 gene amplifying by Nested-PCR and sequencing were performed if viral load persisted for >21 days on appropriate antiviral treatment. Results: (1) All patients and almost all donors except one were CMV seropositive before HSCT. In the URD transplantation cohort, 44 patients (31.88%) experienced CMV reactivation infection pre-HSCT, as did 28 patients (27.45%) in the sibling HSCT cohort. (2) 134 (55.8%) patients had experienced early CMV infection (within day + 100) with a median onset of 27 days (range 2-64) after HSCT. 20.9% developed infection during granulocytopenia and 79.1% did so after neutrophil recovery. The majority of patients developed CMV-positive antigenemia without disease, only 9 (6.7%) patients developed CMV disease (7 patients with CMV pneumonia and 2 patients with CMV enteritis). Unrelated donors (p=0.044, RR=1.440), patients experiencing CMV reactivation pre-HSCT (p=0.012, RR = 0.632) were risk factors for early CMV infection. (3) Patients experiencing early CMV infection were with faster clearance of viremia after GCV treatment. 73 (30.4%) patients developed ≥1 episode of CMV viremia beyond day + 100 (late infection) and persistent reactivation occurred in 30 (41.1%) patients. However in those persistent infection cases, we had not detected the 7 most common GCV-resistance mutations at UL97 codons (M460V/I, H520Q, C592G, A594V, L595S, and C603W). Two novel mutations CMV UL97 H602Q and D605E were found, especially D605E mutation occurred in 16 (53.3%) patients. Conclusion: CMV UL97 D605E and H602Q mutations may contribute to persistent CMV reactivation in allo-HSCT setting.

Introduction: Adenovirus infections are a major cause of morbidity and mortality in pediatric transplantation of hematopoietic stem cells (HSCT). The intestinal tract is the most common cause of systemic adenovirus infections in this population. Early detection and quantifi cation of adenoviral infections in stool for predicting the risk of dissemination represent a major challenge Objective: To defi ne the threshold of adenoviral DNA viral load (VL) in stool associated with a risk of dissemination in blood. Methods: Between September 2010 and April 2011, all patients hospitalized in HSCT unit were weekly tested for adenovirus in blood and stool. Detection and quantifi cation of adenoviruses were performed by real-time PCR with a limit of quantifi cation of 200 copies/ml (Argene, Verniolle, France). The predictive value of dissemination in blood was determined by VL in pairs of plasma and stool as close as possible (median = 2 days). Results: Among the 51 patients 28 had a digestive infection, 16 of the 28 patients presented positive detection in blood and 9 with a plasma VL > 10000 copies/ml. The predictive value was calculated from 72 pairs plasma/stool from 26 patients. VL stool was signifi cantly higher in viremic patients (8.62 x10 8 copies/ ml) than non-viremic patients (2.74 x10 8 copies/ml) (p <10 -5 ). The threshold of 100 000 copies/ml in stool is associated with a risk of dissemination in blood with sensitivity and specifi city were respectively 96% [CI95%:78%-100%] and 65% [50%-78%]. The positive and negative predictive values were respectively 56% [40%-72%] and 97% [84%-100%]. Conclusion: In a pediatric population of HSC transplants, the detection of adenovirus in blood is preceded by a digestive replication. Quantifi cation in stool is predictive of risk of dissemination in blood. These results must be considered for the monitoring of adenovirus infections and pre-emptive treatment in patients at risk of disseminated infections. 

Introduction: An overall incidence of 10-20% of graft failure (GF) is reported after cord blood transplantation (CBT). The decision and timing to initiate search for an alternative graft and proceed to the 2nd transplant is particularly intriguing. After transplantation, the probability of engraftment increases progressively to a maximum and then gradually decreases. We analysed engraftment kinetics after CBT to develop an evidence-based strategy that would facilitate the decision process for a 2nd transplant. Patients and Methods: We investigated engraftment kinetics in a population of 1268 patients who received single, unrelated CBT for acute leukaemia (AL) in complete remission (CR) after myeloablative conditioning regimen. All patients were transplanted at EBMT Centers and reported to the Eurocord Registry from 1994 to 2010. Ratio of lymphoid/myeloid leukaemias was 813/455, refl ecting a major proportion of paediatric patients over adults (929/338). Patients were transplanted in fi rst (47%), second (45%), or third or subsequent remission (8%), respectively. Median (range) age at transplant was 8.7 (0.3-63) years. Median TNC count at freezing was 5.1 (1.1-41.83) x10 7 / Kg. The probability density model was used, that described the probability to engraft at time points after CBT and considered competing events, such as early deaths. Results: The cumulative incidence of engraftment was 86% at 60 days with a median engraftment time of 24 days. The probability to engraft peaked at day 21 and gradually decreased thereafter to day 65. The residual probability to engraft was 21.5% on day 31 (half of the median) and dropped to 5% on day 42. NRM was 26%, 29%, 37% for patients engrafting before day-32, from day-32-42 and over day-42, respectively. Graft failure was reported in 166 patients and data on treatment were available for 118. Forty-one died untreated at a median of 51 days, 10 experienced an autologous recovery and 77 underwent a second transplant (34% autologous rescue). with an overall survival of 30% at 2 years. Conclusions: The probability of engraftment after single CBT for AL in CR peaks at day 21 and decreases rapidly after day 31. Rescue actions, such as the search of another graft, should be considered beyond this time and a 2nd transplant should be carried out soon after day 42. This model can be applied to any type of transplantation and is particularly useful in the setting of high risk of graft failure, such as CBT. Due to a risk of relapse of underlying disease in pts transplanted with progressive malignancy, the use of ATG, incorporated within the conditioning prior to allo-SCT, is still controversial. We report here on a study of 245 consecutive patients transplanted between Jan 1999 and Dec 2009 for progressive MDS. Patients: Inclusion criteria included patients aged over 18 who received allo-SCT from either a sibling (n=153) or HLA-matched unrelated donor (10/10) (n=86) for progressive MDS or AML/RAEB-t. Results: 239 pts were analyzed, including 154 males. According to the WHO classifi cation at diagnosis, 85 pts had RA/RARS/ RCMD, 86 RAEB1, 62 REAB2 and 6 RAEB-t/AML. Sixty-six pts had progressed to a more advanced disease before allo-SCT. At diagnosis, 102 pts had an IPSS int-2 or higher. Cytogenetic was recorded as favorable (n=109), inter (n=61), unfavorable (n=63) and missing (n=6). Disease status at transplant was established as follows: relapsed or refractory disease (n=106) and untreated or stable disease without hematological improvement (n=133). Median age at transplantation was 53 years (20-70). Pts received myeloablative conditioning (n=105) and nonmyeloablative (n=134) including busulfan-based regimens (n=127), TBI-based regimens (n=92) or other alkylating-agentbased regimens (n=20). In this series, 95 patients (40%) received ATG as part of conditioning ('ATG' group), whereas 144 did not ('no-ATG' group). As of April 1st 2011, median followup in survivors was 50 months (IQR, 33-92) with 59 pts died of relapse and 77 of TRM. The estimated 3-yr OS and EFS was respectively 42%, and 32%. The probability of relapse, OS and EFS at 3 years was not signifi cantly different between the two groups. In contrast, the cumulative incidence of grade 2-4 aGVHD was 48% in the no-ATG group and 30% ATG group (P<.001) and the cumulative incidence of grade 3-4 aGVHD was 24% and 11% respectively (P<.001). Although the cumu-lative incidence of chronic GVHD was similar in the no-ATG and ATG groups (64% vs 46%, p=.15), a trend for a lower TRM was observed in the ATG group (22% vs 31%, p=.06). In multivariate analysis, the absence of ATG was the strongest parameter associated with an increased risk of acute grade 2-4 [HR = 2.28, 95% CI: 1.39-3.74, p=.001] and grade 3-4 GVHD [HR = 2.19, 95% CI: 1.04-4.61, p=.035]. Conclusion: The addition of ATG to the conditioning, resulted in a decreased incidence of aGVHD without increasing relapse rates and compromising survival.

Validation of the classical EBMT score and the modifi ed EBMT score in patients undergoing allogeneic haematopoietic transplantation with reduced-intensity conditioning

Introduction: The EBMT risk score has been widely validated in different settings. Recently, some modifi cations to the index have been suggested in order to improve its predictive capacity (Gratwohl. BMT. 2011). The modifi cations of the EBMT score deserve further validation in specifi c settings. Methods: All patients receiving an allogeneic hematopoietic cell transplantation with reduced-intensity conditioning (allo-RIC) in two Spanish centers between 1999 and 2008 were included. RIC consisted of fl udarabine in combination with melphalan (70-140 mg/m 2 ) or busulfan (8-10 mg/kg). The median followup for survivors was 51 months (range 3-123) The classical EBMT score (cEBMT) was calculated as originally defi ned. The modifi ed EBMT score (mEBMT) included an extra point for patients >60 years and neglected the time interval from diagnosis to transplant since its impact varies among hematological diseases. Results: A total of 306 recipients (75% from HLA identical siblings) were included. Most frequent diseases were AML/MDS (n=99, 32%) and NHL (n=63, 21%). The median cEBMT and mEBMT scores were 4 (range 2-7) and 4 (range 2-6). The probability of 100day non-relapse mortality (NRM), 4y-NRM and 4y-overall survival (OS) for the whole cohort were 13% (95%CI 10-16), 36% (95%CI 31-41) and 44% (95%CI 41-47), respectively. The only item from the score with signifi cant impact on NRM and OS in the multivariate analysis was donor type (HR 1.7 [1.1-2.6], p=0.01 and HR 1.4 [1-2], p=0.04, respectively). Additionally, CMV negative patients transplanted from CMV negative donors showed higher OS (HR 0.5 [0.3-0.9], p=0.03) and lower NRM (HR 0.5 [0.2-1], p=0.07). The 4y OS for the cEBMT categories were: score 2= 46%, 3=55%, 4=43%, 5=46%, 6=30% and 7=20% (p=0.06), while applying the mEBMT were: score 2=62%, 3=53%, 4=40%, 5=29%, 6=17% (p=0.001). NRM at 100 days using the mEBMT range from 5% (score 2) to 48% (score 6) (p= 0.004) with steadily increasing HR for each additional score point, while the predictive capacity of the cEBMT was less accurate, specially for the lower scores (2=10%, 3=7%, 4=9%). Both the cEBMT and the mEBMT showed similar predictive capacity for 4y-NRM. The addition of CMV serostatus to the scores did not improve their predictive capacity. Conclusion: In the allo-RIC setting, both the cEBMT and mEBMT predict NRM and OS although the latter seems better. The addition of other items (like CMV status) to the score needs further evaluation in large registry-based studies. Introduction: It is important to identify prognostic indicators which may predict outcome of hematopoietic stem cell transplantation (HCT). Preconditioning serum level of C-reactive protein (CRP) was recently established as a novel prognostic factor in patients transplanted for CML in 1st chronic phase. Patients and Methods: In this study we tested the value of CRP together with other prognostic factors in 256 patients who underwent myeloablative HCT for AML, ALL, MDS and advanced stage CML from February 1993 to April 2011. Recipients of sibling stem cells (121, 47%) received cyclophosphamide & TBI conditioning. In unrelated transplants (135, 53%) in vivo T-cell depletion with anti CD52 antibody was used in addition. Serum CRP levels (normal range 0-9 mg/ L) were measured at a median of 11 days before stem cell infusion whilst patients were well and off antibiotics. Patients' comorbidities were defi ned and assigned weights (1-3) by the HCT comorbidity index (HCT-CI) and disease stage was assessed in accordance with EBMT criteria with 72 patients (28%) classifi ed as early, 139 (54%) as intermediate and 45 (18%) as late stage. Results: In univariate analysis, factors associated with day 100 nonrelapse mortality (NRM) were recipient age and CMV status, disease stage, HCT comorbidity index (HCT-CI), and preconditioning CRP level. Multivariate analysis showed that recipient CMV IgG positivity, late disease stage, HCT-CI>2 and elevated levels of CRP (>9 mg/L) independently predicted increased NRM. Based on the relative risks of these 4 factors we developed a 4-point score for prediction of NRM ( Figure) . Patients scoring 0 points had a probability of NRM of 4% (95% CI: 1-16), those with 1 point 14% (95% CI: 9-23), with 2 points 29% (95% CI: 19-44) and with 3-4 points 60% (95% CI:42-86). Factors associated with survival in univariate analysis were: type of donor (sibling versus unrelated), disease stage, HCT-CI, and preconditioning CRP level. In multivariate analysis only

S28 late disease stage, HCT-CI>2 and elevated levels of CRP (>9 mg/L) independently predicted inferior survival. Probabilities of 5-year and 10-year survival in patients without any of these factors were 53 and 47% compared to 27 and 23% in patients who had one or more of the factors. Conclusion: These results further establish preconditioning levels of CRP as a key prognostic variable for allogeneic HCT outcomes, and together with other well documented factors, may signifi cantly aid the selection of patients who could benefi t from HCT. Telomere attrition induces cell senescence and apoptosis. Dyskeratosis congenita, an inherited type of aplastic anemia caused by mutations in the telomerase complex genes and accelerated telomere shortening, is associated with high treatment-related mortality (TRM) after hematopoietic stem cell transplantation (HSCT). We hypothesized that age-adjusted pre-transplant telomere length predicted TRM after HSCT. Between 2000 and 2005, 178 consecutive patients underwent HSCT from HLA-identical sibling donors after myeloablative conditioning regimens (TBI in 57 patients), mainly for hematological malignancies (n=153), all performed in a single center. The stem cell source was bone marrow (BM) in 128 cases and peripheral blood (PB) in 50 cases. Median age at transplant was 32 years (range 3-65). Graft-versus-host disease (GvHD) prophylaxis mostly consisted of cyclosporine and methotrexate (n=149, 84%). Before HSCT, blood lymphocytes were obtained from each donor-recipient pair. Telomere length was measured by real time quantitative PCR. We determined the normal age distribution of telomere length using a group of 17*3 healthy French hematopoietic stem cell donors (f= − 0.00833*age+1.522) as a control group. We then calculated the pre-transplant recipient age-adjusted telomere length in comparison to controls. Ageadjusted pre-transplant telomere lengths were analyzed for correlation with clinical outcomes post HSCT using competing risk in univariate and multivariate analyses (Fine and Gray). After age-adjustment, patients' telomere length distribution was similar among all 4 quartiles except for disease severity. There was no correlation between telomere length and engraftment, GvHD, or relapse. The overall survival was 62% at 5 years (95CI 54%-70%). After a median follow-up of 51 months (range, 1-121 months), 43 patiens had died due to TRM. The TRM rate inversely correlated with telomere length. TRM in patients in the fi rst (lowest telomere length) quartile was signifi cantly higher than in patients with longer telomeres (p=0.017) ( Figure 1 ). In multivariate analysis, age of the recipients (HR: 1.1, 95% CI [.0-1.1, p=0.0001] and age-adjusted telomere length [HR: 0.4, 95% CI [0.2-0.8, p=0.01]) were independently associated with TRM. No association was found between donor telomere length and outcome post HSCT. In conclusion, age-adjusted recipient pre-transplant telomere length is an independent biological marker of TRM after HSCT. Systemic iron overload (SIO) has been described as an adverse prognostic factor in patients undergoing allogeneic stem cell transplantation (allo-SCT). However, almost all existing studies relied on interference prone surrogate markers like ferritin and reached a variety of different conclusions. Further, optimal surrogate parameter thresholds for risk stratifi cation remain to be defi ned. We aimed at assessing the diagnostic and prognostic impact of liver iron content (LIC) measured by magnetic resonance imageing (MRI) in a cohort of 81 AML and MDS patients undergoing allo-SCT. MRI-based assessment of LIC was included into the routine workup of all our AML and MDS patients undergoing allo-SCT. Correlations between serological parameters and LIC were assessed by Spearmans rank correlation coeffi cient (r). Categorial variables were compared using Fisher's excact test. Post-transplant outcomes including the occurrence of GVHD as well as NRM and overall survival (OS) were analyzed by means of competing events statistics and Kaplan-Meier estimates, respectively. Both ferritin and RBC count were signifi cantly (p<0.001) correlated with LIC (r=0.594 and r=0.702, respectively). With areas under the receiver-operator characteristic of 0.784 and 0.823 both parameters seemed to be adequate predictors of a LIC ≥125 μmol/g. While a threshold of ≥20 RBC predicted for an elevated LIC with a sensitivity of 70.0% and a specifi city of 81.8%, a ferritin ≥1000 ng/ml albeit being very sensitive (95.5%), provided only poor specifi city (27.0%). As a consequence we decided to increase the ferritin threshold to 2500 ng/ml, which resulted in a much better balance between sensitivity (63.6%) and specifi city (81.1%. None of the three SIO parameters was associated with an increased risk of aGvHD and only LIC but not ferritin or a transfusion burden predicted for an increased cumulative incidence (CI) of NRM ( Figure 1 ). Indeed, multivariate analysis confi rmed LIC as an independent risk factor for NRM (HR 1.008 for every 1 μmol/g increase, p=0.005). This translated into a shorter OS in patients with an LIC ≥125 μmol/g ( Figure 2 ) which was confi rmed in multivariate Cox-regression analysis (HR: 3.481, p=0.021). We conclude that MRI-based measurement of LIC but not ferritin is an independent negative prognostic factor for post-transplant outcome of AML and MDS patients. Background: There are numerous data indicating that endothelial injury plays a central role in the activation of immune system following allogeneic haematopoietic stem cell transplantation (alloHCT) and precedes non-infectious complications of vascular origin and graft-versus-host disease (GvHD). The aim of the study was to evaluate a possible correlation between the changes of plasma endothelial dysfunction markers and the occurrence of early non-infectious complications and acute GvHD (aGvHD) after alloHCT. Patients and Methods: Plasma levels of von Willebrand factor antigen (vWFAg), trombomodulin (TM) and activity of factor VIII (fVIII) were measured in 63 pts transplanted with allogeneic stem cells after myeloablative (43 pts) or reduced-intensity (20 pts) conditioning regimen for AML (26 pts), ALL (18 pts), myeloproliferative neoplasms (15 pts) and lymphoma (4 pts). Endo-thelial dysfunction markers were measured before conditioning regimen (day -10), on the day of stem cells infusion (day 0) and on the day +10 and +30 after alloHCT. Results: The TM level on the day 0 was signifi cantly higher in the group of pts who developed early non-infectious complications (n=7 including 5 pts with multiorgan failure, 1 patient with VOD and 1 patient with IPS) compared to the group without those complications (n=56) (7,5 ng/ml vs 3,7 ng/ml; p 0.03) ( Figure 1 ). The concentration of vWFAg was signifi cantly elevated on the day 0, +10 and +30 in comparison to the day -10 in the whole group of patients. However, the relative increase in the concentration of vWF Ag on the day +10 in comparison to the baseline concentration on the day -10 tended to be higher in the group of pts developing grade II-IV aGvHD thereafter (n=18) compared to the group of pts without clinically signifi cant aGvHD (n=40) (mean relative increase in concentration 3,7 vs 1,8; p 0.059). Conclusions: The association between markers of endothelial dysfunction and the occurrence of early non-infectious complications and aGvHD after alloHCT was found. TM level on the day of stem cell infusion, that may refl ect endothelial injury induced by conditioning regimen, is associated with the occurrence of early non-infectious complications. The relative increase in vWFAg concentration early after alloHCT tends to be associated with the development of aGvHD. Further studies evaluating the potential role of these markers as early predictors of non-infectious complications and aGvHD after alloHCT are warranted. (1). T-alfa-1 administration caused no adverse effects whatsoever. Immune reconstitution was assessed by limiting dilution analyses of frequencies of CD4+ T cells that were specifi c for Aspergillus, Candida, Cytomegalovirus, Adenovirus, Herpes Simplex Virus, Varicella-Zoster Virus, Toxoplasma antigens. Normal donor values ranged from 600 to 1200/10e6 plated cells. Control transplant recipients acquired such pathogen-specifi c T cell responses from month 3 onwards in frequencies that ranged from 50 to 250/10e6 plated cells. In patients who received T alfa 1, pathogen-specifi c T cells appeared as early as 1 month after transplant in signifi cantly higher frequencies which soon ranged from 250 to 500/10e6 plated cells. The cumulative incidence of non-relapse mortality (NRM) (mainly infection-related) was 33% in controls vs 10% in Thymosin-treated patients (p=0.02). As a consequence of the improved TRM, Event-Free Survival was better in Thymosin treated patients (41% vs 21% in controls; p=0.06). Finally, multivariable analyses that included diagnoses, disease status at transplant, conditioning regimen and donor lymphocyte infusions showed Thymosin treatment was a signifi cant independent factor predicting a lower incidence of NRM (p=0.04) which tended to provide better survival (p=0.1). In HLA-haploidentical stem cell transplantation we showed adoptive immunotherapy with naturally occurring T regulatory cells (nTregs) followed by conventional T cells (Tcons) preven-ted acute and chronic graft-versus-host-disease (GvHD), favoured lymphoid reconstitution and immunity against pathogens (Di Ianni et al., Blood 2011). A major concern is whether a graftversus-leukemia (GvL) effect is maintained since FoxP3+Tregs can also suppress immune response to tumour. 49 patients (39 AML, 8 ALL, 1 Biphenotipic AL, 1 NHL) received a conditioning regimen including TBI, thiotepa, fl udarabine and cyclophosphamide (28 patients in the 1st clinical trial) or alemtuzumab (21 patients in the 2nd clinical trial), followed by an infusion of nTregs (day -4; mean 2.6x10 6 /kg ±0.8 SD; FoxP3+ cells 92% ±8 SD) and on day 0 CD34+ cells (mean 9.8x10 6 /kg ±3.8 SD) and Tcons (mean 1.1x10 6 /kg ±0.5 SD). No post-transplant GvHD prophylaxis was given. 47/49 patients achieved a rapid and sustained full donor-type engraftment. 5/45 (11%) valuable patients developed acute GvHD. CD4+ and CD8+ peripheral blood counts reached 200/μL on days 70 (range 37-189 days) and 51 (range 21-147) respectively. We observed a rapid development of a wide T-cell repertoire and high frequencies of specifi c CD4+ and CD8+ clones for opportunistic pathogens. Treg immunotherapy did not compromise posttransplant generation of donor-vs-recipient alloreactive natural killer (NK) cell repertoires which was faster than controls and with an enhanced alloreactivity against KIR-ligand mismatched targets. The cumulative incidence of relapse was 0.076 as only 3 high risk patients have relapsed to date at a median follow-up of 16 months (range 4-37). Preservation of alloreactive NK cell repertoires plays a key role in controlling leukaemia relapse. However more than 50% of our patients were transplanted from a non NK alloreactive donors, so we can hypothesize that high numbers of Tcons in the absence of post-transplant immunosuppression will also exert a GvL effect. Objective: Mixed chimerism (MC) after stem cell transplantation (SCT) for acute lymphoblastic leukemia (ALL) is known to predict relapse of the original disease in most cases. Immunotherapy (IT), including immunosuppressive withdrawal and donor lymphocytes infusion (DLI), is a possible strategy to prevent relapse in patients (pts) with MC. The aim of the present study is to analyse the outcome of a pediatric cohort of pts receiving SCT for ALL according to their chimerism status. Moreover, the impact of IT in pts with MC and the relationship between MC and minimal residual disease (MRD) after SCT are described. Patients and Methods: 68 pts diagnosed with ALL received SCT from matched related or unrelated donors after myeloablative conditioning (TBI and VP16) at the University Children's Hospital in Frankfurt, between Jan 1st, 2005 and Dec 31st, 2010. Chimerism analysis was performed using STR based quantitative PCR. Investigations were performed weekly until day 200 and monthly thereafter. 30 pts (44.2%) presented MC after day + 30 from SCT. All children with MC were pre-emptively treated by immunosuppression withdrawal, eventually followed by DLI administration. Results: The whole cohort had a probability of event free survival (pEFS) of 75% and a probability of overall survival (pOS) of 80.9%. Transplant related mortality in the all cohort was 13.2%. Pts with persistent full donor chimerism had a pEFS of 92.1% while those with MC achieved a pEFS of 40.6 (p=0.001). All patients with MC could be pre-emptively treated; 19/30 patients (=63%) converted to complete chimerism and remained in CR. 10 out of 22 relapsed after withdrawal of immunosuppression. 8 pts received further DLI. 7 out of these 8 pts responded and remain in CR whereas 1 patient relapsed. No severe GvHD and no cytopenia were reported after DLI administration. MRD data from bone marrow samples after SCT were available for 44 pts. 25 pts were MRD negative, among which 1 relapsed with isolated extra-medullary disease while exhibiting full donor chimerism. 8 pts had low MRD levels (<1x10 -4 ) and none of them relapsed. 11 pts were MRD positive (≥1x10 -4 ) and exhibited MC, 9 of them relapsed, in 7 cases MRD positivity preceded MC detection in bone marrow. Conclusions: Consequent analysis of chimerism offers the possibility to identify pts at highest relapse risk. Even though the numbers of this cohort are small to draw fi rm conclusions, preemptive immunotherapy may positively infl uence outcome. Despite profound T-cell depletion by in vitro (20 mg) and in vivo (30 mg) Alemtuzumab (ALT) for acute GvHD prevention, no early CMV disease and a low incidence of 2% of EBV associated PTLD was seen in a cohort of 41 patients receiving allogeneic stem cell transplantation (alloSCT) after a reduced intensity, fl udarabin and busulphan based conditioning regimen. We hypothesized that virus specifi c memory T cells are capable of surviving circulating ALT. In this study we analyzed the recovery of CD4 and CD8 T cells and CMV and EBV specifi c memory cells in relation to circulating antibody levels. Three weeks after alloSCT, low numbers of circulating CD8 and CD4 T cells were found (median 0.6 and 0.3 M/L, respectively). These numbers increased after 6 weeks to 27.6 CD4 and 9.4 CD8 T cells M/L, and after 9 weeks to 61.5 CD4 and 19.0 CD8 T cells M/L. At these time points circulating ALT was found in most patients (median levels 0.39, 0.05 and 0.02 ug/ml, after 3, 6 and 9 weeks, respectively). Six weeks after alloSCT in 28/29 patients circulating CD4 T cells lacking CD52 expression were present (median 69% of CD4 cells, range 0-93%) and in 27/29 patients CD52 negative CD8 T cells were found (median 46% of CD8 cells, range 0-99%). Based on concurrent FLAER negativity, we demonstrated that the CD52 negative T cells were GPI anchor defi cient. Because we also detected signifi cant numbers of circulating CD8 T cells with CD52 expression despite functional levels of circulating ALT, we examined CD52 expression levels and in-vitro sensitivity to ALT-mediated complement-dependent cytotoxicity (CDC) of CD4 and CD8 T cells. CD52 expression on CD8 T cells was lower than on CD4 T cells, resulting in relative protection of CD52 positive CD8 T cells against ALTmediated CDC. Using tetramer staining, cytotoxicity assays and analysis of cytokine production, we demonstrated that both CD52 negative and CD52 positive CMV and EBV specifi c CD8 T cells were functional. In conclusion, after TCD using low dose ALT, virus specifi c memory T cells are able to survive due to various escape mechanisms including CD52 negative GPI defi cient escape variants in both the CD4 and CD8 compartments. Moreover, a proportion of CD8 T cells was capable of escaping from ALTinduced CDC due to lower expression of CD52 as compared to CD4 T cells. All escape variants contained fully functional virus-specifi c T cells, which may explain the low incidence of viral complications after low dose ALT-induced TCD. The infl uence of donor parity on the outcome of HLA matched hematopoietic stem cell transplantation (HSCT) is poorly understood. Through exchange of hematopoietic and nonhematopoietic cells, pregnancy frequently leads to the stable engraftment of small numbers of fetal cells; these blood-borne or tissue-residing microchimeric cells may be accompanied by functionally different types of maternal alloreactive T cells, including minor Histocompatibility antigen-specifi c cytotoxic (T-CTL) or regulatory T cells (T-REG). We investigated whether male microchimerism detectable in peripheral blood cells collected after the delivery of male offspring is correlated with HY-specifi c regulatory T cells activity in 57 healthy women with a well documented obstetric and family history. A real-time quantitative polymerase chain reaction was used to determine the presence of a Y chromosome-specifi c multicopy locus (DYS-1) in 5 different highly purifi ed hematopoietic cell types, i.e. T cells, B cells, monocytes, myeloid dendritic cells and/or granulocytes. The trans-vivo Delayed Type Hypersensitivity reaction was used to determine the presence or absence of HY peptide-specifi c T-REG in the same peripheral blood sample. Y chromosome positive leucocytes were detected in 16 out of 35 women with male offspring, but also in 4 out of 8 donors with female offspring and in 4 out of 14 nulliparous females. T-REG specifi c for HY peptides presented by HLA class II molecules were detected in 10 out of 24 women; their HLA class I-restricted counterparts were identifi ed in 17 out of 32 donors analyzed. These presumably CD4 or CD8 positive T-REG were detected irrespective of the presence or absence of circulating male microchimeric cells. Our study indicates that HY-specifi c T-REG appear to be a common occurrence in healthy women, irrespective of their pregnancy history. It remains to be studied whether blood-borne male cells, tissue-residing male cells or both are necessary for the induction and maintenance of this male-specifi c tolerance. Introduction: The human leukocyte antigen-G (HLA-G) is an important tolerogeneic molecule, initially shown to mediate maternal tolerance of the "semi-allogeneic" fetus during pregnancy. We have previously identifi ed a small subset of HLA-G+ T cells and a sizable population of HLA-G+ monocytic cells (CD14+/CD16low/HLA-DRlow) circulating in peripheral blood (PB) of healthy individuals with potent immunoregulatory and immunotolerogeneic properties. Objectives: To investigate the evolution of naturally occurring HLA-G+ myeloid (CD14+) and lymphoid (CD3+) populations after allogeneic hematopoietic cell transplantation (HCT). Methods: PB from 27 unselected transplant patients were analyzed for the presence of HLA-G+ cells by fl ow cytometry. FACS sorted HLA-G+ cells were used as third party cells in mixed lymphocyte cultures (MLC) and CFSE cell proliferation assays to analyze their immunosuppressive properties. Results: HLA-G+ cells in PB of patients after HCT (8.1%±0.8) was signifi cantly increased, as compared to healthy donors (mean 2.5-fold, p<0.0001) or to respective percentages before HCT (mean 2.8-fold, p=0.01). Interestingly, there was a more pronounced expansion of CD3+HLA-G+ cells (5.4 fold increase) as compared to that of CD14+HLA-G+ cells (2.2 fold increase). S32 CD3+HLA-G+ cells were at median 8.7%±0.8 in transplant patients, as compared to both healthy subjects (1.6±0.4, p<0.0001) or pre-transplantation values (1.1±0.2, p=0.005). By delineating patients according to the time after transplantation, we found that HLA-G+ cells, both CD3+ and CD14+, were detected early after transplantation (d<100) and continued to be high at later time periods (100<d<365, d>365). In allogeneic MLC assays, FACS-sorted CD3+HLA-G+ and CD14+ HLA-G+ cells from transplanted patients were able to suppress T-cell responses (p=0.002 and p=0.0005, respectively). Although in our study we couldn't fi nd any correlation between graft-versus-host disease (GvHD) and PB HLA-G+ cells, interestingly, HLA-G+ cells were detected in signifi cant amounts in skin biopsies and their frequency correlated with GvHD (p=0.03). Conclusions: Overall, we believe that HLA-G+ cells belong to the normal repertoire of the immune system and after allo-HCT increase acting as a shield against immune aggression, especially in target organs. Since these cells are easy to isolate from blood, naturally occurring HLA-G+ cells may constitute a novel strategy for adoptive cell immunosuppressive therapy. Introduction: Wilms Tumor gene (WT1) overexpression is described in several oncological diseases including acute myeloid (AML) leukemias. The majority of AML patients don't have a suitable specifi c molecular marker for monitoring minimal residual disease (MRD). Quantifi cation of WT1 in bone marrow samples can be useful as a marker of MRD after allogeneic stem cell transplantation (SCT) and can predict AML relapse. Methods and Results: We have evaluated, sequentially and using a quantitative RT-PCR technique (Leukemia NET method), WT1 expression in 50 consecutive AML patients that overexpressed WT1 at diagnosis and that underwent allogeneic SCT at our centre. The cDNA level of WT-1 was detected in bone marrow samples at diagnosis at the time of transplant and after the allogeneic SCT. Samples of diagnosis showed high WT1 expression levels in all cases with a mean of 5570 (SD 4055) copies of WT1/10000 Abl, median 4600 (range 658-23913) copies WT1/10000 Abl. At transplant 34 pts (68%) were in complete cytologic remission (CcR) and 16 (32%) had refractory or relapsed AML. Bone marrow samples from pts in CcR at BMT showed signifi cantly lower WT1 expression levels (mean 88±130), compared to the samples from pts with relapsed or refractory disease (mean 5727±4265) (P < 0,001). After BMT a rapid decline of WT1 expression levels was observed in all pts that achieved and/or maintained a condition of CcR, especially in those that were in CcR at SCT. After a median follow up of 11 mths from transplant, 10 out 50 pts relapsed (20 %) and all of them had high expression levels of WT1 before the cytological relapse. Three of these pts were successfully reinduced with DLI ± chemotherapy with a rapid reduction of WT1 levels. Conclusions: 1) In our experience there is a complete concordance between WT1 expression levels (measured by quantitative RT-PCR) and status of AML before and after SCT. 2) Our study confi rms that longitudinal quantitative evaluation of WT1 after SCT may be useful as a non-specifi c leukemia marker (NSLM) for monitoring MRD and as a predictor of AML relapse.

3) Based on these results cases with an increase of WT1 levels after SCT and without GVHD should be candidate to DLI and/or discontinuation of immunosuppressive therapy.

Scientifi c activity report 1. Introduction: The Severe Aplastic Anaemia Working Party (SAAWP) reports on AA and other rare acquired and inherited bone marrow failure disorders (see Figure 1 ). The AA database is the only "disease specifi c" database within the EBMT as data are not only collected on patients receiving transplant but also other forms of therapy, including immunosuppressive therapy. Aplastic Anemia's occurrence in childhood is estimated lower than in older ages. Since the data base of the SAA of the EBMT contains data on a relevant number of children with acquired AA (AAA) we sought to investigate the outcome of this disease in the age between 0 and 12 years. To this end we looked into the SAAWP EBMT database and evaluated the overall survival (OS) and event free survival (EFS) in 776 patients aged 0-12 years (median fup 2.4 min 0-max 11.3 years) in relation to the period of diagnosis (1990) (1991) (1992) (1993) (1994) (1995) (1996) (1997) (1998) (1999) (2000) (2001) (2002) (2003) (2004) (2005) (2006) (2007) (2008) (2009) and the received therapy (only IS, fi rst line HSCT and HSCT after failed IS).

The 3 year probability OS and EFS were 86% and 83% respectively which reproduced those seen in period 1990-1990 (87% OS and 85% EFS) leading to the speculation that both diagnostic and therapeutic tools, including supportive treatment, were at a level hardly susceptible of further improvement. Focusing on post 2000, 3 year OS after fi rst line IS therapy was 97% vs 86 % for fi rst line HSCT vs 83% for HSCT performed after having failed IS (p ns). EFS at 3 years after fi rst line IS was 86 % vs 83 % for fi rst line HSCT vs 81% for HSCT performed after having failed IS (p ns). Since fi rst line HSCTs included both matched sibling donor (MSD) and alternative donor transplants (ALT i.e. all other type of donors except from matched sibling donor), we compared OS and EFS of these procedures and found that MSD did signifi cantly better than ALT (OS at 3 yrs 91% vs 78% p0.001; EFS 88% vs 76%, p0.001) donor HSCT.

