Effect of zonisamide on the pharmacokinetics and pharmacodynamics of a combination ethinyl estradiol-norethindrone oral contraceptive in healthy women

CLINICAL THERAPEUTICs®/VoL.2 6 , N o . 12, 2 0 0 4

Effect of Zonisamide on the Pharmacokinetics and Pharmacodynamics of a Combination Ethinyl EstradiolNorethindrone Oral Contraceptive in Healthy W o m e n Sue G. Griffith, MD, PhD, MRCP, 1 and Yuqing Dai, PhD 2 ~Elan Pharmaceuticals, Inc., San Diego, California, and 2i3 5tatprobe, Ann Arbor, Michigan

ABSTRACT

Background: Severalantiepilepticdrugs have clinicallysignificantpharmacokinetic interactionswith oralcontraceptives(OCs) that may result in contraceptive failure. Objective: The aim of this study was to assess the effect of zonisamide on the pharmacokinetics of the individual components of a combination OC (ethinyl estradiol tEE] 0.035 mg and norethindrone [NOR] 1 mg) and on pharmacodynamic variables that may be increased in the event of reduced contraceptive efficacy (concentrations of serum luteinizing hormone [LH], follicle-stimulating hormone [FSH], and progesterone). Methods: This was a single-center, open-label, 1-sequence, crossover study. Healthy, premenopausal women received the combination OC for three 28-day cycles (combination OC for 21 days, followed by placebo for 7 days). Following stabilization on the OC during the first cycle, blood was collected during cycle 2 for the determination of serum EE and NOR profiles (day 14) and serum LH, FSH, and progesterone concentrations (days 13-15). Starting on day 15 of cycle 2, zonisamide was administered orally at 100 mg/d and titrated to a target dose of 400 mg/d. EE and NOR profiles and serum LH, FSH, and progesterone concentrations were obtained again in cycle 3 (in the presence of zonisamide) and compared with those from cycle 2 (in the absence of zonisamide). Results: Thirty-seven healthy premenopausal women (mean age, 26.1 years [range, 18-51 years]; mean body weight, 65.5 kg [range, 50.4-93.1 kg]; mean height, 165.8 cm [range, 152.4-182.9 cm]) received >1 dose of zonisamide. Of the 33 subjects (89.2%) who completed the stud> 26 (78.8%) underwent titration to a stable zonisamide dose of 400 mg/d. For EE, the mean (SD) AUC over a 24-hour dosing interval (AUC~) was 1139 (317) pg.h/mL in cycle 2 and 1143 (312) pg.h/mL in cycle 3; the mean CmaXin the respective cycles was 133 (39) and 141 (46) pg/mL. For NOR, the corresponding values were 140 (48) and 159 (46) ng.h/mL for AUC~ and 21 (5.4) and 23 (6.7) ng/mL for CmaX. The 90% CIs for the geometric mean ratios (cycle 3:cycle 2) for AUCv and Cma~ fell within the accepted range for lack of interaction (0.80-1.25). There were no increases in LH, FSH, or progesterone concentrations between cycle 2 and cycle 3. Conclusions: In these healthy volunteers, steady-state zonisamide dosing had no clinically significant effect on the pharmacokinetics of EE or NOR. There was no pharmacodynamic evidence that zonisamide is likely to reduce the contraceptive effectiveness of OCs containing EE and NOR. (Clin Ther. 2004;26:2056-2065) Copyright © 2004 Excerpta Medica, Inc. Key words: zonisamide, ethinyl estradiol, norethindrone, oral contraceptive, pharmacokinetics, drug interaction. AcceptedJor publication October21, 2004. Printed in the USA. Reproduction in whole or part is not permitted.

2.056

doi:l O.I016/i.clinthera.2004. I 1.019 0149 2918/04/$19.00

Copyright © 2004 Excerpta Hedica, Inc.

