Recent studies have revealed that PWWE have a higher risk of obstetric complications, especially mild preeclampsia, than PWNE [5-6]. The findings of our study compared PWWE/PWNE and proved that women with epilepsy had higher chances of developing complications, such as PrH, preeclampsia, vaginal bleeding, oligohydramnios, polyhydramnios, miscarriage, and stillbirth. These results are consistent with an American retrospective cohort study from 2007 to 2011 conducted in 20% of community hospitals that observed a statistically significant association between PrH (adjusted OR= 1.30, 95% CI, 1.27– 1.33), preeclampsia (adjusted OR=1.59, 95% CI, 1.54–1.63), and stillbirth (adjusted OR=1.27, 95% CI, 1.17–1.38) [9]. A population-based cohort study in Sweden from 1997 through 2011 observed a statistically significant association between preeclampsia (adjusted OR=1.24, 95% CI, 1.07–1.43) and stillbirth (adjusted OR=1.55, 95% CI, 1.05–2.30) [19].
Following the same rationale, two Norwegian population-based cohort studies compared PWWE/PWNE; their findings were similar to those of our study in terms of mild preeclampsia (3.6% and 3.8%, respectively) [5-6].
Our study showed that most PWWE used ASM for pharmacological treatment of epilepsy during pregnancy. The Australian Antiepileptic Drugs in Pregnancy Registry (APR), conducted over the 20-year period from 1998 to 2018, showed similar results: out of the 2,148 pregnancies evaluated, 1,972 (91.8%) involved ASM use, while 176 (8.2%) women were not exposed to pharmacological treatment in the first trimester of pregnancy [20]. In contrast, the UKIEPR, whose prospective cohort study was conducted between December 1996 and August 2016, analyzed the outcomes of 9,247 pregnancies, of which 6,785 (73.4%) involved ASM use in monotherapy, 1,858 (20.1%) involved polytherapy regimens, and 604 (6.5%) did not involve ASM use [21].
The use of older ASM are not advocated during pregnancy; however, these ASMs might be indispensable for seizure control in women with severe epilepsy [15]. ASMs with lower rates of MCMs, such as LTG, levetiracetan (LEV), and oxcarbazepine (OXC), are often preferred for pregnant women with epilepsy [4]. However, in our study, the small number of PWWE using LTG, and no pregnant women using OXC and LEV, revealed that the high-risk pregnancy reference centers in Alagoas did not usually prescribe these drugs (S1 Table). It is important to emphasize that these drugs are not freely distributed by the SUS in Alagoas, a state characterized by a socio-demographic profile of vulnerability, which may have influenced the low rate of prescription of these drugs by the neurologists who attended these centers.
Carbazepine (CBZ), the second most prescribed drug in this study, might be associated with a higher risk of teratogenicity. CBZ has teratogenic capacity, but to a lesser extent than VPA and topiramate [10]. In the present study, a small number of PWWE used CBZ; however, no association between this drug and neonatal outcomes was found. (Table S3) These results corroborate the Swedish Cohort study from 1996 to 2013, which found no statistically significant association between CBZ use and LTG, preterm delivery <37 weeks, and low birth weight (OR=1.1; 95% CI=0.7–1.7, OR=1.3; 95% CI=0.7–2.6, respectively) [11].
The 20-year Australian Registry regarding ASM, as well as the Systematic Review and Meta-Analysis by Viale et al. (2015), found an association of spontaneous abortions in women with epilepsy who were prescribed ASM compared to those who were not prescribed ASM, which differs from our study, which found no statistically significant evidence for this association (OR=1.45; 95% CI=1.12–1.87 and OR 1.54; 95% CI=1.02–2.32, respectively) [20-22].
In contrast, the study by de Lima Leite et al. (2022), which used the same population cohort as our study, found a statistically significant difference in the rates of miscarriage between PWWE/PWNE; however, we found no association between this outcome and the use or non-use of ASM [23].
In terms of seizure types, GTCSs are associated with risks to the foetus and pregnant women. GTCSs are more worrisome and have been reported to cause several complications, including prenatal hypoxia and ischemia in areas of the brain, placental infarction, and intrauterine intracranial hemorrhage with fetal death [24]. In our study, we observed an association between the GTGS and stillbirth. Other epileptic seizures are probably less harmful, but may be associated with intrauterine growth retardation and premature delivery. In Battino's study, out of 33.4% of PWWE, 15.2% had GTCSs, 21 had status epilepticus, one perinatal death was documented, and no maternal deaths were noted [25]. Our study also did not register any maternal deaths, and in relation to status epilepticus, we found almost twice as many cases as in Battino's study, with 24% of PWWE experiencing GTCS and 52 experiencing status epilepticus. This result was higher than that found in EURAP, with only 36 and 21 cases of status epilepticus, respectively [6-26]. This suggests that previous studies have reported at confusing data which were all obtained from high-income countries; therefore they differ from data obtained from countries with more limited healthcare resources, such as those reported in our study, including on the use of older ASMs, that probably do not control seizures.
Finally, although some ASMs interfere with folic acid, data related to the effects of folic acid supplementation on pregnancy outcomes in women with epilepsy remain inconclusive [27]. Reports from prospective epilepsy pregnancy registries in the UK, Poland, and EURAP showed no association between folic acid use and a lower risk of MCM [6-28-29]. A Finnish population-based cohort study showed better results with folic acid supplementation in 43.6% of PWWE [7]. In the present study, folic acid and ferrous sulfate were prescribed for a small number of PWWE, which may be related to the risk of miscarriage. These findings are similar to those in a study by Asadi et al. (OR=2,6; 95% CI=1.2–5.6) [17]