Oncology phase III trials have a high failure rate, leading to high development costs. The Clinical Trials Design Task Force of the Investigational Drug Steering Committee of the NCI Cancer Therapy and Evaluation Program developed Recommendations regarding the design of phase II trials. We report here on the results of a Concordance Group review charged with documenting whether concordance rates improved after the publication of the Recommendations. One hundred and fifty-five trials were reviewed. Letter of Intents (LOI) from the post-Recommendation period were more likely to be randomized (44% vs. 34%) and biomarker selected (19% vs. 10%). Single-arm studies using time-to-event endpoints (benchmarked against historical data) were similar, as was the type of tumor. There was a significant improvement in the rate of concordance, with 74% of LOIs scored as concordant compared with 58% before the Recommendations (P = 0.042). This included a marked decrease in the use of single-arm designs to evaluate the activity of drug combinations (19% vs. 5%, P = 0.009). There were areas for which clarification was warranted, including the need for protocols to include further development plans, the use of realistic benchmarks, the careful evaluation of historical controls, and the use of a standard treatment option as a control. Ongoing critical evaluation of current trial design methodology and the development of new Guidelines when appropriate will continue to improve drug development ensuring that safe and effective cancer therapeutics are made available to our patients as quickly and efficiently as possible.
©2015 American Association for Cancer Research.