The rationale for pregnancy registers for women with epilepsy

The burden of epilepsy for those with the disorder is significant. For women of childbearing age, the uncertainty surrounding their ability to bear children who are free of the disorder, without birth defects, and cognitively and psychologically normal adds to this burden. Although factors other than medication exposure influence these questions, there is no doubt that antiepileptic drugs (AEDs) used to prevent seizures, such as valproate, have a significant, and possibly preventable, role in teratogenicity.1 Pregnancy registers are now showing promising results in elucidating this role and influencing changes in practice for the benefit of women with epilepsy and their children.

Some detailed information on the relative risks associated with AEDs has emerged over the past three decades,2 but it has largely been based on small-scale retrospective studies, with various and incomplete methods of recording data, no set protocols, and other shortcomings. It was clear that better information regarding teratogenicity, preferably from prospective studies, was needed. It was also clear that the expectant mother, as well as the infant, should be a primary consideration. The process of studying pregnancies in women with epilepsy should start well before conception and requires extensive consultation with the expectant mother about the planned management of her pregnancy and medication administration.3

In the late 1990s, these issues provided the rationale for setting up registers of pregnant women with epilepsy who were taking AEDs. There are now several international collaborative, independent and pharmaceutical company-initiated registers. The latter are concerned with single drugs and are not prospective. The collaborative and independent registers generally aim to collect prospective observational data from participating women according to an extensive protocol, with the data computerised for subsequent analysis. None of the registers dictate treatment and all have ethics approval, as well as informed consent from the participating women.4 The major registers are the International Registry of Antiepileptic Drugs and Pregnancy (EURAP),4 which includes data from 46 countries in Europe and elsewhere, and registers in North America, Denmark and the United Kingdom that publish reports independently.

The Australian Pregnancy Register of Antiepileptic Drugs for Women in Pregnancy with Epilepsy and Allied Conditions (the Australian Pregnancy Register), established in 1999, is affiliated with EURAP, but also publishes its findings independently. It collects data from women who have volunteered to participate through a series of five interviews held at various times during pregnancy and after the infant’s birth.5

Over the past decade, all these registers have contributed considerable knowledge, improved prescribing practices and, although they were initially intended to focus on teratogenicity, have been extended to examine maternal wellbeing and seizure control, and cognition of the offspring. In terms of teratogenicity, the data collected in the registers can be used to assess the contributing roles of heredity, social factors, substance misuse, alcohol consumption, social status, intake of other medications, and accurately defined type and activity of epilepsy. The registers represent prospective studies on treatment efficacy and compliance, seizure freedom before pregnancy, interactions between AEDs and hormones, the role of folate supplementation, and many other factors involved in producing normal pregnancies and outcomes.

The international registers have not used untreated control groups until recently — the North American register has used historical controls but is now enrolling a control group of pregnant untreated women, while EURAP compares the effects of different drugs. The Australian Pregnancy Register has from the outset collected data from a control group of untreated women with epilepsy, comprising about 10 per cent of the total, as well as (less successfully) women receiving AEDs for non-epileptic indications such as pain or bipolar disorder. Although there is no ideal control group, collecting data from untreated women with epilepsy provides an important comparison baseline.6

Recent analyses of data from the Australian Pregnancy Register have examined the role of AEDs in teratogenicity. For several decades, the use of AED polytherapy has been enshrined in the international literature as being harmful to the fetus, but our recent analysis of register data casts doubt on this, suggesting that it is the specific composition of polytherapy that is critical, not intake of multiple drugs per se.7 Most recently, analysis of register data has focused on dose issues that are associated with most of the AEDs, and examined the question of whether lower doses of drugs such as valproate may be effective in achieving seizure control without posing a higher risk of teratogenicity than other, less effective drugs.8 The role of AEDs in teratogenicity has become even more complicated as a series of new second-generation drugs have become available, because it takes a long time with many participants to define their role compared with the traditional drugs.9

Findings from the Australian Pregnancy Register have shown that seizure freedom before conception is demonstrably important in predicting the course of future pregnancies; the longer a woman is seizure-free, the better the outlook. The question of repeated pregnancies in women who have had a malformed baby while taking an AED, and advice to women contemplating extending their family, has also been studied.

Findings such as these are of immediate importance to women and their babies, contribute to medical knowledge and have demonstrably altered prescribing practices in Australia and internationally. Recent data indicate that, while prescribing of valproate has risen in the general Australian population, possibly as a result of increasing use for patients with psychiatric illness, especially bipolar disorder, there has been a fall in the number of prescriptions and doses of valproate for women of childbearing age.10 This change in prescribing, which was influenced by register data, has been associated with a fall in fetal malformation rates.11

Increasing the number of women enrolled in the Australian Pregnancy Register is highly desirable to continue study of these important and complex topics. Pregnancy is an important health issue, and we must all collaborate to make it safer for women and their children.