Impact of HLA-Mismatch in Unrelated Donor Hematopoietic Stem Cell Transplantation: A Meta-Analysis

Impact of HLA-Mismatch in Unrelated Donor Hematopoietic Stem Cell Transplantation: A Meta-Analysis AJH Natasha Kekre, 1 * Kimberley S. Mak, 2 Konrad H. Stopsack, 3 Moritz Binder, 3 Kazusa Ishii, 4 Elsa Brånvall, 5 and Corey S. Cutler 6 The magnitude of risk associated with 9/10 mismatched unrelated donor (MMURD) hematopoietic stem cell transplantation and that of mismatches at the individual HLA loci remain unclear. We performed a metaanalysis to assess the difference in clinical outcomes between matched unrelated donor (MUD) and MMURD transplantation. A comprehensive search of Medline and Embase for manuscripts regarding transplantation outcomes in primarily adult patients with hematologic malignancies was performed. The pooled effect estimates were calculated using DerSimonian-Laird random effects models. A total of 13 studies were included, reporting on 13,446 transplants. 9/10 MMURD transplantation was associated with worse overall survival compared to 10/10 MUD transplantation (pooled HR: 1.27, 95% CI: 1.12 1.45; n 5 7 studies). Mismatch at HLA-A, -B, or -C was associated with significantly worse overall survival compared to MUD transplantation, while there was no significant difference associated with -DQ or -DPB1 mismatch. Inferior survival associated with HLA-DRB1 mismatch could not be ruled out. Data on acute and chronic graft-versus-host disease were scarce but favored MUD transplantation. In summary, this meta-analysis of the available literature favored MUD over MMURD transplantation in hematologic malignancies and further quantifies the risks associated with specific HLA-allele mismatches. Am. J. Hematol. 91:551 555, 2016. VC 2016 Wiley Periodicals, Inc. Introduction Hematopoietic stem cell transplantation (HSCT) remains the only curative option for many patients with hematologic malignancies. The best donor for HSCT is a human leukocyte antigen (HLA)-matched related or unrelated donor. Unfortunately, based on average family size, less than 30% of patients will have a matched sibling donor [1]. As ethnic diversity increases in Europe and North America, strategies to identify alternative stem cell sources in the absence of an adult matched unrelated donor (MUD) are needed. Currently, a commonly used strategy is to find a mismatched unrelated donor (MMURD). Although often necessary to pursue a MMURD when a MUD donor is not available, the extent to which survival and complications are impacted by this approach remains unclear. A better understanding and quantification of these risks is needed to inform decisions when options such as umbilical cord blood or haploidentical donors are available [2]. The main concern with MMURD transplant remains graft-versus-host disease (GVHD), a common complication that is closely linked to HLAincompatibility. The estimated risk of acute GVHD in MMURD transplant has varied greatly across studies, with cumulative incidences from 13 to 69% [3 6]. The HLA disparity that has been associated with GVHD in MMURD transplantation may also lead to more graft-versus-tumor effect, which is thought to be part of the mechanism of preventing disease relapse after HSCT. Two studies from the Center for International Blood and Marrow Transplant Registry addressed the risk of mismatched HSCT and reported different results. A large registry study of almost 4000 patients reported increased mortality (RR: 1.25; 95% CI: 1.13 1.38) for a single 7/8 allele or antigen mismatch at HLA-A, -B, -C or -DRB1 compared with 8/8 MUD transplantation [7]. A second analysis examined allele and antigen mismatches separately with no increase in mortality (RR: 1.11; 95% CI: 0.91 1.35) for single allele mismatches compared to MUD transplantation [8]. The graft sources were different with mostly bone marrow in the former and peripheral blood stem cells in the latter. Some older studies have also reported increased mortality specifically with HLA-C mismatches [7,9], whereas other studies have not demonstrated this association [6,8,10]. Additional Supporting Information may be found in the online version of this article. 