Arsela Prelaj, Vanja Miskovic, Matteo Sacco, Alberto Ferrarin, Cristina Maria Licciardello, Leonardo Provenzano, Margherita Favali, Ludovica Lerma, Aleksandra Zec, Andrea Spagnoletti, Monica Ganzinelli, Daniele Lorenzini, Beshoy Guirges, Luca Invernizzi, Cecilia Silvestri, Laura Mazzeo, Marco Meazza Prina, Giulia Corrao, Margherita Ruggirello, Andra Diana Dumitrascu, Rosa Maria Di Mauro, Dario Monzani, Gabriella Pravettoni, Michele Zanitti, Davide Macocchi, Moreno Bruno Marino, Chiara Cavalli, Rebecca Romanò, Claudia Giani, Samuel G Armato, Alessandra Esposito, Christine M Bestvina, Maria Spector, Bogot R Naama, Reham Basheer, Adi Lahiani Hafzadi, Laila Roisman, Iris Watermann, Marlen Szewczyk, Till Olchers, Heinz Richter, Constantin Blanke-Roeser, Costanza Siniscalchi, Anna Di Lello, Teresa Arangoa, Valentina Bartolomeo, Nikolaos Spathas, Evangelos Sarris, Elena Fountzilas, Aina Arbusà Roca, Rocio Caro-Consuegra, Patricia Iranzo, Melissa Fernández-Pinto, Jose Rodríguez-Morató, Luca Agnelli, Mario Occhipinti, Marta Brambilla, Teresa Beninato, Claudia Proto, Sokol Kosta, Michele Pio Di Palma, Eliana Rulli, Stefan Steurer, Ronald Simon, Michael Willis, Giancarlo Pruneri, Filippo De Braud, Marcello Restelli, Enriqueta Felip, Nir Peled, Alexander T Pearson, Helena Linardou, Martin Reck, Giuseppe Lo Russo, Francesco Trovò, Alessandra Laura Giulia Pedrocchi, Marina Chiara Garassino, I3LUNG Consortium
Despite a decade in, immunotherapy (IO) treatment selection in non-small cell lung cancer (NSCLC) remains largely guided by subgroup analyses and imperfect programmed death ligand 1 (PD-L1) and clinical scores. To our knowledge, I3LUNG ( NCT05537922 ) is currently the largest international, real-world, multimodal, artificial intelligence (AI)-based study, enrolling 2,396 patients. We integrated real-world clinical and blood (CB) data, computed tomography (CT) images, digital pathology (DP), and genomics into machine learning early fusion (MLEF) and deep learning intermediate fusion (DLIF) models. Machine learning (ML) and deep learning (DL) CB-only models achieved consistent performance across outcomes with area under the curve (AUC) up to 0.77 in the test (TEST) set. Performance drop in external validation (EXVAL) likely reflects population differences (AUC range: 0.55-0.72). AI models significantly surpassed PD-L1, Eastern Cooperative Oncology Group performance status (ECOG PS), neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase (LDH) and Lung Immune Prognostic Index (LIPI) score in the independent TEST set. The clinical usability study showed that lung expert and nonexpert physicians improved their prediction with the explainable AI (XAI) ML CB-only based tool. Although multimodal integration with MLEF (CB+CT+DP) was associated with higher performance, its incremental benefit remains uncertain, not translated in TEST and EXVAL. The I3LUNG project is a pioneering framework showing the clinical usefulness of AI tools. A prospective validation of the decision support system (both CB and multimodal) is currently undergoing in more than 2,000 patients.