R.H. Rose (1), M. Chetty (1), M. Jamei (1), A. Rostami-Hodjegan (1,2)
(1) Simcyp Ltd, Sheffield, UK (2) University of Manchester, Manchester, UK
Objectives: The operational model of agonism [1] is a mechanism-based pharmacodynamic (MBPD) model that incorporates drug specific parameters, available from in vitro experiments, and system specific parameters that can be estimated using in vivo data. In principle, once a MBPD model has been established for a drug and linked to a mechanistic PBPK model, the PD response for a drug that shares the same mechanism of action can be predicted by changing only the drug specific parameters. The recent success in application of systems approach in predicting pharmacokinetics has led many to believe a similar strategy for prediction of PD aspects should be adopted [2]. This study aimed to establish a MBPD model to describe the hypnotic effects of triazolam, as measured by change in beta-EEG amplitude, and to use this to predict the hypnotic response to zolpidem since the hypnotic effects of both drugs are mediated via α1 subunit containing GABAA receptors.
Methods: Simulations of PK and PD were performed using the default triazolam and zolpidem compound files within the Simcyp Simulator (v11.1). In vitro KD values for triazolam and zolpidem and their relative efficacy were identified from review of the literature. The Simcyp Parameter Estimation module was used to estimate the system dependent parameters of the operational model using published concentration-response data from healthy volunteer studies of triazolam [3-6]. The quality of the parameter estimates was tested by the ability of the operational model to predict the PD response following interaction with ketoconazole [5,7] before predicting the response to zolpidem.
Results: The operational model developed using the parameter estimates predicted the PD response to 0.125 and 0.25mg triazolam with and without ketoconazole DDI reasonably well. The maximal response (Rmax) to zolpidem was predicted well by changing only the PBPK model input to the operational model and the in vitro KD. The simulated/observed ratios for Rmax were 0.86-0.91 for 5mg and 10mg zolpidem [3,5,8]. However, the duration of response was overestimated.
Conclusions: The MBPD model developed for triazolam was reasonably successful in predicting the maximal EEG response to zolpidem. Availability of the MBPD models within a user-friendly environment may facilitate wider use of this approach in the prediction of the dose of a drug candidate that produces equivalent clinical efficacy to a well-studied drug with a similar mechanism of action.
References:
[1] Black and Leff (1983) Proc R Soc Lond B Biol Sci 220: 141-62
[2] Atkinson and Lyster (2010) Clin Pharmacol Ther 88: 3-6
[3] Greenblatt et al., (2000) J Pharmacol Exp Ther 293: 435-43
[4] Greenblatt et al., (1994) Clin Pharmacol Ther 56: 100-11
[5] Greenblatt et al., (1998) Clin Pharmacol Ther 64: 661-71
[6] Greenblatt et al., (2004) Clin Pharmacol Ther 76: 467-79
[7] von Moltke et al., (1996) J Pharmacol Exp Ther 276: 370-9
[8] Greenblatt et al., (2006) J Clin Pharmacol 46: 1469-80
Reference: PAGE 21 (2012) Abstr 2412 [www.page-meeting.org/?abstract=2412]
Poster: Other Modelling Applications