Search

Introduction The National Cord Blood Program has a large inventory of clinical cord blood units (CBUs) available for transplant (more than 60,000). As a small proportion is distributed each year, new collections should theoretically replenish the inventory, but it is unclear which characteristics should be prioritized for banking. Objective Our goal was to characterize distributed CBUs (in terms of cellularity, race/ethnicity, and match), compare them with the inventory, and devise a collection strategy. Methods Shipments from January 2015 to March 2022 were reviewed (n = 1,046). Race/ethnicity information on CBUs was obtained at the time of collection and was available for 55,934 stored CBUs and 1,020 distributed CBUs. It was also obtained for 394 recipients. HLA match was determined as HLA-A and B antigens and DRB1 allele match (/6), and allelic match as HLA-A, B, C, DRB1 allele match (/8). Total nucleated cells (TNC) and CD34 cells were counted at banking of the CBU, and cell doses (per kg of recipient’s weight) were calculated at shipment. Results The proportion of White Non-Hispanic (WNH) CBUs distributed (38%) was lower than what is present in the inventory (46%), but match levels were higher among these units (Figure 1A and 1B), with 55% of WNH CBUs being 5/8 or higher, vs 40% for minorities’ CBUs. Race match between CBU and recipient led to higher matched transplants, but 41% of race mismatched CBUs still reached 6/8 or higher allele match. The median TNC was 102 × 107 for stored CBUs and 167 × 107 for distributed CBUs (Figure 1C). Better matched units have been selected with lower cell contents (Figure 1D). Discussion CBUs broaden transplant options across ethnic groups, and much focus has been placed on collecting minorities’ CBUs. However, our analysis showed that WNH CBUs are still the largest distributed population, with higher allele match to the recipient (predicting better transplant outcomes), so these CBUs should not be excluded from collection strategies. From the data presented here, the goal for new collections should be to obtain CBUs with higher TNC and CD34 (spreading across all races and ethnicities). These would replenish the pool of high cellularity CBUs that are more often selected for transplant.

Cord blood (CB) stem cell transplantation offers a greater tolerance to HLA mismatches compared to adult-derived stem cell transplants (i.e., bone marrow or peripheral blood stem cells). Indeed, 4/6 or 5/8 HLA-matched CB transplantations are regularly performed for patients lacking a matched unrelated donor. Unfortunately, most banked CB units contain a stem cell dose that is too small to treat adult patients, resulting in only 4% to 5% of available CB units offering an adequate cell dose for prompt engraftment for adult patients. Ex vivo stem cell expansion appears to be an attractive strategy to circumvent this cell dose issue, while also enabling the selection of better HLA-matched CB units. In this study, we retrospectively performed HLA matching simulations to assess how the minimal cell content requirements associated with UM171 CB expansion may improve usability of existing CB unit inventories and donor availability for patients of different races and ethnicities. We analyzed a dataset of 58,971 adults for whom a donor search was initiated through the National Marrow Donor Program Be The Match registry against 142,942 CB units from major U.S. public CB banks listed on the Be The Match registry. Our results show that by enabling selection of smaller CB units, UM171-expanded CB transplantation increases donor availability from 72% to 84% for all patients compared to single unmanipulated CB transplantation. Furthermore, the low cell dose criteria for UM171-expanded CB also increases donor availability compared to double CB transplantation, while enabling better HLA matching between donor and recipient. UM171 expanded CB appears particularly beneficial for racial and ethnic minority patients as CB availability increases from 53% to 78% for African Americans, from 66% to 85% for Hispanics, and from 68% to 84% for Asians and Pacific Islanders, compared to single unmanipulated CB transplantation. In addition, UM171 expansion dramatically improves usability of CB units currently in inventories, as only 4.3% and 0.6% of banked CBs have sufficient cell doses for a 70 kg and 100 kg patient, respectively. UM171 raises this proportion to 53.8% and 20.2%, respectively, making CB banks potentially more cost effective. In conclusion, UM171 expansion allows the use of smaller CB units while also improving access to transplantation for racial and ethnic minorities.

