How to run survival-endpoint phase II trials with 25% to 40% fewer patients.
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
2544 Background: New molecular information, whether arising from expression profiling, identification of key mutations, or proteomic analyses is making the personalization of therapy increasingly feasible. This leads to the fragmentation of common disease entities into multiple molecular subtypes. Having sufficient numbers of patients to demonstrate the effectiveness of novel targeted therapies for each subtype has therefore become challenging, essentially impeding progress. Methods: Use of sample size determination software allows for comparison of phase II trials with a target survival proportion at a given time T based upon Kaplan-Meier versus exponential survival methods. Results: The use of exponential survival fitting methods, in lieu of the currently implemented trial designs, to single arm phase II studies can routinely reduce patient numbers by 25-40% (Table) without compromising the statistical power. The biggest advantage is that survival information both before and after time T reduces the variability for the exponential fit estimate. This approach will lead to the saving of financial and patient resources for researchers. Moreover, the time saved will accelerate the approval of agents making the proposed trial designs ethically attractive. However, this approach is only applicable if the exponential distribution assumption behind the approach is valid. Conclusions: The exponential survival fitting method should replace the current standard approach in all instances unless the exponential distribution assumption is known to be false. Detailed examples and analysis will be presented. [Table: see text]
