Case studies: Clinical trials with multiple objectives
Multiple objectives
Multiplicity issues arise in all modern Phase III trials due to the evaluation of several clinical endpoints, several doses or regimens of an experimental treatment or several pre-defined patient populations. The resulting multiplicity problems can be organized by reviewing the sources of multiplicity in each trial, please see examples below.
Mediana has developed state-of-the-art multiplicity adjustment strategies in dozens of Phase III trials with multiple objectives and also built custom software tools (Windows-based or web-based applications with a graphical-user interface) to evaluate their performance under multiple treatment effect scenarios and identify an optimal strategy in each trial. The multiplicity adjustment strategies were accepted by the FDA and EMA without any changes or comments.
For information on the custom software tools developed to address complex multiplicity issues in these Phase III trials, see the custom software page.
Single source of multiplicity
A detailed comparison of candidate multiplicity adjustment strategies have been performed in several Phase III trials where multiplicity was induced by the analysis of several clinical endpoints (primary efficacy endpoint and key secondary efficacy endpoints). Extensive simulations are conducted to select the most powerful and robust multiplicity adjustment strategy in a trial. Examples include clinical trials in patients with lumbar disc herniation, myasthenia gravis or refractory acute pericarditis.
Two sources of multiplicity
Two sources of multiplicity are found in virtually all Phase III trials that investigate the efficacy and safety of several doses of an experimental treatment. The multiple dose-control comparisons define the first source of multiplicity and the evaluation of multiple efficacy endpoints defines the second source of multiplicity. Multiplicity adjustment strategies, known as gatekeeping procedures, are developed to address both sources of multiplicity and simulation-based assessments are required to identify the most appropriate gatekeeping procedure for each trial. Efficient gatekeeping procedures have been developed for numerous neuroscience, oncology and other trials.
Three sources of multiplicity
Very complex settings with three sources of multiplicity arise in Phase III trials with adaptive or group-sequential designs (and thus there are multiple decision points) where several dose-control comparisons are carried out with respect to several efficacy endpoints. A Phase III trial in patients with stable atherosclerotic vascular disease serves as the most recent example of a trial with three sources of multiplicity. Two doses of an investigative drug were tested in this trial using one primary and three secondary endpoints and the evaluations were performed at two interim analyses and the final analysis. The trial was stopped at the very first interim analysis due to a very strong treatment effect.