Thorough QTc Studies: Patients at Heart... PhUSE 2013 Brussels - PowerPoint PPT Presentation

Thorough QTc Studies: Patients at Heart... PhUSE 2013 – Brussels Marc Derycke/Jens-Otto Andreas 16.OCT.2013

2 Agenda ן I ן D ן The Thorough QTc The Thorough QTc ן A ן C

Introduction Early detection of potential serious Adverse Events (AEs) is a key element of patient safety. One of these risks is drug-induced cardiac rhythm modification. A model study type, called Thorough QTc (TQT), has been developed to identify such alterations. How real-life trials cope with the strict conditions of TQT studies populations?

4 Agenda ן I ן D ן The Thorough QTc The Thorough QTc ן A ן C

Drug-Induced Cardiac Rhythm Modifications Unexpected troubles of cardiac rhythm caused by a drug. ( à Delay in ventricular repolarization). Effect measured by QT interval prolongation on ECG.

QT interval prolongation QT interval = time for ventricle depolarization § (contraction) and repolarization (relaxation) § Varies with heart rate à requires a correction for this, giving the QTc.

Drug-Induced Cardiac Rhythm Modifications (Cont.) Drugs can cause delays in cardiac repolarization. Effect measured by QT interval prolongation on ECG. Favors development of ventricular arrhythmias. à Could degenerate into ventricular fibrillation, then sudden death. Often discovered well after regulatory approval.

8 Agenda ן I ן D ן The Thorough QTc The Thorough QTc ן A ן C

The Thorough QTc Study Response: ICH-developed, FDA-endorsed Guidance document: E14 Guidance on Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential for Non-Antiarrhythmic Drugs Defines the concept of “Thorough QTc Study” (TQT study), able to exclude proarrhythmic risk of a drug. Threshold level of regulatory concern : mean effect of the study drug on QTc prolongation around 5ms. à Upper bound of one-sided 95% CI for the time-matched mean effect of the drug on baseline and placebo-corrected QTc would reach +10ms, for at least one timepoint of the study.

The Thorough QTc Study (Cont.) Positive TQT study Negative TQT study Difference in Δ QTc Difference in Δ QTc 12 12 10 10 +10ms ΔΔ QTc - mean (+/- 90% CI) 8 8 ΔΔ QTc - mean +/- 90% CI 6 6 4 4 2 2 0 0 -2 -2 -4 -4 -6 -6 0 4 8 12 16 20 24 0 4 8 12 16 20 24 Time (h) Time (h) Negative TQT study (no effect on repolarization) à No need for extensive ECG safety evaluations during drug development.

The Thorough QTc Study (Cont.) Generally carried out on healthy volunteers (ease of ECG recordings, better control of autonomic conditions and responses to environment). Placebo controlled (effect of study conduct on ECG). Uses maximum tolerated dose of the study drug. Adjusted for baseline.

The Thorough QTc Study (Cont.) Includes ECG assay sensitivity testing – to show the study is able to detect tiny changes in the QTc interval duration. Uses a positive control (drug known to induce QTc prolongation similar in amplitude to the one of the 5ms threshold). à TQT studies need to follow a very strict and complex set of conditions to provide valid conclusions. BUT What to do if some of these stringent conditions could not be met?

13 Agenda ן I ן D ן The Thorough QTc The Thorough QTc ן A ן C

A Real-Life Example TQT study for dopamine agonist rotigotine in treatment of Parkinson’s disease (PD). Supratherapeutic doses not tolerated by healthy volunteers à need to recruit Advanced PD patients. PD affects the quality of ECG recordings : - muscle tremors - skin conditions - disease-related autonomic disturbances.

