Distinct HLA-DPB1 Matching of Unrelated Donors for Malignant and Non-Malignant Indications Is Critical in Pediatric Hematopoietic Stem Cell Transplantation

Hematopoietic stem cell transplantation (HSCT) has become a standard procedure to reconstitute a malignant or defective hematopoietic system for a series of indications, especially when other therapy options are not feasible. Indications are typically malignant transformation within the hemopoietic system, as leukemia or lymphoma, but can, especially by pediatric patients, also comprise non-malignant indications like severe immune defects, hemoglobinopathies, metabolic disorders, or neurometabolic diseases. So far, most studies comprise only adult patients or cohorts where children are included, in a malignant indication for HSCT. The optimal and first choice is a HLA-identical (matched) sibling donor (MSD), however, only 25-30% of patients will have such a donor. The current "gold standard" for matching unrelated donors (MURD) is 10/10 HLA alleles match, which comprises a match on three loci in HLA class I (A,B,C; exons 2-3) and two loci in class II loci (DRB1, DQB1; exon 2). As a consequence of the engraftment of a new donor-based immune system graft-versus-host disease (GvHD) might occur and the overall survival becomes worse as the number of mismatched alleles increased. Very recently, matching of HLA-DPB1 is increasingly considered of importance. HLA-DPB1 is a hypomorphic gene locus and matching can be simiplified by the T-cell epitope (TCE)-algorithm, where the HLA-DPB1 alleles are divided into three distinct immunogenic groups according to their direct cytotoxicity toward T-cells. Current guidelines in Germany conclude that donors with identical or permissive matching should be preferably selected as donors for HSCT. So far, the role of HLA-DPB1 matching in children with malignant and non-malignant indications has not been analyzed. In this study we performed a retrospective study of 251 patients (median age 8.8 years) who received HSCT between 2003 and 2013 at the Charite children9s hospital (main cohort). 86 patients have received a MURD and 65 patients a matched related donor (MRD). The 5 year overall survival (OS) was 0.86 in MRD compared to 0.72 in MURD (p=0.06569). We found an OS of 0.74 when a permissive donor was chosen compared to 0.645 for non-permissive donors. The MRD cohort was further divided according to the type of indication for HSCT. Our data clearly demonstrate that in children undergoing HSCT due to a non-malignant indication, a permissive donor has a significantly better OS (0.84) compared to non-permissive donors (0.50, p=0.02882). In contrast, for patients with a malignant indication for HSCT a non-permissive donor was a better choice (0.76) compared to permissive donors (0.65, p=0.83657). This clear finding was unexpected and is in opposite of the current recommendations for selecting the best suited donors for these indications. In order to corroborate our findings, we adapted our study to 115 patients (meadian age 9.1 years) that received a HSCT at the Children9s University Hospital Wuerzburg between 2005 and 2015 (control cohort). In our main cohort 71% of MURD had bone marrow-derived stem cells and 36% a malignant indication, in the control cohort 92% of donors donated mobilized peripheral stem cells and 71% of patients had a malignant indication. Despite these differences, the TCE-permissive donors had a statistically worse 5 year OS (0.73) compared to the non-permissive donors (1.0, p=0.02026). This finding was even strengthened when 129 MURD of both centers were combined: 75 patients had a better OS with a non-permissive donor (0.88) compared to a permissive donor (0.66; p=0.115; see Figure) only for a malignant indication. For a malignant indication the permissive donor was the better choice (0.86 to 0.71; p=0.19). While the low numbers of patients in the two analyzed centers is due to our focus on an exclusive pediatric patient cohort, the main findings were consistent between the two centers. Therefore, despite the overall low numbers, our data strongly imply that the careful selection of a mismatched, non-permissive donor for pediatric HSCT is a better choice with an underlying malignant indication. In contrast, permissive donors should be chosen for non-malignant indications. We suggest that these findings are to be confirmed in prospective studies and the deviated rules should be considered for selection of suitable donors in children. Disclosures No relevant conflicts of interest to declare.