When we compared the following 3 fi rst line therapy: IS, MSD and ALT donor transplants along with HSCT done after having failed IS, we found an OS advantage for IS (97%) over MSD HSCT (91%) over ALT HSCT (78%), and that OS for HSCT post IS was (83%) (p for the 4 comparisons 0.0001). EFS after MSD HSCT was 88%, after IS wad 86% after ALT HSCT was 76% whereas HSCT post failed provided and EFS of 81% (p for the 4 comparisons 0.01). Taken together these data indicate that front line IS is an excellent therapy for children with AAA, and that in case of failure of this treatment HCST is a very good salvage option. Given that HSCT is able to restore haematopoiesis more completely and more durably than IS, if a matched donor is available the preference of front line HSCT over front line IS can be justifi ed. 

Allogeneic hematopoietic stem cell transplantation (HSCT) is one of the most important treatment options for patients (pts) with high-risk hematological diseases who have an HLA-identical sibling or matched alternative donors. In order to offer the transplant option to patients lacking a suitable matched donor, alternative options, such as unrelated cord blood transplantation (UCBT) or family haploidentical stem cell transplantation (haplo-HSCT), were developed in the last 15 years with promising results. Information is scarce on outcomes after haplo-HSCT or UCBT for pts with severe aplastic anemia (SAA): registry surveys will better clarify the benefi t of alternative transplantation in this setting. On behalf of the SAAWP and PDWP we here report the European Blood and Marrow Transplantation Group (EBMT) survey on outcomes after haplo-HSCT for pts with SAA. We retrospectively analyzed 73 pts (45/73 male, median age at SAA diagnosis 9.78 years, at transplant 12 years, 18/73 adults) with SAA who received a haplo-SCT in 48 centers between 1976 and 2011 -57/73 between 2000 and 2011. Thirty-one centers performed only 1 haplo-SCT, 13 centers 2 haplo-SCT, 3 centers 3 haplo-SCT, 1 center 7 haplo-SCT. Haplo-SCT was used as the fi rst transplant option for 59 pts, as the second for 12 pts and 2 pts received haplo-SCT as third or fourth options. The median time from SAA diagnosis to haplo-SCT was 324 days (range 10-5373). The conditioning was mostly non-myeloablative; TBI was utilized in 16 cases. Ex vivo T cell manipulation was performed in 44 cases. Neutrophil engraftment was documented in 58% of pts at a median time of +14 days (range 3-45). Graft loss was seen in three pts. Six out of 48 pts with documented engraftment experienced acute GvHD grade I, 5 grade II, 4 grade III 4 grade IV and 1 with unknown grade. Twenty-nine pts are alive (median follow-up 1015 days, range 0-6589) -6/29 received transplant before 2000, 23/29 after 2000. Median time to exitus was 93 days (range 4-2710) -main causes of death being infections (24/44), GvHD (6/44), hemorrhage (6/44). At a median follow-up of 33 months the estimated probability of 3-year overall survival (OS) was 37%. There was a trend for improved OS for pts transplanted in centrs who performed at least 3 haplo-HSCT compared to other sites (58% versus 30%, p=0.0578). No statistically difference was documented for transplants performed before and after 1999, year of widespread diffusion of cell-selection devices.

To our knowledge, our survey provides information on haplo-HSCT for pts with SAA for the fi rst time. Development of new strategies on haplo-HSCT will further improve outcome results in this setting.

The care of the human immunodefi ciency virus (HIV)-infected patients has changed dramatically since the widespread use of combined antiretroviral therapy (cART) in 1996. Current cART regimens suppress viral replication, provide signifi cant immune reconstitution, and have resulted in a substantial and dramatic decrease in acquired immunodefi ciency syndrome (AIDS)related complications and deaths in both adults and children. Those patients are now experiencing long-term follow-up and notably auto immune diseases have been described in this context. Severe Aplastic Anemia (SAA) is an immune-mediated disorder in most cases with active destruction of hematopoietic cells by effector T lymphocytes. The association of HIV-infected patients and SAA is exceptional. There are only very few documented cases of allogeneic stem cell transplantation (SCT) or immunosuppressive therapy in HIV patients with acquired SAA. The aim of this study is to collect all cases of HIV-infected patients who developed an acquired SAA to better characterize this entity and eventually its specifi c treatment. ). In multivariate Cox analysis, the 7-year nonrelapse mortality incidence (NRM) was infl uenced by the conditioning regimen in a time-varying manner: up to day +100, the cause-specifi c hazard for NRM was signifi cantly lower for the RIC subgroups (hazard ratio [HR], 4.3; 95% confi dence interval [95% CI], 4.3-11.1; P<0.01); but a land-mark analysis beyond day +100 showed that the 7-year cause-specifi c hazard for NRM was increased after NMA conditioning (HR, 2.4; 95% CI, 1.2-10.4; P=0.04). The 7-year relapse was also infl uenced by the conditioning regimen in a time-varying manner: up to month +12, the relapse incidence was signifi cantly higher after NMA conditioning (HR, 2.4; 95% CI, 1.2-10.4; P=0.04 ); but a land-mark analysis beyond +12 months showed that the 7-year relapse incidence was not infl uenced by the conditioning intensity. The 7-year probabili-ties of progression-free and overall survival were signifi cantly lower only in the NMA subgroup (Multivariate P<0.01; and 0.01, respectively). In conclusion, the similar 7-year survival after IntermRIC and SMC is encouraging, since IntermRIC patients were older (age >50 years in 75% vs. 15%, respectively, P<0.001) and had higher EBMT risk scores pretransplantation. However, 7-year outcomes were worse in NMA recipients. Our data suggest that prospective studies comparing RIC regimens (especially IntermRIC) with ConvMC are appropriate in patients with AML and RAEB who are in a non-advanced disease status and are eligible for either convMC or RIC regimens. Introduction: Stem cell transplantation will continue to be a treatment option for patients with chronic myeloid leukaemia, despite the introduction of tyrosine kinase inhibitors. However, many patients will have received prior therapy with TKIs, including Nilotinib or Dasatinib at the time of allogeneic stem cell transplantation. While the use of Imatinib prior to stem cell transplantation seems to have no adverse impact on the outcome of allogeneic stem cell transplantation little is known on the impact of prior use of second generation TK inhibitors. Objective: To evaluate the infl uence of prior treatment with second generation tyrosine-kinase inhibitors (TKIs) in CML patients on engraftment rates and treatment-related mortality during and after treatment with allo-HSCT. Secondary objectives: to evaluate the effect of prior 2nd generation TKIs on treatment-related toxicity after treatment with allo-HSCT in CML (i.e. incidence and severity of acute and chronic GVHD, SOS (VOD) of the liver, pulmonary complications, and causes of treatmentrelated mortality); to evaluate overall and disease-free survival; and to evaluate the relapse rate. Methodology: All transplant-related data will be collected prospectively, whereas data on prior TKI therapies are recorded retrospectively. The study started at the beginning of 2010. A total of 450 patients are envisaged for analysis. Today, 68 centres have registered for participation. Together they expect to be able to submit 475 patients for this study, of which 163 patients have been included so far. Transplant dates range from November 2009 to December 2011. For two patients, the allotransplant was not the fi rst transplant. Preliminary results: Data entry has been done for 128 patients. In a very preliminary analysis of those 128 patients, 4 patients showed a graft failure. 2/3 of the patients were transplanted in CP and 73% were transplanted beyond 12 mo after diagnosis. The median time from diagnosis to transplant was 21 months and the median age at transplant was 47 years (18-67y). Sixtyseven percent of pts were male and 33% female. The majority (55%) received transplants from matched unrelated volunteer donors. At the CLWP session in Geneva all data received by that date will be presented. Centres that are interested in participating in this study can contact the EBMT Data Offi ce in Leiden via clwpebmt@lumc.nl. Data quality is a general concern in the context of clinical studies. With respect to CLL a double challenge exists: First, current data requested in the disease classifi cation sheet in minimum essential data A (MED A) are insuffi cient to describe the indication for allogeneic HCT. This is because information on cytogenetics and pre-treatment (especially purine-analogue refractoriness) has not been defi ned as minimal essential data in past times. MED B data is only sparse in the database. Second, missing data hinder the analysis of important clinical aspects. This double challenge is addressed in three ways: 1) Prospective non-interventional studies shall guarantee better future data quality through limited MED B information and active data request in small disease entities like 17p-CLL, T-PLL or Richter›s syndrome. 2) A data quality initiative based on a reduced MED B form will be launched for fi rst allogeneic transplants dating back until 2000.

3) The disease classifi cation sheet in MED A CLL will be redefi ned together with CIBMTR and the amount of information requested in MED B will be reduced.

Hematopoietic stem cell transplantation (HSCT) is a constantly evolving procedure. Due to many improvements its safety is increasing, which translates into lower mortality rates and expansion of new indications. A number of novel approaches such as reduced intensity or non-myeloablative conditioning regimens have made HSCT feasible even in elderly patients with comorbidities. Therefore some patients who would not have been considered as candidates for HSCT before undergo this procedure nowadays. The number of surviving patients who have undergone a solid organ transplantation (SOT) is growing. Similar to healthy people, they may also suffer from a number of hematopoietic disorders which are indications for HSCT. Moreover, the probability of developing such disorders may be increased due to immunosuppression. Despite the increased risk associated with a HSCT procedure in these patients, apparently a signifi cant number of patients have undergone autoSCT or alloSCT following SOT in Europe. We decided to perform a retrospective study within the Complications Subcommittee of the Chronic Leukemia Working Party in order to document the outcome of HSCT in these patients and to investigate how the HSCT procedure has affected the functions of transplanted organ. All patients of all ages who have undergone alloSCT or autoSCT following SOT can be included in this study. The centers participating in this study are asked to fi ll in the MED-A form and a shortened MED-B form for the HSCT (prefi lled based on already reported data). In addition, a specifi c MED-C form of the SOT (separate forms for heart, kidney, liver and other organ transplants) has to be fi lled in. The centers in which recipients of solid organ transplant have ever undergone either auto-or alloSCT are encouraged to include their patients. Centres that are interested in participation should contact the Leiden Data Offi ce via clwpebmt@lumc.nl.

A non-interventional prospective study to evaluate the effect of transfusions and iron toxicity in patients with myelodysplastic syndromes treated with allogeneic stem cell transplantation E. Cremers, A. van Objectives: To prospectively evaluate the effect of iron toxicity on treatment outcome after myeloablative (MAC) and reduced intensity (RIC) allogeneic SCT, in previously untreated adults with high risk MDS. This report describes correlations between pre-transplant parameters at time of SCT and 100 days after SCT. Methods: The data collection by the CLWP of the EBMT, started in March 2010. This survey collected prospectively data on diagnosis, SCT, iron parameters, transfusions, comorbidity and treatment outcome. A composite iron score based on serum iron levels, transferrin saturation and serum ferritin levels is used to defi ne a high risk group with serum ferritin >1,500 ng/ml, and/or transferrin saturation >80% and/or serum iron >200 mg/l. Results: From the 134 patients registered so far, additional data is available on 94 patients. Eighty-three patients had MDS and 9 patients MPS, of whom 59 were male. WHO-classifi cations at time of SCT are 7 patients with RA/RARS, 8 with RCMD, 21 with RAEB-1, 28 with RAEB-2 and 1 with sAML. Forty-one patients received a MAC and 53 a RIC regime. After conditioning 31 patients received stem cells from a sibling, 2 from a mismatched relative and 61 patients from an unrelated donor. Time between diagnosis and SCT was <6 months for 36 patients and >6 months for 58 patients. In 41 patients GvHD occurred (grade I-IV). Variables analyzed in 69 patients (day 100 data available) were iron load at day of SCT and +100 days, amount of red blood cell transfusions prior to SCT and between SCT and +100 days and occurrence of GvHD. We observed a strong positive correlation between ferritin levels pre-SCT and +100 days (r=0.64; p= 0.001). Similarly, a positive correlation exists between the iron score pre-SCT and +100 days (p=0.006; OR=18.00). Despite the preliminary nature of this analysis there was a strong indication that the increased transfusion burden prior to SCT is positively correlated with the occurrence of aGvHD (p=0.039; OR=4.13). So far, we found no correlation between transfusion burden pre-SCT and +100 days (Spearman r=0.004; p=0.98) and no correlation between transfusion burden and iron score at time of SCT (p=0.226; OR=2.30). The relation between mortality before day 100 and transfusion load was not signifi cant (p=0.145; OR=3.73). Conclusion: this prospective observational study is running well. Preliminary data indicate that we may expect suffi cient data to achieve the primary objectives of this study.

Graft versus host disease (GVHD) remains one of the main obstacles to broader application of alloge-neic haematopoietic stem cell transplantation (HSCT). It can be prevented with selected methods, but often at the expense of an increased risk of relapse and rejection. To advance our knowledge and de-fi ne optimal approaches, correct information on the value of the different strategies and methods in GVHD prevention and treatment is essential. A recent survey of the EBMT documented once again that GVHD prophylaxis and treatment strategies and techniques used in allogeneic stem cell transplantation are very heterogenous. Even in superfi cially similar protocols, the details in practical work differ con-siderably. There are no widely accepted standard practices for the prevention and management of GVHD. Centre effects have been demonstrated in transplantation outcomes. They are attributed to centre size, macroeconomic aspects and JACIE accreditation status. Differences in outcome might as well be, at least partly, due to differences in treatment strategies and transplantation techniques. A prerequisite to learn about this is to have a common comparator. For this purpose a current "standard" practice needs to be defi ned. These guidelines could lead towards a more uniform practice and they could be used as baseline when evaluating the impact of different components of the transplant proc-ess. Based on available literature and current practice, a joint consensus committee of the EBMT and ELN is working on a proposal for the implementation of minimal EBMT/ELN standards in this fi eld. Identifi cation of NK cells with high antileukemic activity will help to improve targeted cellular therapy strategies, especially in the context of haploidentical stem cell transplantation. We screened the NK cells of mismatched healthy donors with regards to their reactivity against ALL blasts to identify such NK cell subsets. Method: NK cells were isolated from 4 healthy donors with CD56+ MACS beads and incubated with 5 different primary pediatric ALL blasts in a 3 hour CD107a assay. Analysis of the NK cells considered the main KIR receptors CD158a, CD158b and CD158e, as well as ILT2, NKG2a and activating receptors NKG2D, NKp30, NKp44 and NKp46. Blasts were analyzed for NKG2D ligands MIC A and B, ULBP1-3 and adhesion molecules CD11a, CD50, CD54, CD58, CD112 and CD155. Degranulation controls were performed with K562 as target cells. Results: Degranulation of the whole NK cell compartment displayed a wide variability (range of CD107a+ cells: 0.6-11.3%) and was infl uenced by type of donor and type of ALL blasts and was not predictable according to common alloreactivity models in this data set. A signifi cant increase (p<0.001, n=20) of the CD107a+ fraction was reached by overnight IL2 stimulation of NK cells. In contrast to the entire NK cell compartment, the subset of KIR-NK cells revealed a high and stable indirect cytotoxicity in terms of CD107a expression throughout all experiments, whereas the KIR+ NK cell subset by trend followed the receptor-ligand model. ALL blasts showed no or low expression of NKG2D ligands: MIC A (0-2.2%), B (0-1.6%), ULBP1-3 (0-1.6%). Adhesion molecules were diversely expressed on the blasts: CD50 (65-99%), CD54 (48-77%), CD58 (12-98%), CD112 (12-63%), CD11a (2.8-34%), CD155 (0-2.2%). There was no correlation of NK cell degranulation with expression of adhesion molecules or NKG2D ligands on the blasts. Conclusion: Our aim was to identify NK cell subsets with high activity against pediatric ALL blasts for potential use in cellular therapies. The KIR-NK cell subset was found to show high and stable degranulation disposition, independent from the target and from the NK cell donor. Moreover donors displayed an individual non-predictable variability in the level of overall NK cell degranulation. This suggests including functional analyses of NK cells in donor selection strategies. Further investigations are necessary to confi rm these data. Infection with human adenoviruses (ADV) can cause lifethreatening infections in pts after allo-SCT and represents a major reason for transplant related mortality (TRM, in historical cohorts after haplo-SCT up to 30%). Graft manipulation by T-cell depletion and the use of immunosuppressives affect immune reconstitution and can extend duration of impairment of immunity after SCT. Delayed reconstitution of immunity increases the risk for viral infections. New preparative regimens such as reduced intensity conditioning (RIC) followed by CD3/19 depleted PBSC transplant (3/19depl-PBSCT) try to overcome these limitations.

To substantiate this we studied the incidence (inc.) and mortality of ADV infection after ped. allo-SCT overall and as a function of graft. 210 transplants have been performed in 200 ped. pts in Frankfurt between '05 and '11. Donor source was 3/19depl-PBSC (n=95) and BM (n=115). Median follow up was 13.9 mths. Weekly post-transpl. ADV-screening was conducted by qPCR in plasma, throats swabs and faeces. Inc. of ADV detection in any compartment at any time-point was 40.0% and sign. higher after 3/19depl-PBSCT compared to BMT; 49.5% vs 32.2% (p=0.016). Inc. was 26.7% ≤d60, 20.6% d61 to 100 and 14.0% >d100. Cumulative inc. of 3y TRM (CI 3y TRM) due to ADV disease was 4.6% (3/19depl-PBSCT 6.9%, BMT 3.0%; n. sign.). In detail, 14.5% of pts (29/200) died due to non relapse mortality. Amongst these 29 pts, 16 (55.2%) were ADV pos.. ADV positivity was sign. more frequent after 3/19depl-PBSCT compared to BMT (84.6% vs 31.3%; p=0.008). Mortality due to ADV disease among TRM pts was 27.6% (3/19depl-PBSCT 38.5%, BMT 18.8%; n. sign.). Overall CI 3y TRM was sign. higher in ADV pos. (n=84) vs neg. (n=126) pts (28.7% vs 13.4%, p=0.03). Taken together, ADV was detected in every second pt after 3/19depl-PBSCT. Although the CI 3y TRM due to ADV disease was low (4.6%), post-transpl. ADV detection sign. increased the risk for subsequent TRM. In summary, ADV represents a severe threat to survival, particularly in pts after 3/19depl-PBSCT. Apparently, more rapid reconstitution of immunity in these pts compared to historical mega-dose CD34 transplants resulted in low mortality due to ADV disease in this high-risk group. Nevertheless, consequent post-transpl. screening for ADV is indispensable. Early and effective treatment eg with Cidofovir or DLI (ADV-specifi c donor T-cells, cytokine-induced killer cells (CIK)) is important for further reduction of ADV-related mortality. 

A retrospective analysis was performed in children and adolescents who received treosulfan for conditioning before allogeneic or autologous HSCT to identify possible dose related toxicity and determine the incidence of engraftment, treatment related mortality (TRM), overall survival (OS) and event free survival (EFS). Patients below 18 years registered in the EBMT database who underwent HSCT between January 2005 and July 2010 for malignant or non malignant disease were eligible. 843 patients met the inclusion criteria and 75% could be included into the analysis (521 allogeneic, 83 autologous). In the group of patients with allogeneic HSCT, 41 were below the age of 6 months and 65 between 6-months and 1 year. 314 were between 1 and 12 years of age and 101 above 12 years. 165 patients suffered from malignancies, 356 patients were transplanted for non malignant diseases. 437 underwent a fi rst HSCT, 87 had a following HSCT. For allogeneic HSCT, the median treosulfan dose was 42 g/m 2 (12-46). 24 patients (5%) received less then 33 g/m 2 . 153 patients (29%) received between 33 and 39 g/m 2 . The majority of patients received a treosulfan dose between 39 and 45 mg/m 2 (332 patients, 62%). 13 patients (2%) received more than 45 mg/m 2 . Results: The most common grade 3 or 4 toxicities were diarrhoea (24%), stomatitis (22%), and SGOT (25%). We did not fi nd any positive correlation between the given dose and the grade 3 or 4 toxicity in the dose range used within the EBMT centres. Acute graft-versus-host disease (GvHD) of grade 3 or 4 occurred in 10% of the patients and the rate of limited and extended chronic GvHD was 13% and 6%, respectively. OS and EFS correlated signifi cantly with diagnoses: 3-year OS was 51% for ALL, 46% for AML, 80% for inherited disorders, and 93% for haemoglobinopathies. OS was signifi cantly higher in patients who received treosulfan for fi rst HSCT. 3-year EFS was 39% for ALL and 40% for AML. In the given dose range, we did not fi nd an association between dose and GvHD, OS, EFS, relapse incidence and TRM. Conclusion: We identifi ed no dose related toxicity, graft failure, GvHD, TRM, or relapse incidence in malignant diseases. These results emphasise the low toxicity profi le of treosulfan, even in heavily pre-treated children and adolescents and in patients undergoing a second HSCT. Data from the pediatric BFM 98 trial show that the majority of children with AML do not benefi t from allogeneic hematopoietic cell transplantation (HCT) in fi rst remission. Therefore, HCT was only recommended for patients achieving a second remission after fi rst relapse. Assuming a dismal prognosis, the potentially toxic triple alkylator regimen consisting of busulfan, cyclophosphamide, and melphalan (BuCyMel) was recommended, based on the practice of the EWOG-MDS 98 trial and the AIEOP AML 2002/02 study. Here we retrospectively analyze the interim data of children who underwent HCT after BuCyMel for AML in CR2 between 1998 and 2009 before the initiation of a prospective, controlled and centrally monitored AML-BFM HCT trial. Out of 152 identifi ed children transplanted in 17 German pediatric transplant centers after achieving CR2, complete data were available for 62 children after a preparative regimen with BuCy-Mel and for 54 receiving a different regimen. For the fi rst group 62 children were identifi ed for analysis. The median age at HCT was 11.9 years [3.0; 18.3]. GVHD-prophylaxis mostly consisted of CSA and short term MTX with or without ATG. Matched sibling donors were used for 23 of 116 analyzed recipients, the remaining received grafts from matched or partially mismatched unrelated donors. The overall survival at 5 years (5y OS) for children achieving CR2 was 61% (standard error 0.07%). Relapse mortality was 21% (13/62 children). Treatment related mortality reached 16% (10/62) and was caused by infection (4/10), GvHD (4/10) and multi organ failure of unknown cause (2/10). Within the same observational period 54 children received a HCT in CR2 using a regimen other than BuCyMel. 5y OS in this group was 35% (19/54 children, standard error 0.06%) with a TRM of 29% (16/62) and mortality from second relapse reaching 30% (17/62). Conclusion: Even taking into account possible selection bias in this retrospective analysis, HCT in CR2 using ByCyMel resulted in a respectable overall survival of 61%. TRM in this cohort was higher than the one reported by the AIEOP group for children in CR1 and seems to be lower than in EWOG-MDS 98. The prospective, controlled and centrally monitored trial AML SCT-BFM 2007 has started to recruit patients in January 2010. This trial aims to generate solid and valid outcome data for children with AML in CR2 or primary refractory disease to serve as a platform for the development of further transplant strategies. Adolescents are treated both by paediatric teams (group 1, defined as centres reporting a majority of pts <18 yrs, or combined centres, with a dedicated paediatric team) and by adult teams (group 2, defi ned as centres reporting pts below and above 18 yrs, without a dedicated paediatric team).The primary end-point of this study was to evaluate the impact of the type of centre, i.e., paediatric or adult, on outcome after HSCT from a MSD, or from a volunteer UD in adolescents with ALL in 2nd CR, defined as pts older than 14 yrs and younger than 18 yrs at HSCT, given allogeneic HSCT from 1995 to 2009 and reported to the EBMT registry. A total of 435 pts (67% males and 33% females in group 1, 70% males and 30% females in group 2, p=ns) were included in the study; 214 pts were transplanted in paediatric centres, whereas 221 pts were transplanted in adult centres. The 2 groups differed in terms of distribution of patient and transplant-related variables: the frequency of very early relapse in group 2 pts was approximately twice as high as in group 1(p=0,001); the median interval from diagnosis to transplant was shorter in group 2 (p=0.003), the median age at diagnosis was higher for group 2, (p=0.0001). A MSD was used in 40% of the pts in group 1 and in 61% of those in group 2; the remaining pts received an UD HSCT (p<0.001); the source of stem cells was bone marrow in 77% of pts in group 1 and in 58% of pts in group 2; the remaining pts received PBSCs (p<0,0001). The transplant conditioning regimen included TBI in 86% and 81% of group 1 and 2, respectively p=ns). As of September 2011,the median follow-up was 49 months.40% of the pts from group 1 and 42% of the pts in group 2 reported acute GVHD grade 2-4, respectively (p=ns). The 2 year cumulative incidence of chronic GvHD was 24% for group 1 and 32% for group 2, respectively (p=ns).

Five-year Kaplan-Meyer estimates of LFS were 43% for group 1, versus 31% for group 2, respectively (p=0.02).The 5 year cumulative incidence of TRM and relapse rate were 31% and 26% for group 1, 26% and 43% for group 2 (p=NS and p= 0,0006). Univariate analysis showed that year of HSCT <2000 (p=0,01), early relapse (P=0.0004) and diagnosis to transplant interval <31 mos (p<0.0001) were poor-risk factors for LFS. Multivariate analysis confi rmed for adolescents treated at paediatric centres a signifi cantly lower relapse rate and a trend towards better LFS, while no difference was observed in chronic GvHD. Methods: blood samples were collected from two children following the infusion of Bu after doses 1,2,3,5 and 9 (n=34). Then, 5 micro liters each of whole blood and plasma were spotted onto Whatman 903 DBS cards and dried at room temperature for about 30 minutes. The entire spots were cut and then analyzed by a validated LC-MS/MS method. Bu was also measured by established gas chromatography coupled to mass spectrometry (GC-MS) using plasma (n=13) to compare both methods. Results: A good correlation was observed between the Bu levels measured by DBS and DPS (r 2 = 0.95; slope=0.84). The Bu levels measured by DPS (r 2 = 0.92; slope=0.95) and DBS (r 2 = 0.91; slope= 0. 80) were correlated with those measured by standard GC-MS method. The levels estimated by DBS were less than those obtained by DPS and GC-MS methods. The hematocrit (Hct) values of two children were in the range of 25.6 to 30.3 %, suggesting the infl uence of Hct on the measured Bu levels by DBS sampling. Therefore,Hct value of the patient must be considered when whole blood sample is used to estimate the plasma levels of Bu. The plasma levels of Bu estimated using the formula "Plasma levels (analyte) = DBS levels analyte/ (1-hematocrit)" were higher than those obtained by DPS or GC-MS suggesting that a fraction of Bu binds to blood cells. Conclusion: This method can be applied with only 5 micro liters of whole blood for routine therapeutic drug monitoring of busulfan with accuracy and faster turnaround times. The use of capillary whole blood collected after a small fi nger prick is under evaluation for its feasibility and accuracy to monitor busulfan levels in pediatric patients. Objectives: In present study, the outcomes of childhood leukemia treated with haplo-HSCT using parent as donor were evaluated and the risk factors for survival were identifi ed. Methods: One hundred consecutive cases from June 2002 to June 2011 in our center were retrospectively analyzed. The median age of patients was 10 (3-14) years old. The diagnosis included AML (43%), ALL (42%), CML (10%), and mixed lineage leukemia (5%). Transplants in early, intermediate, and advanced disease were 47%, 32% and 21%. HLA mismatched at 1, 2, 3 loci was 7%, 22%, 71%, respectively. Transplants in father to son, father to daughter, mother to son, and mother to daughter were 40%, 7%, 27%, 26% respectively. Major clinical characteristics among these four arms were similar. All patients received unmanipulated combined marrow and peripheral blood stem cells for transplant after conditioning with busulfan and cyclophosphamide (Cy)/Cy and total body irradiation plus antithymocyte globulin. Results: Durable hematopoietic reconstitution was seen in 98% of recipients. Engraftment failure occurred in 2 cases of father to daughter transplants. One-hundred-day transplant-related mortality (TRM) was only 4%. The cumulative incidences of grade II to IV acute graft-versus-host disease (aGvHD) and chronic GvHD (cGvHD) were 33% and 22%, respectively. With the median follow-up of 32 (3-108) months, one-year and 5year overall survival (OS) rates for all patients were 75.9% and 68.5%. Five-year OS rates for patients in early, intermediate and advanced disease were 79.5%, 72.0% and 52.6% (p=0.022). Five-year OS of transplants in father to son, father to daughter, mother to son ,and mother to daughter were 81.5%, 36.2%, 63.4%, 66.8%, respectively (p=0.028). For the statistical symposium: There has been a recurring interest in modelling survival data in the presence of 'cured' individuals. 'Cure' is defi ned on a population level and means that there is no longer excess mortality compared to the general population. In a paediatric population the baseline mortality may be very close to zero. Here, given suffi cient follow-up, survival curves that reach a stable plateau indicate the presence of long-term survivors or cured individuals. This is quite common in paediatric oncology, where Cure-models have been developed to analyse failure time data with long-term survivors. In a more general setting, a particular type of Cure-model that incorporates the expected background mortality can be used. The aim of this work is to compare several types of Cure-model and contrast them to standard survival models. Their use, values, and limitations will be illustrated on data in the fi eld of paediatric stem-cell transplantation. Standard methods of survival analysis, like Cox-regression or the log-rank test, rely on the proportional hazards assumption. The results from Cox and Cure models are very similar when the hazard ratio is constant over time. However, with Cox-regression, it is not possible to distinguish between 1) patients who are cured and are long-term survivors and 2) patients who will die from their disease although a prolongation of survival can be achieved. With non-proportional hazards, the results from Coxmodels may be misleading and have no clear interpretation. In non-proportional hazard situations, Cox and Cure-models give different results. One can fi nd examples on non-proportional hazards in the comparison of treatments with different intensities. One treatment may produce more early toxic events but is more effi cient in preventing disease recurrence. Cure models separately investigate the impact of covariates on 1) the proportion of cured individuals and 2) their impact of failure times of non-cured individuals. The former is often the main interest in paediatric oncology. The estimates have a clear clinically meaningful interpretation congruent to binary regression and survival analysis. However, Cure-models fail when there are no long-term survivors or cures and the follow up is inadequate. Thus, Cure models can provide additional insights. They enable the investigation of predictors for long-term survival without relying on the proportional hazard assumption and should thus be considered more often.

Modelling the GvHD-GvL effect: statistical approaches incorporating clinical aspects L.C. de Wreede LUMC (Leiden, NL)

The immunotherapeutic potential of allogeneic haematopoietic stem cell transplants (SCT), summarized as 'Graft-versus-Leukemia'-effect (GvL), is essential to current practice, and especially exploited in the contexts of Donor Lymphocyte Infusions (DLI) and Reduced Intensity Conditioning. GvL is strongly associated with Graft-versus-Host-Disease (GvHD). Therefore, in clinical studies GvL can best be analysed through the impact of acute and chronic GvHD on relapse incidence (RI) and nonrelapse mortality (NRM). This talk will present several issues to take into account in such studies. Firstly, the relevance, potential and possible disadvantages of using data from the EBMT registry will briefl y be discussed. We will describe some methods to reduce missingness in these data. Secondly, several clinical considerations leading to different modelling choices will be discussed: how to separate aGvHD from cGvHD? How to model different degrees of GvHD? Is it possible to separate the impact of GvHD and immunosuppressive treatment? How should we classify patients who receive a second SCT or a DLI? How do we compare patients with different diseases, patient and treatment characteristics? Thirdly, we will present two statistical models and several refi nements to analyse these data. The fi rst of these is the Cox model for the cause-specifi c hazards for RI and NRM in which both aGvHD and cGvHD are entered as time-dependent variables. This can be extended to a model with time-dependent effects of the occurrence of GvHD or to a landmarking approach in which prediction starts only some time after SCT. The second model is the multi-state model, in which not only the impact of GvHD on RI and NRM can be studied, but also, e.g., the probability of being alive with cGvHD compared for patients with different baseline characteristics. Such a model also allows to study new outcomes, such as "time spent free from immunosuppression and without relapse" as a measure of treatment success. No statistical differences were detected between studied groups (single Tx, double Tx and triple Tx) acc. to age at Tx, gender, disease, donor type, CD34+/kg recipient body weight, cause of death, ANC and platelet recovery except for graft source (PB was used more frequent in the group with >1 Tx when compared to the group with single Tx: 81.5% vs. 62.9%; p=0.036). Both overall survival (OS) and disease-free survival (DFS) were comparable between the group with single Tx and the group with 2 or 3 Tx (OS: 59.2 ± 4.5 for single Tx, 44.2% ± 11.4 for 2nd Tx and 66.7% ± 19.2 for 3rd Tx; DFS: 62.1% ± 4.5 for single Tx, 43.6% ± 11.3 for 2nd Tx and 50.0% ± 20.4 for 3rd Tx). The frequency of aGvHD III-IV was comparable among groups (22.0% for single Tx, 0.0% for 2nd Tx and 16.7%; for 3rd Tx, p=0.173) as well as for chronic GvHD (19.7% for single Tx, 28.6% for 2nd Tx and 16.7% for 3rd Tx, p=0.628). There were no relevant differences between limited and extensive chronic GvHD between the group with 1 Tx and the group with >1 Tx. OS in relation to the disease type was as presented: sAML: 66.7% ± 19.2; N=6, AML: 62.8% ± 5.3; N=94, MDS: 57.1% ± 8.2; N=39, JMML: 32.8 ± 11.0; N=20, p=0.152. Subsequent transplants using reduced toxicity regimens (incl. Treosulfan) offer a curative option for children with AML or MDS/JMML relapsing posttransplant. Haploidentical transplants remain an alternative only for selected pts, for whom unrelated donors are not available. Objectives: The principal toxicity of molecular radiotherapy with 131-Iodine meta-iodobenzylguanidine (mIBG) is dose-dependent myelosuppression. Conventional mIBG therapy gives a whole body radiation absorbed dose (WBD) of around 2Gy, which allows spontaneous count recovery. This study evaluates the use, in a multi-centre setting, of peripheral blood stem cell (PBSC) support to allow safe dosimetry guided escalation of the administered activity of mIBG, given together with topotecan, also myelosuppressive, as a radiosensitizer, to give a prescribed WBD, guided by in vivo dosimetry, of 4Gy in two fractions. Methods: Patients with refractory or relapsed high risk neuroblastoma underwent PBSC harvest. The treatment schedule of mIBG and topotecan in neuroblastoma, MATIN, incorporates two administrations of mIBG prescribed to give a WBD of 4 Gy, with concomitant topotecan. PBSC return was when the residual radioactivity fell below 30 MBq, around two weeks after the second mIBG administration. Results: MATIN has been used 5 European centres in 69 patients with neuroblastoma, 44 male, 25 female, median age 6 years. 46 patients had refractory or progressive disease without prior myeloablative therapy (MAT), 23 had relapsed after MAT. Toxicity: In 2 patients, the full treatment was not given because of adverse events. There was 1 treatment-related death. 5 patients failed to regain normal platelet counts. Further therapy: In 61% of refractory patients, further potentially curative treatment including MAT was delivered. In six patients, repeat MATIN was given. Survival: For all patients, three year event free survival (EFS) was 0.16 (± 0.05) and overall survival (OS) was 0.26 (± 0.06). For patients without prior MAT, EFS and OS were 0.25 (±0.07) and 0.37 (±0.09). For relapsed patients, EFS was 0 and OS 0.07 (±0.07). These differences are statistically signifi cant (P=0.002). Conclusion: We have demonstrated that the MATIN schedule has an acceptable morbidity and mortality profi le in a group of neuroblastoma patients with a very poor prognosis. In very high risk patients with refractory disease, MATIN enabled further, potentially curative, treatment to be given resulting in encouraging survival rates. The MATIN schedule will be further evaluated in a randomised trial. Neutrophil, and platelet recoveries, grade 2-4 acute GVHD, and TRM, incidence were 91%, 88%, 9% and 8%, respectively.

