Pregnancy Registries in Epilepsy
Address correspondence and reprint requests to Dr. E. Beghi at Istituto “Mario Negri,” Via Eritrea, 62, 20157 Milano, Italy. E-mail: firstname.lastname@example.org
Summary: The risk of major malformations in the offspring of mothers with epilepsy receiving antiepileptic drugs is 4–8% compared to 2–4% in the general population. Risk factors include daily dose and polytherapy. Selected drugs have been found to be associated with a higher risk of specific malformations (congenital heart defects and cleft palate with phenytoin and barbiturates; neural tube defects with valproate and carbamazepine). Although some of these findings are unquestionable, several questions are still unsolved, depending the characteristics of the target populations, the small samples of patients, and the design and limiting factors of the published reports. In the last decade, pregnancy registries have been activated by collaborative groups of physicians in Europe (EURAP), North America (NAREP), Australia and India (the latter two recently merged into EURAP), to enroll a large number of exposed women to be monitored prospectively with standardized methods, and by three pharmaceutical companies marketing lamotrigine, gabapentin and vigabatrin, as part of their post-marketing surveillance. Even though the structure of these registries and the target populations should theoretically result in the identification of a sufficient number of women exposed to different drugs and examined for the occurrence of malformations of any type and severity, the implementation of a common database with information from the existing registries may provide valuable information in a shorter time period. Although differences between some of the registries limit the possibility to pool data, a gradual development of a collaboration is highly desirable to discuss a list of design issues and assess to what extent and how data could be compared and organized.
The teratogenicity of the antiepileptic drugs (AEDs) is a relevant issue. Based on the published reports, the risk of major malformations in the offspring of mothers with epilepsy receiving AEDs ranges from 4 to 8% compared with 2–4% in the general population (1–8). Risk factors include a higher daily dose and polytherapy. In addition, the use of selected drugs has been found in some studies to be associated with a higher risk of specific malformations [congenital heart defects and cleft palate with phenobarbital (PB), phenytoin (PHT), and primadone (PRM); neural tube defects with valproate (VPA) and carbamazepine (CBZ)](9–12). Although some of these findings are unquestionable, there are still several unsolved questions. First, the comparative role of each drug is still ill defined, and the data regarding new AEDs are still incomplete. Second, the teratogenicity attributable to monotherapy and to the use of low daily doses is as yet unsettled. Third, although the occurrence of major malformations is better defined, there is still disagreement on the incidence of minor anomalies. Fourth, the adverse effects of drugs on the long-term psychomotor development of the offspring are as yet unknown. Finally, it is still unclear which is the fraction of the risk of malformations and fetal death attributable to the treatment of epilepsy and which is the fraction attributable to the underlying disease and the occurrence of seizures during pregnancy. Tonic–clonic seizures and status epilepticus have been reported to be associated with fetal asphyxia and death (13,14). However, recent studies found no increase in the risk of embryopathy in infants of mothers with epilepsy untreated during pregnancy (8,15,16).
The characteristics of the target populations, the small sample size, and the different study designs are limiting factors in the published reports, which may explain these unsolved questions.
For these reasons, pregnancy registries have been promoted and activated in several countries by teams of researchers and drug companies. The main aim of these registries is to enroll a large number of exposed women who could be subjected to prospective monitoring and to a standardized recording of the data regarding AED treatment and the occurrence of malformations.
PROBLEMS RELATED TO THE ASSESSMENT OF CAUSE–EFFECT RELATION BETWEEN DRUG EXPOSURE AND MALFORMATIONS
The characteristics of an ideal pregnancy registry are depicted in Table 1. In brief, to assure validity of the results, a registry should be population based. An accurate denominator (i.e., a registry of all women with epilepsy in the population who may or may not be treated) is crucial to estimate the risk of malformations attributable to AEDs. The modalities of data collection should be predefined and in line with the requirements for a correct assessment of cause–effect relation between drug exposure and the occurrence of malformations. Follow-up after birth should be prolonged to detect the occurrence of delayed malformations in children seemingly normal at birth. In addition, a sizable sample of patients should be recruited to evaluate less-frequent exposures and different treatment modalities. Finally, the risk of malformations attributable to epilepsy should be more properly estimated comparing treated and untreated women with epilepsy and controls without epilepsy.
