Effect of alosetron on theophylline pharmacokinetics

Abstract

Aims

To examine the potential for alosetron to alter the pharmacokinetics of theophylline by inhibiting its metabolism, as suggested by in vitro and in vivo effects on CYP1A2 activity.

Methods

Ten healthy female volunteers received theophylline 200 mg twice daily alone for 8 days and with alosetron 1 mg twice daily for 15 days in this randomized, placebo-controlled, two-way-crossover study.

Results

Alosetron had no significant effect on theophylline plasma concentrations (C_max≈9 µg ml⁻¹, AUC≈90 µg ml⁻¹ h) or oral formation clearance of three major metabolites produced via CYP1A2: 3-methylxanthine, 1-methylurate and 1,3-dimethylurate (5, 7 and 16 ml min⁻¹, respectively). Concomitant administration of alosetron and theophylline was well tolerated.

Conclusions

The absence of a clinical drug interaction involving inhibition of theophylline metabolism by alosetron was not predicted by in vitro and in vivo metabolic probe data.

Introduction

Lotronex® (alosetron hydrochloride) is a potent and selective 5-HT₃ receptor antagonist recently approved in the U.S., at a dose of 1 mg twice daily, to relieve irritable bowel syndrome (IBS) pain and discomfort, and to improve bowel function in females with IBS whose predominant bowel habit is diarrhoea [1]. Oral administration of alosetron is followed by rapid absorption (t_max 1 h) with a bioavailability of 50–60%. Elimination of alosetron is also rapid (t_½ 1.5 h), predominantly through metabolism by various cytochrome P450 (CYP) enzymes including CYP1A2.

The effect of alosetron on various CYP enzymes was examined to assess the potential for metabolic drug interactions. This was accomplished in three steps using enzyme-specific substrates, progressing from in vitro human microsome experiments, to in vivo metabolic probe substrates, and finally to a clinical pharmacokinetic drug interaction study. In vitro, alosetron 2 µm (100-fold higher than peak therapeutic serum concentrations [2]) produced approximately 50% inhibition of CYP1A2-mediated ethoxyresorufin-O-deethylation (unpublished data). In vivo [2], therapeutic concentrations of alosetron produced 30% inhibition of CYP1A2-mediated [3] caffeine demethylation to 1,7-dimethylxanthine at serum caffeine concentrations of 0.15–15 mm. This moderate inhibition of CYP1A2 activity suggested that alosetron might inhibit the metabolism of drugs that are substrates of this enzyme under therapeutic conditions. Therefore, these findings were pursued further in a clinical pharmacokinetic drug interaction study with theophylline, another CYP1A2 substrate.

Theophylline is a bronchodilator with a narrow therapeutic index (5–20 mg l⁻¹ or 28–110 µm) [4]. Numerous factors, including concurrent drug therapy, may alter theophylline concentration [5]. Theophylline is primarily eliminated by hepatic metabolism, including 8-hydroxylation to 1,3-dimethylurate (1,3-DMU: 45–55% of total clearance), N₁-demethylation to 3-methylxanthine (3-MX: 10–13% of total clearance), and N₃-demethylation to 1-methylxanthine (20–25% of total clearance), all of which are primarily mediated by CYP1A2 [6]. The 1-methylxanthine metabolite exclusively undergoes subsequent oxidation by xanthine oxidase to 1-methylurate (1-MU) [7], which is excreted in urine along with 1,3-DMU and 3-MX.

Methods

Study design and subjects

This randomized, double-blind, placebo-controlled, repeat-dose, two-period, crossover study (S3BA1004) was conducted at L.A.B. Pharmacological Research International Inc., Vaudreuil, Québec, Canada. The study was approved by the L.A.B. Institutional Review Board, and all subjects provided written informed consent to participate. Ten subjects were recruited to provide 90% power to detect a 30% difference in theophylline AUC between treatments, and all completed the study. Subjects were healthy, nonsmoking, females aged 18–44 (mean 31) years, and weighing 52–73 (mean 62) kg (body mass indices of 21–29 kg m⁻²).

