Which contraception for women with epilepsy?

Seizure (2008) 17, 145—150

www.elsevier.com/locate/yseiz

Which contraception for women with epilepsy? Anneliese M. Schwenkhagen a,b,c,*, Stefan R.G. Stodieck c a

Gyna ¨kologicum Hamburg, Altonaerstr. 59, 20357 Hamburg, Germany University of Schleswig Holstein, Campus Lu ¨ beck, Department of Obstetrics and Gynaecology, Ratzeburger Allee 160, 23538 Lu ¨ beck, Germany c Epilepsy Centre Hamburg, Bodelschwinghstrasse 24, 22337 Hamburg, Germany b

KEYWORDS Contraception; Combined hormonal contraceptives; Epilepsy-enzyme inducing antiepileptic drugs; IUD—IUS

Summary Clinical decision making which contraceptive regimen is optimal for an individual woman with epilepsy is one of the most challenging tasks when taking care of women with epilepsy. The bidirectional interactive potential of antiepileptic drugs (AEDs) and hormonal contraceptives needs to be taken into account. Enzyme inducing (EI)-AEDs may reduce the contraceptive efficacy of hormonal contraceptives. If combined oral contraceptives (COCs) are used in combination with EI-AEDs, it is recommended to choose a COC containing a high progestin dose, well above the dose needed to inhibit ovulation, and to take the COC pill continuously (‘‘long cycle therapy’’). But even with the continuous intake of a COC containing a higher progestin dose contraceptive safety cannot be guaranteed, thus additional contraceptive protection may be recommended. Progestin-only pills (POPs) are likely to be ineffective, if used in combination with EI-AEDs. Subdermal progestogen implants are not recommended in patients on EI-AEDs, because of published high failure rates. Depot medroxyprogesterone-acetate (MPA) injections appear to be effective, however they may not be first choice due to serious side effects (delayed return to fertility, impaired bone health). The use of intrauterine devices is an alternative method of contraception in the majority of women, with the advantage of no relevant drug—drug interactions. The levonorgestrel intrauterine system (IUS) appears to be effective, even in women taking EI-AEDs. Likelihood of serious side effects is low in the IUS users. # 2007 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

Introduction Shortly after oral contraceptives became available, contraceptive failures due to pharmacokinetic * Corresponding author at: Gyna ¨kologicum Hamburg, Altonaerstr. 59, 20357 Hamburg, Germany. E-mail address: [email protected] (A.M. Schwenkhagen).

interaction between oral contraceptives and enzyme inducing antiepileptic drugs (EI-AEDs) were reported. Newer data show that it is not only the decrease in contraceptive safety that needs to be taken into account when taking care of women with epilepsy on antiepileptic drugs (AEDs) and hormonal contraceptives, but also the influence of the hormonal contraceptives on the efficacy of antiepileptic drugs. Various mechanisms may explain these

1059-1311/$ — see front matter # 2007 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.seizure.2007.11.013

146

A.M. Schwenkhagen, S.R.G. Stodieck

interactions, e.g. the induction or inhibition of drug metabolism by the hepatic Cytochrome P450 (CYP) isoenzyme system or the Uridine-5-diphosphate-glucuronosyltransferase (UGT)-system. However, many physicians and patients are unaware of this potential for interactions between enzyme-inducing AEDs and hormonal contraception, which may result in contraceptive failure or a decreased seizure control.

