High resolution HLA class I and II typing and CTLp frequency in unrelated donor transplantation: a single-institution retrospective study of 69 BMTs

The results of unrelated donor transplantation (URD-BMT) are difficult to a nalyze since the continuous advances in HLA typing technology allow the det ection of new mismatches unknown at the time of transplantation. We sought to confirm that matched recipient-donor pairs are in fact often mismatched when advanced HLA typing techniques are used, We retrospectively studied th e impact of the results of high resolution HLA typing for HLA class I (-A, -B, -C) and HLA class II (-DR, -DQ, -DP) loci, and cytotoxic T lymphocyte p recursor (CTLp) frequency, on the outcome of 69 URD-BMT procedures, At the time of transplant, six (6/69) and two (2/69) donor-recipient pairs were mi smatched for HLA classI (-A and -B by serology) and HLA class II, respectiv ely, while one pair was mismatched for both HLA class I and II. Using high resolution DNA typing, HLA class I mismatches were found in 31 (45%) pairs and HLA class II mismatches in nine (13%) pairs, Twenty-three of the 69 pai rs were BLA-C mismatched. Low CTLp frequencies were found among the 19 HLA class I matched pairs tested, and also in 5/14 mismatched pairs (of whom th ree had severe aGVHD), The overall survival of the cohort was 28 +/- 6%. Am ong the 33 patients who were fully matched with their donors, the survival rate was 66% in the 18 patients with a standard hematological risk and 9% i n the 15 high risk patients. Only two of the 33 patients developed severe a GVHD, and only one had graft rejection. Among the 36 mismatched pairs, the survival rate was 31% in the 13 patients with a standard hematological risk and 8% in the 23 high risk patients. Sixteen of these 36 patients died fro m severe aGVHD and four had graft failure or rejection, Three of the 10 pat ients with only an HLA-C mismatch died from severe aGVHD, and two had graft rejection. In conclusion: (1) donor-recipient matching based on high resol ution HLA class I and II DNA typing is associated with significantly better outcome after URD-BMT; (2) the results of URD-BMT with classical GVHD prev ention are comparable to those of gene-identical BMT when donor and recipie nt are fully matched for HLA-A, -B, -C, -DRB1 and DQB1 on the basis of high resolution typing; (3) CTLp frequencies do not correlate constantly with H LA class I matching, and our results fail to show that CTLp assay can disti nguish between permissable and non-permissable class I mismatches; (4) clin ical trials involving donor-recipient pairs with known HLA class I mismatch es are needed to improve aGVHD prevention without increasing graft failure rate.