A Real-Life Example (Cont.) 24h ECGs Study design: Day Day Day Day Day 21/22 28/29 35/36 42/43 14/15 Baseline Day Day 32 39 Day 32 Day 39 Placebo M/P P/M Days Days Days Days Days Days 1 to 7 7 to 14 14 to 21 21 to 28 28 to 35 35 to 42 R 4mg/24h 8mg 12mg 16mg 20mg 24mg Rotigotine Day Day 32 P 39 P Baseline ECGs § Parallel-group trial, double-blind and placebo- controlled. § Positive-controlled (moxifloxacin 400mg) for sensitivity analysis.

A Real-Life Example (Cont.) ECG recording and processing: Continuous recording of 10-s ECG samples over 24h. During daytime, measured in a standard semi-reclined position (to reach stable heart rate) for 5 minutes at regular interval. At nighttime, 20-minutes recordings every 30 minutes. Within each recording session, 5 samples preceded by stable heart rate period and with the lowest noise level are selected.

A Real-Life Example (Cont.) With the selected records, construction of a median representative beat used to determine QT interval through pattern-matching algorithm. Median representative beat constructed from filtered signal allows reliable QT interval measurement (vertical lines indicate Q-onset and T-offset points). Review of the automated measurements: manual calculation by 2 independent cardiologists, reconciled by a senior cardiologist if in persistent disagreement. Heart rate measured across baseline recordings determine correction factors for QT interval (daytime / nighttime).

A Real-Life Example (Cont.) Statistical analysis: Parallel-group comparison of rotigotine vs. placebo based on time-matched changes from baseline (average of days − 2 and − 1) at each daytime and nighttime data point ( ΔΔ QTc values). Non-inferiority comparisons of rotigotine vs. placebo using one-sided 95% CI (upper limits of two-sided 90% CI) for the time-matched changes from baseline at each data point. Point estimates and CI were calculated using analysis of covariance, with effects for site, treatment, gender, age, and time-matched baseline QTc as covariates, for each data point separately. If upper CI limit for the comparison of rotigotine vs. placebo below +10ms at each data point à rotigotine has no effect on QTc.

A Real-Life Example (Cont.) Analysis of variance was used to compare the positive control vs. placebo at each data point and calculate point estimates of the respective two-sided 95% CI for the difference in effects between use of placebo and moxifloxacin. Sensitivity of the assay considered established : - one-sided 95% CI (lower limit of two-sided 90% CI) at the maximum effect excludes the possibility of QTc increase < 5ms - profile of QTc changes corresponds to the expected drop in moxifloxacin plasma levels after the infusion.

A Real-Life Example (Cont.) Results: ΔΔ QTc differences (rotigotine minus placebo) in the time-matched changes from baseline at individual data points (intersection union test) à All differences were close to zero. In particular, all upper limits of the 90% CIs were below +5ms, and all lower limits of the 90% CI were above − 5ms.

A Real-Life Example (Cont.) Established sensitivity: • Lower limit of the confidence interval of mean QTc differences between moxifloxacin and placebo in the changes from baseline ( ΔΔ QTc) at peak effect excludes the possibility of an increase < 5ms. • ΔΔ QTc drops along the moxifloxacin plasma levels after the infusion.

A Real-Life Example (Cont.) The placebo infusion also appeared to have a small QTc effect (environmental effect). Variability of the QTc data was very small in both placebo and rotigotine arms despite the long study duration, confirming the very tight ECG measurements.

23 Agenda ן I ן D ן The Thorough QTc The Thorough QTc ן A ן C

Conclusion Thorough ECG studies can be performed in patients with advanced-stage Parkinson’s disease and, therefore, probably also in other patient populations with diseases that have profound effects on the quality of ECG recordings - or other pharmacological compounds that are not tolerable in healthy subjects. Success of this study setup was attributable to the diligent care exercised during the clinical conduct of the study and the attention given to ECG processing. Rotigotine application up to the supratherapeutic doses of 24 mg/24 h did not induce any QTc interval changes, thereby indicating that the drug is not likely to have any cardiac-repolarization impact.

25 Questions?

Thanks!