Introduction: Despite several therapeutic efforts, lung cancer remains a highly lethal disease. Novel therapeutic approaches encompass immune-checkpoint inhibitors, targeted therapeutics and antibody-drug conjugates, with different results. Several studies have been aimed at identifying biomarkers able to predict benefit from these therapies and create a prediction model of response, despite this there is a lack of information to help clinicians in the choice of therapy for lung cancer patients with advanced disease. This is primarily due to the complexity of lung cancer biology, where a single or few biomarkers are not sufficient to provide enough predictive capability to explain biologic differences; other reasons include the paucity of data collected by single studies performed in heterogeneous unmatched cohorts and the methodology of analysis. In fact, classical statistical methods are unable to analyze and integrate the magnitude of information from multiple biological and clinical sources (eg, genomics, transcriptomics, and radiomics).
Methods and objectives: APOLLO11 is an Italian multicentre, observational study involving patients with a diagnosis of advanced lung cancer (NSCLC and SCLC) treated with innovative therapies. Retrospective and prospective collection of multiomic data, such as tissue- (eg, for genomic, transcriptomic analysis) and blood-based biologic material (eg, ctDNA, PBMC), in addition to clinical and radiological data (eg, for radiomic analysis) will be collected. The overall aim of the project is to build a consortium integrating different datasets and a virtual biobank from participating Italian lung cancer centers. To face with the large amount of data provided, AI and ML techniques will be applied will be applied to manage this large dataset in an effort to build an R-Model, integrating retrospective and prospective population-based data. The ultimate goal is to create a tool able to help physicians and patients to make treatment decisions.
Conclusion: APOLLO11 aims to propose a breakthrough approach in lung cancer research, replacing the old, monocentric viewpoint towards a multicomprehensive, multiomic, multicenter model. Multicenter cancer datasets incorporating common virtual biobank and new methodologic approaches including artificial intelligence, machine learning up to deep learning is the road to the future in oncology launched by this project.
Keywords: Artificial intelligence; Multiomics; Non–small cell lung cancer; Personalized medicine; Virtual biobank.
Copyright © 2023 Elsevier Inc. All rights reserved.