- Top of page
Summary: Purpose: To study the interaction between lamotrigine (LTG) and hormonal contraception.
Methods: LTG serum concentrations of female patients using either no hormonal contraception (n = 18), an ethinyl estradiol (EE)-containing (n = 11), or a progestogen (PG)-only–containing compound (n = 16) were analyzed. Patients were recruited prospectively, and blood samples were drawn during drug fasting and at steady-state conditions. Comedication with enzyme inducers, valproate, topiramate, or sertraline was not allowed. Some patients changed groups and thus served as their own controls. Samples were analyzed by a gas chromatography/ mass spectroscopy method. The Mann–Whitney U test was used for statistical comparison of the groups.
Results: The LTG serum concentration-to-dose ratio (CDR), expressed as (mg/L)/(mg/d) was significantly lower in women using EE than in the control group (mean ± SD, 0.010 ± 0.004 vs. 0.017 ± 0.006; p = 0.003). The CDR in women using PG was 0.02 ± 0.007, which was not statistically different from controls. No difference was found in CDR between women using either oral, topical, or parenteral PG. Five women switched from the control to the EE group and experienced a considerable reduction in CDR. An increase of the CDR toward control level was seen in the two women who changed from EE to PG.
Conclusions: It is the EE component of oral contraceptives that interacts with LTG. The PG-only compounds did not alter LTG serum concentrations in this study. These findings should be considered when counselling women with epilepsy in the childbearing ages.
Lamotrigine (LTG) is a broad-spectrum antiepileptic drug (AED) that is increasingly used worldwide for both epileptic and psychiatric disorders. Currently, much attention is directed toward possible endocrine side effects of AEDs in women (1). LTG has so far not been reported to be associated with such disturbances and has therefore been regarded as a preferable choice in this patient group (1,2). However, its interaction potential with sexual hormones has still not been extensively studied. Women of childbearing age often use hormonal contraception, and it has recently been reported that combined oral contraceptives reduce LTG serum concentrations by ≤64% (3,4), which may require dose adjustment of LTG. Most oral contraceptives contain a combination of an estrogen derivative, ethinyl estradiol (EE), and a progestogen (PG). In addition to oral compounds, hormone-containing parenteral, intravaginal, and intrauterine products are on the market. We wanted to know whether the interaction between LTG and hormonal contraception is due to the EE or the PG component. Moreover, we wanted to investigate whether this interaction is restricted to oral application of the hormones or whether it also can be seen with topical/parenteral methods.
- Top of page
This was a three-armed, open, prospective study. LTG serum concentrations of consecutively enrolled female patients of childbearing age using no hormonal contraception, an EE-containing compound, or a PG-only-containing compound were analyzed.
Some patients participated in more than one group and thus served as their own controls.
Concomitant treatment with drugs known or suspected to reduce (enzyme inducers, topiramate) or increase (valproate, sertraline) LTG serum concentrations was not allowed (1,5–7). Patients with liver or kidney disease, known compliance problems, or a history of drug abuse were not included. Blood samples were drawn during drug fasting in the morning and at steady-state conditions, but not standardized in relation to the menstrual cycle. Time intervals between blood samplings in patients who changed group were ≥1 month, and their LTG doses remained unchanged.
All samples were analyzed with a liquid chromatography/mass spectrometry (LC/MS) method. LTG was extracted from 0.1 ml serum with 0.5 ml dichloromethane/isopropanol (9:1) after addition of internal standard solution (50 μl of 10 μg/ml minoxidil in methanol) and alkalinization with 0.1 ml 0.2M sodium bicarbonate. After mixing and centrifugation, the organic extract was transferred to vials and injected on an Agilent MSD 1100 LC/MS system (Agilent, Palo Alto, CA, U.S.A.). The LC/MS system consisted of a G1379A degasser, a G1311A quaternary pump, a G1313A autosampler, a G1316A column oven, and a G1946A mass spectrometer. Separation was performed on a SB-LC18 Zorbax (30 × 4.6 mm, 3 μm) column with a mobile phase consisting of methanol/ammonium acetate, 45:55. LTG was monitored after positive electrospray ionization at m/z 256.0, and the internal standard minoxidil, at m/z 210.2. The calibrated range was from 0.13 to 25.6 μg/ml. Six quality-control samples covering the range from 0.5 to 11.5 μg/ml were analyzed with every batch of unknown samples. Between-day relative standard deviation calculated from quality-control samples was >9% at 0.5 μg/ml and 5% at 9 μg/ml. The limit of quantification of the method was <0.06 μg/ml.
