PHARMACOKINETICS AND DISPOSITION

The effect of fluvoxamine on the pharmacokinetics,safety, and tolerability of ramosetron in healthy subjects

Takeshi Kadokura & Martin den Adel &Walter J. J. Krauwinkel & Tetsuo Takeshige &

Akito Nishida

Received: 25 October 2007 /Accepted: 26 January 2008 /Published online: 26 April 2008# Springer-Verlag 2008

AbstractObjective To investigate the effects of multiple doses offluvoxamine on the pharmacokinetics, safety, and tolerabil-ity of a single oral 10-μg dose of ramosetron.Methods This was a single-center, open, one-sequencecross-over study. On Day 1, healthy male and femalesubjects were administered a single dose of 10 μgramosetron. Dosing of fluvoxamine started with an initialmorning dose of 50 mg on Day 3, followed by a twice daily(12-h interval) dosing of 50 mg on Days 4–12. Themorning dose on Day 11 was administered in combinationwith a single dose of 10 μg ramosetron.Results Co-administration of fluvoxamine with ramosetronresulted in an increase in the Cmax and AUC0-inf of

ramosetron by 1.42-fold (90% CI 1.35–1.49) and 2.78-fold(90% CI 2.53–3.05), respectively.Conclusion Co-administration of the CYP1A2 inhibitor flu-voxamine with ramosetron resulted in an interaction. Howev-er, the safety data collected during the study do not indicatethat this interaction will cause any major safety concerns.

Keywords CYP1A2 . Drug interaction . Fluvoxamine .

Pharmacokinetics . Ramosetron

Introduction

Irritable bowel syndrome (IBS) is a functional disease withpersisting gastrointestinal symptoms, mainly abdominal dis-comfort/pain and defecation abnormality, not accompanied byan organic disease. Although IBS is not a lethal disease,patients with IBS suffer restriction of their behavior dependingon the symptoms, and their social activity is disturbed so thatthe quality of life (QOL) of IBS patients has been reported to besignificantly lowered [1]. 5-HT3 receptors have been identi-fied on the sensory neurons of the gut, and these mediatereflexes that control gastrointestinal motility and secretion,bowel function and the perception of pain [2]. In patients withdiarrhea-predominant IBS (d-IBS), 5-HT3 receptor antago-nists increase colonic compliance, slow colonic transit,improve stool consistency and, therefore, represent valuabletherapeutic compounds for treatment [3–5].

Ramosetron (ramosetron hydrochloride) is a selective 5-HT3 receptor antagonist. It has been on the market since 1996in Japan and a number of other Asian countries as an anti-emetic drug for cancer patients receiving chemotherapy. Itspotential in the treatment of d-IBS was recently assessed in acohort of Japanese patients who were given 5 μg ramosetronhydrochloride once daily for 12 weeks. The results revealed

Eur J Clin Pharmacol (2008) 64:691–695DOI 10.1007/s00228-008-0466-x

T. Kadokura (*)Clinical Pharmacology, Astellas Pharma Inc.,3-17-1, Hasune, Itabashi-ku,Tokyo 174-8612, Japane-mail: takeshi.kadokura@jp.astellas.com

M. den Adel :W. J. J. KrauwinkelExploratory Development, Astellas Pharma Europe B.V.,Elisabethhof 19, P.O. Box 108 2350 AC, Leiderdorp,The Netherlands

T. TakeshigeDrug Metabolism Research Laboratories, Astellas Pharma Inc.,2-1-6, Kashima, Yodogawa-ku,Osaka 532-8514, Japan

A. NishidaProject Management, Astellas Pharma Inc.,3-17-1, Hasune, Itabashi-ku,Tokyo 174-8612, Japan

that ramosetron hydrochloride was effective in relieving ab-dominal discomfort and pain as well as other IBS symptoms.

When ramosetron hydrochloride was orally administeredin a dose of 0.4–0.6 mg, the time to reach the maximumplasma concentration (tmax) occurred between 2.17 and2.67 h after administration, and the plasma half-life (t½)ranged from 4.93 to 5.52 h. The maximum concentration(Cmax) and the area under the concentration–time curve(AUC0-inf) increased nearly in proportion to the dose, and theabsolute bioavailability was 53.0–59.0% and nearly constantamong doses. The urinary excretion rate of unchanged drugwas approximately 8–13%. When ramosetron hydrochloridewas administered in a dose of 0.6 mg twice daily for 7 days,there was no effect on the plasma concentration of theunchanged drug by repeated dosing [6].