Study treatments

Subjects received each of two treatments in a randomly assigned order, separated by a 14 day washout. Treatments consisted of either 1 mg alosetron (Lotronex®, Glaxo Wellcome Inc., Canada) or placebo orally twice daily for 15.5 days, starting on the morning of day 1 and continuing through the morning of day 16. For the first 7.5 days of each treatment, subjects took the medication at home and recorded the time of each dose in a diary. On the evening of day 8, subjects returned to the clinic and started administration of theophylline (THEO-DUR®, Astra Pharma Inc., Canada) 200 mg orally twice daily for 8 days. Subjects remained in the clinic for the last 8 days of each treatment.

Study procedures

Subjects were screened for participation based on medical history and 12-lead electrocardiogram, along with physical examinations, vital signs, clinical laboratory tests, and pregnancy tests. Clinical laboratory tests and pregnancy tests were repeated after the study. Females of childbearing potential were to use adequate contraception throughout the study. Any adverse events occurring during the study were recorded. Compliance with outpatient alosetron dosing was checked by taking a blood sample prior to the start of dosing with theophylline.

Throughout the study, subjects were restricted from ingesting agents that might alter the metabolism of either alosetron or theophylline: other medications or charbroiled foods or cruciferous vegetables or any tobacco product starting 7 days prior to study, methylxanthine-containing foods or beverages starting 72 h prior to study, and alcohol or grapefruit juice starting 48 h prior to study. Subjects fasted from the prior midnight through 4 h after the last theophylline dose on day 15, and received identical meals thereafter in each treatment period.

Sample collection and bioanalysis

Blood samples (5 ml) for measurement of theophylline were drawn prior to, and 1, 1.5, 2, 3, 4, 6, 8, 12, 14, 24, and 36 h after the last dose of theophylline on the morning of day 15 in each treatment period. Blood was anticoagulated with heparin, centrifuged at 2000 rev min⁻¹ for 7 min at 4 ° C, and the resultant plasma was stored at −20 ° C for bioanalysis. Plasma samples were analysed for theophylline using a high performance liquid chromatography (h.p.l.c.) assay (PPD-Development, Richmond, VA). Theophylline and internal standard, 7-(β-hydroxypropyl)-theophylline (11.25 µg), were isolated from plasma following addition (250 µL) of buffer (potassium phosphate 0.5 m, pH 7) and liquid–liquid extraction with 8 ml of methylene chloride : isopropyl alcohol (9 : 1, v : v). The organic layer was removed, evaporated to dryness, and the residue reconstituted in 500 µL mobile phase consisting of water : acetonitrile : methanol : 1 m tetrabutylammonium hydroxide [containing tris(hydoxymethyl)aminomethane, pH 7] (91 : 3.5 : 3.5 : 2, v : v : v : v). A 100 µl aliquot was eluted through a reverse-phase Beckman Ultrasphere™ IP, 150 × 4.6 mm column and a Supelco C18 guard column at 1.1 ml min⁻¹, followed by u.v. detection at 274 nm. The method was validated over a 0.05–25 μg ml⁻¹ range. Quality control samples demonstrated interassay precision within 5% and accuracy within 6% of nominal.

Urine was collected prior to the first theophylline dose on the evening of day 8, and over the 0–4, 4–8, and 8–12 h intervals after the last dose of theophylline on the morning of day 15 in each treatment period. Aliquots (5 ml) from each urine collection were stored in polypropylene tubes at −20 ° C for bioanalysis. Urine samples were analysed for theophylline, 3-MX, 1-MU, and 1,3-DMU, using an h.p.l.c. assay (PPD-Development, Richmond, VA, USA). These four analytes and internal standard, dyphylline (50 µg), were isolated from urine (50 µl) after addition of 0.1 m tetrabutyl ammonium hydrogen sulphate (50 µl), saturated ammonium sulphate (1 ml), and 0.1 m carbonate buffer (50 µl), followed by liquid–liquid extraction with 5 ml methylene chloride: ethyl acetate : isopropanol (45 : 45 : 10, v : v : v). The organic layer was removed, evaporated to dryness, and the residue reconstituted in mobile phase consisting of acetonitrile : 10 mm sodium acetate : 0.5 mm n-decylamine, pH 4 (1 : 99, v  :  v). A 50 µL aliquot was eluted through a reverse-phase Beckman Ultrasphere™ IP, 150 × 4.6 mm column at 30 ° C and 2.5 ml min⁻¹, followed by u.v. detection at 280 nm. The method was validated over a 4–100 µg ml⁻¹ range for all analytes. Quality control samples demonstrated interassay precision within 15% and accuracy within 20% of nominal for all analytes at all concentrations.

