Abstract
INTRODUCTION: A significant challenge in treating non-small cell lung cancer (NSCLC) is its inherent resistance to radiation therapy, leading to poor patient prognosis. This study aimed to identify key genes influencing radiotherapy resistance in NSCLC through comprehensive bioinformatics analysis. METHODS: A total of 103 common genes were identified, enriched in critical biological pathways such as coagulation, complement activation, growth factor activity, and cytokine signaling. Using advanced machine learning techniques like SVM-RFE, LASSO regression, and random forest algorithms, four pivotal genes-TGFBI, FAS, PTK6, and FA2H-were identified. RESULTS: TGFBI showed the strongest correlation with NSCLC prognosis as indicated by a diagnostic nomogram. Additionally, significant differences in immune cell infiltration, particularly involving naive B cells and M0 macrophages, were noted between high-risk and low-risk patients. DISCUSSION: The study suggests that targeting pathways regulating macrophage polarization or enhancing naive B cell activation could play a crucial role in addressing radiotherapy resistance. The findings highlight the potential therapeutic targets, thereby advancing the understanding of the molecular mechanisms underlying radiotherapy resistance in NSCLC, with implications for improving patient management and outcomes.