Digoxin, a drug of low therapeutic index, is commonly prescribed in the elderly. In most patients more than 80% of digoxin is excreted unchanged in the urine. Interactions with drugs which affect digoxin renal clearance have been identified, some of which may be due to effects on tubular secretion. As there is the potential for an interaction between digoxin and aciclovir following oral valaciclovir, we have studied the pharmacokinetics of the drugs alone and in combination in healthy volunteers.

Subjects

Twelve healthy volunteers (seven males, five females) of mean age 31  years (range 22–44 years) and mean weight 75  kg (range 51–99  kg) participated in the study. Exclusion criteria included evidence of a cardiac conduction disorder on a 12-lead ECG and an estimated creatinine clearance <70  ml  min⁻¹. All subjects gave written informed consent, and the protocol was approved by the Wellcome Protocol Review Committee and the King’s Healthcare Research Ethics Committee.

Study design

This open, randomized, four-period crossover study was carried out at the Wellcome Clinical Investigations Unit, King’s College Hospital, London. Volunteers fasted overnight prior to drug administration on the first blood sampling day. Oral valaciclovir, 1000  mg, was administered on one occasion, and on another, with digoxin, 0.75  mg being given 12  h before and 0.75  mg being given with valaciclovir. Digoxin was not dosed to steady state because of the ethical concerns of giving multiple doses to healthy volunteers. Blood was sampled before and 15, 30, 45 and 60  min and 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, 6.0, 8.0 and 12  h after the valaciclovir dose for plasma aciclovir assay. Urine was also collected up to 12  h, weighed and a 10  ml aliquot was frozen at −20°  C for aciclovir assay.

Digoxin was administered as 2×0.75  mg doses alone on one occasion, and on another, with oral valaciclovir, 1000  mg three times daily for 8 days, with valaciclovir started 12  h before the first digoxin dose. The valaciclovir daily dose was that recommended for the treatment of shingles. Blood samples were taken before and 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 6.0, 8.0, 12, 24, 32, 48, 72, 96 and 168  h following the second digoxin dose for plasma digoxin assay. Urine was collected up to 24  h after the second digoxin dose, weighed, and a 10  ml aliquot was frozen at −20°  C for digoxin assay.

Adverse experiences were recorded, 12-lead ECGs were performed before dosing and at 24  h after the second digoxin dose, and continuous lead II ECG monitoring was performed on all subjects for 24  h after digoxin doses.

Assays

Plasma and urine aciclovir determinations were made using double antibody radioimmunoassay (r.i.a.), which is a modification of the original method. The lower limits of quantitation (LLOQs) were 0.01  μg  ml⁻¹ (0.04  μm ) for plasma and 0.025  μg  ml⁻¹ (0.11  μm ) for urine, with separate inter- and intra-assay precision shown by coefficients of variation (CVs) of <10% for plasma and <11% for urine.

Plasma and urine digoxin determinations were made using a competitive r.i.a. procedure (Phoenix International Life Sciences Inc.). The LLOQ was 0.10  ng  ml⁻¹ for both plasma and urine, with combined inter- and intra-assay precision shown by CVs of ≤10% for plasma and ≤12% for urine.

‘bh‘rr‘cc20p0,2,10.5,6\‘el7\

‘cl12\Pharmacokinetic and statistical analysis

‘ol17\‘cc20p0,1,10,12\‘eh‘nd8\‘tageqend‘tagNon-compartmental pharmacokinetic parameters were determined for plasma aciclovir and digoxin. The area under the plasma concentration-time curve from zero to the last measurable plasma concentration AUC(0, t ) was estimated by the linear trapezoidal rule. AUC(0, ∞) was calculated as AUC(0, t )+Ct/λ_z where C_t is the last quantifiable concentration at time t. λ_z was obtained by log linear regression, using the terminal portion of the log of the concentration-time curve. Digoxin C_max and AUC(0, ∞) were corrected for pre-dose concentration (C₀ ) as the plasma profile after the second dose was not at steady state (half-life of approximately 40  h). Digoxin AUC(0, 24  h ) was also measured, but was not corrected. The elimination half-life (t_1/2 ) was calculated as ln 2/λ_z. The apparent volume of distribution, V_z/F, was calculated as Dose. AUC(0, ∞)/λ_z. The aciclovir dose was calculated from the molar equivalents in the valaciclovir dose. Urinary recovery (Ae) was calculated as the product of urine concentrations and weights (assuming density of 1  g  ml⁻¹ ). Renal clearance (CL_R ) was calculated as Ae/AUC where Ae is the amount of drug excreted unchanged in urine, and AUC and Ae were measured over 12 and 24  h for aciclovir and digoxin respectively.

