Time-matched Baseline Research Articles

Impact of baseline ECG collection on the planning, analysis and interpretation of ‘thorough’ QT trials

The current guidelines, ICH E14, for the evaluation of non-antiarrhythmic compounds require a 'thorough' QT study (TQT) conducted during clinical development (ICH Guidance for Industry E14, 2005). Owing to the regulatory choice of margin (10 ms), the TQT studies must be conducted to rigorous standards to ensure that variability is minimized. Some of the key sources of variation can be controlled by use of randomization, crossover design, standardization of electrocardiogram (ECG) recording conditions and collection of replicate ECGs at each time point. However, one of the key factors in these studies is the baseline measurement, which if not controlled and consistent across studies could lead to significant misinterpretation. In this article, we examine three types of baseline methods widely used in the TQT studies to derive a change from baseline in QTc (time-matched, time-averaged and pre-dose-averaged baseline). We discuss the impact of the baseline values on the guidance-recommended 'largest time-matched' analyses. Using simulation we have shown the impact of these baseline approaches on the type I error and power for both crossover and parallel group designs. In this article, we show that the power of study decreases as the number of time points tested in TQT study increases. A time-matched baseline method is recommended by several authors (Drug Saf. 2005; 28(2):115-125, Health Canada guidance document: guide for the analysis and review of QT/QTc interval data, 2006) due to the existence of the circadian rhythm in QT. However, the impact of the time-matched baseline method on statistical inference and sample size should be considered carefully during the design of TQT study. The time-averaged baseline had the highest power in comparison with other baseline approaches.

Evaluation of the impact of lopinavir/ritonavir (LPV/r) and ritonavir (RTV) on QTcF: results of a thorough QT study

Methods Study drugs were dosed for three days to achieve maximal exposure, as RTV-mediated CYP3A4 inhibition is complete and induction is minimal at Day 3. LPV/r was administered at 400/100 mg BID and at supratherapeutic dose of 800/200 mg BID. RTV was dosed at 400 mg BID. Digital EKGs were performed in triplicate on Study Day 3 and compared to time-matched baseline. No effect on QTcF was concluded if the 95% upper confidence bound for the difference of drug mean from placebo mean was less than 10 msec for all times of measurement (Per ICH Guidance). Pharmacokinetic samples were obtained and exposure and QTcF relationship was analyzed.

Thorough QT Study with Recommended and Supratherapeutic Doses of Tolterodine

The objective of our study was to determine the QTc effects of tolterodine. A crossover-design thorough QT study of recommended (2 mg twice daily) and supratherapeutic (4 mg twice daily) doses of tolterodine, moxifloxacin (400 mg once daily), and placebo was performed. Electrocardiograms (ECGs) and pharmacokinetic samples were obtained on days 1-4; time-matched baseline ECGs were taken on day 0. Mean placebo-subtracted change from baseline Fridericia-corrected QT (QTcF) during peak drug exposure on day 4 was the primary end point. Mean QTcF prolongation of moxifloxacin was 8.9 ms (machine-read) and 19.3 ms (manual-read). At recommended and supratherapeutic tolterodine doses, mean QTcF prolongation was 1.2 and 5.6 ms (machine-read), respectively, and 5.0 and 11.8 ms (manual-read), respectively. The QTc effect of tolterodine was lower than moxifloxacin. No subject receiving tolterodine exceeded the clinically relevant thresholds of 500 ms absolute QTc or 60 ms change from baseline. In conclusion, tolterodine does not have a clinically significant effect on QT interval.

