Anti-PD-1 first-line monotherapy is indicated for patients with advanced non-small-cell lung cancer (NSCLC), non-targetable driver mutations and >50% PDL1 expression. Still, the majority of patients do not respond or become resistant after initial response. Identifying patients who will benefit from anti-PD-1 monotherapy or need more aggressive combinatorial treatment is an unmet need. Our aim is to develop the first three-dimensional bronchoschopic biopsies-on-chip (BronchoBOCs) to predict real-time responses to PD1-blockade in NSCLC patients. We will conduct a prospective multicentric exploratory clinical trial by studying in parallel clinical benefits of NSCLC patients to a-PD1 monotherapy matched to responses of their BronchoBOCs. We will acquire profiles, by using mass spectrometry proteomics (LC-MS/MS), FACS, cytometric bead arrays and imaging. Machine learning approaches will be allied to BronchoBOC profiles and clinical data to identify the best predictive biomarker patterns.