Relapse incidence was 37%. Incidence of VOD, hemorrhagic cystitis (HC) and lung toxicity were 11%, 26% and 9%, respectively. In univariate analysis, a Bu Css higher than the median (>603 ng/ml) was associated with higher TRM (P=0.01), and HC (P=0.03) incidences. Overall (OS) and event-free survivals (EFS) were 69% and 55%. Bu Css infl uenced OS and EFS in univariate analysis with a lower OS and EFS with a higher Css (P<0.001 for both - Figure 1 ). In multivariate analysis, a higher Bu Css was associated with a higher TRM (HR=2.76; P<0.001), a lower EFS (HR=1.88; P<0.001) and OS (HR=2.12; P=0.001).

Our study suggests that outcome of pediatric patients receiving iv Bu can be predicted according to its pharmacokinetics. A prospective multicentric EBMT study is ongoing to confi rm these results. A reduced fi rst dose strategy should be investigated as well as the impact of pharmacogenetics. Busulfan (BU) is a key compound in conditioning myeloablative regimens for children undergoing hematopoietic stem cell transplantation (HSCT). There are wide interindividual differences in BU pharmacokinetics, which might increase the risk of venoocclusive disease (VOD), graft rejection and disease relapse. As BU is mainly metabolized by glutathione S-transferase (GST), it is possible that functional polymorphisms in GST genes may contribute to the variability in BU pharmacokinetics and/or HSCT outcomes. Our recent pilot study interrogating the polymorphisms in GST gene and pharmacokinetics after the fi rst dose of intravenous BU showed a signifi cant association between GSTM1 null genotype and higher drug exposure. The objective of the present study was to extend this analysis to higher number of patients (n=74), to evaluate the association between GSTA1, GSTP1 and GSTM1 genotypes and drug exposure and to evaluate whether GST genotype affect HSCT outcome in terms of effi cacy and toxicity. The genotypes in GSTA1 (C-69T, A-513G, G-631T, C-1142G), GSTM1 (deletion) and GSTP1 gene (A1578G, C2293T) were obtained and association analyses were performed with individual polymorphisms as well as with GSTA1 haplotypes resolved by PHASE software. Results seem to successfully replicate those obtained in pilot study for GSTM1 genotype. GSTA1 and GSTP1 polymorphisms analyzed individually were not associated with fi rst dose BU pharmacokinetics. However, haplotype analysis indicated that particular GSTA1 haplotypes affect drug exposure. The haplotype *A2 (C-1142, G-631, A-513, C-69), found in 23% of patients, was associated with lower drug levels, lower dose adjusted drug concentrations (PAUC=0.02; PCmax=0.003; PCss=0.02) and higher drug clearance (p=0.007). This fi nding is in accordance with functional studies suggesting higher GSTA1 enzymatic activity in GSTA1 *A carriers. Gene-dosage effect was also observed with drug exposure and clearance further affected in additive manner (PAUC=0.009; PCmax=0.003; PCss, =0.01, PCl=0.004). Patients with two copies of haplotype *A2 had better event free survival (p=0.04). In contrast, homozygous individuals for the GSTA1 T-69 or G-1142 allele defi ning haplotype *B and haplotype *B1, respectively, had higher occurrence of VOD (HR=10.5, p=0.002).

The results obtained open the possibility to tailor the fi rst BU dose according to patient genetics. 

Treosulfan (TREO) is increasingly used in pediatric HSCT because of its low acute toxicity. Little is known about the optimal dose required for children of different age groups. We conducted a pharmacology feasibility study to identify logistical and practical problems, especially because infants were to be included in the study. Transplantations: 15 children from 3 centers were included. 7 children under 10 kg BW (age 3.8-17.7, mean 10mo) received 3x12 g/m 2 and 8 children over 10 kg BW (age 1y4mo-14y, mean 5y2mo) received 3x14 g/m 2 of TREO over 2 hours on 3 consecutive days. TREO was measured at 8 different times (0-12 h) after administration on days 1 and 3. The "pushpull" method was used for blood sampling to minimize blood loss to 0.5 ml per time point. A method to measure TREO concentration in 0.5 ml of diluted plasma was established and validated. Urine was collected for renal TREO excretion. Initial pharmacokinetic data were obtained. Feasibility issues identifi ed: Obtaining exactly 0.5 ml of blood and pipetting it quickly into the prepared 0.5 ml of acid citrate solution was more diffi cult in practice than expected, leading to inexact pipetting. Nevertheless, the correct concentration of TREO could still be calculated from the total sample volume. When the push-pull method was used on central

S42 venous catheters, frequent catheter malfunctions and occlusions were seen, especially in infants. Quinton type implanted catheters were unproblematic. Pharmacology: The mean AUC in infants receiving 3x12 g/m 2 TREO was higher than in the older children receiving 3x14 g/m 2 (2203 vs 1507 mg/l × h, p=0.04). Volume of distribution at steady state (Vss) was lower in infants (9.4 vs 16.2l, p=0,02) than in older children. Half-life (mean 1.31 vs 1.59 hrs), percentage of renal elimination (60.2 vs 59.4%), and mean residence times (2.79 vs 2.94 hrs) were not different. There was no drug accumulation between days 1 and 3 owing to the low half-life of TREO that resulted in negligible trough levels. AUC profi les on days 1 and 3 were almost superimposable. Conclusion: A TREO pharmacology study in children that includes infants is feasible in a multicentric setting using the methods and considering the caveats described above. The high AUCs and low vss seen in infants compared to older children underline the necessity of further investigations that also include clinical parameters. Supported by BMBF grant Nr. 01E00802 IFB-TX Hannover and Medac GmbH. The development of solid phase assay for detection of anti-HLA antibodies has revolutionized our understanding of the role of these antibodies to predict antibody mediated rejection and graft survival in solid organ transplantation. In haematopoietic stem cell transplantation, infl uence of anti-HLA antibodies on graft outcome is still a matter of debate. Because of an adequate HLA matching in most cases between recipients and donors, incidence of donor specifi c anti-HLA antibodies (DSA) is low. Development of cord blood transplantation (CBT) with low level of HLA matching strongly increases the risk to detect the presence of anti-HLA antibodies by using high sensitive techniques. Presence of DSA might be associated with graft failure, delayed engraftment or persistant thrombocytopenia that could jeopardizes global survival. In pediatric patients, very limited data is available. In this study, we analyzed the presence of anti-HLA antibodies in a cohort of 55 children receiving a CBT with a conditioning regimen of Busulfan and Cyclophosphamide. Using luminex screening assays, we have detected before CBT the presence of anti-HLA antibody in 42% of recipients (23/55) among them 15 (27%) were positive for class I, 17 (31%) were positive for class II and 9 (16%) were positive for both anti-HLA antibodies. Each positive test was confi rmed by an analysis of anti-HLA antibody specifi city by luminex single antigen assay. The titer of these antibodies was defi ned using the Mean Fluorescence Intensity (MFI). In 8 patients, anti-HLA antibodies were directed against HLA of the donor. Presence of anti-HLA antibody class I signifi cantly infl uenced neutrophil recovery (p=0.021, OR=13.09). Platelet recovery showed a tendency to be prolonged in the presence of anti-HLA antibodies of class II, with MFI>2,000 (p=0.06). Interestingly, presence of DSA was associated with a higher incidence of acute GvHD (p=0.03). Finally, overall survival after CBT was not affected by anti-HLA antibody except in presence of anti-HLA class II but at high MFI (>5,000, p<0.001). In summary, our data suggests that anti-HLA antibodies of class I and class II are associated with signifi cant clinical events. A therapeutical strategy to reduce the level of anti-HLA antibody before CBT in children should be considered.

Objectives: Graft-versus-host disease (GVHD) is one of the major complications infl uencing morbidity and mortality after allogenic stem cell transplantation (SCT). In this retrospective analysis over 10 years, we evaluate the outcome of extracorporeal photochemotherapy (ECP) as a treatment modality of GVHD. Patients and Methods: From 1999 to 2011, 38 patients, consisting of 15 (f/m 7/8) with acute GVHD (aGVHD) and 23 (f/m 11/12) with chronic GVHD (cGVHD), were treated with ECP. Patients weighing less than 40 kg were available for the offl ine system (n=16). Very small or critically ill children were treated with mini ECP (n=1). And additionally there was the closed or offl ine system for patients weighing more than 40 kg (n=12). Some patients were treated with both the inline and offl ine system (n=7) or with fi rst mini ECP followed by the offl ine method (n=2). The median age at the fi rst ECP was 7 years for patients with aGVHD and 13 years for patients with cGVHD, the median body weight was 27 and 38 kilograms. Fifteen patients were suffering from aGVHD with manifestations in skin (n=14), oral mucosa (n=3), liver (n=6) and gut (n=4). Twenty-three patients had ECP because of their cGVHD, the affected tissues were: skin (n=17), oral mucosa (n=16), liver (n=8), gut (n=4), lung (n=4), joints (n=10), eyes (n=9), muscles (n=2) and the central nervous system (n=1). Results: In a GVHD 11/15 (73%) patients responded with improvement overall and 11/14 (79%) in skin, 0/3 (0%) in oral mucosa, 4/6 (67%) in liver and 3/4 (75%) in gut involvement after a median of 15 cycles. Chronic GVHD improved in 18/23 (78%) patients. Skin improved in 13/17 (76%), liver in 7/9 (88%), gut in 2/4 (50%) and lung in 1/4 (25%) after a median of 26 cycles. The overall survival was 89% and 94% for patients responding to ECP and 0% and 40% for non-responders. Conclusion: Outcome is dependent on response to ECP. ECP enables steroid tapering and shows good responses for acute and chronic GVHD. In spite of using different methods to perform ECP, the outcome doesn't seem to be strongly dependent on it. Even after 10 years the survivors of GVHD, treated by ECp showed no specifi c side effects regarding to ECP. ECP in children seems to be very save, even in long term outcome.

Background: Relapse after allo-SCT in pts with AML or MDS is associated with a poor prognosis and therapeutic options are limited. Thus, there is need for novel strategies which ideally target the leukemic clone and enhance a GvL effect. Aza might S43 provide these properties, and retrospective studies in this context were encouraging. Design/Methods: We here report the fi nal results from a prospective single-arm EBMT phase II trial investigating the combination of Aza+DLI as 1st salvage therapy in pts with AML or MDS with hematological relapse after allo-SCT. Treatment contained up to 8 cycles Aza (100 mg/m 2 /d d1-5, every 28 d) followed by DLI with increasing dosages (1-5x10 6 -1-5x10 8 cells/kg) after every 2ndAza cycle. Results: A total of 30 pts (19 female/11 male, median age 55 y) from 6 centres were included. The majority of pts (n=28) suffered from AML, while 2 pts had MDS or MDS/MPN. At transplant, 16 pts (53%) had active disease and 14 pts (47%) were in remission. Following allo-SCT, hematological relapse evolved in all pts (median BM blasts: 34%, median chimerism: 67%) at a median time of 175 d (19-1699) . A median of 3 courses Aza (range 1-8) were administered, and 22 of 30 pts (73%) received DLI (median: 1, range: 1-4, median CD3 dose 5x106/kg/DLI). ORR was 47% (CR/CRi 7 pts/ PR 2 pts SD 5 pts). Median time and number of Aza cycles to best response were 79 d (range 28-299) and 3 cycles (range 1-8). Of the 7 pts with CR, 5 pts continue in CR for a median of 777 d (range 461-890), while 1 pt relapsed after 396 d and 1 pt died from GvHD. By Dec 2011 median follow-up of surviving pts is 817 d (range 732-974) and 5 pts (17%) are currently alive. Twelve pts have died due to progressive disease, while 7 pts died during (2 infection, 1 bleeding) or after the end of therapy (1 GvHD, 2 infection, 1 bleeding). Median OS of all pts is 117 d (66-168). Patients with CR/CRi had a signifi cant longer OS than pts not reaching CR/CRi (n. r. vs. 83 d, p<.001). A lower blast count and adverse cytogenetics positively correlated with the likelihood of achieving CR. Eleven pts (37%) developed aGvHD, while cGvHD was observed in 5 pts (17%). Cytopenias grade III/IV occurred in all pts, but were considered to be drug-related in only 11 pts (37%). The most common drug-related grade III/IV non-hematological toxicity were infections. Conclusion: Aza and DLI as 1st salvage therapy is safe and active in pts with AML/MDS who relapse after allo-SCT inducing durable remissions in a subgroup of pts. AZA has become a reference tttof IPSS higher risk MDS, but its roleremain uncertain.We report on 462 consecutive MDS pts who underwent SCT between 1999-2009. Inclusion criteria were age >18, dg of MDS according to the FAB or WHO criteria, SCT from sibling (n=254) or MUD HLA (10/10(n=163)). The SC source was blood (n=294) or marrow (n=123) and the conditioning was either myeloablative (n=184) or of reduced intensity (n=233). 462 fi les were analyzed. 255M,162F. At dg, WHO was,112 RA/ RARS/RCMD,137 RAEB1,140 RAEB2 and 28 RAEBt/AML; Cytogenetic IPSS were fav(n=215), interm(n=106), unfav(n=91); 204 pts IPSS Int-2 or High. 144pts had progressed before allo-SCT. Prior to SCT, 66 pts had received AZA, alone (n=49), AZA +ICT (n=17). The 328 remaining pts received BSC (n=171) or ICT (n=180). In AZA groups, the drug was started 38 to 941d after dg and discontinued 6 to 438d before SCT with a median number of 4 cycles. 288 patients were considered responders at SCT, while 242 were SCT with progressive disease. Median age at SCT was 54 years (18.3-70.7). Pts received myeloabla-tive (n=150) or nonmyeloablative (NMA) (n=267) conditioning with BM (n=123) or PBSC (n=294) as source SC. Compared with other treatment groups, patients belonging to AZA-alonegroup were older (p=.0001), had more often Int-2 and high IPSS (p=.0001) and one third of them (33%) had poor cytogenetics (p=.07). They received more often NMA conditioning (p=.0001) from MUD (p=.0007). As of April, 2011, median FU was 48 mo (2.4-146.4. The estimated 3-y OS and EFS were 59%, 35%, 45%, 53% (p=.10); and 53%, 29%, 40%, 54% (p=.11) in the AZA-alone, AZAchemo, iCT and BSC groups, respectively. The estimated 3yTRM was 35%, 21%, 34% and 28% (p=.036) and 3y relapse was 31%, 35%, 38% and 27% in the AZA-alone, AZA-chemo, iCT and BSC groups (p=.54). The 3y OS and EFS AZA group were better than the iCT alone; the difference was not statistically signifi cant. This difference was statistically different when we compare the 3-y OS and EFS between the AZA group and AZA + iCT with P=0.027 and 0.032 respectively. Multivariate analyses confi rmed the infl uence of prior ttt with an unfav outcome in pts in AZA-chemo group on OS and EFS (p=.027 and p=.032, respectively). AZA ttt is valid therapeutic approach, but mainly in pts receiving AZA alone with an outcome at least equivalent to pts receiving induction CT prior SCT. SCT in pts who required AZA+CT had less satisfactory outcomes, possibly refl ecting additional toxicity and/or more resistant disease.

In order to study if alloSCT can improve the dismal natural course of poor-risk CLL, we aimed at assessing the impact of alloSCT by measuring its outcome from the time of donor search indication rather than from the time of transplant, thereby including those patients who failed to proceed to allografting for any reason. Study and Design and Patients: In a single centre retrospective analysis, course and outcome of all patients with CLL referred for evaluation of alloSCT indication between June 2005 and July 2011 were recorded. Primary endpoint was overall survival (OS) of those patients for whom a 9/10 or 10/10 matched donor could be found within 3 months compared with that of patients without donor, measured from the 3-month landmark after donor search initiation. Results: Altogether 117 patients with a median age of 53 (37-69) years were referred in the six year time period. An indication for donor search was seen in 98 patients (84%). According to EBMT criteria, indication was (1) symptomatic 17p-in 31 patients, (2) fl udarabine refractoriness in 27, and (3) early relapse after intensive pretreatment in 35 patients. Another 5 patients were considered because of Richter's transformation. Of the 98 patients with indication, 8 patients refused donor search. A donor could be identifi ed for 72 (80%) of the remaining 90 patients. Of these, 21 (29%) did not proceed to transplant because of CLL progression (9) or other reasons (12), whereas 51 patients underwent RIC alloSCT from a sibling (17) or unrelated donor (34). With a median observation time of 20 (0-73) months, OS at 2 years from start of search was 69% for all 98 patients who had a search indication. There was no impact of type of EBMT criterion (1 vs 2 vs 3) and age on the outcome of patients with search indication. With regard to the primary endpoint, OS at 2 years from the 3-month landmark after start of search (excluding all patients who died within this time frame for any reason) was signifi cantly better in patients with a donor than in those without (83% vs 56%; p=0.007).

2-year OS of the 51 patients actually transplanted 91%.

Conclusions: Survival of patients with poor-risk CLL and allo-SCT indication according to the EBMT consensus for whom a donor cannot be found is signifi cantly inferior compared to patients with a donor. This observation provides fi rst comparative evidence that alloSCT indeed may have the potential to improve the natural course of poor-risk CLL.

Chemotherapy-refractory stages of chronic lymphocytic leukemia (CLL) represent an indication for allogeneic hematopoietic cell transplantation (HCT). Indirect evidence exists, that lower tumor load is associated with better outcome. Alemtuzumab has demonstrated activity in chemo-refractory disease and to prevent graft-versus-host disease (GVHD). Yet, treatment with alemtuzumab prior to HCT has been associated with inferior outcome in one large retrospective analysis. Here, results from a phase II trial are presented where we systematically investigated treatment with alemtuzumab followed by a washout period and subsequent reduced intensity conditioning (RIC) prior to allogeneic HCT. Methods: All patients were scheduled for one month of cytoreductive pretreatment with 3 x 30 mg alemtuzumab weekly. The washout period after the last dose of alemtuzumab was increased from two weeks to one month during the study. Antibody levels were measured on the day of HCT and T-cell engraftment was monitored. RIC consisted of fl udarabine (150 mg/m2) and busulfan (8 mg/kg) in combination with CSA/MTX as GVHD-prophylaxis. Results: 62 patients with relapsed or refractory CLL and a median age of 56 years were included. Twenty-four patients (39%) suffered from refractory disease, 12 patients (19%) had a del(17p), 19 patients (31%) had treatment failure within 2 years. Two patients died early after pre-treatment due to progressive disease and did not proceed to HCT. Donors were HLA-identical siblings for 26 and HLA-compatible unrelated donors (one mismatch accepted) for 34 patients. No primary or secondary graft failure occurred. Interference of persisting antibody levels with T-cell engraftment could be demonstrated. The cumulative incidence of acute GVHD grades II to IV at day +100 was 42% and of extensive chronic GVHD at one year was 56%. With a median follow-up of 35 months (16 to 61 months) 3-year overall survival and progression-free survival were 62% (95% CI, 48% to 76%) and 52% (95% CI, 38% to 66%). Day +100 non-relapse mortality was 2%. At 3 years non-relapse mortality and relapse incidence were 24% (95% CI, 12% to 36%) and 24% (95% CI, 13% to 35%), respectively. Conclusions: The long-term results of this study argue in favor of that strategy for patients with refractory disease. However, unless in vivo T-cell depletion is the goal, a washout period of a minimum of one month after the last dose of alemtuzumab should be kept and T-cell engraftment needs to be monitored.

Background: Despite strides in our understanding of clinical, genetic, and molecular aspects of the disease, coupled with development of more effective therapies, MM remains incurable. Combining benefi ts of cytoreductive-therapy with adoptive immunotherapy forms the basis of a tandem auto-allo HCT treatment strategy in patients (pts) with MM. Confl icting results, however, have been reported when an auto-allo HCT approach has been compared to an auto-auto HCT strategy, which relies solely on dose-intensity of chemo-or chemoradio-therapy to eradicate disease. Methods: A systematic search of MEDLINE database (1966-Nov 5, 2011), and pertinent conference proceedings, identifi ed 203 publications. Any randomized controlled trial (RCT) comparing an auto-allo HCT to tandem auto-auto HCT were eligible. A meta-analysis was conducted as per the standards recommended by Cochrane Collaboration. Results: Six RCT (full-length manuscript=4, abstract=2) met inclusion criteria. All studies allocated pts to auto-allo HCT if an HLA-matched sibling donor was available (genetic randomization). The 6 included trials enrolled 1798 pts. Three studies enrolled 1229 (auto-allo=414, auto-auto=815) pts on intention-to-treat (ITT) principle. Three remaining trials reported outcomes on 569 (auto-allo=273, auto-auto=296) pts using a per-protocol approach. In all but one study, groups were balanced for presence of disease-related prognostic factors. Non-relapse mortality (NRM) was signifi cantly worse with an auto-allo HCT approach [ 

Introduction: Autologous transplantation has a defi ned role in the upfront treatment of multiple myeloma (MM) but nearly all patients will eventually relapse. Allogeneic transplantation carries the benefi t of graft versus myeloma (GVM) effect but also high risk of treatment related mortality (TRM). Trials randomizing patients with newly diagnosed MM to tandem autologous (TA) or the combination of autologous + RIC allogeneic transplantation (AR) based on the availability of an HLA-matched donor have yielded heterogeneous results. We performed a meta-analysis of all reported trials comparing TA with AR transplantation in newly diagnosed MM to summarize the available evidence supporting the upfront use AR in multiple myeloma. Methods: We utilized a comprehensive search strategy to identify all trials meeting the following entry criteria: prospective trial, inclusion only of newly diagnosed patients (typically after conventional induction therapy), subjects undergoing similar induction therapy and fi rst autologous transplantation in both arms, assignment to RIC allogeneic transplantation or a second autologous transplantation based exclusively on the availability or not of a HLA matched donor, use of RIC and report of at least overall survival (OS) and/or progression-free survival (PFS) as endpoint. OS analyzed per intention to treat (ITT) was the primary endpoint. PFS, complete response (CR) rate and rate of TRM were secondary endpoints. Results: We identifi ed 6 trials meeting entry criteria. There were 1192 patients assigned to TA and 630 patients to AR. Relative to TA, AR patients were more likely to experience TRM (OR = 3.1, 95% CI = 2.1 to 4.5), but not more likely to achieve CR (OR = 1.2, 95% CI = 0.9 to 1.7). There was no statistically signifi cant difference in PFS within the fi rst 36 months (HR=1.1, 95% CI=0.9 to 1.3) or beyond 36 months from transplant (HR=0.9, 95% CI=0.7 to 1.3). Similarly, there was no signifi cant difference in OS between the arms within 36 months (HR=1.2, 95% CI=0.9-1.5) or beyond 36 months (HR=1.1, 95% CI=0.7-1.6). Conclusion: We conclude that AR is associated with substantially higher TRM but with no improvement in CR rate, PFS or OS over TA transplantation in the initial management of patients with MM. Allogeneic transplantation in this setting remains experimental. Innovative approaches are needed to improve the safety and effi cacy of this strategy.

Introduction: In some patients (pat) with Multiple Myeloma (MM) long-term survival can be achieved after autologous (auto) or allogeneic (allo) stem cell transplantation (SCT). We prospectively investigated the incidence and impact of molecular remission (molCR) on long-term outcome after auto-allo tandem approach in MM (NCT 00781170). Methods: From 4/2000 to 10/2008, 73 pat with median age of 50 years (r 29-64) and advanced stage II/III MM were included. Conditioning for auto SCT was melphalan 200 mg/m 2 given over 2 days. After 2-3 months, dose-reduced conditioning with melphalan 140 mg/m 2 , fl udarabine 180 mg/m 2 and ATG Frese-nius® (Fresenius, Bad Homburg, Germany 10 mg/kg for related and 20 mg/kg for unrelated donors on day -3, -2,-1) followed by allo SCT was performed.

Remission was according to modifi ed EBMT criteria. In 46 patients with CR or nCR minimal residual disease was monitored by patient-specifi c primers (n=20) or highly sensitive plasma cell chimerism (n=26). Sensitivity was 10-4 to 10-5 for plasma cell chimerism and 10-5 for nested PCR with patient-specifi c primers respectively. Results: Overall 44 pat (60%) achieved a CR according to EBMT criteria with negative immunofi xation. 8% achieved a VGPR and 18% a partial remission. 3% had progressive disease and 11% were not evaluable for determination of remission. MolCR was observed in 30 pat, in 15 pat the molecular markers were sustained negative while in 15 pat molecular markers were only intermittently negative, resulting in an overall molCR rate of 46% and a stable sustained molCR rate of 23%. After a median follow-up at 7 years the 5 year progressivefree survival was 29% (95% CI: 17-41%). Patients without del 13q14 had a signifi cantly better PFS than those with del 13q14 (5 y: 56% vs. 17%, p=0.02). Patients who achieved CR after transplantation had improved PFS in comparison to non-CR patients (5 y: 41% vs. 28%, p=0.008). Patients with sustained negative molCR had a 5 year PFS of 85% vs. 31% for mixed molCR (p=0.003).

The 5 year overall survival was 52% (95% CI: 40-64%). Patients without del 13q14 had signifi cantly improved 5 year survival (75% vs. 40%. p=0.02). Patients who achieved a sustained molCR had a 5 year OS of 91% while those with mixed molCR resulted in a 5 year OS of 87% (p=0.06). Conclusions: The study underlines the importance of the depth of remission and that achieving molCR is associated with longterm freedom from disease and potential cure of MM. 

Allogeneic stem cell transplantation (aSCT) is the only curative treatment for myelofi brosis. Here we present a long-term follow up of patients with myelofi brosis treated with reduced-intensity aSCT) in the prospective multicenter study conducted by the MDS subcommittee of the Chronic Leukemia Working Party of the EBMT (study registration NCT 00599547). From 2002 to 2007, a total of 103 patients with primary (63 pts) or post-polycythemia vera and -essential thrombocythemia myelofi brosis (40 pts) from seventeen transplantation centers in three nations were included. There were 62 males and 41 females with a median age of 55 years (range, 32-68 years). Risk profi le according to Lille score was low risk with constitutional symptoms (17%), intermediate risk (53%) and high risk (30%). All but three of the patients received peripheral stem cells as stem cell source from either related (n=33) or unrelated donor (n=70) and a conditioning with Busulfan (10 mg/kg orally or 8mg/kg intravenously) ,Fludarabin (180 mg/m²) and ATG according a previously published protocol. After a median follow up of 60 months (range 9-109 months), 45% of patients had a chronic graft vs. host disease which was extensive in the half of cases. The 5 years and 8 years estimated overall survival (OS) was 68% and 65%, respectively with a stable plateau after 5,3 years follow up (Figure-1) . Estimated 5-year disease-free survival was 40%. The cumulative incidence of relapse/progression at 3 and 5 years was 22% and 28% and the non-relapse mortality at 1 and at 3 years was 18% ands 21%, respectively. In multivariate analysis advanced age >55years (HR: 4.69, p=0.001), absence of JAK2V617F mutation (HR: 2.50, p=0.02), mismatched donor (HR: 3.62, p=0.002) were signifi cant independent predictors for reduced OS. During follow up relapse/ progression occurred in 28 patients. Twenty one of them were treated with donor-lymphocyte infusions (DLI) and/or a second allogeneic transplantation (n=11). Fifteen of those were durably rescued with this treatment and were at the last follow up alive for a median interval of 49 months (range 25-61 months) after relapse/progression, whereas another one received recently a second transplantation for relapse and is still alive for a follow up of 4 months. This update of a prospective trial using reduced intensity conditioning followed by allogeneic stem cell transplantation for myelofi brosis confi rmed a very good long-term OS. Aim: To compare effi cacy of allo-HSCT with reduced intensity conditioning (RIC) and myeloablative conditioning (MAC) in children and adolescents with very high risk ALL (VHR ALL). Patients and Methods: 102 patients (pts) 1 to 21 (median 12) y.o. with VHR ALL underwent allo-HSCT between 2000 and 2010. VHR ALL was diagnosed due to late response to chemotherapy (induction failure), poor-risk cytogenetics (t(9:22), t(4:11)), MRD(>10 -2 ), infants with 11q23 rearrangement, short fi rst remission, primary resistance or resistant relapse. Allo-HSCT with RIC was performed in 31pts (RIC group) with Karnofsky (Lansky) <80%, history of therapy toxicity or ongoing infection: 20pts in 1st or 2nd complete remission (CR), 11 pts in resistant relapse. MAC was used in 71pts (MAC-group): 41pts in 1st or 2nd CR, 30pts were in more than 2nd CR or resistant relapse. RIC consisted of Flu 150 mg/m 2 and Mel 140 mg/m 2 (with or without ATG) or Flu 150 mg/m 2 and Bu 8 mg/kg with or without ATG. MAC consisted of Bu 16 mg/kg (or Treo 36-48 mg/ m 2 ) and Cph 120 mg/kg with or without ATG. Thirty pts received allo-HSCT from matched related donor (RIC, n=7), 72 pts from matched unrelated donor (RIC, n=24). Results (Table 1 ): In RIC-group the pts engrafted on median D +18 (13-31). Patients with 1st or 2nd CR receiving allo-HSCT with RIC had the 7-years overall survival (OS) and event-free survival (EFS) of 51% and 36%, 6 of these pts died within 100 days from HSCT (2 of infection and 4 of aGVHD). Six pts (31%) relapsed, 5 of them achieved CR after chemotherapy and DLI or withdrawal of immunosuppression, 1 patient died of disease progression. Among patients undergoing allo-HSCT with RIC in resistant relapse the CR was achieved in 5 of 11 pts. Nevertheless, the 1-year OS was 0%. Three patients died of infection, 1 of aGVHD and 7 of progression. MAC group pts engrafted on median D+21 (10-49). In pts receiving allo-HSCT with MAC the 7-years OS and EFS were 57% and 46%. Ten (26%) pts relapsed, but 2 of them achieved CR after chemotherapy and DLI, 9 pts died of relapse, 5 of aGVHD, 2 of infection, 1 from transplant-related toxicity, 1 due to engraftment failure. Among 30 pts in resistant relapse or more than 2nd CR receiving allo-HSCT with MAC 4 are currently in CR, the other died from relapse (15), infection (7) or aGVHD (4).

Conclusion: Allo-HSCT with RIC is effective in young (<21 y.o.) patients with VHR ALL and its results are comparable with those of allo-HSCT with MAC. Allogeneic stem cell transplantation is the only curative option for many hematological malignancies, but toxicity of conditioning is still a limiting factor. Treosulfan is a bifunctional alkylating agent with reduced extra-hematologic toxicity profi le. We evaluated the effi cacy of Treosulfan-based conditioning regimen. We analyzed 183 pts transplanted between 2004 and 2011 in 12 Italian centers. Median age 51 years (21-71), most common disease was acute leukemia (45%), 50% of pts were transplanted in complete remission. Median Sorror score was 1 (0-7). Donor was unrelated (UD) in 57% of pts; source of stem cells was: BM 8%, PBSC 81%, CB 11% of pts. Conditioning: Treosulfan (14 g/msq for 3 days) and Fludarabine (30 mg/msq for 5 days). In vivo T and B-cell depletion was performed by ATG-Fresenius (10 mg/kg 3 days) and Rituximab (single 500 mg dose) only in pts receiving an UD. GvHD prophylaxis: Cyclosporine A and short course of Methotrexate, excepting CB. Median CD34+ infused for BM-PBSC was 6x10 6 /kg (1-17), median TNC for CB was 3.2x10 7 /kg (1.2-6; 11 pts received double CB). Median follow-up was 35 months (3-84). At day 60 cumulative incidence (CI) of neutrophil engraftment was 93±2%, median time 17 days. In 164 evaluable pts molecular chimerism at day 60 was full-donor in 88%. Median grade of regimen-related toxicity (CTC score) was 0 (0-4), the most frequent toxicity was a transient rise of bilirubin (78%). Grades II-IV aGvHD developed in 58 pts and 100 days-CI was 28±3%. Among 148 pts at risk, 54 developed chronic GvHD and 3 years-CI was 31±4%, aGvHD (time dependent variable) increased the risk of cGvHD (p=0.007). At 3 yrs CI of relapse was 30±3% (n=53); no risk factors found. At 3 yrs DFS was 45±4%, in multivariate analysis factors associated with a decreased DFS were: use of CB (HR=2; p=0.02), active disease at transplant (HR=2, p=0.001). At 3 yrs CI of TRM was 25±3%; in multivariate analysis use of CB (HR=2.9, p=0.005), active disease (HR=2.7, p=0.002), aGvHD (HR=3.5, p<0.001) were associated with increased TRM. Overall survival at 3 yrs was 50±4% and risk factors associated with decreased OS were: use of CB (HR=2.8, p=0.001), active disease (HR=2.3, p<0.001), aGvHD (HR=1.9, p=0.005), sorror>0 (HR=1.6, p=0.05). 86 pts died after transplant, 40% for relapse and 60% for transplant related causes. Treosulfan is an effective myeloablative drug with low extrahaematological toxicities. Strategies to reduce aGvHD are warranted to deceased TRM. Background: Recent cooperative group studies for children with high-risk neuroblastoma (NB) report survival rates approaching 45% by multiple treatment modalities including surgery, radiation therapy and high-dose chemotherapy followed by autologous stem cell transplantation (SCT). Disease relapse remains the most common cause of treatment failure and the survival of patients with relapse or chemo-refractory disease has been reported less than 20%. Recent advances in unrelated cord blood transplantation (UCBT) have provided increased chances for patients with hematological malignancies. Some case reports suggested graft versus NB effects after allogeneic SCT. However, only small series of UCBT for NB has been reported. We retrospectively analyzed the clinical outcome of UCBT for patients with NB in Japan. Patients and Methods: We analyzed the clinical outcome of 75 children (47 boys and 28 girls) with NB (age, <18 years) who received UCBT in Japan, and registered to the Transplant Registry Unifi ed Management Program (TRUMP). Overall survival (OS) and progression free survival (PFS) were analyzed using Kaplan-Meier estimates. Results: The median age at transplantation was 3 years (0-11 years). Clinical stage of the patients was all stage IV (International Neuroblastoma Staging System) except for one stage III. Median follow-up period after UCBT was 401 days (from 8 to 4417 days). The probability of 3-year PFS and 3-year OS of the patients was 44.3% ± 6.8% and 43.9% ± 7.0% respectively. Of note, 3y PFS of the patients who achieved the 2nd complete response (CR) after relapse or the 1st partial response (PR: primary tumor has decreased by 50% to 90% decrease in the size of all measurable lesions) was 38.5% ± 8.6%, although this is signifi cantly worse than that of the patients who obtained the 1st CR or very good partial response (VGPR: Primary tumor has decreased by 90% to 99%, and no evidence of metastatic disease) (51.5% ± 9.5%, p=0.03). Causes of deaths after UCBT were relapse or progression of tumor (n=22) and treatment related mortality (n=13). Multivariate analysis for PFS or OS revealed that disease status at CBT (1st CR+VGPR), number of infused cells (over 4x10e7/kg patient's body weight) and GVHD prophylaxis (cyclosporine/tacrolimus+methotrexate) were signifi cant better covariate factors. Conclusion: These results suggest that UCBT could be the option for relapsed or chemotherapy resistant NB. Cumulative incidence of grade 2-4 acute at day 100 and chronic GVHD at 6 months were 55% and 44% respectively. One-year overall (OS) survival was 29% [95CI: . No factors (protocol, number lines of previous treatments, PS or disease status at Allo-SCT) were found to be associated with either response or OS. Finally, despite an encouraging response rate of 46% in these patients with advanced refractory metastatic RCC, outcome remains poor with short response duration and no long term survivor. However, the recent development of antiangiogenic therapies remains unsatisfactory. Allo-SCT to induce response associated with these agents to prolong response duration deserves to be study in selected population.