Data analysis

Plasma theophylline concentration data were analysed by noncompartmental methods using WinNonlin™ Professional, Version 1.5 (Scientific Consultants Inc., Cary, NC, USA). Area under the concentration-time curve within a 12 h steady-state dosing interval (AUCτ) was determined using linear trapezoidal interpolation. Terminal elimination half-life (t_½) was determined using log-linear regression of terminal concentration vs time to 36 h postdose. Peak-to-trough fluctuation in theophylline steady-state concentrations was determined by calculating (C_max−C_min)/C_max. Renal clearance of theophylline (CL_R) and the oral formation clearances of each metabolite (CLf/F) were determined as the ratio of urinary recovery to AUC during a 12 h steady-state dosing interval.

With the exception of t_max, all pharmacokinetic parameters were log_e-transformed and analysed by analysis of variance (anova) for a standard 2-period, 2-sequence crossover design, allowing for subject, sequence, period, and treatment effects. The Wilcoxon signed-rank test for paired data was used to analyse t_max, and the median treatment difference was determined.

Results

Prior and concomitant dosing with alosetron did not alter theophylline plasma concentrations in a clinically significant manner (Figure 1). No statistically significant differences were observed in plasma theophylline C_max, t_max, C_min, t_½, AUCτ, CL_R or CLf/F of its three metabolites (1-MX, 3-MU, 1,3-DMU) produced by CYP1A2 (Table 1). A statistically significant (P = 0.017) but small increase in theophylline steady-state fluctuation index from 21% to 32% was observed in the presence of alosetron (Table 1). The 10 subjects in this study provided 80% power to detect a 9% change in theophylline AUCτ.

Figure 1.

Mean steady-state plasma theophylline concentrations (n = 10) after administration of theophylline 200 mg twice daily, with prior and concomitant administration of either placebo twice daily (○) or alosetron 1 mg twice daily (•).

Table 1.

Theophylline pharmacokinetic parameters.

	Theophylline + placeboa	Theophylline+alosetrona	Treatment comparsionb
Cmax (µg ml−1)	8.57 (7.20, 10.22)	9.30 (7.81, 11.09)	1.09 (0.97, 1.22)
tmax (h)	6.0 (1.0, 8.4)	3.0 (0.5, 8.0)	−1.50 (−2.76, −0.25)
AUCτ (µg ml−1 h)	91.2 (76.5, 108.7)	92.9 (78.0, 110.8)	1.02 (0.94, 1.10)
t½ (h)	8.15 (6.88, 9.65)	8.75 (7.39, 10.37)	1.07 (0.96, 1.20)
Cmin (µg ml−1)	6.67 (5.43, 8.20)	6.16 (5.01, 7.57)	0.92 (0.85, 1.00)
Fluctuation index	0.21 (0.17, 0.26)	0.32 (0.26, 0.40)	1.54 (1.18, 2.02)
CLR (ml min−1)	5.55 (4.37, 7.05)	6.37 (5.01, 8.09)	1.15 (1.01, 1.31)
3-MX CLf/F (ml min−1)	5.11 (3.53, 7.41)	4.79 (3.30, 6.95)	0.94 (0.83, 1.06)
1-MU CLf/F (ml min−1)	6.85 (4.75, 9.89)	7.24 (5.02, 10.45)	1.06 (0.93, 1.20)
1,3-DMU CLf/F (ml min−1)	16.5 (13.7, 19.8)	16.4 (13.7, 19.7)	1.00 (0.92, 1.08)

^ageometric LS mean (95% CI), except for t_max: median (range).

