With a continuous decrease in family size, identification of an HLA identical sibling donor for allogeneic hematopoietic stem cell transplantation (HSCT) is often unsuccessful. High resolution HLA-Typing of HLA class I and II loci has opened new avenues for identification of the optimal unrelated donor. Yet, the role of HLA-C matching in HSCT remains controversial. This is particularly true for pediatric transplantation recipients in whom currently little data are available. Here we report on the role of HLA-C matching in a single center cohort of children aged 0–18 years (median: 9,6 years; n=32) who underwent HSCT from an unrelated (n=28) or non-sibling family (n=4) donor for hematopoietic malignancies with a typical disease distribution for pediatric transplant recipients (11 ALL, 6 AML, 8 MDS, 3 JMML, 3 CML, 1 NHL). Patients (pts) with acute leukemia not in remission were excluded from the study. Donor selection was performed prospectively on the basis of HLA A, B low resolution and DR, DQ high resolution-typing with 4 donors exhibiting an isolated HLA-A or -B MM. Donors mismatched for HLA-DRB1 and -DQB1 (n=2) were excluded from the evaluation. In retrospective analysis, the number of donors with additional 1–2 HLA-C antigen MM was high (12/32 pts, 38%). Upon evaluation of the influence of HLA-C MM on critical transplantation parameters, significant correlation between HLA-C MM and severe acute GvHD (aGvHD III–IV; p=0,012) was documented. Of 31 pts evaluable for aGvHD, 7/12 (58%) pts in the HLA-C-MM group suffered from severe aGvHD vs. 2/19 (10%) pts without an HLA-C MM. There was no correlation with severe aGvHD and patient age, sex, underlying disease, stem cell source nor HLA-C KIR epitope or HLA-A/B MM. While none of the 4 pts with isolated HLA-A/B MM showed severe aGVHD, 3/5 pts with combined HLA-A/B plus -C MM exhibited grade III–IV aGvHD. In those pts evaluable for chronic GvHD there was no difference between pts transplanted from an HLA-C MM vs. -matched donor. As in other studies, in our transplantation cohort, aGvHDIII–IV is significantly correlated with transplant-related mortality (TRM) with 6/9 (66%) pts with aGvHDIII–IV dead of complications vs. 3/23 (13%) with no or low grade aGvHD (p=0,007). Accordingly TRM is also increased in HLA-C MM pts with 5/12 (42%) pts vs 5/20 (25%) pts in the HLA-C matched group. High TRM is counterbalanced by a complete absence of relapses in the HLA-C MM group compared to 4/20 relapses in the HLA-C matched group resulting in comparable overall survival with 53+/−16% vs. 69+/−11% after a median observation time of 3.5 vs. 2.9 years, respectively (p=0.59). The high risk of GvHDIII–IV in pts. with HLA-C MM documents the necessity for HLA-C typing for optimal donor selection in T-cell replete transplants. Our data demonstrate that in a T cell-replete transplant setting, increased risk for severe aGvHD may neutralize the beneficial effect that an HLA-C-based KIR MM might have on the graft-versus-leukemia effect. In the absence of an ideally matched donor, awareness of the HLA-C constellation will assist with optimization of immunsuppressive prophylaxis.

Supported by the German Cancer Aid and the Elterninitiative Kinderkrebsklinik e.V., Düsseldorf

Author notes

Corresponding author