Analyzing treatment effects on cardiovascular safety (QT)

Abstract

All medications are expected to undergo cardiovascular safety evaluations in the beginning of development as early as possible to evaluate their effect on cardiac repolarization, an effect that can be measured as prolongation of the QT interval on the surface electrocardiogram (ECG). QT prolongation is considered to be the best available predictor of risk for dysrhythmia and sudden death. This thesis primarily focuses on assessing the effect of a drug X at steady state compared with the placebo group on the length of the QT interval corrected for heart rate using linear derived method (QTcLD). The second objective is to study the effect of baseline imbalances on the estimated treatment effect. In this randomized placebo-controlled trial, imbalances in the QTcLD interval distribution at baseline values and covariates between randomized groups were observed. To take this into account four approaches were performed. Properties and challenges of each method were compared and discussed. The analyses were carried into two parts: first the observations at 4 hour 30 minutes (time point 4.5) after dosing were analyzed. Then repeated measures analysis was performed to account for all time point measurements. At time point 4.5 both doses (12 mg and 18 mg) of drug X statistically and clinically prolonged the QTcLD interval. However when all time points were considered neither 12 mg nor 18 mg of drug X at steady state had effect on the prolongation of the length of the QTcLD interval. Baseline imbalance between treatment groups in baseline was influent the results, hence adjusting in the statistical analysis for baseline variables was performed.