Establishment of a prognostic risk assessment model for acute myeloid leukemia based on NK-related genes
MEI Dianfeng1, XIANG Dan1, MAO Chenchen1, ZHOU Haixia2, XUE Xiangyang1, ZHU Shanli1.
1.Institute of Molecular Viruses and Immunology, Department of Microbiology and Immunology, Wenzhou Medical University, Wenzhou 325035, China; 2.Pediatric Research Room, the Second Affiliated Hospital & Yuying Children’s Hospital of Wenzhou Medical University, Wenzhou Children’s Hematological Tumor Disease Research Key Laboratory, Wenzhou 325027, China
MEI Dianfeng,XIANG Dan,MAO Chenchen, et al. Establishment of a prognostic risk assessment model for acute myeloid leukemia based on NK-related genes[J]. JOURNAL OF WEZHOU MEDICAL UNIVERSITY, 2024, 54(7): 563-573.
Abstract：Objective: To develop a novel risk scoring model for the prognosis of AML patients by examining the expression patterns of NK cell-related genes (NRGs) in acute myeloid leukemia (AML) and their association with prognosis. Methods: AML transcriptome data from TCGA database were retrieved by bioinformatics techniques. Utilizing the distinct expression patterns of NRGs, patients diagnosed with AML were categorized with unsupervised clustering methods. Subsequently, survival disparities, functional enrichment of genes, and immune-related assessments were conducted on distinct subgroups of AML patients. NRGs was utilized to develop a prognostic assessment model by employing Cox survival analysis and LASSO regression analysis focusing on the uniquely expressed genes within specific subgroups. Ultimately, bone marrow samples were collected from individuals diagnosed with acute myeloid leukemia for transcriptome sequencing in order to validate the association between various genes and AML within the context of prognostic evaluation. Results:Univariate Cox screening of NRGs related to prognosis was done, and based on their expression levels AML patients were divided into 3 subgroups. An analysis of the survival, enrichment, immunoinfiltrating cells, and immune checkpoint in subgroup 2 indicated a more favorable clinical outlook, and the immune microenvironment has been significantly remodeled. Furthermore, based on the distinctively expressed genes of the three subgroups,a risk score model was developed by Cox combined with LASSO analysis, which was significantly correlated with the clinical prognosis of AML. Transcriptome sequencing revealed that of the six AML risk genes in the risk model, the expression levels of ADGRG1, LSP1 and PDE7B were upregulated in relapsed patients, compared with remission patients, while the protective gene TBX1 revealed decreased expression levels in relapsed patients. Conclusion: The risk assessment model of AML patients constructed in this investigation represents a novel and autonomous prognostic evaluation approach, demonstrating favorable predictive capabilities.
