Allele Matching At HLA-C or DRB1 Is Associated with Improved Survival After Reduced Intensity Double Umbilical Cord Blood Transplantation

Abstract 2010

Mismatch at HLA-C has been associated with increased transplant-related mortality in unrelated bone marrow and peripheral blood stem cell transplantation as well as single myeloablative umbilical cord blood transplantation (UCBT) in children, but the impact of HLA-C mismatch in adult double reduced-intensity conditioning (RIC) UCBT is unknown. Matching at HLA-C is not routinely considered in cord unit selection. We studied the effect of HLA-C matching in 125 patients who underwent double UCBT for hematologic malignancy at Dana-Farber Cancer Institute and Massachusetts General Hospital from 2003 through 2010. The median age was 49 years (range 16–69), and the diagnoses included acute leukemia (45%), MDS (12%), lymphoma (27%), myeloproliferative neoplasms (2%), and others (13%). Data on HLA-C match were recorded but not used in the UCB unit selection strategy. UCB unit selection criteria were a 4/6 allele level A, B, DR match with the patient and other UCB unit. 82% of patients received a fludarabine/melphalan/antithymocyte globulin RIC, and 62% received sirolimus-based graft vs host disease (GVHD) prophylaxis. The median follow-up time among survivors was 32 months (range 6–73). The degree of allele matching at HLA-C (donor to both cords) was 0/4 in 14 patients (11%), 1/4 in 21 patients (17%), 2/4 in 62 patients (50%), 3/4 in 20 patients (16%), and 4/4 in 8 patients(6.4%). Patients who received 2 UCB units both with > 6/8 match at HLA-A,-B,-C, and -DRB1 had improved survival (3 year overall survival (OS) 56% vs 29%, p= 0.01). There was a significant correlation between degree of matching at HLA-C and the frequency of neutrophil engraftment (ANC > 500 by day 42) (0/4=79%, 1/4=76%, 2/4=71%, 3/4=80%, 4/4=100%; p=0.004). Similarly, matching at HLA-C was significantly correlated with platelet engraftment (plt>20,000 by day 100) (0/4=50%, 1/4=52%, 2/4=57%, 3/4=70%, 4/4=100%; p=0.0004). Matching at HLA-A,-B, or –DRB1 did not correlate with engraftment. There was no effect of matching at HLA-C on relapse, acute GVHD, or chronic GVHD. A full match at HLA-C (4 alleles) was associated with improved survival (3-year OS 67% vs 33% with less than full match, p=0.05) but there only 8 patients who received 2 HLA-C matched UCB units. Degree of match individually at HLA-A,-B, or DRB1 was not alone associated with survival. When various combinations of match were examined, full matching at either HLA-C or full matching at HLA-DRB1 (with less than full matching at HLA-C), compared to full matching at neither locus, was associated with improved 3-year OS (67% vs. 42% vs. 24%, p=0.03). HLA-C match did not predict the dominant UCB unit. In summary, (1) patients who received closer HLA allele-level matched UCB units had improved survival after RIC double UCBT; (2) matching at HLA-C in RIC double UCBT may be associated with earlier neutrophil and platelet engraftment; (3) survival may be improved when patients received UCB units fully matched at HLA-C or fully matched at HLA-DRB1 (if less than fully matched at HLA-C) compared to recipients of units fully matched at neither locus; and (4) matching at HLA-A,-B, or DRB1 alone did not correlate with differences in engraftment or survival. These data are limited by the small number of patients that were compatible at HLA-C but warrant examination in a larger cohort to determine the role of matching at HLA-C in UCB unit selection algorithms.