Women With Epilepsy and Control Women: Two Peas in a WEPOD

Commentary

Fertility and Birth Outcomes in Women With Epilepsy Seeking Pregnancy.

Pennell PB, French JA, Harden CL, Davis A, Bagiella E, Andreopoulos E, Lau C, Llewellyn N, Barnard S, Allien S. JAMA Neurol [published online ahead of print April 30, 2018]. doi:10.1001/jamaneurol.20180646.

IMPORTANCE: Prior studies report lower birth rates for women with epilepsy (WWE) but have been unable to differentiate between biological and social contributions. To our knowledge, we do not have data to inform WWE seeking pregnancy if their likelihood of achieving pregnancy is biologically reduced compared with their peers. OBJECTIVE: To determine if WWE without a prior diagnosis of infertility or related disorders are as likely to achieve pregnancy within 12 months as their peers without epilepsy. DESIGN, SETTING, AND PARTICIPANTS: The Women With Epilepsy: Pregnancy Outcomes and Deliveries study is an observational cohort study comparing fertility in WWE with fertility in control women (CW) without epilepsy. Participants were enrolled at 4 academic medical centers and observed up to 21 months from November 2010 to May 2015. Women seeking pregnancy aged 18 to 40 years were enrolled within 6 months of discontinuing contraception. Exclusion criteria included tobacco use and a prior diagnosis of infertility or disorders that lower fertility. Eighteen WWE and 47 CW declined the study, and 40 WWE and 170 CW did not meet study criteria. The Women With Epilepsy: Pregnancy Outcomes and Deliveries electronic diary app was used to capture data on medications, seizures, sexual activity, and menses. Data were analyzed from November 2015 to June 2017. MAIN OUTCOMES AND MEASURES: The primary outcome was proportion of women who achieved pregnancy within 12 months after enrollment. Secondary outcomes were time to pregnancy using a proportional hazard model, pregnancy outcomes, sexual activity, ovulatory rates, and analysis of epilepsy factors in WWE. All outcomes were planned prior to data collection except for time to pregnancy. RESULTS: Of the 197 women included in the study, 142 (72.1%) were white, and the mean (SD) age was 31.9 (3.5) years among the 89 WWE and 31.1 (4.2) among the 108 CW. Among 89 WWE, 54 (60.7%) achieved pregnancy vs 65 (60.2%) among 108 CW. Median time to pregnancy was no different between the groups after controlling for key covariates (WWE: median, 6.0 months; 95% CI, 3.8–10.1; CW: median, 9.0 months; 95% CI, 6.5–11.2; P = .30). Sexual activity and ovulatory rates were similar in WWE and CW. Forty-four of 54 pregnancies (81.5%) in WWE and 53 of 65 pregnancies (81.5%) in CW resulted in live births. No epilepsy factors were significant. CONCLUSIONS AND RELEVANCE: Women with epilepsy seeking pregnancy without prior known infertility or related disorders have similar likelihood of achieving pregnancy, time to pregnancy, and live birth rates compared with their peers without epilepsy.

Women with epilepsy (WWE) have specific challenges as hormonal fluctuations influence the seizure threshold, and antiseizure medications can reduce the efficacy of contraception and increase fetal malformations. In addition, breastfeeding and bone mineral health are important issues in this population, as are aspects of fertility and reproduction. As regards fertility, some studies have reported lower rates among WWE, while others have reported no differences between WWE and reference groups. A Kenyan study, for example, concluded that fertility rates among WWE were reduced by two-thirds compared with the general population of women in Kenya (1). The proposed causes of lower fertility rates among WWE are uncontrolled seizures, enzyme-inducing anticonvulsants, or simply lower rates of marriage or pursuit of pregnancy.

The conflicting results of population-based studies of birth rates in WWE may have resulted from variable sample sizes. A Finnish study of over 14,000 patients with newly diagnosed epilepsy, with a reference cohort of more than 29,000 cases, concluded that the birth rate was lower among WWE, with the hazard ratio of 0.88 (2). The study also found lower birth rates among men with epilepsy, with a hazard ratio of 0.58. Another population-based study from Norway analyzed data from 176 women and found age-specific birth rates. Before age 20, no differences in birth rates were seen between WWE and women without epilepsy; whereas, after age 25, birth rates were lower in WWE compared with the same age group of women without epilepsy (3). On the other hand, the Northern Finland Birth Cohort study of 222 individuals with epilepsy concluded that men and women with epilepsy were similar to the reference group in the number of children born, though active epilepsy during adulthood was associated with fewer births, a risk eliminated by seizure remission (4). However, these studies did not account for the desire to achieve pregnancy.

