[Clinical trials in Radiation Oncology: Methodologic approach].

In the context of the clinical development of radiation oncology, the specificities of curative treatments and the necessary follow-ups for the acute and late tolerance evaluation require rigourous and up-dated methodological approaches given the limited feasibility of some studies to demonstrate their effectiveness. Indeed, the diversity of treatments in terms of delivery, type of radiation and multiple technologies render difficult the medical assessment. Although the randomized controlled trial is the gold standard for demonstrating the causal link of the treatment effect size, a state of the art of current limits is presented and proposals for new methodological approaches are discussed as alternative or complementary possibilities. Co-primary endpoints or pragmatic composite endpoints are to be used with adequate statistical analyses, the use of Bayesian methods, the re-use of observational data for the external control arms identification and the development of Real World Data registers is to be preferred to respond to this colossal challenge.