Use of Survival Analysis and Machine Learning in Stage III Non-Small Cell Lung Cancer.

The present study analyses a clinical cohort of patients with stage III non-small cell lung cancer (NSCLC), with a particular emphasis on the evaluation of the combined treatment of chemoradiotherapy and Durvalumab maintenance. The clinical efficacy of the treatment was assessed by applying survival analysis techniques (Kaplan-Meier and Cox proportional hazards models) and comparing the results with those of the PACIFIC trial. Furthermore, clinical subgroups were explored using machine learning techniques. The findings demonstrate a substantial benefit of immunotherapy on both overall survival (OS) and progression-free survival (PFS). The median OS for patients receiving Durvalumab reached 37.3 months versus 22.97 months for those without, with a 5-year survival rate of 47.9%. Median PFS improved from 12.83 to 22.10 months, with 5-year PFS rates of 47.47% vs. 29.06%. Furthermore, Clustering analysis revealed distinct clinical profiles, with the subgroup matching the PACIFIC trial scheme achieving the best outcomes.