Extrapolation of clinical trial data is being accepted increasingly by regulatory agencies as a means of generating data in diverse situations during drug development process. Under certain circumstances, data can be extrapolated to a different population, a different but related indication, and different but similar product. We consider here the problem of confidence interval estimation for safety data (e.g., adverse events, immunogenicity occurrences) using the concept of estimand [Akacha, M., et. al. (2017)] and a mixed models approach under an extrapolation setting.
Clustered safety data arising in clinical trials come with varying cluster sizes. We propose an approach to construction of confidence interval and evaluate its performance, considering estimators obtained by different weighting schemes of the clusters (equally weighted, with weights proportional to cluster size, and with minimum variance weights). We then evaluate the performance of this approach using simulated data under varying scenarios.
In conclusion, we see that the approach is a useful means for extrapolating both efficacy and safety data (e.g., from adult to pediatric population or from one indication to another) and thus aids not only signal detection but risk-benefit evaluation as well.
Daniel Bonzo is Vice President and Global Head of Biometry at LFB. His research interests include methods for analyzing mixed data and genomics data and design of experiments. Prior to his involvement in the pharmaceutical and biotechnology industries, he was Associate Professor at the School of Statistics of the University of the Philippines in Diliman. He also served as consultant in the Philippines’ semiconductor and telecommunication industries. He obtained his PhD in 2003 from the University of the Philippines – Diliman and did his post-doctoral research in 2005 at Rikkyo University in Tokyo.