The Influence of Body Composition on Ethanol Pharmacokinetics using a Rate Dependent Extraction Model
Nick Holford (1), Yu Jiang (2), Daryl J. Murry (2), Timothy L. Brown(3), Gary Milavetz (2)
(1) Department of Pharmacology and Clinical Pharmacology, University of Auckland, New Zealand (2) College of Pharmacy, University of Iowa, USA (3) National Advanced Driving Simulator, Center for Computer Aided Design, University of Iowa, USA
Objectives: 1) to apply a model capable of linking first pass hepatic extraction with ethanol absorption rate in order to identify hepatic mixed-order, first-order and non-hepatic first order elimination processes for ethanol.
2) to explore the effect of body composition on ethanol disposition parameters.
Methods: 108 subjects were dosed orally to achieve a target peak blood ethanol concentration of 650 mg/L and 1150 mg/L using a randomized, crossover design. A total of 6025 breath samples were measured using an Alco-Sensor IV breathalyzer. Breath ethanol was converted to equivalent blood concentrations for pharmacokinetic analysis. A semi-mechanistic rate dependent extraction model [1] with zero-order input to the gut with subsequent first order absorption was used to describe the data. Between subject variability (BSV) and between occasion variability (BOV) were tested on all parameters. Portal blood flow (Qpv) [2] was predicted from fat free mass (FFM) which was used to predict hepatic vein concentration (Chv) and hepatic intrinsic clearance (CLI). Portal vein concentration (Cpv) was predicted from Qpv and ethanol absorption rate into the portal vein.
Chv=Cpv*Qpv/(Qpv+CLI)
CLI=Vmax/(Km+Chv) + CLFO
Allometric scaling using different size metrics was used for volume of distribution (V), the maximum metabolizing capacity (Vmax), hepatic first order clearance (CLFO), and non-hepatic first order clearance (CLNH). The size metrics were total body weight (TBW), FFM [3] and normal fat mass (NFM) [4].
NFM=FFM+Ffat*(TBW-FFM)
Censored observations were included by using Beal’s M3 method [5]. Data were analyzed using NONMEM 7.3.0.
Results: The bootstrap average values for the population parameters are shown in the table. Model selection guided by objective function value (OFV) suggested the existence of CLNH but the 95% bootstrap confidence interval included zero for both CLFO and CLNH when they were both included in the model.
|
Parameter |
Units |
Population Estimate |
BSV |
BOV |
|
Volume |
L/70kg NFM |
38.6 |
0.091 |
0.149 |
|
Vmax |
g/h/70kg TBW |
15.8 |
0.258 |
0.309 |
|
Km |
mg/L blood |
62.5 |
1.22 |
0.438 |
|
CL non-hepatic |
L/h/70kg TBW |
0.13 |
0.166 |
- |
|
First-order from gut |
1/h |
8.83 |
- |
1.21 |
|
Zero-order input to gut |
h |
0.301 |
- |
0.492 |
|
Ffat for volume |
- |
0.458 |
- |
- |
BSV and BOV calculated from sqrt(NONMEM omega)
Conclusions: A rate dependent extraction model improves model fitting compared with a simple mixed order model. Normal fat mass was determined to be the best size descriptor for V and total body weight for Vmax.
References:
[1] Holford, N.H.G. Complex PK/PD models--an alcoholic experience. Int J Clin Pharmacol Ther, 1997. 35(10): p. 465-8.
[2] Carlisle KM, Halliwell M, Read AE, Wells PN. Estimation of total hepatic blood flow by duplex ultrasound. Gut. 1992;33(1):92-7.
[3] Janmahasatian, S., Duffull, S. B., Ash, S., Ward, L. C., Byrne, N. M., & Green, B. (2005). Quantification of lean bodyweight. Clin Pharmacokinet. 2005; 44(10); 1051-1065.
[4] Anderson BJ, Holford NHG. Mechanistic basis of using body size and maturation to predict clearance in humans. Drug Metab Pharmacokinet. 2009;24(1):25-36.
[5] Beal SL. Ways to fit a PK model with some data below the quantification limit. Journal of Pharmacokinetics & Pharmacodynamics. 2001;28(5):481-504.