Objective: The prognosis of patients with STS who present or relapsed with distant metastases has not improved despite intensifi cation of chemotherapy with or without stem cells rescue. The use of allo-SCT has been initiated for these patients under the assumption of the graft versus sarcoma effect (GvT). In Germany there are two ongoing prospective trials (Frankfurt and Tuebingen) on the role of haploidentical SCT in patients with STS. Additionally, many centers performed allo-SCT in high risk STS patients outside prospective studies. Allo-SCT is regarded mainly as immunotherapy exerting its activity against residual disease. There is, therefore, still controversy over what impact on prognosis an aggressive local therapy (LT) to all metastatic sites, if feasible would have for the outcome after allo-SCT and what would be the optimal time for LT. Methods and Results: To answer this question we have analysed sites of relapse after allo-SCT in 61 patients with primary (18) or secondary (43) metastatic STS registered in the CWS Studies and SCT registry between 1990 and 2010. 43 pts were >10 yrs of age. Different busulfan and melphalan based conditioning regimen were used. 18 patients were treated in prospective trials in Frankfurt or Tuebingen. 27 allografts were received from MRD, 30 from MMRD, 2 from MUD and 2 from MMUD. 29 pts were in CR or VGPR the time point of SCT, other had stable (28) or progressive (4) macroscopic tumour manifestation (PR).18 out of 29 pts who were in CR/VGPR at SCT relapsed in a median time of 8 months (range 2-41). 8/29 patients died in CR of therapy related reasons (TRD), 6 maintained their CR for a median time of 42 months (5-136) after SCT. Only 2 patients relapsed in previous tumor sites (primary tumor or metastases), 5 revealed mixed relapses in old and new localisations, 10 patients developed new metastatic sites (soft tissue, bone, abdomen). Local therapy consisted in the majority of patients of RTX with or without surgery. 8 out of 28 patients with PR prior to SCT achieved CR at day 100 after SCT. 3 died of TRD, 4 relapsed subsequently mainly in old tumour localisation. Conclusion: Only patients with limited tumour burden might benefi t from GvT. The tumour spread after SCT suggest that local control of macroscopic manifestations is a prerequisite but microscopic dissemination seems to be the main obstacle for cure. Additional post-transplant immunotherapeutic approaches are needed to better control the disease. Human mesenchymal stem and progenitor cells (MSPCs) from various tissues are evaluated in clinical trials for bone and marrow regeneration. Better understanding MSPC functionality in vivo could help to select optimal cells for specifi c disorders thereby increasing clinical effi ciency. Here we demonstrate that epigenetic predisposition of bone marrow (BM) MSPCs facilitates cartilage, bone and hematopoietic microenvironment formation in vivo. BM, white adipose tissue (WAT), umbilical cord (UC) skeletal precursors and skin fi broblasts were subjected to surface immune phenotyping before adipo-, chondro-and osteogenic differentiation in vitro. In vivo mesenchymal tissue formation was tested by transplanting MSPCs subcutaneously into NSG mice. Sequential cartilage, bone and BM formation was evaluated by in vivo imaging and histology. Infi ltration of host hematopoiesis was studied by analyzing lin-/sca1+/c-kit+ (LSK) cells using polychromatic fl ow cytometry. Epigenetic profi ling comparing bone forming with non bone-forming MPSCs was done using a whole genome methylation array. Translation of epigenetic regulation into gene transcription was evaluated by qRT-PCR. MSPC from all sources show an almost identical immunephenotype. Effi cient chondrogenic differentiation of BM MSPCs can distinguish this source in vitro. Transplantation studies in vivo suggest that only BM MSPCs are capable to form cartilage and bone through endochondral ossifi cation thereby generating a niche for hematopoietic invasion. Comparing methylation profi les, BM MSPCs show a unique signature and cluster separately. Most differentially methylated genes include transcription factors critically involved in endochondral ossifi cation. CpG island hypomethylation results in increased transcription into mRNA. BM is the only tissue containing precursors with true endochondral differentiation potential. This is refl ected by a BM-specifi c epigenetic signature predisposing these cells to form cartilage and bone. In addition this developmental process is critically involved in hematopoietic niche formation. Our data suggests a more critical evaluation of in vivo differentiation capacity mainly supporting the use of BM-derived MSPCs for bone and marrow regeneration. Introduction: It has previously been shown by us and others that stable therapeutic human neo-vasculogenesis in vivo depends on co-transplantation of pericytes or their mesenchymal stem/ progenitor cells (MSPCs) with endothelial cells or endothelial colony-forming progenitor cells (ECFCs) (Reinisch et al., Blood, 2009 ). Here we describe a novel strategy for unraveling the mediators of neo-vasculogenesis by proteomic profi ling of the early signaling signature in vivo providing us with tools to develop strategies for therapeutic intervention as well as regenerative applications. Methods: MSPC and ECFCs were transplanted subcutaneously in matrigel plugs alone or at a ratio of 20:80 into immune defi cient NSG mice. Implants were harvested 24h after transplantation for proteomic profi ling using KAM 1.3 antibody microarray (www.kinexus.ca) testing over 800 signaling and phospho-proteins. The state of vessel formation and stability of the transplants were verifi ed by histological follow-up of corresponding explants for 2 and 8 weeks after transplantation. Therapeutic targets were selected from antibody microarray based on differential display and were used for in vitro toxicity and viability assays as well as in vivo modulation of therapeutic vasculogenesis. Results and Discussion: Results confi rmed that co-transplantation of ECFCs with MSPCs was most effi cient for forming stable perfused human vessels. "ECFC only" plugs showed vessel formation after transplantation of higher cell number and later in the time course after transplantation. Early chondrogenesis was observed in "MSPC only" plugs after 8 weeks. Protein microarray data analysis revealed signifi cant alteration of components including (1) caspases, DAXX and P53 involved in death-associated pathways, (2) ERB, MAPK, mTOR and TGF-ß signalling, (3) Focal adhesion, VEGF, JAK-STAT and Wnt signaling. Selected targets are currently validated by in vitro and in vivo blocking experiments. Conclusion: Transplantation of more than one purifi ed cell type is needed for tissue engineering and vascular regeneration in vivo. Proteomic profi ling unraveled independent and partially overlapping signaling networks involved in the complex process of vascular regeneration.

Objectives: The manufacturing of red blood cells (RBC) from human induced pluripotent stem cells (iPS) offers the potential to generate large quantities of patient's specifi c RBC from an unlimited source. However, in vitro recapitulation of physiological erythropoiesis from embryonic pluripotent stem cells (ESC), which includes mesoderm induction, generation of hematopoietic stem cells (HSC), followed by erythroid maturation, hemoglobin switching and enucleation, remains a challenge. Methods: We used human ESC (H1) and iPS cells from human fi broblasts, neuronal stem cells and CD34+ HSC for the ex vivo generation of erythroid cells. First, ESC/iPS cells were allowed to form embryoid bodies (EB) under cytokine stimulation for >15 days. Thereafter, EB were dissociated and single cells were applied to a three-step protocol for human erythropoiesis over 18-25 days. Hematopoietic maturation was analysed by colony formation, fl ow cytometry (CD34, CD45, CD36, glycophorin A (GPA)), microscopy after May-Grünwald-Giemsa staining and hemoglobin analysis. Results: The used EB protocol allowed for the generation of 3%-15% CD34+ and CD45+ cells from human ESC/iPS cells, able to form erythroid and myeloid colonies in semisolid cultures. These hematopoietic cells further developed into erythroid precursors as determined by >80% expression of CD36 and GPA, followed by maturation into normoblasts and partially enucleated RBC. Manufactured cells mainly contained embryonic and fetal hemoglobin. Compared to ESC, fi rst results indicate a lower proliferation capacity of iPS cells, whereas maturation into RBC was partially higher. Conclusions: We were able recapitulate the multistep development of RBC from human ESC and iPS cells of different origin. This might contribute as a fi rst step towards the large-scale ex vivo generation of patient's specifi c RBC from human iPS cells for transfusion purposes. Furthermore, the iPS related hematopoiesis and erythropoiesis might serve as a model for studying various hematopoietic diseases. Mesenchymal stem/progenitor cells (MSPC) are defi ned by their capacity to differentiate into bone, cartilage and fat cells in vitro. However, their limited engraftment potential in vivo led to the assumption that the transient therapeutic benefi ts observed after MSPC transplantation are predominantly due to paracrine effects. Based on preliminary observations indicating superior functionality of MSPC expanded under the aegis of human platelet lysate (HPL), we hypothesised that the lack of MSPC engraftment and differentiation in vivo can be reverted by human platelet-derived factors. We compared human bone marrow MSPC expanded in HPLsupplemented culture medium to MSPC derived in foetal bovine serum (FBS). Both cell types can differentiate into osteo-, adipo-and chondrocytes in vitro. However, 3D chondrogenesis in vitro of HPL-MSPC creates >3 times heavier cartilage fragments containing more hypertrophic chondrocytes than those generated with FBS-MSPC, suggesting that chondrogenesis is favoured by human platelet-derived factors. In a new model of spontaneous bone formation in vivo, HPL-MSPC from all 17 donors form bone through an endochondral mechanism after inclusion in Matrigel™ and subcutaneous implantation in immune-defi cient NSG mice, without requiring osteogenic pre-induction. Remarkably, >60% of these human ossicles attract mouse bone marrow, indicating that HPL-MSPC can establish a bone marrow-supporting niche. In contrast, only 20% of FBS-MSPC lines are able to form bone. Detailed phenotypic analysis reveals that the stem cell marker SSEA-4 is expressed at signifi cantly higher levels on HPL-MSPC than on FBS-MSPC, suggesting that MSPC stemness is maintained by humanised culture. Higher SSEA-4 expression of human MSPC in vitro also exquisitely correlates with attraction of mouse haematopoiesis in vivo. In addition, HPL-MSPC could be re-isolated and re-expanded to large numbers from mouse marrow-populated implants and formed bone again in secondary human ossicles, suggesting that stemness is also maintained in vivo. To elucidate the underlying mechanism, HPL-MSPC were treated in vitro with Imatinib, which inhibits PDGF-R gamma phosphorylation. This results in a drop of SSEA-4 surface expression similar to that observed in FBS cultures and in a loss of cartilage and bone differentiation potential in vivo.

Our data indicate that PDGF-R gamma signalling is chiefl y implicated in the maintenance of human MSPC stemness as well as bone and marrow niche forming competence during humanised expansion.

Methods: Human UCMSC were isolated, expanded and infused into 15 refractory SLE patients (NCT00698191, phase II). Clinical effi cacy was evaluated. The percentages of Treg and Th17 cells in peripheral blood were detected by fl ow cytometry. Peripheral cytokines were determined by ELISA and real time PCR, respectively. Peripheral blood mononuclear cells (PBMC) from 11 active SLE were co-cultured with UCMSC to examine the changes of Treg and Th17 cells and supernatant cytokines. Patients: naturally occurring Treg (nTreg) were isolated from peripheral blood and co-cultured with UCMSC in vitro in the presence of anti CD3/28 microbeads and recombinant IL-2. Results: Signifi cant clinical effi cacy was found by UCMSC transplantation. Peripheral blood CD4+CD25+Foxp3+T cells percentage showed obvious increase at 1 week (2.32%±1.40%), 1 month (2.03%±1.30%), 3 months (2.41%±2.06%) and 6 months (2.81%±0.69%) visits ( all p<0.05 vs pre-MSCT 1.58%±1.02%), in parallel with signifi cant decline of CD3+CD8-IL17A+T cells after MSCT (1.30%±1.02% at 1 week, 0.73%±0.60% at 1 month, 0.56%±0.60% at 3 months, 0.52%±0.40% at 6 months, all p<0.05 vs pre-MSCT 1.69%±1.13%). Real time PCR showed that gene expression of RORgammat decreased, while TGFbeta1 increased signifi cantly 1 week after transplantation. Plasma concentration of TNFalpha and IL17A decreased while TGFbeta1 increased. The co-culture of UCMSC with SLE PBMC resulted in a statistical increase of CD4+CD25+Foxp3+T cells and reduce of CD3+CD8-IL17A+T cells (all p<0.05 at UCMSC: PBMC 1:1, 1:10, 1:50 ratio), but without a dose dependent manner. Supernatant TGFbeta1 and IL6 increased but TNFalpha decreased signifi cantly in the co-culture system. The addition of TGFbeta1 or IL6 inhibitor to the culture system statistically abrogated the upregulation of Treg, while had no effect on Th17 cells. The absolute number of nTreg in the presence of UCMSC increased compared to that of nTreg alone, and showed signifi cant enhanced expression of Foxp3 and GITR. Additionally, CFSE dilution suggested that UC-MSC can obviously promote the proliferation of SLE nTreg in vitro. Conclusions: UCMSC transplantation markedly regulated Treg and Th17 cells in SLE patients, which may be one of the mechanisms for its therapeutic potential in refractory SLE. Objective: In experimental acute graft-versus-host disease (GvHD), adoptive transfer of natural FoxP3+ regulatory T cells (Tregs) has proven its potential to prevent a lethal outcome of alloreactive attack. However, it is so far not clear by which mechanism Tregs achieve this goal. Here we address the role of interferon-gamma for donor Treg-mediated GvHDsuppression. Methods: We used blocking anti-interferon-gamma mAb and interferon-gamma-defi cient (Ifng-/-) donor mice in Treg cotransplantation studies in a C57BL/6 into BALB/c mouse model of acute GvHD. GvHD severity was monitored by survival, clinical score and histological analysis. Results: We found that Tregs expressing the transcription factor FoxP3 produced large amounts of interferon-gamma during acute GVHD. Intracellular staining identifi ed alloreactive donor Tregs to be a major source of interferon-gamma after transplantation. In fact, more than 50% of allogeneic but not residual endogenous FoxP3+ Tregs produced interferon-gamma. Notably, blocking of Interferon-gamma with specifi c mAb led to death from GVHD. Furthermore, only wild type Tregs, but not Tregs from Ifng-/-donor mice could prevent fatal GVHD suggesting that Treg-intrinsic interferon-gamma production was required for their protective function. Conclusion: Our data imply that interferon-gamma, in particular in highly infl ammatory Th1 cytokine dominated environments such as acute GVHD, can be produced by natural Tregs and has essential immune-regulatory functions. Persisting antigen-presenting cells (APC) play a pivotal role as stimulators for donor-derived T cells in acute Graft versus host disease (GvHD). Strategies to attenuate GvHD by targeting host-APC (e.g. antibodies) are often species-specifi c and preclinical models are lacking. It has been demonstrated by others that functional human B cells can be established in mice reconstituted with human hematopoietic stem cells (HSC). Here, we focused on the establishment of further APC populations in mice reconstituted with human HSC. HSC were injected into sub-lethally irradiated NOD-scid gamma-/-(NSG) mice. Blood and skin samples were taken fortnightly. After 6 weeks, animals were sacrifi ced for obtaining bone marrow (BM) and spleen. Reconstitution of human APC was surveyed for up to 36 weeks by using different cohorts. The engraftment rate in 13 independent experiments was 70%. B cells, NK cells, monocytes, and HLA DR+/CD11c+ myeloid dendritic cells (DC) in BM and spleen were detected after 6 weeks. After 12 weeks, plasmacytoid DC were also present in BM. T cells were not found before week 18. We stained skin biopsies for the presence of human CD1a / CD207-positive Langerhans cells (LC) as surrogates for tissue-resident DC. Human LC were present in murine epidermis beyond week 6.

In an attempt to test whether the engrafted human APC function as targets for allo-reactive T cells, we generated naïve CD8-positive T-cell lines from partially HLA-matched third party donor by mixed lymphocyte cultures. In fi rst experiments, these T cells (10 7 cells/animal) were injected into mice that had been reconstituted with human HSC previously. Following transfer, the established human hematopoiesis was partially depleted. In mice 8 weeks after HSC-transfer, only B cells were reduced. In contrast, in mice 14 weeks after HSC-transfer, mDCs and monocytes were also depleted. These preliminary fi ndings indicate that beyond human B cells, monocytes as well as mDCs in NSG-mice engrafted with human HSC might function as targets for allo-reactive T cells. In summary, human APC engraft NSG mice and the engraftment of DC is not restricted to primary hematopoietic organs, as shown for epidermal LC. Allo-reactive CD8 T-cells deplete human APC in vivo which indicates their functionality as T-cell targets. Still, mDC and monocytes seem to develop their functionality later than B cells. To improve our mode, factors accelerating APC-development (e.g. FLT3) are currently under investigation.

Bronchiolitis obliterans (BO) after HSCT is a terrifying lateonset non-infectious pulmonary complication (LONIPC) leading to increased morbidity and mortality. Until now very limited therapeutic approaches are available and clinical course depends on diagnosis at an early stage. Of 1779 patients (pts) who underwent allo-HSCT at our center between January 2000 and December 2010 68 pts (39 male, 29 female; median age at transplant 45 years) were diagnosed BO at a median onset of 18 (range 7-178) months after HSCT. In this cohort 18 pts received grafts from HLA-identical siblings (26%), 32 pts from matched (47%) and 18 pts from mismatched (26%) unrelated donors. Transplants consisted of unmanipulated peripheral blood stem cells (n=63; 93%) or bone marrow (n=5; 7%). Based on pulmonary function impairment BO was characterized "severe" in 29 (43%), "moderate" in 29 (43%) and "mild" in 10 (15%) pts. Concomitant chronic GVHD was observed in 65 (96%) BO pts. The calculated 5-year cumulative incidence of BO was 6.8%. A total of 18 BO pts (26%) died corresponding to a calculated 5-year mortality of 22%. In the BO cohort pts younger than 30 years of age at HSCT and pts with a smoking history showed signifi cantly decreased 5-year survival (p=0,05 and p=0,039). A signifi cantly higher proportion of BO pts (69%) compared to non-BO HSCT recipients (51%) was transplanted from ABO incompatible donors. BO pts demonstrated a markedly transient dip of CD4+CD25+ T cells of median 35% of the previous levels at the time of BO onset. Blood quantitative measures of the hypoxia-inducible factor 1 (HIF-1a) expression were signifi cantly higher in BO patients compared to healthy controls (p=0,02) or chronic GVHD pts without lung involvement (p=0,007). Serum immunoglobulin G (IgG) levels were signifi cantly decreased at onset of BO compared to IgG values of HSCT recipients without BO at the corresponding post-transplant time points (5,5 g/l vs. 9,1 g/l, p=0,033). The median exhaled NO concentrations were lower in BO pts than in HSCT pts without BO. By using a threshold level of 15 ppb of exhaled NO a signifi cant discrimination between BO and non-BO pts (p=0,009) was enabled.

Our data demonstrate that BO is a serious LONIPC with substantial mortality. Certain cellular and molecular blood markers and proteins but also breath gas analyses might be useful for an early diagnosis. Background: In Australia, autologous haematopoietic stem cell transplantation (HSCT) is used to treat a growing number of patients with cancer each year, particularly patients of older age. While long-term survival has been studied extensively in allogeneic HSCT recipients, few studies have examined longterm survival in autologous HSCT recipients. Aim: To evaluate rates of mortality in 2-year survivors of autologous HSCT. Method: Over 13000 HSCT recipients treated from 1992-2007 in Australia were assembled for the CAST study. Deaths following HSCT were identifi ed from the Australasian Bone Marrow Transplant Recipient Registry and through data linkage with the National Death Index. Overall survival estimates and standardised mortality ratios (SMRs) were calculated for 4547 patients aged 15 or older who survived at least 2 years after autologous HSCT. Results: By the end of 2007, 1466 deaths had occurred among 2-year survivors of autologous HSCT. Approximately 70% of deaths were due to relapse or progression of disease. Relapse remained the most common cause of death beyond 10 years from HSCT. Overall, the survival probability was 56% at 10 years from HSCT. 10-year survival probabilities by diagnosis ranged from 34% for patients with multiple myeloma to 90% for patients with testicular cancer. In the interval 2 to 5 years from HSCT, observed mortality was more than 17 times higher than expected based on rates in the age-and sex-matched Australian general population (SMR=17.9, 95%CI=16.9-19.0). While observed mortality moved closer to expected mortality in the intervals 6 to 10 years (SMR=10.2, 95%CI=9.1-11.5) and 11 or more years from HSCT (SMR=5.9, 95%CI=4.2-8.4), observed mortality remained signifi cantly increased across all periods examined. For all diagnoses, rates of observed mortality were signifi cantly greater than expected in the general population. SMRs by diagnosis ranged from 7.4 (95%CI=5.6-9.7) for patients with acute myeloid leukaemia to 35.9 (95%CI=32.8-43.0) for patients with breast cancer. Conclusion: Mortality in 2-year survivors of autologous HSCT is signifi cantly increased compared to the Australian general population, even beyond 10 years from HSCT. With the majority of deaths occurring from relapse, prevention of disease recurrence is the greatest challenge for autologous HSCT recipients. SCT is associated with substantial mortality during the fi rst 2 years after SCT whereas afterwards survival often reaches plateau. QoL is increasingly recognized as an important longterm end-point. The pattern of late events and QoL has been S52 reported following myeloablative conditioning (MAC) but is not well defi ned in reduced-intensity (RIC) setting. To explore late outcomes we retrospectively analyzed SCT results among 726 patients (pts) transplanted between 1/2000 and 8/2009. 246 pts were alive and disease-free 2 years after SCT. Median age was 51 years (17-72). Diagnoses included AML/MDS (n=131), ALL (n=24), lymphatic diseases (n=48), CML/MPD (n=29), nonmalignant (n=14). Conditioning was MAC (n=72), RIC (n=118) or reduced-toxicity myeloablative conditioning (RTC, n=56). 172 pts had cGVHD, graded moderate-severe (mod-sev) in 44 and 29% of pts after MAC and RIC/RTC, respectively (p=0.03). 68 and 43% were still on immune suppressive therapy (IST) 2 years after SCT, respectively (p=0.001). With median followup of 68 months after SCT (25-140), the probability of pts surviving disease-free 2 years after SCT to remain disease-free for the next 5 years was 84 and 82% after MAC and RIC/RTC, respectively (p=NS). There were 35 deaths beyond 2 years, 15 due to relapse and 20 due to non-relapse causes (NRM). The cumulative incidence of NRM was 7% after both MAC and RIC/RTC. The cumulative incidence of relapse was 9% and 11%, respectively (p=NS). Age>55 and mod-sev cGVHD predicted shorter survival, HR 2.1 (p=0.07) and 2.6 (p=0.006), respectively. Mod-sev cGVHD predicted NRM, HR 5.2 (p=0.001). Advanced disease predicted relapse, HR 2.6 (p=0.004). The probability of stopping IST by 8 years was 59 and 75% after MAC and RIC/RTC, respectively (p=0.001). Median IST duration was 30 and 20 months, respectively (p=0.05). QOL was assessed by the EORTC QLQ-C30 questionnaire. Mean QoL score was 69, 66 and 65 after MAC, RIC and RTC, respectively. Low QoL score was reported by 15, 14 and 19%, respectively (p=NS). Continuous need for IST and depression correlated with a low score while healthy lifestyle (return to work, physical and sexual activity) and academic education correlated with high scores. In conclusion, late outcome is similar after MAC and RIC/RTC. Late NRM is similar although cGVHD is less severe and IST duration shorter after RIC/RTC. Pts who are disease-free 2 years after SCT, particularly younger pts with no mod-sev cGVHD can expect good survival and QoL. Background: RIC allo-SCT is increasingly used in elderly or frail patients (pts) not eligible for myeloablative conditioning. While the natural history of LTC and QOL are rather well described in the myeloablative allo-SCT setting, data is still sparse in the RIC setting. This single centre survey analyzed the outcome and features of LTC and QOL in 110 RIC allo-SCT pts who survived for a minimum of 2 years (y) after allo-SCT. QOL was assessed in a cross-sectional study using the EORTC QLQ-C30 questionnaire and the FACT-BMT questionnaire. Results: The K-M estimate of overall survival (OS) was 93% (95%CI, 88-99%) and 81% (95%CI, 71-94%) at 5 and 10 y, respectively. The primary cause of death was relapse found in 4 pts (4%). Six pts (5%) died of non-relapse-related causes as secondary malignancies (n=3), chronic GVHD (cGVHD; n=2) and infection (n=1). With a median follow-up of 4.6 y (range, 2-12.1), cGVHD was the most prevalent LTC with a cumulative incidence (CInc) of 66% (95%CI, 57-74) at 10 y. In multivariate analysis, a mismatched unrelated donor allo-SCT was the strongest independent risk factor associated with the development of LTC (RR=4.06, 95%CI 1.81-9.10, p=0.002).

In this series, 61 pts (55%) responded to the different questionnaires. Overall, in the EORTC QLQ-C30 questionnaire 70% of these pts reported a good to very good QOL, with a mean global QOL group score of 71.7 (SD, 20.8). Compared to the group without cGVHD, pts with cGVHD had signifi cantly lower QOL. Similarly, in the FACT-BMT questionnaire the scores in all subscales indicated good QOL. However, there was statistically signifi cant differences between the two groups in average QOL scores in the physical and functional well-being and BMT subscales, as well as for the FACT-G, FACT-TOI and FACT-BMT total composite scores. Moreover, the QOL survey revealed that 29% of working age-patients could return to full-time work. Finally, 2 childbearing age women (18%) were reported to have had pregnancies. Conclusion: Pts who are alive beyond 2 y after RIC allo-SCT have a high probability of cure with excellent OS and good QOL. However, cGVHD remains an issue requiring long-term appropriate psychological support. Also, the natural history of LTC after RIC appears to be different from that described in the standard myeloablative setting, warranting more research in this fi eld and lifelong surveillance through a close partnership between the transplant center and organ-specifi c specialities. Secondary malignancies are a known complication in longterm survivors after SCT. The incidence has been reported mostly after myeloablative conditioning (MAC). Fludarabine (F)-based RIC and RTC are associated with lower incidence of early SCT-related complications, however due to the so far relatively limited long-term follow-up, the incidence and riskfactors for secondary malignancies following these regimens is not well defi ned. We retrospectively reviewed a single institution database of 931 allogeneic SCTs given over 12 years to identify patients (pts) with secondary malignancies. Conditioning regimens included standard MAC (n=257, TBI-based in 111), F-based RIC [n=449, including F with intermediate-dose busulfan (FB2) and F with melphalan (FM)] or F-based RTC [n=225, including F with high-dose busulfan (FB4) and F with treosulfan (FT)]. 21 pts had secondary malignancies including squamous cell carcinoma of the skin (n=5), penis (n=1) vagina (n=1), tongue (n=1) and esophagus (n=2), colon cancer (n=3), breast cancer (n=2), pancreatic cancer (n=2), metastatic cancer of unknown primary (n=1), melanoma (n=1), metastatic sarcoma (n=1), Kaposi sarcoma (n=1). Median age at SCT was 53 years (29-70). 19 pts were given F-based RIC/RTC and none had TBI. The median time from SCT was 49 months (7-138). 17 pts had prior chronic GVHD, 10 moderate-severe, 14 were still on immune-suppression at diagnosis of secondary malignancy. The 10-year cumulative incidence was 5.3% (95%CI, 3.2-8.7%). It was 2.2%, 8.8%, 5.5% 7.8% and 5.5% after BuCy, FB2, FB4, FM and FT respectively (p=0.05 for BuCy Vs. RIC/ RTC). Multivariate analysis identifi ed RIC/RTC and moderatesevere chronic GVHD as adverse prognostic factors with HR 5.9 (p=0.03) and 2.7 (p=0.04), respectively. Pts were treated with surgery for localized tumors and with chemo-radiotherapy or palliative therapy for metastatic disease. Currently, 17 pts are alive and 4 have died, all of them had advanced solid tumor at presentation. 10-year cumulative incidence of death due to secondary malignancy was 1.6%. In conclusion, secondary malignancies are rare but signifi cant complication after allogeneic SCT. Curative approach is feasible in a subset of pts. The combination of F and intermediate to high-dose alkylating agents may be associated with higher risk for second malignancies than standard therapy with high-dose alkylator and/or TBI. Larger studies with a larger number of events are needed to confi rm these observations. Background: Metabolic syndrome (MS) is diagnosed in up to 23% of patients treated with haematopoietic stem cell trans-plantation (HSCT) and 30-40% of children with obesity. The purpose of the study was to determine the genetic background of the syndrome in children treated with HSCT. Patients and Methods: Thirty patients (pts.) 1,5-19,4 (average 9,9) years old, 24 boys and 6 girls, diagnosed with ALL-15, ID-5, AA-4, AML-4, MDS-1, NBL-1 treated with HSCT (autologous-1) according to EBMT protocols were included to the study. The blood was collected before and 7 months in average after HSCT (19 pts.). The control group was composed of 29 children 3-18 (average 13) years old, 15 boys and 14 girls with obesity. There were no signs and symptoms of the primary disease in children after HSCT, all of them completed immunosuppressive treatment. BMI before HSCT was 13,3-27,4 (average 19,1), BMI percentile 0,1-98,3 (average 56,2). BMI after HSCT was

S54 9,7-28 (average 18,7), BMI percentile 0,1-87,8 (average 51,3). BMI in the obesity group was 21,8-38,4 (average 31,3), BMI percentile 94,2-100 (average 98). Blood samples (0.5 ml) were drawn from all the study participants. Subsequently, total RNA extraction was performed and microarray analysis was conducted with GeneChip Human Gene 1.0 ST Arrays (Affymetrix) according to the manufacturer's protocol. We decided to compare the patient without overweight treated with HSCT (low risk of activation of genes responsible for MS development) with children with obesity (high risk of activation of genes responsible for MS development). Results: BMI percentile <85 was noticed in 20 and 15 patients before and after HSCT respectively. Among 200 signifi cantly differentially expressed genes when comparing children before HSCT with BMI<85 with control group, 108 were underexpressed and 92 overexpressed in the control group. Only higher than 1.5 (Fold Change) differences in gene expression were analyzed. Among others, the IGFBP2 gene was underexpressed in the control group. This gene encodes IGF-binding protein, which inhibits IGF-mediated growth and developmental rates. This relationship was not observed after HSCT. Conclusions: As there is now substantial evidence linking IGFBP-2 with nutritional status and insulin sensitivity, changes in expression of IGFBP2 gene in patients after HSCT may be involved in development of metabolic syndrome after HSCT. Study sponsored by national grant number K/PBW/000520. Long term survivors of SCT are at increased risk from cardiovascular (CV) disease. We evaluated CV risk in 109 recipients (62 males, 47 females) surviving >5 yr after SCT. Ninety-nine patients received ≥12 Gy TBI conditioning and 10 received non-myeloablative conditioning followed by bone marrow (15) or T cell depleted peripheral blood SCT (94). SCT indications were CML (56), acute leukemia (30), MDS (13), and others (10) . Median ages at SCT, 5 and 10 yr follow-up were 34, 40 and 46 yrs, respectively. Ninety-four individuals developed cGVHD and 36 required immunosuppressive treatment for >3 yrs. Survivors were followed up at 5, 7, 10, 15 yrs postSCT and had >90% Karnofsky performance status at the time of analysis. PostSCT CV events were uncommon: 2 survivors had percutaneous coronary interventions and 1 had recurrent pulmonary emboli. Serial electrocardiograms revealed 6 new ischemic changes and 28 new rhythm/conduction abnormalities. Serial echocardiograms showed no change from baseline. At 5 and 10 yrs postSCT respectively, 44% and 52% of survivors had abnormal lipid profiles. Despite being younger, 23% of survivors met the Adult Treatment Panel III threshold for dyslipidemia treatment at 5 yrs post SCT, which is higher than the 15% at 45-54 yrs in the general population (Multi-Ethnic Study of Atherosclerosis). There were signifi cant increases in prevalence of hypertension (p<0.001), diabetes (p=0.018), elevations in C-reactive protein (CRP) (p<0.001) and body mass index (p=0.044) postSCT compared to baseline. The Framingham general CV risk score of males at 5 yrs postSCT predicted a 10 yr risk of CV events -double the expected norm (median 10.4% vs. 5.4%). The calculated median heart/vascular age was 8 yrs ahead of patients' chronological age. Females received hormone replacement therapy postSCT and none had increased CV risk scores. To explore CV risk factors, we examined presence and persistence of cGVHD beyond 3 yrs and CRP as an infl ammatory marker. cGVHD was not associated with CV risk at any time point but CRP correlated signifi cantly with the CV risk at 5 yrs postSCT (p=0.02). All risk factors stabilized between 5 and 10 yrs postSCT. These studies highlight a signifi cant increased risk of CV events in male SCT recipients in their second and third decade after SCT.Identifying mechanisms underlying the infl ammatory processes associated with CV risk would be a fi rst step in developing effective preventative management of heart disease in SCT recipients. Background: Four patients under radiotherapy for prostate cancer were accidentally overexposed to irradiation which led to severe therapy-resistant diarrheal and painful hemorrhagic colitis and recto-vesical/recto-prostatic fi stula. Mitigation of lethal intestinal injury, following high doses of irradiation, can be achieved by injection of Mesenchymal Stromal Cells (MSC). MSC accelerate functional recovery of the intestine and dampen the systemic infl ammatory response by reducing the levels of proinfl ammatory cytokines in radiation-induced gastrointestinal syndrome. Objective: To provide compassionate therapy for refractory irradiation-induced colitis and diminish pain, haemorrhage and diarrhea, allogeneic bone marrow (BM)-derived MSC from family donors were intravenously infused to four patients. Material: MSC were obtained by culture from BM aspirates of their children. Three patients received 5x10 6 /kg, 2.6x10 6 /kg and 5x10 6 /kg MSC, respectively, by intravenous infusion. Before, one month and six months after MSC therapy the clinical parameters pain, haemorrhage and fi stulisation (MRI, colonoscopy) were evaluated. One patient was retreated (0.8x10 6 /kg of MSC) after 6 months. Recently a fourth patient was treated by two MSC injections (1.85x10 6 /kg and 2.4x10 6 /kg one week later). In this patient, lymphocyte subsets and frequency of diarrhea were monitored. Results: No toxicity occurred. After MSC therapy two patients revealed a substantiated clinical response for pain and hemorrhage. In one patient pain reappeared after 6 months and again substantially responded on a second MSC infusion (Figure 1) . In one patient a beginning fi stulisation process could be stopped. In this case a stable remission has been observed for more than 3 years of follow-up. The fourth patient underwent two injections of MSC: The frequency of painful diarrhea diminished from 5-7 per day to 3 per day after the fi rst MSC injection and stabilized at 2-3 per day after the second injection. In parallel, after the MSC injections, an increase of NK cells and B lymphocytes and a decline of CD4+ and CD8+ T lymphocytes was observed (Figure 2 ). Prostate cancer remained in stable complete remission in all patients. Conclusion: In four patients with refractory irradiation-induced colitis systemic MSC therapy was safe and effective on pain, diarrhea, haemorrhage and infl ammation. MSC may represent a safe therapeutic measure for patients under radiotherapy overirradiated.