Effect of AEDs on hormonal contraceptives AEDs that may potentially impair contraceptive safety of hormonal contraceptives via an increased clearance of the synthetic steroids include the ‘‘strong CYP3A inducers’’ carbamazepine, phenytoin, phenobarbital and primidone as well as the ‘‘mild CYP3A inducers’’ oxcarbazepine, topiramate and felbamate (Table 1). Although earlier data did not show an effect of lamotrigine on contraceptive safety, a recent pharmacokinetic study in healthy women revealed a clinically relevant influence of 300 mg lamotrigine daily on a combined oral contraceptive (COC) with 30 mg ethinylestradiol and 150 mg levonorgestrel: the area under the curve (AUC) and the maximal plasma concentration (Cmax) of the levonorgestrel decreased whereas the ethinylestradiol pharmacokinetics were unchanged by lamotrigine. FSH and LH (luteinizing hormone) concentrations increased (by 4.7-fold and 3.4-fold) demonstrating a reduced suppression of the hypothalamic—pituitarian axis.1 Although measurement of serum progesterone showed no evidence of ovulation, contraceptive safety cannot be guaranteed, especially if higher lamotrigine doses and other COCs containing different progestins are used. Combined oral contraceptives have two components: an estrogen, usually ethinylestradiol, and a progestin. The metabolism of ethinylestradiol appears to be fairly well understood. However, Table 1 Effect of AEDs on hormonal contraceptives Contraceptive efficacy is potentially impaired

Contraceptive efficacy is likely not to be impaired

Carbamazepine Felbamate Lamotrigine Oxcarbazepine Phenobarbital Phenytoin Primidone Rufinamide Topiramate

Gabapentin Levetiracetam Pregabalin Tiagabin Valproate Vigabatrin Zonisamide

much less is known about the metabolism of the various progestins that are used in today’s COCs. Therefore, results of a drug—drug interaction study with a COC pill containing a specific progestin cannot be transferred to other hormonal contraceptives that contain a different progestin. This also applies to the impact of AEDs in higher daily doses as those used in published interaction studies.

Effect of hormonal contraceptives on AEDs Most drug—drug interaction studies have focused on the effect of AEDs on oral contraceptive safety. Much less is known about the result of a coprescription of hormonal contraceptives on AEDs, which is surprising, since it is known for a long time that oral contraceptives have a strong influence on drug metabolizing enzymes. Lamotrigine was the first AED that was evaluated in this respect: comedication with an ethinylestradiol containing hormonal contraceptive results in clinically relevant decrease of lamotrigine levels. In COCs, during the period ‘‘on the pill’’ lamotrigine levels decrease by approximately 50%, followed by an increase of lamotrigine levels in the contraceptive-free week up to 80— 100% of the baseline lamotrigine level. This is often clinically relevant and may result in an increased risk of seizure recurrence especially in week 2 and 3 on the pill or in concentration-dependent adverse effects at the end of the pill-free interval.2—5 These fluctuations are most likely due to an induction of UGT1A4, the enzyme responsible for the glucoronidation of lamotrigine, by ethinylestradiol. Valproate levels also seem to be reduced by the concomitant use of hormonal contraceptives.6,7 Just as with lamotrigine the magnitude of observed fluctuations of the valproate levels appear to vary interindividually.

A dogma revisited: is the prescription of COCs with 50 mg ethinylestradiol adequate to improve contraceptive safety? Unwanted pregnancies due to contraceptive failure of the chosen method should be avoided. In order to improve contraceptive efficacy it is often recommended that women on EI-AEDs should use COCs with at least 50 mg ethinylestradiol. However, this advice is problematic for two reasons: (1) there are no published data to prove the efficacy of this strategy, but reported pregnancies of women on

