Inhibition Constants, Inhibitor Concentrations and the Prediction of Inhibitory Drug Drug Interactions: Pitfalls, Progress and Promise

Strategies and standards for predicting the likelihood of pharmacokinetically significant inhibitory drug-drug interactions for drug development purposes which rely primarily on projected in vivo concentrations of cytochrome P450 (CYP) or transporter inhibitors, [I], and in vitro estimates of their inhibitory constants, K(i), were specified in several commentaries based upon a conference held by the European Federation of Pharmaceutical Sciences (EUFEPS) several years ago. Since then the application of those strategies and standards has met with varying degrees of success. Many of the vexing issues that were identified in the EUFEPS Conference Report remain, while other issues are systematically being resolved. This article briefly reviews the underlying strategy in the prediction of the significance of inhibitory DDIs using [I]/K(i) ratios; some of the difficulties or pitfalls associated with the predictive application of [I]/K(i) ratios; and some of the recent refinements of the general strategy.

Lack of pharmacokinetic interaction between rofecoxib and methotrexate in rheumatoid arthritis patients

Rofecoxib is a highly selective and potent inhibitor of cyclooxgenase-2 (COX-2). Methotrexate is a disease-modifying agent with a narrow therapeutic index frequently prescribed for the management of rheumatoid arthritis. The objective of this study was to investigate the influence of clinical doses of rofecoxib on the pharmacokinetics of methotrexate in patients with rheumatoid arthritis. This was a randomized, double-blind, placebo-controlled study in 25 rheumatoid arthritis patients on stable doses of methotrexate. Patients received oral methotrexate (7.5 to 20 mg) on days -1, 7, 14, and 21. Nineteen patients received rofecoxib 12.5, 25, and 50 mg once daily on days 1 to 7, 8 to 14, and 15 to 21, respectively. Six patients received placebo on days 1 to 21 only to maintain a double-blinded design for assessment of adverse experiences. Plasma and urine samples were analyzed for methotrexate and its major although inactive metabolite, 7-hydroxymethotrexate. The AUC(0-infinity) geometric mean ratios (GMR) and their 90% confidence intervals (90% CI) (rofecoxib + methotrexate/methotrexate alone) for day 7/day -1, day 14/day -1, and day 21/day -1, for rofecoxib 12.5, 25, and 50 mg, were 1.03 (0.93, 1.14), 1.02 (0.92, 1.12), and 1.06 (0.96, 1.17), respectively (p > 0.2 for all comparisons to day -1). All AUC(0-infinity), GMR and Cmax GMR 90% CIs fell within the predefined comparability limits of (0.80, 1.25). Similar results were observed for renal clearance of methotrexate and 7-hydroxymethotrexate at the highest dose of rofecoxib tested (50 mg). It was concluded that rofecoxib at doses of 12.5, 25, and 50 mg once daily has no effect on the plasma concentrations or renal clearance (tested at the highest dose of rofecoxib) of methotrexate in rheumatoid arthritis patients.

The use of in vitro methods to predict in vivo pharmacokinetics and drug interactions

With the dramatic change underway in the process of drug discovery and development it has become increasingly important to define, both qualitatively and quantitatively, the dispositional features of new chemical entities (NCEs) as early in the process as possible. To that end strategies have emerged that are designed to enable reasonable predictions about a NCE's absorption from the gastrointestinal tract, systemic bioavailability and likelihood for significant pre-systemic clearance, character of metabolic processing both within the gastrointestinal tract and the liver, in vivo pharmacokinetics (PK), and likelihood for clinically significant interactions with other drugs. To some extent these strategies have embraced interspecies allometric scaling in which findings in animals are extrapolated to predict outcomes in humans. However, a greater emphasis in recent years has been placed on predicting human PK and the likelihood of clinically significant drug-drug interactions for NCEs solely from in vitro experiments. These general strategies have been methodologically streamlined so that hundreds or even thousands of experiments on a given NCE can be conducted within several days. Dispositional data from these pre-clinical experiments is useful for rapidly identifying potential marketing advantages for NCEs, and for screening out those substances that should not be placed into more expensive and labor-intensive animal experiments or brought to clinical trial. The key issue in these strategies is the accuracy with which pre-clinical findings predict clinical outcomes. Based largely on retrospective analyses the current state of the art exhibits a high percentage of useful predictions. However, there are many examples in which the prediction of either human PK or clinical drug-drug interactions from pre-clinical data has failed. The reasons for inaccurate predictions are manifold, and may include the actual in vitro methodology used, inappropriate model selection, and errant scale-up factors. Additionally, in vitro methods may fail to account for complex hepatobiliary processing including transport phenomena and Phase II metabolism. Progress has been made in establishing humanized methodologies that accurately describe these processes, with a view toward reconstituting the contributions of each into a more complex and accurate depiction and prediction of in vivo PK and drug-interaction potential.

Norastemizole Sepracor

Sepracor and Janssen are developing the histamine H1 antagonist, norastemizole (an active metabolite of Johnson & Johnson's Hismanal), for the potential, non-sedating treatment of allergy. Sepracor expects to file an NDA with the FDA by the fourth quarter of 2000 [337315,358429]. As of September 1999, Sepracor was conducting two large-scale phase III seasonal allergic rhinitis studies [340260]. Sepracor expects norastemizole to be the most potent non-sedating histamine, with equal or more rapid onset of action than other therapies [229516]. Norastemizole is 13- to 16-fold more potent as an H1 antagonist than astemizole and 20- to 40-fold more potent in inhibiting histamine-induced bronchoconstriction. Following a single dose of norastemizole (25 mg p.o.), there is significant attenuation of histamine-induced wheal and flare responses within 30 min. The drug's major advantage is its lack of cardiotoxicity or interactions with other drugs that increase the risk of developing serious arrhythmias [301469]. In July 2000, Morgan Stanley Dean Witter predicted filing for FDA approval for allergic rhinitis during the first half of 2001, and a partnership announcement around the time of this NDA filing. The analysts also forecast European sales of $8.3 million in 2002, rising to $16.7 m by 2005 [384868].

