Re-engineering the process of cancer clinical research in data driven health care systems.

The current framework surrounding the clinical development of medicines has provided patients with many innovative therapies whose safety, quality and efficacy have been empirically demonstrated. However, many important aspects relating to the use of these therapies in real-life conditions are often neglected during the clinical development stage, including how to combine them with other existing treatments, how they compare to alternatives, how they will perform in populations that were not studied in clinical trials (children, elderly people, smokers, pregnant women, patients with renal or hepatic impairment), how long they have to be administered to achieve the desired effects and whether a lower dose could produce the same results. It is clear that there is a research gap between the pre-approval development of medicines and their post-approval use in real-life clinical practice, making it difficult for clinicians to decide on how to best treat their patients. In oncology, this gap is particularly problematic, as the rise of precision medicine has led to expensive therapies being granted marketing authorization and reimbursement without sufficient real-world evidence to confirm their effectiveness. This PhD project focuses on potential approaches to bridge the gap by exploring how pragmatic clinical trials, big data analytics and new technologies such as artificial intelligence and machine learning could be utilized to generate the necessary real-world evidence that is currently lacking. The aims are to 1) characterize to which extent these methods could be harnessed to fill the real-world clinical research vacuum, 2) outline their past, current and future impact on regulatory and clinical decision-making, 3) clarify the views of key stakeholders in the drug development process with respect to these strategies, and 4) investigate the ethical, economic and regulatory dimensions of real-life clinical research incorporating such strategies.