^bgeometric LS mean ratio (90% CI), except for t_max: median difference (90% CI).

Both treatments were well tolerated. No serious adverse events were reported, and no clinically significant changes in clinical laboratory tests or vital signs were observed.

Discussion

Alosetron was shown in vitro and in vivo to moderately inhibit CYP1A2 activity (unpublished data). Based on these observations, it was important to examine the potential effect of alosetron on a representative substrate of this enzyme administered clinically.

The clinical study was conducted exclusively in females because alosetron is approved for treatment of diarrhoea-predominant IBS in females. In addition, higher serum concentrations (lower clearance) of both theophylline [5] and alosetron [8] have been reported in females, which may be related in part to lower CYP1A2 activity [9] compared with males. For these reasons, investigating this potential drug interaction in females appeared to be more clinically relevant.

The results of this study demonstrated that alosetron does not alter the metabolism of theophylline, neither elevating plasma theophylline concentration nor decreasing oral formation clearance of its three metabolites, 3-MX, 1-MU, and 1,3-DMU. The absence of an effect by alosetron on theophylline metabolism in vivo contrasted with previous observations that alosetron moderately inhibited CYP1A2-mediated ethoxyresorufin O-deethylation in vitro and caffeine demethylation in vivo. This discrepancy may be partially explained by incomplete dependence of theophylline on CYP1A2 for elimination. Caffeine demethylation to 1,7-DMX [10] and theophylline metabolism to 1,3-DMU, 3-MX, and 1-MU [6] constitute similar dose-fractions (≈0.8) metabolized with similarly high affinities (K_m≈0.3 mm) by CYP1A2. However, in contrast to the singular dependence of in vivo caffeine demethylation on CYP1A2, the high concentrations of theophylline in this study (6–9 µg ml⁻¹ or 33–50 µm) may have allowed a greater contribution from other enzymes with lower affinity, such as CYP2E1, which mediates hydroxylation of theophylline to 1,3-DMU (K_m≈15 mm) [6, 10, 11]. Although incomplete dependence of theophylline on CYP1A2 may partially explain the observed in vitro/in vivo discrepancy, more is needed to account for the complete lack of inhibition of CYP1A2-dependent theophylline metabolism. An additional explanation is the extent to which alosetron could produce such inhibition. In vitro inhibition of CYP1A2 by alosetron was only moderate, and achieved at a total concentration in microsomes (2 µm or 570 ng ml⁻¹) that was apparently too high to translate to in vivo therapeutic conditions where peak alosetron plasma concentrations, nearly 30-fold lower, do not exceed 20 ng ml⁻¹ or 70 nm. Thus, it appears that although alosetron produced moderate inhibition of CYP1A2 in vitro, it is unlikely to produce a clinically significant effect on the metabolism of drugs by CYP1A2 in vivo.

An unexpected observation in this study was the apparent effect of alosetron on fluctuation in theophylline steady-state concentrations derived from the sustained-release tablet employed. Although steady-state fluctuation index was statistically significantly increased, neither C_max nor C_min were significantly altered in the presence of alosetron. It is not clear how alosetron might be implicated in such an effect, although its pharmacological effect of slowing proximal colonic transit might be considered [12]. The small increase observed in theophylline renal clearance might also be contributory. However, this small increase in fluctuation and the lack of a notable elevation in peak concentrations, generally associated with toxicity, are unlikely to produce clinically significant consequences.

The effect of theophylline on alosetron pharmacokinetics was not assessed in this study because the limited contribution of CYP1A2 (10%) to alosetron metabolism, together with the wide therapeutic index of alosetron, renders any inhibition of such a minor pathway unlikely to be clinically significant.

In summary, alosetron did not alter the metabolism of theophylline, in contrast to its effect on in vitro and in vivo probe substrate metabolism. Therefore, in vitro studies showing inhibition with inhibitor concentrations greatly in excess of therapeutic concentrations may not be predictive of potential clinical drug interactions.