Differential kinetics of phenytoin in elderly patients

The elderly have a relatively high risk of developing adverse drug reactions. Phenytoin continues to be a preferred drug for treating generalised tonic-clonic seizures in the elderly and simple partial seizures that generalise. Phenytoin is eliminated almost entirely by hepatic oxidation. The principle enzymes responsible are cytochrome P450 (CYP)2C9 and CYP2C19. CYP2C9 is saturated by therapeutic doses of phenytoin, and at steady state both enzymes are probably operant in most people. The nonlinear pharmacokinetics of phenytoin make it a difficult drug for which to establish safe and effective administration regimens. An important area of inquiry is whether the differential disposition kinetics of phenytoin in the elderly render its administration an even more difficult challenge. Moreover, since the elderly are generally subject to more polypharmacy than younger adults, are they, as a result, subject to either more frequent or more severe drug interactions with phenytoin than younger adults? In order to examine these issues we were interested in learning the extent to which old age might affect the plasma protein binding of phenytoin, its hepatic metabolism and, ultimately, its pharmacokinetic profile. With regard to the latter we looked carefully at the methods that have been used to characterise the disposition kinetics of phenytoin in general, and in the elderly, in particular. There are many conflicting findings with regard to the effect of age on the disposition kinetics of phenytoin. However, the strategies used for estimating kinetic parameters for phenytoin [viz the maximum rate of metabolism/elimination (Vmax) and the Michaelis-Menton constant (Km)] exhibit deficiencies that could account for some of the disparate findings. Certainly, more careful prospective studies focusing on the effects of age on phenytoin disposition kinetics are warranted. However, in light of the information currently available, no special attention need be paid to the initiation of phenytoin administration in elderly patients who are taking multiple anticonvulsants. On the other hand, for the elderly receiving phenytoin monotherapy, the initiation of phenytoin administration should occur at lower doses than would be customary for younger adults, and phenytoin blood concentrations should be appropriately monitored in order to evaluate individual Vmax and Km values for informed dosage adjustments.

Cyclooxygenase-2 inhibition by rofecoxib reverses naturally occurring fever in humans

Cyclooxygenase (COX) exists as constitutive (COX-1) and inducible (COX-2) isoforms. Nonsteroidal antiinflammatory drugs (NSAIDs) such as ibuprofen and diclofenac inhibit both COX-1 and COX-2. The role of COX-2 in the genesis of fever in monkeys and humans was examined with use of the specific COX-2 inhibitor rofecoxib. Rofecoxib was administered to monkeys made febrile by 6 microg/kg intravenous lipopolysaccharide. Induced pyrexia was followed by oral rofecoxib (1 or 3 mg/kg), diclofenac (3 mg/kg), or vehicle. Rofecoxib and diclofenac rapidly reversed the elevated temperature (P < .05 versus vehicle for 3 mg/kg rofecoxib and diclofenac at 70 to 90 minutes after dosing). A single-dose, parallel-group, double-blind randomized trial was conducted in 94 patients with fever caused by a viral-type illness. Mean baseline temperature was similar for all groups (-38.5 degrees C). Patients received oral doses of 12.5 mg rofecoxib, 25 mg rofecoxib, 400 mg ibuprofen, or placebo and the mean +/- SE change in oral temperature at 4 hours after dosing was -0.97 degrees C +/- 0.11 degrees C, -1.19 degrees C +/- 0.09 degrees C, -1.20 degrees C +/- 0.11 degrees C, and 0.01 C +/- 0.17 C, respectively (P < .001 for active treatments versus placebo). Specific inhibition of COX-2 by rofecoxib results in antipyretic activity in monkeys and humans comparable to dual COX-1/COX-2 inhibitors such as diclofenac or ibuprofen. The data support the hypothesis that it is the COX-2 isoform that is primarily involved in the genesis of fever in humans.

An investigation of the relationship between estrogen, estrogen metabolites and blood cholesterol levels in ovariectomized rats

17 beta-Estradiol (E2) has long been known for protecting against coronary heart disease by lowering cholesterol levels in premenopausal women. A recent study in our laboratory suggested that two hydroxylated metabolites of E2 possess similar hypocholesterolemic effects in male rats. This effect has been further investigated with additional estrogen metabolites in ovariectomized rats with a view toward mimicking the true postmenopausal situation in humans. Their effects in reproductive tissues were also evaluated histologically. Fundamentally, the following issues were addressed: (1) Do oxidized metabolites of estradiol lower total cholesterol levels? (2) Can a hypocholesterolemic effect be achieved without eliciting estrogenic activities on reproductive tissues? The results of this investigation showed that a number of oxygenated metabolites of estradiol can lower cholesterol levels. Among them, 4-hydroxyestradiol (4-OHE2) produced a striking hypocholesterolemic effect and a substantial uterotropic effect. 2-Hydroxyestradiol (2-OHE2), 2-methoxyestradiol (2-meoE2) and 2-methoxyestrone (2-meoE1) produced a significant decrease in cholesterol levels at doses that did not produce significant uterotropic effects.