Which contraception for women with epilepsy? EI-AEDs with COCs containing 50 mg ethinylestradiol; (2) the recommendation is not plausible, since 50 mg ethinylestradiol is still far below the dose that is needed to inhibit ovulation (100 mg). What can we do instead? Two alternatives most likely to be superior to the old recommendation of using ‘‘high dose’’ COCs with 50 mg ethinylestradiol are:  the use of a COC containing a progestin dose well above the dose needed to inhibit ovulation and  the continuous use of the COC without a pill-free interval (‘‘long cycle’’). To understand why this strategy is probably very effective one needs to know how modern oral contraceptives work. The main contraceptive mechanism of COCs is ovulation inhibition. COCs suppress the hypothalamic—pituitarian—ovarian axes resulting in an impaired follicular growth and anovulation. Ovulation inhibition is accomplished by disturbing the gonadotropin secretion at the hypothalamic and pituitarian level with an inhibition of the preovulatory LH (luteinizing hormone) peak. Additionally, synthetic progestins may directly influence ovarian function by a direct inhibition of the ovarian steroid biosyntheses. The reduced ovarian and follicular activity is reflected in low estradiol and absent progesterone serum concentrations. Modern COCs have two components: ethinylestradiol and a progestin. Both are on their own able to inhibit ovulation. However, in modern COCs ovulation inhibition is mainly achieved by the progestin and not by ethinylestradiol. The typical daily progestin dose in today’s COCs is about 1.5—2 times the ovulation-inhibiting dose (Table 2). Modern commonly prescribed oral COCs contain only between 20 and 30 mg ethinylestradiol, which is far below the dose that would be needed to inhibit ovulation (about 100 mg).9 In these COCs the estrogen is mainly responsible for cycle control, the regular shedding Table 2 Ovulation-inhibiting doses (without additional estrogen)8 Progestin

mg/day

Chlormadinone acetate Cyproterone acetate Desogestrel/3-keto-desogestrel Dienogest Drospirenone Gestodene Levonorgestrel Norethisterone Norethisterone acetate Nomegestrol acetate

1.7 1.0 0.06 1.0 2.0 0.04 0.06 0.4 0.5 5.0

147 of the endometrium. It also reinforces the ovulation inhibiting effect of the progestin, adding some extra contraceptive safety. Thus, the use of a contraceptive containing 50 mg of ethinylestradiol does not necessarily ensure contraceptive safety, if used in combination with EI-AEDs that induce the metabolism of the progestin as well. Nevertheless it may improve cycle control. To further improve contraceptive efficacy (additional to the use of a COC that contains a progestin well above the dose needed to inhibit ovulation), it is also recommended to use the COC without a pillfree interval. If the COC is used the traditional way (3 weeks on the COC, followed by 1 week off) FSH and LH secretion recommence instantly after the last day of pill intake, which explains the rapid return to fertility after the discontinuation of a COC. This immediate reactivation of the hypothalamic—pituitarian—ovarian axes leads to the prompt return of a follicular growth. During each 7-day pillfree interval the development of new follicles starts over again. This can be demonstrated by sequential ultrasound measurements of the increasing follicular activity and the assessment of the rising estradiol serum concentrations. With the 1st day back on the pill, the follicular growth is disturbed. However it always takes a couple of days of uninterrupted use of the COC in adequate doses to ensure a definite arrest of follicular development and anovulation. This explains why missed pills in the 1st week of the COC use (or progestin levels to too low to properly inhibit this process, as it may happen due to a comedication with EI-AEDs) may in particular promote ‘‘escape ovulations’’. If the COC is continuously used without a pill-free interval gonadotropin secretion and ovarian function is continuously suppressed, which results in an enormous increase in contraceptive efficacy.10 Additional to these effects of the ‘‘long cycle’’ use of COCs, contraceptive safety is further enhanced by other contraceptive mechanisms, e.g. the uninterrupted thickening of the cervical mucus (also referred to as ‘‘minipill effect’’) or the continuous alteration of the endometrium. Even if EI-AEDs happen to reduce the COCinduced inhibition of ovarian function and even if escape ovulation occurs, these mechanisms still may prevent pregnancy. Although there is no doubt, that the contraceptive efficacy is improved by the continuous use of a COC that contains a progestin in a high dose (well above the dose that would be needed to inhibit ovulation), full oral contraceptive safety cannot be guaranteed in women with epilepsy taking strong EI-AEDs. Thus additional contraceptive protection (such as barrier methods as condoms) may be worthwhile.

