Comparison of Statistical Models Adjusting for Baseline in the Analysis of Parallel-Group Thorough QT/QTc Studies

A powerful test for the maximum treatment effect in thorough QT/QTc studies

Parallel-group thorough QT/QTc studies focus on the change of QT/QTc values at several time-matched points from a pretreatment day (baseline) to a posttreatment day for different groups of treatment. The International Council for Harmonisation E14 stresses that QTc prolongation beyond a threshold represents high cardiac risk and calls for a test on the largest time-matched treatment effect (QTc prolongation). QT/QTc analysis usually assumes a jointly multivariate normal (MVN) distribution of pretreatment and posttreatment QT/QTc values, with a blocked compound symmetry covariance matrix. Existing methods use an analysis of covariance (ANCOVA) model including day-averaged baseline as a covariate to deal with the MVN model. However, the ANCOVA model tends to underestimate the variation of the estimator for treatment effects, resulting in the inflation of empirical type I error rate when testing whether the largest QTc prolongation is beyond a threshold. In this article, we propose two new methods to estimate the time-matched treatment effects under the MVN model, including maximum likelihood estimation and ordinary-least-square-based two-stage estimation. These two methods take advantage of the covariance structure and are asymptotically efficient. Based on these estimators, powerful tests for QT/QTc prolongation are constructed. Simulation shows that the proposed estimators have smaller mean square error, and the tests can control the type I error rate with high power. The proposed methods are applied on testing the carryover effect of diltiazem to inhibit dofetilide in a randomized phase 1 trial.

Baseline selection in concentration-QTc modeling: impact on assay sensitivity

The baseline selection in concentration-QTc (C-QTc) modeling is not well studied in the literature. Time-matched baseline and pre-dose baseline have been commonly used as a covariate in C-QTc modeling for parallel and crossover study, respectively. It has been showed that the C-QTc model using time-matched baseline has a low chance of showing assay sensitivity in parallel study. To better understand the impacts of baseline section in C-QTc, we examined the original and subsampled moxifloxacin and placebo data from more than 50 of TQT studies submitted to FDA with regard to assay sensitivity. Our analyses show that baseline selection (time-matched, pre-dose, average) has an impact on prediction from C-QTc modeling and the impact depends on study design (parallel, crossover). The impact to categorical table of ΔQTc is unlikely to alter the interpretation of the outlier category (ΔQTc>60) that corresponds to the regulatory concern. The results presented here can guide C-QTc study design as well as baseline selection in C-QTc modeling.

Clinical ECG Assessment

With the adoption of the ICH E14 guidance, the thorough QT/QTc (TQT) study has become the focus of clinical assessment of an NCE's effects on ECG parameters. The TQT study is used as a guide to the liability of a drug to cause proarrhythmias on the basis of delayed cardiac repolarization. Around 300 TQT studies have been performed since 2005 and through interactions between sponsors and regulators, especially FDA's Interdisciplinary Review Team (IRT) for QT studies. These studies can today be performed more effectively and with great confidence in the generated data. This chapter will discuss technical features and the design and analysis of TQT studies, how assay sensitivity is demonstrated, and examples from recently conducted studies. ECG assessment for drugs that cannot be safely given to healthy volunteers is also addressed, and examples from studies in cancer patients and in healthy volunteers with tyrosine kinase inhibitors are discussed. The TQT study is resource intensive and designed to solely evaluate whether an NCE prolongs the QTc interval. If data with similar confidence can be generated from other studies that are routinely performed as part of the clinical development, this would represent a more optimal use of human resources. Methods and approaches to increase the confidence in ECG data derived from "early QT assessment" in single-ascending/multiple-ascending dose studies are therefore discussed, and a path toward replacing the TQT study using these approaches will be outlined.

Circadian Variation and Baseline Definition in Parallel-Group Thorough QT Studies

For thorough QT (TQT) studies employing a parallel-group design, there has been a clear regulatory preference for the use of time-matched, rather than time-averaged, baseline values to account for circadian variation when estimating the magnitude of the drug effect on the QT interval. In this paper, both historical data from parallel-group TQT studies and simulated data from assumed circadian models are utilized to comprehensively assess the performance characteristics of 3 repeated-measures analysis of covariance models. The results indicate that each analysis model performs adequately in the absence of an observed time-matched baseline imbalance between the treatment groups. However, the analysis model with time-matched baseline as a covariate performs poorly under the setting of an observed time-matched baseline imbalance between the treatment groups. The analysis model with time-averaged baseline as a covariate and the analysis model with both time-matched and time-averaged baselines as covariates provide unbiased estimates of the treatment difference and properly control the type I error rate, regardless of an observed time-matched baseline imbalance or within-patient variation in circadian parameters.