Influence of some novel N-substituted azoles and pyridines on rat hepatic CYP3A activity

A series of N-substituted heteroaromatic compounds structurally related to clotrimazole was synthesized, and the effects of these compounds on ethosuximide clearance in rats were determined as a measure of their abilities to induce cytochrome P4503A (CYP3A) activity. Ethosuximide clearance and in vitro erythromycin N-demethylase activity were shown to correlate. In this series, imidazole or other related heteroaromatic "head groups" were linked to triphenylmethane or other phenylmethane derivatives. Within the series, it was found that 1-triphenylmethane-substituted imidazoles elicited the greatest increase in CYP3A activity, and that among the triphenylmethyl-substituted imidazoles, the highest activities were achieved by the substitution of F- or Cl- in either the meta or para position of one of the phenyl rings. Diphenylmethyl-substituted pyridine was effectively devoid of activity. Compounds eliciting the largest increase in CYP3A activity (viz. 1-[(3-fluorophenyl)diphenylmethyl]imidazole, 1-[(4-fluorophenyl)diphenylmethyl]imidazole, and 1-[tri-(4-fluorophenyl)methyl]imidazole) produced little or no increase in ethoxyresorufin O-dealkylase (EROD) activity (i.e. CYP1A), whereas benzylimidazole, which elicited only a small increase in CYP3A activity, produced an almost 9-fold increase in CYP1A activity. For a series of eleven compounds exhibiting a wide range of influence on CYP3A activity, a positive correlation was found between ethosuximide clearance and hepatic CYP3A mRNA levels.

Ethosuximide is primarily metabolized by CYP3A when incubated with isolated rat liver microsomes

The cytochrome P450 (CYP) subfamily responsible for ethosuximide metabolism was investigated by HPLC assay of ethosuximide incubations with isolated rat liver microsomes from control rats and from rats treated with inducing agents to enrich hepatic microsomes in selected CYP isoforms. Inducing agents included beta-naphthoflavone (BNF, CYP1A inducer), phenobarbital (PB, CYP2B/2C/3A), isoniazid (INH, CYP2E1), clotrimazole (CTZ, CYP3A), clofibrate (CLO, CYP4A), and an imidazole CTZ-analog known as CDD3543 (CYP3A). Incubations with BNF, INH, CTZ, and control microsomes showed significantly (p<0.05) more metabolite produced by CTZ microsomes vs. BNF, INH, and control microsomes at 10, 30, 60, and 120 min incubation. Ethosuximide metabolite levels generated by CTZ microsomes at 120 min were 36.5 times those of control microsomes. Correspondingly, ethosuximide concentrations were significantly (p<0.05) lower for incubations with the CTZ microsomes compared with BNF, INH, and control microsomes at 60 and 120 min. Sixty-minute incubations with all microsome groups exhibited significantly (p<0.05) higher metabolite formation rates (nmol/nmol CYP/min) for CTZ (11.8x control) and PB (9.6x control) microsomes vs. all other groups. Antibody inhibition experiments demonstrated ethosuximide metabolite levels for PB microsomes were not affected by CYP2B1 antibodies, whereas CYP3A2 antibodies reduced metabolite levels for both PB and CTZ microsomes by over 80%. These results indicate CYP3A is primarily responsible for ethosuximide metabolism in rats.

The influence of anesthetic agents on rat hepatic cytochromes P450 in vivo

The objective of this study was to identify which anesthetics when used acutely will affect cytochrome P450 (CYP) activity in male Sprague-Dawley rats in vivo. The anesthetics tested were fentanyl citrate, alpha-chloralose, ketamine, urethane (ethyl carbamate), halothane, and ether. CO2 anesthesia was used as the control comparator. Theophylline was used as a probe for CYP1A activity, phenobarbital for CYP2B/2C, flecainide for CYP2D1, and ethosuximide for CYP3A activity. All probes were administered via tail vein injection after anesthetic-induced loss of the righting reflex. Single sample probe clearances were estimated, and used as an index of CYP activity. Fentanyl citrate, alpha-chloralose, halothane, and ether did not have statistically significant effects on any of the CYP activities. Ketamine did not significantly affect CYP1 or CYP2B/2C activity. However, it decreased the clearance of flecainide (i.e. CYP2D1 activity) by 13.4% (p < 0.001) and the clearance of ethosuximide (i.e. CYP3A activity) by 17.6% (p < 0.0001). Urethane increased the clearance of theophylline by 91.5% (p < 0.0001), and decreased the clearance of ethosuximide by 40.5% (p < 0.0001) though it did not affect CYP2B/2C or CYP2D1 activities significantly. From this data, we conclude that a single dose of ketamine mildly inhibits the activity of CYP2D1 and CYP3A, and a single dose of urethane strongly inhibits CYP3A but increases CYP1A activity.

Exercise training does not alter cytochrome P-450 content and microsomal metabolism

The purpose of this investigation was to determine whether increased endurance exercise capacity alters total hepatic cytochrome P-450 content and cytochrome P-450 (CYP1A and CYP2B) mediated hepatic microsomal mixed-function oxidase drug metabolism. Twenty adult male Sprague-Dawley rats were randomly assigned to either a control (C) or an endurance trained group (ET). ET rats were progressively trained 5 d.wk-1 for 11 wk. Both C and ET rats were administered in random order single posttraining doses of probe drugs theophylline (probe for CYP1A) and antipyrine (probe for CYP2B). Soleus muscle citrate synthase activity of ET rats was significantly greater (P < 0.01) than for C rats (mean +/- SD; C, 26.4 +/- 1.3 mumol.g-1.min-1; ET, 46.1 +/- 2.7). In contrast, total liver cytochrome P-450 content was not significantly different (P > 0.01) among C and ET rats (mean +/- SD; C, 0.554 +/- 0.055 nmol.mg-1 liver protein; ET, 0.604 +/- 0.080). Likewise, the posttraining C and ET single-sample plasma clearances of theophylline (mean +/- SD; C, 1.89 +/- 0.360 1.h-1.kg-1 total liver weight; ET, 2.08 +/- 0.49) and antipyrine (mean +/- SD; C, 6.44 +/- 1.56 1.h-1.kg-1 total liver weight; ET, 6.51 +/- 1.02) were not significantly different (P > 0.01). Therefore, it was concluded that strenuous endurance training of 11 wk duration did not alter total hepatic cytochrome P-450 content or CYP1A or CYP2B activity.

