Prediction of acute myeloid leukemia prognosis based on autophagy features and characterization of its immune microenvironment.

Background: Autophagy promotes the survival of acute myeloid leukemia (AML) cells by removing damaged organelles and proteins and protecting them from stress-induced apoptosis. Although many studies have identified candidate autophagy genes associated with AML prognosis, there are still great challenges in predicting the survival prognosis of AML patients. Therefore, it is necessary to identify more novel autophagy gene markers to improve the prognosis of AML by utilizing information at the molecular level.

Methods: In this study, the Random Forest, SVM and XGBoost algorithms were utilized to identify autophagy genes linked to prognosis, respectively. Subsequently, six autophagy genes (TSC2, CALCOCO2, BAG3, UBQLN4, ULK1 and DAPK1) that were significantly associated with patients' overall survival (OS) were identified using Lasso-Cox regression analysis. A prediction model incorporating these autophagy genes was then developed. In addition, the immunological microenvironment analysis of autophagy genes was performed in this study.

Results: The experimental results showed that the predictive model had good predictive ability. After adjusting for clinicopathologic parameters, this feature proved an independent prognostic predictor and was validated in an external AML sample set. Analysis of differentially expressed genes in patients in the high-risk and low-risk groups showed that these genes were enriched in immune-related pathways such as humoral immune response, T cell differentiation in thymus and lymphocyte differentiation. Then immune infiltration analysis of autophagy genes in patients showed that the cellular abundance of T cells CD4+ memory activated, NK cells activated and T cells CD4+ in the high-risk group was significantly lower than that in the low-risk group.

Conclusion: This study systematically analyzed autophagy-related genes (ARGs) and developed prognostic predictors related to OS for patients with AML, thus more accurately assessing the prognosis of AML patients. This not only helps to improve the prognostic assessment and therapeutic outcome of patients, but may also provide new help for future research and clinical applications.