Cytochrome P450 2C9 phenotyping using low-dose tolbutamide

OBJECTIVES

The hypoglycaemic drug tolbutamide is used for assessment of CYP2C9 activity in vivo. However, therapeutically active doses of 500 mg bear the risk of hypoglycaemia, and a tolbutamide-derived parameter based on a single plasma or urine concentration reflecting CYP2C9 activity accurately is lacking.

METHODS

We examined tolbutamide and its metabolites 4'-hydroxy-tolbutamide and carboxytolbutamide in plasma and urine of 26 healthy, male volunteers up to 24 h after intake of 125 mg tolbutamide using liquid chromatography-tandem mass spectrometry. CYP2C9 genotypes were determined by sequencing of exons 3 and 7. Raw plasma and urine data were compared with pharmacokinetic parameters, CYP2C9 genotypes, and data from a study in 23 volunteers with all six CYP2C9*1-*3 combinations who received 500 mg tolbutamide.

RESULTS

Plasma clearance and tolbutamide plasma concentrations 24 h after drug intake reflected the genotypes: 0.85 l/h and 1.70 microg/ml (95% confidence interval, CI, 0.80-0.89 l/h and 1.50-1.90 microg/ml) for CYP2C9*1 homozygotes (n=15), 0.77 l/h and 2.14 microg/ml (95%CI, 0.67-0.88 l/h and 1.64-2.63 microg/ml) for *1/*2 genotypes (n=7), 0.60 l/h and 3.13 microg/ml (95%CI, 0.58-0.62 l/h and 2.68-3.58 microg/ml) for *1/*3 genotypes (n=3), and 0.57 l/h and 3.27 microg/ml in the single *2/*2 carrier. Natural logarithms of tolbutamide plasma concentrations 24 h after intake correlated to plasma clearance (r(2)=0.84, P<0.0000001). This correlation was confirmed in the comparison data set (r(2)=0.97, P<0.0000001).

CONCLUSIONS

A low dose of 125 mg tolbutamide can safely and accurately be used for CYP2C9 phenotyping. As a simple metric for CYP2C9 activity, we propose to determine tolbutamide in plasma 24 h after drug intake.

Tolbutamide, flurbiprofen, and losartan as probes of CYP2C9 activity in humans

The metabolic activity of CYP2C9 in 16 subjects expressing four different genotypes (CYP2C9*1/*1, *1/*2, *1/*3, and *2/*2) was evaluated. Single oral doses of tolbutamide, flurbiprofen, and losartan were administered in a randomized, crossover design. Plasma and urine were collected over 24 hours. The urinary metabolic ratio and amount of metabolite(s) excreted were correlated with formation clearance. The formation clearance of tolbutamide to its CYP2C9-mediated metabolites demonstrated a stronger association with genotype compared to flurbiprofen and losartan, respectively (r2 = 0.64 vs. 0.53 vs. 0.42). A statistically significant correlation was observed between formation clearance of tolbutamide and the 0- to 12-hour urinary amount of 4'-hydroxytolbutamide and carboxytolbutamide (r = 0.84). Compared to tolbutamide, the correlations observed between the respective measures of flurbiprofen and losartan metabolism were not as strong. Tolbutamide is a better CYP2C9 probe than flurbiprofen and losartan, and the 0- to 12-hour amount of 4'-hydroxytolbutamide and carboxytolbutamide is the best urinary measure of its metabolism.

Effects of CYP2C19 and CYP2C9 genetic polymorphisms on the disposition of and blood glucose lowering response to tolbutamide in humans

