Evaluation of novel treatments through clinical trials remains the backbone of oncological clinical research, but only a minor portion have been tested in Phase III trials. The continued publication of underpowered trials provides an ongoing need for meta-analyses to detect clinically significant outcomes. Although tumor relapse and survival are important issues and easily measured outcomes in trials, they are often not the most relevant indicators for treatment success. As diagnostic technologies and treatments continue to advance, methodologies defining high quality studies have been established, but still enthusiasm to adopt novel technologies that leads to studies holding well-described bias that do not aid the rational use of the studied test. Global awareness of such bias and standard research methodology is the clue toward iconic studies giving rational supporting novel cancer treatments and patients' support.
Keywords: Cancer genomics; Cox proportional hazards regression Analysis; Predictive oncomarkers; Prognostic cancer factors; Recursive partitioning.
Copyright © 2021 Elsevier Inc. All rights reserved.