Previous studies have shown that ramosetron is mainlymetabolized by liver P450, and its four major metabolitesfound in urine, with the exception of the conjugates, werepharmacologically active [7]. The main cytochrome P(CYP) isozymes involved in the metabolism of ramosetronare CYP1A2 and CYP2D6, and it is possible that thepharmacokinetics of ramosetron are affected by CYP1A2and CYP2D6 inhibitors. Since the drug fluvoxamine isexpected to be commonly used in the target population oframosetron and also has CYP1A2-inhibiting properties, it isimportant to carry out a clinical investigation of the effectof fluvoxamine on the pharmacokinetics, safety andtolerability of ramosetron.

Methods

Study design

This was a single-center, open, one-sequence cross-overstudy. The aim was to evaluate the effects of fluvoxamine(Fevarin), given 50 mg twice daily in steady state, on thepharmacokinetics, safety and tolerability of a single 10-μgdose of ramosetron in healthy male and female subjects. Atotal of 24 subjects (12 subjects per gender group) wereenrolled in the study. Sample size was calculated usingprevious results (unpublished data). Based on previousstudies, the clinically efficacious dose of ramosetron wasexpected to be in the range of 5–10 μg once daily. Therefore,the higher single dose of 10 μg ramosetron used here wasselected in order to study the increased exposure oframosetron in terms of pharmacokinetics and safety parame-ters when combined with fluvoxamine. As the use of fluvox-amine beyond daily doses of 100 mg was expected to result inside effects in healthy subjects, such as nausea, and henceincrease the frequency of dropout, it was considered justifiableto select a daily dose of 100 mg fluvoxamine (e.g. 50mg twicedaily) from the therapeutic dose range of fluvoxamine. In

addition, the recommended daily dose for adults sufferingfrom depression is 100 mg. All subjects were informed of thenature and purpose of the study, and their written informedconsent was obtained before screening.

The subjects were admitted 1 day prior to first drugadministration (Day 0). On Day 1, the subjects wereadministered a single dose of 10 μg ramosetron. Startingon the morning of Day 3, the subjects were administeredfluvoxamine to reach steady state. For safety and tolerabilityreasons, dosing started with an initial morning dose of 50 mgon Day 3, followed by a twice-daily (12-h interval) dosing of50 mg on Days 4–12. The morning dose on Day 11 wasadministered in combination with a single dose of 10 μgramosetron. The subjects were discharged in the morning ofDay 13. Approximately 1–2 weeks after discharge, thesubjects returned to the unit for a post-study visit.

The study was performed at FOCUS Clinical DrugDevelopment GmbH (Neuss, Germany), in accordance withthe International Conference on Harmonization Guidelinesfor Good Clinical Practice (GCP) and with the principles ofthe Declaration of Helsinki (1996 version). The study wasapproved by the Independent Ethics Committee (IEC), andall subjects gave written, informed consent before partici-pating in the study.

Bioanalysis

For the determination of the plasma concentrations oframosetron and its metabolites (M-1, M-3, M-4 and M-8)in the presence and absence of fluvoxamine, 10-mLsamples of venous blood were removed from the arm ofthe subject at pre-dose and 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12, 16,24, 32, 40 and 48 h post-dose on Days 1 and 11. For theanalysis of fluvoxamine in plasma, samples of venousblood (5 mL) were taken at the pre-morning dose on Days5, 7 and 9–12 and on Day 13, 12 h after the lastadministration of fluvoxamine. Samples were collected instandard, appropriately labeled, blood-sampling tubes(Vacutainer; Becton Dickinson, Franklin Lakes, NJ) con-taining Li-heparin as anticoagulant. The samples wereprocessed within 15 min of collection: plasma was obtainedby centrifugation at approximately 1500 g for 15 min atapproximately 4°C. Plasma samples were stored in light-protected conditions at −20°C or below, within 30 min ofcentrifugation until shipping. Plasma concentrations oframosetron and its metabolites were determined by meansof a validated high sensitivity liquid chromatography-tandem mass spectrometry (LC-MS/MS) method, with0.88 pg/mL being the lower limit of quantification forramosetron and its metabolites (M-1, M-3, M-4 and M-8).

Plasma concentrations of fluvoxamine were determined bymeans of a validated LC-MS/MS method, with 0.20 ng/mLbeing the lower limit of quantification for fluvoxamine.