To correct for varying daily doses, the concentration-to-dose ratio (CDR) was calculated by dividing serum concentration by total daily dose. By accepting an α-error of 0.05 and a β-error of 0.10, the minimal number of participants necessary to detect a difference greater than the usual fluctuation (defined by us as >33% deviation from the mean LTG CDR value of the control group) was calculated to be seven in the test groups and 13 in the control group. The nonparametric Mann–Whitney U test was applied for comparison of the EE and the PG groups vs. control, respectively. The study was approved by the Regional Committee for Ethics in Medical Research. All patients gave written informed consent.
- Top of page
Forty-five women, age 17 to 44 years, were included in the study, 18 without hormonal contraception, 11 using combined contraceptives containing EE, and 16 using PG-only compounds. All comedications with LTG, including the administration mode of the contraceptive method, are shown in Table 1. No significant differences were found in mean age or body mass index (BMI) or mean daily LTG dose between groups, although the doses were somewhat lower in the EE group (225 ± 193 mg) than in the control group and the PG group (323 ± 182 mg and 291 ± 110 mg, respectively).
Table 1. Comedications with lamotrigine
|Group||No.||Contraceptive method||Other drugs (no.)|
|Administration||Compounds and doses|
|6||−||−||LEV (4); ZNS (1); GBP (1)|
|EE||3||Oral||EE, 35 μg; cyproterone acetate, 2 mg||LEV (1)|
|3||Oral||EE, 30/40/30 μg; levonorgestrel, 50/75/125 μg (triphasic)||−|
|2||Oral||EE, 30 μg; drospirenone, 3 mg||−|
|1||Oral||EE, 30 μg; desogestrel, 150 μg||−|
|1||Oral||EE, 30 μg; levonorgestrel, 150 μg||−|
|1||Vaginal ring||EE, 15 μg + etonogestrel, 120 μg/24 h||−|
|PG||3||Oral||Desogestrel, 75 μg||−|
|1||Oral||Noretisteron, 0.35 mg||−|
|7||Subdermal||Etonogestrel, 68 mg||LEV (1), ZNS (1)|
|1||Subdermal||Levonorgestrel, 36 mg||−|
|1||IM||Depot-medroxyprogesterone, 150 mg every 12 wk||−|
|3||Intrauterine||levonorgestrel, 20 μg/24 h||LEV (1)|
EE users had clearly lower serum concentrations at identical doses, compared with controls and PG users. The mean serum concentrations were 5.6 ± 3.1 mg/L (control group), 2.0 ± 1.3 mg/L (EE group), and 5.4 ± 2.1 mg/L (PG group). The difference between the EE group and controls was highly significant (p < 0.001).
The ranges, medians, and percentiles of the dose-corrected serum concentrations, expressed as serum concentration-to-dose ratio (CDR), are shown in Fig. 1. The difference between the EE group and the control group was highly significant (p = 0.003), whereas no statistical difference was found between the controls and the PG group. The patients in the latter group had been using PG for a median period of 8 weeks (range, 4 to 208 weeks) before blood sampling.
Figure 1. Box plot showing the median (horizontal line), interquartile range (lower and upper edge of the box), and total range of the serum concentration-to-dose ratio of the three study groups. EE, ethinyl estradiol; PG, progestogen. ***p = 0.003 (EE group compared with controls).
Download figure to PowerPoint
Figure 2 shows the change in CDR of patients taking part in more than one group. Switches occurred in either direction along the X-axis in individual patients, and the time interval between measurements was ≥4 weeks. LTG serum concentrations of patients who switched from EE to PG increased to control value levels. Patient A first switched from control to the EE group, and then from the EE to the PG group. Accordingly, her CDR decreased from 0.026 to 0.011, and then returned to 0.022. One patient of the control group (patient D) was later treated with a topical EE–containing vaginal ring. Her CDR decreased from 0.023 to 0.012, whereas her LTG dose remained unchanged.