Pharmacokinetic parameters following digoxin or valaciclovir alone or in combination were subjected to analysis of variance. All parameters except t_max were log-transformed prior to analysis. Data were back-transformed to provide a point estimate with 95% confidence interval (CI) for the ratio in pharmacokinetic parameters between treatments. Differences between t_max medians were estimated (with 95% CIs) using a method based on the Wilcoxon Signed Rank test.

The search for alternative routes of cobalamin administration began soon after cyanocobalamin was isolated and introduced for parenteral use in 1948. Nasal inhalation, insufflation or instillation of cyanocobalamin were proposed in the early 1950. The formulations for nasal inhalation or instillation consisted of low concentrations of cyanocobalamin in isotonic saline solution or lactose powder. Although these formulations were reported to be effective in the treatment of pernicious anaemia, none of these proposals found a follow-up in clinical practice, in industry or in the scientific literature. Recently a gel for nasal cyanocobalamin instillation has become commercially available. The gel, in a dose of 400–500  μg cyanocobalamin, has been claimed to be safe and effective. We are not aware of studies that show the safety and efficacy of the nasal gel in treating cobalamin deficiency. A practical problem with the use of a gel is that it sometimes dries up and sticks to the unit-dose tubes. In the sixties oral cyanocobalamin administration for the treatment of pernicious anaemia was proposed. Despite the claim that doses of 150  μg to 1000  μg daily are effective, it is rarely used.

A disadvantage of all these forms of treatment is that they contain cyanocobalamin while hydroxocobalamin is the drug of choice. Since the fifth World Health Organization Model List of Essential Drugs, hydroxocobalamin, and not cyanocobalamin, has been the cobalamin included in the list. Hydroxocobalamin binds more extensively to plasma proteins and has a longer half-life in the body than cyanocobalamin. As a result, hydroxocobalamin is better retained in the body and therefore requires less frequent dosage. Moreover, cyanocobalamin is contraindicated in patients with tobacco and tropical amblyopia and optic neuropathy in pernicious anaemia. A plea for the withdrawal of cyanocobalamin has been made.

Recently, a formulation for the nasal administration of hydroxocobalamin has been developed. Here we present the results of our exploratory study on the absorption of nasally administered hydroxocobalamin in healthy elderly adults. To our knowledge this is the first report to document nasal hydroxocobalamin absorption in humans.

Subjects

Healthy elderly adults were recruited from the general population by means of advertisements in local newspapers. The following exclusion criteria were applied: use of vitamin supplements containing cobalamin, use of nasal medication, or acute or chronic rhinitis. Ten subjects were selected, six females and four males, age 71.6±4.6 year (mean±s.d.).

Protocol

The study protocol was approved by the Committee for Experimental Research with Humans of the University Hospital Nijmegen. All subjects gave written informed consent. On the first test day subjects were randomly allocated to a dose of 750  μg or 1500  μg hydroxocobalamin. On the second test day (after 163±10 days, range 147–175 days) subjects received the alternate dose. Blood samples were obtained through an indwelling cannula inserted into a forearm vein of each subject. After a blood sample was collected the cannula was flushed with 1  ml of heparinized salt solution (150 IE heparin in 0.9% NaCl solution) in order to maintain patency. Before each blood sample was collected the first 1–2  ml blood were discarded. EDTA-plasma was collected before and after drug administration at the following time points: 0, 10, 20, 30, 40, 60, 120, 180 and 240  min.

The hydroxocobalamin formulation contained 750  μg hydroxocobalamin per 70  μl in a preserved solution and the nasal spray device was a metered pump (Valois, France, VP 7/70). The investigator administered 750  μg (one puff) or 1500  μg (one puff in each nostril) hydroxocobalamin nasally. The nasal dispenser was weighed before and after administration of the drug to ensure that the dose had actually been given.

