QTc data sensitivity: Is animal screening important to ensure reduced data variability?

Abstract

The ICH E14/S7B Q&As recommend a positive control study to demonstrate the assay sensitivity at a Test Site to detect changes in ECG parameters, namely QTc to enable translatability to the clinic. The objective of this study was to demonstrate the effectiveness of QT prolongation detection using telemetry implants (Physiotel: Data Sciences International) and manually restrained (snapshot) technology in non-human primates (NHP) dosed with moxifloxacin. Two Latin square positive control groups (cohorts 1 and 2) were simultaneously dosed (i.e., same room, drug formulation, technical staff, etc.) with moxifloxacin; 8 NHPs implanted with either the M-11 telemetry device or the L-21 telemetry device. Doses occurred every 7 days for a total of 4 doses and cardiovascular data were collected for 24 h on each occasion on unrestrained animals. Both groups showed measurable QT/QTc prolongation but a higher than anticipated least significant difference (LSD) of 22.4 msec for cohort 1 and 12.60 msec for cohort 2. It was determined that some of the animals had inherently longer QTc intervals than other animals leading to larger individual animal variability. Both cohorts had a maximum standard deviation (SD) of 25.68 and 14.62, respectively. To accurately determine sensitivity capabilities, subjects with similar QT intervals and the same dosing rotation were grouped together in a hybrid Latin square consisting of 2 animals from cohort 1 and 2 animals from cohort 2. The re-assigned animals with similar QT variability (maximum SD = 11.54) resulted in an LSD of 9.8 msec. Pre-screening subjects for inherent ECG waveform variability could allow for greater data sensitivity ensuring adherence to expected industry standard sensitivity. In addition to implanted telemetry, the same 8 subjects had manually restrained, snapshot data collected. All subjects demonstrated similar QT/QTc values from predose to postdose with no observation of QT variability (QTc of 281.4 to 272.7 msec respectively) and as such, could not detect measurable changes in QTc, compared to unrestrained techniques, which did provide measurable changes even with the highest calculated SD. Collectively, these data demonstrate diminished sensitivity of manually restrained, snapshot data as opposed to telemetry data from unrestrained, freely moving animals.