Effect of Topiramate or Carbamazepine on the Pharmacokinetics of an Oral Contraceptive Containing Norethindrone and Ethinyl Estradiol in Healthy Obese and Nonobese Female Subjects

Authors


Address correspondence and reprint requests to Dr. M. Bialer at Department of Pharmaceutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Ein Karem, POB 12065, Jerusalem, 91120, Israel. E-mail: bialer@md.huji.ac.il

Abstract

Summary:  Purpose: To study the pharmacokinetics of a combination oral contraceptive (OC) containing norethindrone and ethinyl estradiol during OC monotherapy, concomitant OC and topiramate (TPM) therapy, and concomitant OC and carbamazepine (CBZ) therapy in order to comparatively evaluate the pharmacokinetic interaction, which may cause contraceptive failure.

Methods: This randomized, open-label, five-group study included two 28-day cycles. Five groups of female subjects received oral doses of ORTHO-NOVUM 1/35 alone (cycle 1) and then concomitant with TPM or CBZ (cycle 2). The treatment groups were group 1, TPM, 50 mg/day; group 2, TPM, 100 mg/day; group 3, TPM, 200 mg/day; group 4, TPM, 200 mg/day (obese women); and group 5, CBZ, 600 mg/day. Group 4 comprised obese women whose body mass index (BMI) was between 30 and 35 kg/m2. The BMI of the remaining four groups was ≤27 kg/m2.

Results: Coadministration of TPM at daily doses of 50, 100, and 200 mg (nonobese) and 200 mg (obese) nonsignificantly (p > 0.05) changed the mean area under the curve (AUC) of ethinyl estradiol by –12%, +5%, –11%, and –9%, respectively, compared with OC monotherapy. A similar nonsignificant difference was observed with the plasma levels and AUC values of norethindrone (p > 0.05). CBZ (600 mg/day) significantly (p < 0.05) decreased the AUC values of norethindrone and ethinyl estradiol by 58% and 42%, respectively, and increased their respective oral clearance by 69% and 127% (p < 0.05). Because CBZ induces CYP 3A–mediated and glucuronide conjugation metabolic pathways, the significant increase in the oral clearance of ethinyl estradiol and norethindrone was anticipated.

Conclusions: TPM, at daily doses of 50–200 mg, does not interact with an OC containing norethindrone and ethinyl estradiol. The lack of the TPM–OC interaction is notable when it is compared with the CBZ–OC interaction.

Interactions between the established antiepileptic drugs (AEDs), carbamazepine (CBZ), phenytoin (PHT), phenobarbital (PB), and combination oral contraceptives (OCs) have been recognized for many years (1–4). CBZ, PHT, and PB (2,5) have been reported to cause contraceptive failure, whereas valproic acid (VPA) has been shown not to affect the activity of OCs (6). In the last decade, the following new AEDs, have been approved for mono- and polytherapy for refractory epilepsy patients: felbamate (FBM), gabapentin (GBP), lamotrigine (LTG), levetiractam (LEV), oxcarbazepine (OCBZ), tiagabine (TGB), topiramate (TPM), vigabatrin (VGB), and zonisamide (ZNS) (7). It should be appreciated that the OC–AED interaction is not restricted only to epilepsy, as old and new AEDs have a broad use in other nonepileptic CNS disorders such as bipolar disorder, neuropathic pain, and migraine prophylaxis.

Ethinyl estradiol is well absorbed but is subject to an extensive first-pass effect on oral administration, with an average oral availability (or absolute bioavailability) of 51 ± 9%. In addition, wide variation occurs in its absolute bioavailability, ranging from 20 to 65%. The principal site of this presystemic elimination is the gut mucous membrane, with sulfate conjugation (sulfation) being the principal metabolic pathway (elimination), accounting for 60% of the first-pass metabolism (8–10). In addition to sulfation, ethinyl estradiol undergoes glucuronide conjugation and cytochrome P450 (CYP)3A-mediated 2-hydroxylation. Once absorbed, the major route of elimination (metabolism) is by hydroxylation (oxidation) at position 2 of the molecule with an average fraction metabolized (fm) of 30%. The CYP3A involvement in ethinyl estradiol metabolism was confirmed by its inhibition by troleandomycin (a CYP3A inhibitor) and induction by rifampin (9). After glucuronide conjugation, ethinyl estradiol undergoes enterohepatic circulation, and consequently, ∼30% of the dose can be recovered in the feces (8–10). A very small fraction of orally administered ethinyl estradiol is excreted unchanged in the urine [fraction excreted unchanged (fe) = 1–5%]. Ethinyl estradiol is subjected to extensive first-pass effect on oral administration; its absolute bioavailability ranges from 20 to 60%(9). The clearance of ethinyl estradiol is 5.4 ± 2.1 ml/min/kg; its volume of distribution is 3.5 ± 1.0 L/kg; and its half-life is 10 ± 6 h (8). The principal metabolic pathway is oxidation at position 2 and hydroxylations at the 4 and 16 positions (9).

