Abstract

Current criteria for the selection of adult volunteer unrelated donors (URD) include matching for HLA alleles determined by high resolution typing methods. However, the impact of URD matching for 8/8 HLA alleles (HLA-A, B, C, DRB1) on survival compared to HLA identical sibling donor bone marrow transplant (BMT) remains undefined. We tested this hypothesis in a homogenous population of patients with good-risk chronic myeloid leukemia (CML) in first chronic phase (CP1) where a strong allogeneic effect and hence a lower risk of relapse is anticipated. In total, 939 URD bone marrow transplant recipients with CML in CP1 were retrospectively typed for HLA -A, B, C, DRB1, DQ and DP alleles. Outcomes were compared to 3514 HLA identical sibling donor bone marrow transplant recipients (CML, CP1). Five-year survival after 8/8 matched URD HCT was slightly, but significantly worse than that after sibling BMT (54% versus 63%). [RR 1.36 (95% CI 1.17–1.59), p=0.0001]. In multivariate analysis (model adjusted for time from diagnosis to transplant, CMV serostatus of donor and recipient, year of transplant, and T-depletion, and stratified on donor and recipient gender mismatch and radiation use), a single allele mismatch at any class I locus or at HLA-DRB1 led to significantly poorer survival. Risks for single class I versus single DRB1 mismatches were similar. Two locus mismatching was associated with higher risk compared to matched or single locus mismatching. In this homogenous cohort of good risk patients with CML in CP1, 5 year survival after transplant from 8/8 allele-matched donors begins to approach the results of sibling donor transplantation. Additive effects of multi-locus mismatching are not well tolerated and should be avoided if possible.

NRelative risk (95% CI)P-value5 year survival
* Antigen or allele mismatch, **allele mismatch 
HLA identical siblings 3514 1.00  63% (61–64%) 
8/8 matched 531 1.36 (1.17–1.59) 0.0001 54% (50–58%) 
7/8 A mismatch* 68 1.90 (1.36–2.65) 0.0002 43% (32–55%) 
7/8 B mismatch* 34 2.77 (1.81–4.23) <0.0001 30% (15–48%) 
7/8 C mismatch* 113 1.88 (1.45–2.43) <0.0001 42% (33–51%) 
7/8 DRB1 mismatch* 37 2.07 (1.37–3.12) 0.0005 37% (23–54%) 
6/8 Class I mismatches* 128 2.66 (2.11–3.35) <0.0001 33% (25–42%) 
6/8 Class I mismatch* and DRB1 mismatch** 28 3.40 (2.23–5.19) <0.0001 21% (8–37%) 
7/8 DRB1 mismatch** vs. 7/8 Class I mismatch* 37/215 1.03 (0.67–1.61) 0.88N/A  
NRelative risk (95% CI)P-value5 year survival
* Antigen or allele mismatch, **allele mismatch 
HLA identical siblings 3514 1.00  63% (61–64%) 
8/8 matched 531 1.36 (1.17–1.59) 0.0001 54% (50–58%) 
7/8 A mismatch* 68 1.90 (1.36–2.65) 0.0002 43% (32–55%) 
7/8 B mismatch* 34 2.77 (1.81–4.23) <0.0001 30% (15–48%) 
7/8 C mismatch* 113 1.88 (1.45–2.43) <0.0001 42% (33–51%) 
7/8 DRB1 mismatch* 37 2.07 (1.37–3.12) 0.0005 37% (23–54%) 
6/8 Class I mismatches* 128 2.66 (2.11–3.35) <0.0001 33% (25–42%) 
6/8 Class I mismatch* and DRB1 mismatch** 28 3.40 (2.23–5.19) <0.0001 21% (8–37%) 
7/8 DRB1 mismatch** vs. 7/8 Class I mismatch* 37/215 1.03 (0.67–1.61) 0.88N/A  

Author notes

Disclosure: No relevant conflicts of interest to declare.