Impact of Allele-Level HLA-Matching on Outcomes after Double Unrelated Cord Blood Transplantation in Adults with Malignant Diseases

Abstract

Introduction: In single unrelated cord blood transplantation (sCBT), an increasing number of HLA-mismatches (MM) has been associated with inferior overall survival (OS) attributed to higher transplant-related mortality (TRM). These findings have prompted current guidelines to endorse high resolution HLA typing in the selection of CB (cord blood) units, including double CBT (dCBT). However, a comprehensive appraisal of the effect of allele-level HLA matching following dCBT was likely hindered by the relatively small sample sizes in previous studies. Herein we aimed to analyze the impact of allele-level HLA-matching on outcomes of a large cohort of dCBT recipients to improve CB unit selection and further explore the unique dCBT biology. Methods: We included patients ≥18 years-old receiving a dCBT between 2006-2019 with the availability of HLA typing minimally at HLA-A, -B, -C at the antigen-level and HLA-DRB1 at the allele-level for the recipient and both CB units. A validated high-resolution imputation tool, HaploStats, was used to adjudicate allele-level match status for a subset of cord and recipient pairs (n = 173) having only antigen-level typing at HLA-A, -B or -C. A sensitivity analysis for OS was performed excluding imputed cases, and results were unchanged. HLA matching was classified according to the CB unit with the greatest degree of HLA mismatch with the patient, meaning that only the most mismatched unit was considered and that the interunit matching was not considered. Continuous covariates were dichotomized to find an optimal cutoff value by using the Contal-O'Quigley method, except for cryopreserved CD34 and total nucleated cell (TNC) count, whose cutoffs were based on previous findings reported by our group. Pairwise comparisons of the numbers of allele MM showed that receiving units with ≥4 MM had the highest negative impact on OS. Cox and Fine-Gray regression models were built for the different outcomes using the backward elimination method (p<.05 to stay). Results: 963 patients met the eligibility criteria, of which 392 received at least one CB unit with 0-3 allele MM while 571 with ≥4 allele MM. Of those receiving at least one unit with ≥4 allele MM, 18% had a maximum HLA disparity of 5/6 at lower HLA-typing resolution, 69% of 4/6 and 13% of ≤3/6. Median age at dCBT was 49 years old. Acute leukemia and myelodysplastic syndrome (MDS) accounted for 70% of the cases. There was a larger proportion of patients receiving CB units with fewer allele MM after 2012. Other patient-, disease- and CB unit-related characteristics are summarized in Table 1. The day-42 cumulative incidence of neutrophil recovery was 83% (95%CI 81-85). The number of allele MM had no significant impact on neutrophil recovery (≥ 4 MM vs 0-3 MM, multivariate HR 0.94 [0.80-1.11], P=.46). The cumulative incidence of TRM at 4 years was 36% (95% CI 32-40) for patients receiving CB units with ≥4 MM vs. 23% (95% CI 19-28) for those with 0-3 MM (multivariate HR 1.58 [95%CI 1.18-2.11], P=.002] - Fig. 1A). Infectious complications were the most frequent cause of TRM in the group receiving dCBT with ≥ 4 allele MM (40%). Relapse, grade II-IV/III-IV acute and chronic GVHD were not significantly different between the two groups. Patients receiving CB units with 0-3 allele MM had a 4-year survival of 54% (95%CI 49-59) compared to 43% (95%CI 39-47) for those receiving units with ≥4 allele MM (multivariate HR 1.40 [1.11-1.78], P=.005 - Fig. 1B). The inferior OS associated with higher allele HLA disparity was only partially mitigated by increasing the TNC count in multivariate analysis (0-3 MM + TNC ≤ 3.9x10ˆ7/kg vs.0-3 MM + TNC > 3.9x10ˆ7/kg, HR 1.50 [0.97-2.32], P=.07; ≥4 MM + TNC > 4.9x10ˆ7/kg vs. 0-3 MM + TNC > 3.9x10ˆ7/kg, HR 1.42 [1.06-1.89], P=.02; ≥4 MM + TNC ≤ 4.9x10ˆ7/kg vs. 0-3 MM + TNC > 3.9x10ˆ7/kg, HR 1.66 [1.24-2.22], P=<.001 - Fig. 1C). Discussion: Our results confirm that allele-level HLA typing is a significant factor for OS following dCBT. The inferior OS associated with the increasing number of allele-level MM was driven by higher TRM. Selection of CB units with 0-3 allele-level MM is recommended. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal