AUTHOR=Zhang Yu , Ye Chenjing , Zhu Haojie , Zhuang Youran , Chen Shaozhen , Weng Yingxi , Ren Jinhua , Luo Xiaofeng , Zheng Jing , Zheng Xiaoyun , Li Jing , Lan Lingqiong , Xie Yongxin , Han Zhongchao , Hu Jianda , Yang Ting TITLE=Association of iKIR-mismatch model and donor aKIRs with better outcome in haploidentical hematopoietic stem cell transplantation for acute myeloid leukemia JOURNAL=Frontiers in Immunology VOLUME=13 YEAR=2023 URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2022.1091188 DOI=10.3389/fimmu.2022.1091188 ISSN=1664-3224 ABSTRACT=Objectives

Killer cell immunoglobulin like receptor (KIR) can trigger the alloreactivity of NK cells. However, there is no clear consensus as to their function. Here, we investigated the potential influence of KIR mismatch and KIR alleles on the outcome of haploidentical hematopoietic stem cell transplantation (haplo-HSCT) in acute myeloid leukemia (AML) patients.

Method

Data from 79 AML patients treated with haplo-HSCT were retrospectively analyzed. HLA-C genotyping was determined by the PCR-rSSO method. KIR, HLA-A and HLA-B genotyping was performed by the PCR-SSP method. Cox proportional hazards model and Kaplan-Meier survival curves were used for analysis.

Results

Both KIR ligand mismatch (KLM) group and KIR receptor-ligand mismatch (RLM) group were associated with a decreased risk in aGVHD and relapse rate (RR), and better overall survival (OS) compared to the KIR ligand matching and receptor-ligand matching groups, respectively (aGVHD: KLM: p=0.047, HR:0.235; RLM: p<0.001, HR:0.129; RR: KLM: p=0.049, HR:0.686, RLM: p=0.017, HR:0.200;OS:KLM: p=0.012, HR: 0.298, RLM: p=0.021, HR:0.301). RLM was more accurate at predicting relapse and aGVHD compared with KLM (aGVHD: p=0.009; RR: p=0.039). Patients with greater number of donor activating KIRs (aKIR) had a lower incidence of aGVHD and relapse, and the benefits correlated with the increase in the number of donor aKIRs (aGVHD: p=0.019, HR:0.156; RR: p=0.037, HR:0.211). Patients with RLM and the highest number of donor aKIRs had the lowest RR, lowest incidence of aGVHD and best OS.

Conclusions

Both KLM and RLM reduced the risk of aGVHD and relapse after haplo-HSCT in AML patients, and RLM showed superiority in predicting HSCT outcome. The synergistic effects of RLM and donor aKIRs can provide a better donor selection strategy to improve haplo-HSCT outcome in AML patients.