The absolute bioavailability of norethindrone is 64 ± 15%, suggesting the presence of a significant first-pass effect. The clearance of norethindrone is 5.9 ± 1.1 ml/min/kg; its volume of distribution is 3.6 ± 1.7 L/kg; and its half-life is 8.5 ± 2.6 h (8). A very small fraction of orally administered norethindrone is excreted unchanged in the urine (fe = 4%), and the drug is eliminated mainly by metabolism. The major metabolic pathways are 5α- and β-reduction and 3α- and β-hydroxylation. Thus, the potential for induction of OC elimination by AEDs and other drugs exists in the CYP3A-mediated pathways (hydroxylations) and in glucuronide conjugation.

Rosenfeld et al. (11) studied the effect of TPM on this same combination OC (ORTHO-NOVUM 1/35, 1 mg norethindrone, and 35 μg ethinyl estradiol, daily) in 12 patients with epilepsy who were treated with VPA monotherapy. The OC was given for 21 days, followed by seven daily doses of inert tablets for four 28-day cycles. After the baseline cycle (cycle 1), escalating doses of TPM (200, 400, and 800 mg/day given in two divided doses) were administered in cycles 2 through 4, respectively. Compared with baseline, no significant change was seen in the pharmacokinetics (PK) of norethindrone in the presence of TPM (200–800 mg/day). However, the oral clearance (CL/F) values of ethinyl estradiol were 15–33% higher for cycles 2 and 4 compared with baseline (p < 0.05). The conclusion of the study was that when prescribing OC for patients receiving TPM, clinicians should consider initial therapy with an OC containing >35 μg ethinyl estradiol.

As more experience with the clinical use of TPM has been gathered in recent years, lower doses have been seen to show improved tolerability, while maintaining efficacy. In many countries, currently recommended initial target doses for monotherapy are 100 mg daily, given in two divided doses. Consequently, and as a follow-up of the study of Rosenfeld et al. (11), the primary objective of the current study was to evaluate the effect of TPM monotherapy at daily doses of 50, 100, and 200 mg, compared with 600 mg daily CBZ, on the PK of norethindrone and ethinyl estradiol after administration of ORTHO-NOVUM 1/35 to healthy female subjects. The secondary objective was to determine the effect of TPM monotherapy (200 mg/day) on the pharmacokinetics of norethindrone and ethinyl estradiol in obese women and to compare it with that in nonobese women. The tertiary objective was to determine the extent of a possible CYP3A4-induction by TPM compared with CBZ, by using the ratio of 6β-hydroxycortisol to cortisol in 24-h urine samples. The preliminary results of this interaction study were recently reported in brief abstract form (12,13) and are described in detail here.

METHODS

Subjects and study protocol

This was a randomized, open-label, five-group study that included two 28-day cycles. Five groups of female subjects received oral doses of ORTHO-NOVUM 1/35 alone (cycle 1), and then concomitant with TPM or CBZ (cycle 2). The treatment groups (by target dosages or obesity designation) were group 1, TPM, 50 mg/day; group 2, TPM, 100 mg/day; group 3, TPM, 200 mg/day; group 4, TPM, 200 mg/day (obese women); and group 5, CBZ, 600 mg/day. Group 4 comprised obese women whose body mass index (BMI) was between 30 and 35 kg/m2. The BMI of the remaining four groups was ≤27 kg/m2. The study was conducted under the requirements outlined in current Good Clinical Practice at Clinical Pharmacology Associates, Miami, Florida, U.S.A. The study protocol and amendments were reviewed by an independent Institutional Review Board (IRB).

The study proceeded as follows: During cycles 1 and 2, all subjects received 1 tablet of ORTHO-NOVUM 1/35 at 8 a.m. on days 1 to 21 and 1 placebo tablet of ORTHO-NOVUM 1/35 at 8 a.m. on days 22 to 28. On day 1 of cycle 2, titration of TPM or CBZ began, to achieve the assigned target dose by day 12 of cycle 2. The target dose of TPM or CBZ was maintained until day 21 of cycle 2 (p.m. dose). ORTHO-NOVUM 1/35 was provided as 28-day DIALPAK tablet dispensers. TPM was administered as 25-mg tablets. CBZ was administered as Tegretol, 200-mg scored tablets, or half of a 200-mg scored tablet.

During cycle 1, subjects were confined to the study unit from the evening of day 18 to the morning of day 22 for pharmacokinetic evaluations. On day 28 of cycle 1 (the day before day 1 of cycle 2), subjects were readmitted to the study unit in the evening for observation during TPM or CBZ dose titration. Subjects remained at the study unit until day 7 of cycle 2 (or longer, at the investigator's discretion). All subjects were confined to the study unit from the evening of day 18 of cycle 2 to the morning of day 22 for pharmacokinetic evaluations.