Introduction Cord blood units (CBU) are tested at banking for markers relevant to donor safety (infectious disease markers) and hematopoietic stem cell transplantation (cellularity and HLA typing). As CBUs are being increasingly used for new, non-transplant purposes, additional markers are sometimes needed for CBU selection. Objective We describe here our procedure for testing CBU, looking for specific NK cell characteristics: KIR haplotype B (associated with more activating KIRs), and Fc gamma receptor III A (FCGR3A) 158 V/V (presence of a homozygous valine in position 158) polymorphism (which has a higher affinity for IgG), as requested by some of our clients for non-transplant studies. Methods First, a list of eligible CBUs was established, based on the order specifications. Testing was requested from an outside lab (Histogenetics, Ossining, New York). As CBU DNA is a precious resource, we wanted to limit the number of new samples to be sent by first testing leftover samples from HLA typing, then sending additional samples to the lab, to reach the required number of CBUs. Testing was done sequentially (Figure 1): first, KIR typing, then FCGR3A polymorphism on the CBU with KIR B haplotype. Results A total of 216 samples were selected for KIR testing (130 from stored DNA and 86 from newly sent DNA); 154 (71%) were found to have at least one B haplotype receptor. For those, FCGR3A polymorphism was tested; 60 samples (39%) were F/F, 70 (46%) F/V, and 23 (15%) V/V (Figure 1), in line with expected frequencies from the literature. Average TAT for KIR testing was 13 days and an additional 18 days for FCGRA3. Discussion Performing additional testing can easily be integrated into a bank’s routine and help select CBU for purposes other than transplantation. When several markers are needed, sequential testing is an efficient and cost-saving method.

We conducted a prospective evaluation of cord blood (CB)–derived adoptive cell therapy, after salvage chemotherapy, for patients with advanced myeloid malignancies and poor prognosis. Previously, we reported safety, feasibility, and preliminary efficacy of this approach. We present updated results in 31 patients who received intensive chemotherapy followed by CB infusion and identify predictors of response. To enhance the antileukemic effect, we selected CB units (CBU) with shared inherited paternal antigens and/or noninherited maternal antigens with the recipients. Twenty-eight patients with acute myeloid leukemia (AML), 2 with myelodysplastic syndrome, and 1 in chronic myeloid leukemia myeloid blast crisis were enrolled; 9 had relapsed after allogeneic transplant. Response was defined as

Adolescents and young adults (AYAs) with cancer are a unique group of patients in terms of disease incidence and biology, outcome, and psychosocial needs. This study aims to correlate the risk of graft-versus-host disease (GvHD) and age in a population of children and young adults with acute leukemia undergoing hematopoietic stem cell transplantation (HSCT) in first complete remission (CR).We analyzed the outcome of 153 consecutive children (15 years), AYAs (15-24 years), and adults (25-35 years) with lymphoblastic or myeloid acute leukemia in first CR who underwent HSCT with matched donors after myeloablative conditioning. GvHD prophylaxis was methotrexate and cyclosporine A (CsA) in all patients.The cumulative incidence of grade II-IV acute GvHD (aGvHD) was significantly higher in AYA patients than in children (subdistribution hazard ratio (SHR), 2.04, p = 0.005) or adults (SHR 1.59, p = 0.048). Both gut and skin aGvHD occurred more frequently in AYA patients. Increasing CsA blood levels with age could not fully account for this difference. No difference in terms of grade III-IV aGvHD was observed. Chronic GvHD was more frequent in AYAs (SHR 2.81, p = 0.007) and adults (SHR 2.31, p = 0.033) than in children. No difference in terms of nonrelated mortality and overall survival was observed among the age subgroups.Since GvHD occurrence is strongly correlated to quality of life, specific attention should be paid to AYAs undergoing HSCT. Further studies should investigate the reasons for the excess of GvHD observed in this population.