Several recent in-vitro data have revealed that CYP2C19, in addition to CYP2C9, is also involved in the 4-methylhydroxylation of tolbutamide. We evaluated the relative contribution of CYP2C9 and CYP2C19 genetic polymorphisms on the disposition of blood glucose lowering response to tolbutamide in normal healthy Korean subjects in order to reappraise tolbutamide as a selective in-vivo probe substrate of CYP2C9 activity. A single oral dose of tolbutamide (500 mg) or placebo was administered to 18 subjects in a single-blind, randomized, crossover study with a 2-week washout period. Twelve subjects (of whom six were CYP2C19 extensive metabolizer (EM) and six were CYP2C19 poor metabolizer (PM) genotype) were of the homozygous wild-type CYP2C9*1 genotype; the other six subjects were of the CYP2C9*1/*3 and CYP2C19 EM genotype. Pharmacokinetic parameters were estimated from plasma and urine concentrations of tolbutamide and 4-hydroxytolbutamide. Serum glucose concentrations were measured before and after oral intake of 100 g dextrose. In subjects heterozygous for the CYP2C9*3 allele, C(max) and AUC of tolbutamide were significantly greater and the plasma half-life significantly longer than those in homozygous CYP2C9*1 subjects. No pharmacokinetic differences were found between CYP2C19 EM and PM genotype subjects. The estimated AUC of the increase in serum glucose after oral intake of 100 g dextrose was 2.7-fold higher in subjects with the wild-type CYP2C9 genotype than in those with CYP2C9*1/*3, but CYP2C19 genetic polymorphism did not alter the blood glucose lowering effect of tolbutamide. The plasma AUC of 4-hydroxytolbutamide and the ratio of 4-hydroxytolbutamide/tolbutamide did not differ significantly between CYP2C19 PM and EM genotype subjects, while these parameters were about twice as high in subjects with the wild-type CYP2C9 genotype than in heterozygous CYP2C9*3 subjects (P < 0.05). Our results strongly suggest that the disposition and hypoglycemic effect of tolbutamide are affected mainly by CYP2C9 genetic polymorphism, but not by CYP2C19 polymorphism. The in-vivo contribution of CYP2C19 to tolbutamide 4-methylhydroxylation appears to be minor in humans. This suggests that, at least in vivo, tolbutamide remains a selective probe for measuring CYP2C9 activity in humans.

In vivo effect of clarithromycin on multiple cytochrome P450s

The in vivo effects of oral clarithromycin administration on the in vivo activity of cytochrome P450 1A2, 2C9, and 2D6 were determined. The cytochrome P450 probes caffeine (CYP1A2), tolbutamide (CYP2C9), and dextromethorphan (CYP2D6) were administered as an oral cocktail prior to and 7 days after oral clarithromycin (500 mg twice daily) administration to 12 healthy male subjects. Blood and urine samples were collected and assayed for each of the compounds and their metabolites using high-performance liquid chromatography. The CYP1A2 indices, oral caffeine clearance (6.2 +/- 3.3 l/h before and 5.7 +/- 4.2 l/h after, p > 0.05) and the 6-h paraxanthine to caffeine serum concentration ratio (0.49 +/- 0.3 before and 0.44 +/- 0.3 after, p > 0.05), were unchanged following clarithromycin dosing. Neither the tolbutamide oral clearance (0.77 +/- 0.28 l/h before and 0.72 +/-0.24 l/h after, p > 0.05) nor the tolbutamide urinary metabolic ratio (779 +/- 294 before and 681 +/- 416 after, p > 0.05) indices of CYP2C9 were altered by clarithromycin administration. In the case of CYP2D6, the dextromethorphan to dextrorphan urinary ratio was not significantly different before (0.021 +/- 0.04) and after (0.024 +/- 0.06) clarithromycin dosing. In conclusion, clarithromycin does not appear to alter the in vivo catalytic activity of CYP1A2, CYP2C9, and CYP2D6 in healthy individuals as assessed by caffeine, tolbutamide, and dextromethorphan, respectively.

Fluvoxamine inhibits the CYP2C9 catalyzed biotransformation of tolbutamide

OBJECTIVE

Our objective was to examine the interaction between fluvoxamine and tolbutamide to confirm that fluvoxamine inhibits CYP2C9.

METHODS

The study was carried out as an open, randomized, crossover design with 14 healthy participants. In period A, all volunteers took 500 mg of tolbutamide orally. In period B, the volunteers were randomly assigned to one of two groups. Each group took either 150 mg or 75 mg of fluvoxamine a day for 5 days (day -3 to day 2). The groups then took 500 mg of tolbutamide as a single dose (day 0). In both periods, blood and urine were sampled at regular intervals. Plasma was analyzed for tolbutamide, and urine was analyzed for tolbutamide and its two metabolites, 4-hydroxytolbutamide and carboxytolbutamide by means of HPLC.

RESULTS

During treatment with fluvoxamine, there was a statistically significant decrease in the median of the total clearance of tolbutamide, from 845 mL/h to 688 mL/h, among the volunteers who received 75 mg/d. There was a reduction that reached borderline statistical significance in the group that received 150 mg/d of tolbutamide. The clearance by means of 4-hydroxytolbutamide and carboxytolbutamide was significantly reduced in both groups (ie, from 901 mL/h to 318 mL/h in the group that received 150 mg of tolbutamide per day and from 723 mL/h to 457 mL/h in the group that received 75 mg of tolbutamide per day). Thus there was a tendency toward a more pronounced inhibition of the 4-hydroxylation during treatment with 150 mg/d of fluvoxamine compared with 75 mg/d, but the difference was not statistically significant.

