V Gotta (1), F Cools(2), K van Ammel(2), D J Gallacher(2), S A G Visser(3), F Sannajust(4), P Morissette(4), M Danhof(1), P H van der Graaf(1)
(1) Systems Pharmacology, Leiden Academic Center of Drug Research (LACDR), Leiden University, Leiden, The Netherlands. (2) Global Safety Pharmacology, Janssen Research & Development, Janssen Pharmaceutica NV, Beerse, Belgium. (3) Quantitative Pharmacology and Pharmacometrics / Merck Research Laboratories, Merck & Co., Inc., Upper Gwynedd, PA, USA (4) SALAR-Safety and Exploratory Pharmacology Department / Merck Research Laboratories, Merck & Co., Inc., Westpoint, PA, USA
Objectives: Drug-induced QTc-interval prolongation (?QTc) is assessed in preclinical cardiovascular safety studies as a surrogate for pro-arrhythmic risk in human. These studies in conscious telemetered dogs (4-8 animals) can be based on varying designs (e.g. dose range, number of doses, route of administration, study duration, PK/PD sampling in same animals or a satellite group, no PK). We aimed to assess the quantitative consistency of pharmacodynamic (PD) relationships derived from such studies (inter-study variability, ISV), and the preclinical-clinical correlation of PD effects.
Methods: A total of 14 studies (moxifloxacin: N=6 studies/n=32 dogs, dofetilide: N=6/n=27, and sotalol: N=2/n=10) were analyzed using population PK/PD modeling – first separately, then in a meta-analysis by pooling data from each compound. Typical PD relationships were derived with 95%CIs from each model. Preclinical PD meta-predictions were used as a reference to evaluate ISV and translational relationships with clinical effects (derived from a systematic literature review). For moxifloxacin, ISV was additionally assessed in a hierarchical random-effects model[1].
Results: The PDs of all 3 drugs was best characterized by sigmoidal Emax-models (meta-models, all with ?QTcmax≈50ms or 20% from baseline). The 95%CIs of 13/14 (93%) study-derived PD predictions comprised the meta-prediction, despite varying structural PK/PD models. ISV of ?QTc-predictions at upper therapeutic exposure was ±30% (range: ±1-69%). Predicted ?QTc increase within unbound therapeutic range was 4–12 ms (moxifloxacin, 2.9–5.6 μM, EC50=11.6 μM), 4–18 ms (dofetilide, 0.4–2 nM, EC50=4.2 nM), and 14-19 ms (sotalol, 3.7–11 μM, EC50=10.1 μM), and was overlapping with clinical %?QTc from baseline. Including an ISV-level in the moxifloxacin meta-model decreased PD BSV by 10-26%, BSV on the hill coefficient decreased most. Resulting ISV (24-39%, RSE>100%) did not exceed BSV (28-37%).
Conclusions: This study provides a first quantitative assessment of ISV in preclinical ?QTc evaluations. Results suggest that consistent predictions can be obtained from highly varying studies by systematic PK/PD analysis, i.e. suitable for translational purpose. Furthermore, a 10% ?QTc-effect[2] seems to correspond already to a half-maximal effect in dog (and to a ≥35ms effect in human), suggesting that this is an unsatisfying study sensitivity target. Utilizing PK/PD analysis can improve the detection of small preclinical ?QTc [3].
References:
[1] Laporte-Simitsidis, S. et al. Inter-study variability in population pharmacokinetic meta-analysis: when and how to estimate it? J. Pharm. Sci. (2000) 89, 155–67.
[2] Sasaki, H., Shimizu, N., Suganami, H. & Yamamoto, K. QT PRODACT: inter-facility variability in electrocardiographic and hemodynamic parameters in conscious dogs and monkeys. J. Pharmacol. Sci. (2005) 99, 513–22.
[3] Gotta, V. et al. Sensitivity of pharmacokinetic-pharmacodynamic analysis for detecting small magnitudes of QTc prolongation in preclinical safety testing. J. Pharmacol. Toxicol. Methods (2014) 72C, 1–10.
Reference: PAGE 24 (2015) Abstr 3398 [www.page-meeting.org/?abstract=3398]
Poster: Drug/Disease modeling - Safety