Table 1. General characteristics of an ideal pregnancy registry
|Calculation of the risk attributable to epilepsy|
|Prospective data collection|
|High-quality data recording|
|Known timing of exposure|
|Detailed treatment schedule and dosing|
|Prolonged follow-up after birth|
|Correct definition of pregnancy outcome|
|Standard definition of malformations|
|Identification of drop-outs|
However, some of these requirements are in conflict with the implementation of a large registry (which can be most easily based on referral patients), with the scarce availability of untreated women with epilepsy, and with the compliance of normal controls. In addition, with few exceptions (e.g., Iceland), automatic population registries are generally unavailable. Finally, women with epilepsy who are untreated during pregnancy are likely to be different from treated women in many aspects that may be of significance for the outcome of the pregnancy.
In the last decade, pregnancy registries have been activated by collaborative groups of physicians in Europe (EURAP), North America, Australia, and India, although recently the Australian and Indian registries have joined EURAP. In parallel, three pharmaceutical companies marketing lamotrigine (LTG), gabapentin (GBP), and vigabatrin (VGB) have started registration of pregnant women with epilepsy receiving those drugs as part of their postmarketing surveillance. Details on the characteristics of these registries along with some preliminary findings were given at the 23rd International Epilepsy Congress in Prague.
COMPARATIVE FEATURES OF THE EXISTING PREGNANCY REGISTRIES
An outline of the main methodologic aspects of a number of pregnancy registries is given in Table 2. To some extent, these fulfill most of the requirements of the ideal pregnancy registry (see Table 1). In most of the registries, the study population consists of women with epilepsy receiving old and new AEDs and enrolled through physicians, and may or may not be representative. The North American registry is the only one not allowing physicians to refer their patients. Women are recruited before the outcome of their pregnancy is known, and surveilled prospectively. The follow-up after birth is 12 months in most registries, and, except for registries run by pharmaceutical companies, all AEDs are being investigated. In EURAP and, to a lesser extent, in the Australian and Indian registries, a detailed list is provided of the main demographic and clinical risk factors documented during pregnancy and of outcome details, including the characteristics of the neonate. Previous obstetric history and pregnancy testing are detailed in all but the GBP registry. General practitioners also are involved in the U.K., EURAP, Australian, and Indian registry, and patients and/or lay associations are invited to report for the U.K., North American, Australian, and Indian registries. Reports of all malformations observed at birth and during follow-up are described as such. Only major malformations, identified by inclusion/exclusion criteria during the first 2 postpartum months, are tabulated by the North American registry, and broad categories of malformations are reported in the GBP registry.
Table 2. Comparative features of existing pregnancy registries
|Source of referral||Neuro|
|Epilepsy population||With all|
|With LTG||With GBP||With VGB||With or|
|<16 wk||Before known|
|No. of contacts||4–5||3||2||4||2||1||1||6|
|Previous obstetric hx||Detailed||Yes||No. pregnancies|
|Prenatal diagnoses||Detailed||Yes||Yes||With outcome||Yes||—||Moderately|
|Risk factors examined|| || || || || || || || |
| Social background||+||+||−||+||−||+||−||+|
| Fam hx birth|
| Other exposures||+||+||+||+||Open-ended|
| Epilepsy syndrome||+||+||−||+||−||−||+||+|
| Sz type/frequency||+||+||+||+||+||+||−||+|
| AED schedule||+||+||+||+||LTG doses|
| Other drugs||+||+||+||+||Timing of|
|Outcome details|| || || || || || || || |
| Termination with|
| Gestational age||+||+||+||+||+||+||+||+|
| Head circumference||+||+||−||+||+||−||−||+|
In the North American registry, data on pregnancy outcomes are collected by a research group separately. Findings indicating a significant relation between malformations and a specific drug are selected and submitted to the manufacturer and organized for publication.
COLLABORATION BETWEEN EXISTING REGISTRIES AND FUTURE PERSPECTIVES
Even though the structure of the existing registries and the size of the target population should theoretically result in the identification of a sufficient number of women exposed to different AEDs and carefully examined for the occurrence of malformations of any type and severity, the expected rate of recruitment is too slow to achieve this goal in a relatively short period. Thus the implementation of a common database including information obtained from the existing pregnancy registries is an initiative that can have substantial impact because of its high scientific value. The workshop at the 23rd International Epilepsy Congress in Prague therefore stimulated collaboration between registries with similar design. Consequently, the Australian and the European registries have now merged, adopting the EURAP protocol, and a collaboration also has been established between the European and U.K. registries. Although differences between some of the registries limit the possibility to pool data, a gradual development of a collaboration is highly desirable. The first step in this process would be the identification of a working group willing to discuss a list of design issues (inclusion/exclusion criteria, mechanisms of enrollment, definitions of malformations, duration of follow-up, review process, etc.) and to analyze in detail to what extent data can be compared and how this should be organized.