The plasma cobalamin concentration was determined by competitive radioisotope binding technique using purified hog intrinsic factor as cobalamin binder (Solid Phase DualCount, Diagnostic Products Corporation, Los Angeles, California). The within-assay coëfficient of variation was 3.1–3.5% and the between-assay coëfficient of variation was 3.3–6.7%. The assay has a detection limit of 37  pmol  l⁻¹. In case the plasma cobalamin concentration fell outside the range of the assay, i.e. >1700  pmol  l⁻¹, the samples were diluted with zero-diluent and assayed again.

The study was based on information derived from the General Practice Research Database (GPRD). Since 1987, over four million residents in the United Kingdom have been enrolled with selected general practitioners (GPs) who use office computers provided by Value Added Medical Products (VAMP Health) and have agreed to provide data for research purposes to the GPRD which is currently owned by the United Kingdom Department of Health. The GPs received 12 months of instruction on the standardized recording of medical information and they agreed to supply anonymized information to academic researchers on an ongoing basis. The information recorded includes patient characteristics, drugs dispensed, clinical diagnoses, notation of referrals to consultants, hospitalizations, certain historical information and other findings, (e.g. smoking status, blood pressure, height and weight). Referral letters from consultants and hospitalizations are kept in a manual file. The general practitioners generate prescriptions directly from the computer, and these are automatically transcribed into the patient’s computer record. The details of each prescription including dose, instructions and quantity are automatically recorded on computer and can be used to determine dose and duration of drug exposure. A modification of the Oxford Medical Information System (OXMIS) classification is used to enter medical diagnoses, and a coded drug dictionary based on the Prescription Pricing Authority’s dictionary is used for the recording of prescriptions. (OXMIS codes were mapped onto ICD codes for the purpose of this study.) Two large validation studies determined that information on all patient referrals and hospitalizations present in the manual medical records in the general practitioners’ offices was recorded on the computer over 90% of the time.

Subjects

This was a case-control study. The base population from which the data were derived was restricted to subjects who used a quinolone antibiotic (n=142 316) at some time between January 1, 1991 and April 30, 1995. The quinolones evaluated were ciprofloxacin, norfloxacin, ofloxacin, nalidixic acid, cinoxacin, enoxacin, acrosoxacin and temafloxacin. In order to be included, a subject had to have at least 18 months of information on drugs prescribed and diagnoses recorded on computer prior to the date of diagnosis. Subjects with a prior diagnosis of degenerative neurologic disease (i.e., stroke, multiple sclerosis), those with chronic liver or renal disease and those who were recorded as being drug addicts were excluded from the study.

Three groups of subjects aged 15–84 years of age were identified from the base population of quinolone users and formed the case groups: (1) persons who committed suicide (OXMIS code 3009D), (2) persons who had a diagnosis of attempted suicide (OXMIS codes L3009P, parasuicide; 9779L, drug overdose suicidal; and 3009C, attempted suicide) and (3) persons who had a diagnosis of suicidal ideation (OXMIS codes 3009BW, suicidal ideation; 3009BT, suicidal thoughts; and 3009BP, suicidal plans). If a subject had more than one of the ‘case’ diagnoses during the study period, only the first such diagnosis was considered. The index date from which exposure was determined was the date that the diagnosis was made.

Controls were derived from an initial random sample of 1000 subjects in the base population who did not have one of the ‘case’ diagnoses during the study period. The same exclusion criteria applied to cases were applied to the controls. A random index date between January 1, 1991 and April 1995 was assigned to each control subject and was considered to be the index date from which exposure was determined.

Based on the suggestion that angiotensin II (ANG II) is the primary antagonist of NO-mediated vasodilation in the balance of factors regulating renal perfusion and the well established role of ANG II in the regulation of renovascular tone in healthy subjects and patients with hypertension, we hypothesised that exogenous l-arginine could influence an ANG II-mediated renal vasoconstriction if counterregulatory activation of NO synthesis is limited by substrate availability. We have therefore studied the effect of pretreatment or concomitant administration of l-arginine on increased renovascular resistance induced by exogenous ANG II-infusion in healthy male subjects.