S.G. Griffith and Y. Dai

INTRODUCTION

Pharmacokinetic drug-drug interactions are well recognized with many antiepileptic drugs (AEDs). Several AEDs (eg, phenytoin, carbamazepine, phenobarbital) are potent inducers of cytochrome P450 (CYP) isozymes--in particular CYP3A4--and increase the clearance of CYP3A4 substrates, which may result in clinically important drug interactions, including reduced efficacy of oral contraceptives (OCs). 1-5 The metabolism of OC components is complex and involves oxidative, reductive, and conjugatire pathways, with formation of active metabolites. Ethinyl estradiol (EE) and norethindrone (NOR), also known as norethisterone, are commonly used as the estrogen and progestin components of OCs. The pharmacokinetics of EE and NOR have been well characterized. EE is rapidly and almost completely absorbed from the gastrointestinal tract, but because of extensive first-pass metabolism in gut mucosa and liver, its absolute bioavailability is only -50%. The Tmax iS -1.5 tO 2.5 hours after dosing. EE is -97% bound to plasma albumin; it does not bind to sex hormone-binding globulin (SHBG) but does induce SHBG synthesis. CYP3A4 is responsible for the 2-hydroxylation of EE, which is the major route of metabolism. This 2-hydroxymetabolite is further transformed by methylation and glucuronidation. EE is excreted in urine and feces as glucuronide and sulfate conjugates and undergoes enterohepatic circulation. The tl/2 of EE is -10 hours. <7 NOR is absorbed from the gastrointestinal tract, undergoing first-pass metabolism (absolute bioavailability, -64%); the plasma Tmax iS ] tO 2 hours after oral administration. NOR exhibits biphasic pharmacokinetics (ie, an initial distribution phase followed by a prolonged elimination phase, with a tl/2 of >8 hours). It is highly protein bound (-60% to albumin, 35% to SHBG), and administration in combination with an estrogen increases the amount of NOR that is bound to SHBG. NOR is metabolized in the liver, with 50% to 80% of the dose excreted in the urine and up to 40% in the feces. Its major metabolic pathways are 5g- and 5~-reduction and 3o~- and 3~-hydroxylation.6-8 Zonisamide* (1,2-benzisoxazole-3-methanesulfonamide) is an AED, classified as a sulfonamide, that is *Trademark: california).

Zonegran ® (Elan

Biopharmaceuticals,

San

Diego,

chemically unrelated to other antiseizure drugs. 9,1° It has been marketed in Japan and South Korea for more than a decade. In 2000, it was approved in the United States for use as adjunctive therapy in the treatment of partial seizures in adults with epilepsy. Zonisamideg pharmacokinetic characteristics have been well documented. 9-n It is metabolized to inactive metabolites primarily through reductive cleavage of the benzisoxazole ring of the parent drug by CYP3A4. Steady-state serum zonisamide concentrations are -10 to 30 pg/mL after doses of 200 to 600 mg/d in adults with epilepsy. In the absence of CYP3A4 inducers, the apparent clearance of zonisamide at steady state is -0.70 L/h after oral administration and the tl/2 is -50 to 60 hours; the respective values are -1.3 L/h and 30 hours when zonisamide is coadministered with AEDs known to be CYP3A4 inducers. Unlike certain other AEDs (eg, phenytoin, carbamazepine, phenobarbital), 1 zonisamide does not appear to induce CYP enzymes or glucuronidation. The present study was designed to assess the effect of steady-state dosing of zonisamide on the steadystate pharmacokinetics of the EE and NOR components of a combination OC (EE 0.035 mg and NOR 1 mg), t and on pharmacodynamic variables that may be increased in the event of reduced contraceptive efficacy (concentrations of serum luteinizing hormone [LH], follicle-stimulating hormone [FSH], and progesterone). SUBJECTS A N D M E T H O D S Inclusion Criteria

Healthy, premenopausal women aged 18 to 55 years who were within 20% of ideal body weight (according to Metropolitan Life height-weight tables), had taken OCs for at least 3 months before screening, and were willing to change their OC to the study OC were eligible for enrollment. All subjects provided written informed consent before undergoing any screening or other study procedures. Eligible subjects were to have no significant current or past diseases, based on the medical history, physical examination, vital signs (blood pressure, heart rate), and clinical laboratory tests (hematology, serum chemistry, and urinalysis). They had to have negative tTrademark: Ortho-Novum ® 1/35 (Ortho-McNeil Pharmaceutical, Inc., Raritan, New Jersey).

2057

CLINICALTHERAPEUTICS@

results on pregnancy testing and screening examinations for HIV, hepatitis B, and hepatitis C. Use of concomitant medications was restricted, and ingestion of grapefruit juice was prohibited throughout the study. Subjects were not permitted to use tobacco products or recreational drugs during the study, and no alcohol was permitted within 48 hours before blood sampling. Subjects agreed to use a barrier method of contraception throughout the study in addition to taking the study OC.