Another important limitation to population-based studies is the inability to account for the comorbidities of epilepsy. For example, a prospective longitudinal study from one University Hospital in Finland found that 60% of woman with active epilepsy had children compared with 77% in the general population (5). In that study, the single most important factor was severe comorbidity, which occurred at rates of 15 to 18 percent among individuals on monotherapy, and obviously at higher rates among those on polytherapy. Therefore, excluding severe comorbid factors reveals similar fertility rates between WWE and controls.

Pennell et al. enrolled WWE, from four academic medical centers, who were seeking pregnancy and who did not have a prior diagnosis of infertility. Enrollment occurred within 6 months of discontinuing contraception, and the outcomes were compared with a control group. The study was designed to assess differences in fertility rates between WWE and a control group when seeking pregnancy and enrolled 89 WWE and 108 controls. The primary aim was the proportion of successful pregnancies within a year of enrollment, while secondary aims included live-birth rates, time-to-pregnancy, as well as rates of sexual activity and ovulation. In addition, the authors explored the role of seizures and medications on fertility among WWE.

Inclusion criteria for both WWE and controls were as follows: age 18 to 40 years, desire to conceive, a steady male partner, and discontinuation of contraception in the previous 6 months. An electronic daily diary app and a cloud data system (Women With Epilepsy: Pregnancy Outcomes and Deliveries [WEPOD] study diary) was used. The app collected information about sexual activity, menstrual bleeding, medication intake, and seizures. Also, ovulation status was assessed by measuring serum progesterone between days 20 and 22 of the menstrual cycle for up to 2 cycles. Lost pregnancies, such as those resulting from miscarriage or ectopic pregnancy, were considered positive pregnancy outcomes.

The authors found no difference in pregnancy rates between WWE (60.7%) and controls (60.2%). Moreover, the median time to achieve pregnancy was not different: 6 months in WWE and 9 months in controls. Both groups achieved similar pregnancy outcomes with 81.5% of each group producing live births; likewise, miscarriage rates were similar, occurring in 14.8% of WWE and 18.5% of controls.

The majority of participants had more than 80% daily tracking of sexual activity, and among these, there was no difference in sexual activity between the two groups. The authors note that this result does not rule out decreased libido in WWE, since the goal of the recruited participants was to conceive, which makes sexual desire a secondary drive. In addition, the two groups were similar in their rate of ovulation: 88.5% among WWE and 81.3% among controls. Since other studies have reported lower pregnancy rates among women with uncontrolled seizures, the WEPOD study assessed that factor and found that active seizures in the preceding 9 months did not influence the chances of pregnancy. Among the WWE group, enzyme-inducing antiseizure medications were associated with lower rates of pregnancy (with a hazard ratio of 0.456); this finding was not statistically significant, possibly because only 18% of WWE were receiving enzyme-inducing anticonvulsants.

A strength of this prospective observational study is its design to eliminate social factors and, thus, its ability to largely rule out the role of biological factors in fertility among WWE. The enrolled women did not have any prior diagnosis of infertility or polycystic ovary syndrome and had planned their pregnancies in advance. Therefore, the addressed scenario is possibly the most common one seen in the epilepsy clinic and, thus, the results are a great asset in counseling a significant proportion of WWE about fertility. Also, the authors used a digital diary to collect information about adherence to medications, seizure occurrence, sexual activity, and menstrual bleeding. This provided a notable strength of this study because the participants were very compliant with the diary.

The small sample size of the subgroup of women receiving multiple antiseizure medications precluded the assessment of whether polytherapy increases the risk of miscarriage. The availability of many antiseizure medications that are not enzyme-inducing may have accounted for the results. Also, only 18% of the enrolled WWE were receiving enzyme-inducing anticonvulsants, making any conclusions about the risk of infertility among those women tentative. While the emergence of many antiseizure medications that do not influence liver enzymes has added a great benefit to epilepsy treatment, clinicians continue to use all available medications in the pharmacoresistant patient. Perhaps future studies can assess the roles of polypharmacy and enzyme-inducing drugs on fertility.