Cancer modeling: From mechanistic to data-driven approaches, and from fundamental insights to clinical applications

Cancer is still one of the major causes of death worldwide. Even if its comprehension is improving continuously, the complexity and heterogeneity of this group of diseases invariably make some cancer cases incurable and lethal. By focusing only on one or two cancerous molecular species simultaneously, traditional in vitro and in vivo approaches do not provide a global view on this disease and are sometimes unable to generate significant insights about cancer. In silico techniques are increasingly used in the oncology domain for their remarkable integration capacity. In basic cancer research, a vast number of mathematical and computational models has been implemented in the past decades, allowing for a better understanding of these complex diseases, generating new hypotheses and predictions, and guiding scientists towards the most impactful experiments. Although clinical uptake of such in silico approaches is still limited, some treatment strategies are currently under investigation in phase I or II clinical trials. Besides being responsible for new therapeutic ideas, in silico models could play a significant role in optimizing clinical trial design and patient stratification. This review provides a non-exhaustive overview of models according to their intrinsic features. In silico contributions to basic cancer science are discussed, using the hallmarks of cancer as a guidance. Subsequently, in silico cancer models, that are a part of currently ongoing clinical trials, are addressed. In a forward-looking section, issues such as the need for adequate regulatory processes related to in silico models, and advances in model technologies are discussed.