Clinical pharmacokinetics of intravenous ethanol : Relationship between the ethanol space and total body water

Sammanfattning: Introduction: Total body water (TBW) is an important parameter in pathological states where the normal regulation of fluid balance is impaired (e.g., during critical illness, congestive heart failure, bum injury and renal insufficiency). Dilution of water isotopes, such as deuterium oxide (D2O), is considered die gold standard method for measuring TBW in humans. However this procedure requires skilled staff, expensive equipment and the results are seldom available in a timely fashion, which makes it less suitable for clinical applications. Ethanol is completely miscible with water and is thought to distribute into the TBW. The ethanol volume of distribution at steady state (Vss) can be estimated by pharmacokinetic analysis of the concentration-time profile. Ethanol can be measured in expired breath with high precision and this non- invasive method of analysis could provide an attractive alternative with the prospect of bedside monitoring of Vss within 4-6 hours. The aim. of this thesis was to develop an appropriate pharmacokinetic model for intravenous ethanol administration and to identify experimental factors that impact on the results. With this background, we compared Vss. determined by ethanol dilution with TBW determined by D20 dilution. The complicated absorption kinetics of ethanol (e.g. variable gastric emptying and first-pass metabolism) was avoided by use of the intravenous route of administration. Material and Methods: Forty-six healthy volunteers (20 women and 26 men) received intravenous infusions of ethanol (0.4- 0.6 g/kg body weight) in 15-60 minutes. The concentration of ethanol was measured in end-expired breath by infrared spectrometry and in blood and urine by headspace gas chromatography. Specimens of blood and breath were obtained at 5-15 min intervals for 3-6 hours post-dosing. The concentration of D20 in plasmawater was measured with isotope ratio mass spectrometry. Subjective feelings of inebriation were measured with a visual analogue scale. A number of pharmacokinetic models were developed and evaluated in the course of this thesis. The effect of eating a meal on ethanol kinetics was investigated in crossover studies by giving the same dose of ethanol in fed and fasted states. The magnitude and time-course of arterio-venous differences in ethanol concentration were determined to assess the importance of local peripheral vasodilatation and vasoconstriction on results. This was achieved by warming the sampling hand in a heating box or cooling the hand in cold water during ethanol infusion experiments. The impact of ~ling site on pharmacokinetic parameters of ethanol was investigated by simultaneous sampling in expired breath, arterial blood and venous blood. Inter- and intra-individual variations in precision of the estimates of Vss. and TBW were studied by making repeat infusions of ethanol a few days apart, and use of analysis of variance. Results and discussion: The two-compartment model with parallel Michaelis-Menten kinetics and first-order rend elimination provided an excellent prediction of the entire concentration-time profiles of ethanol and proved to be theoretically superior to the other kinetic models tested. A good reproducibility was obtained for all pharmacokinetic parameters, especially for Vss and the maximal metabolic rate (Vmax). Eating a meal increased the rate of ethanol metabolism by 30-60% and this confounding factor needs to be considered in clinical studies when Vss. for ethanol is estimated. Peripheral vasoconstriction caused a lowering of the concentrations of ethanol in venous blood, which also impacts on the pharmacokinetic parameters of ethanol. Measuring ~1 in breath agreed more closely with arterial blood concentrations than with venous blood. The pharmacokinetic parameters of ethanol could be measured with equally high precision in expired breath and venous blood. An ethanol dose of 0.4 g/kg body weight infused in 15 minutes led to unacceptable inebriation in some volunteer subjects. The precision of estimating Vss by ethanol dilution was about the same as for measuring TBW by D20 dilution (SD 0.8-.1.1 litres). However, me observed a systematic bias of -13% between Vss and TBW (D20). A likely explanation for this finding might be that water in the body, at least in part, is structured in such a way that ethanol is prevented from complete equilibration. Conclusions: A two-compartment model with parallel Michaelis-Menten and first-order renal elimination gave an excellent fit to the concentration-time profile of ethanol after intravenous ~on. Feeding state and peripheral circulation affected the estimated model parameters. Sampling and analysis of breath provided results with the same high precision as venous blood for estimating ethanol Vss. The non-invasive nature of breath sampling makes it more suitable for clinical purposes. The systematic bias of-13% between the Vss and TBW by D20 dilution suggests that ethanol does not distribute uniformly into the TBW. The reason for this discrepancy remains to be established.