Platelet-lysate-expanded mesenchymal stromal cells for the treatment of steroid resistant GvHD: clinical results and biological readouts in a multicentre study G. Lucchini (1) 

Objective: Mesenchymal stromal cells (MSC) are multipotent cells with broad immunomodulatory properties. Their role in the treatment of graft versus host disease (GvHD) has been intensively investigated. The present study reports the outcome of a cohort of patients (pts) receiving MSCs as part of GvHD treatment. Methods: Pts exhibiting acute or chronic grade II-IV steroidresistant GvHD after allogeneic stem cells transplantation were eligible to the study. Enrolled pts received third-party, bone marrow derived, platelet lysate-expanded MSCs for a minimum of 3 doses. Median dose was 1x10 6 /cells per kg body weight. The suggested infusion interval was 5-7 days. Clinical response to treatment was assessed 28 days after the last MSC infusion. GvHD was graded according to NIH criteria. Before MSC infusion and weekly till day + 28 after the last MSC infusion, blood samples were collected to analyse lymphocyte subsets and GvHD activity biomarkers. Results: From August 2009 to September 2011 23 pts (13 adults, 10 pediatric) aged 1-58 years received MSCs. GvHD presented as acute in 20 cases and chronic in 3, it involved a single organ in 12 pts and multiple organs in 11 cases. GvHD was defi ned as grade II in 7 pts, grade III in 12 and grade IV in 4 pts. In this series overall response to MSC infusion was 60,8%, and complete response was 17,3%. None of the pts affected by chronic GvHD benefi ted from the treatment. Both skin and gut GvHD presented a good response rate, skin showing an earlier response (2-4 days) compared to gut (5-7 days). Pts with multiple organ involvement and higher GvHD grade seemed to show a worse response to treatment. 12 out of 23

S56 treated pts are alive; 5 pts died from GvHD complications, 3 pts died from infectious events, and 2 relapsed. Immunobiological data support the idea that MSC are able, in responding pts, to increase T regulatory cells, as demonstrated by the progressively decreasing in Th1/Treg andTh17/Treg ratio. Biomarkers of GvHD activity correlated with MSC response. Conclusions: The present study underlines the safety of PLexpanded MSC use in children and adults. MSC effi cacy seems to be greater as early as the administration is performed, in acute rather than in chronic GvHD, as well as in milder from of the disease involving skin and gut, particularly in pediatric pts. Biological readouts corroborate the clinical observations and hint at the ability of MSCs to increase T regulatory cells as mechanism of action against GvHD. CMV, Adenovirus (Ad) and EBV are viral pathogens causing morbidity and mortality in patients after HSCT and cord blood transplantation (CBT). We have shown that adoptive immunotherapy with peripheral blood (PB) donor derived multivirusspecifi c Cytotoxic T Lymphocytes (mCTLs) directed against EBV, CMV and Ad can effectively prevent and treat the clinical manifestations of these viruses after HSCT. We have now extended these studies by expanding mCTL from CB to restore cellular immunity to CMV, EBV and Ad simultaneously after CBT. The development of mCTLs for patients undergoing CBT requires the priming and extensive expansion of naïve T cells rather than the more limited and simple direct expansion of pre-existing memory T cell populations from virus-experienced donors. We have developed a novel protocol utilizing an initial round of stimulation with autologous CB-derived dendritic cells transduced with a recombinant Ad5f35 vector carrying a transgene for the immunodominant CMV antigen, pp65 (Ad5f35pp65) in the presence of IL-7, IL-12 and IL-15. This is followed by 2 rounds of weekly stimulation with autologous Ad5f35pp65transduced EBV-LCL in the presence of IL-15 or IL-2. So far we have enrolled 7 patients and treated 4 patients on this phase I study where the CTLs were generated from the 20% fraction of a fractionated CB unit. All 7 patients engrafted neutrophils and platelets <30 and <60 days post CBT respectively. CTL were infused to 4 patients from 63-146 days post CBT at doses of 5x10e6/m 2 to 1x10e7/m 2 . No infusion-related toxicities or GvHD was observed. Within 2 weeks of CTL, all patients had detectable EBV-specifi c T cells in their PB that persisted >1 year post CTL. Three patients remain free of CMV, EBV, and Ad reactivation up to 1 year post CBT. One patient was transiently viremic for Ad and CMV but after 2 doses of CTL permanently cleared the CMV and Ad from the PB and stool respectively. He remains asymptomatic and virus free >1 year post-CBT. Analysis of this patient's PB showed a concomitant rise in CMV and Ad-specifi c T cells by 4 weeks. Further, using TCR deep sequencing we have shown that T cell clones present in the CTL lines of all patients, but not in their PB pre infusion, expand up to 14 fold post infusion. In summary, administration of low doses of CB derived virus-specifi c CTL to patients after CBT has so far been safe and can facilitate reconstitution of virusspecifi c T cells and control viral reactivation/infection.

Outcomes of CD19-directed multivirus specifi c cytotoxic T lymphocyte therapy for patients with relapsed B cell malignancies after allogeneic haematopoietic stem cell transplantation C. Bollard (1) 

Allogeneic hematopoietic stem cell transplant (HSCT) may increase long term disease-free survival in patients with Bcell malignancies, but delayed immune reconstitution associated with the procedure is associated with viral infections and disease relapse. We hypothesized that a single T-cell platform mediating both antiviral and antileukemic activity could benefi t these patients. We prepared CTL with specifi cities through native receptors directed towards EBV/CMV/adenovirus (Ad), then engineered them to express a chimeric antigen receptor (CAR) targeting CD19. We used donor-derived antigen presenting cells expressing Ad antigens and transgenic CMVpp65 following transduction with the Ad vector Ad5f35CMVpp65. Multi virus (MV)-specifi c CTL were then transduced with a retroviral vector encoding CAR-CD19.28zeta. Safety:6 patients were infused with 1.5 to 4.5x10e7 cells/m2 without infusion related toxicity. One CLL patient developed fever, diarrhea, hypotension 4 weeks post CTL. Findings were consistent with ileitis at a known site of disease. Biopsy of the gut showed an absence of normal and malignant B cells, and the presence of CAR-CD19.28z T cells. Persistence: There was a predictable decline of T cells in peripheral blood (PB) following infusion. However, persistence is documented by their presence in disease sites (GIT and bone marrow) up to 9 weeks. Anti-viral activity: Patient 1 had Ad positivity in stool, which resolved without antiviral treatment. No other patient developed viral infections post CTL. Anti-tumor activity: Of the two patients with Ph+ ALL, one patient had 4% blasts detectable in the PB at the time of CTL which cleared within 2 weeks post CTL. She subsequently relapsed and then died of disease 7 months post CTL. The other patient had a transient decrease in blast count to 0.2% but ultimately progressed after 4 weeks. 4 CLL patients have been treated. The fi rst CLL patient had resolution of lymphadenopathy within 2 weeks but following the disappearance of CTLs from PB, progressed and died after 2 months. The second CLL patient has had stable disease for over 6 months with an infl ux of T-cells in his bone marrow. The remaining 2 CLL patients are still early but by 6 weeks both have had reductions in their CLL counts without toxicity. These results provide encouraging evidence of the safety, persistence, and effectiveness of monoculture CD19CAR multivirus T cells in the treatment of high risk B-cell malignancies post allogeneic HSCT. Only one patient suffered from grade III skin and gut acute GVHD. We observed 6 CR (5 AML, 1MM), 6 PR (3 AML, 2 MM and 1 HD) and 4 NR (4 AML) and 2 PD (1 NHL, 1 MM) in the 18 evaluable patients. Overall disease response rate was 66% and 78% at the highest dose level. With a median follow up of 270 days (range 80-712), 13 patients (72%) are alive and of those 7 (54%) in CR while 6 (46%) with active disease; 5 (28%) patients died, 4 for PD and one of fungal lung infection. One year OS and PFS were of 70% and 48%, respectively. Conclusions: The sequential infusion of low dose DLI and CIK cells proved feasible, safe and effective with an MTD that has not been reached. We are now looking at further increase the total amount of CIK cells infused to enhance the overall clinical activity of this cellular therapy approach. Background: Decompensated alcoholic liver disease carries a poor prognosis, due in part to an impaired liver regeneration. We explored in a randomized trial the potential of G-CSF stimulated autologous bone marrow stem cells (aBMT) to improve liver insuffi ciency over a 3-month period after clinical decompensation. Patients/Methods: 58 patients (M/F: 34/24, mean age 54 yrs) admitted for decompensated alcoholic cirrhosis (mean MELD 19) and alcoholic steatohepatitis were randomized early after admission to 2 treatment arms. Standard medical therapy (SMT) alone (n=30), was compared to combined SMT with aBMT (n=28). aBMT included a 5-day G-CSF (10 mcg/kg sc daily) mobilization prior to BM aspiration (~90 ml) under propofol sedation, BM MNC isolation (Ficoll®) and reinfusion within the same day in the hepatic artery. A second liver biopsy was performed at 4 weeks to assess histological changes. The MELD score and clinical outcome were determined at 1, 2 and 3 months of followup. The primary endpoint was a decrease of > 3 on the MELD score at 3 months as compared to the initial score. Results: Patients characteristics were similar at baseline. BM could be administered in all but 2 patients. BM infused contained a mean of 0.47x10e8 MNC/kg. During follow-up, 6 patients died (2 in aBMT group, 4 in SMT group), and 9 patients (aBMT: 4; SMT: 5) resumed alcohol consumption. Adverse events were equally distributed in both groups, unrelated to aBMT. The primary endpoint was reached in 19/28 (66%) aBMT patients, and 17/30 (57%) SMT patients (see table) .This difference was not statistically signifi cantly different. Regenerations markers at histology and in serum showed similar evolution over time. Conclusions: Autologous BMT in vulnerable patients hospitalized for decompensated alcoholic liver disease was well tolerated but did not improve liver function nor stimulated liver cell proliferation over a 3-month period. These is in contrast but not in opposition with our previous study in which repetitive G-CSF injections without aBMT infusion were associated with hepatic progenitor proliferation at day 7. Thus long-term liver regeneration is more diffi cult to achieve than short-term hepatic progenitor stimulation, which is readily achievable with growth factor administration. The present results suggest either an insuffi cient stimulus or a resistance of the diseased liver to regenerate. S58 reported. We report here preliminary results of a pilot study of myeloablative dual UCB transplantation where one UCB unit is expanded ex vivo using the NiCord technology. The NiCord UCB graft consists of an expanded AC133+ and a re-frozen unexpanded AC133-fraction. HPCs were expanded for 21 days in media containing hematopoietic cytokines supplemented with nicotinamide (NAM), a vitamin B3 derivative that delays differentiation and improves engraftment effi ciency of human HPCs. Patients were conditioned with TBI (1350cGy) and fl udarabine 160 mg/m 2 . GVHD prophylaxis consisted of tacrolimus and mycophenolate mofetil. Five patients (med age 57; range 41-61) with AML (n=3) or MDS (n=2) received a NiCord-containing dual UCB graft. The unmanipulated unit was HLA 4/6 (n=4) or 5/6 (n=1) matched with the patient and contained a median cryopreserved TNC dose of 3 x 10 7 /kg (range 2.3-3.9). The expanded unit was 4/6 matched (n=3) or 5/6 HLA-matched (n=2) with the patient and contained a median cryopreserved TNC dose of 2.2 x 10 7 /kg (range 1.7-3.4). After expansion of the AC133+ cell fraction, the TNC and CD34+ cell dose increased a median 486 fold (range 321-577 fold) and 82 fold (range 46-100 fold), respectively. Patients received the NiCord unit that contained a median TNC and CD34+ cell dose of 2.5x10 7 /kg (range 1.0-4.3) and 3.0x10 6 /kg (range 1.6-3.9), respectively. The transplanted negative fraction contained a median of 1.3 x 10 6 CD3+ T-cells/ kg (range 0.8-2.8). All patients are durably engrafted, with a median time to ANC>500 of 14 days (range 11-26) and platelets >20,000 of 33 days (range 30-49). Peripheral blood CD3+ and CD15+ cell chimerism from the 5 patients at the time of last followup are reported. Pt. 1, day +280; CD3 (NiCord-2%, unmanipulated-80%, host-18%), CD15 (NiCord-64%, unmanipulated-36%). Pt. 2, day +180; CD3 (NiCord-100%), CD15 (NiCord-100%). Pt. 3, day +120; CD3 (NiCord-92%, host-8%), CD15 (NiCord-90%, host-10%). Pt. 4 and 5, day +60; CD3 (unmanipulated-100%), CD15 (unmanipulated-100%). There were no infusion-related or unexpected adverse events and all patients are alive and well without GVHD. These data demonstrate prompt and prolonged engraftment of both myeloid and lymphoid cells derived from expanded HPCs (NiCord) in 3 of 5 patients treated to date. Longer follow-up is required to confi rm durability. Background: EBMT is a leading society in the fi eld of hematopoietic stem cell transplantation that aims, among other things, to foster clinical research among the affi liated centres.It would help the EBMT to identify potential areas of interest for cooperative research if it was known which topics its own members fi nd important enough to investigate in clinical studies. Objectives: To assess the type and quality of investigator contribution to EBMT 2012 in the fi eld of infectious diseases. Methods: The abstracts were categorized as follows: type (prospective, retrospective, case report, meta-analysis), origin (single centre, multicentre), population (adult, paediatric, adult and paediatric), type of transplant (autologous, allogeneic, autologous and allogeneic), infection (bacteria, fungi, virus, other), and main objective (epidemiology, diagnosis, prophylaxis, therapy, adoptive immunotherapy, immunity/immune recovery, risk factor/survival, other).

Results: 77 (8%) of the 930 abstracts submitted by physicians (8%) were had infectious complications as their topic. 45 (58%) were retrospective studies, 23 (30%) were prospective studies, 7 were case reports, and there was 1 in-vitro study and 1 metaanalysis. Only 2 abstracts reported the results of multicentre studies, 1 from EBMT-IDWP and 1 from GITMO. 36 abstracts (47%) concerned adults, 17 (22%) were focused on paediatrics, 8 on mixed populations and the population was not defi ned in 16 abstracts (21%). 52 abstracts (67.5%) focused on allogeneic transplant, 5 (6.5%) on autologous transplant, 11 (14%) on allogeneic/autologous transplant, and the or was not applicable in 1 abstract. Viral infections were studied in 37 (48%) abstracts, fungal infections in 18 (23%), bacterial infections in 8 (10%), and various infections in 10 (13%) with pneumonia and vaccination being the focus of 2 and 1 abstracts, respectively. (1 was not applicable) The objective of 11 (14%) abstracts was epidemiology, 11 (14%) risk factors/survival, diagnosis 13 (17%), prophylaxis 16 (21%), immunity/immune recovery 9 (12%), therapy 9 (12%), adoptive immunotherapy 2 (3%,) the remaining 6 having miscellaneous objectives. Conclusion: Most abstracts reported retrospective studies involving single-centres, with viral and fungal infections representing the most frequent areas of interest. This information should prove useful in helping plan EBMT-wide prospective, multicentre studies. . The Conference objective was to elaborate guidelines for the prevention, diagnosis and management of infectious complications arising during the treatment of leukaemia and following hematopoietic stem cell transplantation (HSCT). Since then, ECIL meetings have been held every 2 years, the last one (ECIL4) being held in September 2011. On each occasion, 50 to 60 experts from the fi elds of haematology, oncology, infectious diseases, microbiology and clinical trials from 28 countries -mostly European, and also Israel and Australia -were invited to participate. Each presentation and set for guidelines were prepared by a designated working group, who adopted the quality of evidence and strength of recommendation grading system of the Centers for Diseases Control and presented their proposals during a plenary session at the ECIL meeting. These were then modifi ed according to the discussions that followed until a consensus was reached. This was then made available to the medical community as a slide set on each of the web sites of the organizations involved: www. ebmt.org. www.eortc.be, www.ELN.org and www.ichs.org. Since ECIL1, 4 bacterial topics, 6 fungal topics, and 3 viral topics have been addressed, and 14 papers have been published or are in press. Additionally, as guidelines by defi nition have rather short life-expectancies, most guidelines, especially the fungal ones, have been updated on a 2-year basis. One of the objectives of ECIL is also to identify new areas of research. For example, at the ECIL4 meeting held in September 2011, 3 new topics were addressed: (1) Management of respiratory virus and adenovirus infections (2) Bacterial resistance in haematology. (3) Specifi c consideration on fungal disease and antifungal treatment in children. The core of the guidelines on these 3 topics will be presented at the IDWP session. We hope that the ECIL guidelines will help HSC transplanters to make rationale, evidence-based choices when managing the infections their patients encounter. Results: The overall EBV-PTLD incidence was 3.2%, and it ranged from 1.2% in MFD-SCT, 2.9% in MMFD/haplo-SCT, 4% in MUD-SCT, 4.1% after CB-HSCT, and to 11.2% in MMUD-SCT. EBV-PTLD occurred at median of 2 months (range, 0.2-81 months) after HSCT. PTLD resolution after rituximab was observed in 69% of patients. By multivariate analysis, risk factors infl uencing the response of PTLD to rituximab were age >30 years and involvement of extra-lymphoid tissue. In the prognostic model, an increasing number of risk factors worsened the outcome: 100-days PTLD survival with 0, 1 or 2 factors: 93%, 66% and 39%, respectively (P<0.001). Immunosuppression tapering was associated with a higher proportion of survival after PTLD (86% v 60%). Chemotherapy did not improve survival in no-RI arm. An increase of serum EBV-DNA-emia during therapy was predictive of poor survival after PTLD, while a decrease after one week predicted good outcome. Conclusions: Over two-thirds of patients with EBV-PTLD responded to treatment with rituximab. Age over 30 years, extra-nodal disease, and absence of early response to rituximab are associated with a poor prognosis. Patients and Methods: 8 patients (ALL-2, SAA-2, AML-2, ABL-1, ES-1), aged 5-19 years, were included to the study. PTLD occurred 22-153 days after allo-HSCT (4 MUD, 3 MMUD, 1 MFD). Diagnosis of PTLD was based on clinical symptoms and quantitative EBV-DNA load in serum and in cerebro-spinal fl uid (CSF), obtained by PCR. Therapy of PTLD: weekly rituximab 375 mg/m 2 iv with simultaneous ITH administration until resolution of clinical symptoms and EBV-DNA-emia negativity. Rituximab in dose 10-30 mg, solved in 3-10 ml of 0.9% NaCl was administrated ITH for 2-10 minutes, 1-12 times in 7 days intervals. Results: All patients had systemic PTLD with CNS involvement. Clinical symptoms of CNS involvement were presented in 7 of 8 patients as symptoms of encephalitis: nausea, fever, loss of memory, consciousness disturbances, fatigue, loss of appetite. In 1 patient, unusual fatigue was the only symptom of CNS involvement. Other symptoms of systemic PTLD included: lymphoid tissue involvement (n=6), hepatomegaly (n=4), and fever (n=5). MRI was performed in 3 patients and revealed: (i) 5x6 mm focal change in occipital cortex of brain in one patient, (ii) 1.5 cm tumor-like change located between pons and cerebellar peduncles, (iii) cortical atrophy. Serum EBV-DNA-emia at diagnosis: 3.44x10E5-7.5x10E9 gc/mL. CSF EBV-DNA load at the diagnosis: 4.10x10E3-3.77x10E7 gc/mL. CSF leukocytosis 2-36/mcL, protein 34.4-191 mg/dL. In 6 patients serum EBV-DNA-emia was negative after 2-5 rituximab doses. CSF EBV-DNA negativization was observed after: 1 rituximab dose in 3 patients, 2 rituximab doses in 2 patients, 3 rituximab doses in one patient and in one patient after 8 rituximab doses. Five patients achieved complete remission, 2 patients achieved partial remission (1 patient without symptoms, slight progression changes in MRI, CSF EBV-DNA positive; 1 without clinical symptoms, tumor-like focus observed in MRI, CSF EBV-DNA negative). One patient has died after clinical progression of PTLD, although EBV-DNA serum load decreased by 2 logs after one-week therapy. Seven patients had good tolerance of rituximab ITH without post-dural puncture symptoms. Adverse events were observed in one patient: seizures during third rituximab ITH infusion. Conclusions: Rituximab intrathecal administration is a safe and effective method of therapy of CNS involvement of PTLD. Further studies of this treatment modality are necessary. In 2009, mutations in the gene for dedicator of cytokinesis 8 (DOCK8) have been identifi ed as the cause of the autosomal recessive variant of Hyper-IgE syndrome. The clinical presentation of this primary combined immunodefi ciency is characterized by eczema, debilitating viral infections of the skin, chronic mucocutaneous candidiasis, pulmonary infections, severe allergies, vascular complications and a high risk for malignancy. While the long-term prognosis of affected patients is not yet clearly defi ned, the high morbidity and mortality of this disease suggest HSCT as a potential curative measure. Based on data from the ongoing DOCK8 therapy survey, we retrospectively studied the outcome of HSCT in patients with DOCK8 mutations. A total of 22 patients from 11 institutions were identifi ed, 4 of whom had been previously reported. Three patients had malignant disease at the time of transplantation (2 cutaneous T-cell lymphomas, 1 EBV-LPD). At a median age of 11 years (3-18) transplantation from a MUD (n=10), MFD (n=1), MMFD (n=2) or MSD (n=9) was carried out. After a median follow-up of 6 months (2-91) the overall survival is 73% (16/22). Three patients had grade III-IV acute GVHD. Causes of death were relapsing malignancy (n=2), GVHD (n=2), graft failure (n=1) and sepsis (n=1). T-cell chimerism at last follow-up was >99% in 12/14 evaluable patients and 50-90% in 2/14. Of the 16 surviving patients all 9 evaluable had complete correction of their immunodefi ciency. Other symptoms such as eczema, allergies, mollusca, bacterial infections, fungal infections and pulmonary function defi cits disappeared or vastly improved in all evaluable patients. Two post transplant malignancies were observed: one thyroid carcinoma 7 years post transplant with a TBI-containing conditioning regimen and one squamous cell carcinoma diagnosed shortly after transplant in previously infl amed skin lesions. Both patients are alive. Longer follow-up will be needed to ascertain that HSCT will correct the malignancy risk, as the DOCK8 molecule has been implicated as a tumor suppressor and is expressed in extrahematopoietic tissues. A survey to defi ne the natural course of disease and to guide treatment recommendations for all patients is currently underway on behalf of ESID and EBMT. In summary, HSCT corrects the immunodefi ciency and other disease manifestations in DOCK8 defi ciency and this treatment should be offered at least to all patients with severe disease manifestations. We have previously showed that a CMV-reactivation after HSCT is associated with a reduced risk for leukemic relapse in pts with AML. Further, Erlach et al. showed in a lymphoma mice transplant model that coinfection with mCMV induced a strong anti-lymphoma effect by induction of apoptosis in lymphoma cells, which consecutively improved OS in mice. This prompted us to investigate the infl uence of replicative CMV infection in 94 (median age 45, 18-70) pts with lymphoma, who received transplants from unrelated (n=67, 71%) or related (n=27, 29%) donors. Pts were transplanted from HLA-ident. (n=74), HLA-MM (n=16) or HLA-haploident. SIB (n=4). 13 pts (14%) were transplanted for indolent lymphoma (FL n=11, CLL n=2), 67 pts (79%) for aggressive lymphoma (B-lineage n=35, T-lineage n=27, transformed n=5), 11 pts (12%) for MCL and 3 pts (3%) for HD. The disease status of pts at HSCT was CR in 20 pts, PR in 40 pts, refractory in 30 pts and untested in 2 pts. 55 pts (59%) received previous autograft and 82 pts (87%) were treated prior to transplant with at least 3 chemotherapy lines. The HCT-CI were 0-2 in 76 pts (81%) and 3+ in 18 pts (19%). Myeloablative conditioning was applied in 60 pts (64%) while 34 pts (36%) received RIC. 68% of pts (n=48) were at risk for CMV reactivation. CMV replication as detected by pp65 antigenemia assay occurred in 34 pts (36%). Taking all competitive risks into account, the cumulative incidence of PFS at 5 yrs after HSCT was 62% (95% CL: 31-45) in pts without as compared to 80% (95% CL: 9-31) in pts with pp65 antigenemia (p<0.018). In multivariate analysis including all effecting factors, CMV replicative status was confi rmed as a strong independent predictor of PFS (HR: 0.29, 95% CL: 0.08-1.00, p<0.049) together with chronic GvHD (HR: 0.32, 95% CL: 0.13-0.80, p<0.016), and chemorefractory (HR: 3.3, 95% CL: 1.28-8.4, p<0.013). The anti-lymphoma effect was detectable across all lymphoma subsets and was most pronounced in pts with chemotherapy refractory lymphoma. How ever, OS rate and aGVHD grade 2-4 did not differ in both groups (52% for pts with CMV-R. vs 51% without, n.s., and 39% vs 35%, n.s, respectively), whereas 5-yr NRM was higher in group with CMV-repl. (42% vs 23%, p=0.05). This is the fi rst report which demonstrates a strong and GVHDindependent effect of CMV replication on the PFS in pts with lymphoma, which deserves further prospective studies. Sterile CBS eye drops were prepared to contain 1.6 ng/ml EGF in a one-day-dose dispensing device, and were administered for one month. Extent of corneal epithelial defect was evaluated in mm2/area, subjective symptoms (OSDI score), Schirmer test I (ST), Break Up Time (BUT), tear osmolarity, corneal esthesiometry (Cochet-Bonnet esthesiometer), conjunctival scraping and imprint cytology with goblet cell (GC) count were performed at baseline (V0), after fi fteen (V1) and thirty (V2, endpoint) days of treatment. Satisfaction and tolerability questionnaires were evaluated at V1 and V2. Results: between January 2010 and April 2011 we treated 17 patients with severe dry eye (all DEWS score 4) after allogeneic HSCT. Seven patients had extensive chronic GVHD. Only 5 (29%) were on systemic immunosuppressive treatment. All patients had been previously treated with autologous serum, whereas half the patients had received topical cyclosporine A. A signifi cant reduction was shown at the endpoint vs. baseline in corneal epithelial damage (media+SD: 13.6±14.2 vs. 36.7±29 mm 2 /area, respectively), discomfort symptoms (OSDI score 24.2±8.8 vs. 42.3±15.4), scraping cytology score (3.9±1.4 vs. 6.6±2.5), tear osmolarity (313.7±6 vs. 324±7.7 mOsm/L) (p always <0.0001) while a signifi cant improvement was shown in corneal esthesiometry (48.2±1.8 vs. 49.7±1.1 nylon/mm/length, p<0.05). The extent of initial corneal damage but not duration of epithelial defects correlated with response (p<0.0001). No patient withdrew or presented adverse event. All patients reported a high satisfaction degree upon eye-drops instillation. Conclusions: Our data demonstrate that CBS eye drops may represent a new therapeutic approach for the healing of severely injured corneal epithelium in patients with chronic GVHD. A longer follow up is needed to confi rm the duration of the effi cacy after CBS treatment. Bronchiolitis obliterans syndrome (BOS) after allogeneic HSCT is a serious manifestation of cGVHD with high mortality. 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 and to elucidate the biology of BOS. 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, presence of another cGVHD manifestation and lack of response to prior therapies. Twenty-one patients have enrolled. One withdrew prior to medication and 17/21 patients have reached the primary endpoint (6 months) on study medication (10 mg qhs). Study participants ranged from 15-64 years, 12/20 female, with baseline FEV1 of 24-73% predicted, and median FEV1/VC 0.5 (0.29-0.78). All patients met criteria for treatment success (<15% decline in 6 months), with FEV1 changes of: increased 5-13% predicted (n=5), stable with change <5% (n=7), and declined 5-13% (n=5). 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: 15/17 improved. Six minute walk test demonstrated 4 signifi cant increases, 11 stable, 2 signifi cant decline. Lung function score was decreased in 1, increased in 4, and stable in remainder. Other manifestations of cGVHD showed: 4/8 GI GVHD improved and 4/8 stable; 3/6 liver cGVHD improved, 3/6 stable, 1 worse, using NIH consensus staging. Of 10 patients enrolled >2 years ago, 70% are alive, with 2 durable improvements (6, 14%) and 2 stable from baseline, compared with historical controls of 44% survival at 2 years. At the primary endpoint, available preliminary data show increases in proportion of CD8 T cells with Cysteinyl receptor expression in patients with improvements in FEV1 suggesting a possible link between receptor expression and disease modifi cation by montelukast. At 1 year, CD4/CD8 ratios improved in 5 and normalized in 3 patients, suggesting improving immune function with prolonged montelukast therapy. Montelukast was well-tolerated with only one grade II probable attributable adverse event during the collection period. These fi ndings suggest that montelukast is a promising therapy for BOS after allogeneic HSCT and that the cysteinyl leukotrienes may be a point of regulation for progressive BOS after HSCT. Results: FEV1 and LFS were weakly to moderately correlated with 8-9 outcome measures (|r| >0.30); DLCO with grip strength, AAS and walk velocity (|r| >0.30), and increasing RS and NIHLs was associated with 11 and 12 outcomes respectively (p<0.05; most p<0.01). Since BOS is a manifestation of lung cGVHD we analyzed components of NIHLs relative to BOS. 91% of pts without BOS can be predicted on the basis of RS alone (symptom score<2) while only 56% with BOS could be correctly identifi ed using only RS. 89% pts without BOS as well as 89% with BOS could be predicted on the basis of FEV1< 59% alone. 63% with and 64% without BOS can be correctly classifi ed by DLCO alone; 80% without BOS could be correctly predicted on the basis of the NIHLs alone as could 93% with BOS (NIHLs≥2). Overall 3yr survival (OS) was 76.2%. In univariate analysis worse RS (p<0.0001), lower FEV1 (p=0.007), DLCO (p=0.002), higher LFS (p=0.002) and NIHLs (p<0.0001) were associated with worse OS. In multivariable analyses only NIHLs retained an association with OS. OS at 3y: 81% (95%CI 73-86%) vs. 31% (95%CI 14-55%) for NIHLs 0-2 vs. 3 respectively. Conclusion: Most components of the NIHLs are associated with important clinical outcomes. Severe NIHLs is predictive of clinical BOS and poor survival. These data support the validity of NIHLs in pts with cGVHD. Background: Bronchiolitis obliterans syndrome (BOS) following allogeneic HCT is a manifestation of chronic graft-versushost disease with poor prognosis and 5-year survival around 20%. Treatment and survival of patients with BOS has not improved over the last 20 years. Evidence regarding the effi cacy of various immunosuppressive therapies in BOS is still sparse. Aim: The purpose of the study was to prospectively diagnose BOS early in the course of the disease by serial assessments of consecutive patients from day 100 after HCT and to compare the effi cacy of adjunct ECP with other immunosuppressive treatments in BOS patients newly diagnosed and evaluated uniformly according to the National Institutes of Health consensus criteria. Patients and Methods: Forty-six patients (24 male, 22 female) with a median age of 41 (range, 19-65) years developed BOS a median of 336 (range, 97-1254) days after HCT. At onset of therapy BOS was mild in 37 (80%) and moderate in 9 (20%) patients. The median FEV1 and the median lung function score (LFS) at onset were 65% (range, 45-77%) and 6 (range, 2-9). Signs of air trapping in HR-CT scans were present in 70% of patients and BO was proven histologically in 21 (46%). Therapy of BOS consisted of steroids with or without calcineurin inhibitors (CNI) or sirolimus in all patients whereas 27 (59%) also received adjunct ECP on 2 days every other week. Overall survival (OS) was analysed after a median follow-up of 35 (range, 12-66) months after HCT and 22 (range, 3-54) after diagnosis of BOS. Results: When assessed after 1 year, 17 of 27 patients (63%) responded to adjunct ECP and 12 of 19 (63%) to other immunosuppressive therapies. Complete (CR) and partial resolutions (PR) were achieved in 11% and 52% of the ECP cohort compared to 16% and 47% of the non-ECP group and not signifi cantly different. Eight patients (80%) given adjunct ECP as fi rst-line therapy achieved a response (CR n=3, PR n=5) compared to 12 (63%) given fi rst-line therapy with CNI and steroids alone (CR n=3, PR n=9). OS at 3 years was 86% for responders and 71% for progressive and stable disease. Patients with progressive BOS had signifi cantly worse OS at 3 years compared to CR (44% vs. 100%, p=0.02) and PR (44% vs. 83%, p=0.006). Conclusion: Early diagnosis and immunosuppressive/ immunomodulatory therapy are able to improve outcome of patients with BOS. Longer follow-up and larger patient numbers are warranted for confi rmation of these promising fi ndings. After our fi rst study in 19 patients, receiving Imatinib as compassionate use for steroid refractory Chronic Graft Versus Host Disease (cGVHD) we conducted a second study in a larger series of patients with steroid-refractory cGVHD, by using the more stringent response criteria suggested by NIH Consensus Conference for cGVHD. Overall 34 patients with cGVHD, refractory to 2 lines of therapy have been enrolled: 23 male,11 female, median age 49 years (28-73). Median duration of cGVHD was 29 (2-148) months; all patients were heavily pretreated and 19 have been previously treated with ECP and/or Rituximab. Main cGVH targets were lung (27) and skin (generalized scleroderma in 15 and localized in 8); 6 patients had Sicca Syndrome and 12 other visceral involvement. Imatinib was administered at low dose (100-200 mg/day) for a median duration of 16 months (3-45); 15 patients are still on Imatinib.The drug was well tolerated and we did not observe toxic deaths; haematological toxicity was mild, (1 patient had grade 3 anemia), while the main grade 3-4 extra-hematological toxicities were: muscle aches (7), dyspnea (4) and skin rash (3). Response rate (RR), evaluated according to NIH criteria, was 45% at 6 months and 48% at 12 months; skin Body Surface Area (BSA) involvement was signifi cantly reduced in 43% of cases, with a reduction of the Rodnan score in 67% of cases; lung functional score (LFS) improved in 47% of patients and Gastrointestinal score in 63%; among 15 patients receiving steroids,11 were able to

S63 stop or signifi cantly reduce steroids. With a median follow-up of 31 months (3-45) 27 patients are alive and 7 patients died: 4 due to GVHD progression and 2 for infections; 17 patients are in continue response at last follow-up, without additional treatments. The RR observed in this second trial, in a larger series of patients with steroid-refractory cGVHD seems slightly inferior, compared to the RR observed in the fi rst trial, probably due to the more stringent response criteria used. Moreover the higher median age and the different kind of patients could have infl uenced these results: the previous study included only patients with skin fi brotic cGVHD and pediatric patients. However the stable response observed after 12 months and the promising outcome (Figure 1 : OS; Figure 2 : EFS), in this very hard to treat set of patients, suggest that Imatinib is a valuable option in patients with steroid-refractory or steroid-dependent cGVHD. This work has been supported by AIFA (Agenzia Italiana del FArmaco). Introduction: Extracorporeal photopheresis (ECP) is an important second line therapeutic intervention in steroid refractory chronic GVHD (cGVHD) with recognised effi cacy as a steroid sparing agent. Few reliable biomarkers predicting ECP response exist. B-cell activating factor (BAFF) has described roles in immature B-cell survival. Elevated BAFF levels reportedly correlate with cGVHD activity and excess BAFF may contribute to cGVHD maintenance. We report that BAFF level following 6 months of ECP therapy predicts the likelihood of disease fl are and capacity for successful steroid taper. Methods: We retrospectively evaluated 28 adult patients undergoing ECP for steroid-refractory, resistant or intolerant cGVHD. ECP was performed using the Therakos XTS TM or Cellex TM devices. ECP treatment schedule was 2-weekly dual treatments for an initial 3 months, then monthly paired treatments until at least 12 months. 24/28 patients were receiving steroids at start of ECP. Skin disease response was assessed using the Modifi ed Rodnan skin scoring system. Extracutaneous organ cGVHD response was assessed by reduction in symptoms as defi ned by NIH criteria. Disease fl are was defi ned as signifi cant symptomatic increase of, or (re)appearance of, GVHD in affected organ(s). Initiation of steroid or other immunosuppressives during treatment was also regarded as loss of disease control. Successful uninterrupted steroid taper was defi ned as the capacity for dose reduction without steroid re-escalation between 3 and 18 months of ECP. Soluble BAFF in patient sera was measured prior to ECP, and at 3, 6, 9 and 12 months of ECP using commercially available enzyme-linked immunosorbent assay. Results: All patients with serum BAFF levels above 4 ng/ml following 6 months of ECP therapy (n=15) experienced a loss of disease control as evidenced by GVHD fl are between 3 and 18 months of ECP. This resulted in re-escalation of steroid dose in 13/15 patients (87%) and steroid introduction in 1 patient. Loss of GVHD control was signifi cantly less common amongst patients with BAFF levels below 4 ng/ml at 6 months of ECP; 6/13 patients experienced disease fl are between 6 & 18 months (P=0.001, Fishers exact) resulting in steroid or cyclosporine re-escalation in 5/13 patients (P=0.002) whilst steroid taper without escalation was possible in 7/11 patients (P=0.001). Conclusions: Our data supports further prospective studies to assess the potential prognostic value of early BAFF measurement in ECP therapy for cGVHD. Background: The most widely used mice model of chronic graftversus-host disease (cGvHD) is an MHC-matched bone marrow transplantation model of sclerodermatous cGvHD. A limitation of that model is that mortality is relatively low, making diffi cult to study the impact of potentially therapeutic compounds. Aims: To develop a more severe model of cGVHD and to assess the impact of Rapamycin administration in that model. Results: Lethally irradiated Balb/C mice were injected with 10x10 6 bone marrow cells and 70x10 6 splenocytes from B10.D2 donor mice. Twenty-one days later, all mice developed cGvHD. For the severe model, donor B10.D2 mice were injected with 0.5x10 6 splenocytes from Balb/C twenty-one days before

S64 transplantation. All mice from the severe model (n=8) died a median of 32 days while 3 of 7 mice in the classical model survived beyond day 52. Mean survival was decreased in the severe model compared to the classical model (32 days versus 37 days; p=0.0185). Recipient mice in the severe group experienced higher weight loss, hair loss and skin fi brosis. Numbers of T lymphocytes (231.9 ± 151.4 versus 951 ± 532.8; p=0.0032) and CD4+ T cells (63.25 ± 41.93 versus 135.0 ± 14.39; p=0.0018) per microliter of blood at day 21 were lower in the severe group than in the classical model. Moreover, number of regulatory T cells (Tregs) was decreased in the severe model (1.250 ± 0.8864 versus 8.000 ± 6.753; p=0.0151). We then investigated whether rapamycin administration could prevent GVHD in the severe model. All (n=8) mice treated with PBS (placebo) died a median of 32 days after transplantation, while 6 of 8 mice given 1 mg/kg/day i.p. rapamycin survived beyond day 52 (p=0.0012). Number of Tregs/μl was higher at day 21 in rapamycin-treated mice than in mice given PBS Recent advances in unrelated cord blood transplantation (UCBT) has provided increased chances for patients with hematological malignancies to receive hematopoietic stem cell transplantation (HSCT). We have investigated the effect of HLA disparity of unrelated cord blood on HSCT outcome in children and in adults separately. 498 children aged 15 years or younger (HLA -A, -B low resolution and -DRB1 high resolution matched, n=82, one locus mismatched, n=222, two loci mismatched, n=158, three loci mismatched, n=36) (median age, 5 years) and 1,880 adult patients (HLA matched, n=71, one locus mismatched, n=309, two loci mismatched, n=1,025, three loci mismatched, n=475) whose age was 16 years or older (median age, 49 years) at the time of transplant were analyzed. Subjects were recipients of single unit UCB as fi rst HSCT with leukemia. Median infused total nucleated cell number was 5.30 x 10 7 /kg in children and 2.52 x 10 7 /kg in adults (p<0.001). With adjusted analyses, in children, HLA two-antigen mismatched UCBT showed signifi cant increased risk of overall mortality (relative risk [RR]=1.61, P=0.042) and transplant-related mortality (RR=3.55, P=0.005) compared to HLA matched. Risk of relapse did not differ signifi cantly. Risk of mortality increased according to the number of mismatched loci (p for trend, 0.043 and 0.002 for overall mortality and TRM) The risk of relapse was not different among HLA disparity groups in children. Risk of grade 2 to 4 acute GVHD was increased in one-(RR=2.18, P=0.003) and two-(RR=2.51, P=0.001) loci mismatched in children. Two-loci mismatched was associated with higher risk of grade 3 to 4 acute GVHD in children (RR=2.45, P=0.041).

