Biomedical subjects
Publications and source records attributed to M Gibaldi.
A digital computer study using simulated data containing random error indicates no difference in the precision and accuracy of parameter estimation when the data are fitted to several different equations.
The purpose of this investigation was to study, by digital computer stimulation, the accumulation kinetics of drugs which exhibit concentration-dependent binding to tissues and either linear (constant free fraction) or concentration-dependent (increasing free action with increasing drug concentration) binding to plasma proteins. It was assumed that elimination rate is proportional to free drug concentration in plasma and that there occurs instantaneous equilibration of drug between vascular and nonvascular spaces. Nonlinear binding can yield, under certain conditions, apparently biexponential plasma concentration-time curves which may be misinterpreted as being representative of a linear and biexponential system. Such misinterpretation would cause the following errors in the prediction of drug accumulation and elimination kinetics during and after constant-rate infusion: (a) the time required to reach steady state may be overestimated, and (b) the prominence of the apparent distribution phase after cessation of infusion may be underestimated. Drugs with linear and nonlinear plasma protein binding characteristics differ with respect to the relaionship between infusion rate and steady-state concentration. This relationship is linear when plasma protein binding is linear. Steady-state concentration increases less than proportionally with increasing infusion rate if plasma protein binding is drug concentration dependent.
Linear relationships were observed between the log of total body clearance (C1B) and simulated serum concentrations (Cmin) six hours after a test dose of chloramphenicol, as well as between C1B and the reciprocal of Cmin. Correlation coefficients for these relationships were 0.988 and 0.977, respectively. Clearance estimates obtained from a single serum sample following a test dose of a drug may prove to be a useful method of predicting dosage requirements for individual patients.
A pharmacokinetic model that incorporates linear binding of drug to plasma proteins and tissue indicates the same relationship between apparent volume of distribution and drug binding as that proposed by Gillette (1971) based on a simple distribution model. Apparent volume of distribution (V) is directly proportional to free fraction of drug in plasma (fp) and indirectly proportional to free fraction of drug in tissue (fT). In the case of a constant fT, a plot of V versus fp will be linear with an intercept equal to plasma volume (Vp). If fT changes with fp, an apparently linear plot may result but the intercept will exceed Vp. An approach to the calculation of fT, a composite binding parameter, is presented and illustrated by comparing the tissue binding of tolbutamide in patients during acute viral hepatitis and upon recovery.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
The effect of chloramphenicol therapy (48 mg/kg/day) on the serum concentrations of phenytoin and phenobarbital was studied in a patient previously stabilized on anticonvulsant medications. Phenytoin, 12 mg/kg/day, and phenobarbital, 5 mg/kg/day resulted in serum concentrations averaging 10.8 microgram/ml before and 30.5 microgram/ml, after chloramphenicol therapy. A reduction in dose of both phenytoin and phenobarbital was required to minimize adverse effects during the course of chloramphenicol therapy. An average daily dose of phenytoin of 9.1 mg/kg resulted in an average serum concentration of 17.8 microgram/ml. A daily dose of phenobarbital of 4.0 mg/kg resulted in an average serum concentration of 37.1 microgram/ml. These changes indicate 50.5% and 40.4% decreases in clearance of phenytoin and phenobarbital. Multiple-dose nonlinear regression analysis of phenytoin and phenobarbital serum concentration data obtained during chloramphenicol therapy indicated a 62.5% and a 29.5% decrease in clearance. Subsequent serum concentration monitoring demonstrated a similar reduction in phenobarbital clearance when chloramphenicol was added to phenobarbital alone.
The pharmacokinetics of hydroflumethiazide after oral administration of a 100-mg dose to 12 volunteers were evaluated to determine whether absorption of the drug is best described by zero-order or first-order absorption. Comparison of the two absorption models was based on three criteria: (1) correlation coefficients; (2) standard deviations of the parameter estimates; and (3) visual fits. Evaluation of the individual results in light of each criterion showed that the zero-order absorption model is the more appropriate one for describing the hydroflumethiazide data.