1 Division of Hematology, the Ottawa Hospital Research Institute, Ottawa, Ontario, Canada; 2 Department of Radiation Oncology, Boston Medical Center, Boston University School of Medicine, Boston, Massachusetts; 3 Harvard T. H. Chan School of Public Health, Boston, Massachusetts; 4 Hematology Branch, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland; 5 Department of Hematology, Karolinska University Hospital, Stockholm, Sweden; 6 Division of Hematologic Malignancies, Dana Farber Cancer Institute, Boston, Massachusetts Parts of this manuscript have been presented at The American Society of Hematology Annual Meeting (December 2014, San Francisco, CA) and at the Annual Meeting of the American Society of Blood and Marrow Transplantation/Center for International Blood and Marrow Transplantation Research (February 2015, San Diego, CA). Conflict of interest: Nothing to report. *Correspondence to: Natasha Kekre, MD MPH FRCPC, Blood and Marrow Transplant Program, Ottawa Hospital Research Institute, University of Ottawa, Ottawa, Ontario, Canada. E-mail: nkekre@toh.on.ca Contract grant sponsors: German Academic Exchange Service (DAAD; to M.B. and K.S.), Fernstr om Foundation (to E.B.), Canadian Institutes of Health Research (to N.K.), and Stem Cell Cyclists of the Pan-Mass Challenge (to C.C.). Received for publication: 10 February 2016; Accepted: 19 February 2016 Am. J. Hematol. 91:551 555, 2016. Published online: 1 March 2016 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/ajh.24342 VC 2016 Wiley Periodicals, Inc. doi:10.1002/ajh.24342 American Journal of Hematology, Vol. 91, No. 6, June 2016 551

Kekre et al. In addition, five-loci mapping with inclusion of HLA-DQB1 has become the standard of care and quantification of the risk associated with HLA mismatches in the era of five-loci mapping is needed. The current standard of care is to use a matched donor when available, indicating that randomized controlled trials comparing MMURD and MUD transplantation are unlikely to be performed. To our knowledge, there have been no systematic reviews or metaanalyses that have comprehensively assessed the difference in outcomes between 9/10 MMURD and 10/10 MUD transplantation. Thus, we performed a meta-analysis to compare the outcomes of MMURD and MUD transplantation addressing overall survival and included an analysis of outcomes for specific HLA allele mismatches. Methods Eligibility criteria. We conducted a systematic search of Medline and Embase as well as a manual search of references in relevant articles through October 9, 2015. The search was not restricted by publication date or language. For the Medline search, both keywords and Medical Subject Heading (MeSH) terms were used, including the following: hematopoietic stem cell transplantation, histocompatibility antigens, match, mismatch, lymphoma, leukemia, mortality, and survival (full search strategy in Appendix 1). Studies identified in this search were manually screened by two independent investigators; conflicts were adjudicated by a third investigator. Studies were included if they reported original data from randomized controlled trials, case control studies, or observational studies on HSCT for hematologic malignancies primarily in adult patients, comparing survival outcomes and/or GVHD for 10/10 versus 9/10 unrelated donor transplantation, using high-resolution HLA-typing. HLA-typing had to include HLA-A, -B, - C, -DRB1 and -DQB1 at a minimum. Studies were excluded if they only reported HSCT outcomes for non-malignant diseases or solid tumors, if only an abstract was available, or if the study analyzed patient populations that were substantially overlapping with other included studies. Data extraction. Using a standardized data extraction form, two independent investigators extracted and tabulated all data. Discrepancies were resolved via referencing the original reports and group discussions. The primary measure