692 Eur J Clin Pharmacol (2008) 64:691–695

Inter- and intra-assay precision and accuracy wereassessed at the lower limit of quantification and at low,medium and high QC concentration level. For ramosetronand its metabolites, inter- and intra-assay precision results(CV%) were less than 18.3% at the lower limit ofquantification and less than 12.5% at low, medium andhigh QC samples. Inter- and intra-assay accuracy resultswere all less than ±14.7% at all concentration levels. Forfluvoxamine, the precision (CV%) was less than 5.7% andthe accuracy was within ±5.3% at all concentration levels.

Pharmacokinetic analysis

Pharmacokinetic analysis of the plasma concentration dataof ramosetron and its metabolites was performed usingWinNonlin® ver. 4.1 (Pharsight Corp., Mountain ViewCA). Non-compartmental analysis was used exclusively.The linear–logarithmic trapezoidal rule was used forcalculating the AUClast (area under the plasma concentra-tion–time curve from time zero to the time of the last

quantifiable concentration) values. The AUC0-inf (areaunder the plasma concentration–time curve from time zeroextrapolated to infinity) was estimated using the equation

AUC0�inf ¼ AUClast þ Clast=kel

where Clast is the predicted plasma concentration of the lastquantifiable sample, and kel is the first-order terminalelimination rate constant determined by ordinary linearregression of the logarithmically transformed concentrationvalues.

Statistical analysis

All 24 participating subjects were included in the dataanalysis. Analysis of variance (ANOVA) of the log-trans-formed AUC0-inf and Cmax was performed using SAS ver.8.2 (SAS Institute, Cary, NC) in order to obtain 90%confidence intervals for the ratio of ramosetron in combi-nation with fluvoxamine relative to ramosetron alone.There would be no evidence for an interaction betweenramosetron and fluvoxamine if the limits of the 90%confidence intervals (90% CI) for both AUC0-inf and Cmax

fell within 0.8 and 1.25.

Results

Demographic, safety and tolerability

A total of 24 Caucasian subjects were enrolled in the study.The mean age, body weight and height of the male subjectswere 32.8 years (range 25–48 years), 74.7 kg (range 67–88 kg)and 175.1 cm (range 167–180 cm), respectively; the meanage, body weight and height of the female subjects were38.2 years (range 24–52 years), 67.9 kg (range 56–83 kg)and 163.8 cm (range 151–175 cm), respectively. All subjectswere treated, and all completed the study. None of thesubjects died during the course of the study, there were no

Fig. 1 Mean plasma concentrations of the unchanged drug–timecurve after oral administration of 10 μg ramosetron to healthy adultmales and females in the presence (filled circles) and absence (opencircles) of fluvoxamine (mean ± SD, n=24)

Table 1 Pharmacokinetic parameters of the unchanged drug in plasma after oral administration of 10 μg ramosetron to healthy adult males andfemales in the presence and absence of fluvoxamine (mean ± SD)

Pharacokinetic parameters Ramosetron 10 μg alone Ramosetron 10 μg + fluvoxamine 100 mg

Male (n=12) Female (n=12) Total (n=24) Male (n=12) Female (n=12) Total (n=24)

Cmax (pg/mL) 33.2±5.9 49.7±12.2 41.5±12.6 50.1±8.2 65.2±11.8 57.6±12.6tmax (h) 2.29±0.96 1.83±0.65 2.06±0.84 2.96±0.81 2.67±0.96 2.81±0.88t½ (h) 6.99±0.86 7.43±1.51 7.21±1.22 13.88±2.42 13.37±2.19 13.63±2.28V/F (L) 315±69 255±90 285±84 211±31 168±33 190±38AUClast (pg.h/mL) 322±84 458±180 390±154 872±164 1077±211 974±213AUC0-inf (pg.h/mL) 335±84 470±181 403±155 965±210 1181±252 1073±252

Cmax, Maximum observed plasma concentration; tmax, time to reach Cmax; t½, apparent terminal half-life; V/F, apparent volume of distribution;AUClast, area under the plasma concentration–time curve from time zero to the time of the last quantifiable concentration; AUC0-inf, area under theplasma concentration–time curve from time zero extrapolated to infinity

Eur J Clin Pharmacol (2008) 64:691–695 693

serious adverse events and no subjects withdrew from thestudy. Analysis of vital signs, electrocardiogram (ECG) andsafety laboratory parameters did not reveal any safetyconcerns.

Pharmacokinetics

In the absence of fluvoxamine, ramosetron was rapidlyabsorbed, with a mean tmax of 2.06 h. After Cmax wasreached, the plasma concentration decreased with a mean t½of 7.21 h (Fig. 1, Table 1). Compared with ramosetron, theplasma concentrations of the metabolites was much lower.In the absence of fluvoxamine, a sufficient number ofsamples contained quantifiable concentrations of M-1, M-3and M-8 to allow at least the determination of tmax, Cmax

and AUClast. The mean AUClast of M-1, M-3 and M-8 inthe absence of fluvoxamine were approximately 8.8, 1.2and 2.8% of that of unchanged drug, respectively (Table 2).

