Acute myeloid leukemia (AML) is a highly heterogeneous malignancy with immunosuppressive tumor microenvironment (TME), which contributes to the development of drug resistance and poor outcomes of immune therapy in AML patients. Therefore, it is imperative to unveil the mechanisms underlying immune suppression in AML TME and identify novel therapeutic targets for immunotherapy. Here, our study identified two immune-related subtypes via RNA-seq analysis, with Cluster 2 significantly associated with multiple biological characteristics and clinical phenotypes, including patients' age, survival and mutational burden. Meanwhile, differential analysis revealed 397 upregulated genes, 75 downregulated genes and 87 prognosis-related differentially expressed genes (DEGs) between Cluster 1 and Cluster 2. Subsequently, LSP1 was identified as the most significant prognostic marker for AML patients. Furthermore, we found that LSP1 expression was significantly correlated with patients' age, class and prognosis and notably affected the cytotoxicity of sorafenib against AML cells. Finally, scRNA-seq analysis revealed a significant correlation between LSP1 and the expression of immune checkpoint molecules in CD8