Sensitivity/reliability of the time-matched baseline subtraction method in assessment of QTc interval prolongation

Time-matched baseline subtraction (TMBS) is one of the recommended methods to assess QT prolongation of drugs. This study evaluates the sensitivity/reliability of TMBS. Individual heart rate corrected QT (QTc) data were obtained from 22 healthy females who had baseline QT on two consecutive days with ≥9 readings over 24h on either day. 1000 re-sample datasets were generated via randomly selecting individual data series from one of the 2 days as the basis for subtraction. For each dataset, the inter-day difference in QTc at time-matched points (ΔQTc) and its maximum (ΔQTc,max) were determined for all subjects. The mean ΔQTc,max for the 22 subjects and its distribution for the 1000 replicates were then computed. Individual QTc time profile showed fluctuations without a common pattern among all subjects or within a subject across the 2 days. No subject had a ΔQTc of >30ms (highest: 24 ms). The within-day standard deviation (SD) for individual subjects ranged 2.5-8.1 ms. Many subjects had even higher between-day SD. For the 1000 replicates, the mean ΔQTc,max ranged 3.8-13.3 ms and SD ranged 4.0–8.4 ms. This investigation raises a question on the sensitivity/reliability of TMBS method. Additional studies/populations will be similarly investigated. Clinical Pharmacology & Therapeutics (2004) 75, P56–P56; doi: 10.1016/j.clpt.2003.11.211

Nonlinear mixed effects modeling of time dependencies in drug-free ECG data from healthy volunteers

Purpose It may be important to understand time-dependencies in ECG data when developing models as a function of drug concentration in longitudinal studies. A model-based analysis was employed to explore and compare the dependence of RR, QT and QTcI (individual rate-corrected QT) on daytime temporal fluctuations. Methods Data were collected from normal, healthy volunteers (32 subjects, 4 occasions). Intercept (mean) and cosine (period, amplitude (A), mean (M) and offset (O)) models were fitted to RR, QT and QTcI data (NONMEM V, FOCE). Results The cosine model was superior to an intercept model for all endpoints. The final model included exponential inter-individual variance on M and O, additive on A. Selected parameter estimates (% SE) are listed below. (See Table) Inter-occasion variability (IOV) for model parameters was appreciable for all endpoints. Even after correcting for heart rate-dependency, QTcI modeling revealed IOV of 0.9% and 67% in M and A, respectively. Conclusion Time-dependency was not completely removed by heart rate correction and substantial IOV exists. This has direct implication when utilizing time-matched baseline adjustments for determining drug effect. Since it may be of interest to qualify QTc changes as small as 5 msec, uncontrolled variability in time-dependent QTcI and random IOV is of sufficient magnitude to potentially result in misinterpretation of drug-induced QT effects. Clinical Pharmacology & Therapeutics (2004) 75, P97–P97; doi: 10.1016/j.clpt.2003.11.371 Table 1. M (msec) A (msec) RR 904 (1.9) 59.5 (9.3) QT 374 (0.94) 10.3 (9.3) QTcI 386 (0.73) 3.9 (19)

Concentration-QTC modeling in healthy volunteers (HV) and in subjects with major depressive disorder (MDD) for a novel antidepressant candidate

Purpose Evaluation of QTc vs. concentration relationship was undertaken to assess QTc prolongation early in drug development for a novel antidepressant. Objective To evaluate the QTc to concentration relationship in HV and MDD subjects at high concentrations and to predict QTc prolongation in MDD subjects for various sources of pharmacokinetic variability. Methods 28 healthy subjects and 31 subjects with MDD were studied. All subjects received 10 mg QD on Day 1 followed by 30 mg on days 2-6 of the antidepressant candidate. In addition, healthy volunteers received 60 mg QD on days 7–9. Triplicate time-matched baseline ECG data on two baseline days were collected. Triplicate and singlet ECG measurements were collected during treatment arms. Delta QTcF (Fridericia) was modeled using NONMEM v.5. Intercept vs. slope/intercept models were employed. Influence of disease state was also tested. Nonparametric bootstrap was used to construct 95% confidence intervals. Results Concentrations were similar in HV and MDD subjects. A linear QtcF vs. concentration relationship best described the data. HV and MDD subjects had similar slope estimates. Slope for HV: 0.0375 (0.0206, 0.0555) and for MDD: 0.03501 (0.0203, 0.0508). Conclusions Due to the similarity of slope in HV and MDD subjects, use of a population PK/PD model based on phase I data allows for the projection of QTcF in such cases as overdose and drug-drug interaction. Clinical Pharmacology & Therapeutics (2004) 75, P88–P88; doi: 10.1016/j.clpt.2003.11.336