Classification of HLA-matching for retrospective analysis of unrelated donor transplantation: revised definitions to predict survival

Abstract

The best unrelated donors (URD) for hematopoietic cell transplantation (HCT) are alleles matched at HLA-A, -B, -C, and DRB1. Earlier studies mostly used incomplete or lower resolution HLA typing for analysis of transplant outcome. To understand the impact of incomplete HLA characterization, we analyzed 14,797 URD HCT (1995-2006) using multivariable regression modeling adjusting for factors affecting survival. Of 21 matching cohorts, we identified 3 groups with significantly different outcomes. Well-matched cases had either no identified HLA mismatch and informative data at 4 loci or allele matching at HLA-A, -B, and -DRB1 (n = 7477, 50% of the population). Partially matched pairs had a defined, single-locus mismatch and/or missing HLA data (n = 4962, 34%). Mismatched cases had > or =2 allele or antigen mismatches (n = 2358, 16%). Multivariate adjusted 5-year survival estimates were: well-matched: 54.1 (95% confidence interval), 52.9-55.4), partially matched: 43.7 (42.3-45.2), and mismatched: 33.4 (32.5-36.5), P < .001. A better matched donor yielded 10%-11% better 5-year survival. Importantly, intermediate resolution -A, -B, and -DRB1 alleles matched "6/6 antigen matched" HCT had survival outcomes within the partially matched cohort. We suggest that these proposed HLA subgroupings be used when complete HLA typing is not available. This improved categorization of HLA matching status allows adjustment for donor-recipient HLA compatibility, and can standardize interpretations of prior URD HCT experience.

Figure 1. All Groups

Estimated 1 year survival for a first transplant, under 18 year old, CMV negative patient with a 25 year old donor. The patient is assumed to be Caucasian with a KPS of 90-100, transplanted in 2003-2006 using myeloablative conditioning and tacrolimus/ methotrexate as GVHD prophylaxis. The 21 groups cluster (per Table 2) as: Well matched (thin/blue lines); partially matched (thick/red lines); and mismatched (medium/green lines).

Figure 2. One Year Survival Estimates for Model Patients in Different HLA Matching Groups

Estimated survival for different patients undergoing a first transplant. Patients are all assumed to be CMV negative, Caucasian with a KPS of 90-100, transplanted in 2003-2006 receiving tacrolimus/methotrexate as GVHD prophylaxis and a 25 year old donor. All transplants are assumed myeloablative except the 55 year old late MDS patient who was assumed to have a reduced intensity transplant. Patients’ age, disease and disease stage are as shown in each panel.

Figure 3. Five Year Survival Estimates for Model Patients in Different HLA Matching Groups

Estimated five year survival for different patients undergoing a first transplant. Patients are assumed to be CMV negative, Caucasian with a KPS of 90-100, transplanted in 2003-2006 receiving tacrolimus/methotrexate as GVHD prophylaxis and a 25 year old donor. All transplants are assumed myeloablative except the 55 year old late MDS patient who was assumed to have a reduced intensity transplant. Patients’ age, disease and disease stage are as shown in each panel.