This study developed and validated a Bayesian joint model integrating longitudinal peripheral blood data, demographic and epidemiologic variables, and multiple biomarkers to predict immunotherapy outcomes in advanced NSCLC using real-world data from 424 patients across three clinical centers. The model, which incorporated blood measurements collected at baseline and during the first treatment cycles, achieved strong predictive performance for progression-free survival at 6, 12, and 24 months in two independent validation cohorts, with AUCs ranging from 0.667 to 0.870. Predicted responder and non-responder groups also showed significantly different progression-free and overall survival outcomes. These results suggest that combining monitored blood-based biomarkers with clinical data in a longitudinal modeling framework improves immunotherapy response prediction compared with approaches based solely on baseline biomarkers.

Ramos-Guerra, A. D., Farina, B., Rubio Pérez, J., Vilalta-Lacarra, A., Zugazagoitia, J., Peces-Barba, G., Seijo, L. M., Paz-Ares, L., Gil-Bazo, I., Dómine Gómez, M., & Ledesma-Carbayo, M. J. (2025). Monitoring peripheral blood data supports the prediction of immunotherapy response in advanced non-small cell lung cancer based on real-world data. Cancer immunology, immunotherapy : CII, 74(4), 120.