of interest was the multivariable-adjusted hazard ratio (HR) for overall mortality for 9/10 MMURD compared to 10/10 MUD transplantation. Studies that only included unadjusted hazard or risk ratios were excluded because of concerns for confounding of the relationship between choosing a mismatched donor and mortality. Secondary outcomes included acute and chronic GVHD, reported as relative risks (RRs) from competing-risk and odds ratios (ORs) from logistic regression models. Subgroup analyses for specific HLA alleles were performed. Statistical methods. Pooled effect estimates (HRs) were calculated using DerSimonian-Laird random-effects models, anticipating heterogeneity between studies due to differences in patient populations, treatments, and study methods. To assess heterogeneity, the I 2 statistic was used to describe the percentage of total variation across studies due to underlying inconsistency rather than chance (with 25 50, 50 75, and >75% indicating low, moderate, and high heterogeneity) [11]. Heterogeneity across studies was further evaluated with meta-regression. Predefined factors of heterogeneity included publication year, follow-up duration, age, sex, graft source, conditioning regimen, T-cell depletion and geographical site of study. We addressed overlap between study populations by exclusion and sensitivity analyses. Other sensitivity analyses evaluated changes in the pooled HR over time and the effect of individual studies on the overall pooled HR. Publication bias was assessed with funnel plots and Egger s regression test [12]. All primary analyses were conducted using the metan plug-in set (University of Oxford, United Kingdom) in Stata version 12.1 (StataCorp, College Station, United States). All tests were two-sided and a P-value of 0.05 or less was considered statistically significant. Results Study selection The titles and abstracts of 3533 studies were assessed, with 13 studies fulfilling the inclusion criteria [13 25], reporting on 13,446 transplants. A full summary of the systematic literature search is shown in Fig. 1. Table I presents the main characteristics of the included studies. Seven of the 13 studies provided an overall survival estimate comparing 9/10 MMURD and 10/10 MUD transplantation [13,16,18,20,22,24,25] The most common diseases were acute leukemia and myelodysplastic syndrome. The median percentage of females among study participants was 40% (range 30 49%), with a median reported age at HSCT of 37 years (range 23 60 years). Seven studies reported the use of in or exvivo T-cell depletion [14,17 21,24]. Of these studies that used T-cell depletion, only three studies reported using T-cell depletion in over 50% of patients [18,21,24]. One study [13] stratified results by disease risk, and these subgroups were treated as separate effect estimates throughout the analyses. Overall survival Figure 1. Study search and selection schema. The pooled analysis demonstrated that 9/10 MMURD transplantation was associated with an increased risk of mortality compared to 10/10 MUD transplantation (HR: 1.27, 95% CI: 1.12 1.45, P < 0.001, n 5 7 studies; Fig. 2). There was low heterogeneity across studies (I 2 5 31%). In univariable meta-regression, none of the predefined putative sources of heterogeneity were significantly associated with overall survival (P 0.477 for all comparisons), including length of follow-up, year of publication, median age, gender proportion, intensity of conditioning regimen, graft source (bone marrow versus peripheral blood stem cells), location of study (United States versus other), and use of T-cell depletion. Neither the funnel plot (Supporting Information Fig. 1) nor Egger s test suggested publication bias (P 5 0.500). The overall estimate was robust to the removal of any one study. Including only studies for which HSCT was mostly performed before the year 2000, the pooled HR was 1.46 (95% CI: 1.00 2.13, P 5 0.048), whereas for HSCT performed after 2000, the pooled HR was 1.23 (95% CI: 1.10 1.36, P < 0.001). These estimates were not significantly different (P 5 0.413). The subgroup analyses of mismatches at individual HLA loci revealed an increased risk of mortality with single allele mismatches at HLA-A (HR 5 1.48, 95% CI: 1.19 1.86, P 5 0.001), HLA-B (HR 5 1.45, 95% CI: 1.20 1.75, P < 0.001) and HLA-C (HR 5 1.58, 95% CI: 1.23 2.01, P < 0.001) when compared with MUD transplantation (Fig. 3). For HLA-DRB1, this association was not significant but did not exclude an increased risk (HR 5 1.16, 95%CI: 0.84 1.59, 552 American Journal of Hematology, Vol. 91, No. 6, June 2016 doi:10.1002/ajh.24342