Short report: a comparative study of the interaction between antacid and H2-receptor antagonists

The influence of concomitant antacid administration on the relative bioavailability of the H2-receptor antagonists cimetidine, famotidine, nizatidine and ranitidine, was investigated in a panel of 21 healthy, adult male volunteers in an eight-way crossover trial. Administration with antacid reduced the bioavailability of all agents tested. The reduction in area under the serum concentration-time curve (AUC) was greatest for cimetidine (23%) and ranitidine (26%) and least for nizatidine (12%) and famotidine (19%). Reductions in peak serum concentration (Cmax) followed a similar pattern. The times of peak serum concentrations were not affected by antacid. Comparison of the relative bioavailability among all drugs tested showed no statistically significant differences in the effect of antacid administration on these agents. However, a high degree of intersubject variability was observed.

Drug interactions of H2-receptor antagonists

Three drug interactions of nizatidine and of other antisecretory agents were studied comparatively. First, the effects of nizatidine, cimetidine and ranitidine on the dispositional kinetics of theophylline were evaluated in chronic obstructive pulmonary disease (COPD) patients. Second, the effect of magnesium/aluminium hydroxide on the relative bioavailability of nizatidine, famotidine, cimetidine and ranitidine was evaluated in healthy volunteers. Finally, the effects of nizatidine and omeprazole on the dispositional kinetics of phenytoin were evaluated in healthy volunteers. Only cimetidine altered the steady-state kinetics of oral theophylline, slowing theophylline clearance by 25%. Each of the H2-receptor antagonists exhibited a modest decline in relative bioavailability when ingested with antacid. Antacid ingestion decreased the bioavailability of famotidine, ranitidine and cimetidine by 20-25%, and the bioavailability of nizatidine by 12%. Each of these effects was statistically significant. Finally, it was found that neither omeprazole nor nizatidine affected the single dose kinetics of phenytoin.

Absence of an inhibitory effect of omeprazole and nizatidine on phenytoin disposition, a marker of CYP2C activity

The effects of omeprazole (40 mg orally per day) and nizatidine (300 mg orally per day) on the disposition of phenytoin (4.5 mg kg(-1) p.o. single dose) were studied in 18 healthy, young adult males. Total and unbound plasma concentrations of phenytoin were measured for 48 h after each dose of phenytoin. Neither treatment altered the disposition kinetics of phenytoin, the hydroxylation of which is mediated specifically by cytochromes P450 of the 2C subfamily.

Use of single sample clearance estimates of cytochrome P450 substrates to characterize human hepatic CYP status in vivo

1. Single sample clearance estimates (CL/F) of orally administered ethosuximide were obtained in four groups of healthy adult subjects. One group was treated with phenobarbital to induce CYP2B/2C and CYP3A activity; one group was treated with rifampin to induce CYP3A activity; one group consisted of cigarette smokers (increased CYP1A activity), and one group was untreated (controls). Ethosuximide CL/F values were slightly, though not significantly, increased among cigarette smokers (12.5% increase) and the phenobarbital group (25% increase), but rifampin treatment resulted in a significant increase (65%). 2. The influence of rifampin treatment on the single sample oral clearances of antipyrine, theophylline, phenytoin, carbamazepine, ethosuximide, quinidine, valproic acid, and lorazepam was investigated to determine whether rifampin induces only CYP3A. Rifampin treatment significantly increased the oral clearance of each drug from 1.4-fold (valproic acid) to 3.4-fold (quinidine). These findings indicate that the inductive effect of rifampin extends well beyond CYP3A.

The effects of cobalt chloride, SKF-525A, and N-(3,5-dichlorophenyl)succinimide on in vivo hepatic mixed function oxidase activity as determined by single-sample plasma clearances

1. Four drugs--antipyrine, theophylline, quinidine, and ethosuximide--were used as probes of in vivo hepatic mixed function oxidase (MFO) activity. Functional MFO activity was evaluated by estimating probe clearances subsequent to pretreatment of rats with either cobalt chloride, SKF-525A, or N-(3,5-dichlorophenyl) succiminide (NDPS). 2. Clearances of each probe were estimated from single plasma concentration measurements. Each pretreatment altered the clearances of this panel of probes in a different way. NDPS pretreatment increased theophylline clearance while slowing quinidine and ethosuximide clearances. SKF-525A slowed all probe clearances except for ethosuximide. Cobalt chloride slowed all probe clearances except for theophylline. 3. The use of multiple probes as substrates for the hepatic cytochrome P-450 system can provide some insight into the functional consequences of xenobiotic exposures on that system. Moreover, xenobiotic-induced functional changes on hepatic MFO when assessed in vivo appear to be modest relative to changes in in vitro activity or hepatic cytochrome P-450 content. This minimally invasive multiprobe method may be useful for assessing xenobiotic influences on human hepatic MFO in vivo.