Study Design This was a single-center, open-label, 1-sequence, crossover study. The study protocol and informed consent form were approved by an independent institutional review board. The study consisted of a screening period, an OC treatment period without coadministered zonisamide, an OC treatment period with coadministered zonisamide, and a posttreatment period. Subjects received the combination OC for two or three 28-day cycles (combination OC once daily on days 1-21 of each cycle and an inert tablet once daily on days 22-28). Cycle 1 was not necessary if the subject was already taking the study OC. OC administration began on the Sunday after the start of menses (cycle 1, day 1), although no pharmacokinetic or pharmacodynamic assessments were conducted during this cycle. Administration of zonisamide began on day 15 of cycle 2 at a dosage of 100 mg PO once daily. The dose was increased by 100 mg each 5 days to a target of 400 mg/d. The titration schedule was as follows: 100 mg each morning for 5 days; 100 mg BID for 5 days; 100 mg each morning and 200 mg each evening for 5 days; and 200 mg BID. If zonisamide was not tolerated at a specific dose level, the dose could be decreased to the last tolerated level. For a subject to remain in the study, the daily zonisamide dose had to be stable at >200 mg/d for >12 days after titration. Pharmacokinetic profiles for EE and NOR in the presence of zonisamide were assessed on the 13th day of cycle 3, after which zonisamide was discontinued. The study OC was continued until the 28-day pack was finished. Study Procedures Screening took place within 14 days before enrollment. Screening assessments included a physical examination, 12-lead electrocardiogram, routine clinical laboratory tests, a serum pregnancy test, urine 2058

screening for alcohol and substances of abuse, and screening tests for hepatitis B, hepatitis C, and HIV. Subjects were instructed to take the OC dose with -240 mL water at the same time each morning and to record the time in a diary. Each 28-day OC cycle immediately followed the completion of the previous one. Subjects were housed in a clinical study unit from day 13 to day 15 of cycle 2, and from the day before until completion of the pharmacokinetic profiles for EE and NOR in cycle 3 (usually from day 13 to day 15). Before the OC dose at which EE and NOR profiles were assessed, subjects underwent a fast of at least 10 hours. Staff at the clinical study unit administered the OC with 240 mL water, and subjects were not permitted to eat, lie down, or sleep for 4 hours afterward. Lunch was served - 4 hours after dosing, and dinner was served -6.5 hours later. The first dose of zonisamide was administered in the clinical study unit on day 15 of cycle 2 with the OC dose. Thereafter, subjects took zonisamide doses themselves, except during the confinement period in cycle 3. At the 100-mg/d dose level, zonisamide was administered each morning as a 100-mg capsule; at all other dose levels, it was administered each morning and evening (-12 hours apart) as one or two 100-mg capsules, depending on the dose level. All doses were taken with -240 mL water, with the morning zonisamide dose taken at the same time as the OC dose. Subjects recorded doses and dosing times in a diary. Subjects returned to the clinical study unit periodically for assessment of compliance with study medications and a review of adverse events (AEs) and concomitant medications. Poststudy assessments (physical examination, vital signs, 12-lead electrocardiogram, routine clinical laboratory tests, serum pregnancy test, and recording of AEs and concomitant medications) were conducted at the end of cycle 3. The following safety measures were assessed throughout the study: AEs, physical examinations, vital signs, and routine clinical laboratory tests. Additional urine screening for alcohol and substances of abuse and serum pregnancy testing were also performed throughout the study.

Sample Collection and Bioanalysis Ethinyl Estradiol and Norethindrone

Blood samples (7 mk peripheral venous blood collected in a plain tube without anticoagulant) for mea-

S.G. Griffith and Y. Dai

surement of serum EE and NOR concentrations in the absence of zonisamide were collected before dosing and at 1, 2, 3, 4, 6, 8, 12, 16, and 24 hours on day 14 of cycle 2. Blood samples for measurement of serum EE and NOR concentrations during coadministration of zonisamide were collected at the same time points on the 13th day of stable zonisamide dosing in cycle 3. This took place on day 14 of cycle 3 in the majority of subjects but could have occurred as late as day 21 of cycle 3, depending on when stabilization at a tolerated dose of zonisamide was achieved. The blood was allowed to clot at room temperature for at least 30 minutes, and the tubes were then centrifuged for 15 minutes, no later than 60 minutes after collection. Serum samples (-3 mL) were transferred by pipette into polypropylene tubes, immediately frozen, and stored at --20°C until assayed. Serum samples were analyzed for EE and NOR by PPD Development (Richmond, Virginia) using a validated capillary gas chromatography/mass spectrometry method. The quantification range for EE was 2 to 1000 pg/mL; for NOR, it was 0.05 to 25 ng/mL. Interassay precision (%CV) was <10.7% for EE and <8.0% for NOR. The assay met US Food and Drug Administration guidelines for linearity, precision, and accuracy. ~2

measurement of serum zonisamide concentrations were collected before the last 3 doses of zonisamide in cycle 3 to determine whether steady state had been reached, and at 12, 84, 156, and 324 hours after the last dose of zonisamide for estimation of the zonisamide tl/2. The blood was allowed to clot at room temperature for at least 30 minutes, and the tubes were then centrifuged for 10 minutes, no later than 90 minutes after collection. Serum samples (-4 mL) were transferred by pipette into polypropylene tubes, immediately frozen, and stored at - - 7 0 ° C until assayed. Serum samples were analyzed for zonisamide by AAI International (Shawnee, Kansas) using a validated method. Zonisamide and the added internal standard, N,N-dimethylzonisamide, were extracted from serum using liquid-liquid extraction, and the extract was then subjected to reverse-phase highperformance liquid chromatography. Zonisamide and N,N-dimethylzonisamide in the effluent were detected and quantitated by ultraviolet absorbance. The quantification range was 0.5 to 64 pg/mL, and the interassay precision (%CV) was <8.6%. The assay met US Food and Drug Administration guidelines for linearity, precision, and accuracy. 12

