Cancer genetics has led to major discoveries, including proto-oncogene and tumor-suppressor concepts, and cancer genomics generated concepts like driver and passenger genes, revealed tumor heterogeneity and clonal evolution. Reconstructing trajectories of tumorigenesis using spatial and single-cell genomics is possible. Patient stratification and prognostic parameters have been improved. Yet, despite these advances, successful translation into targeted therapies has been scarce and mostly limited to kinase inhibitors. Here, we argue that current cancer research may be on the wrong track, by considering cancer more as a “monogenic” disease, trying to extract common information from thousands of patients, while not properly considering complexity and individual diversity. We suggest to empower a systems cancer approach which reconstructs the information network that has been altered by the tumorigenic events, to analyze hierarchies and predict (druggable) key nodes that could interfere with/block the aberrant information transfer. We also argue that the inter-individual variability between patients of similar cohorts is too high to extract common polygenic network information from large numbers of patients and argue in favor of an individualized approach. The analysis we propose would require a structured multinational and multidisciplinary effort, in which clinicians, and cancer, developmental, cell and computational biologists together with mathematicians and informaticians develop dynamic regulatory networks which integrate the entire information transfer in and between cells and organs in (patho)physiological conditions, revealing hierarchies and available drugs to interfere with key regulators. Based on this blueprint, the altered information transfer in individual cancers could be modeled and possible targeted (combo)therapies proposed.

This article is protected by copyright. All rights reserved.