Subject population

After approval of the study protocol by the Ethics Committee of Vienna University School of Medicine and written informed consent, 17 healthy, non-smoking, drug-free male volunteers between 20 and 32 years of age were studied. Each subject passed a screening examination that included history and physical examination, 12-lead electrocardiogram, complete blood cell count with differential, 24  h creatinine clearance and urinalysis, urine drug screen, serum electrolytes, bilirubin, BUN, creatinine, cholesterol, triglycerides, γ-glutamyltransferase, glucose, lactate dehydrogenase, ASAT, ALAT, total protein, activated partial thromboplastin time, thrombin time, hepatitis A, B, C serologic tests and human immunodeficiency virus antibody tests.

Study protocol

The study was conducted according to two different trial protocols. Both protocols followed a double-blind, randomised, two-way cross-over design with washout periods between study days of at least 5 days.

The study had a randomized cross-over design and three frusemide doses of 10, 25 and 40  mg were administered on separate study days with intervals of at least 1 week. To standardize experimental circumstances, no medications were allowed within 1 week before each study day. The subjects were asked to refrain from alcohol and extreme physical activity for 3 days before the start of each experiment. Standardized meals were provided the day before and during each study day with a total content of 159  mmol sodium and 81  mmol potassium  day⁻¹ and caffeinated drinks such as coffee or tea were not allowed during this time. Urine was collected for 24  h on the day before each study day to assess adherence to the diet and to have an estimation of basal diuresis. The subjects fasted overnight and the study started in the morning at the closure of the 24  h urine collection. A cannula was inserted into an antecubital vein of each arm and blood samples were taken to measure basal levels of plasma sodium, potassium and chloride. Then, each subject emptied his bladder after which the administration of frusemide was started (at 0  h). A rapid infusion of 10, 25 or 40  mg of frusemide (Furix^®, Nycomed Pharma, Oslo, Norway), diluted with saline solution to a total volume of 10  ml, was given intravenously over 5  min. The subjects provided urine samples by voiding at 5  min intervals for the first hour and at 15  min intervals for another 4  h after dosing. Urine losses of each period were replaced volume for volume using an intravenous isotonic rehydration fluid to prevent depletion of volume and electrolytes. This solution was prepared by the hospital pharmacy and contained 0.45% NaCl and 2.5% glucose. The fluid was administered using two to four infusion pumps, depending on the urinary volume produced during the preceding interval (IMED 960 volumetric infusion pump with a 9200 accuset, Imed Corporation, San Diego, USA). The subjects remained fasting throughout the 5  h study period. Blood samples were taken to measure plasma sodium, potassium and chloride at 2.5  h and 5  h after dosing. Plasma samples were stored at −20°  C. The urine volumes were weighed and aliquots were carefully protected from light and stored in plastic tubes at −70°  C, until analyzed for frusemide and sodium.

Clearly, minimising PONV is an important clinical goal and most strategies to alleviate it have revolved around the use of antiemetic medication, especially in patients receiving opioid analgesics for postoperative pain management. It has long been speculated that rapid changes in blood opioid concentration are more likely to cause unpleasant side-effects than more gradual changes in concentration but simple manoeuvres to alleviate PONV, such as altering the mode of analgesic drug administration, appear not to have been investigated. Some devices used for PCA allow the duration over which a dose is delivered to be either ‘bolus’ (usually delivered over around 40  s) or over a longer period, typically 5 min. Increasing the bolus dose delivery duration time will decrease the rate of change of blood drug concentration and the maximum blood-drug concentration (C_max ) whilst prolonging the time to C_max (t_max ). This study was performed to test the hypothesis that increasing the duration of delivery of morphine would decrease the incidence of PONV.

This work was presented as a poster at the 11th World Congress of Anaesthesiologists, Sydney April 17, 1996.

