Critical Review of Umbrella, Basket, and Platform Designs for Oncology Clinical Trials
Until recently, clinical development of cancer drugs has been conducted separately for different histological types of tumors, because histological type was the primary known determinant of drug responsiveness. With the development of targeted therapy, this histology focus has been supplemented by focus on genomic alterations in the tumor. The enrichment design, illustrated in Figure1, has been used extensively over the past 15 years for phase III registration studies. Patients whose tumors do not contain the genomic alteration, which would render them likely to be sensitive to the test drug, are excluded from the randomization. If the drug is shown to be effective for those whose tumors contain the genomic alteration, it could possibly be tested for the other patients if the drug has multiple targets. In the United States, approval of a drug for a genomic marker restricted population is dependent on the availability of a US Food and Drug Administration approved test for the genomic alteration. The enrichment design can require many fewer patients to be randomized relative to the “all comers” randomized design.1
In this article, we will describe some of the more recently developed designs for genomically driven clinical trials in oncology.