Abstract
To optimize patient selection for first-line immune-checkpoint inhibitor plus chemotherapy (ICI-Chemo) in advanced non-small cell lung cancer (NSCLC), we developed a 9-gene Immune-Chemotherapy Prediction Score (ICPscore) from the phase 3 ORIENT-11 trial. A high ICPscore identified patients with markedly superior progression-free and overall survival from ICI-Chemo versus chemotherapy alone, whereas low-scorers derived minimal benefit, outperforming existing biomarkers such as PD-L1. Its predictive value for immunotherapy efficacy was further validated in the OAK (NSCLC) and IMvigor210 (metastatic urothelial cancer) cohorts. Multi-cohort and single-cell analyses linked a high ICPscore to an immune-active tumor microenvironment characterized by myeloid cell activation, thus providing a biological rationale for its enhanced performance and positioning it as a robust tool for treatment personalization.