Meta-analysis of mismatched versus matched HSCT TABLE I. Characteristics of All Included Studies Author Year Population N Data Source HSCT Follow-up Age Female (%) Petersdorf 2004 North American 948 Single-center 1985-2003 NR 37 (2-55) 44 Tiercy 2004 European 114 Multi-center 1988-1999 47 (11-124) 34 (1-62) 37 Flomenberg 2004 North American 1874 Registry 1988-1996 NR 42 (18-63) NR Loiseau 2007 European 334 Multi-center 1993-2003 36 (1-144) 23 (1-56) 46 Ludajic 2008 European 161 Multi-center 1995-2007 42 (8-132) 23 (0-51) 46 Crocchiolo 2009 European 805 Registry 1999-2006 25 (3-85) 41 (18-65) 44 Kawase 2009 Asian 4643 Registry 1993-2005 NR 26 (1-69) 40 Walter 2010 North American 85 Single-center 1996-2007 NR 33 (2-52) 50 Shaw 2010 European 488 Registry 1996-2006 55 (3-132) 46 (0-74) 37 F urst 2013 European 2646 Multicenter 1997-2010 24 (NR) 51 (18-75) 40 Passweg 2015 European 802 Registry 2000-2013 NR 44 (1-71) 37 Robin 2015 European 502 Registry 2002-2010 NR 60 (20-76) 39 Smith 2016 North American 44 Single-center 2007-2013 25 (11-81) 56 (32-68) 30 Data are given as median (range). NR: Not reported. Follow-up is reported in months, age at HSCT in years. Figure 2. Hazard ratios for overall survival of 9/10 MMURD versus 10/10 matched unrelated donor transplantation. study [20] reported a RR of 2.55 (95% CI 0.78 8.33) for grade III-IV acute GVHD and 1.45 (95% CI: 0.76 2.76) for chronic GVHD. Similarly, only one to two studies reported acute GVHD outcomes for specific allele mismatches; these are summarized in Supporting Information Table I. There was insufficient data to address chronic GVHD at HLA-specific alleles. Figure 3. Hazard ratios for overall survival according to specific allele mismatches (HR 5 hazard ratio; P 5 P-value for pooled risk estimate 6¼ 1.0; I 2 5 heterogeneity). P 5 0.363). There was no difference in survival with mismatches at HLA-DQB1 (HR 5 0.95, 95% CI: 0.74 1.21, P 5 0.668) and HLA- DPB1 (HR 5 0.99, 95% CI: 0.89 1.10, P 5 0.816). Graft-versus-host disease When comparing 9/10 MMURD to 10/10 MUD transplantation, two studies reported grade II-IV acute GVHD outcomes [18,20], with a pooled HR of 1.45 (95% CI: 0.97 2.17; P 5 0.069; I 2 5 0%). One Discussion Given the curative potential of allogeneic HSCT, the use of MMURD transplantation has increased the availability of HSCT to patients without a matched donor. The results of this meta-analysis suggest that 9/10 MMURD transplantation is associated with a 27% increased risk of overall mortality compared to 10/10 MUD transplantation. There was a low amount of heterogeneity between studies for overall survival. Pre-defined factors known to influence HSCT outcomes, including T-cell depletion, were not significantly associated with this outcome. However, meta-regression may have lacked power to detect factors for heterogeneity given the limited number of included studies. Heterogeneity likely reflects differences in disease risk at the time of HSCT and other patient factors. Only a few studies reported on individual allele mismatches. Amongst 9/10 HLA MMURD transplantation, HLA-A, -B, or -C doi:10.1002/ajh.24342 American Journal of Hematology, Vol. 91, No. 6, June 2016 553