The addition of fluvoxamine clearly affected the phar-macokinetics of ramosetron (Fig. 1, Table 1). In themajority of subjects, tmax increased slightly, but in twosubjects, the tmax was decreased. A consistent increase inCmax (23 subjects) and AUC0-inf (all subjects) wasobserved. The ratio of the least square mean (LSM) of

treatment effect was 1.42 (90% CI 1.35–1.49) for Cmax and2.78 (90% CI 2.53–3.05) for AUC0-inf (Table 3). The meanvalue of t½ increased from 7.21 to 13.63 h. The plasmaconcentrations of metabolites decreased, and only M-1 wasquantifiable in four subjects. The concentrations of M-4were below the level of quantification (LOQ) in all subjectsfor both treatments.

Mean (SD) Ctrough values of fluvoxamine ranged between41.9 (46.4) ng/mL on Day 11 and 43.5 (50.6) ng/mL on Day13, indicating that steady state was achieved when thesecond dose of ramosetron was administered.

Discussion

Previous in-vitro metabolism studies have shown that bothCYP1A2 and CYP2D6 are able to metabolize ramosetroninto a number of metabolites. Since the drug fluvoxamine isexpected to be commonly used in the target population oframosetron and also has CYP1A2-inhibiting properties, itwas considered useful to investigate potential interactionsbetween such concomitant medication and ramosetron. Wetherefore evaluated the effects of multiple doses of fluvox-amine on the pharmacokinetics of a single oral 10-μg dose

Table 2 Pharmacokinetic parameters of metabolites in plasma after oral administration of 10 μg ramosetron to healthy adult males and females inthe presence and absence of fluvoxamine (mean ± SD)

Metabolites Parameter Ramosetron 10 μg alone Ramosetron 10 μg + fluvoxamine 100 mg

Male (n=12) Female(n=12)

Total (n=24) Male(n=12)

Female (n=12) Total (n=24)

M-1 Cmax (pg/mL) 4.02±0.79 4.38±0.71 4.20±0.76 1.21 (n=1) 1.54±0.22 (n=3) 1.46±0.25 (n=4)tmax (h) 2.83±0.78 2.79±0.78 2.81±0.78 12.00 (n=1) 9.35±2.30 (n=3) 10.01±2.30 (n=4)AUClast

(pg.h/mL)31.6±6.8 37.3±10.2 34.5±8.9 8.2 (n=1) 12.5±6.7 (n=3) 11.4±5.9 (n=4)

M-3 Cmax (pg/mL) 1.53±0.43 (n=11) 1.64±0.30 1.58±0.36 (n=23) NA NA NAtmax (h) 2.64±0.50 (n=11) 2.04±0.75 2.33±0.70 (n=23) NA NA NAAUClast

(pg.h/mL)4.5±3.0 (n=11) 4.7±1.8 4.6±2.4 (n=23) NA NA NA

M-8 Cmax (pg/mL) 2.33±0.45 2.87±0.61 2.60±0.59 NA NA NAtmax (h) 2.50±0.74 2.21±0.89 2.35±0.81 NA NA NAAUClast

(pg.h/mL)10.0±2.9 12.2±4.6 11.1±3.9 NA NA NA

NA, Not applicable

Table 3 Pharmacokinetic interaction between ramosetron and fluvoxamine

PK parameter Treatment LSMa LSM ratio 90%CI lower 90%CI upper

Cmax Ramosetron alone 39.8 1.42 1.35 1.49Ramosetron with fluvoxamine 56.4

AUC0-inf Ramosetron alone 376.1 2.78 2.53 3.05Ramosetron with fluvoxamine 1044.7

LSM, Least square mean; log-transformed results are back-transformed to the original scale; 90% CI, 90% confidence interval

