Allogeneic Transplantation: When is a Mismatch Better than a Match?

Abstract

An accepted fact in allogeneic transplantation from unrelated donors, supported by multiple publications, including large numbers of transplantation patients, is that matching for human leukocyte antigen (HLA) is critical to ensure the best outcomes for patients. The current gold standard is an unrelated donor matched for 10/10 (or 8/8) HLA alleles (at HLA-A, -B, -C, -DRB1 þ/ DQB1). Graft-versus-host disease (GVHD) is increased in the HLA-mismatched setting, and overall survival becomes significantlyworse as the number of mismatched alleles at these loci increases [1,2]. Interestingly, data now suggest that an HLA mismatch may actually be tolerated, or even beneficial, in some transplant settings. This has been shown inparticular forHLA-DPB1where mismatching for this locus at allelic level reduces the risk of disease relapse, both inaTcelledepleted [3]andaTcellereplete setting [4]. Even more provocative is the observation, based on functional studies, that beingmatched ormismatched for DPB1 T cell epitope (TCE) groups provides a better prediction of transplant outcomes, including overall survival, than does consideration of DPB1 allele level matching alone [5]. It is well demonstrated in both the in vivo and in vitro setting that HLA-DPB1 is able to elicit an allogeneic response in a specific manner. This was originally shown in the setting of GVHD [6] but has since also been shown in the setting of graft rejection (using the TCE model) [7]. Proof of principle that HLA-DPB1 mismatches could be a specific target of the graft-versus-leukemia (GVL) effect was shown by Ibisch et al., who demonstrated that primary leukemia blasts both expressed variable levels of DPB1 on the cells surface and could be killed by DPB1-specific T cells [8]. Taking this work into the clinic, Dr. Falkenburg’s group [9] previously showed they could detect polyclonal HLA-DPB1 specific T cells in a patient with B-cell chronic lymphocytic leukaemia who responded to donor lymphocyte infusion (DLI) (from a DPB1mismatched donor) [9]. In this issue, Rutten and colleagues extend these findings in 24 patients undergoing transplantation from a 10/10 HLAmatched, DPB1-mismatched unrelated donor [10]. The authors develop an elegant assay for quantifying HLADPB1especific responses by measuring CD137þ CD4þ T cells