Kekre et al. mismatches all confer similarly increased risk of mortality when compared with MUD transplantation. Mismatches at HLA-DRB1 may be associated with increased mortality, although this did not attain statistical significance. The point estimates for HLA-DQB1 or -DPB1 did not suggest higher mortality with mismatches at these loci, although an up to 21% higher hazard of mortality with HLA-DQB1 mismatches cannot be excluded. The reported overall estimates for HLA- DPB1 do not account for the differential effects of permissive and non-permissive mismatches [26]. This analysis specifically examines HLA mismatching in patients with full typing of HLA-A, -B, -C, -DRB1 and -DQB1. This is a distinction from earlier registry studies only reporting on MMURD based on HLA-A, -B, -C, and -DRB1, without measuring and accounting for HLA-DQ mismatching. The results of our analysis are similar to those observed in these prior registry-based studies that reported an increased risk of overall mortality with HLA mismatch between 25% and 50% [18,21]. This suggests that the magnitude of risk identified in patients with 9/10 MMURD compared with 10/10 MUD transplantation has not changed considerably in the era of five-loci HLA testing when compared to the previous four-loci testing. The concern with MMURD transplantation has always been the increased risk of GVHD, knowing that the activation of donor T-cells against the recipient s HLA antigens is closely related to the degree of HLA-incompatibility. As evidenced by the low number of studies in our analysis, this is insufficiently quantified in the current literature when using five-loci HLA testing. Although exclusion of any single study of the original 16 studies did not change the effect estimates significantly, the concern for substantial overlap of several of the published patient populations remained. Inclusion of overlapping populations may have resulted in variance underestimation and narrowed confidence intervals, therefore three articles reporting similar results were excluded from the final analysis [26 28]. Unmeasured confounding is an inherent limitation of observational studies, and systematic differences between patients with MMURD and MUD are possible. For example, it is conceivable that patients undergoing MMURD transplantation have a higher risk of disease relapse necessitating a mismatched transplant. We therefore only included studies that attempted to minimize confounding by reporting multivariable-adjusted effect measures. It is appropriate that individual studies adjusted for different covariates, since different confounders may be relevant in different study populations. To our knowledge, this is the first meta-analysis to report the outcomes of 9/10 MMURD transplantation compared with 10/10 MUD transplantation. Our results quantify and specify the negative impact on survival that single HLA-allele mismatches confer and provide helpful information when discussing the risks of MMURD and MUD transplantation with patients. Acknowledgments The authors thank Dr. Paul A. Bain, Countway Library of Medicine, Harvard Medical School, Boston, MA, USA for his valuable assistance with developing the search strategy, and Ann Farrell, Mayo Clinic Library, Rochester, MN, USA for assistance with literature search. The authors thank Dr. C. Chung Hsieh from the Harvard T.H. Chan School of Public Health for editing an earlier version of the manuscript. They also thank Dr. Olga Stromberg, Karolinska University Hospital, Stockholm, Sweden for translation of an original article in Russian, as well as Dr. David Chi Wai Chung, Chinese University of Hong Kong, Dr. William Mak, University of Hong Kong, and Dr. Yiyi Yan, Mayo Clinic Rochester, for retrieval and translation of articles in Chinese during the literature review process. Appendix 1: Detailed search strategy PubMed ( Stem Cell Transplantation [Mesh:noexp] OR Hematopoietic Stem Cell Transplantation [mesh] OR stem cell transplant*[tiab]) AND ( Histocompatibility Antigens [Mesh] OR histocompatibility[tiab] OR hla[tiab]) AND (match*[tiab] OR mismatch*[tiab] OR mis match*[tiab] OR compatible[tiab] OR compatibility[tiab]) AND ( Lymphoma [Mesh] OR lymphoma*[tiab] OR Hodgkin*[tiab] OR Leukemia [mesh:noexp] OR Leukemia, Myeloid [Mesh] OR Leukemia, Lymphoid [Mesh] OR Leukemia, Plasma Cell [Mesh] OR leukemia*[tiab] OR