Luteinizing Hormone, Follicle-Stimulating Hormone, and Progesterone Blood samples (7 mL peripheral venous blood collected in a plain serum separator tube without anticoagulant) for measurement of LH, FSH, and progesterone concentrations were collected each morning on days 13, 14, and 15 of cycles 2 and 3. The blood was allowed to clot at room temperature for at least 30 minutes, and the tubes were then centrifuged for 10 minutes, no later than 60 minutes after collection. Serum samples were stored at 2°C to 10°C until assayed at the clinical laboratory of PRACS Institute (Fargo, North Dakota) by paramagnetic-particle chemiluminescent immunoassays using the Access Immunoassay System (Beckman Instruments, Inc., Fullerton, California). The limits of detection were 0.2 to 200 mIU/mL for LH, 0.2 to 250 mIU/mL for FSH, and 0.08 to 40 ng/mL for progesterone.

Pharmacokinetic

Parameters

The following noncompartmental pharmacokinetic parameters for EE and NOR were derived from serum concentration-time data using WinNonlin Enterprise version 3.3 (Pharsight, Mountain View, California): observed Cmin; observed Cmax; Tmax; AUC over a 24hour dosing interval (AUC~), estimated by the trapezoidal rule; apparent oral clearance (C1/F), calculated from the quotient of the oral dose and the AUC~; and the apparent volume of distribution during the terminal elimination phase based on steady-state data (V/F), calculated as the quotient of the oral dose and the product of the elimination rate constant (k z) and the AUC v where k was estimated by linear regression of the terminal linear portion of log-transformed concentration-time data. The tl/2 was estimated for zonisamide, calculated as loge2/)v~. Statistical Analysis

Zonisamide Blood samples (10 mL peripheral venous blood collected in a plain tube without anticoagulant) for

It was estimated that a sample size of 36 healthy female subjects would provide 30 evaluable subjects, but an additional 6 subjects could be enrolled if nec2059

CLINICALTHERAPEUTICS®

essary. Pharmacokinetic measures for EE and NOR obtained during OC dosing in the presence of zonisamide (cycle 3) were compared with those obtained during OC dosing alone (cycle 2). Geometric mean ratios (cycle 3:cycle 2) and the associated 2-sided 90% CIs were calculated based on the t distribution of mean log-transformed pharmacokinetic parameters (AUCv and Cmax). A repeated-measures linear model based on trough zonisamide levels before the last 3 zonisamide doses in cycle 3 was used to assess the attainment of steady state. The slope of the regression line fitting the 3 serum zonisamide concentrations was expected to be zero if steady state had been reached. The statistical test of the hypothesis that the slope was equal to zero was 2-sided, and significance was set at 0.05. A 2-sided, paired t test at the 5% level of significance was used to compare within-subject changes in LH, FSH, and progesterone concentrations in cycles 2 and 3, and 2-sided 95% CIs based on the t distribution were constructed for mean changes. RESULTS Study Subjects Forty-one subjects were enrolled in the study, and 37 received zonisamide. Eleven (26.8%) subjects were already taking the study OC and therefore did not require cycle 1. Four (10.8%) of the 37 subjects who received zonisamide withdrew from the study, 2 because of an AE and 2 at their request. The mean age of the 37 subjects who received zonisamide was 26.1 years (range, 18-51 years), and all were white. Mean body weight was 65.5 kg (range, 50.4-93.1 kg), and mean height was 165.8 cm (range, 152.4-182.9 cm). Thirty-three (89.2%) of the 37 subjects who received zonisamide completed the study. Of those subjects who completed the study, 26 (78.8%) were stabilized on zonisamide 400 mg/d, 3 (9.1%) were stabilized on 300 mg/d, and 4 (12.1%) were stabilized on 200 mg/d at the time EE and NOR profiles were assessed in cycle 3. Four subjects missed or took the wrong dose of zonisamide on 1 occasion; 1 subject missed a dose of OC but took it the following day. Because of the timing of the errors in relation to blood sampling, none of these dosing deviations should have affected any of the pharmacokinetic or pharmacodynamic measures. Six subjects had other minor protocol deviations, but

