Experimental Design and Sample Size for Long-Term Carcinogenicity Studies: Statistical Considerations

• Activity Number: 31
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Lei Shu*+ and Lanju Zhang and Sherry Morgan
• Companies: AbbVie and AbbVie and AbbVie
• Keywords: Experiment Design ; Sample Size ; Power

Abstract

The assessment of the safety of a pharmaceutical product often includes the study of carcinogenic risk in 2-year rodent bioassays. Differential or excess mortality among the study groups may lead to an early termination of one or more study groups in order to maintain sufficient power for statistical test. From recent communication with regulatory agency (FDA-CAC), there are specific sample size requirements on both control group and dose groups. To better comply with the regulation and smartly use our limited resources, we proposed to use a statistical simulation method to evaluate the risk of early termination. We used an example to show how to compare two plans by the simulation method. The simulation method can also be used to simultaneously evaluate multiple end points, for example, the risk of termination together with the power of the study.