Myelodysplastic Syndromes [Mesh] OR myelodysplas*[tiab] OR mds[tiab] OR Myeloproliferative Disorders [Mesh] OR myeloproliferative[tiab] OR myelofibrosis[tiab] OR cml[tiab] OR mf[tiab] OR aml[tiab] OR cll[tiab] OR Multiple Myeloma [Mesh] OR myeloma[tiab] OR malignan*[tiab]) AND ( Mortality [Mesh] OR mortality [sh] OR mortalit*[tiab] OR deaths[tiab] OR Survival [Mesh] OR Survival Analysis [Mesh] OR survival[tiab] OR kaplan meier[tiab] OR proportional hazard*[tiab] OR Graft vs Host Disease [Mesh] OR gvhd[tiab] OR graft vs host [tiab] OR graft versus host [tiab]) Embase stem cell transplantation /exp OR (( stem cell OR stem cells ) NEAR/2 transplant*):ti,ab ( histocompatibility antigen /exp OR histocompatibility:ti,ab OR hla:ab,ti) AND (match*:ti,ab OR mismatch*:ti,ab OR mis match :ti,ab OR mis matching :ti,ab OR mis matched :ti,ab OR mis matches :ti,ab OR compatible:ti,ab OR compatibility:ti,ab) lymphoma /exp OR leukemia /exp OR myelodysplastic syndrome / exp OR myeloproliferative disorder /exp OR lymphoma*:ti,ab OR Hodgkin*:ti,ab OR leukemia*:ti,ab OR myelodysplas*:ti,ab OR mds:ti,ab OR myeloproliferative:ti,ab OR myelofibrosis:ti,ab OR cml:ti,ab OR mf:ti,ab OR aml:ti,ab OR cll:ti,ab OR myeloma:ti,ab OR malignan*:ti,ab mortality /exp OR proportional hazards model /exp OR Kaplan Meier method /exp OR graft versus host reaction /exp OR mortalit*:- ti,ab OR death*:ti,ab OR survival:ti,ab OR kaplan meier :ti,ab OR (proportional NEAR/1 hazard*):ti,ab OR (hazard NEAR/1 (model* OR ratio*)):ti,ab OR gvhd:ti,ab OR graft vs host :ti,ab OR graft versus host :ti,ab References 1. Ballen KK, King RJ, Chitphakdithai P, et al. The national marrow donor program 20 years of unrelated donor hematopoietic cell transplantation. Biol Blood Marrow Transplant 2008;14:2 7. 2. Kekre N, Antin JH. Hematopoietic stem cell transplantation donor sources in the 21st century: Choosing the ideal donor when a perfect match does not exist. Blood 2014;124:334 343. 3. Ciurea SO, Saliba RM, Rondon G, et al. Outcomes of patients with myeloid malignancies treated with allogeneic hematopoietic stem cell transplantation from matched unrelated donors compared with one human leukocyte antigen mismatched related donors using HLA typing at 10 loci. Biol Blood Marrow Transplant 2011;17: 923 929. 4. Nakamae H, Storer BE, Storb R, et al. Low-dose total body irradiation and fludarabine conditioning for HLA class I-mismatched donor stem cell transplantation and immunologic recovery in patients with hematologic malignancies: A multicenter trial. Biol Blood Marrow Transplant 2010;16:384 394. 5. Mehta J, Singhal S, Gee AP, et al. Bone marrow transplantation from partially HLA-mismatched family donors for acute leukemia: Single-center experience of 201 patients. Bone Marrow Transplant 2004;33:389 396. 6. Hauzenberger D, Schaffer M, Ringden O, et al. Outcome of haematopoietic stem cell transplantation in patients transplanted with matched unrelated donors vs allele-mismatched donors: A single centre study. Tissue Antigens 2008;72: 549 558. 7. Lee SJ, Klein J, Haagenson M, et al. High-resolution donor-recipient HLA matching contributes to the success of unrelated donor marrow transplantation. Blood 2007;110:4576 4583. 8. Woolfrey A, Klein JP, Haagenson M, et al. HLA-C antigen mismatch is associated with worse outcome in unrelated donor peripheral blood stem cell transplantation. Biol Blood Marrow Transplant 2011;17:885 892. 9. Ho VT, Kim HT, Aldridge J, et al. Use of matched unrelated donors compared with matched related donors is associated with lower 554 American Journal of Hematology, Vol. 91, No. 6, June 2016 doi:10.1002/ajh.24342