MULTDOS, a computer method to curve fit data obtained on multiple dosing, was used with either the 1969 or 1974 version of the NONLIN program to compare the pharmacokinetic parameters of ethosuximide during repetitive administration of single or divided daily doses. Elimination rate constants, excretion rate constants, and apparent volumes of distribution were similar between the two dosing regimens and essentially identical between the two nonlinear regression programs.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
A digital computer curve-fitting method, designed to estimate pharmacokinetic model constants by utilizing all drug concentration-time data collected during repetitive dosing studies, was applied to data manifesting systematic dose-to-dose variability in one or another of the pharmacokinetic parameters. The method accurately determined dose-to-dose changes in absorption or elimination rate constants or in the apparent volume of distribution, and it would be useful for detecting phenomena such as self-induction and self-inhibition that may occur during multiple-dose administration. The method can also be used to analyze multiple-dose data of drugs exhibiting capacity-limited elimination and to obtain estimates of the Michaelis-Menten parameters.
A plasma flow rate-limited pharmacokinetic model was developed to describe the distribution of digoxin to the heart, liver, kidneys, skeletal muscle, and GI tract in the rat. The model also provides for renal, hepatic (metabolic and biliary), and GI clearance as well as for biliary and GI secretion and GI reabsorption of digoxin. Predicted concentrations of digoxin in the heart, liver, skeletal muscle, and plasma were consistent with experimental observations in conscious rats after an intravenous dose. The model was extended to describe digoxin concentrations in the plasma of bile duct-ligated rats and ureter-ligated rats, simply by modifying appropriate clearance parameters. Excellent agreement was obtained between predicted and observed urinary excretion rates of digoxin for 12 hr after in intravenous dose to normal and bile duct-ligated rats.
The antisecretory (determined from salivary flow rates) and antimotility (determined from riboflavin absorption) effects of usually recommended doses of propantheline, hexocyclium, and isopropamide were compared in four adult volunteers. Both propantheline and hexocyclium significantly decreased salivary flow and increased riboflavin absorption. Although the usual dose of propantheline was about twice as effective as the usual dose of hexocyclium in suppressing salivary flow, these doses produced comparable effects on riboflavin absorption. Isopropamide had little or no effect on either the salivary flow rate or riboflavin absorption. Propantheline and hexocyclium elicited little effect on salivary flow when administered after a meal. Prolonged-release dosage forms of these drugs produced effects comparable to those produced by much smaller doses in conventional tablets and gave no indication of providing prolonged anticholinergic effects.
Concomitant oral administration of salicylamide (200 mg/kg) and 3H-terbutaline (1 mg/kg) to rats with ligated bile ducts decreased absorption of terbutaline from the gut from 73 to 56% as measured by urinary excretion of radioactivity in 48 hr. No increase in the fraction of terbutaline excreted unchanged was observed, suggesting that salicylamide does not substantially inhibit the conjugation of terbutaline with glucuronic acid. An increase in the fraction of terbutaline excreted unchanged observed in normal animals may result from enhanced excretion of terbutaline glucuronide into bile rather than from inhibition of conjugation.
A physiologically based pharmacokinetic model for digoxin disposition developed in the rat was modified to account for the interspecies differences in tissue-to-plasma digoxin concentration ratios and applied to the dog. The model provided a quantitative assessment of the time course of digoxin concentrations in dog plasma, various tissues, and urine. It also predicted the effect of renal failure on digoxin pharmacokinetics in the dog. An attempt to scale the dog model to humans by simply considering differences in organ volumes, organ flow rates, and digoxin clearances was partially successful. Good predictions of plasma digoxin concentration and urinary digoxin excretion after a single dose and of steady-state plasma, heart, and skeletal muscle digoxin concentrations were obtained. However, the model predicted considerably higher kidney digoxin concentrations than are actually found. Although the model adequately characterized the time course of digoxin concentrations in patients with moderate renal impairment, it provided a relatively poor fit to that observed in anuric patients.