Fifty-five healthy subjects between the ages of 18 and 40 years completed the study and were included in the comparative pharmacokinetic analysis (TPM, 50 mg/day, n = 11; TPM, 100 mg/day, n = 10; TPM, 200 mg/day, n = 12; TPM, 200 mg/day (obese), n = 12; and CBZ, 600 mg/day, n = 10). All subjects had to be taking an OC containing norethindrone and ethinyl estradiol for ≥2 months before the beginning of the study. Subjects who had been taking a norethindrone- and ethinyl estradiol–containing OC other than ORTHO-NOVUM 1/35 were switched to ORTHO-NOVUM 1/35 on day 1 of cycle 1. Women were excluded from participation in the study if they were pregnant or nursing, wished to become pregnant within 2 months after administration of study drug, were taking hormone-replacement therapy, were taking a carbonic anyhdrase inhibitor, had a history of thromboembolus or clotting disorder, or had a family history of epilepsy.

Sample collection and analytic procedures

On day 20 of cycles 1 and 2 for all five groups, 10-ml blood samples were collected immediately before and at the following times after the morning dose of ORTHO-NOVUM 1/35: 1, 2, 3, 4, 6, 8, 12, 16, and 24 h. Blood samples were drawn by direct veinpuncture or through an indwelling peripheral venous heparin lock catheter and collected into vacuum tubes containing ethylenediamine tetraacetic acid (EDTA) anticoagulant. After centrifugation, plasma was pipetted into a separate plastic tube and frozen at –20°C until analyzed.

Plasma samples were analyzed for norethindrone and ethinyl estradiol concentrations by a validated gas chromatography/mass spectroscopy (GC/MS), negative-ion method at PPD Development (Richmond, VA, U.S.A.) The quantification range was 0.05 ng/ml [limit of quantification (LOQ)] to 25 ng/ml for norethindrone and 2 pg/ml (LOQ) to 1,000 pg/ml for ethinyl estradiol. The interassay precision was <9.2% and <10.7% coefficient of variation for norethindrone and ethinyl estradiol, respectively.

As an enzyme-induction assessment to determine and quantify the extent of cytochrome P450 (CYP) induction, CYP3A induction status was evaluated by measurement of the ratio of 6β-hydroxycortisol to cortisol in 24-h urine samples collected on days 21–22 of cycles 1 (baseline) and 2 (TPM or CBZ treatment). Urine samples were analyzed for 6β-hydroxycortisol and cortisol concentrations by using a validated HPLC–MS/MS method at MDS Pharma Services, Inc. (St-Laurent, Montreal, Quebec, Canada). The quantification range was 2–500 ng/ml for cortisol and 50 to 1,923 ng/ml for 6β-hydroxycortisol. The LOQ was 2 and 50 ng/ml for cortisol and 6β-hydroxycortisol, respectively. The interassay precision was <9.5% and <6.6% coefficient of variation for cortisol and 6β-hydroxycortisol, respectively.

Pharmacokinetic parameters and statistical analyses

The following pharmacokinetic parameters for each component of ORTHO-NOVUM 1/35, norethindrone and ethinyl estradiol, were estimated from their respective plasma-concentration data: (a) peak plasma concentration (Cmax); (b) time to reach Cmax (tmax); (c) area under the plasma drug (norethindrone or ethinyl estradiol) concentration–time curve over a dosing interval (24 h) (AUC), estimated by linear trapezoidal summation before Cmax, and log-linear trapezoidal summation after Cmax; and (d) oral clearance (CL/F) of norethindrone and ethinyl estradiol, calculated from the quotient of the oral dose and AUC.

The pharmacokinetic parameters (Cmax, AUC, and CL/F) of norethindrone and ethinyl estradiol were compared between the two treatments (with and without TPM or CBZ) and within each group (groups 1–5) by using ANOVA (analysis of variance). The pharmacokinetic parameters of interest were analyzed on log-transformed values. An ANOVA model was fit to one of the log-transformed pharmacokinetic parameters of interest (Cmax, AUC, or CL/F) with group, subject nested within group, treatment (with and without TPM or CBZ), and Group by treatment interaction as factors. Linear contrasts were set up to compare the two treatments (with and without TPM or CBZ) within each group (groups 1–5). The ratio of means and the 90% confidence intervals for the ratio of mean pharmacokinetic parameters from test treatment (with TPM or CBZ) to reference treatment (without TPM or CBZ) were constructed for each group (groups 1–5). Individual group analysis was done by appropriate linear contrast by using the ANOVA model with group, subject within group, treatment, and group by treatment as factors. In addition, the “poolability” of all TPM groups (groups 1–4: TPM, 50 mg, 100 mg, 200 mg nonobese, and 200 mg obese) was evaluated by using simultaneous linear contrasts in the ANOVA model. If the TPM groups were “poolable” statistically (at α = 10%), then the treatment effect (with and without TPM) within the pooled TPM group was compared by using a linear contrast in the ANOVA. The ratio of the means and the 90% confidence intervals for the ratio of the mean pharmacokinetic parameters from the two treatments (with and without TPM) were constructed separately for each TPM dose and for the pooled TPM group.