148

How to deal with the decrease of lamotrigine levels if used in combination with ethinylestradiol containing hormonal contraceptives? Ethinylestradiol containing hormonal contraceptives (COCs, vaginal ring (Nuvaring1) or patch (Evra1) may lead to clinically relevant lamotrigine level fluctuations accompanied by a worsening of seizure control. Simple lamotrigine dose adjustment in order to improve seizure control may lead to an overdosing and dose-dependent side effects in the contraceptive-free interval. The easiest way to stabilize lamotrigine levels and subsequently seizure control and to avoid this overdosing is the continuous use of the hormonal contraceptive without a free interval. Since the ethinylestradiolinduced decrease of lamotrigine levels may occur very rapidly only a few days after the hormonal contraceptive was started, it appears to be reasonable to start lamotrigine dose adjustment shortly after the initiation of the hormonal contraception. In the majority of cases approximately a doubling of the initial lamotrigine dose is needed to stabilize lamotrigine levels.

Are progestin-only methods suitable for women with epilepsy on EI-AEDs? In progestin-only methods the mechanism of contraceptive action is dose-dependent. The low dose progestin-only preparations, the progestin-only pills (POPs), that are also referred to as minipills, contain progestins below ovulation-inhibiting dose, e.g. 30 mg of levonorgestrel/day. Thus, ovulation is not consistently inhibited. POPs exert their contraceptive effect mainly through the peripheral actions of progestins: a thickening of the cervical mucus that leads to impaired sperm penetration, an adverse effect on the endometrium and a decrease in tubal motility. Since the main mechanism of action is peripheral, POPs need to be taken continuously without a pill-free interval. Because of the very low doses used, POPs are likely to be ineffective in women on EI-AEDs. The intermediate dose progestin-only methods allow some follicular development, but inhibit ovulation in almost all cycles. Typical examples are Cerazette1 (a pill that contains 75 mg desogestrel p.o./day), or the two progestin implants Implanon1 or Jadelle1. Implanon1 is a single rod contraceptive implant that is inserted underneath the skin of the upper arm. It contains 68 mg etonogestrel (active 3keto-metabolite of desogestrel) which is released

A.M. Schwenkhagen, S.R.G. Stodieck over 3 years. Jadelle1 consists of two rods containing 75 mg levonorgestrel each, which is released over a period of up to 5 years. Even though the implants provide highly reliable contraception in the general female population not on EI-AEDs, there are published cases of contraceptive failures with these methods.11—14 Thus, these intermediate dose progestin-only methods are not recommended in women with epilepsy on EI-AEDs. The high dose injectable progestin-only formulations inhibit ovulation. The two most widely used preparations are depot medroxyprogestroneacetate (DMPA) and norethisterone enanthate (NET-ET). These very high dose progestin-only depot formulations are considered to provide effective contraception in most publications, although clinical trials to support this are lacking. Some authors recommend shortening the interval between each injection (e.g. from 12 to 10 weeks) in women on EIAEDs. Although it is plausible that these high dose progestin-only depot preparations provide contraceptive efficacy even if the patient is treated with EI-AEDs they may not be first choice because of a number of serious side effects: over the last couple of years an increasing body of evidence was published showing that particularly the long-term use of DMPA may adversely affect bone mineral density, thus possibly promoting the development of osteoporosis.15—18 Peak bone mass may not be achieved in adolescent girls. Other worrisome side effects include a delayed return to fertility19 for many months, in some cases even more than 1 year. Women on DMPA may also complain about weight gain, skin problems like acne or hair loss, and depressive symptoms. Thus the decision whether to use depot preparations in women on EI-AEDs should take into account all the risks and benefits of this contraceptive choice for the individual patient.

Are IUDs or the levonorgestrel-IUS an alternative for women with epilepsy? The classic intrauterine copper or silver device (IUD) or the newer levonorgestrel releasing intrauterine system (IUS) (Mirena1) as it is called by the manufacturing company may be an alternative for women with epilepsy an EI-AEDs. The IUS has steroid reservoir that releases 20 mg of levonorgestrel in the uterine cavity. Over time the release rate decreases slowly to 15 mg/day. The device is approved for 5 years. Most of the contraceptive effect is mediated via local mechanisms: the high tissue concentration of levonorgestrel in the endometrium results in a suppression of endometrial growth, which in turn