Meta-analysis of mismatched versus matched HSCT relapse and superior progression-free survival after reduced-intensity conditioning hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 2011;17:1196 1204. 10. Sasazuki T, Juji T, Morishima Y, et al. Effect of matching of class I HLA alleles on clinical outcome after transplantation of hematopoietic stem cells from an unrelated donor. Japan Marrow Donor Program. N Engl J Med 1998;339: 1177 1185. 11. Higgins JPT, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557 560. 12. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629 634. 13. Petersdorf EW, Anasetti C, Martin PJ, et al. Limits of HLA mismatching in unrelated hematopoietic cell transplantation. Blood 2004;104: 2976 2980. 14. Tiercy J-M, Passweg J, van Biezen A, et al. Isolated HLA-C mismatches in unrelated donor transplantation for CML. Bone Marrow Transplant 2004;34:249 255. 15. Flomenberg N, Baxter-Lowe LA, Confer D, et al. Impact of HLA class I and class II highresolution matching on outcomes of unrelated donor bone marrow transplantation: HLA-C mismatching is associated with a strong adverse effect on transplantation outcome. Blood 2004; 104:1923 1930. 16. Loiseau P, Busson M, Balere M-L, et al. HLA Association with hematopoietic stem cell transplantation outcome: The number of mismatches at HLA-A, -B, -C, -DRB1, or -DQB1 is strongly associated with overall survival. Biol Blood Marrow Transplant 2007;13:965 974. 17. Ludajic K, Balavarca Y, Bickeb oller H, et al. Impact of HLA-DPB1 allelic and single amino acid mismatches on HSCT. Br J Haematol 2008; 142:436 443. 18. Crocchiolo R, Ciceri F, Fleischhauer K, et al. HLA matching affects clinical outcome of adult patients undergoing haematopoietic SCT from unrelated donors: A study from the Gruppo Italiano Trapianto di Midollo Osseo and Italian Bone Marrow Donor Registry. Bone Marrow Transplant 2009;44:571 577. 19. Kawase T, Morishima Y, Matsuo K, et al. Highrisk HLA allele mismatch combinations responsible for severe acute graft-versus-host disease and implication for its molecular mechanism. Blood 2007;110:2235 2241. 20. Walter RB, Pagel JM, Gooley TA, et al. Comparison of matched unrelated and matched related donor myeloablative hematopoietic cell transplantation for adults with acute myeloid leukemia in first remission. Leukemia 2010;24: 1276 1282. 21. Shaw BE, Mayor NP, Russell NH, et al. Diverging effects of HLA-DPB1 matching status on outcome following unrelated donor transplantation depending on disease stage and the degree of matching for other HLA alleles. Leukemia 2010;24:58 65. 22. F urst D, M uller C, Vucinic V, et al. High-resolution HLA matching in hematopoietic stem cell transplantation: A retrospective collaborative analysis. Blood 2013;122:3220 3229. 23. Passweg JR, Schanz U, Chalandon Y, et al. High-resolution HLA matching in unrelated donor transplantation in Switzerland: Differential impact of class I and class II mismatches may reflect selection of nonimmunogenic or weakly immunogenic DRB1/DQB1 disparities. Bone Marrow Transplant 2015;50:1201 1205. 24. Robin M, Ruggeri A, Labopin M, et al. Comparison of unrelated cord blood and peripheral blood stem cell transplantation in adults with myelodysplastic syndrome after reducedintensity conditioning regimen: A collaborative study from Eurocord (Cord blood Committee of Cellular Therapy & Immunobiology Working Party of EBMT) and Chronic Malignancies Working Party. Biol Blood Marrow Transplant 2015;21:489 495. 25. Smith E, Devlin SM, Kosuri S, et al. CD34- selected allogeneic hematopoietic stem cell transplantation for patients with relapsed, highrisk multiple myeloma. Biol Blood Marrow Transplant 2016;22:258 267. 26. Fleischhauer K, Shaw BE, Gooley T, et al. Effect of T-cell-epitope matching at HLA-DPB1 in recipients of unrelated-donor haemopoietic-cell transplantation: a retrospective study. Lancet Oncol. 2012;13:366 374. 27. Morishima Y, Sasazuki T, Inoko H, et al. The clinical significance of human leukocyte antigen (HLA) allele compatibility in patients receiving a marrow transplant from serologically HLA-A, HLA-B, and HLA-DR matched unrelated donors. Blood 2002;99:4200 4206. 28. Wade JA, Hurley CK, Takemoto SK, et al. HLA mismatching within or outside of cross-reactive groups (CREGs) is associated with similar outcomes after unrelated hematopoietic stem cell transplantation. Blood. 2007;109:4064 4070. doi:10.1002/ajh.24342 American Journal of Hematology, Vol. 91, No. 6, June 2016 555