RESULTS

The mean (SD) plasma concentration–time profile of norethindrone and ethinyl estradiol obtained after oral administration of ORTHO-NOVUM 1/35 with and without TPM (100 or 200 mg, daily) are presented in Figs. 1–4. The mean (SD) parameters of norethindrone and ethinyl estradiol given alone or after TPM coadministration are presented in Tables 1 and 2. The log-transformed Cmax and AUC ratio of the means of norethindrone and ethinyl estradiol and their 90% confidence interval for each TPM dose are presented in Tables 3 and 4. Statistical analysis confirmed that the Cmax and AUC values of all TPM groups were “poolable,” and thus, the log-transformed Cmax and AUC ratio of the means of norethindrone and ethinyl estradiol and their 90% confidence interval for the pooled TPM doses are presented in Table 5. The 90% confidence intervals for the pooled TPM data were within the 80–125% boundaries, whereas the confidence interval of each TPM dose deviated from this range, but it narrowed and became closer to (or within) the 80–125% range after exclusion of a couple of outliers (Tables 3 and 4).

Figure 1.

Mean (SD) plasma concentration–time profiles of norethindrone after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of topiramate (100 mg/day TPM, nonobese).

Figure 2.

Mean (SD) plasma concentration–time profiles of norethindrone after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of topiramate (200 mg/day TPM, nonobese, and 200 mg/day, obese).

Figure 3.

Mean (SD) plasma concentration–time profiles of ethinyl estradiol after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of topiramate (100 mg/day TPM, nonobese).

Figure 4.

Mean (SD) plasma concentration–time profiles of ethinyl estradiol after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of topiramate (200 mg/day TPM, nonobese, and 200 mg/day, obese).

Table 1.  Norethindrone pharmacokinetic parameters and associated ANOVA results after coadministration of topiramate in healthy nonobese or healthy obese subjects
  ORTHO-NOVUM 1/35b  
TreatmentParameterAlone
(cycle 1)
With topiramate
(cycle 2)
p valuec% change in mean
(cycle 2–cycle 1)/cycle 1
  • a

     Median value.

  • b

     Arithmetic mean (SD).

  • c

     Analysis on log-transformed data.

  • d

     Values after excluding outlier data from three subjects.

Topiramate, 50 mg/day (nonobese)tmax (h)a11 
 Cmax (ng/ml)16.3 (5.2)15.6 (5.5)0.8155−4.2
 AUC (ng · h/ml)119 (49)108 (41)0.6728−8.8
 CL/F (L/h)9.8 (3.8)10.5 (3.8)0.67317.0
Topiramate, 100 mg/day (nonobese)tmax (h)a11 
 Cmax (ng/ml)15.7 (8.8)15.9 (4.8)0.19971.1
 AUC (ng · h/ml)100 (63)108 (48)0.08447.9
 CL/F (L/h)39.2 (83)12.0 (7.9)0.0844−69.2
 CL/F (L/h)d12.9 (9.6)11.7 (8.3)0.6451−9.4
Topiramate, 200 mg/day (nonobese)tmax (h)a11 
 Cmax (ng/ml)17.1 (7.9)16.8 (9.2)0.4251−1.9
 AUC (ng · h/ml)112 (56)98.5 (47)0.3560−11.8
 CL/F (L/h)14.0 (15)20.1 (31)0.354843.5
 CL/F (L/h)d14.2 (16)11.3 (6.3)0.8152−20.5
Topiramate, 200 mg/day (obese)tmax (h)a11 
 Cmax (ng/ml)11.4 (3.8)11.7 (3.5)0.84502.4
 AUC (ng · h/ml)79.2 (43)91.4 (38)0.288315.5
 CL/F (L/h)16.1 (8.0)12.8 (5.4)0.2886−20.5
Table 2.  Ethinyl estradiol pharmacokinetic parameters and associated ANOVA results after coadministration of topiramate in healthy nonobese or healthy obese subjects
  ORTHO-NOVUM 1/35b  
TreatmentParameterAlone
(cycle 1)
With topiramate
(cycle 2)
p valuec% change in mean
(cycle 2–cycle 1)/cycle 1
  • a

     Median value.

  • b

     Arithmetic mean (SD).

  • c

     Analysis on log-transformed data.

  • d

     Values after excluding outlier data from three subjects.

