Nonparametric Approaches to Assess the Risk of QT Prolongation of an Investigational New Drug

• Activity Number: 240
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Kristin Ashley Linn*+ and Sujit Kumar Ghosh
• Companies: North Carolina State University and North Carolina State University
• Keywords: Bayesian inference ; Monte Carlo simulation ; QT interval ; QT prolongation ; Linear mixed models ; Nonparametric

Abstract

The QT interval is the portion of the heart's electrical cycle that corresponds to ventricular depolarization and subsequent repolarization. Some non-antiarrhythmic drugs may cause QT interval prolongation, which is associated with a potentially fatal condition called torsades de pointes. In this paper, we use a Gaussian process framework to estimate the maximum difference from baseline of the corrected QT interval (QTc). The mean of the Gaussian process is approximated by a sequence of Bernstein polynomials, which is shown to result in a simple linear mixed model framework. The maximum difference between the mean functions is then estimated to characterize the effect of treatment on the QTc interval. The results of a simulation study demonstrate that our proposed method has several advantages over existing methods including decreased bias and mean squared error. We apply our method to a real data set obtained from a 'thorough QT/QTc' study conducted by GlaxoSmithKline.