Figure 2. Lamotrigine serum concentration-to-dose ratios in individual patients (A–H) who changed groups. Contraceptive methods were as follows. A: Ethinyl estradiol (EE) + levonorgestrel, depot-medroxyprogesterone IM. B, C: Etonogestrel subdermal implant. D: Etonogestrel vaginal ring. E, F: EE + cyproteronacetate. G: EE + levonorgestrel, etonogestrel subdermal implant. H: EE + cyproteronacetate.
Download figure to PowerPoint
- Top of page
This study suggests that PG-only contraceptive compounds do not reduce LTG serum concentrations. In accordance with previous findings (3,4), our results show that the use of combined oral contraceptives is associated with considerably reduced LTG serum levels. Interestingly, reduced LTG concentrations also were demonstrated in one patient using an EE-containing vaginal ring (see Fig. 2). Although the mean EE serum concentrations achieved with this device generally are somewhat lower than those with conventional oral contraceptives, the average minimal EE concentrations are quite similar (8).
Although no reports indicate significant fluctuations of the LTG serum concentrations throughout the physiologic menstrual cycle, it has recently been shown that LTG concentrations may increase considerably toward the end of the pill-free week in women taking combined oral contraception (9). In the previous studies by Sabers et al. (3,4), and in the present study, blood sampling was not standardized in relation to the menstrual cycle. Nevertheless, the effect of EE on the LTG serum concentrations could be clearly demonstrated. The more patients who were in the pill-free period, the more would our results underestimate the true effect of EE on the LTG serum concentration. With respect to PG, it should be noted that the PG compounds used in our study were administered on a continuous long-term basis without regular drug-free intervals.
LTG is metabolized mainly via glucuronidation by uridine-diphosphate glucuronosyltransferase (UGT) 1A4 at the N2 position and then eliminated via the kidneys (10). EE, which itself is glucuronidated by UGT1A1, is a known inducer of UGTs and may thereby increase the clearance of other glucuronidated drugs (11). Hence it is reasonable to assume that the reduction of LTG serum concentration also is due to increased glucuronidation, leading to accelerated renal elimination of LTG. This remains to be proven by studies of the LTG metabolite pattern. The observed reduction in LTG-CDR by >50% is of comparable magnitude to the effect of carbamazepine comedication (12) and may thus require dose adjustment in many patients.
PG-only compounds did not influence LTG serum concentrations, regardless of the administration mode, which either was oral, subdermal, intramuscular, or intrauterine in our study. Moreover, we could not detect any differences related to the various application forms, but the number of patients was limited. Marked differences are found in the pharmacokinetic and pharmacodynamic properties of the various PGs on the market. Only limited information is available on their dose–response relation in humans, and only progestational effects have been studied extensively. Nevertheless, all PG-only compounds in this study, except the intrauterine device used by three participants, were administered systemically, and all of them have demonstrated contraceptive efficacy at the doses used, indicating biologically active serum concentrations (13). Moreover, no differences in LTG-CDR were related to the administration mode of the PG compounds in our study. Removal of the three patients using an intrauterine drug-delivery system caused only marginal statistical changes. Hence all PG users were considered a uniform group in the final statistical analysis.
We also could also not see any association between the CDR and the duration of PG treatment, which ranged from 1 to 6 months (median, 8 weeks), a period that should have been sufficiently long to allow any potential effect on glucuronidating liver enzymes to develop. In contrast to our results, it was recently reported that the oral PG compound, desogestrel, increased LTG serum concentrations in seven of 10 patients (14). However, the effect varied considerably (0–96%), and similar effects of PG on other drugs have, to our knowledge, not previously been reported in the literature.
Recently, LTG was shown to reduce the area under the concentration–time curve (AUC) of the PG component of some combined oral contraceptives, as in levonorgestrel, by 19% (9). The clinical relevance of this interaction remains to be elucidated. Given the comparatively more brittle mechanism of PG-only pills exclusively targeting the cervical mucus and the endometrium, the risk for contraceptive failures should be considered in women who also use LTG. However, the novel 75-μg desogestrel-only pill differs from other PG-only pills in having a more robust mechanism, providing consistent ovulation inhibition with a performance appearing to be very similar to that of combined oral contraceptives (15).
In conclusion, our study adds further information on the interaction potential of LTG that should be considered when counselling women with epilepsy of childbearing age.