Topiramate, 50 mg/day (nonobese)tmax (h)a11 
 Cmax (pg/ml)139 (46)128 (44)0.6177−7.6
 AUC (pg · h/ml)1,349 (413)1,187 (421)0.2949−12.0
 CL/F (L/h)28.2 (8.2)33.9 (15)0.294820.3
Topiramate, 100 mg/day (nonobese)tmax (h)a11 
 Cmax (pg/ml)129 (56)135 (27)0.24505.3
 AUC (pg · h/ml)1,113 (429)1,177 (294)0.30795.8
 CL/F (L/h)47.1 (53)31.6 (8.7)0.3080−32.9
 CL/F (L/h)d30.4 (7.7)31.1 (9.1)0.84232.3
Topiramate, 200 mg/day (nonobese)tmax (h)a11 
 Cmax (pg/ml)130 (34)111 (40)0.1004−15.2
 AUC (pg · h/ml)1,113 (298)987 (331)0.1578−11.3
 CL/F (L/h)33.5 (8.9)51.3 (60)0.157753.1
 CL/F (L/h)d33.7 (9.3)34.2 (7.5)0.70581.5
Topiramate, 200 mg/day (obese)tmax (h)a11 
 Cmax (pg/ml)95.4 (40)85.2 (17)0.7248−10.7
 AUC (pg · h/ml)911 (433)826 (148)0.8231−9.4
 CL/F (L/h)45.1 (18)44.0 (10)0.8228−2.3
Table 3.  Norethindrone natural log transformed Cmax and AUC parameter ratio of means and 90% confidence intervals for the ratio of means for treatment (with topiramate) to reference (without topiramate) for each topiramate and carbamazepine treatment groups
 Geometric means 90% confidence interval for the
ratio of the means
ParameterTreatment
(with topiramate)
Reference
(without topiramate)
Estimated ratio (%)
of geometric mean
Lower bounda
(% reference)
Upper bounda
(% reference)
  • a

     Bounds as percentage of the reference treatment (with topiramate/without topiramate) geometric mean.

  • b

     After excluding outlier data from three subjects.

TPM, 50 mg/day     
 Cmax (ng/ml)15169564141
 AUC (ng · h/ml)1021109268126
 Cmaxb (ng/ml)15169579113
 AUCb (ng · h/ml)1021109277111
TPM 100 mg/day TPM, TPM     
 Cmax (ng/ml)151113891210
 AUC (ng · h/ml)9769141102195
 Cmaxb (ng/mL)15169275112
 AUCb (ng · h/ml)1019510686130
TPM 200 mg/day     
 Cmax (ng/ml)13158357122
 AUC (ng · h/ml)80958563114
 Cmaxb (ng/ml)171510991131
 AUCb (ng · h/ml)999610385134
TPM, 200 mg/day (obese)     
 Cmax (ng/ml)111110571153
 AUC (ng · h/ml)857012190163
 Cmaxb (ng/ml)111110588124
 AUCb (ng · h/ml)8570121101144
CBZ 600 mg/day     
 Cmax (ng/ml)1015634296
 AUC (ng · h/ml)47100473465
 Cmaxb (ng/ml)1116695684
 AUCb (ng · h/ml)54108504061
Table 4.  Ethinyl estradiol natural log transformed Cmax and AUC parameter ratio of means and 90% confidence intervals for the ratio of means for treatment (with topiramate) to reference (without topiramate) for each topiramate and carbamazepine treatment groups
 Geometric means 90% confidence interval for the
ratio of the means
ParameterTreatment
(with topiramate)
Reference
(without topiramate)
Estimated ratio (%)
of geometric mean
Lower bounda
(% reference)
Upper bounda
(% reference)
  • a

     Bounds as percentage of the reference treatment (with topiramate/without topiramate) geometric mean.

  • b

     After excluding outlier data from three subjects.

TPM 50 mg/day     
 Cmax (pg/ml)1211329168123
 AUC (pg · h/ml)1,1121,2938668109
 Cmaxb (pg/ml)1211329180105
 AUCb (pg · h/ml)1,1121,293867598
TPM 100 mg/day     
 Cmax (pg/ml)13310712491168
 AUC (pg · h/ml)1,14398011791150
 Cmaxb (pg/ml)1311369683112
 AUCb (pg · h/ml)1,1651,1859885114
TPM 200 mg/day     
 Cmax (pg/ml)951267657100
 AUC (pg · h/ml)8861,0788265103
 Cmaxb (pg/ml)1171239583108
 AUCb (pg · h/ml)1,0441,0769785111
TPM 200 mg/day (obese)     
 Cmax (pg/ml)83889471125
 AUC (pg · h/ml)8128379777122
 Cmaxb (pg/ml)83889483107
 AUCb (pg · h/ml)8118379786110
CBZ 600 mg/day     
 Cmax (pg/ml)861127756104
 AUC (pg · h/ml)5411,005544269
 Cmaxb (pg/ml)97115857399
 AUCb (pg · h/ml)6501,051625471
Table 5.  Norethindrone and ethinyl estradiol natural log-transformed Cmax and AUC parameter ratio of means and 90% confidence intervals for the ratio of means for treatment (with topiramate) to reference (without topiramate) for pooled topiramate treatment groups
 Geometric means 90% confidence interval for the
ratio of the means
ParameterTreatment
(with topiramate)
Reference
(without topiramate)
Estimated ratio (%)
of geometric mean
Lower bounda
(% reference)
Upper bounda
(% reference)
  • a

     Bounds as percentage of the reference treatment (with topiramate/without topiramate) geometric mean.