Which contraception for women with epilepsy? leads to a substantial reduction in duration and amount of menstrual bleeding. Many women experience amenorrhoea, which should not be judged as an unwanted side effect but an external sign that system is in proper position. Other contraceptive mechanisms include cervical mucus hostility which gives additional protection against the progression of sexual transmitted infections to pelvic inflammatory disease (PID) with possible later fertility problems.20 Since the IUS exerts its contraceptive properties mainly by local uterine mechanisms it is on theorethical grounds very unlikely that EI-AEDs impair its contraceptive efficacy. This is supported by an observational series on women using the IUS Mirena1 concurrently with antiepileptic and other enzyme-inducing drugs. The authors of the study concluded, that if there is any increased pregnancy risk, it falls within ‘‘acceptable bounds’’.21 In contrast to usual COCs the levonorgestrel-IUS does not seem to influence lamotrigine plasma levels (own observation in six patiens). Thus the IUS (and to a lesser extent the IUDs) are a good alternative to the classical hormonal contraceptives in women with epilepsy, especially if they are taking enzyme inducers or lamotrigine. Yet many health care providers are reluctant to recommend IUDs or the IUS due to myths and misconceptions, e.g. the belief that IUDs cause PID or infertility and should consequently not be used in younger women who have not given birth. However, studies indicate that there is only a small increased short-term risk of PID in IUD users during the 1st weeks after insertion. After this time, PID risk appears comparable to that of women not using an IUD. Furthermore studies show that the biggest risk factor for PID and resulting tubal sterility in young women using IUDs is their and their partner’s sexual behaviour (i.e. promiscuity). Newer clinical trials suggest that there is little or no good evidence that the prior use of a copper-containing IUD increases the risk for impaired fertility.22—24 Keeping in mind the IUSs pronounced effect on the cervical mucus, which is likely to prevent the ascend and progression of vaginal and cervical infections to endometritis and PID, health care suppliers should no longer withhold the IUS because of their fear to promote infertility.

Emergency contraception High doses of progestin after intercourse can prevent pregnancy in a majority of women without contraception. There are no data on what doses are needed in women on EI-AEDs, but current guidelines recommend a higher dose (1.5 mg levonorges-

149 trel instead of 0.75 mg as soon as possible followed by 0.75 mg 12 h later).

Conclusion A neurologist should never forget to ask a women, if and what type of contraception she is using before prescribing an AED. The decision, which AED to use, or which contraceptive method is optimal for an individual woman with epilepsy depends on a variety of factors. The bidirectional interactive potential of AEDs and hormonal contraceptives needs to be taken into account when counselling women with epilepsy. Alternatives such as the use of an IUD or IUS should be kept in mind if standard hormonal contraceptives to not appear to offer the contraceptive safety, that is wanted in an individual patient or if potential side effects of the contraceptive method do not warrant their use.

References 1. Sidhu J, Job S, Singh S, Philipson R. The pharmacokinetic and pharmacodynamic consequences of the co-administration of lamotrigine and a combined oral contraceptive in healthy female subjects. Br J Clin Pharmacol February 2006;61(2): 191—9. 2. Sabers A, Buchholt JM, Uldall P, Hansen EL. Lamotrigine plasma levels reduced by oral contraceptives. Epilepsy Res November 2001;47(1—2):151—4. 3. Sabers A, Ohman I, Christensen J, Tomson T. Oral contraceptives reduce lamotrigine plasma levels. Neurology August 26, 2003;61(4):570—1. 4. Stodieck SRG, Schwenkhagen A. Lamotrigine plasma levels and combined monophasic oral contraceptives (COC) or a contraceptive vaginla ring. A prospective study in 30 women. Epilepsia 2004;45:187. 5. Christensen J, Petrenaite V, Atterman J, Sidenius P, Ohman I, Tomson T, et al. Oral contraceptives induce lamotrigine metabolism: evidence from a double-blind, placebo-controlled trial. Epilepsia March 2007;48(3):484—9. 6. Herzog AG, Farina EL, Blum AS. Serum valproate levels with oral contraceptive use. Epilepsia June 2005;46(6):970—1. 7. Galimberti CA, Mazzucchelli I, Arbasino C, Canevini MP, Fattore C, Perucca E. Increased apparent oral clearance of valproic acid during intake of combined contraceptive steroids in women with epilepsy. Epilepsia September 2006;47 (9):1569—72. 8. Kuhl H. Pharmacology of estrogens and progestogens: influence of different routes of administration. Climacteric August 2005;8(Suppl. 1):3—63. 9. Taubert H-D, Kuhl H. Beeinflussung reproduktiver Funktionen durch synthetische Hormone. I. Kontrazeption mit Hor¨r die Praxis. 2nd ed. Stuttgart, New monen: ein Leitfaden fu York: Thieme; 1995. pp. 66—119. 10. Spona J, Elstein M, Feichtinger W, Sullivan H, Ludicke F, Muller U, et al. Shorter pill-free interval in combined oral contraceptives decreases follicular development. Contraception August 1996;54(2):71—7.

