Identification and validation of a prognostic risk-scoring model based on the level of TIM-3 expression in acute myeloid leukemia

Abstract

Abstract Acute myeloid leukemia (AML) is characterized by an unfavorable prognosis due to the presence of self-renewing leukemic stem cells (LSCs). The presence of T-cell immunoglobulin mucin-3 (TIM-3) on the surface of LSCs has been observed in various types of human AML, exerting an impact on the prognostic outcome. Exploring the hub genes associated with varying levels of TIM-3 expression offers a valuable approach to enhance our understanding of the underlying mechanisms involving TIM-3 and to identify potential prognostic indicators in AML. Nevertheless, to date, no research studies have reported a prognostic model that relies on the level of TIM-3 expression. In our study, we screen the hub-genes based on different expression level of TIM-3 through WGCNA. The prognostic risk-scoring model was constructed based on hub-genes. The results show the risk prognostic model has extraordinary ability to predict prognosis in both the training and validation sets. The high-risk group present poor prognosis with mutation of NPM1, TP53 (Multiple Hit) and FLT3(multiple hit), while IDH2 (Missense Mutation), MUC16 (Multiple Hit/Missense Mutation) occur mutation in low-risk group presenting favorite prognosis than high-risk group. Leukocyte cell–cell adhesion, regulation of T cell activation and I-κB kinase/NF-κB signaling enriched in high-risk group, involving in HSCs or LSCs anchoring to BM, which implicated in LSCs survival and chemotherapy resistance. B7-H3 (CD276) and CD276 would be the potential immune targets in high-risk group. The risk score model may help in distinguishing immune and molecular characteristics, predicting patient outcomes.