Exploring the potential of multiomics liquid biopsy testing in the clinical setting of lung cancer

Abstract

The transformative role of artificial intelligence (AI) and multiomics could enhance the diagnostic and prognostic capabilities of liquid biopsy (LB) for lung cancer (LC). Despite advances, the transition from tissue biopsies to more sophisticated, non-invasive methods like LB has been impeded by challenges such as the heterogeneity of biomarkers and the low concentration of tumour-related analytes. The advent of multiomics - enabled by deep learning algorithms - offers a solution by allowing the simultaneous analysis of various analytes across multiple biological fluids, presenting a paradigm shift in cancer diagnostics. Through multi-marker, multi-analyte and multi-source approaches, this review showcases how AI and multiomics are identifying clinically valuable biomarker combinations that correlate with patients' health statuses. However, the path towards clinical implementation is fraught with challenges, including study reproducibility and lack of methodological standardization, thus necessitating urgent solutions to solve these common issues.A flow diagram to visualize how multiomics approaches can be split into multi-marker, multi-analyte and multi-source approach; then, their link to AI, to decrypt and use in the clinical setting the messages hidden within them. The combined use of Artificial Intelligence (AI) and multiomics could improve the diagnosis and prognosis of Lung Cancer (LC) via Liquid Biopsy (LB); through multi-marker, multi-analyte, and multi-source analysis, the way is paved for the achievement of these goals, once tested through appropriate large-scale multi-center studies.image