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

UNLABELLED: The results of unrelated donor transplantation (URD-BMT) are difficult to analyze since the continuous advances in HLA typing technology allow the detection 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 the 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 precursor (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 mismatched for HLA class I (-A and -B by serology) and HLA class II, respectively, 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 pairs were HLA-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 three had severe aGVHD). The overall survival of the cohort was 28 +/- 6%. Among 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% in the 15 high risk patients. Only two of the 33 patients developed severe aGVHD, 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 from severe aGVHD and four had graft failure or rejection. Three of the 10 patients with only an HLA-C mismatch died from severe aGVHD, and two had graft rejection.

IN CONCLUSION: (1) donor-recipient matching based on high resolution 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 prevention are comparable to those of geno-identical BMT when donor and recipient 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 HLA class I matching, and our results fail to show that CTLp assay can distinguish between permissible and non-permissible class I mismatches; (4) clinical trials involving donor-recipient pairs with known HLA class I mismatches are needed to improve aGVHD prevention without increasing graft failure rate.