This study aims to integrate multi-dimensional biomarkers (including baseline data indicators, genetic features, hematological data, and imaging data) with artificial intelligence (AI) technology to construct a predictive model for efficacy differences between immunotherapy + chemotherapy (triple therapy) and immunotherapy + chemotherapy ± anti-angiogenic agents (quadruple therapy) in non-small cell lung cancer (NSCLC) patients. Specific objectives include:
Refining Current Understanding: Existing knowledge generally suggests that “immunotherapy + chemotherapy” is slightly less effective than “immunotherapy + chemotherapy + anti-angiogenic agents.” This study systematically investigates the subpopulations in which “immunotherapy + chemotherapy” outperforms the quadruple therapy.
In-Depth Analysis of Special Subgroups: A detailed exploration of factors influencing efficacy, such as the expression positivity/rate of immune inhibitory molecules (post EGFR/ALK resistance), vascular endothelial growth factor (VEGF) expression, clonal evolution, DNA repair deficiency, chemotherapy sensitivity, and immunosuppressive microenvironment, to address gaps in current precision treatment strategies.
Optimizing Prediction Model Accuracy: Utilizing AI-based multi-modal fusion technology to overcome limitations of traditional statistical methods and develop clinically accessible stratification tools for individualized treatment decision-making.