Forty-eight patients aged between 34 and 74 years (mean age 47 years; s.d.=8) undergoing total abdominal hysterectomies were selected for the study. All patients approached agreed to participate. All patients had similar systemic pathology (ASA 1–2) and had renal and liver functioning within normal physiological limits. Patients who could not understand the English language or the concept of PCA, those with a history of mental illness and those receiving monoamine oxidase inhibitors (MAOI) within 14 days prior to the study were excluded from selection. Four patients were withdrawn from the study: one patient was not using PCA because she had very little pain, one patient disliked morphine and requested pethidine, one patient underwent a vaginal rather than abdominal hysterectomy, and one patient underwent a laparotomy. Two patients received intra- or post-operative ketorolac, a third patient received ondansetron and another patient had insufficient data recordings taken. These patients were not included in the statistical analysis.

Most clinically used α₁-adrenoceptor antagonists, e.g. terazosin, were originally developed for the treatment of arterial hypertension and do not discriminate α₁-adrenoceptor subtypes. In contrast, the novel antagonist, tamsulosin, was specifically developed for the symptomatic treatment of benign prostatic hyperplasia and has about 15-fold selectivity for α_1A- over α_1B-adrenoceptors with intermediate affinity for α_1D-adrenoceptors. Interestingly placebo-controlled clinical trials with terazosin have reported larger incidences of unwanted side effects than trials with tamsulosin but a direct comparison has not been performed.

The present study was designed to gain further insight into a potential role of α₁-adrenoceptor subtype selectivity in the action of tamsulosin and terazosin. This was done by a pharmacokinetic analysis of receptor binding following a single, oral dose of tamsulosin (0.4  mg) and terazosin (5  mg) in a placebo-controlled, single-blind, randomized, three-way cross-over study. For this purpose we have developed a new radioreceptor assay using cloned human α₁-adrenoceptor subtypes stably expressed in rat-1 fibroblasts. With this assay we have assessed binding to each of the three subtypes at 1, 3, 5, 7, 10 and 23.5  h following drug intake. Specific analysis of tamsulosin and terazosin plasma concentrations by specific h.p.l.c. analysis was performed in comparison.

The study protocol was approved by the ethics committee at the University of Essen Medical School. Ten healthy male subjects (median age: 26.5 years, range: 22–36 years) participated after having given informed written consent. Each subject completed 3 study days during which they received in a single-blinded, randomized cross-over manner 1 tablet of 5  mg terazosin (purchased as Flotrin^® from a German pharmacy), 1 capsule of 0.4  mg tamsulosin modified release formulation (Omnic^®, provided by Yamanouchi Europe B.V., Leiderdorp, Netherlands), or 1 capsule of placebo matching the tamsulosin capsule. Study days were at least 7 days apart to secure sufficient drug washout.

On each study day the subjects reported to the laboratory at 07.00  h after an overnight fast. An indwelling catheter was placed into a forearm vein for blood withdrawals. The subjects remained in the supine position until after the last blood withdrawal 24  h after drug ingestion. Blood samples were taken 2  h before and 1, 3, 5, 7, 10 and 23.5  h after drug intake. The blood samples were collected into EDTA-coated tubes to prevent coagulation. They were stored on ice immediately. Plasma samples were generated by centrifugation and stored at −20°  C until analysis, i.e. up to 4 months. Each blood sample was used for h.p.l.c. and radioreceptor analysis. A light snack was allowed after the 7  h blood withdrawal and a pizza or pasta dish after the 10  h withdrawal. While no other food was given until after the last blood withdrawal, drinking of water was allowed ad libitum.

Progress in surgical treatment requires improvement of perioperative management. Percutaneous transhepatic portography (PTP) and the insertion of portal vein catheters are useful procedures in liver surgery, and make possible clinical pharmacokinetic assessment of liver function.

We hypothesized that the clearance of ICG following portal vein injection (CLpv) would be more useful than that after peripheral vein injection (CLiv) for evaluating the intrinsic clearance of ICG and hepatic blood flow.

Eight patients were studied. The study was approved by the Institutional Human Research Committee of Osaka City University Hospital, and informed consent was obtained from each patient. Preoperative laboratory data revealed mild liver dysfunction or cirrhosis in each patient. Hepatic cirrhosis was confirmed by histological examination of liver specimens.

