Abstract
BACKGROUND: The survival rate of hepatocellular carcinoma (HCC) is low and the prognosis is poor. The basement membrane (BM) is increasingly recognized as a significant player in HCC, although its potential impact on diagnosis, prognosis, and treatment remains unclear. The study aims to investigate the potential of BM-related genes as novel prognostic biomarkers in HCC. METHODS: In this study, we utilized BM-associated genes to classify cohorts of HCC from publicly available database The Cancer Genome Atlas (TCGA) into BM-associated clusters. Through machine learning regression processes, we established a 31-gene molecular classifier known as the BM-based prognostic signature (BMS), based on differentially expressed genes. RESULTS: We then assessed the prognostic efficacy of BMS and observed its superiority as a prognostic indicator for HCC, either alone or when combined with other clinical factors. Furthermore, we identified distinct biological behaviors and immune features within different risk subgroups identified by BMS. Specifically, high-risk patients exhibited suppressed immune infiltration and worse overall prognosis. Lastly, we conducted a search for potential therapeutic drugs such as fluvastatin and navitoclax tailored to the high-risk subgroup identified by BMS, aiming to enhance personalized treatment strategies for HCC in the future. CONCLUSIONS: Our findings highlight the potential of BM-associated molecular classifiers like BMS in improving the diagnosis, prognosis, and treatment of HCC.