CONCLUSION

Fluvoxamine is a moderate inhibitor of CYP2C9 in vivo.

Quantitative determination of tolbutamide and its metabolites in human plasma and urine by high-performance liquid chromatography and UV detection

An isocratic, high-performance liquid chromatography method has been developed for simultaneous determination of the oral antidiabetic tolbutamide and two of its metabolites, 4-hydroxytolbutamide and carboxytolbutamide, in human plasma and urine. The method was based on simple one-step liquid-liquid extraction with tertiary-butyl methyl ether as extraction solvent. The chromatographic eluent was 23:77 (v/v) methanol: 0.01 M aqueous sodium acetate buffer pH 3.0, and the UV detection was performed at a wavelength of 230 nm. The limit of detection was 0.1 microM for tolbutamide in plasma and 1.5 microM, 0.5 microM, and 0.75 microM for carboxytolbutamide, 4-hydroxytolbutamide, and tolbutamide, respectively, in urine. The limit of quantitation was 0.5 micro for tolbutamide in plasma and 2 microM, 0.75 microM, and 1.25 microM for carboxytolbutamide, 4-hydroxytolbutamide, and tolbutamide, respectively, in urine. The overall mean recoveries ranged from 91% to 109% for tolbutamide in plasma and from 80% to 98% in urine for all three compounds.

Pharmacokinetics of tolbutamide in ethnic Chinese

AIMS

Ethnic differences in drug disposition have been described for many drugs. Despite the widespread use of tolbutamide in Asian populations, the pharmacokinetics of tolbutamide, a CYP2C9 substrate, have not been described in ethnic Chinese.

METHODS

The pharmacokinetics of tolbutamide (500 mg orally) were studied in 10 young, healthy volunteers (seven male/three female; age 21-29 years), each of whom had four ethnic Chinese grandparents. Plasma concentrations of tolbutamide were measured for 32 h post-dose by high performance liquid chromatography. The concentrations of hydroxytolbutamide and carboxytolbutamide were also measured in urine for 32 h post-dose. Noncompartmental pharmacokinetic parameters were calculated using standard equations and compared with those previously reported in Caucasian subjects using the Mann-Whitney U test.

RESULTS

Pharmacokinetic parameters in Chinese (mean+/-s.d.) including Cmax (63+/-11 microg ml(-1)), tmax (median 3.3 h; range 1.6-6.0 h), V/F (9.1+/-1.7 l) and t1/2, (9.1 h; harmonic mean) were similar to the values in Caucasians. CL/F (637+/-88 ml h(-1)) was higher in Chinese than Caucasians. The urinary recoveries of hydroxytolbutamide (13+/-1% of dose) and carboxytolbutamide (68+/-5% of dose) and the partial apparent metabolic clearance (0.15+/-0.02 ml min(-1) kg(-1)) in Chinese were comparable with Caucasians.

CONCLUSIONS

The pharmacokinetics of tolbutamide have been described in ethnic Chinese and the disposition is similar to that reported in Caucasians. This study suggests that there is no substantial ethnic difference in the tolbutamide hydroxylase activity of CYP2C9.

Prolonged elimination of tolbutamide in a premature newborn with hyperinsulinaemic hypoglycaemia

So far, gestational diabetes treated with tolbutamide has never been associated with severe hypoglycaemia in the newborn when the mother's diabetes was well controlled. We report a case of a premature neonate, gestational age 34 weeks, with severe and long-standing hypoglycaemia from birth. The mother had well-controlled gestational diabetes, treated with tolbutamide from the 24th week of gestation until delivery. The neonate had inappropriately high levels of serum proinsulin, insulin and C-peptide relative to blood glucose concentrations. From day 19 after birth, the levels were normalized. Serum tolbutamide was 140.6 micromol/l (38 microg/ml) at 3 h after birth. Zero-order kinetics were seen during the first 90 postnatal hours. The half-life of serum tolbutamide decreased from 46 to 6 h. It is suggested that tolbutamide, when given to the mother until delivery, may cause severe and prolonged hyperinsulinaemic hypoglycaemia in premature neonates. The initially prolonged tolbutamide half-lives and zero-order kinetics suggest immaturity of hepatic elimination during the first 2 days of postnatal life.

