The transition from phase II to phase III studies

Abstract

6514 Background: Phase II trials are performed to detect potential anti-tumor effects of a new treatment and should be used to decide whether to proceed to a phase 3 trial or not. However, many phase 2 trials never lead to a phase 3 trial despite encouraging results. Here we sought to determine how often (i) positive phase 2 trials have led to phase 3 trials, and (ii) how often phase 2 trials were designed to lead to a phase 3 trial. Methods: We reviewed 200 phase 2 trials, presented at ASCO meetings in 1995–1996, and 2006, selecting randomly 20 abstracts with encouraging results for 5 cancer sites (breast, lung, GI, GU, Gyn) in each time period. For those presented in 1995–1996, we searched systematically for subsequent randomized studies where one treatment arm was similar to that in the phase 2 study. For those presented in 2006, a questionnaire was sent to authors asking whether they recommend evaluating the regimen in a phase 3 trial, whether a phase 3 trial is planned and whether resources (budget, patients, drugs) are available to conduct a phase 3 trial. Results: Ten years after presenting phase 2 trials with positive results, only 13 regimens have been evaluated in a phase 3 trial. Of 100 investigators who presented a phase 2 trial in 2006, 42 returned the questionnaire, 36 confirmed that the results met criteria of efficacy and 25 thought the regimen should be evaluated in a phase 3 trial. Only 10 investigators plan to undertake a phase 3 trial, and 8 stated they had resources to do so. Reasons for not planning a phase 3 study included insufficient efficacy (7), insufficient access to patients (5) or financial support (5), lack of interest from colleagues (6), and lack of support from the company (8). Conclusions: Few (∼13%) phase 2 trials with promising activity are followed by phase 3 trials and this is not increasing with time. Reasons include lack of resources such as money, drugs and patients. Many of these limitations are known when planning the phase 2 study, implying that many phase 2 trials are not planned as precursors of phase 3 trials. Resources spent on such trials would be better applied to practice-changing phase 3 trials. No significant financial relationships to disclose.