The use of single sample clearance estimates to probe hepatic drug metabolism: handprinting the influence of phenobarbitone on human hepatic drug metabolism

1. Single sample clearance estimates, CL, were calculated for seven drugs employed as probes of human hepatic drug-metabolizing enzymes. Clearance estimates were calculated in healthy young adult male volunteers either taking no pretreatment, or taking phenobarbitone (PB) 100 mg nightly for 3 nights. This intermittent regimen (3 nights on, followed by 4 nights off) was repeated for at least 3 consecutive weeks prior to challenge with an individual probe. 2. Valproic acid was selected as a probe of both peroxisomal and microsomal beta-oxidase activity; antipyrine, phenytoin, quinidine, and carbamazepine were selected as probes of hepatic mixed-function oxidases (MFO), and lorazepam as a probe for UDP-glucuronosyl transferase activity. 3. Clearances of all probes except lorazepam, theophylline and phenytoin were approximately 20-30% faster in PB-treated than in control subjects; however, only in the case of carbamazepine did the increased clearance approach statistical significance. Neither phenytoin nor theophylline clearances were increased by PB. 4. A clearance index (probe CL for PB-treated subjects divided by probe CL for untreated subjects) was calculated for each probe, and an ordinal transformation of the log of the resultant ratio was plotted for each probe giving rise to a 'handprint' of the effect of PB on drug-metabolizing activity.

The use of single sample clearance estimates to probe hepatic drug metabolism: handprinting the influence of cigarette smoking on human hepatic drug metabolism

1. Conditions were examined under which estimates of drug clearance made from a single measurement of plasma concentration effectively represented multiple-sample estimates of clearance for quinidine, valproic acid, unbound valproic acid, and lorazepam. When plasma concentrations were measured at various post-dose times, both individual and mean values of single-sample clearance estimates, CL, corresponded closely to multiple-sample clearance estimates. Best post-dose sampling times were: quinidine, 8 h; valproic acid, 24 h; and lorazepam, 24 h. 2. Single-sample clearance estimates, CL, were calculated for seven drugs employed as probes of human hepatic drug-metabolizing enzymes. Valproic acid was used to probe microsomal and peroxisomal beta-oxidase activity; antipyrine, phenytoin, quinidine, carbamazepine, and theophylline were used as probes of hepatic mixed-function oxidases (MFO), and lorazepam as a probe for UDP-glucuronosyltransferase activity. 3. A clearance index (CI, namely probe CL for smokers divided by probe CL for non-smokers) was calculated for each probe. The effect of cigarette smoking (and presumably polycyclic aromatic hydrocarbon exposure) on all probe CL values was consolidated and plotted as the logarithm of the CI to produce a handprint of drug metabolizing enzyme activity for cigarette smokers. 4. Only theophylline CL was significantly faster among smokers than non-smokers (P less than 0.01). 5. We conclude that the use of multiple probes of MFO activity when given in a single-dose, single-sample protocol for structuring handprints represents a minimally invasive and useful approach to characterize xenobiotic-mediated effects on hepatic MFO.

The use of single-sample clearance estimates to probe hepatic drug metabolism in rats. IV. A model for possible application to phenotyping xenobiotic influences on human drug metabolism

1. Conditions were examined under which estimates of drug clearance made from a single measurement of plasma concentration effectively represented multi-sample estimates of clearance for carbamazepine, quinidine, and paracetamol. When plasma concentrations were measured at various post-dose times, both individual and mean values of single-sample clearance estimates (CL) corresponded closely to multi-sample clearance estimates. Best post-dose sampling times were: carbamazepine, 3 h; quinidine, 10 h; and paracetamol, 6 h. 2. Single sample clearance estimates, CL, were calculated for seven drugs employed as probes of hepatic drug-metabolizing activity in rats. Valproic acid was investigated as a probe of microsomal and peroxisomal oxidases; antipyrine, theophylline, ethosuximide, carbamazepine and quinidine as probes of hepatic mixed-function oxidases (MFO), and paracetamol as a probe for UDP-glucuronosyltransferase activity. 3. A clearance index (CI, namely, probe CL after xenobiotic pretreatment divided by control probe CL) was calculated for each probe. Eight pretreatments were used: phenobarbital (PB), beta-naphthoflavone (beta NF), polychlorinated biphenyls (PCB), rifampin, pregnenolone-16 alpha-carbonitrile (PCN), clofibric acid, cimetidine, and piperonyl butoxide. The effect of each xenobiotic pretreatment on all probe CL values was consolidated and plotted as the logarithm of the CI, and a distinct pattern or handprint evolved for each pretreatment. 4. We conclude that the use of multiple single-sample probes of hepatic MFO activity can be useful in structuring handprints to characterize xenobiotic-mediated effects on hepatic MFO. This minimally invasive in vivo approach may have application in investigating and possibly phenotyping MFO activity in human subpopulations that are subject to sustained exposure to particular xenobiotics.

The suitability of carbamazepine as a single-sample probe of human mixed function oxidase activity

1. Carbamazepine concentrations measured in plasma and plasma ultrafiltrates by fluorescence polarization immunoassay (f.p.i.a.) from 0 to 48 h post-dose were used to calculate CL, V, clearance of plasma unbound drug (CLunb), Vunb; mean (+/- SD) values for these were: 0.017 l/kg/h (+/- 0.004), 1.05 l/kg (+/- 0.14), 0.058 l/h/kg (+/- 0.012), and 4.28 l/kg (+/- 0.41), respectively. 2. A single-dose, single-sample procedure for estimating carbamazepine oral clearance was evaluated with a view to using carbamazepine as a probe in screening for host factor influences on human drug metabolism. Single sample estimates of carbamazepine clearance (CL) were closest to multiple sample values for carbamazepine clearance (CL) when blood samples were collected 48 h after carbamazepine ingestion. This was the case for plasma total carbamazepine and plasma unbound carbamazepine. 3. A value of 1.1 l/kg was used for V in calculating all single sample estimates of clearance (CL), and a value of 4.3 l/kg was used to calculate single sample estimates of clearance of plasma unbound drug (CLunb). The mean prediction error (MPE) was negatively biased for CL and CLunb values calculated from single carbamazepine concentrations in samples collected at 24, 36, and 48 h post-dose. MPE was less than 5% errant for CL and less than 1% errant for CLunb when those parameters were calculated from 48 h concentrations of plasma total carbamazepine or plasma unbound carbamazepine, respectively. Root mean squared error (r.m.s.e.) was lost lowest when 48 h post-dose samples were used for single-sample clearance estimates.