In absence of available matched-related or unrelated donor (MUD), mismatched unrelated donors (MMUD) and unrelated cord blood (UCB) are both considered to be suitable donors, with similar post-transplant overall survival. In most of these retrospective comparisons, HLA typing of adult donors was performed at eight loci. The aim of this study was to compare the outcome of patients transplanted from UCB (N = 64) with those transplanted from 9/10-HLA MMUD (N = 84) or 10/10-HLA MUD (N = 196). In multivariate analysis, UCB was associated with less Grade II–IV acute GVHD in comparison with MUD (aHR 1.97, 95% CI 1.19–3.27, P = 0.009) and MMUD transplants (aHR 1.79, 95% CI 1.02–3.15, P = 0.042), while the cumulative incidence of chronic GVHD was not significantly different between the three groups. Overall survival (OS), non-relapse mortality, and relapse were not different between MMUD and UCB transplantation, whereas OS was impaired after UCB in comparison with MUD (aHR 0.65, 95% CI 0.43–0.99, P = 0.043). Factors also impacting OS were the donor/recipient CMV serostatus (Donor−/Recipient+ aHR 1.76, 95% CI 1.23–2.52, P = 0.002 compared with D−/R−), the donor/recipient gender combination (Female/Male versus other combinations aHR 1.57, 95% CI 1.11–2.22, P = 0.012) and disease risk (aHR 1.58, 95% CI 1.05–2.38, P = 0.027 for high vs. low risk disease). Our data confirm that UCB and 9/10-HLA MMUD are both relevant alternative options when no 10/10-HLA donor is available. Donor/recipient gender combination and CMV serostatus had a significant impact on survival and may be taken into account, along with donor type, in the setting of MMUD and UCB transplants. Am. J. Hematol. 90:897–903, 2015. © 2015 Wiley Periodicals, Inc.

Pregnancy is the only natural source of anti-HLA immunization. The exact frequency of this immunization remains undetermined as prior studies either used methods with a low sensitivity or were performed late after delivery. We present here the first study on women at delivery evaluating anti-HLA immunization by Luminex. We also attempted to isolate factors influencing immunization, such as soluble HLA-G (sHLA-G) levels and genetic polymorphisms. With Luminex, anti-HLA immunization was observed in 54.4% of the women. As expected, immunization frequency increased with the number of children, reaching 74% in women with >2 deliveries. Among immunized women, strong cytotoxic Ab (as detected by Complement Dependent Cytotoxicity) were associated with a lower level of sHLA-G. None of the studied polymorphisms influenced immunization rate in the whole cohort. Among 94 first pregnant women with no history of miscarriage, the -174 IL-6 gene promoter mutation (G/C) appeared more frequently in non immunized women (69% vs. 45% in immunized ones, p = 0.02). Lastly, the occurrence of a miscarriage before the first live delivery significantly decreased immunization. These results may help to understand mechanisms of pregnancy induced immunization. They also have an impact in the management of previous pregnant women requiring organ or hematopoietic stem cell transplantation.

