Assessment of Interspecies Differences in Drug-Induced QTc Interval Prolongation in Cynomolgus Monkeys, Dogs and Humans

Abstract

Background and purpose: The selection of the most suitable animal species and subsequent translation of the concentration-effect relationship to humans are critical steps for accurate assessment of the pro-arrhythmic risk of candidate molecules. The objective of this investigation was to assess quantitatively the differences in the QTc prolonging effects of moxifloxacin between cynomolgus monkeys, dogs and humans. The impact of interspecies differences is also illustrated for a new candidate molecule.

Experimental approach: Pharmacokinetic data and ECG recordings from pre-clinical protocols in monkeys and dogs and from a phase I trial in healthy subjects were identified for the purpose of this analysis. A previously established Bayesian model describing the combined effect of heart rate, circadian variation and drug effect on the QT interval was used to describe the pharmacokinetic-pharmacodynamic relationships. The probability of a ≥ 10 ms increase in QT was derived as measure of the pro-arrhythmic effect.

Key results: For moxifloxacin, the concentrations associated with a 50% probability of QT prolongation ≥ 10 ms (Cp50) varied from 20.3 to 6.4 and 2.6 μM in dogs, monkeys and humans, respectively. For NCE05, these values were 0.4 μM vs 2.0 μM for monkeys and humans, respectively.

Conclusions and implications: Our findings reveal significant interspecies differences in the QT-prolonging effect of moxifloxacin. In addition to the dissimilarity in pharmacokinetics across species, it is likely that differences in pharmacodynamics also play an important role. It appears that, regardless of the animal model used, a translation function is needed to predict concentration-effect relationships in humans.

Keywords: PKPD modelling; QT interval prolongation; cardiovascular safety; drug development; interspecies differences.

Fig. 1

Predicted and observed pharmacokinetic profiles of moxifloxacin after administration of different dose levels to dogs (left panels), monkeys (mid panels) and humans (right panels). The upper panels show examples of the individually predicted concentrations. Mid panels: observed (symbols) and population predicted (lines) concentrations. Lower panels: goodness-of-fit plots depicting the observed vs predicted concentrations (symbols). The solid line represents the identity line

Fig. 2

(Upper panels) Examples of individually predicted concentrations of NCE05. Lower panels depict the observed (symbols) and mean predicted (lines) concentrations in monkeys and humans. Pre-clinical doses ranged from 25 mg/kg (dashed line/△) to 40 mg/kg (dotted line/+), whereas healthy subjects received doses of 4 mg (dashed line/△), 14 mg (dotted line/+) or 30 mg (dash dotted line/×)

Fig. 3

Model performance and predicted QT profiles after administration of placebo and different doses of moxifloxacin to dogs (left panels), cynomolgus monkeys (mid panels) and humans (right panels). Observations are indicated by symbols, population predictions by lines. The solid line in the lower panels represents the identity line. In dogs: ○ (grey) and ^_____ are pre–dose values; △ (yellow) and ^{_ _ _ _} are placebo; + (greenish) and - - - - 3 mg/kg, x(slate grey) and - ^_ - ^_ 10 mg; ◊(light blue) and ^{__ __ __} 30 mg. In monkeys: × (black)/solid line are placebo; * (red)/dashed line 90 mg/kg. In humans: dashed line and symbols depict effects of a 400 mg dose

Fig. 4

Model performance and predicted QT profiles after administration of placebo and different doses of NCE05 to cynomolgus monkeys (left panels) and healthy subjects (right panels). Observations are indicated by symbols, population predictions by lines. The solid line in the lower panels represents the identity line. In monkeys: * (green)/solid line are placebo; △ (red)/dashed line 25 mg/kg; + (blue)/dotted line 40 mg/kg. In humans: * (green)/solid line 1 mg; △ (red)/dashed line 4 mg; + (blue)/dotted line 14 mg; × (black)/dashed-dotted line 30 mg

Fig. 5

(Upper panel) Observed vs model predicted QT interval for moxifloxacin (left) and NCE05 (right). Black circles, grey crosses and slate grey triangles represent the experimental observations in dogs, cynomolgus monkeys and humans, respectively. (Lower panel) Comparison of the risk of drug-induced QTc prolongation across species. Dotted line: calculated values for conscious dogs; dashed grey line: calculated values for monkeys; solid, black line: calculated values for humans. The thick black line indicates the observed Cmax range of the clinical study