2060

none of these deviations were considered to have affected the outcome measures. Approximately one third of subjects used concomitant medications during the study (most commonly ibuprofen), but none of these medications was considered to have affected the outcome measures. Ethinyl Estradiol and Norethindrone Pharmacokinetics The pharmacokinetics of EE and NOR were assessed in all 37 subjects before administration of zonisamide (cycle 2) and in 33 (89.2%) of these 37 subjects after steady-state dosing of zonisamide (cycle 3). Mean serum EE and NOR concentrations at each time point were similar before and after zonisamide dosing, as can be seen from the concentration-time curves for EE (Figure 1) and NOR (Figure 2). Summary statistics for the pharmacokinetic parameters, including geometric mean ratios (cycle 3:cycle 2) and 90% CIs for AUCv and Cmax, are shown in Table I for EE and Table II for NOR. The 90% CIs for the ratios of EE AUCv (0.92-1.02) and Cmax (0.95--1.09) and NOR AUCv (1.06-1.15) and Cmax (1.02--1.12) fell well within the accepted range for lack of interaction (0.80-1.2513). Hence, the results indicate that steady-state dosing of zonisamide had no clinically relevant effect on the pharmacokinetics of EE or NOR. Zonisamide Pharmacokinetics Statistical analysis of trough serum zonisamide concentrations collected before the last 3 doses of zonisamide indicated no significant trend over time, confirming that steady-state concentrations had been achieved. Mean (SD) trough concentrations before the last 3 doses of zonisamide in the 26 subjects taking zonisamide 400 mg/d at the end of the dosing period were 23.41 (8.91), 24.13 (8.07), and 23.34 (7.21) l~g/mL. The mean tl/2 of zonisamide was 36.2 (8.1) hours (range, 22.6-60.3 hours), shorter than has been reported previously in the absence of enzymeinducing AEDs (-50-60 hoursM), although trough concentrations were broadly similar to those reported at this dose level in healthy subjects (-26 t~g/mLl~). Follicle-Stimulating Hormone, Luteinizing Hormone, and Progesterone Concentrations Pairwise statistical comparisons of serum LH and FSH concentrations in samples collected on days 13,

S.G. Griffith and Y. Dai

- I - Cycle 2 (n = 37) -~- Cycle 3 (n = 33) 200

-

180i

160-

i

140 E v

o_

120-

b

g

100-

(0 E E

80604020-

0

~ I

I

I

0

4

8

I

12

I

16

I

20

I

24

Time (h) profiles f o r ethinyl estradiol after multiple doses o f a combination ethinyl estradiol-norethindrone oral contraceptive in the absence (cycle 2) and presence of zonisamide at steady state

Figure I. Mean (SD) s e r u m c o n c e n t r a t i o n - t i m e

(cycle 3).

14, and 15 of cycle 3 during coadministration of zonisamide with corresponding samples collected in the absence of zonisamide during cycle 2 showed no significant differences in circulating LH or FSH concentrations, and the 95% CIs for the differences included zero (Table III). The analyses of serum progesterone concentrations showed no significant differences in circulating progesterone concentrations on days 14 and 15 of cycle 3 compared with days 14 and 15 of cycle 2, and the 95% CIs for the differences included zero (Table III). However, progesterone concentrations were significantly lower on day 13 of cycle 3 (in the presence of zonisamide) compared with day 13 of cycle 2 (P = 0.012); this is the opposite of what would be expected if zonisamide reduced the contraceptive effect. One subject had slightly elevated serum progesterone concentrations (>3 ng/mL) on days 13 to 15 during

cycle 2 (range, 3.50-3.81 ng/mL), but concentrations did not increase after administration of zonisamide in cycle 3 (range, 2.85-3.86 ng/mL); this subject~ serum EE and NOR levels also were unchanged between cycles 2 and 3. Based on the 3 pharmacodynamic variables measured, there was no indication that zonisamide was likely to reduce the contraceptive effectiveness of the study OC. Safety Assessm e nts

Treatment-emergent AEs (events occurring at any time after the first dose of zonisamide) were reported in 34 (91.9%) of the 37 subjects who received zonisamide along with the OC. Overall, the most frequently reported treatment-emergent AEs (>20%) were headache (21 subjects [56.8%]), nausea (16 [43.2%]), asthenia (9 [24.3%]), and dizziness (9 [24.3%]). The maximum intensity of any AE was

2061

CLINICAL THERAPEUTICS@

-0- Cycle 2 (n = 37) -O- Cycle 3 (n = 33) 30-

25

E

20

15-

8 L)

E 2

10-

5

0

I 0

I

I

I

I

I

I

4

8

12

I6

20

24

Time (h) Figure 2. Mean (SD) serum concentration-time profiles for norethindrone after multiple doses of a combination ethinyl estradiol-norethindrone oral contraceptive in the absence (cycle 2) and presence of zonisamide at steady state (cycle 3).

