From nuclear to a global family: more donors for MDS

In this issue of Blood, Saber et al report similar survival rates for myelodysplastic syndrome (MDS) patients after allogeneic stem cell transplantation (ASCT) for both 8 of 8 matched unrelated and HLA-identical donors, but worse survival rates for 7 of 8 matched unrelated donors.¹  Historically, best results in ASCT for MDS have been obtained with HLA-identical sibling donors, but recent improvement of stem cell transplantation from unrelated donors has led to an increased donor pool for MDS patients.

MDS is a heterogeneous group of clonal stem cell disorders characterized by cytopenias, abnormal blasts, and the risk of transformation into acute myelogenous leukemia (AML). In the United States, MDS affects 75 per 100 000 persons >65 years of age per year.²  Allogeneic stem cell transplantation (ASCT) is considered to be a curative treatment option for MDS patients, and its role in younger patients is well established. However, its broader use has been hampered by the inherent complications of the transplant procedure leading to therapy-related morbidity and mortality. Especially stem cell transplantation with unrelated donors for MDS has resulted in a nonrelapse mortality of >40%, as reported in early register studies by the European Group for Blood and Marrow Transplantation (EBMT).³  Even in more recent years when the NMDP reported some improvement in disease-free survival for matched unrelated donors (MUD) transplantation, results were still less than encouraging.⁴ 

The introduction of less toxic conditioning regimens and the development of unrelated donor registries (with now more than 18 million potential volunteer donors worldwide) have rapidly increased the numbers of allogeneic stem cell transplantations for hematological malignancies and especially for MDS. In the register of EBMT the numbers of ASCT for MDS/AML have increased from 737 in 2002 to 1636 in 2010, and the percentage of MDS patients aged 60 years and older has increased from 10% to 33%.⁵  According to the Center of International Blood and Marrow Transplant Research (CIBMTR), MDS has become the third most frequent indication for ASCT (www.cibmtr.org).⁶  In general, results of unrelated stem cell transplantation have been improved by careful donor selection using high-resolution allele typing of HLA-A, HLA-B, HLA-C and HLA-DR (8 of 8). More recently, in AML, similar survival rates for HLA-identical and 8 of 8 and 7 of 8 matched unrelated donor stem cell transplantations have been reported.^7,8  Not surprisingly, the EBMT survey of 2011 showed 54% of all ASCT in Europe were done with unrelated donors in contrast with 39% with HLA-identical siblings.⁹  Therefore, the article from Saber et al,¹  on behalf of CIBMTR, is timely for the MDS scientific community that investigates the impact of 8 of 8 matched unrelated and 7 of 8 mismatched unrelated donor cell transplantation in comparison with HLA-identical sibling transplantation. This report is in line with the described increase in the use of unrelated donors in MDS as 75% of the included patients have been transplanted from unrelated donors. The obtained results from AML studies confer in part that transplantation of MDS patients with 8 of 8 matched unrelated donors has resulted in similar survival rates to transplantation from HLA-identical siblings. This, however, contrasts with the AML transplant ASCT from 7 of 8 matched unrelated donors that resulted in a worse outcome for MDS.

The main message from the study is that if no HLA-identical sibling is available, then a fully matched (8 of 8) unrelated donor can be used as a donor source with a similar outcome; but if the unrelated donor is only 7 of 8 allele matched, then the treating physician should be aware of a worse outcome. This is certainly good news for MDS patients, increasing the number of those for whom a curative transplant approach can be offered because only about 30% will have an HLA-identical sibling. However, a fly in the ointment remains for MDS patients searching for a suitable unrelated donor. Despite comparable survival rates between 8 of 8 MUDs and HLA-identical siblings in the multivariate analysis, therapy-related mortality remains significantly higher in both 8 of 8 and 7 of 8 MUD groups than in the HLA-identical sibling group in the univariate as well as the multivariate analysis. This is most likely due to a higher incidence of graft-versus-host disease after unrelated donor transplantation. After adjustment for other risk factors, the 3-year adjusted probability of overall survival was higher after HLA-identical sibling (47%) than after 8 of 8 MUD (38%) and 7 of 8 MUD (31%) transplantation, suggesting that an HLA-identical sibling remains the donor of choice for ASCT in MDS patients.

Why are the results different from the recently reported AML CIBMTR study? One reason could be a selection bias, which is supported by significant different median times from diagnosis to transplant between HLA-identical sibling donors (6 months) to 8 of 8 MUD (8 months) and 7 of 8 MUD donors (10 months). The delay of the transplant with unrelated donors, which can result in disease progression, can be caused by a time-consuming donor search, but also by a physician’s choice to postpone the transplant until signs of disease progression occur. The major drawback of this large registry study is the lack of sufficient data, such as transfusion dependency and frequency, and the international prognostic scoring system including cytogenetics and comorbidity index at time of transplantation, which would be helpful to compare different donor sources regarding disease-specific and other transplant specific risk factors. The insight into pathogenesis of MDS and the field of ASCT, as well as conventional therapy, are rapidly moving so that the role of ASCT from related and unrelated donors needs to be more determined by prospective trials, including nontransplant approaches, which are now ongoing (NCT 01404741) or planned in Europe and the United States.

In summary, the CIBMTR study of Saber et al,¹  in comparison with historical reports, demonstrated a marked improvement of unrelated stem cell transplantation for MDS patients, but the results of 7 of 8 MUD remain worse and those of 8 of 8 MUD are comparable only with HLA-identical sibling. However, reflecting recent progress, it is likely that in the near future the optimal donor will be in the “global family” rather than in the “nuclear family,” because the results of unrelated stem cell transplantation will be further improved by selecting young MUD for elderly patients,¹⁰  selecting donor beyond HLA, such as selection by using the killer-immunoglobulin-like receptor (KIR) system,¹¹  and better control of graft-versus-host disease, which remains a major cause of treatment-related mortality. Until this becomes a reality, an HLA-identical sibling remains the donor of choice.