Each patient underwent PTP under local infiltrative anaesthesia supplemented by intravenous analgesics (Stage I). ICG (25  mg) was injected into a peripheral vein before PTP, and then into the portal vein after PTP. ICG was given as a bolus with at least 30  min separating the two injections. Arterial blood samples were withdrawn through the radial artery for h.p.l.c measurement of ICG concentrations in plasma at 2, 5, 7, 10, 12, 15, 17, 20, 25, and 30  min after each injection. One week later, enflurane was given to patients on the day of operation. Secobarbitone (100  mg) and 0.5  mg of atropine sulphate were given intramuscularly 1  h before operation. Anaesthesia was induced with 5  mg  kg⁻¹ of thiopentone and patients were intubated using 0.1  mg  kg⁻¹ of vecuronium bromide. The muscle relaxant was supplemented during operation when necessary, and its effect was reversed using 2  mg of neostigmine with 1  mg of atropine. No other medication was given perioperatively to any patient. ICG (25  mg) was given as a bolus into a branch of the portal vein (a mesenteric vein) and then into a peripheral vein during anaesthesia, 30  min apart between two injections (Stage II) and after partial hepatectomy (Stage III). Twelve hours after surgery, the same amount of ICG was administered via a peripheral vein to all patients, and also into the portal vein in three of the eight patients (Stage IV). Blood samples were obtained using the same schedule as during PTP.

Non-compartmental analysis was applied to the ICG time-concentration data obtained. The concentration measured after the second injection was corrected by three steps, since the drug remaining after the first injection might have affected the concentration of the drug after the second injection. The first step was to determine whether the time-concentration curve after the first injection was first-order. The second step was to calculate the concentration remaining after the first injection just before the second injection. The third step was to subtract the estimated residual after the first injection from the observed concentration at each time point after the second injection. The area under the curve (AUC) after the injection was calculated by adding two areas, the area from the injection to the last measured point according to the trapezoidal rule, and that from the last measured point to infinite time by extrapolation using the seven points before the final one. The clearance (CL) was obtained by dividing the dose by the AUC. The mean residence time (MRT) was calculated as the second moment of the time-concentration data. The volume of distribution (V  ) was calculated by multiplying CL by MRT. Plasma clearance was converted to whole blood clearance using the haematocrit for each period of ICG injection. The extraction ratio was calculated as 1 minus the quotient of the AUCs.

Pharmacokinetic parameter data were statistically analyzed using ANOVA with Scheffe’s F-test.

Cholestyramine is an anion exchange resin with lipid-lowering properties. Taken orally, it is not absorbed from the gastrointestinal tract but is excreted unchanged in the faeces. Cholestyramine lowers LDL-cholesterol by binding to bile acids and preventing their enterohepatic circulation.

Therefore, because of the physicochemical properties of troglitazone, investigations were undertaken to determine the influence of concomitant administration of cholestyramine on the pharmacokinetics of troglitazone.

Preclinical studies

An in vitro study using [¹⁴C]-troglitazone, in conditions mimicking the small intestine in vivo, was conducted to determine the adsorption of troglitazone by cholestyramine. [¹⁴C]-troglitazone was dissolved in ethanol to give a solution used to spike 3 and 500  μg  ml⁻¹ incubates of solutions of human serum albumin (HSA) prepared in Sorensen’s buffer pH  6. HSA was used to increase the aqueous solubility of the drug and to mimic food proteins in the gut. The percentage loss (P) of soluble radiolabel from each test incubate compared with the appropriate control was used as a measure of absorption. This was determined from the following equation: where Rc=mean radioactivity concentration in the control incubate supernatant (d  min⁻¹/200  μl) and Rt=mean radioactivity concentration in the corresponding test incubate supernatant (d  min⁻¹/200  μl).

Cholestyramine (4×1  g) was administered once, 1  h prior to receiving a single oral tablet of troglitazone 200  mg, and on three occasions 1, 2 and 4  h post-dose, as an aqueous suspension (1  g in 10  ml water) by gavage to 11 Beagle dogs. Blood sampling was carried out from the cephalic vein at 0.5, 1, 2, 3, 4, 6, 8, 10 and 24  h following administration. Plasma samples were assayed by reversed phase h.p.l.c. with u.v. detection as previously described. A paired t-test was used to compare previously obtained (control) AUC (area under the plasma concentration time curve), C_max (maximum observed plasma concentration) and t_max (time at which C_max was reached) values for each dog with corresponding values obtained from this study.