mild for 23 subjects (62.2%), moderate for 9 (24.3%), and severe for 2 (5.4%). No deaths or serious AEs occurred. Thirty-three subjects (89.2%) experienced AEs considered by the investigator to be related to zonisamide. Two subjects withdrew because of AEs considered by the investigator to be related to zonisamide. One subject was withdrawn due to a rash that started on the 26th day of zonisamide dosing (while the subject was taking zonisamide 400 mg/d) and resolved after 9 days. The other subject was withdrawn because of renal calculi, which became symptomatic on the 14th day of zonisamide dosing (while the subject was taking zonisamide 300 mg/d). This subject received analgesics as an outpatient and passed 2 ureteric stones. Another subject developed neutropenia (leukocyte count 2000/mm s, neutrophil differential 9.3%) that was noted after completion of 2062

zonisamide dosing (up to 400 mg/d). This AE was considered by the investigator to be related to a viral infection rather than to study treatment, and it resolved spontaneously. No other clinically important changes were noted in clinical laboratory measures or other safety parameters. DISCUSSION

At therapeutic doses, several commonly used AEDs are potent inducers of CYP3A4. This results in increased clearance of the components of OCs and may result in contraceptive failure of various OCs. 2-s,ls The findings of this study indicate that steady-state dosing of zonisamide at therapeutic levels had no clinically relevant effect on the pharmacokinetics of either EE or NOR in healthy female subjects taking a combination OC preparation. The 90% CIs for the geometric mean ratios of AUC~ and Cmax

S.G. Griffith and Y. Dai

Table I. Summary statistics for ethinyl estradiol pharmacokinetic parameters in the absence (cycle 2) and presence (cycle 3) of zonisamide.

Table II. Summary statistics for norethindrone pharmacokinetic parameters in the absence (cycle 2) and presence (cycle 3) of zonisamide.

Geometric Mean Ratio, Cycle 3:Cycle 2 (90% CI) (n = 33)

Geometric Mean Ratio, Cycle 3:Cycle 2 (90% CI) (n = 33)

Parameter AUC~, pg'h/mL Mean (SD) Median

Range

Cycle 2 (n = 37)

Cycle 3 (n = 33)

0.97 (0.9~ 1.02)

625 2060 569 1882

Mean (SD) Median Range

133 (39) 123 26 85

141 (46) 134 68 275

Cmin, pg/mL Mean (SD) Median Range

19 (9.2) 19 C>40.8

19 (7.2) 18 0 36.9

I. I I 6

I. I 0 2

CI/F, Uh Mean (SD) Median

Range

(n = 37)

Cycle 3 (n = 33)

AUC~, ng'h/mL 1139 (317) 1143 (312) I 122 I 126

Cm~×,pg/mL

Tm~×, h Median Range

Cycle 2 Parameter

33 (9.3) 31

1~56

1.02 (0.95 1.09)

33 (10.3) 3I 19 62

VJF, L

Mean (SD) Median

562 (196) 575 (402) 536 499

Range

203 1372

135 2500

Mean (SD)

140 (48)

159 (46)

Median Range

135 78 294

153 91 319

Cma×, ng/mL Mean (SD) Median Range

21 (5.4) 20 13~i I

23 (6.7)

Cmin,ng/mL Mean (SD) Median Range

1.9 (0.96) 2.3 (0.80) 1.7 2.3 0.03.2 0.5 3.8

Tma×, h Median Range CI/F, Uh Mean (SD) Median Range Vz/F, L Mean (SD) Median Range

I.II (I.06 1.15)

1.07 (i.02 1.12)

22 I I ~i8

I. I IQ

I. I I 3

7.9 (2.36)

6.7 (I.74)

7.4

6.5

3.4 12.8

3.1 I 1.0

I I I (40.2) 105 32 221

93 (32.7) 82 29 173

AUC, = AUC over a 24-hour dosing interval; CI/F = oral clearance ;V/F = apparent volume of distribution during the terminal elimination phase.

AUC, = AUC over a 24-hour dosing interval; CI/F = oral clearance ;V/F = apparent volume of distribution during the terminal elimination phase.

for both EE and NOR (range, 0.92-1.15) were contained within the accepted range for lack of interaction (0.80-1.2513). Other pharmacokinetic parameters for EE and NOR were similar in cycle 2 (in the absence of zonisamide) and cycle 3 (during coadministration of zonisamide). Similarly, there were no

includes AEDs such as phenytoin, carbamazepine, and phenobarbital--are potent inducers of the CYP isozymes responsible for drug metabolism and significantly increase the clearance of various drugsl,<15,17-21; they also increase SHBG concentrations. When drugs of this group are administered with OCs, both mechanisms may contribute to reductions in unbound levels of the estrogen and progestin components of the OC. Drugs in the second group--which includes AEDs such as topiramate, =,23 felbamate, 2~ and oxcarbazepine25,26--are less potent inducers of CYP isozymes but cause some increase in the apparent systemic clearance of OC steroids. Higher-dose OCs must be used to provide

changes in serum LH, FSH, or progesterone concentrations to suggest that coadministration of zonis-

amide would compromise the effectiveness of OCs containing EE or NOR. Ragueneau-Majlessi et a116 have categorized AEDs into 3 groups based on their potential to cause enzyme induction resulting in pharmacokinetic drug interactions. Drugs in the first g r o u p - - w h i c h