Optimal dosing of phenytoin: an evaluation of the timing and appropriateness of serum level monitoring

The appropriateness of all serum phenytoin concentrations measured at this 294-bed, tertiary care institution over a 7-week period were evaluated. Each serum level was determined to be justified or unjustified based on either pharmacokinetic or clinical criteria. A total of 234 serum levels were measured in 58 patients. Thirty-eight (16%) of the levels were justified on a pharmacokinetic basis, whereas 70 (30%) were justified for clinical assessment of the patient. One hundred thirty (56%) of the serum levels could not be justified on any grounds. Also identified were distinct clinical situations for 13 patients for which monitoring of serum levels was warranted but not performed. Most phenytoin serum level measurements at this institution were not reasonably justified and may have been unnecessary. A yearly reduction in laboratory costs in excess of $9,000 was predicted if such serum levels had been ordered only when clearly indicated.

Accuracy of two equations in determining normalized phenytoin concentrations

The accuracy of two equations in normalizing total phenytoin concentrations in the presence of renal failure or hypoalbuminemia was evaluated in 11 renal failure and 23 hypoalbuminemic patients. Blood samples were obtained from hospitalized patients receiving phenytoin and were assayed for free and total phenytoin concentrations. Estimated normalized phenytoin concentrations based on free drug concentration were compared statistically with normalized concentrations calculated from the two equations via Student's t-test. The equation for normalizing phenytoin concentrations in hypoalbuminemic patients significantly underpredicted normalized phenytoin concentrations 15.7 +/- 8.5 versus 19.9 +/- 12.1 mg/L (p less than 0.001). In patients with renal failure, the mean phenytoin concentration from the respective equations and that based on free concentration were 14.1 +/- 6.2 and 14.0 +/- 7.9 mg/L, respectively. However, in 5 of 11 renal failure patients the equation resulted in over- or underprediction by at least 25 percent. Neither equation should be used clinically to normalize phenytoin concentrations in these patient populations.

Predicting the ciprofloxacin-theophylline interaction from single plasma theophylline measurements

The effect of ciprofloxacin treatment on theophylline clearance was evaluated with a theophylline multiple dose, multiple sample protocol and with a single dose, single sample protocol. The object was to determine whether a single dose, single sample protocol for estimating theophylline clearance could be used as a screening strategy for evaluating host factor influences on theophylline clearance. Ciprofloxacin (750 mg per os) was administered every 12 h for nine doses in the multidose study and every 12 h for seven doses in the single dose protocol. Subjects were sixteen healthy, non-smoking young adult males. The oral clearance of theophylline at steady state, (CL/F)ss, decreased from a mean (+/- s.d.) value of 0.035 (+/- 0.008) 1 h-1 kg-1 to 0.024 (+/- 0.004) 1 h-1 kg-1 during ciprofloxacin treatment. Single sample estimates of theophylline clearance, CL/F, similarly decreased from 0.040 (+/- 0.014) 1 h-1 kg-1 to 0.018 (+/- 0.008) 1 h-1 kg-1. Mean theophylline clearances were significantly different when comparing control with ciprofloxacin treatment means (P less than 0.01), but were not different when comparing single sample vs multiple sample clearances for a given treatment (i.e. control or ciprofloxacin). It is concluded that a single dose, single sample strategy may be used in screening for host-factor influences on theophylline clearance.

Multiple-dose pharmacokinetics of intravenously administered cefoperazone and sulbactam when given in combination to infected, seriously ill, elderly patients

The pharmacokinetics of cefoperazone and sulbactam in combination were evaluated in six, elderly, seriously ill patients treated with the drug combination for intra-abdominal infections. After giving informed consent, three males and three females aged 63.5 to 77.5 (mean 67.9) years and weighing 54.5 to 86.8 (mean, 67.6) kg were treated with cefoperazone (2.0 g) and sulbactam (1.0 g) infused intravenously every 12 h for at least 5 days. Cefoperazone and sulbactam pharmacokinetics were characterized on both days 1 and 5 of treatment. Eleven serial blood samples were obtained just prior to and following dose 1 on days 1 and 5 of treatment. Mean estimates of cefoperazone maximal concentration in plasma (Cmax), area under the curve of drug concentration in plasma versus time (AUC), half-life (t 1/2), apparent volume of distribution by the area method (Varea), apparent volume of distribution at steady state (Vss), and total body clearance (CL) for day 1 (day 5) were 297.5 237.5) micrograms/ml, 1,247 (1,063) micrograms.h/ml, 7.0 (4.9) h, 16.1 (13.4) liter, 13.1 (14.4) liter, and 28.9 (34.2) ml/min, respectively. Day 1 (day 5) mean values for sulbactam Cmax, AUC, t 1/2, Varea, Vss, and CL were 110.3 (78.0) micrograms/ml, 228 (217) micrograms.h/ml, 3.4 (2.5) h, 26.1 (18.5) liter, 18.9 (15.4) liter, and 97 (94) ml/min, respectively. Both drugs evidenced slower elimination and greater pharmacokinetic variability in these patients compared with values previously reported for normal volunteers. As patients improved during the course of therapy, the only pharmacokinetic parameter significantly changed between days 1 and 5 was a shortened sulbactam t 1/2. Our inability to find substantial evidence of pharmacokinetic normalization may have been related to sample size and study duration. Both drugs were present in potentially therapeutic concentrations for the entire 12-h dosing interval, but without undue accumulation from days 1 to 5.

