Choice of baseline in a multiple-dose thorough QT study (TQTS) - effect on analysis of moxifloxacin-induced QTc prolongation

Background: The primary endpoint of a thorough QT study (TQTS) is the change from baseline in QT corrected (QTc) measured on electrocardiograms (ECG) tracings. It has been suggested that during a crossover study, the time-matched or predose baseline could be recorded. The choice of method for baseline ECG collection may influence the results and the cost of the TQTS. Objective: The objective of our study was to compare the collection of a time-matched baseline before each period (TM (EACH)), an average of all the time-matched baseline (TMMEAN), a time-matched baseline before period 1 (TMP1) and a predose baseline (PDEACH) on QT interval prolongation induced by moxifloxacin, in a 30 subjects, 5 arm cross-over TQTS. Results: Moxifloxacin induced a similar significant increase in QT corrected using Fridericia's formula (QTc) (lower limit of the confidence interval excluded 5 ms) at all time-points between 0.5 hours and 12 hours with all baseline methods. TMEACH was associated with a lower within subject variability (SD 5.59 ms) than TMMEAN and TMP1 (5.90 and 6.90 ms, respectively). PDEACH was associated with the highest variability (7.41 ms). Conclusion: The collection of ECG tracings during a full baseline day before each period was associated with the lowest variability. However, the two more cost effective designs (TMP1 and PDEACH) were sufficient, in this small TQTS, to significantly detect moxifloxacin.