3488 A comparison between the Rolling 6 and 3+3 dose escalation study designs for phase 1 clinical trials

Abstract

OBJECTIVES/SPECIFIC AIMS: The development of new anti-cancer agents for children requires an inherently longer timeline than in adults. The 3+3 study design for Phase 1 dose escalation trials is commonly used to estimate the maximum tolerated dose and assess safety. The Rolling 6 study design was developed to shorten the study conduct timeline. METHODS/STUDY POPULATION: This study compares twenty Phase 1 COG Pilot and Phase 1 Consortium trials that employed the Rolling 6 design with hypothetical results under the assumption that a 3+3 design had been executed. The number of evaluable patients required to complete the study, number of DLTs, number of inevaluable patients, overall study duration, time suspended to enrollment (i.e., waiting for DLT evaluation), and DLT risk are compared between study designs using Wilcoxon’s signed rank test. RESULTS/ANTICIPATED RESULTS: The Rolling 6 study design required less time to complete the studies compared with 3+3 design (median 273 vs. 297 days, P = 0.01). In general, the Rolling 6 study design required more patients, had more inevaluable patients, and there were more dose limiting toxicity (DLT) events. However, there was no significant difference in DLT risk (median 0.15 vs. 0.17, P = 0.72). DISCUSSION/SIGNIFICANCE OF IMPACT: The Rolling 6 study design effectively shortens the study conduct timeline compared with the traditional 3+3 design for Phase 1 COG Pilot and Phase 1 Consortium trials without increasing the risk of toxicity.