Background: CBU selection for transplant is based on HLA match and potency/quality determinations associated with the speed of hematopoietic engraftment. Total Nucleated Cell (TNC) count, viable (v) CD45+ and CD34+ cells and Colony Forming Units (CFU) have been used as surrogate markers of hematopoietic stem cells. However, besides TNC measurements, most of the assays still have poor standardization among laboratories. The NCBP has manufactured over 60,000 clinical CBU (as of 07/2018). Of those, 32,413 CBU had pre-cryopreservation samples tested prospectively (period: 03/2008-04/2018) using validated assays for TNC (Sysmex XE-2100), vCD45+ and vCD34+ cells (single platform flow cytometry using ISHAGE strategy and 7-AAD for viability) and HRDI-CFU assay (High Resolution Digital Imaging CFU strategy, Albano et al, Blood 2011;118:485). Among those, 754 CBU have been shipped to US Transplant Centers (TC) for single or double unit transplants. Prior to release for clinical use, CBU had also quality/potency tests conducted prospectively in post-thaw samples from attached segments, using the same assays. Aims and study design: 1) to evaluate the quality/potency of 754 NCBP CBU provided for transplantation (period: 03/24/2008-12/31/2017) by pre-cryopreservation characteristics, post-thaw segment evaluations and their correlations, and 2) to assess their impact on engraftment in two patient subsets after myeloablation: cohort 1, N=99 patients (pts), who received single CBU transplants, and cohort 2, N=139 pts with double unit grafts, who engrafted with the NCBP CBU. All laboratory evaluations were performed by NCBP. Clinical data were provided by CIBMTR; only pts with information on time to engraftment (ANC>500) and established donor chimerism were analyzed. Results: The 754 CBU selected by TC had quality/potency parameters showing a strong correlation between cell numbers and function pre-cryopreservation and post-thaw (Table 1; Pearson correlation). Segment post-thaw median recoveries were 73% (SD: 21.8%) for vCD45+ cells; 69% (SD: 20.7%) for vCD34+ cells and 60% (SD: 22.9%) for CFU, with a strong correlation between pre-cryopreservation and post-thaw values (Pair T Test: 0.74; 0.87 and 0.79, respectively; all P500 was 16 days (range: 7-45). Cohort 2 had 85% adult pts with average age 33 years, average weight 69 kg (range: 11-129) and median time to ANC>500 19 days (range: 6-68). Engrafting CBU median pre-cryopreservation TNC was 2.6x10^7/kg and vCD34+ 1.4x10^5/kg. CBU pre-cryopreservation and segment quality/potency markers showed very good correlation also (Table 1). In cohort 1 all pre-cryopreservation cell doses and segment vCD34+ and CFU correlated with neutrophil engraftment (Table 2). Pts with earlier engraftment (≤16 vs >17 days) received CBU with significantly higher doses of TNC, vCD34+, vCD45+ and CFU (pre-cryopreservation and post-thaw tests). In the recipients of double CB transplants, the engrafting CBU TNC and vCD45+ cell doses did not correlate with neutrophil engraftment, but the vCD34+ and CFU did (Table 2). Similarly, pts with earlier engraftment (≤19 vs >20 days) received CBU with higher doses of vCD34+ and CFU, but not TNC or CD45+. Conclusions: Prospective testing of HPC, CB products manufactured by a single CB Bank over a 10-year period and provided for transplant show consistent, highly significant correlation among pre-cryopreservation graft characteristics and segment measurements. Pre-cryopreservation values were predictive of time to ANC>500 in myeloablated recipients; in those cases, segment evaluation did not improve the correlations. Notably, vCD34+ and CFU correlations are consistently maintained after cryopreservation and thaw ensuring a HPC, CB product with reliable performance for transplantation, cell therapies or use with expansion technologies. Disclosures No relevant conflicts of interest to declare.

International audience; BACKGROUND: Cytomegalovirus (CMV) has a role in chronic rejection and graft loss in kidney transplant (KTx) and lung transplant (LTx) recipients. In addition, donor CMV seropositivity is an independent risk factor for renal graft loss. The anti-CMV response might modulate this risk. Expression of programmed cell death 1 (PD-1), a receptor involved in viral-specific T-cell exhaustion, is influenced by a single nucleotide polymorphism called PD-1.3 (wild-type allele G, variant allele A). METHODS: We performed a retrospective study to assess the impact of PD-1.3 on graft outcome in donor CMV seropositive (D+) and donor CMV seronegative (D-) KTx and LTx. We also performed a case-control study to evaluate the anti-CMVpp65 response according to genotype. RESULTS: PD-1.3 was determined in 1,119 KTx and 181 LTx. In 481 D+ KTx, A allele carriers (24%) experienced significantly less graft failure compared with GG carriers (p = 0.001). Multivariate analysis showed that this association was independent of donor and recipient age, acute rejection episodes, and number of human leukocyte antigen mismatches (hazard ratio, 0.381; 95% confidence interval, 0.209-0.696; p = 0.002). Analysis in 85 D+ LTx showed similar results: A allele carriers had better survival (hazard ratio, 0.302; 95% confidence interval, 0.128-0.716; p = 0.006) and greater 6-month forced expiratory volume (71% ± 17% vs 54% ± 16%, p = 0.001). In D- recipients, PD-1.3 did not affect KTx or LTx outcome. Finally, AA recipients had a stronger anti-CMVpp65 T-cell response than matched GG recipients (p = 0.003). CONCLUSIONS: The A variant allele in PD-1.3 single nucleotide polymorphism improved graft survival in kidney and lung transplant recipients receiving grafts from CMV-positive donors.

Human leukocyte antigen (HLA) and non-HLA immunogenetic factors impact the outcome of hematopoietic stem cell transplantation (HSCT). While in early studies of cord blood transplantation (CBT) the major histocompatibility complex immunogenetic barrier was considered less important in some cases than in HSCT from unrelated donors, and preference was given to the number of cells infused as a success criteria, the most recent data emphasize the critical role for HLA matching. This affects the choice of strategy for immunogenetic typing for cord blood banking. Other immunogenetic systems may be of interest in the context of CBT and warrant further investigation. Finally the role of antibodies is also discussed as well as the overall immunogenetic assessment (HLA and beyond HLA). It will provide a tailored and personalized choice of the best donor for each individual patient.

