HLA factors in transplantation for nonmalignant hematologic disorders

In this issue of Blood, a study by Horan and colleagues shows that differences in the HLA alleles of patients and unrelated donors in hematopoietic stem cell transplantation (HSCT) for nonmalignant diseases result in increased risk for adverse treatment outcome.¹  This is the largest dataset examined so far for the evaluation of HLA mismatches in HSCT for nonmalignant diseases. It includes predominantly pediatric patients diagnosed with 39 diseases. Many patients received nonmyeloablative conditioning; a significant proportion of the infused grafts were depleted of T-lymphocytes; 6 diseases account for 77% of the cases.

Currently, the criteria used for selection of unrelated donors for nonmalignant diseases derive from studies performed in HSCT for patients with malignant diseases² ; in the latter it was found that the HLA mismatches associated with patient mortality. The study by Horan et al shows that HLA mismatches in HLA-A, -B, -C, and DRB1 loci also have a significant impact in outcome of HSCT for nonmalignant diseases. This study provides useful insights that can be applied to the definition of unrelated donor selection criteria for nonmalignant diseases. The nonmalignant disease cohort¹  differs significantly from those examining HSCT for hematologic malignancies²  in age, conditioning regimens, and graft composition. In the nonmalignant disease study the incidence of graft failure was at least 2 to 3 times higher than in the cohorts of malignant diseases. In nonmalignant disease transplantation, many patient deaths (29.8%) were associated with graft failure. The multivariate analyses showed that the occurrence of a single or a double mismatch in HLA-A, -B, -C, or DRB1 loci associated with graft failure; 2 HLA mismatches were also associated with mortality. The single HLA mismatch associated with patient death only in the univariate analysis. Interestingly, in nonmalignant disease HSCT, the mismatched HLA loci did not associate with any type or grade of graft-versus-host disease (GVHD). Horan and colleagues noted that the absence of an association between HLA mismatch and acute GVHD was likely because most of the patients received a lymphocyte-depleting antibody and/or received an ex vivo T cell–depleted graft. These findings contrast with those made in HSCT for malignant diseases; in the latter, the HLA mismatches associate risks for acute GVHD, nonrelapse mortality, or treatment-related mortality. Therefore, the causes of death in HSCT for malignant and nonmalignant diseases appear to differ significantly. The increased incidence of graft failure in nonmalignant disease transplantation may be caused by preserved or enhanced patient immune competence as they have not received chemotherapy prior to the preparative regimen.

This study was limited in sample size and a more detailed analysis of the impact in individual HLA loci or of the type of mismatch was not possible. Horan et al indicate that differences at either of the classically matched HLA loci (A, B, C, DRB1) at either allele or antigen level mismatches appear to have an equivalent impact in outcome. In nonmalignant disease study the mismatches in HLA-DQ and DP loci were not associated with poor outcomes. However, an adverse effect of mismatches in these loci cannot be fully excluded because the group of patients presenting mismatches in HLA-A, -B, -C, or DRB1 loci had more mismatches in HLA-DQ and/or DP loci.

In the study by Horan et al the impact of preformed allo-antibodies in the patients' serum that could react with the mismatched donor antigens (DSAs) was not evaluated. Other studies of HSCT in malignant diseases showed an association between DSA and primary graft failure (PGF).^3,4  Typically, many nonmalignant disease patients have preserved or enhanced immune function and often receive blood transfusions; it is likely that in nonmalignant diseases the incidence of humoral HLA allo-immunization may be even higher than in patients with hematologic malignancies. Therefore, in the nonmalignant disease HSCT setting, the presence of anti-HLA donor antigens may be more significant in outcome by increasing the risk of PGF. Future nonmalignant disease studies should be performed to directly investigate the impact of DSA on outcome. It is suggested that the evaluation of anti-HLA antibodies reactive with donor antigens should now be performed prospectively in HSCT for nonmalignant diseases for the selection of unrelated donors.

Although humoral sensitization plays a significant role in graft rejection, the patient's T-cell HLA allo-reactivity may also cause graft failure. In HSCT for malignant diseases, 30% to 40% of PGFs present anti-HLA antibodies reactive with the mismatched donor antigens^3,4 ; it can be argued that a significant proportion of the remaining PGF cases may have resulted from rejection mediated by patients' T-cell lymphocytes. In the nonmalignant disease HSCT study,¹  HLA mismatches associate with graft failure at both allele and antigen levels. It can be argued that because most of the HLA allele level mismatches are not recognized by allo-antibodies, the negative impact of these mismatches may result from T-cell allo-recognition.

The present study sets grounds for further investigation of HLA matching in nonmalignant diseases, specifically in factors determining risk for graft failure. Investigations in this area may allow expanding HSC sources from either unrelated cord blood units or from donors presenting a mismatch in 1 HLA haplotype. These graft sources are easily and rapidly available for children with nonmalignant diseases; however, due to higher risk of graft failure they are not always preferred. Therefore, further insight into histcompatibility factors may allow the development of strategies for identifying patient/donor pairs with low risk for graft rejection; the resulting criteria may allow finding allogeneic HSC sources for almost all patients.