Tolbutamide hydroxylation in humans: lack of bimodality in 106 healthy subjects

The tolbutamide hydroxylation capacity was studied in 106 healthy unrelated volunteers from the Australian population. Following a 500 mg oral dose of tolbutamide, the ratio of metabolites (hydroxytolbutamide plus carboxytolbutamide) to unchanged tolbutamide excreted in urine from 6 to 12 h post-dose (urinary metabolic ratio, MR) was determined. Metabolic ratio values did not appear bimodally distributed, even following various transformations of the data (i.e. Log10, inverse, Log10 inverse). A poor metabolizer (PM) subject from a previous clinical study, however, could be distinguished (MR value 159) from the above subjects (MR value range 324-3033), particularly from the histogram plot of inverse tolbutamide metabolic ratio. The poor metabolizer's parents had metabolic ratio values (526 and 478) that were at the lower end of the range of metabolic ratios obtained from the population study, and may indicate that they both have a heterozygous genotype and that a recessive form of inheritance is most likely. As the hydroxylations of tolbutamide and phenytoin are closely linked, the incidence of slow tolbutamide metabolizers is likely to be similar to that for phenytoin (about 1:500) and this is consistent with the failure to detect a single poor tolbutamide metabolizer in our random sample of 106 individuals.

A new aspect of tolbutamide metabolism in the rabbit: the role of 1-butyl-3-(p-formylphenyl)sulphonylurea

We investigated the metabolism of tolbutamide by using synthetic 1-butyl-3-(p-formylphenyl)sulphonylurea (ATB), an intermediate in the metabolic pathway of tolbutamide. ATB (40 mg kg-1) administered intravenously to rabbits was oxidized to 1-butyl-3-(p-carboxyphenyl)sulphonylurea (CTB) and also reduced to 1-butyl-3-(p-hydroxymethylphenyl)sulphonylurea (HMTB). Therefore, it is likely that in the metabolism of tolbutamide, the oxidation of HMTB to ATB involved the reverse reaction, suggesting the reduction of ATB to HMTB. The oxidation of ATB to CTB was inhibited by disulfiram pretreatment. ATB was detected in the blood following intravenous administration of HMTB in rabbits pretreated with disulfiram. These results, confirm that ATB is an intermediate in the oxidative metabolism of tolbutamide in the rabbit.

Excretion of tolbutamide metabolites in young and old subjects

Tolbutamide (1 g/70 kg) was administered as a single intravenous dose to 31 healthy, non-smoking, drug-free males between 23 and 87 years old and the total amounts of hydroxy and carboxytolbutamide excreted in 24 h were measured. There was a significant decrease in the urinary recovery of both metabolites with age. The reason for these findings is not known at the present time and may be associated with the decrease in creatinine clearance observed in these subjects or other changes in the pharmacokinetics of tolbutamide which are currently being investigated.

Validation of the tolbutamide metabolic ratio for population screening with use of sulfaphenazole to produce model phenotypic poor metabolizers

The present study has validated kinetically a convenient method to measure tolbutamide hydroxylation capacity in human beings by use of urinary metabolic ratios. The known in vivo and in vitro inhibitory properties of sulfaphenazole were used to convert control phase subjects to phenotypically "poor" metabolizers of tolbutamide. Six healthy subjects were given a single 500 mg oral dose of tolbutamide with and without sulfaphenazole, 500 mg every 12 hours. Tolbutamide, hydroxytolbutamide, and carboxytolbutamide in urine were determined by newly developed HPLC procedures. Plasma tolbutamide clearance and half-life were measured, as were the metabolic ratio (hydroxytolbutamide + carboxytolbutamide/tolbutamide) in successive 6-hour urine collections. The mean tolbutamide plasma clearance decreased from 0.196 +/- 0.026 ml/min/kg without sulfaphenazole to 0.039 +/- 0.009 ml/min kg with sulfaphenazole, and the mean half-life of tolbutamide increased from 7.28 +/- 0.89 hours to 38.76 +/- 13.30 hours. The metabolic ratio determined in the 6 to 12 hour urine collection period decreased from 794.0 +/- 86.6 to 126.0 +/- 79.3, and this collection period also gave the best separation of subjects between phases. There was a good correlation between tolbutamide plasma clearance and metabolic ratio (rs = 0.853, p less than 0.01, n = 12) and between the percentage decrease in plasma tolbutamide clearance and the percentage decrease in metabolic ratio (r = 0.932, p less than 0.01, n = 6). The tolbutamide urinary metabolic ratio therefore effectively distinguishes tolbutamide hydroxylase activity in "normal" subjects and in those converted to model phenotypically "poor" metabolizers by sulfaphenazole.

