Association between Phase II Trials Design and Successful Subsequent Phase III Trials

Sabrina Vari (1), Vanja Vaccaro (1), Emilio Bria (2), Michele Milella (1), Diana Giannarelli (1)

(1) Medical Oncology, Regina Elena National Cancer Institute, Italy
(2) Medical Oncology, University of Verona, Italy

Background
The process of developing drugs is a challenge often leading to a resource consumption contributing to determine the high costs of the few molecules that gain the approval. Besides costs also ethical aspects have to be considered during drug process because of patients waiting for the development of effective drugs. So phase II design is crucial to foreseen the results of a phase III trial.

Objective
To investigate phase II features associated with subsequent phase III successful trials in advanced/metastatic solid tumors.

Methods
We systematically reviewed through MEDLINE literature from 2011 to date searching for phase III trial on novel (targeted or immuno) therapies or new drug combination or new indication. Once phase III trials were selected according to established criteria the preceding phase II trials were retrieved. Features of phase II studies considered in the analysis included primary endpoint, randomization, number of participating centers, sample size, follow-up length and sponsor type. Association was measured with chisquare test and logistic model.

Results
We selected 232 trials, 122 with Overall Survival (OS) as primary endpoint, 107 with Progression Free Survival (PFS), three trials were designed with alternative endpoints. A positive result was found in 102 trials (43.9%), 45 (44.1%) designed on OS and 57 (55.9%) on PFS. Negative studies were 130, 77 (59.2%) considered OS as the primary endpoint. Analysis on the association of these results with phase II trial is still ongoing.

Conclusion
In a phase III setting the choice of primary endpoint and effect size is crucial. OS is considered the gold standard but can be biased by subsequent therapies and crossover effect. A well designed multicenter phase II trial can help to avoid time and resources consumption during drug development.