HLA-B Leader Mismatch Positively Correlates with Survival Rate in Haplo and Cord Blood Transplantation for Acute Leukemia in Japanese Patients

Introduction： HLA class I genes contain seven exons. Exons 2, 3, and 4 generate α1, α2, and α3 extracellular domains, respectively, and exons 5, 6, and 7 encode the transmembrane region and cytoplasmic tail, respectively. Exon 1 is not involved in those synthesis, but encodes the leader peptide. Position 21 of exon 1 is invariant in HLA-A and HLA-C but dimorphic in HLA-B. Depending on the cytosine/thymine difference at this site, different HLA-B leader peptides containing either methionine [M] or threonine [T] at the second position when translated. HLA-B leader dimorphism is associated with NK cell recognition of HLA-E/NKG2A inhibitory responses, and [M] alleles have higher HLA-E expression than [T] alleles. In HLA-B-1 mismatch-unrelated allogeneic hematopoietic stem cell transplant (allo-HSCT), the HLA-B leader recipient/donor mismatch group has been reported to have a higher risk of acute GVHD than matched groups. In addition, acute GVHD is more common with HLA-DR 1 mismatched allo-HSCT in patients with [M] alleles. However, the effect of HLA-B leader mismatch on haploidentical peripheral blood stem cell transplantation (haplo-PBSCT) and cord blood transplantation (CBT) is controversial. In addition, the effects of HLA-leader mismatch may vary among races. We evaluated the impact of HLA-B leader mismatch on CBT and haplo-PBSCT for acute leukemia (AL) in Japanese patients.

Methods： We retrospectively analyzed 66 patients with AL (44 with AML, 20 with ALL (18 with B-ALL, 2 with T-ALL), and 2 with MPAL) who underwent first allo-HSCT (haplo-PBSCT or CBT) in our hospital from2006 to 2021.

The median age of the patients was 38 years (range:17-64). The HCT-CI scores were 0 (low) in 29 patients, 1-2 (Int) in 18, and >2 (high) in 19. Forty-six patients were transplanted in CR and 20 were transplanted in non-CR. Thirty patients received haplo-PBSCT (18 with ATG and 12 with PTCY), and 36 received CBT. GVHD prophylaxis was FK506+MTX (15-10-10 mg/m²) in 54 patients and MMF+FK506+CY in 12. The median observation period was 405 [36-5842] days. The HLA-B leader ([M] or [T]) was determined for each allele in the recipient and donor and divided into three groups (TT, MT, and MM) using the HLA-B Leader Assessment Tool (BLEAT).

Overall survival (OS), non-relapse mortality (NRM), and event-free survival (EFS) rates were assessed. A relapse or death was defined as an event.

Results： The HLA-B leader phenotypes of the recipients included 47 (71%) with TT, 17 (26%) with MT, and 2 (3%) with MM; 68% were HLA-B leader-matched and 32% were mismatched (TT to MT 7, MT to TT 9, MM to MT 2, MT to MM, 3). Two-year NRM was lower in the mismatched group than in the matched group (0% vs. 22.7%, p=0.014), and the two-year OS was significantly better in the mismatched group (80.7% vs. 47.2%, p=0.01). The two-year cumulative incidence of relapse (CIR) was comparable (matched: 38.0% versus mismatched: 38.5%, p=0.78) between the two groups. The frequency of grade II-IV acute GVHD was comparable between the two groups in all patients, but in CBT, was higher in the mismatched group than in the matched group (54.5% vs. 16.0%, respectively, p=0.03). In the multivariate analysis adjusted for CR status (CR or non-CR), HCT-CI (>2), CBT, and HLA-B, leader mismatch was an independent favorable factor for OS (HR:0.24, 95%CI:0.09-0.66, p<0.01).

Conclusion： HLA-B leader mismatch positively correlated with the survival rate. Evaluation of the HLA-B leader might be a useful marker for donor selection in first haplo-PBSCT and CBT for AL.

No relevant conflicts of interest to declare.