2063

CLINICALTHERAPEUTICS®

Table III. Summary of changes in concentrations of serum luteinizing hormone (LH), follicle-stimulating hormone (FSH), and progesterone between cycle 3 of oral contraceptive administration (in the presence of zonisamide) and cycle 2 (in the absence of zonisamide). Change in Parameter,Cycle 3

Parameter

Day 13 (n = 33)

Day 14 (n = 33)

Cycle 2 Day 15 (n = 33)

LH, mlU/mL 95% CI P*

0.13 ~).62 to 0.88 0.727

0.31 0.06 1.09 to 0.48 ~).63 to 0.74 0.431 0.87 I

FSH, mlU/mL 95% CI P*

0.11 0.05 0.12 q).61 to 0.84 q).61 to 0.71 q).49 to 0.74 0.752 0.873 0.686

Progesterone, ng/mL q). 18 O.13 0.06 95% CI q).32 to q).OZi q).29 to 0.04 q). 19 to 0.06 P* 0.012 O.133 0.308 *Paired t test.

effective contraception in women using these AEDs. Agents in the third group--which includes AEDs such as gabapentin, 27 tiagabine, 28 vigabatrin, 29 sodium valproate, 3° lamotrigine, 31 and levetiracetam 1 6 do not appear to induce CYP isozymes and do not significantly alter the pharmacokinetics of the estrogen or progestin components of OCs. Based on the results of the present study, zonisamide can be included in this third group of AEDs. The majority of subjects in this study successfully underwent titration to the target zonisamide dose of 400 mg/d. Because this was a study in healthy subjects, the titration schedule was faster than is recommended for patients with epilepsy. Despite rapid titration, zonisamide was reasonably well tolerated at the target dose of 400 mg/d when taken with a combination OC containing EE 0.035 mg/d and NOR 1 mg/d, although the incidence of AEs was somewhat higher than when dose escalation is slower. CONCLUSIONS Coadministration of zonisamide with the combination OC did not affect the pharmacokinetics of EE and NOR and did not increase serum concentrations of LH, FSH, or progesterone. Based on the findings of

2064

this study, there is no indication that zonisamide is likely to reduce the effectiveness of OCs containing EE and NOR. ACKNOWLEDGHENTS The authors acknowledge the study directors, Jaymin Shah, PhD, and Rose Kovelesky, PhD, for writing the protocol and implementing the clinical phase of the study, and the investigators and staff at PRACS Institute, Fargo, North Dakota, where the study was conducted. REFERENCES 1. Levy RH, Mather GG. Metabolic enzymes and antiepileptic drug interactions. Adv Neurol. 1998;76:49-55. 2. Kenyon IE. Unplanned pregnancy m an epileptic. BMJ. 1972;1:686-687. 3. Coulam CB, Annegers JE Do anticonvulsants reduce the efficacy of oral contraceptives? gpilepsia. 1979;20: 519-525. 4. Wilbur K, Ensom MH. Pharmacokinetic drug interactions between oral contraceptives and second-generation anticonvulsants. Clin Pharmacokinet. 2000;38:355-365. 5. Crawford P Interactions between antiepileptic drugs and hormonal contraception. CN5 Drugs. 2002;16: 263-272. 6. Golzieher JW Pharmacokinetics and metabolism of ethinyl estrogens. In: Golzieher JW, Fotherby K, eds. Pharmacology of the Contraceptive Steroids. New York: Raven Press; 1994:127-151. 7. Emery MG. Estrogens and progestins. In: Levy RH, Thummel KE, Trager WF, et al, eds. Metabolic Drug Interactions. Philadelphia: Lippincott Williams & Wilkins; 2000:511-528. 8. Sweetman S, ed. Martindale: The CompleteDrug Reference. 33rd ed. Chicago: Pharmaceutical Press; 2002. 9. Peters DH, Sorkm EM. Zonisamide. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in epilepsy Drugs. 1993;45: 760-787. 10. Leppick IE. Zonisamide. Epilepsia. 1999;40(Suppl 5): $23-$29. 11. Shah J, Shellenberger K, Canafax DM. Zonisamide: Chemistry, biotransformation, and pharmacokinetics. In: Levy RH, Mattson RH, Meldrum BS, Perucca E, eds. Antiepileptic Drugs. 5th ed. Philadelphia: Lippmcott Williams & Wilkins; 2002:873-879. 12. US Dept of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and

S.G. Griffith and Y. Dai

Research, and

Center for Veterinary Medicine.