Clinical study

Study methodology has been described elsewhere. Twelve healthy subjects each received a single oral dose of troglitazone 400  mg alone and with cholestyramine 12  g (3×4  g sachets) in random order as part of an open, two-way crossover study. Troglitazone was taken 30  min after a standardized breakfast and cholestyramine was given as an aqueous solution 1  h after ingestion of troglitazone. Blood samples were taken pre-dose and at 15, 30, 45, 60, and 90  min and 2, 3, 4, 6, 8, 12, 24, and 48  h post-dose.

Pharmacokinetic parameters, C_max, t_max, t_1/2 (elimination half-life) and AUC were calculated for troglitazone and its main metabolites as described previously.

The normal parametric method was not used in these analyses because one subject was reported with no absorption of troglitazone and therefore, a zero value for AUC; thus log transformation was not possible. The pharmacokinetic parameter values for AUC, C_max, and t_max for troglitazone, sulphate and quinone metabolites were analysed using Koch’s non-parametric method based on the Wilcoxon Rank Sum test. An estimate of the median difference between troglitazone with cholestyramine and troglitazone alone, together with a 95% confidence interval was calculated. Subjects for whom it was not possible to calculate t_max due to poor absorption were excluded from the analysis of t_max. The analysis of t_1/2 for troglitazone and sulphate metabolite was performed following log transformation using analysis of variance allowing for subjects, periods and treatments. Subjects for whom it was not possible to calculate t_1/2 due to a poorly defined terminal elimination phase were excluded from the analysis of t_1/2.

Troglitazone is a white to pale yellow crystalline powder (pK_a 6.1, 12.0) which is practically insoluble in water (solubility approximately 0.02  mg  ml⁻¹ ). It has an oil/water partition coefficient of 2.4 at pH  7 and is absorbed largely via the small intestine. Following oral administration, troglitazone is primarily metabolised to a sulphate conjugate and an oxidative quinone metabolite. A glucuronide conjugate is also formed. Troglitazone has a half-life of approximately 10–15  h and that of its major metabolite is approximately 20  h [unpublished data]. A high degree of variability in area under the plasma concentration-time curve (AUC) occurs (approximate CV of 30%), [unpublished data].

It is important therefore to assess the effect of food on the gastrointestinal absorption of troglitazone. This study was designed to investigate the effect of concomitant food intake and food intake 30  min prior to dosing on the pharmacokinetic profile of troglitazone.

hirteen healthy males (mean age 31 years, range 20–41 years) were enrolled in this open, randomized, three-way, cross-over study. They weighed between 64.4  kg and 95.2  kg (mean 79.1  kg), and ranged in height from 1.70–1.86  m (mean 1.76  m). Each subject received one tablet of troglitazone (400  mg) following an overnight fast 1) alone, 2) concomitantly at the start of a standardized diabetic breakfast and 3) 30  min after a standardized diabetic breakfast. All treatments were given with water (200  ml) and were separated by a minimum of 7 days. Apart from the standardized breakfast (approximately 64  g carbohydrate, 25  g fat, 20  g protein and 600 calories) of cornflakes (35  g), skimmed milk (150  ml), a slice of wholemeal toast (25  g) and butter (25  g) and an apple, food was not allowed until 4  h after dosing; a light lunch was then provided.

All subjects were free from any significant medical condition, received no regular medication for 4 weeks prior to the study and no drug treatment for 48  h prior to the study. Subjects were not allowed to consume food or drink (apart from water) from 22.00  h on the night prior to the study. Water was permitted up to midnight prior to dosing. Smoking, alcohol and strenuous exercise were not permitted from the evening prior to drug administration to after the last blood sample on each occasion. Subjects returned between 7 and 10 days following the end of the study for laboratory safety tests.

The study protocol was reviewed and approved by the Ethics Review Committee of Glaxo Wellcome Research and Development Ltd. All subjects gave written consent to participate and the study was conducted at the department of Clinical Pharmacology, Glaxo Wellcome Research and Development Ltd, Ware, Hertfordshire, UK in accordance with the provisions of the revised Declaration of Helsinki (1964).

