Abstract
BACKGROUND: Acute myeloid leukemia (AML) is a common hematologic malignancy with a generally unfavorable prognosis. Cuprotosis as a new form of programmed cell death has been shown to play an important role in tumorigenesis and progression; However, the relationship between cuprotosis and the prognosis of AML patients remains unclear. METHODS: Transcriptomic and genomics data, along with clinical information, were obtained from the TCGA and GEO databases. Especially, unsupervised clustering and machining learning were used to identify molecular subtypes and cuprotosis-related risk scores respectively. Kaplan-Meier analysis, univariate and multivariate Cox regression, and Receiver Operator Characteristic curve (ROC) were performed to assess the prognosis based on cuprotosis-related genes (CRGs). Moreover, multiple algorithms were used to evaluate immunological heterogeneity among patients with different risk scores. For in vitro analysis, the expression of genes involved in CRGs was detected by Quantitative Reverse Transcription Polymerase (qRT-PCR) in AML patients. RESULTS: Transcriptomic and genome data indicated the immense heterogeneity in the CRGs landscape of normal and tumor samples. Cuprotosis subtype A and cuprotosis regulatory subtype B in the genomics map and biological characteristics were significantly different from the other groups. Furthermore, these two subtypes had lower risk scores and longer survival times compared to other groups. Cox analyses indicated that risk score was an independent prognostic factor for AML patients. In addition, our risk score could be an indicator of survival outcomes in immunotherapy datasets. CONCLUSIONS: Our study demonstrates the potential of CRGs in guiding the prognosis, treatment, and immunological characteristics of AML patients.