The use of single sample clearance estimates to probe hepatic drug metabolism in rats. I

1. Conditions were examined under which estimates of drug clearance made from a single measurement of plasma concentration effectively represented multisample estimates of clearance. When plasma concentrations were measured at various post-dose times, both individual and mean values of single sample clearance estimates, Cl, corresponded closely to multisample clearance estimates, Cl, and significant differences between Cl and Cl could not be detected. 2. Best post-dose sampling times were: theophylline, 6 h; phenytoin, 2 h; valproic acid, 20 min; antipyrine, 4 h; and S-warfarin, 48 h. 3. When theophylline clearance was evaluated by both multisample and single sample experiments during diethyl ether versus urethane anaesthesia, clearances were about 50% slower for ether-anaesthetized rats. This outcome was qualitatively and quantitatively the same regardless of whether single sample or multiple sample clearances were estimated, and single sample theophylline clearances were virtually identical to multisample clearances under both anaesthetic conditions. 4. We conclude that multiple drugs can be potentially useful for probing hepatic drug metabolizing activity in rats when using a single plasma measurement to estimate clearance. An appropriate array of such probes might effectively be used to handprint host-factor influences on drug metabolizing activity.

The use of single sample clearance estimates to probe hepatic drug metabolism in rats. II

1. Single sample clearance estimates, Cl, were calculated for each of five drugs employed as probes of hepatic drug-metabolizing activity in rats. Probe drugs were theophylline, phenytoin, valproic acid, antipyrine, and S-warfarin. Cl values were calculated for each probe in animals pretreated with phenobarbital, isosafrole, beta-naphthoflavone, or clofibrate. Control animals were pretreated with vehicle only. 2. A clearance index (c.i., probe Cl after pretreatment divided by probe Cl control) was calculated for each probe and each pretreatment regimen, and data were consolidated to give different probe-based handprints of the pretreatment effects. 3. S-Warfarin was the least specific probe as its c.i. was greater than 1.0 subsequent to each pretreatment. Theophylline appeared to be the most selective probe since its c.i. deviated significantly from unity (3.56) only after beta-naphthoflavone pretreatment. Phenytoin exhibited c.i. values less than unity after each pretreatment indicating that it may not, when used as a single sample probe of hepatic drug-metabolizing activity, effectively discriminate between inductive or inhibitory effects of xenobiotics. 4. Multi-probe-based handprints of hepatic drug-metabolizing activity structured from simple single sample estimates of probe clearance have potential in the rapid screening of xenobiotic-induced alterations of drug-metabolizing enzyme activity.

Use of limited laboratory-acquired therapeutic drug monitoring data in determining individualized drug requirements

Routinely acquired therapeutic drug monitoring (TDM) data from 220 patients were used to estimate patient-specific pharmacokinetic parameters for the following drugs: aminoglycoside antibiotics (gentamicin and tobramycin), digoxin, theophylline, carbamazepine, procainamide, phenobarbital, and quinidine. A microcomputer based set of algorithms operating on two relatively unconstrained TDM values estimated pharmacokinetic parameters with which future TDM levels were forecast. Mean prediction errors (mpe) and root mean squared errors (rmse) were used as measures of predictive performance. Values of mpe deviated from zero by less than one microgram per l for digoxin and by less than one microgram per l for all other drugs. Values of rmse were also small when viewed in the context of the respective therapeutic plasma concentration ranges.

Use of three probes to assess the influence of sex on hepatic drug metabolism

Clearances of phenytoin (PHT), ethosuximide, and theophylline were estimated by a single-dose, single-sample strategy in healthy, young adult men and women. PHT concentrations were measured in salivary ultrafiltrates, ethosuximide concentrations were measured in saliva, and theophylline concentrations were measured in plasma. Estimates of the clearances of these three probes of hepatic drug-metabolizing enzymes revealed a trend toward slightly lower clearances among women compared to men: mean PHT clearance 15% lower, ethosuximide clearance 27% lower, and theophylline clearance 14% lower. However, only differences in ethosuximide clearances were statistically significant. The use of oral contraceptive steroid (OCS) drugs on PHT and ethosuximide clearance was examined, and no significant effects were detected.

Inability of ibuprofen to alter single dose phenytoin disposition

The effect of ibuprofen treatment (1600 mg day-1 per os for 7 days) on single dose (300 mg per os) dispositional parameters of phenytoin (DPH) was evaluated in 10 healthy males. Ibuprofen slightly increased the free fraction of DPH from mean (+/- s.d.) value of 6.9 (+/- 0.7)% to 7.8 (+/- 0.3)% (P less than 0.01). DPH total clearance (CL/F) was not significantly changed going from a value of 0.027 (+/- 0.009) 1 h-1 kg-1 to 0.032 (+/- 0.014) 1 h-1 kg-1 (0.05 less than P less than 0.1). Intrinsic unbound DPH clearance (CLuint/F) was unaffected by ibuprofen treatment. Neither DPH's apparent volume of distribution (V/F) nor the apparent volume of distribution of unbound DPH (Vunb/F) were significantly altered by ibuprofen treatment. The slight changes in DPH plasma protein binding that are due to concurrent ibuprofen treatment in doses of 1600 mg day-1 are not likely to produce important changes in DPH dispositional parameters.