Aim Use of match probabilities is of great interest to reduce costs and delays in the search for a hematopoietic stem cell donor. Our aim was to estimate the accuracy and relevance of allele match probabilities provided by BMDW for patients (pts) with a challenging cord blood unit (CBU) search. Methods We reviewed pts having a new search in our CB bank (60 000 CBU, search for 4–6/6 match at HLA-A, -B intermediate resolution, -DRB1 high resolution; period: 1/1/2017–1/15/2017). Twenty pts were identified as having a challenging search, defined as Results Of the 209 CBU identified, 168 were categorized as having a potential allele match of 6/8 or higher, with match probabilities calculated. Only 24 (11%) were correctly assigned: one 8/8 allele match, one 7/8 and 22 6/8. A higher proportion of WNH CBU was in the correctly assigned group: 56% vs 31% in the wrongly assigned group, p = 0.008. The average probability of match for actually matched CBU was higher than for not matched CBU (69.3% vs 3.3%, p Conclusions Match probabilities provided by BMDW are a very useful tool to predict the actual compatibility. However, predictions were inaccurate for many not WNH CBU and compatible grafts were identified despite low match probabilities. In pts with challenging searches, low match probability should not prevent allele typing of the CBU, since lower match (5/8 or 4/8) can still be an acceptable transplant option.

Continuing a project presented at the 15th International HLA and Immunogenetics Workshop (IHIWS) on the rarity of HLA alleles, we sought to expand the number of data sources and bioinformatics tools available in the Allele Frequencies Net Database website (AFND, www.allelefrequencies.net). In this 16th IHIWS Rare Alleles project, HLA alleles described in the latest IMGT/HLA Database (release 3.8.0) were queried against different sources including data from registries (stem cell) and from 74 different laboratories around the world. We demonstrated that approximately 40% of the alleles officially named in the IMGT/HLA Database have been reported only once across all different sources. To facilitate the large-scale analysis of rare alleles, we have produced an online tool called the Rare Allele Detector that simplifies the detection of alleles that are considered to be 'very rare', 'rare' or 'frequent'. Tools and associated data can be accessed via the www.allelefrequencies.net website.

International audience; Anti-HLA antibody (Ab) monitoring is essential for the follow-up of transplant patients. However, it can be affected by drugs, especially Ab infused as a conditioning regimen for transplantation. ATG Fresenius, commonly used in this setting, is a polyclonal rabbit Ab raised against the Jurkat human T cell line (HLA-A3, 32; B7, 35). We report here the de novo detection by CDC and flow cytometry (Luminex) of anti-HLA-A3 Ab in the serum of kidney recipients treated with ATG Fresenius. The Ab, of rabbit origin, was detected in every assessable patient (n = 16), with the exception of the HLA-A3 recipients and/or recipients receiving an HLA-A3 graft, before becoming undetectable, at latest, at day 102 after transplantation. It is of major importance that transplantation monitoring laboratories bear in mind the possibility of therapeutic Ab detection when interpreting anti-HLA Ab results.

Sirolimus enhances the effect of apoptotic cell infusion on hematopoietic engraftment and tolerance induction

International audience; Hyperacute rejection is an early complication of transplantation, caused by the presence in the recipient of pre-formed donor-directed antibodies (Ab). We report a case of fatal early graft dysfunction after lung transplantation in a female recipient with no anti-HLA Ab detected before transplantation. Further immunologic studies revealed the presence, in the pre-transplant serum, of low-titer anti-HLA Ab directed against the donor's HLA, detectable only by flow-cytometry screening for panel-reactive antibodies. This observation emphasizes the importance of sensitive Ab detection in patients awaiting lung transplantation. Women should be specifically targeted for such sensitive Ab detection.

