Intrauterine Death in Pregnancy: Some Reassurance, yet Some Caution

Commentary

Antiepileptic Drugs and Intrauterine Death: A Prospective Observational Study From EURAP.

Tomson T, Battino D, Bonizzoni E, Craig JJ, Lindhout D, Perucca E, Sabers A, Thomas SV, Vajda F; EURAP Study Group. Neurology 2015;85:580–588.

OBJECTIVE: To compare the risk of spontaneous abortions and stillbirth associated with maternal use of different antiepileptic drugs (AEDs). METHODS: The EURAP registry is an observational international cohort study primarily designed to determine the risk of major congenital malformations (MCMs) after prenatal AED exposure. Using EURAP data, we prospectively monitored pregnancies exposed to the 6 most common AED monotherapies and to polytherapy. Intrauterine death (spontaneous abortion and stillbirth combined) was the primary endpoint. RESULTS: Of 7,055 pregnancies exposed to monotherapy with lamotrigine (n = 1,910), carbamazepine (n = 1,713), valproic acid (n = 1,171), levetiracetam (n = 324), oxcarbazepine (n = 262), or phenobarbital (n = 260), and to polytherapy (n = 1,415), 632 ended in intrauterine deaths (592 spontaneous abortions and 40 stillbirths). Rates of intrauterine death were similar across the different monotherapies (8.2%; 95% confidence interval [CI] 7.5%–8.9%), higher with polytherapy (12.1%; 95% CI 10.5%–13.9%), but showed no relationship with AED dose in monotherapy at conception. Multivariable analysis including 11 covariates in addition to the different AED exposures showed that the risk was greater with polytherapy vs monotherapy (risk ratio [RR] 1.38; 95% CI 1.14–1.66), parental history of MCMs (RR 1.92; 1.20–3.07), maternal age (RR 1.06; 1.04–1.07), and number of previous intrauterine deaths (RR 1.09; 1.00–1.19). The risk was greater with early enrollment and decreased with later gestational week at enrollment (RR 0.84; 0.82–0.86). CONCLUSIONS: The most important risk factors for intrauterine death in pregnancies of women with epilepsy include maternal exposure to AED polytherapy and the presence of MCMs in at least one of the parents.

The recent publication from the EURAP pregnancy registry documents the rate of intrauterine death in pregnancies of women with epilepsy who are taking antiepileptic drugs (AEDs). The reassuring aspect of their findings is that the overall rates were low, ever lower than in the general population. They considered pregnancies losses up to 24 weeks as the outcome of interest, which included spontaneous miscarriages and stillbirths. Fetal losses after 24 weeks of gestation were not included as outcomes. These two separate adverse pregnancy outcomes were differentiated on timing of their occurrence. Intrauterine deaths occurring up to 20 weeks gestation were designated as spontaneous miscarriages, and intrauterine deaths after this until to 24 weeks of gestation were considered stillbirths.

The ability of the study design to answer the query posed of the dataset deserves consideration. Women were enrolled in EURAP after pregnancy up to 16 weeks of gestation with a median gestational age at enrollment of 8 weeks. From a recent study of spontaneous miscarriage, most miscarriages occurred prior to 12 weeks gestation with a peak at 7 weeks (1). In this large dataset of 4,539 women, which included women enrolled prior to conception, the overall rate of spontaneous miscarriage was 13% (1). We can probably assume, therefore, that the EURAP study slightly underestimates the overall spontaneous miscarriage rates by missing early pregnancies and underestimates stillbirth rates by not including those that occurred later in pregnancy. The authors do acknowledge these limitations of their study and, consistent with the expected timing of intrauterine loss, report a higher risk associated with early enrollment in the study. They also report 15 perinatal deaths in their population, but the number of actual stillbirths that occurred outside of the analysis window was not reported. The interpretation of these findings and the message for patients should probably be in relative terms: AED monotherapy during pregnancy probably does not increase the risk of spontaneous miscarriage compared to the general population. For the three AEDs in which there were adequate numbers of subjects taking varied daily doses (lamotrigine, valproate, carbamazepine), there was no effect of dose on outcome. The rate of intrauterine deaths with these three monotherapies was around 8% for each, with 95% confidence intervals of 7 to 10%.

The finding of AED polytherapy increasing the risk by 40% over monotherapy, with an absolute rate of 12% and an upper limit of the 95% confidence interval at 14% is less reassuring. This too is probably an underestimate. Twenty percent of the all 7055 pregnancies assessed in this study were to mothers taking polytherapy; these were various combinations of the AEDs overall, with no predominant AED combination. Across the monotherapy and polytherapy exposure groups, there were 10 to 27 times more spontaneous miscarriages than stillbirths, indicating that the increased risk is primarily in early pregnancy.

The authors found that unmodifiable risk factors, such as a history of previous miscarriage and maternal age, were risk factors for intrauterine death, as reported in the general population (2). Notably, the occurrence of maternal generalized tonic clonic seizures during the first trimester did not impart increased risk. The greatest risk, was associated with a family history of major congenital malformations (MCMs), which doubled the risk. This was exclusive of concurrent maternal uterine deformities that would increase the risk of fetal loss by a direct structural mechanism. The risk factor of parental history of congenital malformations is not specifically reported in the general population, while consanguinity and known genetic and chromosomal defects are well-described as risk factors (3, 4). One study reported a possible association but analyzed the outcomes in the other direction: the mothers of offspring who had neural tube defects and cardiac malformations were more likely to have had previous miscarriages (5).

Folic acid use did not alter the risk, and the authors refuted a previous report from a smaller subset of their data, which suggested that folic acid use was important in preventing spontaneous miscarriage in women with epilepsy (6). It seems the only clear importance of folic acid use for risk reduction in this population is its association with improved cognitive outcome in the offspring (7).

The question is then raised: is this family history of MCM risk factor specific for the epilepsy population? The authors discuss a genetic vulnerability revealed by their work, but it is not all-or-none, of course. Or is it a risk factor anchored by the genetic vulnerability and unmasked by AED exposure? We cannot know the answer to this question from this data since there is no untreated group of women with epilepsy in this study.

One recent and fascinating study indicates that AEDs alter placental function and each one studied (phenytoin, valproic acid, carbamazepine, levetiracetam, lamotrigine) had differing effects on placental transport mechanisms (8). In this study, human trophoblast cells were exposed to AED concentrations at a clinically therapeutic range. Each AED altered placental expression of carriers for folate and thyroid hormones as well as carriers involved in the elimination of potential toxins from the fetus, including breast cancer resistance protein and multidrug resistance proteins. It is quite plausible, therefore, that alteration in placental transport is more deviant when two AEDs are in play vs one AED. These alterations could contribute to the increased risk of intrauterine death and MCMs with maternal polytherapy use.

In summary, the report is overall reassuring for most women with epilepsy who are pregnant or planning pregnancy since they often take AED monotherapy. For those taking polytherapy compared to monotherapy, or with a family history of MCMs vs none, the risk of intrauterine death is relatively higher, approximately 40% and 100% higher, respectively. However, the absolute rates are quite low, at 10 to 12 percent, within the accepted rate for the general population. Therefore, these risk factors should probably not alarm patients during office visit discussions. But it should be kept in mind that the tenet of aiming for monotherapy becomes more important in this clinical setting.

The unique risk of intrauterine death in the epilepsy sub-population with a family history of MCM is intriguing. Further understanding of AED effects on placental functioning may also be important when unraveling the risks to pregnancies in women with epilepsy.