M. Chetty (1), T. Cain (1), R.H. Rose (1), M. Jamei (1), A. Rostami Hodjegan (1,2)
(1) Simcyp Ltd, Sheffield; UK (2) University of Manchester, Manchester, UK
Objectives: Induction of CYP1A2 by cigarette smoke is well established. A recent report suggests the possibility of predicting consequential pharmacokinetic (PK) differences using in vitro-in vivo extrapolation (IVIVE) combined with physiologically-based PK (PBPK) [1]. Although reports on PK differences between non-smokers and smokers are abundant, studies on the associated pharmacodynamic (PD) responses are infrequent. This study aimed to predict differences in pharmacological response resulting from PK differences in passive smokers and heavy smokers, using theophylline as an example of a CYP1A2 substrate, with forced expiratory volume (FEV1) as a marker of response.
Methods: The Simcyp Simulator (V11.1) was used to simulate the PK/PD profiles of ten trials with ten subjects using the study design and PD parameters (Emax model with an effect compartment) from a PK/PD study of Caucasian non-smokers with moderate respiratory dysfunction [2]. A population of heavy smokers (> 20 cigarettes/day) was modelled with an increased CYP1A2 abundance of 94 pmol P450/mg protein (CV 43%) [1].Similarly, a population of passive smokers was modelled using data on the effects of passive smoking on theophylline clearance [3]. The PK of theophylline in heavy and passive smokers were simulated and compared with clinical data [3, 4]. Based on the PD model in non-smokers and PK differences in the three groups, PD responses in individuals exposed to cigarette smoke were simulated. Models did not consider the direct effect of cigarette smoke on FEV1.
Results: Both the PK/PD models in non-smokers and the simulated PK profiles in heavy smokers and passive smokers predicted the clinical data adequately. Clearance in heavy smokers and passive smokers was on average 1.72 and 1.44 fold higher than in non-smokers, respectively. The area under the concentration-response curve corrected for baseline (AUCRcorr) in both heavy smokers and passive smokers was lower than in non-smokers (by on average 20% and 12%, respectively).
Conclusions: The Simcyp PBPK/PD model was able to differentiate the responses due to PK differences in smokers and non-smokers. A shortcoming of this PD model is the lack of baseline data to account for direct effects of cigarette smoke on FEV1. However, similar models using alternate PD markers can be developed and used to predict dosage adjustments in candidate drug molecules that are metabolised predominantly by CYP1A2, with a known PD profile in non-smokers.
References:
[1] Plowchalk DR and Roland-Yeo K. (2012) Eur J Clin Pharmac (online)
[2] Flores-Murrieta FJ et. al., (1999) Proc West Pharmacol Soc 42: 3
[3] Matsunga S et. al. (1989) Clin Pharmac Ther 46: 399
[4] Gardner MJ et. al. (1983) Br J Clin Pharmac 16: 271
Reference: PAGE 21 (2012) Abstr 2447 [www.page-meeting.org/?abstract=2447]
Poster: Other Modelling Applications