In Adults, the risk of mortality did not increase with the number of mismatched loci (RR=0.98, P=0.924 for one-locus mismatched, RR=0.88, P=0.423 for two-loci mismatched, and RR=0.95, P=0.746 for three-loci mismatched for overall mortality). In adults, risk of relapse was signifi cantly decreased in two-loci mismatched (RR=0.67 P=0.029). Risk of TRM, grade 2 to 4 or grade 3 to 4 acute GVHD did not differ among HLA disparity groups in adults. Effect of HLA disparity on transplant outcomes were different between children and adults. In children, increased number of mismatched HLA loci correlated with increased risk of mortality. In adults, there was no increase in mortality with increase in the number of mismatched HLA loci. Late engraftment is a risk factor for CBT in adults. To overcome this limitation in 1999 we developed the "dual transplant" method (1): co-infusion of only one CB unit and highly T-depleted mobilized CD34+ (HP) cells from an adult third party donor (TPD), haploidentical or not. Here we report data from 98 adults (61/37 M/F) median weight 70 Kg (42-111) transplanted in 3 Spanish centers to treat high risk hematological malignancies (HM). Indications: Acute Leukemia (AL) 85 (43 AML, 32 ALL), other 13. Transplant products cellularity and HLA compatibility data shown in Table 1 ; 105 CB units were used as a 2nd unit was required for 7 patients: 2 rejections; 4 graft failures due to lack of viable HP; 1 relapse. Conditioning: For 49: 10 Gy fractionated TBI, Fludarabine 60 mg/m, CTX 120 mg/kg and ATG; other 49 received Busulfan 3,2 x2-3 mg/kg instead of TBI. Post-Tx treatment: G-CSF, Prednisone 1 mg/Kg 8-14 days and CsA till full CB chimerism. Engraftment data shown in Table 1 . The TPD did not take in 8 cases: 1 due to very early CB engraftment; 6 seemingly due to recipient allosensitation against TPD (mother or husband). Other 6 who had engraftment failure/rejection of the initial CBT had sustained TPD graft until the take of a second CBT, given after 33-94 days preceded by a 2nd condi tioning (Fludarabina+ATG± 2 Gy TBI or Thiotepa+Fludarbine+ATG). Morbi-mortality and survival data on incidence of TRM, relapses and GVHD are shown in Table 1 . In no case were TPD cells involved in aGVHD lesions. Most common infections were CMV (72 episodes), declining after 3-4 months with 6 CMV-diseases causing 5 deaths. Other 2 deaths were due to toxoplasmosis and 1 each to EBV-PTLS, leishmaniasis and scedosporium prolifi cans infection (related to long pre-tx neutropenia). Conclusions: Time to neutrophil recovery after dual transplants is consistently shorter than reported for other approaches (ex-vivo expansion, double CBT, intrabone infusion), resulting in low risk of early infections. This and the favorable data on aGVHD and relapses (i.e. GVT) contribute to less hospital days, what together with the procedure relative low cost translates into favorable costs/results ratio. The procedure may allow selection of CBT units prioritizing HLA compatibility to cell content and use of a fraction of CBT Cord blood as a source of stem cells has been a successful addition to the fi eld of allogeneic stem cell transplantation (SCT). The drawback is that patients suffer from a longer period of compromised immunity. We analyzed T cell Receptor Excision Circles (TRECs), Immunoglobulin G (IgG), immunoglobulin M (IgM) levels, lineage-specifi c chimerism analysis and patient outcome after cord blood transplantation (CBT) in 50 patients transplanted at our center. Engraftment of neutrophils was seen in 88% of the patients at a median time of 29 days (range 3-79). Complete donor chimerism (DC) within the CD19+, CD3+, and CD33+ cell lineages was seen in 74%, 72%, and 76% of the patients, respectively. DC was associated with acute graft-versus-host disease (GVHD) grades II-IV for the CD3+ cell lineage (p=0.01) and, in multivariate analysis, with total body irradiation (TBI) for all cell lineages (p<0.01). Overall survival (OS) at one and fi ve years was 55% and 43%. Non-malignant diseases was associated with better 5-year OS (72%) than malignancies (28%, p=0.026). In multivariate analysis, a negative correlation was seen between OS and age, hazard ratio (HR) 1.04 (95% CI: 1.02-1.06, p<0.001); acute GVHD grades III-IV, HR 3.43 (95% CI: 1.95-6.02, p<0.001); and mesenchymal stem cell treatment 2.66 (95% CI: 1.11-6.35, p=0.027). Transplant-related mortality (TRM) at 100 days and one year was 16% and 30%. The incidence of acute GVHD grades II-IV was 34%. Acute GVHD grades III-IV was associated with AB0 incompatibility (HR 2.61, p=0.05) and myeloablative conditioning (HR 4.17, p=0.047). We found that TREC levels after CBT were lower in adults, patients with malignant disease, patients with myeloablative conditioning, and in patients with a lower nucleated cell dose in the graft. In addition mesenchymal stem cells (MSC) as co-infusion at the time of CBT had a negative effect on TREC reconstitution which may be one explanation for the decreased survival in MSC treated patients. Reduced IgM and IgG levels post CBT were associated with older patient age, a major AB0 mismatch, and infusion of mesenchymal stem cells. Our results highlight the importance of close monitoring of the reconstitution of the immune system after cord blood

transplantation. In addition it shows a potentially new suppressive effect of MSCs on the immune system. The impacts of allelic HLA matching in patients (pts) with AML and MDS who receive PBSC after a RIC remain to be addressed. In this study, we aim to compare the impact of the donor type in this setting: HLA identical sibling (S) versus high resolution HLA matched 10/10 unrelated donor (MUD). From 01/01 to 12/10, 108 pts with AML (n=63) and MDS (n=45) received PBSC after RIC in our center, either from S (n=69) or MUD (n=39). RIC was fl udarabine based in 95% of pts. Engraftment, acute and chronic GvHD, transplantation-related mortality (TRM), relapse rate (RR) and overall survival (OS) at 3 years were compared according to type of donor: S or MUD. WHO classifi cation for MDS was RAEB1 (24%), RAEB2 (36%), MDS transformed into secondary AML (20%), CMML2 (9%), RA (4%), or other (7%). Disease risk was assumed by cytogenetic (MRC for AML, IPSS for MDS) and EBMT score (good risk: CR1 for AML or MDS or untreated MDS, intermediate risk: CR2 for AML, CR2 or partial remission for MDS, poor risk: all other status). Cytogenetic was poor, intermediate or good for 21, 74 and 5% of AML and 24, 36 and 40% of MDS, respectively. EBMT score at time of HSCT was poor, intermediate or good for 29, 7, 64% of MDS and 11, 21, 68% of AML, respectively. Pts characteristics according to type of donor were similar for age (median 57 years), gender and disease distribution. Particularly, disease risks were comparable in 2 groups. Conversely, conditioning regimen (more ATG in MUD), donor age (younger for MUD) and number of CD34+ cells infused (higher in MUD) were different. All pts engrafted. The cumulative incidence of acute GvHD was 40% with S and 44% for MUD (p=0.58). The cumulative incidence of chronic GvHD was 49% with S and 45% with MUD (p=0.66). No signifi cant risk factor was associated with acute or chronic GvHD. TRM was 17% and 22% with S and MUD, respectively (p=0.55). Adjusting for age, MDS was the only factor increasing TRM (HR 3.4; p=0.02). RR was 46% with S and 30% with MUD (p=0.28). OS was 44% with S (95%CI: 33-61) and 50% with MUD (95%CI: 35-71) (Figure 1 ). Cytogenetic and EBMT score did not signifi cantly infl uence the OS. After adjustment for age, cytogenetic risk, ATG and number of CD34+ cells infused, donor type still did not infl uence OS. In an homogeneous group of pts with AML or MDS receiving PBSC after RIC, survival was strictly identical if the donor was an HLA matched sibling donor or MUD including after adjustment for variables potentially impacting survival. Delayed hematopoietic recovery and graft failure (GF) are critical complications of cord blood transplantation (CBT) and are associated with TNC or CD34 cell dose and HLA disparities. Other factors such as patients's HLA-antibodies (Ab) may impact neutrophil recovery. To analyze the effect of anti-HLA-Ab on CBT outcomes we analyzed 206 pts who underwent CBT after RIC regimen from 2000-2010. Median follow-up was 36 months. Eighty-two percent of pts were transplanted for malignancies, 40% had single and 60% double-CBT. Thirty percent of CBU had 0-1 HLA mismatch (A, B, DRB1), 67% received CyFluTBI2Gy and median infused TNC was 3.7x10 7 /Kg. Pre-transplant serum was tested for HLA-Ab with a panel of fl uorescent beads coated with single HLA-antigen using Lumi-nexTM platform. Results were interpreted as fl uorescence intensity (MFI) against donor-specifi c mismatch. Overall 48 pts (23%) had anti-HLA-Ab before CBT and those were donor specifi c anti-HLA-Ab (DSA) in 16 pts. Among the 16 pts with DSA (11 females, 5 males), 9 had single and 7 double-CBT (none had DSA directed to both CB units). Seven pts had DSA vs to HLA-Class-I, 5 vs to HLA-Class-II and 4 to both HLA-Class-I and-II. DSA threshold ranged from 1620-17629 MFI. Cumulative incidence (CI) of day-60 neutrophil engraftment was 76%. It was 44% for recipients with DSA and 81% in pts without DSA (p=0.006). There was no difference for pts with anti-HLA-Ab non donor-specifi c (77% vs 69%). Multivariate model showed DSA (RR 2.7, p=0.01) and CBT before 2008 (RR 1.49, p=0.03) independently associated with GF. Seven pts with DSA engrafted, 4 after double-CBT and chimerism analysis showed the engraftment of the CBU with DSA in 1 case. Among 50 pts who failed engraftment,, 9 (20%) pts had DSA specifi c for donor HLA-Class-I (n=4) or Class-II (n=2) or both Class-I and Class-II (n=3). CI of platelet recovery at day-180 was 62%, 12 of 16 patients with DSA did not achieve platelet recovery. CI of 1-year TRM was 35%. DSA was associated with higher TRM (p=0.002). Overall survival at 3-years was 44%, it was 41% and 45% for pts with non-malignant and malignant disease respectively. OS was 47% for recipients without DSA and 25% for those with DSA, p=0.006. In multivariate analysis, the absence of DSA was the only factor associated with better survival (RR 2.41, p=0.005) Donorspecifi c anti-HLA-Ab in recipients of CBT is associated with failed engraftment and lower survival. Screening for DSA may be included in the algorithm of donor choice. Rationale: The NMDP manages the largest international registry of volunteers registered to donate stem cells to unrelated patients with blood-related diseases. Key stages leading to donation include joining the registry, confi rmatory typing (CT), and fi nal work up. CT-stage has emerged as a critical decisionpoint in the process because most donor attrition occurs at this point. In addition, although more than 65% of whites continue toward donation at CT-stage, only 40% of members of all other ethnic groups continue. Higher attrition among ethnic minority donors disproportionately disadvantages ethnic minority patients searching for donors. Our goal was to examine characteristics of potential donors that might be associated with CT-stage decision and/or race/ethnicity. Methods: We conducted telephone interviews with a stratifi ed random sample of registry members who had initially matched a patient, were contacted at CT-stage, and who decided either to continue toward donation (N=843) or to opt-out of the registry (N=224). Measures included demographics, culturallyrelated variables (e.g., medical mistrust, religious objections to donation), psychosocial variables (e.g., emotional distress, self-esteem) and donation-related variables (e.g., ambivalence about donation, donation-related concerns). Analyses examined differences in these variables by CT-stage decision and racial/ethnic group. Results: Those who opted-out of the registry at CT-stage had more religious objections to donation (F=51.9, p<.001), more mistrust of the medical system (F=5.0, p<.05), less trust that the stem cells would be used appropriately (F=32.1, p<.001), more emotional distress (F=29.6, p<.001), more ambivalence (F=430.3, p<.001), and more medical concerns (F=56.4, p<.001). Ethnic group analyses indicated that whites, compared to minority groups, had fewer religious objections to donation (t=2.4, p<.05), less mistrust of the medical system (t=2.5, p<.05), and more trust that the stem cells would be used appropriately (t=4.2, p<.001). Asian Pacifi c Islanders had more emotional distress (t=2.7, p<.01) and were more ambivalent about donation (t=2.1, p<.05) than other groups. Conclusions: These fi ndings have important implications for the management of donors, particularly donors belonging to ethnic minority groups. They suggest that communications with donors at key contact points including recruitment and CT-stage could be tailored to address concerns specifi c to each group. Approximately 1 in 15 patients will undergo a second allogeneic haematopoietic cell (HPC) transplant, the majority using the same donor. Anthony Nolan carried out a retrospective study of subsequent HPC donations made by its donors over a sixyear period, from 2005 to 2011, with the aim of predicting which donors were more likely to be called for a second donation. Methods: The initial transplants of those patients requiring a subsequent donation were examined: fi ve key donor characteristics (route of harvest, HLA match, age, gender and CMV status) were analysed, and their incidence compared to that found in all HPC donations provided in the same period. Results: 2105 HPC donations were made during this time, of which 117 (5.6%) were subsequent donations. 93 (79.5%) of these were to the same patient and the following results are for this group. The median time between donations was 198 days (range 39-4016 days). Main disease categories included acute myeloid leukemia (28%, n=26), acute lymphoblastic leukemia (14%, n=13), myelodysplasia (12.9%, n=12), non-Hodgkin lymphoma (10.8%, n=10), aplastic anaemia (8.5%, n=8) and chronic myeloid leukaemia (7%, n=7) -this composition differed signifi cantly from the overall sample (Chi-square goodness of fi t test, p<0.001). Indications for second allogeneic transplant included primary graft failure (PGF, 7.4%, n=7), secondary graft failure (SGF, 58.1%, n=54), disease relapse (REL, 31.2%, n=29) and others (3.2%, n=3 ). Mean cell doses at initial transplant were 2.88x10 8 /kg total nucleated cells for bone marrow and 7.7x10 6 /kg CD34+ cells for peripheral blood stem cells (PBSC). Bone marrow was more common as the source of stem cells at initial transplant in PGF (71.4%, p=.012), SGF (50%, p<0.001) and REL (48.2%, p=0.005) when compared to the background frequency (24.6%, weighted mean). There was a trend towards a greater frequency of HLA mismatched donors in the PGF group only (p=0.093). Donor age >30, CMV positivity and female gender were not found to be more frequent. Conclusion: Although small, this study suggests that the donation of bone marrow confers a greater likelihood of a donor being called for a second HPC donation to the same patient when compared to PBSC collection (8.01% vs 2.3%, p<0.001). These fi ndings have implications not only for harvest physicians when consenting donors for bone marrow harvest, but also for transplant centres when considering the most appropriate HPC source for their patients. Bu/CY-based conditioning for HSCT may induce short term toxicity, e.g. VOD and long-term sequelae, e.g. infertility. Reduced toxicity conditioning regimens providing both suffi cient myeloid donor engraftment and reduced toxicity are currently investigated. 58 pediatric (0.3-17 ys) and 16 adult (18-39 ys) patients (n=55 male; n=19 female) with non-malignant diseases (CGD n=45, SCID n=8, HLH n=3; WAS n=3; XLP/CID n=4, NEMO/OP n=2, Fucosidosis/ALD n=4, Thalassemia n=3, refractory anemia/ PNH n=2) from 15 centers underwent conditioning comprising 180 mg/sqm fl udarabine (d-8 to -3), low dose or targeted busulfan (d-5 to -2) and serotherapy with rabbit ATG (Fresenius: 40 mg/kg or Genzyme: 7.5 mg/kg) in MSD or unrelated CBT or alemtuzumab (0.5-0.6 mg/kg; d -8 to -6) in MUD/MUD HSCT. HSCT was performed at a median age 8.75 years (0.3-39 ys) using MSD (n=29), MRD (n=4), MUD/MMUD (n=40) and haploidentical (n=1) donors. BM (n=59), PBSC (n=11) and CB (n=4) served as stem cell sources. GVHD-prophylaxis comprised CSA (until d+160) and mainly MMF (until d +100). Pharmakokinetic monitoring for busulfan was performed in 52 patients targeting at a submyeloablative AUC of 45-65 mg/Lxh for patients with PID and refractory cytopenias and a myeloblative AUC of 75-95 mg/Lxh for hemoglobinopathy, metabolic disease, WAS and CGD/MDS after gene therapy. In 22 patients where PK monitoring was not feasible reduced dose (8 mg/kg orally or 6.4 mg/kg iv.; n=16) or full dose busulfan (16 mg/kg; n=6) was administered. N=5 cases of mild VOD responding well to therapy were observed. The rate of aGVHD grade III-IV and limited cGvHD was 4% (4 of 74) and 13% (10 of 74), respectively. Neutrophil/platelet engraftment occurred in 71/74 (96%) patients after a median of +19 and +22 days, respectively. 4 patients had autologous reconstitution/non-engraftment and were successfully retransplanted. One patient received a successful stem cell top-up due to prolonged pancytopenia. After a median follow-up of 15 months (range 4-102 mo), the overall and event free survival rates are 95% (70/74) and 89% (66/74), respectively. All surviving patients exhibit a >90% myeloid donor chimerism except one OP patient with stable 70% myeloid chimerism. This protocol provides excellent short term toxicity and myloid engraftment features with high cure rates. Targeting busulfan administration is of major importance to steer therapy and to ensure myloid engraftment while reducing short-and long-term toxicity. 

Infantile osteopetrosis is a rare heterogenous genetic disorder caused by mutations in one of several genes, which impair osteoclast function. Haematopoietic stem cell transplantation (HSCT) is the only treatment that can improve the phenotype. In patients without an HLA-matched family or unrelated donor, HLA-haploidentical HSCT has been performed with success rates between 24% and about 60%. Here we report on a recent survey of HLA-haploidentical transplantation in patients with osteopetrosis. Twenty-one patients with infantile osteopetrosis were transplanted at the University Hospitals of Ulm, Paris, Goteborg and Jerusalem between 2001 and 2009. DNA analysis revealed biallelic mutations in the TCIRG1 gene (n=16), and the CLCN7 gene (n=4), respectively. In one patient no causative mutation could be detected (genes sequenced: TCIRG1, CLCN7 RANK, RANKL, OSTM1). The age at transplant was 2 to 72 (median 15) months. The conditioning regimen was based on busulfan and fl udarabin in all patients combined with cyclophosphamide in 7 and thiotepa in 14 patients, respectively. ATG or Campath-1H was used as prophylaxis for rejection and graft-versus-host disease. G-CSF mobilized cells from a parent were used as stem cell source after CD34 positive selection. Toxicity of the conditioning regimen was high with severe VOD or complications making ICU treatment necessary in 13/21 patients (62). Eleven patients (52%) rejected their fi rst graft and/or showed an autologous reconstitution after prolonged cytopenia. The risk for rejection or autologous reconstitution was associated with age older than 10 months (7/8 patients rejected) and less than 10 x 10 6 CD34+ cells/kg in the graft (5/5 patients did not engraft). However, 8 of 11 patients could be rescued by a second HSCT after rejection and survived despite severe complications and prolonged aplasia. Notably, three patients aged 72, 42 and 16 months at their fi rst HSCT died because of graft failure despite repeated transplantations. Eighteen of 21 patients (86%) survived with a median follow up of 72 months (3-161 months). These data indicate that HLA-haploidentical HSCT in osteopetrosis is an approach associated with severe complications, in particular a high risk for rejection in patients beyond the age of 10 months, but can be performed with an improved and quite high success rate compared to previous studies. Allogeneic BMT still remains the only treatment potentially able to defi nitively cure thalassemia major. We evaluated the outcome of patients transplanted from an unrelated donor (UD), selected using stringent criteria of compatibility after high-resolution molecular typing of HLA class I/II loci, analyzing factors with a prognostic relevance. We studied 122 patients (96 children and 26 adults, 71M/51F, age range at time of BMT 1-35 years, median 10.5). 15 patients had positive HCV serology. 39 pediatric patients belonged to the class I of risk of the Pesaro classifi cation, 41 to the class II and 16 to the class III. 106 donor/recipient pairs were matched for the HLA-A,-B,-C,-DRB1 loci; 16 donor/recipient pairs had a single HLA-class I allelic disparity. Patients received a myeloablative regimen based on busulfan/treosulfan, thiotepa and either cyclophosphamide (54 patients), or fl udarabine (68 patients). All patients received fresh, unmanipulated BM cells (median dose of nucleated cells infused: 5x10 8 /kg; range 1.4-15). GvHD prophylaxis was homogeneous in all patients and consisted of CsA, short-term MTX and ATG (2.5 mg/kg from day -4 to -2). 50 patients at risk (28%) developed grade II-IV acute GvHD with 16 of them (13%) experiencing grade III-IV acute GvHD. 14 patients at risk (13%) had chronic GvHD, which was limited in 8 cases and extensive in 6. 20 patients died and 16 patients had either primary (11 patients) or secondary (5 patients) graft failure. Acute and chronic GvHD were the main causes of death accounting for 6 and 4 fatal events respectively; graft failure was responsible of 2 fatal events. The 5-year OS and thalassemia-free survival (TFS) were 84% (95%CI, 76-90) and 75% (95%CI, 66-81) respectively. The OS of children belonging to class I, II and III of the Pesaro classifi cation was 97% (95%CI, 83-99), 85% (95%CI, 70-93) and 81% (95%CI, 51-93), respectively, while that of adults was 65% (95%CI, 44-80) (p<0.01). In multivariate analysis, only occurrence of grade III-IV acute GvHD and older patient's age predicted a poor outcome (HR 7.02, 95%CI 2.76-17.83, p=0.001 and 1.10, 95%CI 1.03-1.18, respectively). These data indicate that, provided that selection of the donor is based on stringent criteria of HLA-compatibility, transplantation from UD is able to cure a large proportion of thalassemia major patients. Refi nements of strategies aimed at preventing both GvHD and graft failure could further improve the chance of cure of these patients. Introduction: Chronic Granulomatous Disease (CGD) is a rare X-linked or autosomal recessive inherited primary immuno-defi ciency leading to multiple opportunistic infections, abscesses and infl ammation. 55% survive to 30 years old. HSCT offers curative treatment in patients with severe and life threatening complications from the disease. Methods: Retrospective analysis of case notes of all CGD patients who underwent HSCT in Newcastle Upon Tyne Hospitals NHS Trust. Results: 41 patients have undergone HSCT for CGD since 1997; 35 survived and 6 died (1 during conditioning). Median age at diagnosis and HSCT were 3.21 (range 0-16) and 8.38 (range 0.75-27.2) years respectively. 3 patients were adults (>18 years) at time of transplant. 2 of 6 patients who died had active fungal infection at time of transplant and 1 other died from pulmonary haemorrhage following aspergilloma. Median age at HSCT of those who survived was 7.17 years (range 0.75-16.9) and those who died was 18 years (range 8.33-27.2). 39 patients had had an infection prior to transplant, the commonest being pneumonia or lung abscesses, lymphadenitis and perianal or gluteal abscesses. Aspergillus spp. were the most common organisms isolated, followed by Staphylococcus aureus. 26 patients had CDG-associated colitis prior to transplantation. 40 patients received HSCT, 2 underwent a further transplant due to graft failure and 3 required top-up stem cell transfusions. 15 patients had GvHD, 15 skin (I-II), 4 liver and 2 gut. 6 patients developed chronic GvHD, 1 of whom died. 2 remain on immunosuppression. 1 patient had recurrence of colitis. 28 patients have normal neutrophil oxidative burst (NOB; >95%) after transplant, 6 have a dual population of neutrophils but are clinically normal and 1 patient has <20% NOB and is awaiting further HSCT 6 years after umbilical cord stem cell transplant. All patients with greater than 2 years follow up have shown a response to primary childhood vaccinations. None require antifungal prophylaxis and 2 are on long term antibiotic prophylaxis. Conclusions: HSCT is curative for CGD with 85% survival in this cohort and good long term graft function, restoring neutrophil function and responding to vaccination. Infl ammatory colitis also resolves post transplant. Survival is related to severity of complications pre-HSCT and hence age at HSCT. HSCT should be considered in the early management of CGD, particularly when there is a well matched donor. Hyper IgE Syndrome (HIES) constitute a rare subset of immunodefi ciencies collectively characterized by a trias of increased serum IgE levels, eczematous rash and recurrent skin and pulmonary infections. Autosomal recessive HIES has been found to be associated with mutations of DOCK8. Hitherto, few reports have been published on patients with DOCK8 HIES treated with hematopoetic stem cell transplantation (HSCT). We report about a girl born to non-consanguineous parents who presented with a severe generalized refractory eczema beginning in the fi rst month of life, multiple food allergies, recurrent sinopulmonary infections, chronic mollusca contagiosa and a chronic HSV1 infection of the eye. Extremely high IgE levels of > 30.000 IU/ml and high values for absolute eosinophils were found. A homozygous mutation in DOCK8 (c.850_851delCT; p.L284fsX292) was identifi ed. Examinations before HSCT revealed a solitary lung nodule. Histology and culture after surgical resection showed aspergillus fumigatus. Following a reduced intensity conditioning regimen consisting of fl udarabine, thiotepa, melphalan and ATG the child received bone marrow stem cells (4,05x10 6 /kg CD34+) from a 10/10 matched HLA identical unrelated donor at the age of 3 years. Immunosupression included cyclosporin and mycofenolatmofetil (until day +30). Engraftment with absolute neutrophil counts > 500/μl occurred on day +19. Transplant complications included acute skin GvHD grade II at day + 26 which responded well to systemic steroids as well as a CMV pneumonitis 4 weeks after SCT which was treated with ganciclovir. Soon after transplantation, the eczematous rash and the HSV 1 infection of the eye disappeared, while the chronic mollusca contagiosa infection persisted for several months. Peripheral blood consistently showed >99% donor chimerism in all cell subsets. Cyclosporine was discontinued at day +180. Six months after transplantation the child is in very good health without signs of infection. The skin appears almost unremarkable with only small remnants of molluscum lesions. Food allergies nearly completely disappeared. IgE levels and absolute numbers of eosinophils have normalized.

In conclusion, we have shown in a patient that DOCK8 hyper IgE syndrome can be treated successfully with HSCT. To more comprehensively evaluate the risk and benefi t of HSCT in DOCK8 defi ciency a study including HSCT centers worldwide is underway registered at the European Society of Immunodefi ciencies (ESID). Background: Wiskott-Aldrich syndrome (WAS) is a rare X-linked immunodefi ciency caused by genetic defects of the WAS protein (WASP) gene. Patients with WAS typically demonstrate micro-thrombocytopeniain the absence of other hematological symptoms. Here we report on 6 male infants with WAS, presenting with an unusual complication, which was initially undistinguishable from juvenile myelomonocytic leukemia (JMML). Methods: The central morphological and laboratory diagnosis of JMML has been performed by the Japanese Society of Pediatric Hematology (JSPS) and the European Working Group of MDS in Childhood (EWOG-MDS) in Japan and Europe, respectively. The laboratory diagnosis of JMML consisted of a mutational analysis of genes of the RAS-signaling pathway including PTPN11, KRAS, NRAS (n=6) and c-CBL (n=1). The diagnosis of WAS was made by the analysis of the intracellular WASP expression by Western blot analysis or fl ow cytometric analysis and a detection of WASP gene mutation. Results: Five Japanese patients and one German patient with an initial suspected diagnosis of JMML were later diagnosed with WAS. Leukocytosis, monocytosis, presence of myeloblasts and hematopoietic precursor cells in peripheral blood and a hypercellular bone marrow with dysplasia were fi rst observed at the median age of 2 months (1-3 months) and persisted for several months. The morphological and hematological pictures are compatible with the diagnosis of JMML.

All patients demonstrated thrombocytopenia since birth and bloody stool noted fi rst soon after birth; the later is typical for WAS but not for JMML. Splenomegaly was observed in one patient only; absence of splenomegaly at diagnosis is highly unusual for JMML. A mutational analysis of genes of the RASsignaling pathway was negative in all 6 patients. Non-hematological features such as eczema and bloody stool fi nally led to the diagnosis of WAS at the median age of 4 months (3 to 9 months); in all caes the diagnosis was confi rmed by the lack of intracellular WASP expression and a detection of WASP gene mutation. Interestingly, mean platelet volume was normal in these patients and 3 of them demonstrated occasional "giant" platelets, which made the straightforward diagnosis of WAS diffi cult. Conclusions: WAS should be considered in male infants presenting with JMML-like features if no molecular markers of JMML can be demonstrated. Clinical information such as bloody stool are helpful for the early consideration of WAS. 

Recovery of immunity after hematopoietic stem cell transplantation (HSCT) remains poorly defi ned. In clinical practice, physicians are left with many uncertainties about when to stop infection prophylaxis and allow outdoor activities. It is unclear how decision making on these issues is infl uenced by the immune status of the HSCT recipient. A multi-round survey among members of the EBMT Inborn Errors and Pediatric Diseases Working Parties was performed using a web-based Delphi procedure, to reach a consensus on issues of infection prophylaxis strategy after pediatric HSCT that are not covered by existing guidelines.

