Daniel M. Jonker (1), Derek Leishman (2), Rob Wallis (2), Peter E. Milligan (2) and E. Niclas Jonsson
(1) Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Uppsala University, Sweden, (2) Pfizer Global Research & Development, Sandwich Laboratories, Sandwich, Kent, UK
Objectives: Drug-induced QT prolongation is an important biomarker for assessing the risk of heart arrhythmias and reflects blockade of the hERG potassium channel. The aim of the present work is to be able to predict the magnitude and variability of QT prolongation in patient populations from in vitro and preclinical studies. Dofetilide is a pure class III antiarrhythmic drug and as such an ideal starting point. In the present study, an integrated population PKPD model was developed for the action of dofetilide at the hERG potassium channel in vitro and in man.
Methods: A joint analysis was performed of 5 clinical studies, including 80 healthy volunteers and 17 patients with ischaemic heart disease. The affinity and activity of dofetilide was determined in recombinant cell cultures expressing the hERG potassium channel.
The analysis was performed in NONMEM in two steps. In the first step, the PK model was developed and covariates were identified in a semi-automated fashion. In the second step, the individually estimated dofetilide concentrations were used to drive the PD model. By including the in vitro dofetilide affinity in the operational model of pharmacological agonism, an estimate was obtained for the efficiency of the transduction from ion channel binding into the QT effect.
Results: A three compartment PK model with first-order absorption after oral administration characterized the time course of dofetilide concentrations well. Three covariates were identified by using stepwise forward selection followed by backward elimination at the 1% significance level: age for clearance, body weight for peripheral volume and body weight for clearance. Fridericia-corrected QT values were adequately described with the operational model including an effect compartment. Based on the in vitro affinity of 4.79 ng/ml, the transducer ratio was estimated at a value of 5, indicating the presence of a modest ion channel reserve. Upon chronic administration of dofetilide, tolerance was found to develop until a maximum was reached after four days.
Conclusion: In this work, the magnitude and variability of dofetilide-induced QT prolongation was estimated with a single mechanism-based PKPD model. This model will be extended to other compounds to explore its capability to predict the degree and variability of QT prolongation in a heterogeneous patient population from preclinical drug characteristics.
Reference: PAGE 13 (2004) Abstr 502 [www.page-meeting.org/?abstract=502]
Poster: poster