  • b

     Values after excluding outlier data from three subjects.

Norethindrone     
 Cmax (ng/ml)13.213.0102.284.0124.3
 AUC (ng · h/ml)89.984.3106.691.2124.6
 Cmaxb (ng/ml)14.114.1100.391.7109.7
 AUCb (ng · h/ml)95.690.8105.395.8115.8
Ethinyl estradiol     
 Cmax (pg/ml)10511293.881.1108.5
 AUC (pg · h/ml)9681,03193.983.4105.7
 Cmaxb (pg/ml)11011794.188.0100.6
 AUCb (pg · h/ml)1,0121,07394.388.4100.7

The coadministration of TPM at daily doses of 50, 100, or 200 mg did not significantly change the plasma profile or pharmacokinetic parameters of norethindrone (change in mean AUC or Cmax, <12%; p > 0.05). The 69% decrease and 44% increase (Table 1) in norethindrone mean oral clearance (CL/F), in the 100- and 200-mg/day groups, respectively, is the result of the sensitivity of this PK parameter to a single outlier in each of these two groups. Similarly, the coadministration of TPM at daily doses of 50, 100, or 200 mg did not significantly change the plasma profile or pharmacokinetic parameters of ethinyl estradiol (p > 0.05). The 33% decrease (100 mg/day TPM) and 53% increase (200 mg/day TPM) in ethinyl estradiol mean oral clearance (CL/F), is the result of the sensitivity of this PK parameter to a single outlier in each of these two groups, and it disappeared after excluding the single outlier. The two outliers had very low AUC levels, probably because of lack of compliance, which resulted in a huge CL/F overestimate that significantly affected the mean CL/F, but not the (high) mean AUC value. Norethindrone and ethinyl estradiol plasma levels from the obese subjects were lower than those in the nonobese because of the lower dose per body weight. Nevertheless, TPM coadministration did not significantly change the plasma levels or the PK parameters of norethindrone or ethinyl estradiol (p > 0.05).

The mean (SD) plasma concentration of norethindrone and ethinyl estradiol after CBZ administration (600 mg daily) is depicted in Figs. 5 and 6. The mean (SD) parameters of norethindrone and ethinyl estradiol given alone or after CBZ coadministration are presented in Table 6. The log-transformed Cmax and AUC ratio of the means of norethindrone and ethinyl estradiol and their 90% confidence interval for the CBZ-treated group are presented in Table 3. These 90% confidence interval values deviated dramatically from the 80–125% boundaries, and in all cases except the Cmax of ethinyl estradiol, did not contain the 100%.

Figure 5.

Mean (SD) plasma concentration–time profiles of norethindrone after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of carbamazepine (600 mg/day CBZ, nonobese).

Figure 6.

Mean (SD) plasma concentration–time profiles of ethinyl estradiol after multiple doses of ORTHO-NOVUM 1/35 with or without coadministration of carbamazepine for group 5 (600 mg/day CBZ, nonobese).

Table 6.  Norethindrone and ethinyl estradiol pharmacokinetic parameters after coadministration of carbamazepine, 600 mg/day, to healthy nonobese subjects
ParameterAlone
(cycle 1)
With carbamazepineb
(cycle 2)
p valuec% change in mean
(cycle 2–cycle 1)/cycle 1
  • a

     Median value.

  • b

     Arithmetic mean (SD).

  • c

     Analysis on log-transformed data.

  • d

     Significant at α = 0.05.

Norethindrone    
 tmax (h)a11 
 Cmax (ng/ml)17.2 (7.5)10.8 (5.2)0.0721−37.3
 AUC (ng · h/ml)126 (77)53.1 (24)0.0003d−57.9
 CL/F (L/h)14.9 (19)25.1 (18)0.0003d68.7
Ethinyl estradiol    
 tmax (h)a11 
 Cmax (pg/ml)117 (34)95.5 (41)0.1505−19.3
 AUC (pg · h/ml)1,062 (345)616 (244)0.0001d−42.0
 CL/F (L/h)37.2 (16)84.3 (91)0.0001d57.5

As an enzyme-induction assessment to determine and quantify the extent of cytochrome P450 (CYP) induction, CYP3A induction status was evaluated by measurement of the ratio of 6β-hydroxycortisol to cortisol in 24-h urine samples collected on days 21–22 of cycles 1 (baseline) and 2 (TPM or CBZ treatment). As depicted in Table 5, the change in this ratio during TPM administration compared with baseline was not statistically significant. In contrast, a statistically significant increase was observed in the CBZ-treated group. The mean increase in the ratio for the CBZ group was 461%(Table 7).

