Presentation Title

AN IMPROVED NONLINEAR MIXED EFFECTS PHARMACOKINETIC MODELOF INTRAVENOUS D9TETRAHYDROCANNABINOL (THC) AND METABOLITES IN VOLUNTEERS

Location

Atrium

Format

Poster

Start Date

14-2-2014 12:00 AM

Abstract

Objective. To develop and evaluate a nonlinear mixed effect model of d9-THC and its two major metabolites. Background. Marijuana use is increasingly becoming more prevalent due to the rise of medical use. A nonlinear mixed effects model to predict plasma concentrations of the major active constituents of marijauna, D9-THC and its two metabolites does not exist. This model could be used in health care and by law enforcement. Methods. Data from 25 subjects administered d9-THC intravenously over 5 minutes and sampled from arterial line in serial fashion from time of dosing to 48 hours was available. Of the 975 available plasma concentrations, analyzed with LC/MS/MS API 4000, 75 were BLQ and not used in the model. Nonlinear mixed effect modeling using Phoenix® NLME 1.2 was applied to the data to determine the typical values of a 3 compartment pharmacokinetic model for the parent THC that provides input to a second 3 compartment model for the first and active metabolite THC-OH, which in turn provides input to the 2 compartment model for the second and inactive metabolite THC-COOH. The model was parameterized using CL and V, residual error was log-additive, the omega matrix was diagonal, initial estimates were taken from NCA analysis of the data. The FOCE-extended least squares algorithm performed best with this data set, standard errors were determined using the central difference Hessian method. The model was optimized in several steps, using traditional assessment techniques: -2LL, AIC, BIC, Conditional WRES Vs pred, ipred, time and observed dose/volume. The model employs actual body weight as a covariate. The “final” model was evaluated using the predictive check option in Phoenix. Results. The data was successfully fit to a 3 stage metabolite model. The model parameters follow: tvV thc 6.05939 L tvV thc2 29.7352 L tvCLd12 thc 51.5397 L tvV thc3 325.58 L tvCLd13 thc 19.0989 L tvV thc-oh 88.6359 L tvCL metabolic thc49.9912L/h tvV thc-cooh 6.51545 L tvCL metabolic thc-oh153.778 L/h tvV thc-oh2 14.4936 L tvCLd12 thcoh 0.725483L/h tvV thc-cooh2 405.939 L tvCLd12 thc-cooh 25.7877 L/h tvCL renal thc-cooh 7.86533 L/h tvV thcoh3 371.384 L/h tvCLd13 thc-oh 109.992 L/h Conclusion. This model could be used to predict THC, THC-OH or THCCOOH concentrations over time. Grants. none

This document is currently not available here.

COinS
 
Feb 14th, 12:00 AM

AN IMPROVED NONLINEAR MIXED EFFECTS PHARMACOKINETIC MODELOF INTRAVENOUS D9TETRAHYDROCANNABINOL (THC) AND METABOLITES IN VOLUNTEERS

Atrium

Objective. To develop and evaluate a nonlinear mixed effect model of d9-THC and its two major metabolites. Background. Marijuana use is increasingly becoming more prevalent due to the rise of medical use. A nonlinear mixed effects model to predict plasma concentrations of the major active constituents of marijauna, D9-THC and its two metabolites does not exist. This model could be used in health care and by law enforcement. Methods. Data from 25 subjects administered d9-THC intravenously over 5 minutes and sampled from arterial line in serial fashion from time of dosing to 48 hours was available. Of the 975 available plasma concentrations, analyzed with LC/MS/MS API 4000, 75 were BLQ and not used in the model. Nonlinear mixed effect modeling using Phoenix® NLME 1.2 was applied to the data to determine the typical values of a 3 compartment pharmacokinetic model for the parent THC that provides input to a second 3 compartment model for the first and active metabolite THC-OH, which in turn provides input to the 2 compartment model for the second and inactive metabolite THC-COOH. The model was parameterized using CL and V, residual error was log-additive, the omega matrix was diagonal, initial estimates were taken from NCA analysis of the data. The FOCE-extended least squares algorithm performed best with this data set, standard errors were determined using the central difference Hessian method. The model was optimized in several steps, using traditional assessment techniques: -2LL, AIC, BIC, Conditional WRES Vs pred, ipred, time and observed dose/volume. The model employs actual body weight as a covariate. The “final” model was evaluated using the predictive check option in Phoenix. Results. The data was successfully fit to a 3 stage metabolite model. The model parameters follow: tvV thc 6.05939 L tvV thc2 29.7352 L tvCLd12 thc 51.5397 L tvV thc3 325.58 L tvCLd13 thc 19.0989 L tvV thc-oh 88.6359 L tvCL metabolic thc49.9912L/h tvV thc-cooh 6.51545 L tvCL metabolic thc-oh153.778 L/h tvV thc-oh2 14.4936 L tvCLd12 thcoh 0.725483L/h tvV thc-cooh2 405.939 L tvCLd12 thc-cooh 25.7877 L/h tvCL renal thc-cooh 7.86533 L/h tvV thcoh3 371.384 L/h tvCLd13 thc-oh 109.992 L/h Conclusion. This model could be used to predict THC, THC-OH or THCCOOH concentrations over time. Grants. none