Development of a prognostic model based on anoikis-related genes for predicting clinical prognosis and immunotherapy of hepatocellular carcinoma

Abstract

Hepatocellular Carcinoma (HCC) is the predominant cause of cancer-related mortality worldwide. The majority of HCC patients are diagnosed at advanced stages of the disease, with a high likelihood of metastasis and unfavorable prognosis. Anoikis resistance is a crucial factor contributing to tumor invasion and metastasis, although its specific role in HCC remains unclear. Based on the results of univariate Cox regression and least absolute shrink-age and selection operator (LASSO) analysis, a subset of anoikis-related genes (ARGs) significantly associated with overall survival (OS) was identified. A multivariate Cox regression analysis subsequently identified PDK4, STK11, and TFDP1 as three prognostic ARGs, which were then used to establish a prognostic risk model. Differences in OS caused by risk stratification in HCC patients were demonstrated. The nomogram analysis indicated that the ARGs prognostic signature served as an independent prognostic predictor. In vitro experiments further confirmed the abnormal expression of selected ARGs in HCC. The association between risk scores and OS was further examined through Kaplan-Meier analysis, CIBERSORT analysis, and single-sample gene set enrichment analysis (ssGSEA). This study is a pioneering effort to integrate multiple ARGs and establish a risk-predictive model, providing a unique perspective for the development of personalized and precise therapeutic strategies for HCC.