Characteristics of a phase II study that predict for a positive phase III trial—A study of 383 clinical trials
6604 Background: To identify the characteristics of a phase II study that predict for a subsequent positive phase III trial. Methods: All phase II studies and subsequent phase III clinical trials on biologics in advanced cancers published from 1985 to 2005 were extracted. Chi-square test and logistic regression models were used for analyses. Results: 383 phase III clinical trials and their preceding phase II studies were identified. 183 (47.8%) phase III trials were “positive” and 200 (52.2%) were negative. 220 trials (57.4%) used biologics alone and 162 (42.3%) used a combination of biologics and chemotherapy. Over the study periods 1985–1990, 1991–1995, 1996–2000, 2001–2005, the percentage of phase II studies that led to positive phase III trials increased from 37.7% to 33.3% to 56.0% to 76.8% (p<0.001). The interval between the publication of phase II and III studies, 0.5–5, 6–10, 11–15, and 16–20 years were also associated with the success of phase III trial, 55.6%, 42.2%, 32.6%, and 10.0%, respectively (p<0.001). Phase II studies from multiple rather than single institutions were more likely to have a successful trial (60.4% vs. 39.4%; p<0.001). The percent of successful trials from pharmaceutical companies was significantly higher compared to academic, cooperative groups, and research institutes (89.5% vs. 44.2%, 45.2%, 46.3%; p=0.002). The publication of the phase II studies in journals with an impact factor of 8 or greater compared to those less than 8 was also predictive (44.1% vs. 58.0%; p=0.024). Phase II studies with a lower attrition rate were also associated with a positive phase III trial (61.1% vs. 41.8%; p=0.025). On multivariable analysis, all factors, except for journal impact factor, were independent predictive factors for a positive phase III trial. Conclusions: In phase II biologic studies, characteristics such as larger number of patients, more recent year of study, multiple vs. single institution participation, shorter time period between publication of phase II to phase III trial, and lower rate of attrition were predictive factors of success in a phase III trial. Investigators need to be cognizant of these phase II study characteristics before designing phase III trials. No significant financial relationships to disclose.