Validation Testing in Thorough QT/QTc Clinical Trials∗

∗This article represents the points of view of the authors. It does not necessarily represent the official position of the U.S. FDA.

Abstract

In order to validate the results of a thorough QT/QTc (the duration of depolarization and repolarization of ventricles or the duration corrected for heart rate) clinical trial, ICH E14 recommended to include a concurrent positive control treatment in the trial. It further recommended that validation is achieved if the positive control has an effect on the mean QT/QTc interval of about 5 ms. Zhang (2008) discussed the intersection-union test approach for testing the validation hypotheses and an alternative global average test approach. In this article, we further discuss the difference and relationship of the two sets of hypotheses and the difference in the efficiencies of the two approaches. We conclude that validation can be achieved if either one test rejects the null hypotheses without inflating the family-wise Type I error rate. However, using both approaches may improve the efficiency in validation assessment.

Key Words:

• Assay sensitivity
• Controlling Type I error rate
• QT prolongation
• Statistical testing

ACKNOWLEDGMENTS

The authors want to thank Dr. Joanne Zhang of the FDA for sharing her manuscript on assay sensitivity tests, for which this article is a follow-up. The authors are also in debt to the referee for encouragement and constructive comments and suggestions for revision.

Notes

∗This article represents the points of view of the authors. It does not necessarily represent the official position of the U.S. FDA.