International audience; Apoptotic leukocytes are endowed with immunomodulatory properties that can be used to enhance hematopoietic engraftment and prevent graft-versus-host disease (GvHD). This apoptotic cell-induced tolerogenic effect is mediated by host macrophages and not recipient dendritic cells or donor phagocytes present in the bone marrow graft as evidenced by selective cell depletion and trafficking experiments. Furthermore, apoptotic cell infusion is associated with TGF-beta-dependent donor CD4+CD25+ T-cell expansion. Such cells have a regulatory phenotype (CD62L(high) and intracellular CTLA-4+), express high levels of forkhead-box transcription factor p3 (Foxp3) mRNA and exert ex vivo suppressive activity through a cell-to-cell contact mechanism. In vivo CD25 depletion after apoptotic cell infusion prevents the apoptotic cell-induced beneficial effects on engraftment and GvHD occurrence. This highlights the role of regulatory T cells in the tolerogenic effect of apoptotic cell infusion. This novel association between apoptosis and regulatory T-cell expansion may also contribute to preventing deleterious autoimmune responses during normal turnover.

Abstract 2000 Background: Over the past 20 years, allogeneic hematopoietic cell transplantation (HCT) has been increasingly performed with peripheral blood stem cells (PB) and gained benefit from better HLA-typing. Similar long-term survival has been suggested after HLA-matched related and unrelated donor HCT. Till now, the optimal strategy for donor selection is still controversial. We evaluated the impact of donor type (10/10 HLA-matched unrelated (MUD) vs. matched related (MRD)) and other donor traits on long term outcomes of patients with hematologic malignancies after PB HCT. Patients and Methods: We analyzed outcomes of 442 consecutive patients with hematologic malignancies who were transplanted with PB either from MUD (n= 164) or MRD (n=278) at our center from 01/2000 to 12/2010. Median patient age was 48 years (range 7–68). Diseases included 122 acute myelogenous leukemias, 62 non-Hodgkin lymphomas, 60 myelodysplastic syndromes, 57 multiple myelomas, 40 acute lymphoblastic leukemias, 37 myeloproliferative disorders, 29 Hodgkin diseases, 20 chronic myeloid leukemias and 15 chronic lymphocytic leukemias. Two-third of patients underwent HCT following reduced intensity conditioning. Graft-versus-host disease (GVHD) prophylaxis consisted mostly in cyclosporine plus MMF or methotrexate. ATG was used in 19% of HCT. We assessed the impact of donor factors (type, age, gender, CMV serologic status and ABO group) on chronic GVHD (cGVHD), relapse, non relapse mortality (NRM) and overall survival (OS). Concerning donor age, as the upper age limit for voluntary PB donation was usually 60 years, we completed our analysis by performing 3 groups according to donor type and age (MUD, MRD Donors: Median donor age was 40 years (range 18–72). Most young donors were MUD ( Results: The median follow-up (FU) was 36 months (range 2–133) and 25% of patients had a FU of at least 60 months. The cumulative incidence (CIf) of cGVHD at 2 years was 58%. In multivariate analysis, sex mismatch (female > male) increased risk of cGVHD (HR: 1.41 [95% CI 1.05–1.88], P=.02) while MRD≥60y resulted in lower risk (HR: 0.48 [0.25–0.94], P=.031). Donor type by itself did not impact on cGVHD (58% with MRD and 59% with MUD). At 5 years, the CIf of relapse was 34% and was higher with MRD than MUD (39% vs. 24%, P=.038). Adjusted for disease risk, conditioning and infused cells count, only MRD≥60y resulted in significant higher risk of relapse than MUD (HR 2.41[1.26–4.62], P=.008) while MRD Conclusion: Donor/recipient gender parity, donor type and age appeared as significant predictive factors of long term outcomes after HLA-matched PB HCT. Nor donor CMV status nor ABO group seemed to impact on outcomes in our cohort. The selection of a sex mismatched donor (female>male) was associated with significant higher risk of cGVHD. Using PB as graft source, HLA 10/10 MUD provided higher NRM but better disease control and similar OS than MRD. Having combined donor type and age, we observed notable poor outcome (high relapse rate and low OS) for patients transplanted with MRD≥60y in our cohort. Given those results, one may question HCT with old MRD when a younger MUD is available. Disclosures: No relevant conflicts of interest to declare.