150 11. Schindlbeck C, Janni W, Friese K. Failure of Implanon contraception in a patient taking carbamazepin for epilepsia. Arch Gynecol Obstet January 2006;273(4):255—6. 12. Shane-McWhorter L, Cerveny JD, MacFarlane LL, Osborn C. Enhanced metabolism of levonorgestrel during phenobarbital treatment and resultant pregnancy. Pharmacotherapy November—December 1998;18(6):1360—4. 13. Haukkamaa M. Contraception by Norplant subdermal capsules is not reliable in epileptic patients on anticonvulsant treatment. Contraception June 1986;33(6):559—65. 14. Odlind V, Olsson SE. Enhanced metabolism of levonorgestrel during phenytoin treatment in a woman with Norplant implants. Contraception March 1986;33(3):257—61. 15. Berenson AB, Breitkopf CR, Grady JJ, Rickert VI, Thomas A. Effects of hormonal contraception on bone mineral density after 24 months of use. Obstet Gynecol May 2004;103(5 Pt 1):899—906. 16. Cromer BA, Stager M, Bonny A, Lazebnik R, Rome E, Ziegler J, et al. Depot medroxyprogesterone acetate, oral contraceptives and bone mineral density in a cohort of adolescent girls. J Adolesc Health December 2004;35(6):434—41. 17. Scholes D, Lacroix AZ, Ott SM, Ichikawa LE, Barlow WE. Bone mineral density in women using depot medroxyprogesterone acetate for contraception. Obstet Gynecol February 1999;93 (2):233—8.

A.M. Schwenkhagen, S.R.G. Stodieck 18. Paiva LC, Pinto-Neto AM, Faundes A. Bone density among long-term users of medroxyprogesterone acetate as a contraceptive. Contraception December 1998;58(6):351—5. 19. Hassan MA, Killick SR. Is previous use of hormonal contraception associated with a detrimental effect on subsequent fecundity? Hum Reprod February 2004;19(2):344—51. 20. Toivonen J, Luukkainen T, Allonen H. Protective effect of intrauterine release of levonorgestrel on pelvic infection: three years’ comparative experience of levonorgestrel- and copper-releasing intrauterine devices. Obstet Gynecol February 1991;77(2):261—4. 21. Bounds W, Guillebaud J. Observational series on women using the contraceptive Mirena concurrently with anti-epileptic and other enzyme-inducing drugs. J Fam Plann Reprod Health Care April 2002;28(2):78—80. 22. Hubacher D, Lara-Ricalde R, Taylor DJ, Guerra-Infante F, Guzman-Rodriguez R. Use of copper intrauterine devices and the risk of tubal infertility among nulligravid women. N Engl J Med August 23, 2001;345(8):561—7. 23. Grimes DA. Intrauterine device and upper-genital-tract infection. Lancet September 2000;356(9234):1013—9. September 16. 24. Hov GG, Skjeldestad FE, Hilstad T. Use of IUD and subsequent fertility-follow-up after participation in a randomized clinical trial. Contraception February 2007;75(2):88—92.