Table 7.  6β-Hydroxycortiso to cortisol ratios and percentage changes and ANOVA p values after oral administration of ORTHO-NOVUM 1/35 with and without coadministration of topiramate or carbamazepine in healthy female volunteers
 6β-Hydroxycortisol/cortisol ratiob  
TreatmentCycle 1Cycle 2p valueaMean % change
(cycle 2–cycle 1)/(cycle 1)
  • a

     Statistical comparison (p values) from comparison of 6β-hydroxycortisol-to-cortisol ratio data.

  • b

     Arithmetic mean (SD).

Topiramate, 50 mg/day4.93 (2.55)5.48 (1.74)0.818033 (59)
Topiramate, 100 mg/day6.26 (2.16)7.43 (1.94)0.521732 (49)
Topiramate, 200 mg/day6.30 (2.83)9.54 (2.37)0.081982 (75)
Topiramate, 200 mg/day (obese)6.83 (2.37)5.86 (2.15)0.7987−8 (17)
Carbamazepine, 600 mg/day5.32 (2.30)25.5 (12.8)0.0001461 (284)

During cycle 1, when the subjects were taking ORTHO-NOVUM 1/35 alone, adverse events occurred in two (17%) subjects in group 4 and one (8%) subject in group 5. When TPM or CBZ was administered concurrently with ORTHO-NOVUM 1/35 during cycle 2, adverse events occurred in all subjects in groups 1, 2, 4, and 5, and in 11 of the 12 subjects in group 3. The most common adverse events among the 45 TPM-treated subjects who participated in cycle 2 were somnolence (78%), headache (51%), and nausea (33%). No clear relation was observed between increasing TPM dose and increasing incidence of any specific adverse event. The incidences of the following events were greater among the TPM-treated subjects than the CBZ-treated subjects: headache (51% vs. 30%); hypoesthesia (27% vs. 0); and dizziness (18% vs. 0). The incidences of the following events were higher among the CBZ-treated subjects than the TPM-treated subjects: nausea (50% vs. 33%); and pruritus (40% vs. 13%).

All of the adverse events reported were mild or moderate in severity, and the majority were considered to be possibly related to TPM or CBZ. No serious adverse event occurred, and most events resolved without treatment. No subject was withdrawn from the study because of an adverse event. No subject in any treatment group had a noteworthy change from prestudy to the final evaluation in laboratory analyte values, vital signs measurements, physical examination findings, or electrocardiograms.

DISCUSSION

In a previous study of adjunctive TPM therapy to VPA, Rosenfeld et al. (11) reported the mean exposure (AUC) of ethinyl estradiol decreased by 18–30% (compared with OC monotherapy) in a dose-related fashion, when ORTHO-NOVUM 1/35 was coadministered with TPM daily doses of 200, 400, or 800 mg. This was due to an increase in its mean oral clearance (CL/F) by 15% (200 mg/day), 18% (400 mg/day), and 33% (800 mg/day) (p < 0.05). However, no significant change occurred in the serum levels or oral clearance of norethindrone (11). In the present study, coadministration of TPM at daily doses of 50, 100, and 200 mg (nonobese) and 200 mg (obese) insignificantly changed the mean AUC of ethinyl estradiol by –12%, +5%, –11%, and –9%, respectively, compared with the OC alone. A similar nonsignificant difference was observed with the plasma levels and AUC values of norethindrone (p > 0.05). Although the 90% confidence interval for the AUC and Cmax ratio (with and without TPM) for each TPM dose deviated from the 80–125% range, it did contain the value of 100%(Tables 3 and 4). Excluding outliers because of compliance problems narrowed the confidence intervals, brought it closer to the 80–125% range, and in the case of ethinyl estradiol (Table 4), it was within the 80–125% range (except for the lowest dose-AUC). The pooled TPM data were wholly contained within the 80–125% boundaries (Table 5). In contrast, the CBZ-treated group deviated dramatically from the 80–125% range, even after excluding outliers, and in all cases (except the Cmax of ethinyl estradiol) did not contain the value of 100%(Tables 3 and 4).

With respect to the change in mean total exposure (AUC) of norethindrone and ethinyl estradiol at the 200 mg/day TPM dose, the results of the present study were similar to the results reported by Rosenfeld et al. (11). In both studies, TPM had little or no effect on norethindrone. In the previous study, TPM (200 mg/day) reduced mean AUC of ethinyl estradiol by 18%, whereas in the present study, mean AUC was reduced to a similar amount, 11%. However, in the previous study, a dose-related reduction of ethinyl estradiol total exposure by concomitant administration of TPM (200, 400, 800 mg/day) was observed, whereas in the present study, a dose-related reduction was not observed. This absence of a dose-related reduction is consistent with the weak (or absence at low doses) enzyme-induction potential of TPM. At TPM doses of ≤200 mg/day, this induction is dose independent and insignificant. At TPM doses >200 mg/day, this induction becomes dose-dependent and occurs to a greater extent.