Guidance for Industry: Bioanalytical Method Validation (May 2001). Available at: http://wwv¢fda.gov/cder/ guidance/4252fnl.pdf. Accessed October 1, 2004. 13. US Dept of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research, and Center for Biologics Evaluation and Research. Guidance /or Industry: In Vivo Drug

Metabolism/Drug Interaction Studies--Study Design, Data Analysis, and Recommendations /or Dosing and Labeling (November 1999). Available at: http:// www.fda.govlcderlguidance/2635fnl.pdf. Accessed October 1, 2004. 14. Kochak GM, Page JG, Buchanan RA, et al. Steady-state pharmacokinetics of zonisamide, an antiepileptic agent for treatment of refractory complex partial seizures. J Clin Pharmacol. 1998;38:166-171. 15. Crawford P, Chadwick DJ, Martin C, et al. The interaction of phenytoin and carbamazepine with combined oral contraceptive steroids. BrJ Clin Pharmacol. 1990; 30:892-896. 16. Ragueneau-Majlessi I, Levy RH, Janik E Levetiracetam does not alter the pharmacokinetics of an oral contraceptive in healthy women. Epilepsia. 2002;43:697-702. 17. Chang 5% McAuley JW Pharmacotherapeutic issues for women of childbearing age with epilepsy Ann Pharmacother. 1998;32:794-801. 18. Back DJ, Bates M, Bowden A, et al. The interaction of phenobarbital and other anticonvulsants with oral contraceptive steroid therapy. Contraception. 1980;22:

495-503. 19. Haukkamaa M. Contraception by Norplant subdermal capsules is not reliable in epileptic patients on anticonvulsant treatment. Contraception. 1986;33:559-565. 20. Odlind V, Olsson SE. Enhanced metabolism of levonorgestrel during phenytoin treatment in a woman with Norplant implants. Contraception. 1986;33:257-261. 21. Shane-McWhorter L, Cerveny JD, MacFarlane LL, Osborn C. Enhanced metabolism of levonorgestrel during phenobarbital treatment and resultant pregnancy. Pharmacotherapy. 1998;18:1360-1364.

22. Doose DR, Wang SS, Padmanabhan M, et al. Effect of topiramate or carbamazepine on the pharmacokinetics of an oral contraceptive containing norethindrone and ethinyl estradiol in healthy obese and nonobese female subjects. Epilepsia. 2003;44:540-549. 23. Rosenfeld WE, Doose DR, Walker SA, Nayak RK. Effect of topiramate on the pharmacokinetics of an oral contraceptive containing norethindrone and ethinyl estradiol in patients with epilepsy Epilepsia. 1997;38: 317-323. 24. Saano V, Glue P, Banfield CR, et al. Effects of felbamate on the pharmacokinetics of a low-dose combination oral contraceptive. Clin Pharmacol Ther. 1995;58:523531. 25. Klosterskov Jensen P, Saano V, Hating P, et al. Possible interaction between oxcarbazepine and an oral contraceptive. Epilepsia. 1992;33:1149-1152. 26. Fattore C, Cipolla C, Gatti G, et al. Induction of ethinylestradiol and levonorgestrel metabolism by oxcarbazepine in healthy women. Epilepsia. 1999;40: 783-787. 27. Eldon MA, Underwood BA, Randinitis EJ, Sedman AJ. Gabapentin does not interact with a contraceptive regimen of norethindrone acetate and ethinyl estradiol. Neurology. 1998;50:1146-1148. 28. Mengel HB, Houston A, Back DJ. An evaluation of the interaction between tiagabine and oral contraceptives in female volunteers. J Pharm Med. 1994;4:141-150. 29. Bartoli A, Gatti G, Cipolla G, et al. A double blind, placebo-controlled study on the effect of vigabatrin on in vivo parameters of hepatic microsomal enzyme induction and on the kinetics of steroid oral contraceptives in healthy female volunteers. Epilepsia. 1997;38: 702-707. 30. Crawford P, Chadwick D, Cleland P, et al. The lack of effect of sodium valproate on the pharmacokinetics of oral contraceptive steroids. Contraception. 1986;33:23-29. 31. Holdich T, Whiteman P, Orme M, et al. Effect of lamotrigine on the pharmacology of the combined oral contraceptive pill. Epilepsia. 1991;32(Suppl 1):96. Abstract.

Address correspondence to: Sue G. Griffith, MD, PhD, MRCP, Elan Pharmaceuticals, Inc., 7475 Lusk Boulevard, San Diego, CA 92121. E-marl: [email protected]

2065