Assessments

Blood samples (5  ml) were taken pre-dose and at the following nominal times post-dose: 15, 30, 45, 60 and 90  min and 2, 3, 4, 8, 12, 24 and 48  h. Blood samples were centrifuged at 1500  g for 10  min and the resultant plasma frozen at −20°  C until required for assay. Plasma samples (300  μl) were assayed by h.p.l.c. with ultraviolet detection (230nm) following liquid/liquid extraction with ethyl acetate:hexane (3ml, 90:10 v/v). The organic phase was dried down under nitrogen and the samples were re-constituted in ethanol (100  μl) prior to injection of an aliquot (30  μl) onto the h.p.l.c. column (YMC-Pack ODS A-314G, 300×6  mm) which was maintained at 35°  C and eluted at a flow rate of 1.2ml  min⁻¹ with a mobile phase consisting of acetonitrile/water/phosphoric acid (60:40:0.08 v/v). Chromatographic peak height ratios of each analyte relative to the internal standard (9-acetylanthracene) were quantified by reference to linear calibration lines prepared freshly on each occasion of assay. A calibration range of 0.1 to 6.4  μg  ml⁻¹ and a limit of quantification of 0.1  μg  ml⁻¹ for both troglitazone and the sulphate metabolite were used. Assay performance was monitored by quantifying three quality control samples in duplicate which were required to be within 15% of their nominal concentrations.

The assay was validated and the intra assay precision ranged from 1.7 to 28% for troglitazone and from 1.1 to 13% for the sulphate with a bias of −12 to +3% for troglitazone and +0.8 to +22% for the sulphate. The inter assay variation was determined for three analyte concentrations on seven occasions and precision varied from 13 to 14.8% for troglitazone (bias +0.4 to +7.8%) and from 4.2 to 16.2% for the sulphate (bias −0.4 to +9.5%). Troglitazone and the sulphate metabolite were stable in plasma for up to 12 months and three freeze-thaw cycles had little effect on their concentrations.

Data analysis

The following parameters were derived for each subject from the plasma troglitazone and sulphate metabolite concentration data. The maximum observed plasma concentration (C_max ) and the time at which C_max was reached (t_max ) were noted directly. For troglitazone the AUC to the last measurable time point (AUC_last ) was calculated by the log-linear trapezoidal rule and the lag-time before appreciable absorption of troglitazone occurred (t_lag ) was recorded as the last timepoint with a value below the limit of quantification prior to absorption. For the sulphate metabolite, the terminal elimination rate constant (λ_z ) was calculated by linear least-square regression using logarithmically transformed points in the terminal phase. The terminal phase plasma half-life (t_1/2 ) was calculated by ln(2)/λ_z. AUC was extrapolated to infinity (AUC_∞ ) by adding the ratio of the last measurable concentration divided by the elimination rate constant. The plasma concentrations of troglitazone were not adequately defined in the terminal phase and therefore λ_z and the corresponding t_1/2 and AUC parameters could not be measured.

The pharmacokinetic parameter values, C_max and AUC for troglitazone and the sulphate metabolite were analysed using analysis of variance allowing for subjects, periods and treatments. Separate tests for treatment by period interaction and for carry-over effects were performed for each of the variables analysed parametrically. Where there was evidence of a carry-over effect, this was included in the statistical model. The pharmacokinetic parameter t_1/2 for the sulphate metabolite was therefore analysed using analysis of variance allowing for effects due to subject, period, carryover and treatment. A log transformation was performed for each parameter in order to satisfy the constant variance assumption for the analysis of variance. Geometric mean values for each of the treatments were evaluated together with ranges. Estimates of pairwise treatment differences were calculated together with 95% confidence intervals.

Values of t_max and t_lag were summarized by the treatment medians and compared between pairs of treatments using the Wilcoxon signed rank test, with corresponding 95% confidence intervals for the estimates of treatment difference derived.

Safety

Adverse events occurring during the study and laboratory parameters for clinical chemistry, haematology and urinalysis were recorded at screening, pre-dose, 48  h post-dose and between 7 and 10 days following the end of the study.