Simple high-performance liquid chromatographic method for the determination of tolbutamide and its metabolites in human plasma and urine using photodiode-array detection

A high-performance liquid chromatographic method has been developed for the determination of tolbutamide and its metabolites in human plasma and urine. The compounds examined were extracted with diethyl ether from the acidified biological fluid. Chlorpropamide was used as internal standard, and 235 nm was chosen as the wavelength for diode-array detection. A study of the relationship between the capacity factor and the mobile phase composition and pH showed that acetonitrile-2-propanol-0.1% orthophosphoric acid (17: 17: 66, v/v) was the best eluent on a C8 reversed-phase column. The method is precise, sensitive and suitable for pharmacological investigations.

Lack of relationship between tolbutamide metabolism and debrisoquine oxidation phenotype

The oxidative metabolism of tolbutamide was studied in 13 healthy subjects of known debrisoquine phenotype. Three were poor (PM) and ten were extensive (EM) metabolisers of debrisoquine. The mean values for total plasma clearance, elimination half-life, and metabolic clearance were 0.26 ml.min-1.kg-1, 3.4 h, and 0.17 ml.min-1. kg-1 in PM subjects and 0.22 ml.min-1.kg-1, 4.3 h and 0.15 ml.min-1.kg-1 in EM subjects. Total urinary recovery (% of dose) and ratio of hydroxy- to carboxytolbutamide were 69.4% and 0.219 respectively in PM subjects and 70.9% and 0.226 in EM subjects. There were no statistically significant differences between EM and PM metabolisers for any of these parameters. In addition there was no correlation between the debrisoquine metabolic ratio and tolbutamide urinary metabolite recovery or plasma clearance. These data indicate that hydroxylation of debrisoquine and tolbutamide are not catalyzed by the same enzyme. The ratio of hydroxy- to carboxytolbutamide in our subjects, and in other recent studies, suggests that some previous publications were inaccurate and their conclusions about the genetic control of tolbutamide metabolism were incorrect.

[Quantitative approach to drug compliance of diabetics]

Many patients do not take their medication according to the instructions of their physician. Patient compliance is particularly poor in the long-term treatment of diseases, which produce only minor symptoms in the patient. The aim of the present study was to quantitate patient compliance in diabetics treated with tolbutamide. The amount of tolbutamide taken by the patient was estimated on the basis of its 24 hours urinary excretion. The collection of the 24 hours urine is a standard procedure in our outpatient clinic for the measurement of glycosuria. The study involved 33 diabetics followed over several months. In those reliable patients, who conscientiously took their medication, we found a good correlation between the prescribed dose and the 24 hours urinary excretion. In contrast, when the data from the entire group of 33 out-patients was evaluated, the correlation between these two parameters of prescribed dose and urinary excretion was poor. The correlation between urinary excretion and the dose admitted by the patient was just as poor. According to our results, only one out of two diabetics took the prescribed dose. After our initial observations, we informed the outpatients about their inclusion in this study. In spite of the fact that the patients knew that they were under observation for compliance, we found large interindividual differences in the urinary excretion of tolbutamide over several months. In contrast, the intraindividual variability was low. We found that interviews with patients are a poor means to detect non-compliance, whereas urinary excretion measurements appear to be more reliable.(ABSTRACT TRUNCATED AT 250 WORDS)

Determination of oral anti-diabetic agents in human body fluids using high-performance liquid chromatography

Two widely prescribed anti-diabetic agents for which no simple assay method was previously available can now be determined by high-performance liquid chromatography using a UV detection system. The two drugs investigated were tolbutamide (a sulphonylurea) and phenformin (a biguanide). Tolbutamide can be assayed directly, after a single extraction step, on a reversed-phase system, illustrating the simplicity of the technique for carrying out analyses on underivatised drug compared with gas chromatography. Phenformin was not so easily chromatographhed using straightforward partition systems; however, by the choice of a suitable ion-pair agent it was possible to chromatograph the underivatised drug in a relatively simple reversed-phase system.