694 Eur J Clin Pharmacol (2008) 64:691–695

of ramosetron in healthy male and female subjects. Ourresults support the important role of CYP1A2 in themetabolism of ramosetron: the Cmax increased 1.42-fold(90% CI 1.35–1.49) and the AUC0-inf increased 2.78-fold(90% CI 2.53–3.05). The treatment-by-gender interactionfor AUC0-inf was not statistically significant; therefore,there was no evidence to support a difference betweengenders in terms of the magnitude of the treatmentdifference. We did observe a statistically significanttreatment-by-gender interaction for Cmax; for both genders,higher Cmax values were observed for ramosetron incombination with fluvoxamine than for ramosetron alone,although a larger difference between treatments wasobserved in the male subjects (the ratio of LSM is 1.51for male subjects and 1.33 for female subjects). The meant½ increased from 7.21 to 13.63 h. Furthermore, apronounced reduction in plasma concentrations was ob-served in the three metabolites M-1, M-3 and M-8. Thein-vivo formation of these metabolites is highly CYP1A2dependent. If only the elimination of ramosetron had beeninhibited, a similar increase in t½ and in AUC0-inf (withonly a minimal change in Cmax) would have been expected.The increase in AUC0-inf, however, was more pronouncedthan the increase in t½. Consequently, it is likely that notonly the elimination was inhibited but also the first-passeffect, resulting in an increase in bioavailability. Thus, in-hibition of CYP1A2 probably resulted in both a decreasedelimination and a higher bioavailability of ramosetron.

The Cmax and AUC0-inf of ramosetron were higher infemale subjects than in male subjects. These differenceswere statistically significant. Even after the effect of weighthad been adjusted for, a gender difference in Cmax andAUC0-inf were statistically significant (data not shown).Although the data do not provide enough information toconclude what was responsible for this difference, theresults suggest that bioavailability may have been higher infemale subjects. A gender difference in pharmacokineticshas been reported for some drugs which were metabolizedby CYP1A2 [8], and the present results clearly indicate thatCYP1A2 plays an important role in ramosetron metabo-lism. A gender difference in the pharmacokinetics has alsobeen reported for alosetron and ondansetron, which havebeen categorized in the same class as ramosetron [9, 10]. A

contribution of CYP1A2 has been reported with these drugs[11, 12].

In conclusion, the co-administration of the CYP1A2inhibitor fluvoxamine with ramosetron resulted in aninteraction in which the Cmax and AUC0-inf of ramosetronwere increased by 1.42-fold and 2.78-fold, respectively.The safety data collected during the study do not indicatethat such interactions between ramosetron and fluvoxaminewill cause any major safety concerns.

References

1. Gralnek IM, Hays RD, Kilbourne A, Naliboff B, Mayer EA(2002) The impact of irritable bowel syndrome on health-relatedquality of life. Gastroenterol 119:654–660

2. Gershon MD (1999) Review article: roles played by 5-hydroxy-tryptamine in the physiology of the bowel. Aliment PharmacolTher 13[Suppl 2]:15–30

3. Delvaux M, Louvel D, Mamet JP, Campos-Oriola R, Frexinos J(1998) Effect of alosetron on responses to colonic distension inpatients with irritable bowel syndrome. Aliment Pharmacol Ther12:849–855

4. Forster JM, Houghton LA, Whorwell PJ (1997) Alosetron showscolonic transit in patients with irritable bowel syndrome (IBS).Gastroenterol 112:A732 (Abstract)

5. Northcutt AR, Camilleri M, Mayer EA et al (1998) Alosetron,a 5HT3-receptor antagonist, is effective in the treatment of fe-male irritable bowel syndrome patients. Gastroenterol 114:A812(Abstract)

6. Nakashima M, Kanemaru M, Miura H, Takeshige T, Atsuta Y(1995) Phase I clinical study of oral ramosetron. Rinsyo-To-Kenkyu 72:264–278

7. Astellas Pharma Inc. Investigator’s Brochure (2005) YM060Ramosetron Hydrochloride

8. Schwartz JB (2003) The influence of sex on pharmacokinetics.Clin Pharmacokinet 42:107–121

9. Koch KM, Palmer JL, Noordin N, Tomlinson JJ, Baidoo C (2002)Sex and age differences in the pharmacokinetics of alosetron. Br JClin Pharmacol 53:238–242

10. Pritchard JF, Bryson JC, Kernodle AE, Benedetti TL, Powell JR(1992) Age and gender effects on ondansetron pharmacokinetics:evaluation of healthy aged volunteers. Clin Pharmacol Ther51:51–55

11. Koch KM, Corrigan BW, Manzo J et al. (2004) Alosetron repeatdose pharmacokinetics, effects on enzyme activities, and influenceof demographic factors. Aliment Pharmacol Ther 20:223–230

12. Dixon CM, Colthup PV, Serabjit-Singh CJ et al. (1995) Multipleforms of cytochrome P450 are involved in the metabolism ofondansetron in humans. Drug Metab Dispos 23:1225–1230

Eur J Clin Pharmacol (2008) 64:691–695 695