Dispositional and pharmacodynamic characteristics of brodifacoum in warfarin-sensitive rats

A dose-response curve for the hypoprothrombinemic effect of brodifacoum 3-[-3(4'-bromobiphenyl-4-yl) 1,2,3,4-tetrahydronaphth-1-yl] -4-hydroxycoumarin, was constructed using doses ranging from 0.1 to 0.33 mg/kg. Brodifacoum exhibited a remarkably steep dose-response curve. Brodifacoum failed to exhibit a dose-dependent effect on the degradation rate constant (kdeg) for prothrombin complex activity (PCA) after a PCA-synthesis-blocking dose of warfarin. Both phenobarbital pretreatment and SKF525A treatment altered to anticoagulant response to brodifacoum. Phenobarbital decreased the anticoagulant effect, whereas SKF525A increased it, suggesting that a substantial portion of brodifacoum-induced hypoprothrombinemia is mediated by brodifacoum itself rather than by metabolites. Finally, rats dosed orally with brodifacoum (0.2 mg/kg p.o.) were sacrificed in groups of 3-5 at various times up to 120 h after the dose. Brodifacoum was assayed in serum, small intestine, and liver by an HPLC method. Brodifacoum disappeared slowly from serum with a half-life of 156 h. Disappearance from small intestine was rapid, for 24 h, but intestinal levels began increasing from 24 to 72 h after the dose. Concentrations in liver were rapidly established, and exceeded serum concentrations by 20-fold. Brodifacoum levels in liver remained relatively constant for 96 h. Sustained liver concentrations after a single dose may partially account for brodifacoum's apparent potency relative to warfarin.

Impact of cimetidine on the pharmacokinetics of theophylline

The effect of cimetidine on the pharmacokinetics of theophylline was evaluated. Twelve healthy men volunteers participated in this two-way crossover study. Six subjects were initially treated with theophylline (control), and six were treated with cimetidine and theophylline (cimetidine). Sixteen days after the start of the first phase, the subjects were crossed over. Aminophylline equivalent to 2 mg/kg theophylline was dissolved in 100 ml 5% dextrose injection and infused into an antecubital or forearm vein over a 20-minute period. Cimetidine was ingested in 300-mg doses every six hours beginning 48 hours before the theophylline infusion and continued until 48 hours after the infusion. Blood samples were collected periodically for 48 hours, and serum theophylline concentrations were determined by liquid chromatography. Theophylline concentration versus time data were computer fitted to a biexponential equation. Total theophylline clearance (CLtot), apparent volume of distribution (V), and elimination half-life (t 1/2 beta) were calculated. Semilogarithmic plots of theophylline disappearance showed a decrease during cimetidine treatment. Mean theophylline t 1/2 beta increased significantly during cimetidine treatment from 5.9 +/- 1.0 to 7.3 +/- 1.5 hr, and CLtot decreased significantly from 0.065 +/- 0.010 to 0.049 +/- 0.012 liter/hr/kg. Mean V was unchanged. Two subjects failed to exhibit decreased theophylline clearances, and theophylline clearance increased in another subject. Cimetidine was found to slow the clearance of theophylline and to extend its half-life. Daily doses of theophylline may require adjustment in patients requiring the coadministration of cimetidine. Cimetidine should be added with caution, and serum theophylline concentrations should be monitored accordingly.

Convenient synthesis of 6-nor-9,10-dihydrolysergic acid methyl ester

6-Nor-9,10-dihydrolysergic acid methyl ester (IV) was prepared by demethylation of 9,10-dihydrolysergic acid methyl ester (II) with 2,2,2-trichloroethyl chloroformate, followed by reduction of the intermediate carbamate (III) with zinc in acetic acid. The 6-ethyl-V and 6-n-propyl-VI derivatives were prepared by alkylation of IV with the appropriate halide. All of the ergoline derivatives were evaluated for stereotyped behavior in rats, with 6-nor-6-ethyl-9,10-dihydrolysergic acid methyl ester (V) being active but much less potent than apomorphine. Compound VI was evaluated for its effect on blood pressure; at a dose of 30 mg/kg ip, it significantly lowered, diastolic pressure in normotensive rats.

Effect of phenylbutazone on the metabolic disposition of warfarin in the dog

Phenylbutazone has been shown to modify the physiologic disposition of warfarin in the dog. Dogs pretreated with phenylbutazone for a week, and permitted to remain drug-free for an additional week, eliminated warfarin from plasma 2-3 times faster than non-pretreated controls with an average half-life of 7.8 hours. In-vitro hepatic microsomal metabolism of warfarin and cytochrome P-450 content were greater for hepatic microsomes derived from phenylbutazone pretreated animals.

Phenylbutazone-warfarin interaction in the dog

The administration of phenylbutazone together with warfarin to dogs resulted in an elevation of the free fraction of warfarin in the plasma from 2-6 to 8-0% thus providing direct support for the notion that phenylbutazone induced inhibition of warfarin binding to plasma proteins. This inhibition as evaluated by a kinetic method was accompanied by a two-fold decrease in the plasma half-life of warfarin from 18-4 h in control animals to 9-6 h in phenylbutazone-treated animals. Marked increases in warfarin-induced hypoprothrombinaemia were observed when at doses up to 8 mg kg-1 (orally) it was given with phenylbutazone (50 mg kg-1, orally). The unbound fraction of warfarin in canine plasma ranged from 1-7 to 4-3% indicating individual differences in the extent of the plasma binding of warfarin in the dog.

Topical application of a chlordane-containing ectoparasiticide: effect on the plasma half-life of warfarin in dogs

Brief exposure of dogs to topical chlordane solutions resulted in a significant and long-lasting decrease in the biological half-life of orally administered warfarin. The effect is presumed to be an expression of chlordane's well-documented inductive effect on hepatic microsomal drug metabolizing enzymes and its long-term storage in fat depots. The facility with which chlordane is absorbed through the intact skin of dogs may render casually-treated animals unsuitable for subsequent pharmacologic study for long periods of time.