Determination of phenylbutazone, tolbutamide and metabolites in plasma and urine using chemical ionization mass spectrometry

Quantitative analytical procedures for the analysis of phenylbutazone and tolbutamide levels in plasma have been developed which involve the addition of deuterium labeled internal standards to plasma followed by extraction and direct sample insertion into a mass spectrometer operating under chemical ionization conditions. Peak height ratios used to calculate plasma levels were determined by using either selected ion monitoring or repetitive scan data. The scan approach was used in a related procedure for the simultaneous determination of tolbutamide and two metabolites from urine. The accuracy, precision and sensitivity of the direct sample insertion approach to drug level measurement has been determined. Examples are given of data obtained in the course of pharmacokinetic studies in which this analytical approach appears to offer advantages in the analysis of multicomponent mixtures encountered in drug-drug interaction studies.

Alteration of drug metabolism in Gilbert's syndrome

The pathophysiology of Gilbert's syndrome was studied by investigating the metabolism of the drug tolbutamide, which is metabolised by the liver but does not undergo glucuronidation. Using rat liver cell supernatant, tolbutamide was shown to bind to the hepatic cytoplasmic Y protein in a manner similar to other organic anions, but not to Z protein. In 31 patients with Gilbert's syndrome the plasma disappearance (plasma half-life, mean +/- SD: 628+/-84 min) and metabolic clearance (7-9+/-1-8 ml/min) were significantly (P less than 0-0005) altered compared with the 13 controls (mean half-life 393+/-26 and mean clearance 13-4+/-1-5). The eight patients with hyperbilirubinaemia due to haemolytic disease showed no difference from the normal control subjects. In three patients with Gilbert's syndrome the cumulative urinary excretion of tolbutamide metabolites, 24 hours after the administration of the drug, was 30% lower than in the controls. In the five patients with Gilbert's syndrome, phenobarbital administration (100 mg/day) produced a significant increase in clearance of the drug from 8-8+/-0-8 to 13-4+/-1-9 ml/min; this was paralleled by a fall in serum bilirubin concentration. The plasma half-life of tolbutamide was similar in Gunn rats and Wistar rats. The results suggest that the metabolic defect(s) of Gilbert's syndrome affects compounds other than bilirubin and that defective uptake is probably the major factor.

Alteration of drug metabolism during cholestasis in man

The morphological and functional alterations of the smooth endoplasmic reticulum of the liver cell related to biliary stasis have brought attention to drug biotransformation during cholestasis. The metabolism of meprobamate, pentobarbital and tolbutamide was assessed in subjects with intrahepatic recurrent cholestasis (3), cholestatic hepatitis (6), extrahepatic biliary obstruction (7) and normal controls (16). In the patients with recurrent intrahepatic cholestasis no differences in drug metabolism were noted as compared to the control group. In cholestatic hepatitis the plasma half-lives of meprobamate (828 +/- 422 min.) and pentobarbital (39+-65) were significantly longer than in in controls (444 +/- 37 and 25.4 +/- 1.1 respectively). Tolbutamide plasma half-life appeared unchanged. The most striking variations were observed in the patients with extrahepatic biliary obstruction. In such cases while meprobamate half-life was unchanged, pentobarbital half-life was significantly prolonged (31.2 +/- 2.5) and the in vitro metabolism of the drug, using liver preparations, was decreased to less than 50% of the control value. In contrast the metabolism of tolbutamide was accelerated as evidenced by a significant decrease of plasma half-life (165 +/- 48 min. versus 384 +/- 76 of the controls) and an enhanced urinary excretion of the drug's metabolites. However the metabolism of tolbutamide in vitro did not show any difference between normal and cholestatic liver. Whatever the mechanism of the peculiar behaviour of tolbutamide in extrahepatic biliary obstruction it seems to be related to the increased bile dalt concentration during cholestasis. In fact the low values of plasma half-life increase significantly either relieving the biliary obstruction or producing a bile salt depletion with cholestyramine. Preliminary results in vitro suggest the bile salt could displace tolbutamide from albumin binding thus increasing the amount of free drug available for biotransformation by the liver. In conclusion cholestasis may affect drug metabolism depending on the degree of biliary stasis, liver cell injury and the type of drug tested. The mechanism could be that of an impaired biotransformation in the smooth endoplasmic reticulum or could involve extrahepatic factors.