APOLLO 11, Consortium of Italian Centers Involved in Treatment of Patients With Lung Cancer Treated With Innovative Therapies: Real World Data and Translational Reaserch

With the advent of several innovative therapies ("target" therapies, immunotherapy (IO) and/or next generation therapies), the identification of prognostic/predictive and/or resistance biomarkers represents the main goals of translational research in advanced lung cancer (aLC). In particular, with the arrival of IO the treatment landscape of advanced Non-Small Cell Lung Cancer (NSCLC) patients (pts), have radically changed prolonging significantly the overall survival (OS). The role of Programmed Death Ligand 1 (PD-L1) remains undefined and to date no other biomarker are used to select patients for IO. Artificial Intelligence (AI) frameworks, which synthesize and correlate information from different data sources, is a potentially highly efficient instrument to construct algorithms reinforcing the individualized prediction. APOLLO 11 main aim is to build a strong Italian long-lasting lung cancer network (in around 48 Italian centres) on real world data and translational research by creating a decentralized long-term national database and a "virtual" multilevel biobank. This will allow to avoid worthless dispersion of single institution data, which often remaine unused and inconclusive due to limited number of patients and low statistical power. Besides, APOLLO 11 will take advantage of the translational research joint effort with the credo "unity is strength". The multilevel structure is designed also for smallest centres of give them possibility to contribute to biobank looking at their internal facilities. APOLLO 11 have the goal to establish the consortium by fully activating at least 20 of the 48 involved centres able to collect data within the database and to guarantee the activation ready for at least 10 biobanks able to collect samples (tumour tissue, blood, stool and urine). Moreover, the key scientific aim of the project is to provide answers about aNSCLC pts treated with IO. In particular, try to find a "software" biomarker able to better predict IO selection compare to standard of care single biomarker (e.g., PD-L1, TMB) using artificial intelligence/machine/deep learning (AI/DL/ML) methodologies. Real world (demographic, histology, PD-L1, molecular, treatment and outcomes information etc), radiomics and where available (from previous researches) multiomic data from all the activated centres will be collected. Furthermore, single-cell RNA sequencing (scRNA seq) analysis will be carry out. Finally, AI/ML/DL will be used to integrate the multisource data, analyse them and develop predictive algorithms of IO efficacy in NSCLC pts. Explainable trustworthy AI (XAI) methodology will be used in order to better understand black box for explaining clinical and biological insight.