With respect to the control of experimental variability, or validity of study design, the present study may be considered better, compared with that reported by Rosenfeld et al. (11). The present study assessed the interaction potential as TPM monotherapy, rather than the previous study's assessment as TPM adjunctive therapy to VPA. In addition, the present study quantified both norethindrone and ethinyl estradiol by the more specific GC/MS analytic technique, compared with the radioimmunoassay technique used in the previous study. For these reasons, the calculated Cmax and AUC 90% confidence interval ratios for norethindrone and ethinyl estradiol (during TPM treatment/without TPM treatment) were tight, and therefore, the results of the present study should be considered reliable with respect to the clearance-induction potential of TPM on norethindrone and ethinyl estradiol.

As stated earlier, the secondary objective of the current study was to determine the effect of TPM (200 mg/day) on the pharmacokinetics of norethindrone and ethinyl estradiol in obese females. With the exception of obese subjects having lower peak plasma concentration (Cmax) and systemic exposure (AUC) as a result of the lower dose per kilogram body weight, the lack of significant effect of TPM on norethindrone and ethinyl estradiol was similar in obese subjects to that in nonobese subjects.

Thus it can be concluded that TPM daily doses of 50–200 mg do not interact with OCs containing norethindrone and ethinyl estradiol. The lack of the TPM–OC interaction is more striking when it is compared with the CBZ–OC interaction observed in the current study. CBZ (600 mg/day) decreased mean AUC values of norethindrone and ethinyl estradiol by 58 and 42%, respectively, and increased their respective mean oral clearance by 69 and 127% (p < 0.05). The increase in the oral clearance is due to an induction by CBZ of the CYP3A-mediated and glucuronide conjugation metabolic pathways of ethinyl estradiol and norethindrone (8–10).

In contrast to CBZ, the change in the 6β-hydroxycortisol/cortisol ratio, which is a marker for CYP3A induction, during TPM administration compared with baseline was not statistically significant, confirming the low induction potential of TPM.

The effect of CBZ (600 mg/day), on the PK of norethindrone and ethinyl estradiol was evaluated in the present study to put the PK results from the TPM-treated groups in perspective. CBZ has been reported to decrease significantly the plasma concentrations of the progestin and estrogen in a combination OC (2–5). In the present study, CBZ (600 mg/day) had a significant effect on the PK of norethindrone and ethinyl estradiol. Mean Cmax and AUC values of norethindrone and ethinyl estradiol were reduced (37.3/19.2%) and (57.9/42.0%) (norethindrone/ethinyl estradiol), respectively. Given the extent of these reductions, CBZ would be expected and has been observed to decrease the efficacy of OCs and to cause contraceptive failure (2–5). The clinical significance of the small reduction in norethindrone and ethinyl estradiol Cmax and AUC as a result of TPM coadministration is not known. However, it is clear that TPM does not result in a reduction similar to that seen with CBZ. The absence of a clinically significant effect of TPM on norethindrone and ethinyl estradiol PK is further suggested by the observation that for both norethindrone and ethinyl estradiol Cmax and AUC values, the 90% confidence interval bounds of parameter ratios (during TPM treatment/without TPM treatment) were within bioequivalence limits or acceptable range for lack of interaction (80–125%) when the TPM-treated group data (groups 1–4) were pooled. At doses ≤200 mg/day, TPM does not have a statistically significant effect on norethindrone and ethinyl estradiol plasma levels.

In a recent article, Ragueneau et al. (14) categorized AEDs into three groups on the basis of their potential to cause induction-drug interactions. Old AEDs, such as PHT, PB, and CBZ may reduce steroid OC levels during concomitant administration because of induction by CYP3A and other metabolizing enzymes and increase in OC-binding globulin levels (3,15). Both mechanisms may contribute to reduced OC levels during concomitant OC–AED administration (14). A second group of AEDs includes TPM (11), OCBZ (16–18), and FBM (19), which have been reported to alter OC plasma concentrations. A third group of AEDs includes GBP (20), LTG (21), LEV (14), TGB (22), VGB (23), and VPA (6), which do not alter the PK of OCs (3,4,14). The current study shows that TPM should be categorized by the dose used. At daily doses of ≤200 mg, TPM belongs to the third group rather than to the second one.

In summary, OC failure rates are high among women taking enzyme-inducing AEDs such as CBZ (3–5). The AUCs of norethindrone and ethinyl estradiol changed by <12% in a non–dose-related fashion during coadministration of TPM at daily doses of 50–200 mg. The clinical implications of these small and nonsignificant changes in norethindrone and ethinyl estradiol are not known. However, considering the significant effect that CBZ coadministration (600 mg/day) had on the reduction in plasma levels and AUC values of norethindrone and ethinyl estradiol, it is clear that TPM does not cause a similar reduction in OCs containing norethindrone or ethinyl estradiol.

Ancillary