The fi rst survey described fi ctional cases covering the spectrum of HSCT conditions and complications in order to establish current policies (n=53 physicians). The survey demonstrated relative agreement when initiating infection prophylaxis, while exhibiting a surprising disagreement on discontinuation of prophylaxis and permitting outdoor activities. In the second round we questioned practical rules on the most important issues, i.e. (re)vaccinations after HSCT, stopping prophylactic drugs (such as PCP and antifungal prophylaxis) and initiating activities that carry infection risk (such as outdoor play, public transportation, attending school/daycare and travel abroad), with an emphasis on immunologic recovery in addition to time after transplant as primary criteria for decision making (n=96 physicians). In the third round the resulting statements were proposed, and approved or disapproved by each center (n=20 SCT centers). The consensus statements are presented (see Table) . S71 virus-specifi c T-cell receptors (TCR) may be an effi cient means to transfer CMV specifi c T-cell function into the patient. For this, we used in vitro transcribed RNA encoding CMV-specifi c TCR for electroporation of human T cells that resulted in potent effector functions against CMV-infected targets for up to one week. Due to the transient expression of the introduced RNA, a potential study protocol would have to include the weekly administration of TCR redirected T cells. However, this approach might be hampered by the induction of serious alloreactivity through the repeated transfer of polyclonal donor T cells. To address this concern, we generated TCR transfected naive and memory T-cell subpopulations. The latter have been reported to induce less alloreactivity due to a more restricted endogenous TCR repertoire. Although both naive and memory T-cell subsets showed comparable TCR expression upon RNA transfection, memory T cells mediated superior cytotoxicity against CMV-infected targets. In addition, we analyzed alloreactivity of the naive and memory T-cell subsets against HLA-mismatched EBV-transformed B-cells in IFNg ELIspot and cytotoxicity assays. As expected, alloreactivity was mainly mediated by T cells of naive phenotype. To investigate the relevance of these fi ndings for a clinical application, we used a mouse model that allows the analysis of T-cell alloreactivity directed against human hematopoiesis. For this, we reconstituted NOD/SCID/IL2Rgcnull (NSG) mice with human CD34+ stem cells and adoptively transferred them with naive and memory CD8 T-cell populations previously stimulated against cells of the stem cell donor. In line with the in vitro data, naive T cells mediated a stronger alloreactivity against the donor hematopoiesis than memory T cells did. This was shown by a signifi cant decrease of human CD45+ hematopoietic cells and B-cells in spleen and bone marrow of the mice. Our data show that human memory CD8 T-cell populations mediate decreased alloreactivity in vitro as well as in vivo. This observation along with strong effector function upon TCR transfer makes memory CD8 T cells promising tools for adoptive immunotherapy. Memory stem T cells (TSCM) with the ability to self renew and the plasticity to differentiate into potent effector cells represent ideal weapons to treat cancer. Nonetheless, clinical-grade proce dures to specifi cally target this T-cell population remain elusive. Here we report that it is possible to differentiate in vitro, vigorously expand and genetically engineer TSCM lymphocytes in clinically compliant conditions starting from naïve precursors. Requirements for the generation of this T-cell subset, best described as CD62L+ CCR7+ CD45RA+ CD45R0+ IL-7Ra+ CD95+, are CD3/CD28 engagement and culture with IL-7 and IL-15. The gene expression profi le and functional phenotype validate this cell population as a distinct memory T lymphocyte subset, hierarchically superior to central memory (TCM) and effector memory lymphocytes. To better defi ne naive-derived TSCM function and potential in a clinically relevant context, we evaluated their ability to mediate xenogenic GvHD in vivo. Indeed, naïve-derived TSCM prove superior expansion and persistence than TCM manipulated with the same protocol, and are able to differentiate into effector cells, mediating a lethal xenogeneic GvHD with the same kinetics and intensity of unmanipulated lymphocytes. Furthermore, gene-modifi ed TSCM are the only T-cell subset able to engraft and mediate GvHD upon serial transplantation, thus demonstrating that selfrenewal capacity is preserved even after ex vivo expansion and genetic manipulation. Finally, we identifi ed and functionally validated the natural counterpart of our gene-modifi ed TSCM population in healthy donors as a rare fraction of naive T cells, recapitulating the CD62L+ CCR7+ CD45RA+ CD45R0+ IL-7R alpha + CD95+ phenotype. These fi ndings pave the way for a rapid clinical translation of cancer adoptive immunotherapy approaches based on ex vivo generated TSCM lymphocytes. We have previously showed that a CMV-replication (CMV-R) after T cell repleted HSCT is associated with a reduced risk for leukemic relapse in AML. Here, we fi rst evaluated in a cohort of 64 pts the effect of a CMV-R on the cumulative relapse incidence (CIR) after in-vivo T cell depleted transplant using Cam-path®, and secondly, the infl uence of a CMV-R on CIR in pts with KIR ligand incompatibilities in a 2nd cohort of pts (n=100) transplanted from HLA-MM SIB or URD. In the 1st cohort (n=64) pts after myeloablative (TBI based conditioning n=56, chemotherapy based conditioning n=8), T cell depleted transplant using Campath® did not benefi t from a CMV-R in regard of CIR. The 5-yr CIR for pts with CMV-R (n=23) after transplant was 57% vs 51% (n.s.) in pts (n=41) in whom a CMV-R was not detected. Pts were transplanted in 1.CR (n=18), 2.CR (n=21) or more progressive stages (n=25) from HLA-identical SIB (n=18) or URD (n=32). Two pts received transplants from a SIB with one HLA -MM and 12 from URDs with one HLA-MM. CMV status of recipient (R) or donors (D) were in 25% R-/D-, 8% R-/ D+; 25% D+/R-and 42% R+/D+. The median age of pts was 46 yrs and 38 yrs for donors. 5-yr OS was statistically not different in both groups. In the 2nd cohort of 100 AML-pts transplanted from HLA-MM URD (n=96) or SIB (n=4) donors a documented CMV-R was associated with a reduced 5-yr CIR of 17.2% (95% confi dence limit [95% CL]: 14-20) compared to 44% (95% CL: 34-54) (p=.04). GVHD prophylaxis was performed with MTX, CSA with or without ATG (n=46). The median age of pts was 49 yrs and 38 yrs for donors. Pts were transplanted in 1.CR (n=40), 2.CR (n=32) or more progressive stages (n=28) from a donor with one HLA-MM (n=85) or 2 HLA-MM (n=15). Although CIR ligand expression is reported to be infl uenced by a CMV-R, here the KIR ligand status had no infl uence on the CIR. Pts with (n=23) or without KIR ligand incompatibility (n=19) towards their donor benefi t both from a CMV-R. In multivariate analysis CMV including competitive factors which infl uence CIR, CMV-R was confi rmed as a independent predictor of relapse (hazard ratio [HR]: 0.12, 95% CL: 0.015-0.95, p<0.04) together with chronic GvHD (HR: 0.18, 95% CL: 0.74-0.79, p<0.018). The reduced risk for CIR translated into superior OS estimates at 3 yrs for pts with CMV-R (n.s. OS: 63% vs. 48%). In conclusion, this report confi rms the strong and independent effect of CMV-R on the leukemic relapse risk in AML-pts after transplant. Objective: Unmanipulated haploidentical blood and marrow transplantation is an important alternative transplantation strategy for patients without a HLA-matched or unrelated donor. Delayed platelet recovery following HSCT is a common problem. Thrombopoietin (TPO) is a naturally endogenous cytokine which can stimulate the differentiation of stem cell into the megakaryocyte, promote megakaryocyte proliferation and polyploidization, then ultimately increase the number of platelet. RHuTPO, as a memic of TPO, has been used to increase the number of platelet for many years, but it has not been evaluated in platelet engraftment of patients after haploidentical stem cell transplantation so far. Design and Methods: In this prospective randomized controlled study, we determined the safety, tolerance and effect of rHuTPO when administered to patients after unmanipulated haploidentical blood and marrow transplantation. All the patients received the same conditioning protocol. For the study group patients, rHuTPO was administered daily hypodermically at a single dose of 15000U, initiated the day after peripheral blood stem cell infusion and continued until platelet recovery to >50,000/ul, or till the 21st day after HSCT. Results: From April 2010 to June 2011, a total of 111 patients undergoing HSCT was randomly assigned to receive rHuTPO (n=55) or not (n=56). Serious adverse effects were not observed during the study period. All patients achieved an absolute netrophil count (ANC) of 500/ul in a median day of 12d (range, 10 to 20 days) vs 12.5d (range, 9 to 20 days) respectively in the study group and controlled group (P=0.990).The accumulative rate of platelet recovery to >20,000/ul was signifi cantly higher in the study group than in the controlled group (P=0.0194), but the difference of platelet recovery to >50,000/ul was not statically signifi cant (P=0.378). During the fi rst 100 days, 18 patients in the study group and 14 patients in the controlled group developed idiopathic secondary post-transplant thrombocytopenia (ISPT) (P=0.448). There was also no statically signifi cant difference in acute graft-versus-host disease, chronic graft-versushost disease, cytomegalovirus infection, other complications following HSCT, and overall survival between the two groups.

Conclusions: Our fi ndings shows that rHuTPO can be used in the patients after unmanipulated haploidentical blood and marrow transplantation to promote platelet engraftment and it is safe and well tolerated. Macrophage migration inhibitory factor (MIF) is a key activator of the innate immune system by initiating the infl ammatory response. A constitutional variant (MIF-173CC) in the promoter region of MIF, which causes high levels of MIF, has a detrimental impact in the outcome of patients with severe sepsis and autoimmune diseases. We analyzed the clinical impact of MIF-173CC in acute myeloid leukemia (AML) patients treated with chemotherapy (n=226), median age 50 years (range, 15-74); and in 675 patients submitted to HLA identical sibling allogeneic stem cell transplantation (allo-SCT), median age 44 years (range, 16-69). Cumulative incidence for acute GvHD III-IV, TRM, relapse and DFS was computed with cmprsk pak-kage and Kaplan-Meier. Functional studies for this variant after infl ammatory and cytomegalovirus (CMV) stimuli were performed in 180 healthy individuals. In the AML group, MIF-173CC was associated with lower DFS (CC: 0% vs GG+GC: 26%, p=0.047) and lower OS (CC: 0% vs GG+GC: 23%, p=0.003). This detrimental effect was confi rmed in patients submitted to allo-SCT. Thus, patients with MIF-173CC had a higher frequency of death by sepsis or infection (CC: 60% vs GG+GC: 20%, p=0.003), and higher TRM (CC: 44% vs GG+GC: 24%, p=0.037). We further analyzed the clinical impact of MIF-173CC in two allo-SCT subgroups: those diagnosed with AML (allo-SCT/AML) (n=253) and those diagnosed with lymphoid malignancies (allo-SCT/Ly) (n=115); in the allo-SCT/AML subgroup, patients with MIF-173CC had also a higher frequency of death by sepsis or infection (CC: 75% vs GG+GC: 15%, p=0.002), higher aGVHD III-IV (CC: 58% vs GG+GC: 10%, p=0.005), and higher TRM (CC: 43%, vs GG+GC: 16%, p=0.048). In the allo-SCT/Ly subgroup, patients with MIF-173CC had higher TRM (CC: 75% vs GG+GC: 27%, p=1.7x10 -9 ) and lower OS (CC: 0%, vs GG+GC: 48%, p=0.013). In all associations, multivariate analysis showed an independent risk for MIF-173CC. Functional studies showed higher infl ammatory response and lower CD8+ cells activity after CMV stimuli for MIF-173CC. In conclusion, the detrimental impact of MIF-173CC is presented in the outcome of AML and allo-SCT patients, and an anomalous infl ammatory response with lower lymphocyte response is found for this variant. Regulatory gamma delta T cells (gamma delta Tregs) is a novel kind of immune cells. In our previous study, we found decitabine could increase the induction of gamma delta Tregs with enhanced immunosuppression. We show here several associated mechanisms. Decitabine-conditioned gamma delta Tregs and gamma delta Tregs without decitabine condition (common gamma delta Tregs) were induced as described previously. CFSE-based assay was performed to test their ability to suppress phytoheagglutinin-activated human peripheral blood mononuclear cells (hPBMCs) proliferation. CFSE-labeled hPBMCs proliferation was signifi cantly reduced when cocultured with gamma delta Tregs for 5 days. The inhibition rates were signifi cantly different between decitabine-conditioned gamma delta Treg group and common gamma delta Treg group ( Figure  1A ). To clarify the underlying mechanisms we performed ELISA and fl owcytomix to measure cytokine levels in the supernatant. We noted an elevated IL-10 and TGF-beta1 levels as well as decreased IFN-gamma and TNF-alfa levels in decitabineconditioned gamma delta Treg group ( Figure 1A) . We further confi rmed elevated IL-10 level in the supernatant was due to increased IL-10 production in decitabine-conditioned gamma delta Tregs by western blot. ICOS levels have been reported to correlate with IL-10 synthesis. We found ICOS expression was up-regulated in decitabine-conditioned gamma delta Tregs ( Figure 1B ). We also found less than 3% absolute number of decitabine-conditioned gamma delta Tregs and 12.6% common gamma delta Tregs were reduced on day 5, respectively, revealing elevated TGF-beta1 level was due to improved Foxp3 stability. Xeno-GVHD was induced by hPBMCs transfusion in NOD/SCID mice. Mice were co-injected with either decitabineconditioned gamma delta Tregs or common gamma delta Tregs at a ratio of 1:1. As a result, hPBMCs transfusion alone induced lethal GVHD with average survival time 25±8 days while survival time were 43±5 days and 58±7 days in mice co-injected with common gamma delta Tregs and decitabine-conditioned gamma delta Tregs, respectively (p<0.05). Mean IL-10 plasma levels on day 30 after hPBMCs transfusion were 178.9±11 pg/ ml and 25.6±7 pg/ml in mice co-injected with decitabine-conditioned gamma delta Tregs and common gamma delta Tregs respectively (p<0.01). Altogether, our fi ndings reveal that decitabine-conditioned gamma delta Tregs provide enhanced protection from GVHD via improved Foxp3 stability and ICOS up-regulation. Targeting malignant B-cell lymphomas by antigen specifi c T-cells is a promising strategy to improve treatment protocols. Due to the lack of lymphoma specifi c antigens, strategies have to rely on self-antigens as targets. This bears the problem of simultaneously eradicating non-malignant B-cells. The loss of non-malignant B-cells seems to be an acceptable side effect

S74 in the era of anti-CD20 antibody treatment. Malignant B-cells are diffi cult targets for adoptive T-cell therapy: (i) because of fast growth, they outpace T-cell responses; (ii) B-cells as antigen presenting cells are considered to render T-cells tolerant; (iii) malignant B-cells frequently possess alterations within the apoptosis machinery. Newly developed lambda-OVA transgenic animals express a full OVA protein in B-cell specifi c fashion. To demonstrate tolerance towards OVA we challenged lambda-OVA animals with OVA expressing B-cell lymphoma cells. Lambda-OVA mice failed to reject OVA expressing lymphomas and died of disease progression whereas wild type littermates displayed delayed outgrowth of lymphomas and selection for antigen loss variants. Splenic B-cells from lambda-OVA transgenic mice stimulate OT-1 TCR-transgenic T-cells specifi c for the immunodominant peptide SIINFEKL to secrete IFN-gamma. B-cells from transgene negative control littermates or T-cells or any other hematopoietic cell from lambda-OVA mice failed to induce IFNgamma secretion by OT-1 cells. Adoptive transfer of 2.5 Mio. OT-1 cells into lambda-OVA mice did not result in expansion and B-cell counts in the peripheral blood remained normal. This suggested induction of peripheral tolerance that remained stable after immunization with OVA protein. In contrast, wild type littermates showed expansion of OT-1 cells after immunization. Treatment mice with 300 mg/kg endoxan (EDX) induced lymphopenia. Adoptive transfer of OT-1 T-cells 3 days after EDX treatment resulted in massive expansion of OT-1 cells after OVA immunization and long term depletion of B-cells from the peripheral blood of lambda-OVA mice, whereas control littermates reconstituted their B-cell compartment within 30 days after EDX treatment. Taken together, we show here that lambda-OVA mice are tolerant towards OVA, display a B-cell specifi c expression of OVA and render antigen specifi c T-cells anergic after adoptive transfer. This anergy can be overcome by induction of lymphopenia using EDX treatment resulting in long term depletion of B-cells in the host.

myeloablative in all cases. Overall survival, leukemia free survival (LFS) and relapse incidence (RI) at 3 years were 57±4%, 49±4% and 45±3%, respectively. Only 4 patients (1.9%) had VOD (moderate-2, severe-2) at median day 16 (range, 10-47). One of the patients died from VOD. Non relapse mortality at 3 years was low 6±1%. In multivariate analysis the only prognostic factor that was found to be signifi cant for OS, LFS, RI and NRM was age >50 vs <50 years with p-value of <0.001, <0.001, <0.006 and <0.001, respectively (47±5%, 38±5%, 52±5%, 10±3% vs 68±5%, 76±4%, 32±5% and 0%, respectively). In summary, these results suggest, that similar to the allogeneic setting, VOD is a very uncommon event after AutoSCT using iv Bu in the conditioning regimen translating into a low NRM incidence. We compared TBI/Cy to I.V Bu/Cy conditioning prior to alloSCT from HLA matched unrelated donors in 169 adult pts with AML in Rel 1. 95 pts were given TBI/Cy and 74 Bu/Cy. Median age was 38 (18-62) and 42 (19-72) years in the TBI/Cy vs. Bu/Cy groups, respectively (P<0.005). FAB classifi cation, cytogenetic risk, time from diagnosis to alloSCT, donor gender and CMV serostatus did not differ between the groups. Median year of alloSCT was 2004 vs. 2007, respectively (P<0.001). ATG was used in 35% vs. 71% in the TBI/Cy and Bu/Cy groups, respectively (P<0.0001). 80% and 78% of the TBI/Cy and Bu/Cy groups received PBSC grafts, while 22% and 20% received BM grafts, respectively (P=0.8). Median follow-up was 23 (range, 1-125) and 27 (1-120) months in the TBI/Cy and Bu/Cy groups, respectively. Engraftment was similar, 17 (10-33) and 16 (6-31) days in the TBI/Cy and Bu/Cy groups, respectively (P=0.23). Similarly, acute GVHD (≥Gr II) incidence did not differ between the 2 groups: 33% vs. 37% for the TBI/Cy vs. Bu/Cy, respectively. Death before day 100 occurred in 38% vs. 25% with TBI/Cy vs. Bu/Cy, respectively (P=0.25). 2y NRM was similar between the 2 groups, 28±5% vs. 19±5%, respectively (P=0.2). 2y relapse rate was 54±5% vs. 50±6%, respectively (P=0.56). Induction of remission post alloSCT was higher with Bu/Cy vs TBI/Cy, 72% vs 54% (P=0.02). 2y LFS was also higher with the Bu/Cy vs TBI/Cy, groups: 23 ± 6% vs. 18±4%, respectively (P=0.045). Similarly, 2y OS was signifi cantly higher with Bu/Cy vs. TBI/Cy 37±6% vs. 21±5%, respectively (P=0.013). The main cause of death was disease relapse: 53% and 60%, with TBI/Cy vs. Bu/Cy, respectively (p=0.49). VOD and infection-related deaths did not differ between the groups. In multivariate analysis the interval from diagnosis to transplant (> vs < 16 months) was the most signifi cant prognostic factor for Rel, LFS and OS 25±8% vs 59±4% (p=0.004), 48±9% vs 17±3% (p=0.002) and 41±7% vs 20±4% (p=0.003), respectively. Age, cytogenetic risk groups and use of ATG were not signifi cant prognostic factors for survival. In all, this observational registry based study suggest that in AML pts in fi rst Rel undergoing unrelated transplantation post transplant iv Bu/Cy vs TBI/Cy incuced higher remission rate which results in better LFS and OS. This advantage in favor of the iv Bu/Cy regimen is also possibly due to a lower overall toxicity and improved capacity for salvage therapy. We investigated the impact of occurrence of GVHD on transplantation outcomes in a large cohort of AML pts given allogeneic PBSC after RIC conditioning. Data from 1859 AML pts in fi rst (n=1439) or second (n=420) CR transplanted between 2000 and 2009 following a RIC regimen at EBMT affi liated centres were analyzed. Pts were given PBSC from HLA-identical sibling (MRD, n=1208), or from HLA-matched unrelated donors (MUD, n=651). ATG was given in 269 (22%) MRD and in 267 (41%) MUD recipients, respectively, while 151 (13%) MRD and 165 (25%) MUD recipients received in-vivo T cell depletion with alemtuzumab. The impact of chronic GVHD (cGVHD) on outcomes was assessed using time-dependent multivariate Cox models and in a landmark analysis at 18 months after transplant. The 3-y cumulative incidence of cGVHD was 47%. Fifty-three percent of patients with cGVHD had extensive cGVHD, while the remaining 47% had limited cGVHD. In multivariate analyses, occurrence of grade II-IV aGVHD was associated with a lower risk of relapse (HR=0.8; P=0.04), a higher risk of chronic (HR=2.2; P<0.001) and extensive chronic GVHD (HR=2.8; P<0.001), a higher risk of NRM (HR=2.4 P<0.001), a worsened LFS (HR=1.3; P=0.01), and a worsened OS (HR=1.5; P<0.001). In multivariate time-dependent analyses, occurrence of limited cGVHD was associated with a lower risk of relapse (HR=0.7; P=0.05), comparable NRM (HR=1.4; P=0.16), comparable LFS (HR=0.9; P=0.29) and better OS (HR=0.5; P<0.001), while occurrence of extensive cGVHD was associated with a lower risk of relapse (HR=0.6; P=0.01), higher NRM (HR=3.2; P<0.001), a trend for worsened LFS (HR=1.3; P=0.06) and comparable OS (HR=0.9; P=0.34). In a landmark analysis in patients who were leukemia-free at 18 months after transplantation (n=776), 2-year relapse, NRM, LFS and OS were 16±2%, 2.5±1%, 82±2%, and 89±2%, respectively, in patients without cGVHD before the landmark time-point, versus 9±1% (P=0.001), 8±1% (P<0.001), 83±2% (P=0.65), and 86±2% (P=0.38), respectively, in patients with cGVHD before the landmark time-point. In conclusion, in this cohort of AML patients transplanted in remission, occurrence of cGVHD was associated with a lower risk of relapse that translated to better OS in patients with limited cGVHD but not in those with extensive cGVHD who experienced higher long term NRM. These results highlight the role of the GVT effect in RIC allo-SCT, but also the need for improving the prevention of severe cGVHD in pts receiving RIC allo-SCT. In vivo T cell depletion of the graft with anti-thymocyte globulin (ATG) or with alemtuzumab has been frequently used in the

Caillot (3), F. Witz (4)

); (4)C.H.R. Hôpitaux de Brabois (Vandoeuvre-Les-Nancy, FR); (5)Erasmus MC-Daniel den Hoed Cancer Centre

The role of autologous HSCT in acute lymphoblastic leukemia (ALL) was subject of several prospective studies, though without clear advantage over conventional-dose chemotherapy. More recent analyses demonstrated that results of autoHSCT depend strongly on the level of minimal residual disease (MRD)

The conditioning regimen was based on TBI in 255 (59%) patients. Peripheral blood was used as a source of stem cells in 345 (78%) cases. With a median follow-up of 78 months, the probability of overall survival at 10 years was 42%±3% for Ph(-) ALL and 17%±4% for Ph(+) ALL (p<0.0001), while leukemia-free survival (LFS) was 35%±3% and 12%±3%, respectively (p<0.0001). The conditioning regimen did not affect the 5 year LFS rates in Ph(-) ALL (45%±4% for TBI-based vs. 46%±5% for CHT-based, p=0.83) nor Ph(+) ALL (17%±5% for TBI-based vs. 15%±5% for CHTbased, p=0.58). Also, no effect of the stem cells source could be demonstrated in Ph(-) ALL (57%±6% for BM vs. 45%±4% for PB, p=0.42) and Ph(+) ALL (18%±8% for BM vs. 16%±4% for PB, p=0.45). Importantly, results of autoHSCT for Ph(+) ALL improved over time with 8%±4% LFS for transplantations performed before and 23%±5% after year 2000 (p=0.03). In contrast, the results did not change in a setting of Ph(-) ALL. In conclusion, a signifi cant proportion of patients with both Ph(+) and Ph(-) ALL treated with autoHSCT remain alive and disease-free after 10 years of follow-up. TBI-and non-TBIbased conditioning regimens appear equally effective. Both peripheral blood and bone marrow may be used as a source of stem cells. Improvement in the setting of Ph(+) ALL might be associated with the introduction of tyrosine kinase inhibitors. O344 Outcomes after autologous stem cell transplantation in AML patients using intravenous busulfan-based conditioning regimen

IL); (2)EBMT Acute Leukemia Working Party and Registry

Hospital Reina Sofi a (Córdoba, ES); (6)Hotel Dieu

Hospital Bretonneau (Tours, FR); (14)Ege University Medical School (Izmir, TR); (15)Hospital Constantiaberg Medi-Clinic (Cape Town, ZA); (16)Rome Transplant Network (Rome, IT)

In multivariate analyses (performed in patients given grafts from 10/10 HLA-matched donors), in comparison to the use of ATG, the use of alemtuzumab was associated with higher NRM (HR=2.5, P=0.025), a statistically non-signifi cant but higher relapse rate (HR=1.7, P=0.18), and signifi cantly worse LFS (HR=0.5, P=0.013) and OS (HR=0.4, P=0.002). In summary, this homogeneous cohort of AML patients transplanted in fi rst CR and given PBSC grafts from unrelated donors, the use of alemtuzumab in comparison with ATG was associated with worse LFS and OS. O348 Improved survival by allogeneic haematopoietic stem cell transplantation versus autologous HSCT or chemotherapy as consolidation therapy in AML CR1 patients aged 40-60 years: the role of reduced-intensity conditioning and leukaemia risk category

)University Hospital

-22) following mobilisation. Three patients achieved the goal of a normal CDAI, no drug therapy and normal upper and lower endoscopy but so did 1 patient following mobilisation alone. Serious Adverse Events were common (n=100 to date) with 42 infective episodes requiring or prolonging hospitalisation, following both mobilisation and conditioning and transplantation. There were 7 episodes of viral (re)activation. Temporary fl are of Crohn's disease activity or a need for surgery occurred in 8 patients. Conclusions: Immunoablation and HSCT appears to be an effective treatment for some patients with Crohn's Disease, although full results will be required for a fi rm conclusion. Risks are signifi cant, making it potentially suitable for only a limited number of patients. Data from the whole trial will be needed to judge whether mobilisation alone has any benefi ts. O362 Preliminary results of ASTIMS, a prospective randomised Phase II EBMT trial on autologous HSCT

L. Vuolo (8), D. Currò (1), L. Roccatagliata (1), M. Filippi

University Gabriele D'Annunzio (Chieti-Pescara, IT); (5)Pescara Hospital (Pescara, IT); (6)OO.RR. Bergamo (Bergamo, IT)

Patients (pts) with AML in CR1 currently qualify for alloHSCT in case of intermediate or poor-risk AML. Earlier, we showed limited benefi t by alloHSCT in pts >40 yrs, due to increased non-relapse mortality (NRM) (Blood 2007; 109:3658). Since, transplant outcome has improved and reduced intensity conditioning (RIC) regimen were introduced in this age category by several centers. The HOVON/SAKK group set out to address the question whether alloHSCT would result in better outcome as compared to autoHSCT or chemotherapy (CT) with integrated comparison of myeloablative (MAB) versus RIC and of autoHSCT versus CT. Patients (n=1105) with AML in CR1, aged 40-60 years, entered in 4 prospective HOVON/SAKK trials, were studied, including 237 pts proceeding to MAB and 144 to RIC alloHSCT. 724 pts were consolidated with either a third cycle of CT (n=470) or autoHSCT (n=254). More pts with poorrisk AML proceeded to alloHSCT than to autoHSCT/CT. Recipients of MAB or RIC were comparable as regards AML-risk and EBMT-risk-score, but differed with respect to use of T-celldepletion (TCD) and year of transplant. A trend towards more chronic GVHD was observed among recipients of RIC alloH-SCT (49% vs 39%, p=0.07). Patients with alloHSCT showed better OS (56%±3) than pts receiving alternative consolidation (46% ±2), p<0.001, irrespective of leukemia risk, and with no difference between autoHSCT and CT. Relapse Free Survival (RFS) at 5 years estimated 55±5% following RIC alloHSCT, as determined by 36±4% relapse and 9±3% NRM. RFS at 5 years following MAB alloHSCT estimated 47±3% with relapse 29±3% and NRM 24±3% at 5 years. Multivariate analysis including TCD and year of transplant showed no signifi cant differences between RIC and MAB pts, as regards OS (HR:0.89, p=0.55), RFS (HR:0.94, p=0.75), NRM (HR:0.67, p=0.26), and relapse (HR:1.05, p=0.87). In conclusion, consolidation by alloHSCT signifi cantly improves outcome as compared to either CT or autoHSCT in AML CR1 pts aged 40-60 years, indicating that alloHSCT should be considered standard consolidation therapy in intermediate and poor-risk AML in CR1 up to the age of 60. Since consolidation with RIC alloHSCT produces results at least as good as those following MAB alloHSCT, a prospective randomized trial of RIC vs MAB, that includes younger pts with AML in CR1 as well, is advocated. Blastic plasmacytoid dendritic cell neoplasm (BPDC), formerly known as blastic NK cell lymphoma, is a rare hematopoietic malignancy preferentially involving the skin, bone marrow and lymph nodes. Most patients relapse very soon after aggressive acute leukaemia like chemotherapy but anecdotal long term remissions after consolidating myeloablative allogeneic stem cell transplantation have been reported. Results: Within the EBMT registry 139 patients could be identifi ed who underwent alloSCT (n=100) or autoSCT (n=39) for the diagnosis of BPDCN. In 74 patients histology and immunphenotyping reports could be obtained and central review confi rmed the diagnosis of BPDC in 39 patients (34 alloSCT, 5 autoSCT). The 34 allografted patients were treated with a reduced intensity conditioning regimen (RIC, n=9) or myeloablative conditioning (MAC, n=25) and 19 of 34 patients were transplanted in CR1. After a median follow up time of 28 months (range: 4-77+ months), 11 patients relapsed of whom 8 died due to disease progression. 9 patients died in the absence of relapse. No relapse occurred later than 27 months after transplant. Median disease free survival (DFS) was 15 months (range: 4-77+ months) and median overall survival (OS) was 22 months (range: 8-77+ months; Figure 1a ). However, long-term remissions of up to 77 months after alloSCT could be observed. Patients allografted in CR1 tended to have a superior DFS (p=0.119) and OS (p=0.057; Figure 1b) . MAC was associated with a better OS (p=0.001) which was attributable to the signifi cantly higher non-relapse mortality (NRM) rate of patients after RIC (p=0.014), who had been signifi cantly older (age RIC: 56 years, age MAC: 36 years, p=0.0014). The relapse rate was not different in patients after RIC and MAC, respectively. However, there was no survivor after RIC.

Three of 5 patients in the autoSCT group were transplanted in CR1 of whom 1 patient relapsed after 8 months, 1 patient experienced treatment related mortality and 1 patient remained in CR for 28 months. The 2 remaining patients had more advanced disease at autoSCT and relapsed 4 and 8 months thereafter. Conclusion: AlloSCT is effective in BPDC and might provide curative potential in this otherwise incurable disease, especially when performed in CR1. However, it remains to be shown if the potential benefi t of alloSCT in BPDC is due to conditioning intensity, or if there is a relevant contribution of graft-versusleukaemia activity. Neutropenic enterocolitis (NEC) is a life threatening complication of leukemic and solid tumors patients (pts) treated with chemotherapy (CHT) with mortality rate up to 21-48%. Perforation occurs in 5%-10% of cases. Early diagnosis is crucial to start conservative medical management (CMM) which appears the optimal strategy for most cases. NEC should be always suspected in Neutropenic pts with abdominal pain (AP), fever (F) and diarrhoea (D). Ultrasound (US) was used to evaluate bowel-wall thickening (BWT), and >4 mm is considered diagnostic of NEC. Authors have proposed objective criteria for immediate surgical treatment. Objective: Evaluate prospectively if US can detect early signs of NEC and guide a prompt treatment (CMM or surgical). Method: in the last 4 years abdominal US was performed as only one symptom (or a combination) appeared within 12h from onset: F and/or D and/or AP in CHT-related neutropenic patients (pts). Results: Out of 890 pts 55 episodes were identifi ed (6.1%). Four pts had 2 episodes of NEC. Disease diagnosis were HD (N=10), ALL (N=3), AML (N=14), MM (N=7) and NHL (N=17). Treatment was intensive CHT (N=22), allogenic transplant (N=3), autologous transplant (N=26). At time of diagnosis symptoms were: F+AP+D in 72.5%, D+P in 15%, F+D in 7.5%, F+P 2.5%, P in 2.5%. F and D alone were never present at diagnosis of NEC. As control group we considered pts with CHT related mucositis and pts restaged with US during neutropenia in absence of symptoms. None of them had BWT. Overall 9 patients died (14.5%). Treatment was CMM in 89% of patients and 85% survived. Six pts underwent surgery, guided by US features, during neutropenia, and 50% are alive. Median BWT was 8.3 mm in surviving pts and 11.1mm in deceased. Median time to response from beginning of CMM was 24h and the fi rst sign of was a decrease in AP. Conclusion: US allowed to detect early signs of NEC and to start prompt treatment in this life threatening complication, which was CMM in 89%. The early recognition and intervention allowed 85% survival rate. US guided surgical intervention with 50% survival rate. Images of US and CT were superimposable. Pts who experience a 2° episode of NEC have 50% chance of dying. A close clinical evaluation of the patient by Physician, Surgeon and Radiologist is mandatory. A prompt US in neutropenic patients as just one symptom appears allow to make early diagnosis of this life threatening complication and guide treatment (conservative or surgical). Background: Febrile neutropenia and sepsis are frequent and life-threatening complications in patients with hematological malignancies. Blood cultures (BC) identify a pathogen in only 20 to 30% of febrile episodes, the culturing and pathogen identifi cation process is lengthy, postponing the start of a pathogentargeted treatment. Thereby a molecular tools able to promptly recognize pathogens causing sepsis even despite ongoing antimicrobial therapy is of potential clinical relevance. Methods: We assessed the diagnostic usefulness of the Light-Cycler SeptiFast test (SF), a PCR-based multiplex assay which can be performed on peripheral blood in hematological patients. In this study, blood samples from febrile oncohaematological patients were tested by traditional blood culture (BacT/Alert 3D; bioMerieux) and by a novel commercial real-time PCR assay (LightCycler SeptiFast; Roche Molecular Systems) performed concomitantly at the onset of febrile neutropenia. Results: A total of 869 blood samples were collected from 273 consecutive patients treated for febrile neutropenia at the San Raffaele Hematology over the last three years. Out of the total 869 episodes examined, positive results were detected in 246 samples by SF (28.3%), and in 143 by BC (16.4%). Together, the two methods identifi ed a total of 345 microorganisms in 312 S78 (36%) episodes: Gram-positive bacterial species (74%), Gramnegative bacterial species (23%), and fungal species (3%). Concordance between the two methods was 73%, mainly due to samples that tested negative by culture but positive using the molecular approach (54% of the total positive samples). The cases positive by SF alone were mostly samples from patients with initial concordant results on samples harvested before the administration of a specifi c antimicrobial therapy, or, importantly, sample positive for a clinically relevant agent such as Aspergillus fumigatus, which is hard to detect by the traditional approaches. Conclusion: The LightCycler SeptiFast test gives new insights into the natural history of infection and in the development of new algorithms for the diagnosis of sepsis. Using SF combined with BC improves microbiological documentation in febrile neutropenia particularly in persistent fever despite antibacterial therapy, when a nonresponding bacterial infection or an invasive fungal infection is suspected; results of SF may lead to a more targeted antibiotic therapy early after the onset of fever. Background: Routine use of quinolone prophylaxis in neutropenic haematological patients is widespread. Although it has been reported to result in signifi cant survival advantages, emergence of bacterial resistance has become a concern. Objectives: The aim of this study was to evaluate the clinical and microbiological impact of discontinuing ciprofl oxacin prophy laxis (CP) in haematological patients with prolonged neutro penia. Subgroup analysis was performed on patients undergoing haematopoietic stem cell transplantation (HSCT). Methods: Between August 2009 and July 2011 a total of 225 patients were included in this retrospective study, of which 75 underwent HSCT. 10 were excluded due to documented infection at the start of the conditioning regimen. Three sequential periods of 8 months were analysed: routine use of CP, discontinuation of CP and reintroduction of CP. Patient characteristics were comparable between groups. Clinical endpoints were occurrence of neutropenic fever, bacteraemia, severe sepsis, septic shock and infection related mortality. Microbiological aspects included bacterial isolates from stool and blood cultures and their resistance pattern. Total per protocol antibiotic use for treatment of neutropenic fever and duration of hospitalisation were also analysed. Results: The incidence of neutropenic fever, bacteraemia, severe sepsis, septic shock and infection related mortality did not differ signifi cantly between the three groups. Routine microbiological screening of stools showed a higher prevalence of pathogens in the group not receiving CP, however these were frequently multisensitive. Quinolone resistance disappeared very rapidly after discontinuation of CP. There was no difference in the presence of ESBL producing isolates. An expected, a rise in the occurrence of gram-negative bacteraemia was observed with discontinuation of CP, but again with a multisensitive pattern. Antibiotic consumption and duration of hospitalisation did not differ signifi cantly between groups. Conclusions: Because of increasing quinolone resistance, routine ciprofl oxacin prophylaxis in HSCT patients was discontinued. This did not lead to an increase in serious infectious complications nor to an increase in antibiotic consumption. Quinolone resistance disappeared very rapidly after discontinuation of ciprofl oxacin prophylaxis and also reappeared very quickly after reintroduction. Analysis of this data has led us to suspend routine ciprofl oxacin prophylaxis indefi nitely. We performed a matched cohort study to assess the effect of Palifermin in patients affected by haematological malignancies and who underwent high dose chemotherapy and autologous stem cell transplant (ASCT). Forty patients were treated with Palifermin and were compared to 80 matched controls. Patients treated by Palifermin and controls were matched for the two factors considered important for infection risk such as diagnosis and lenght of neutropenia after HSC infusion. Patients Treated with Palifermin had a diagnosis on Multiple Myeloma (n=23), Non Hodgkin Lymphoma (n=8), Acute Leukemia (n=6), Chronic Lymphoid Leukemia (n. 1) or Amyloidosis (n. 2), 35% of patients were in advanced phase of their disease while 65% in early phase. Twenty-six patients (65%) received HD-PAM (Melphalan 200 mg/m 2 ) and 14 patients ( 32.5%) BEAM, BU-CY or HD-Thiotepa based conditioning. Patients treated with Palifermin compared to controls had a lower rate of FUO (p=0.02). Rate of "infections not related to CVC" (sum of FUO + gram negative bacteremia + pneumonia) were also signifi cantly reduced in patients trated by Palifermin, 25% versus 50% (P=0.01) and the protective effect of Palifermin remained signifi cant also in a multiple logistic regression model [P=0.03] adjusting for potential confounders. The odds ratio of severe mucositis was 30% lower in Palifermin treated patients than in controls (odds ratio: 0.70, 95% CI: 0.33-1.51) but this difference was not signifi cant (P=NS). Fewer patients treated by Palifermin had severe gastrointestinal toxicity (12% versus 65%, P<0.001), morphine utilization (12.5% versus 38.7%, P<0.001), need for total parental nutrition (TPN) (10% versus 68.7%, P<0.001) and blood products transfusions requirement (P=0.04). Average economical costs related to the sum of inpatient stay, TPN, systemic antifungal treatment and blood products, were lower in Palifermin treated patients than in controls (11.985 EUROs versus 15.717, P=0.002) and in Palifermin treated patients fully compensated for the cost of this drug. 75% of patients treated by Palifermin experienced skin rash that did not required interruption of administration. Overall survival in the two groups when stratifi ed for disease status, was not different (p=0.1). In conclusion, in patients undergoing High Dose Chemotherapy and PBSC transplantation, Palifermin reduces gastrointestinal toxicity, resource utilization and "infections not related to CVC". As per expert consensus, this MTC used an unconstrained baseline to account for potential heterogeneity in study designs and patient populations. Using posterior credible intervals estimated from the MTC, we evaluated the following: 1) log-odds of developing proven/probable IFI at Day 180 for ITR, POS and VOR relative to FLU; 2) the probability that each agent was superior to FLU for this endpoint; and 3) the probability that each agent was the best of all agents.