Abstract 3135 Background: When an HLA-matched donor is not available for allogeneic Hematopoietic Stem Cell Transplantation (HSCT), the use of an alternative HLA-mismatched stem cell source may be considered. In Europe, HLA compatibility for HSCT is calculated for 10 alleles (HLA-A, -B, -C, -DRB1, -DQB1). The aim of this study was to retrospectively compare outcomes after transplantation from 10/10 HLA-MUD, 9/10 HLA-MMUD and UCB performed at the Hospital Saint Louis (Paris, Fr). Methods: Patients receiving a first allogeneic transplantation from a 10/10, 9/10 UD or UCB from 2000 to 2011 were included. High resolution HLA typing was performed by PCR SSO and SSP for HLA loci: A, B, C, DRB1 and DQB1. The following variables were studied as risk factors for transplant outcomes: disease, disease risk (standard/high risk), age at transplant, gender, donor/recipient sex-matching, ABO matching, donor/recipient CMV status, conditioning regimen, and use of anti-thymoglobulin (ATG). Results: 355 consecutive patients with hematologic malignancies were analyzed. One hundred and ninety-six were transplanted with MUD, 84 with MMUD (mismatches for HLA-A: 16%, -B: 16%, -C: 39%, -DRB1: 8%, -DQB1: 21%) and 75 with UCB (52% with single and 48% with two UCB unit; 87% of all UCB transplants were 4–6/6 HLA-matched). Median patient age was 31 (range: 5–55). Patient characteristics differed between the 3 groups: (i) median age at transplant: 36 in MUD, 31 in MMUD, 22 in UCB (p Cumulative incidences of grade II-IV and grade III-IV acute GvHD disease (aGvHD) were 61% (66% for MUD, 60% for MMUD and 48% for UCB) and 17% (17%, 24% and 15%), respectively. Three-year cumulative incidence of chronic GvHD (cGvHD) was 46% (51% for MUD, 49% for MMUD and 29% for UCB). Three-year NRM was 34% (28% for MUD, 31% for MMUD and 51% for UCB). Graft failure occurred in 15% of UCB patient, 8% in MMUD group and 3% in MUD group (significant difference between UCB and MUD: OR=5.44, 95%CI 1.93–15.3, p=0.001). Multivariate analysis is summarized in table 1. It showed that MMUD tented to have a higher incidence of aGvHD III-IV than MUD and UCB. UCB had a lower incidence of cGvHD than MMUD. No significant effect of HSCT source on NRM was demonstrated. Conclusion: Compared to MMUD, UCB-HSCT induces less cGvHD and non significantly increased NRM. Compared to MUD, MMUD and UCB-HSCT did not result in a clear increase of NRM. UCB and 9/10 HLA-MMUD are both suitable stem cell sources for patients who cannot benefit from 10/10 HLA-MUD transplant. Disclosures: No relevant conflicts of interest to declare.

We have recently showed in a murine bone marrow transplantation (BMT) model that apoptotic cell infusion simultaneously with a bone marrow (BM) graft favors hematopoietic engraftment, and induces a long term tolerance restricted to BM cell allo-antigens. This effect is observed whatever the origin of apoptotic cells and does not lead to the development of autoimmunity or GvHD. This favorable effect on engraftment was obtained without immunosuppressive (IS) drugs. In human BMT, IS treatment is always required. The aim of this study was to investigate the interactions between apoptotic cells and IS drugs, especially since some of them might interfere with other strategies of tolerance induction. The roles of ciclosporin A (CSA), mycophenolate mofetil (MMF), sirolimus (Rapa) and LF150195 (LF), a deoxyspergualin analog, were tested. Recipient Balb/c mice (H2d) were sublethally irradiated (6 Gy) 1 day before receiving 106 BM cells with or without apoptotic cells (5x106 γ-irradiated splenocytes) from donor FvB mice (H2q). Drugs were injected IP: CSA 50mg/kg/d from day 0 (d0) to d14, MMF 20mg/kg/d from d0 to d14, Rapa 1.5mg/kg/d from d0 to d14, and LF 0.625mg/kg/d from d0 to d10. Engraftment was assessed by flow cytometry analysis in peripheral blood. Skin grafts from BM donor mice were performed >70 days after BMT, to assess the influence of IS drugs on our tolerance induction protocol. CSA treatment didn’t favor engraftment in mice receiving BM alone (22% engraftment vs 20% without IS drug, p=NS), and induced only transient chimerism (