JAAC Heuberger (1), Z. Guan (1), O. Oyetayo (1), L. Klumpers (1), J. M. A. van Gerven (1), A. F. Cohen (1), J. Freijer (1)
(1) Centre for Human Drug Research, Leiden, the Netherlands
Objectives: To evaluate the Prior subroutine in NONMEM® by comparing model results after using a Prior versus the standard method of pooling datasets. Both methods were used to develop a multi-compartment pharmacokinetic (PK) model for THC using clinical data, incorporating three major administration routes (oral, pulmonary, IV) and the long term PK of the cannabinoid tetrahydrocannabinol (THC).
Methods: For the PK modeling NONMEM® 7.2[1] was used. THC PK data were obtained from two inhalation[2,3], one oral[4] and a cross-over inhalation and IV administration studies[5]. The pooled data contains 910 THC plasma concentrations from 84 healthy volunteers. Using the traditional method, four datasets were pooled and modelled simultaneously. Secondly, three of the datasets were pooled and modeled. Results of this latter model were subsequently applied in the Prior subroutine to fit the fourth dataset, one of the inhalation studies. The results were then compared to the results of the traditional method.
Results: Both developed models accurately describe and predict THC plasma concentrations up to 48 hours after oral, pulmonary or intravenous dosing. Results for the standard model approach (with results of the Prior approach) show that THC has a fast initial and intermediate half-life, a relatively long apparent terminal half-life of 21 h (18 h), with a clearance of 38.9 L/h (38.1 L/h), systemic bioavailability 28.4 (26.3) % or 22.4 (25.6) % after inhalation with and without nose clip respectively. Remaining parameter estimates were also nearly identical, with a central volume of distribution of 6.17 (5.70) L, and rate constants to and from peripheral compartments of 1.3 (1.15), 0.04 (0.045), 4.10 (4.26) and 1.04 (1.16) /h. The Ka (1.99 /h), lag time (0.198 h) and oral bioavailability were identical, as the data fitted with the prior subroutine contained no oral input. Moreover, runtimes were reduced 17 times using the Prior subroutine.
Conclusions: Both model approaches describe the PK of THC accurately, with consistent parameter estimates. They can be applied to predict concentration-time profiles of THC after different dosing regimens (e.g. accumulation) for the mostly used administration routes. Also, the models can improve decision making in future clinical trials of (novel) cannabinoids. Model comparison showed that the Prior function in NONMEM® can be applied to fit prior results with new additional data rather than fitting a new model to an increased dataset, which will hence speed up model development.
References:
[1] Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA
[2] Kleinloog D, Liem-Moolenaar M, Jacobs G, Klaassen E, de KM, Hijman R et al. Does olanzapine inhibit the psychomimetic effects of Delta(9)-tetrahydrocannabinol? J Psychopharmacol 2012; 26:1307-1316.
[3] Klumpers LE, Cole DM, Khalili-Mahani N, Soeter RP, Te Beek ET, Rombouts SA et al. Manipulating brain connectivity with delta(9)-tetrahydrocannabinol: a pharmacological resting state FMRI study. Neuroimage 2012; 63:1701-1711.
[4] Klumpers LE, Beumer TL, van Hasselt JG, Lipplaa A, Karger LB, Kleinloog HD et al. Novel Delta(9) -tetrahydrocannabinol formulation Namisol(R) has beneficial pharmacokinetics and promising pharmacodynamic effects. Br J Clin Pharmacol 2011. [5] Ohlsson A, Lindgren JE, Wahlen A, Agurell S, Hollister LE, Gillespie HK. Single dose kinetics of deuterium labelled delta 1-tetrahydrocannabinol in heavy and light cannabis users. Biomed Mass Spectrom 1982; 9:6-10.
Reference: PAGE 23 () Abstr 3031 [www.page-meeting.org/?abstract=3031]
Poster: Methodology - New Modelling Approaches