Results: Three open-label and 2 double-blind RCTs, randomizing 2147 patients in total, were included. Based on the MTC estimates, VOR most reduced the probability of proven/probable IFI (yeasts and moulds) at day 180 relative to FLU, closely followed by ITR and POS (see Table) . Conclusion: The results suggest that mould-active azoles are more effective than FLU for preventing overall IFI incidence in alloHCT recipients post-transplant, presumably due to FLU's lack of mould activity. However, the available data did not allow us to clearly distinguish between ITR, POS and VOR in this respect. If mould coverage is desired, other considerations, such as long-term tolerability, cost and ease of use, may therefore be important when selecting the most appropriate azole for IFI prophylaxis in this setting -at least until additional data on comparative effi cacy become available.

Can adequate prophylactic blood levels of itraconazole be achieved with capsule formulation co-administered with an acidic drink?

Background: Invasive fungal infection (IFI) is a leading cause of infectious death post stem cell transplantation (SCT). Antifungal prophylaxis is a key component of fungal management strategies and has been shown to improve outcome in these patients. For SCT patients with low risk of IFI, oral itraconazole is one the main antifungal prophylactic agents used. A previous audit of practice on our unit showed intolerance to itraconazole suspension was 36%. As the alternatives to itraconazole suspension are far more costly, this represented a substantial greater cost to the unit for equivalent prophylactic cover. Itraconazole capsules are not recommended in the setting of SCT due to variable bioavailability; however anecdotal reports suggested adequate blood levels could be achieved if capsules were co-administered with an acidic drink. There was no evidence base for this, and as the practice had begun to be used on the ward, this study attempts to establish whether this method of administration is a viable option for IFI prophylaxis. Objectives: To test the null hypothesis that states 'itraconazole capsules do not achieve adequate drug serum levels and achieve inferior levels compared to itraconazole suspension'. Method: Observational retrospective cohort study comparing two groups of SCT patients (n=189) from the Bristol Bone Marrow Transplant unit. The fi rst arm (n=92) of the study received 'gold standard' itraconazole suspension; the second arm (n=97) received itraconazole capsules, co-administered with an acidic drink. Serum itraconazole levels were then objectively measured to see whether the therapeutic prophylactic level had been achieved. Results: Therapeutic levels of itraconazole were achieved in 84.5% (82/97) of patients in the capsule group, compared to 61.9% (57/92) of patients in the suspension group. The capsule administration method was not inferior, but superior, to the suspension method. A Chi-square test of the frequencies found this result to be highly statistically signifi cant (p=0.0004). A regression analysis looked at whether sex or age could predict the level: sex was not a predictor (p=0.64); age was just outside the level for statistical signifi cance (p=0.055). Conclusion: Adequate prophylactic blood levels of itraconazole can be achieved when the capsule is co-administered with an acidic drink. This represents a viable cost-effective option offering additional fl exibility for the prophylaxis of IFI in stem cell transplantation patients. Invasive aspergillosis (IA) continues to be a major cause of morbidity and mortality in patients undergoing alloSCT. The aim of the study was to analyze molecular risk of IA development and outcome in pts qualifi ed for alloSCT, based on gene polymorphism studies. Material and Methods: Two new tetra-primer ARMS-PCR's (Amplifi cation Refractory Mutation System) were designed using Lasergene Primer Select software for TLR4 and UGTA1 SNP. IFGN and TNFR2 polymorphisms were screened with published primers but using new PCR conditions, common to all reactions. Gene polymorphisms were analyzed in a cohort of 111 pts with acute leukemias (79), chronic leukemias (10) or lymphoid malignances (22) and their sibling (53) or unrelated (58) donors. Pts were conditioned with myeloablative (50) or reduced intensity (61) regimen and grafted with 3.8 (0.9-5.2)x10 6 /kg of CD34+ cells. Standard defi nitions for neutropenic, bacterial, fungal and viral infections were used. Results: Neutrophil recovery occurred in all pts at a median 21 (range; 12-38) days. Infectious complications included neutropenic fever in 102 (92%), CMV infection in 45 (40.5%) and documented IA in 22 (19%) pts (proven-5 and probable-17). Acute GVHD 2-4 grade was seen in 41 (37%), while chronic extensive in 9 (8%) pts. The only pretransplant factor (donor/ recipient gene polymorphisms, age, sex, diagnosis, disease stage) signifi cant for development of documented IA by logistic regression analysis was recipient TLR4 SNP (p=0.023, HR 4.0). Among transplant-related factors (dose and source of stem cells, intensity of conditioning, neutropenia duration, GVHD) signifi cant for IA development was GVHD occurrence (p=0.005, HR=5.1). Over a median follow-up of 14.5 (range; 1.5-65.5) months, 89 (80%) pts were alive with median survival 16.5 (95%CI=1.5-66.6) months. Among 22 (20%) deaths, 10 were related to IA. Factors signifi cant for death from infections by univariate analysis was GVHD occurrence (p=0.007, HR 3.9) documented IA (p<0.001, HR 7.0) and CMV infection (p=0.034, HR 2.5). Factors predicting death from infection by multivariate analysis was only diagnosis of documented IA (p=0.003, HR 23.3). No factors predicting the outcome of IA pts were identifi ed. Conclusion: High mortality rate of IA after alloSCT remains a concern. TLR4 1063A>G SNP strongly predicts IA development. IA risk analysis based on gene polymorphism may help in stratifi cation of pts to offer an individualized treatment. Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) is considered the choice treatment for patients (pts) with Fanconi Anemia (FA). We evaluated the risk factors for late mortality and secondary malignancies in 1-year (yr) survivors in the largest cohort of FA pts post-HSCT ever studied, so far. Pts with FA reported to the EBMT group alive 1 yr after a matched allogeneic HSCT were reviewed. Donor and recipient were matched if HLA A and B were identical at the generic level and HLA DRB1 at the allelic level. Cord blood as source of stem cells was excluded. Data was analyzed using proportional hazards and proportional cause-specifi c hazards models. Between May 1972 and January 2009, 509 FA pts alive 1 yr post-HSCT were included (273 male). Median age at HSCT was 9 yrs (range, 10 months to 44 yrs). 77% of pts had received stem cells from a related donor and bone marrow (80%) was the main source of stem cells. Irradiation was used in 27% of the cohort, while fl udarabine-based regimen was used in 29%. T-cell depletion was used in 41%. In January 2010, 15% (n=74) of the pts had died. Median age at death was 19 yrs. With a median FU of 6 yrs (1 to 28 years), the probability for survival after HSCT was 49% at 20 yrs (95%CI 38-65). The main causes of death were secondary malignancies in 52% of cases (89% of solid tumor) and treatment related mortality in 21%. Cumulative incidence of death and secondary cancer are presented in Figure 1 (89% of solid tumors) . Pts should be transplanted before the age of 10 with bone marrow as source of stem cells. Chronic GvHD still emerges as a major cause for both secondary malignancies and mortality. This study also highlights the need for very long-term FU for FA pts after HSCT. Clinical characteristics of the two groups were as follows: SIB vs UD grafts had comparable age (20 vs 21 median age, p=0.1) but were grafted earlier (152 vs 607 median days, p<0.00001), had less frequently anti-thymocyte globulin (ATG) in the conditioning regimen (50% vs 61%, p<0.0001) and had less frequently radiation conditioning (25% vs 46%, p<0.00001), and more frequently received marrow as a stem cell source (64% vs 55%, p=0.002). In univariate analysis SIB transplants were superior to UD grafts (79% vs 63%, p=0.007). However patients receiving an UD grafts were grafted signifi cantly later, and we thus tested survival in patients grafted ≤/> 6 months. Other variables signifi cant in univariate analysis were patient age (</> 20), interval Dx-Tx(</> 180 days), stem cell source (BM vs PB) and the use of ATG. The actuarial survival of SIB transplants (n=584) (79%), was superior to UD grafts (n=77) (63%) in patients with and interval Dx-TX of ≤180 days, but not in patients who had an interval Dx-Tx of over 180 days (SIBs; n=373 vs UDs; n=433, p=0.4) (69% for both). We then ran a multivariate COX analysis with age, stem cell source, ATG, donor type, interval Dx-Tx. Favourable predictors for survival were, the use of BM as a stem cell source (RR 0.5, p<0.00001), an interval diagnosis-transplant (Dx-Tx) <180 days (RR 0.6, p=0.006), patient age <20 years (RR 0.6, p<0.0001), anti-thymocyte globulin (ATG) in the conditioning (RR 0.7, p=0.01). Donor type (SIB vs UD) was not signifi cant (RR= 0.8, p=0.2). In conclusion we have shown in this study that the overall outcome of UD transplants is getting close to results achieved with SIB grafts; this is particularly true for patients with a longer interval from diagnosis (>180 days). When donor groups are corrected for age, interval Dx-Tx, use of ATG and use of stem cell source, the outcome of UDs and SIBs is comparable. Background: Thymoglobuline (TG), a rabbit anti-thymocyte globulin (ATG), has been shown to be as effective as Lymphoglobuline (LG), a horse ATG as second line immunosuppressive therapy (IST) for aplastic anemia (AA) in previous studies. However, recent prospective studies comparing horse ATG and rabbit ATG have shown that rabbit ATG was inferior to horse ATG as fi rst line treatment as measured by hematologic response rate and overall survival. These confl icting results were mainly derived from studies for adult AA. We retrospectively analyzed the clinical outcome of 453 children with AA who received LG or TG as fi rst line IST between 1992 and 2010.

[O357][O358]

Patients and Methods: The eligibility criteria were newly diagnosed (within 6 months after diagnosis) severe AA children (younger than 18 years) who did not receive specifi c therapies for AA. Two hundred ninety seven children received IST consisted of LG (15 mg/kg/day for 5 days) and cyclosporine (6 mg/kg/day for at least more than 6 months). One hundred fi fty six children received IST consisted of TG (2.3-5.0 mg/kg/day for 5days) and cyclosporine. The median age (8 years vs 7 years) and disease duration before initiating of treatments (16 days vs 18 days) were comparable between LG group and TG group, respectively. However, LG group suffered from more severe disease than TG group (very SAA/SAA; 166/297 vs 57/156). Results: In the LG group, the response rate at 3 and 6 months were 136/294 (45.6%) and 178/292 (61.3%), respectively. In the TG group, the response rate at 3 and 6 months were 66/151 (43.7%) and 87/141 (61.7%), respectively. The response rates were comparable between two groups (p=0.94). However, 10year overall survival rates in the TG group was signifi cantly inferior than the LG group (92.1% vs 84.7%, p=0.01). Because the number of children who rescued by bone marrow transplantation (BMT) was larger in LG than the TG, we compared BMT free survival rate between two groups. When children who received BMT were censored at the time of BMT, survival rates were also signifi cantly different between two groups (94.4% vs 90.3%, p=0.01). Multivariate analysis for BMT free survival rates confi rmed the previous result. Conclusion: These results suggest that use of TG may be reasonable for children with SAA when horse ATG is not available.

Objectives: The management of paroxysmal nocturnal hemoglobinuria (PNH) has drastically changed since the introduction of the fi rst complement inhibitor eculizumab. However, the possible development of C3-mediated extravascular hemolysis (Risitano et al, Blood 2009) may limit the hematological benefi t in some PNH patients. Here we provide preclinical data of a novel strategy for membrane-targeted delivery of complement inhibition. Methods: TT30 (Alexion Pharm.) is a 65 kDa recombinant human fusion protein consisting of the iC3b/C3d-binding region of complement receptor 2 (CR2) and of the inhibitory domain of the complement alternative pathway (CAP) regulator factor H (fH). Here we investigate the effect of TT30 on PNH erythrocytes in an in vitro model consisting of an extended acidifi ed serum assay (EASA). Results: We initially showed that PNH erythrocytes undergo lysis when incubated with normal, ABO-matched, human serum (HNS), while sera containing eculizumab (obtained from PNH patients on eculizumab treatment) resulted in a partial inhibition of hemolysis. However, in these latter conditions, C3-opsonization of all PNH erythrocytes eventually occurs within 24h. In the EASA, TT30 demonstrated a dose-dependent complete inhibition of hemolysis; furthermore, it completely prevents C3opsonization of PNH erythrocytes. These data were obtained with different proportions of erythrocytes, ranging from 2 to 33% (normal-like). TT30 was twice more potent than recombinant fH (either of the full fH or of the moiety included in TT30), and an anti-CR2 monoclonal antibody (mAb) disrupting TT30 binding to C3 halved the effect of TT30. The supposed binding of TT30 to PNH erythrocytes was directly shown by fl ow cytometry (using both anti-fH and anti-CR2 mAbs), fi nally proving that TT30 works in a membrane-bound fashion.By supplying a membrane-targeted activity of fH, TT30 is able to disable CAP activation as soon as it starts due to spontaneous C3-tickover with subsequent C3-binding to erythrocyte surface. In PNH patients, TT30 would result in a complete protection from CAP-mediated intravascular hemolysis, as well as in prevention of possible extravascular hemolysis secondary to progressive C3-opsonization. Conclusion: We report a novel strategy of membrane-targeted selective CAP inhibition. Our in vitro observations that TT30 inhibits both intravascular hemolysis and C3-opsonization of PNH RBCs provided the rationale for the ongoing fi rst-inhumans phase I study in PNH patients.

Autologous Stem Cell Transplantation International Multiple Sclerosis (ASTIMS) is a multicentre, prospective, randomized, single blinded phase II study that compares the activity of autologous stem cell transplantation (AHSCT): BEAM, carmustine, cytosine-arabinoside, etoposide and melphalan) versus Mitoxantrone for the treatment of patients with severe multiple sclerosis (MS) unresponsive to conventional therapy. Inclusion criteria were clinically and laboratory defi nite MS, with an EDSS between 3.5 and 6,5, 18 to 50 years of age, with a secondary progressive (SP) or relapsing remitting (RR) severe clinical course in the last year, defi ned as a deterioration of at least 1 point at EDSS or 0.5 if EDSS was > 5.5, despite conventional therapy and the presence of one or more Gd enhancing areas at MRI. The primary endpoint is the number of new T2 weighted lesions at MRI at 48 months from the randomization, while secondary endpoints are clinical progression, relapses, early and late adverse effects. The study started in 2004 and in November 2009, with 21 enrolled cases, 9 in the AHSCT and 12 in the Mitoxantrone arm, the Steering Committee decided to terminate the accrual of patients, having considered in particular, the diffi culties in the enrolment of cases. Large differences in the possible side effects between the two arms, the refusal of patients to be randomized, the absence of MRI activity in the majority of SP MS patients were among the main reasons for the problems in the patients accrual. The randomized patients were followed in order to obtain the planned MRI and clinical data. Adverse events occurred in both the treatment arms, but serious adverse events were reported only in the AHSCT group under the monitoring of the DSMB. Trial safety was regularly monitored by the DSMB. An interim analysis was carried out in the fi rst 10 randomized cases, that showed that 2 years after the completion of treatment, a mean number of 3.17 new T2 lesions in the Mitoxantrone arm vs. 0.5 in the AHSCT cases (p<0.033). Four cases were lost at follow up. From December 2011, 15 out of 17 evaluable patients were followed for a period longer than 48 months. Mean follow up is now 68 months (range 26-93) and fi nal MRI and clinical evaluation will be analysed at the beginning of 2012.

Background: The autologous hematopoietic stem cell transplantation in early diabetes type 1 is a new treatment modality of this autoimmune disorder. Applied early in the course of the disease has a potential to stop the destruction of beta cells and allow patients to regain exogenous insulin independency. We provide here the follow up data of 15 patients after hematopoietic stem cell transplantation in early diabetes type 1. Methods: The patients were mobilized with cyclophosphamide and G-CSF. The conditioning consisted of cyclophosphamide (50 mg/kg/day on days -5,-4,-3, -2 prior to transplantation) and antithymocyte globulin (Thymoglobulin -of 0.5 mg/kg/day given on day -5, and 1.0 mg/kg/day given on days -4,-3, -2 and -1). Results: The mean time of observation as of October 2011 was 26 months (range 19-40 months). No severe complications were observed during the transplantation and in the post transplantation period. Fourteen out of 15 patients became independent of exogenous insulin after the transplantation -median time without exogenous insulin for all these patients was 18,5 months (range 6-40 months). Nine out of 15 (60%) remain in remission of diabetes (exogenous insulin free) with median follow up of 26 months (range 18,5-40 months). The patients who returned to insulin have reduced requirement compared to the one prior to transplantation (mean 0,20 IU/kg versus 0,34 IU/kg). The average HbA1c concentration was reduced in all patients after the transplantation. Conclusion: The remission of diabetes after hematopoietic stem cell transplantation can be achieved in almost all patients with over 60% of patients remaining insulin-free 2 years after transplantation. Plerixafor is a specifi c, reversible antagonist of the CXCR4 receptor, which can be used in combination with G-CSF and chemotherapy for mobilization of autologous peripheral blood hematopoietic progenitor cells (auto-PBSCT). In this study, we collected data from 211 patients with various hematological disorders (multiple myeloma -MM, n 81, non Hodgkin lymphoma, n. 105, Hodgkin's lymphoma, n 25). 144 patients were considered proven poor mobilizer, since they have had a previous mobilization failure, and 65 were classifi ed as predicted poor mobilizers according to GITMO criteria (BMT, 2011, May 30:1-10). All patients received plerixafor (Mozobil-Genzyme-Sanofi ) plus G-CSF with or without chemotherapy in 23 Italian centres. A total of 140 patients (68%) collected ≥ 2 x 10 6 CD34+ cells/ kg. The collection yield was signifi cantly higher in MM patients (82%) than in HL (70%) and NHL (57%). The target value of > 20 CD34+cell/uL was reached in 65% of the cases after a S84 median of 4.5 fold increase in CD34+ cells/uL following the use of plerixafor. The statistical analysis showed that previous treatment with radiotherapy, lenalidomide, melphalan, the number of lines of chemotherapy, age as well as leukocyte and erythrocyte counts at the time of the mobilization regimen had no signifi cant effect on the effi ciency of collection of CD34+ cells, whereas thrombocytopenia (less than 100 x 10 9 /L at the time of the mobilisation attempt) and previous use of fl udarabine predicted poor mobilisation with plerixafor. Eighty % of the patients who mobilized successfully were transplanted with CD34+ cells mobilised with the combination of G-CSF and plerixafor. Short term recovery of platelet and neutrophil counts after auto-PBSCT, and safety profi le post-PBSCT were found to be comparable with an historical control group transplanted with CD34+ cells which were mobilsed with growth factors and chemotherapy. These data suggest that plerixafor is safe, and may allow the rescue of lymphoma and MM patients, who need auto-PBSCT but failed CD34+ cells mobilization. 

Background: Busulfan (Bu) based conditioning regimens were developed for use in Autologous Stem Cell Transplantation (ASCT) for Hodgkin Lymphoma (HL) with the goal to reduce the risk of relapse without increasing toxicity seen with conventional regimens such as Carmustine, Etoposide, Cytarabine, Melphelan (BEAM). At the Ohio State University, Bu was combined with cyclophosphamide (CY) and VP-16 (BUCYVP) and became the standard conditioning regimen here for 15 years. In 2005, we switched to BEAM regimen. Methods: We retrospectively analyzed 179 patients (pts) with relapsed or refractory HL who underwent ASCT between 1992 and 2010. All pts received at least one chemotherapy regimen prior to ASCT. Kaplan-Meier estimates were used to analyze progression free survival (PFS), overall survival (OS). Cumulative incidence of relapse (CIR) was measured from transplant date until relapse, treating deaths as competing risks using Gray's test. Results: One hundred eleven pts received BUCYVP (62%) and 68 (38%) pts received BEAM as a conditioning regimen prior to ASCT. Median age at transplant was 33 years with 1:1 M:F ratio for BEAM and 2:1 for BUCYVP. Sixty two percent of BEAM pts were diagnosed at an advanced stage vs. 43% BUCYVP pts, with majority in both regiments receiving 2 lines of chemotherapy prior to transplant. Median follow-up from diagnosis and from transplant was longer for BUCYVP at 86 months (range 11-331) and 45 (0-212) months vs. 51 (18-347) and 20 (3-59) months for BEAM respectively. Pts receiving BUCYVP had higher CIR at 6 and 12 months compared to those receiving BEAM, (p=0.04). Cumulative incidence of nonrelapse mortality (NRM) at 12 months was also higher for BUCYVP (p=0.004). Median PFS at 1 year was 61% for BUCYVP and 85% for BEAM (p<0.001) and at 3 years was 41% and 72% respectively (p<0.001). Median OS from diagnosis was also longer at both 3 and 5 years for pts receiving BEAM, 92% and 84% when compared to pts receiving BUCYVP, 81% and 67% respectively (p=0.012). Adverse events were more frequent in the BUCYVP cohort. Veno-occlusive disease was signifi cantly more common in BUCYVP-treated pts, 19% BUCYVP pts vs. 0% BEAM pts (p=0.002). Bacterial infections were observed in 21% of BUCYVP and 13% of BEAM pts (p=0.29). Conclusion: Results of this retrospective study support the use of BEAM as the standard conditioning regimen prior to ASCT for HL. More stringent Bu-based regimen appear to be more toxic without conveying a survival advantage. Patients and methods: The database of EBMT was used to identify patients with EATL who had received autologous and/ or allogeneic SCT in 2000-2007. All centres reporting these patients were contacted to obtain more detailed information in regard to histopathological report/pathology review and to receive information in regard to treatment before stem cell transplantation as well as the most recent follow-up. Results: Eighty-fi ve transplanted EATL patients were identifi ed from the registry. Seventy-three patients had received ASCT and 12 patients allogeneic SCT. Histopathological review data and follow-up data were available in 22 ASCT patients and are included in this report. There were 14 females (64%) and 8 males with a median age of 55 years (49-60). Half of the patients had a history of celiac disease. Ten patients had stage IV disease. The median number of treatment lines before ASCT was 1. Eleven patients (50%) were in fi rst complete or partial remission at the time of ASCT. BEAM was the most commonly used high-dose regimen (17 patients, 77%). The median follow-up time was 45 months from ASCT for the living patients. Thirteen patients (59%) have experienced a relapse or progression with a median time of only 4 months from ASCT. The median disease-free survival (DFS) was 9 months and overall survival (OS) 15 months from ASCT, respectively. OS, DFS, cumulative incidence of relapse and non-relapse mortality (NRM) at 2 years were 45%, 40%, 55% and 4%. Conclusions: ASCT is feasible in selected patients with a low NRM. About 40% transplanted patients achieve a long-term remission. This fi gure seems to better compared to historical series of EATL patients treated without ASCT although patient selection factors have to be taken into account. As a high number of early relapses observed after ASCT, more effective induction therapy is needed in order to improve the outcome of EATL. ASCT is a treatment option in transplant-eligible patients who respond to initial therapy. 

Aims: To defi ne the role of HDT and ASCT in the upfront treatment of HIV-NHL at high risk, in terms of effi cacy and toxicity. We report the interim results of a multicenter study including HDT and ASCT as consolidation after fi rst-line treatment of HIV-NHL at high-risk. Patients and Methods: Eligibility: untreated aggressive HIV-NHL (including DLBCL, plasmablastic, anaplastic); aa-IPI 2-3; age<61; CD4>50/mcL; availability of effective HAART. Burkitt and CNS lymphoma are excluded. Patients (pts) receive R-CHOP-21 (CHOP alone for CD20 neg lymphoma) x6 cycles and, if responsive, SC collection after CTX 4 gr/ms+G-CSF and BEAM. Pts receive HAART during the entire program. Overall survival (OS) at 2 ys is the primary endpoint. Results: Since 2007 to Nov 2011 18 pts entered the study. 13 had DLBCL, 4 plasmablastic and 1 anaplastic. Median age 46.5 ys; PS >1 56%; stage III/IV 22%/78%; LDH >n.v. 94%; aaIPI 2/3 44%/56%. 72% had a prior history of HIV-positivity; 67% were on HAART at NHL diagnosis (detectable HIV 61%); CD4 283 (58-571); HCV pos 50%. 16/18 pts are evaluable for (R)-CHOP response (2 on treatment): 1 pt died with hepatic failure, 2 had prolonged cytopenia (1 with severe hepatic toxicity) that lead to withdrawal of therapy and 13 completed (R)-CHOP therapy; 9 pts had complete remission (CR), 3 partial remission (PR) and 1 disease progression (PD) (ORR 75%, CR 56%, PR 19%, according to ITT). All responsive pts collected CD34+ cells; 2 had early PD and 1 went off protocol because of cardiac EF <50%. Finally, 9 pts received ASCT according to protocol. Treatment-related toxicities (7 pts evaluable) included 2 grade II and 5 grade III GI and 1 grade III hepatic toxicity; 5 FUO, 1 sigmoiditis and 1 VZV infection. TRM 0%. All transplanted pts are alive and relapsefree after 32 ms . No OI are reported. 2-years PFS and OS of the entire series were 66% and 72% (f-up 24 ms, 2-58). Conclusions: This is the fi rst prospective trial addressing the role of HDT and ASCT in fi rst line treatment of HIV-LNH. The procedure was well tolerated and the clinical results highly encouraging. Interim evaluation of OS in this very high risk series of pts is satisfactory and accrual is ongoing. S86 allogeneic stem cell transplantation (allo-SCT) in these patients is unclear. Patients and Methods: We conducted a Phase II trial to evaluate safety and effi cacy of combining total skin electron beam (TSEB) radiation with allo-SCT in patients with advanced disease. Between 06/01 and 12/10, a total of 34 (females=20 & males=14) CTCL patients underwent an allo-SCT. Median age at transplant was 53 years (20-73) and median number of prior treatments was 3 (range 1-7). Thirty-fi ve percent (12/34) had transformed to diffuse large cell lymphoma (DLCL) and 35% had SS. Ninety one percent of patients received TSEB radiation prior to transplant and 31 (91%) received the conditioning regimen fl udarabine, melphalan ± thymoglobulin. Graft-versushost-disease (GVHD) prophylaxis was with tacrolimus/ methotrexate. Median time from diagnosis to transplant was 23 months (range 4-111). Results: Thirty patients engrafted, 2 had autologous reconstitution. Median follow up of surviving patients is 32 months (range, 2-115). The cumulative incidence of acute grade II-IV GVHD and chronic GVHD at 3 years was 41% (95% CI 27-62) and 44% (95% CI 29-65) respectively. The Kaplan-Meier estimate of OS and current progression-free survival (CPFS) at 3 years was 61% (95% CI 40-76%) and 50% (95% CI 32-68%) respectively. Sixteen patients had disease progression/relapse after transplant, 7 of whom achieved another CR after immunomodulation, second allo-SCT, and/or additional chemo therapy. These patients are alive at median follow-up of 18 months (range 2-108) since progression. The cumulative incidence of progression at 3 years for patients with transformed disease was 67% vs. 34% for those who had not transformed (HR 2.5; P=0.07). Patients with SS had a lower rate of progression 37% vs. 50%, HR 0.5; P=0.2) although this did not reach statistical signifi cance likely due to small numbers. At the time of this report 22 patients are alive (19 in CR and 1 with stable disease). Conclusion: This large, single center prospective study of allo-SCT in patients with advanced CTCL demonstrates long-term disease control in a subset of patients. However, relapse is common. Maintenance therapy after transplant or novel conditioning regimens may result in further improvement of these results. TSEB radiation therapy followed by allo-SCT in patients with advanced SS may impact OS.

Allogeneic stem cell transplantation for mantle-cell lymphoma -a report from the prospective trials #060 and #074 of the East German study group haematology/ oncology (OSHO) W. Krüger (1) Introduction: Mantle cell lymphoma (MCL) has a poor prognosis under conventional therapy. AlloSCT is here a promising approach. Methods: Two prospective trials were conducted to investigate the effi cacy of chemotherapy-based conditioning followed by allo-SCT for treatment of MCL: Trial #074 was open for patients with de-novo MCL and #060 for patients requiring salvage therapy. At least a pR was mandatory for proceeding to alloSCT. Conditioning consisted of treosulfan (12 g/m²) and fl udarabin (150 mg/ m²). Busulphan plus cyclophosphamide was optional for younger patients. ATG was given prior to mismatched or mud SCT. Results: 39 mainly male (n=31) patients with a median age of 59y (33-69) have been recruited into both trials (#060: n=15, #074: n=24). In de-novo patients the median MIPI was 5 (2-9). Salvage patients were pre-treated with 8 (6-13) cycles chemotherapy. (Re)-Induction prior to TX consisted mainly of R-CHOP or R-DHAP. 33 patients proceeded to alloTX from mrd or mud. 2 patients died from progressive disease prior to TX, 1 patient had no donor, in one case the diagnosis was revised and 2 patients were withdrawn. 26 patients (79%) were conditioned with Treo/Flu and 7 (21%) with Bu/Cy. 76% (n=25) of patients received a graft from an unrelated donor. Toxicity was moderate and incidence of acute GvHD was 51%. 26 patients are well with a median KI of 100% (range 50%-100%) and alive after SCT with a median follow-up of 18 months (0,3-114) in CCR without differences between both trials. 6 patients have died from infections and cerebral bleeding (n=1) and one from infection related to an acute GvHD IV° due to DLI for relapse (blastic variant) in CR of MCL. Molecular analyses showed a 2-4log reduction of circulating lymphoma cells after chemotherapy alone. Blood became negative by qPCR after allo-SCT in all 5 patients analysed so far. An intermediate increase of circulating lymphoma cells in 3 patients was successfully treated by rituximab, withdrawal of Cy-A and DLI. Conclusion: Allo-SCT is a standard salvage therapy for suitable patients and an option for patients with de-novo MCL. Longterm remissions can be reached and negativity of mrd analyses by qPCR strongly supports curative potential of allo-SCT. Close monitoring of minimal disease allows steering of immunotherapy 'on demand' and helps to avoid an overtreatment with possible associated adverse effects. GvL-effects have curative potential even in the case of relapsed blastic variant of MCL. Patients with T-cell lymphomas have a dismal prognosis when a disease relapse occurs after front line treatment. Here, the outcome of 24 patients with T-and NK-cell lymphomas treated S87 consecutively at our center in a rather uniform approach with allogeneic stem cell transplantation is presented. Results: Patients in the age range of 11 to 65 years were included after a median of 2 prior lines of therapy with these T-NHL subtypes: ALCL-ALC+, ALCL-ALK-, EATL, extranodal NK/ T-cell nasal type, T-PLL, AITL, T-LBL and PTCL-NOS. With few exceptions, the CLAEG regimen, consisting of cladribine, cytosine arabinoside, etoposide, and mandatory G-CSF was administered for remission induction or early consolidation. As soon as possible after recovery the patients were conditioned for allogeneic transplantation with BEAM and alemtuzumab (Faulkner et al., Blood 103:428, 2004). The protocol called for a fi xed dose of 50 mg of the CD52-antibody alemtuzumab in order to provide GvHD prophylaxis and to facilitate tumor control as well. Most patients (n=19) received transplants from unrelated donors, including 5 with one or more mismatches. Toxicity of the transplant procedure was moderate, although infections were frequent and the leading cause of death in two patients. After rapid engraftment only 6 patients showed equal or greater grade II acute GvHD, however one of them died of grade IV disease. Six cases of chronic GvHD occurred. In 5 patients donor lymphocyte infusions were applied. While 12 patients had never achieved a CR before transplantation, 20 of 22 evaluable patients reached a CR after transplantation. With a median follow up of 1252 days (range: day 90 to 1871), 12 patients are in continuous complete remission. For overall survival see the enclosed fi gure. In this group of advanced relapsed or refractory T-NHL allogeneic stem cell transplantation with BEAM-alemtuzumab conditioning is able to induce long-term remissions in a considerable proportion of patients, when performed rapidly after intense remission induction. Joint Session EBMT / WMDA 385 Impact of earth quake/nuclear accident on stem cell transplants in Japan Y. Kodera, S. Chiba, S. Kato, S. Taniguchi for JSHCT, JMDP and JCBBN On JSHCT: The head quarter located in Nagoya was intact. Attacked region was the eastern region of the main island of Japan where 17 hematopoietic stem cell transplant teams located. The earthquake minimally infl uenced autologous and related HSCT. The infl uence to unrelated HSCT would be described later. After June, the HSCT activities of this region returned to usual level. JSHCT made a quick search for the capability to accept recipients and donors from the attacked region. Finally 107 teams through the country expressed the will to accept them. JSHCT organized Nuclear Accident Committee and announced two statements including the support for autologous stem cell storage for potential victims of acute radiation injury. JSHCT sent representatives to Paris, London, Chicago and Minneapolis to report our situation. Monthly report of JSHCT showed no signifi cant decrease of related HSCT in this period comparing to the last year. On JMDP: The head quarter located in Tokyo was partially damaged but the function was maintained. On March 11-14, 43 recipients including 1 recipient in the attacked region were under the preconditioning but the recipient successfully received marrow from Tokyo area. Forty-two of 43 corresponding donors were confi rmed to be safe but 2 including the one who could not be confi rmed its safety out of 43 were switched to CBSCT. Several recipients of 22 under the fi nal preparation moved to other areas to receive unrelated HSCT. New coordination in the region was postponed until June. JMDP set the standards of urgent coordination for the potential victims of acute radiation injury. Monthly report from JMDP showed slight decrease of unrelated HSCT from volunteer donors in March but it recovered in April. On JCBBN: The head quarter located in Tokyo was intact. One of 11 local banks located in the eastern region (Miyagi CBS Bank) and its building got damage without the damage of nitrogen tanks. Allocation of CBS from that bank was postponed until June mostly because of the transportation problems. The other banks were basically intact and monthly report of JCBBN showed certain increase of CBSCT in March. Summary: Although the mega-earthquake/tsunami was almost catastrophic in the eastern region of the main island, its infl uence to stem cell transplant activity was not critical at the country level. The nuclear accident brought us the opportunity to re-consider the role of HSCT to acute radiation injury.[O370]