ABSTRACT
Innovative methodologies are always proposed for and applied to the development of oncology therapies, not only because cancer is a major cause of death and reduced life expectancy with high unmet medical needs, but also due to the complexity of cancer pathogenesis which demands innovative approaches to mitigate risk and optimize resources. In clinical trials, the classic statistical techniques in analyzing right-censored TTE data are commonly based on the non-parametric Kaplan Meier (KM) product limit estimation of survival function or the semi-parametric Cox proportional hazards regression. Putter et al. provided a comprehensive tutorial on statistical methods in analyzing competing risks with an emphasis on practical issues in clinical trials, and Koller et al. discussed the translational aspects of competing risks to clinical research. Response-adaptive randomization is always combined with adaptive enrichment design, based on which, randomization probability could be updated with accruing within-study data in a continuous learning process.
