|
|
| |||||||||||||||||||||||||||||||||
Background The frequency of major malformations, growth retardation, and hypoplasia of the midface and fingers, known as anticonvulsant embryopathy, is increased in infants exposed to anticonvulsant drugs in utero. However, whether the abnormalities are caused by the maternal epilepsy itself or by exposure to anticonvulsant drugs is not known.
Methods We screened 128,049 pregnant women at delivery to identify three groups of infants: those exposed to anticonvulsant drugs, those unexposed to anticonvulsant drugs but with a maternal history of seizures, and those unexposed to anticonvulsant drugs with no maternal history of seizures (control group). The infants were examined systematically for the presence of major malformations, signs of hypoplasia of the midface and fingers, microcephaly, and small body size.
Results The combined frequency of anticonvulsant embryopathy was higher in 223 infants exposed to one anticonvulsant drug than in 508 control infants (20.6 percent vs. 8.5 percent; odds ratio, 2.8; 95 percent confidence interval, 1.1 to 9.7). The frequency was also higher in 93 infants exposed to two or more anticonvulsant drugs than in the controls (28.0 percent vs. 8.5 percent; odds ratio, 4.2; 95 percent confidence interval, 1.1 to 5.1). The 98 infants whose mothers had a history of epilepsy but took no anticonvulsant drugs during the pregnancy did not have a higher frequency of those abnormalities than the control infants.
Conclusions A distinctive pattern of physical abnormalities in infants of mothers with epilepsy is associated with the use of anticonvulsant drugs during pregnancy, rather than with epilepsy itself.
Anticonvulsant drugs1 taken by pregnant women to prevent seizures are among the most common causes of potential harm to the fetus. In the 1970s and 1980s, the anticonvulsant drugs used most frequently to prevent seizures — phenobarbital, phenytoin, and carbamazepine — were found to cause major malformations, microcephaly, growth retardation, and distinctive minor abnormalities of the face and fingers in infants exposed to them during pregnancy.2,3,4,5,6,7,8
However, medical textbooks9,10,11 have suggested that these defects are caused by other factors, such as genetic abnormalities that cause the mother's epilepsy and are inherited by the fetus.12 To elucidate this issue, we conducted a cohort study of three groups of infants: those whose mothers took anticonvulsant drugs during the pregnancy, those whose mothers had epilepsy but took no anticonvulsant drugs during the pregnancy, and those whose mothers had no history of epilepsy and took no anticonvulsant drugs during the pregnancy (the control group).
Methods
Study Design
This study was conducted from 1986 to 1993 at five maternity hospitals in the Boston area: Brigham and Women's Hospital, Beth Israel Hospital, St. Margaret's Hospital, St. Elizabeth's Hospital, and Newton–Wellesley Hospital. Potential subjects were identified in the labor and delivery suites by nurses who asked the women if they had taken any medication for seizures during the pregnancy and if they had ever had a seizure.13 Women who answered yes to either question were then interviewed, with the approval of their obstetricians and nurses, to inform them about the study and to determine whether they qualified for inclusion. Women were excluded if they did not speak English, had a multiple-gestation pregnancy, or had another potentially teratogenic factor, such as type 1 diabetes mellitus. If the women qualified, they were asked to enroll and to give written informed consent. The study protocol was reviewed and approved annually by the institutional review board at each participating hospital. If a mother chose not to enroll, the results of the pediatrician's examination of her infant were reviewed to obtain birth weight, length, head circumference, and the presence or absence of any major malformations; these infants were not examined by a study physician.
The women enrolled in the study were asked to provide demographic data and to complete questionnaires administered by a research assistant to determine why they were taking anticonvulsant drugs (e.g., epilepsy or bipolar disorder); the dosage of each anticonvulsant drug; the characteristics of the seizures, their frequency during the pregnancy, and whether the women lost consciousness during seizures; and the family history with respect to epilepsy. With the written authorization of each woman, the results of all diagnostic tests (e.g., magnetic resonance images, electroencephalograms, and skull radiographs) and the dosages and serum concentrations of any anticonvulsant drugs were obtained. The responses to the questions and the results of the diagnostic evaluations of the women were reviewed by the study epileptologist to determine the type of epilepsy and its apparent cause with the use of an international classification of epilepsy.14
For each of the infants born to the enrolled women (i.e., either an infant exposed to anticonvulsant drugs or an infant not exposed to anticonvulsant drugs whose mother reported having had epilepsy), a control was recruited from the 10 infants born closest in time to him or her. Selecting randomly from this group of infants, we approached each mother until one was enrolled for each index infant. The same questionnaire was administered to the mothers of the control infants, with a separate consent form.
Examination of the Infants
The infants in all three groups were examined by a study physician; this physician was unaware of the exposure status of the infant during 93 percent of the examinations. The protocol for the physical examination listed 53 minor physical features to be recorded as present or absent. The dermal-ridge patterns (i.e., loops, whorls, and arches) on all fingers were recorded with magnification by otoscope, and 29 measurements were made, including the circumference and bitemporal width of the head, the inner canthal distance, and the length of the nose and upper lip (measured with a plastic ruler, a tape measure scored in millimeters, or sliding calipers; Seritex, Carlstadt, N.J.). The nose was measured from the lowest point in the depression of the bridge of the nose to the level of the alae nasi. The upper lip was measured from the base of the nasal septum to the upper edge of the vermilion border.
The outcomes of interest were major malformations, microcephaly, growth retardation, and hypoplasia of the midface and fingers. Major malformations were defined as structural abnormalities with surgical, medical, or cosmetic importance (identified during the first five days of life). Features that were not classified as major malformations are listed in Table 1.
|
Statistical Analysis
Individual logistic-regression analyses were conducted for each of the main outcomes. The correlation between outcome and each category of exposure (no maternal history of seizure or exposure to drugs, maternal history of seizures without exposure to drugs, exposure to one drug, exposure to more than one drug, and exposure to all drugs) was adjusted for maternal cigarette smoking (half a pack or more per day); ingestion of alcohol (14.8 ml [0.5 oz] on two or more occasions per week); self-reported use of cocaine or other illicit drugs; loss of consciousness during a seizure, presence of a febrile illness with a temperature above 39°C (102°F) for 48 hours, or presence of an important medical illness (e.g., multiple sclerosis); maternal or paternal head size or height more than 2 SD below the mean; or the presence of a major malformation in a first-degree relative.
We used two logistic-regression techniques to evaluate the interrelated outcomes: collapsed logistic-regression analyses relating the probability of at least one of the outcomes to the category of exposure, and generalized estimating equations assessing a global effect of anticonvulsant drugs on each set of outcomes after adjustment for confounders and for correlation among outcomes measured in the same subject.19
Results
By screening 128,049 women in labor and delivery suites, we identified 509 who had taken one or more anticonvulsant drugs during pregnancy, 386 of whom had taken one drug and 123 of whom had taken two or more drugs (30 women who had switched from one drug to another were included in the latter group) (Table 2). Among the 386 women who had taken only one anticonvulsant drug, 35 had taken the drug for medical conditions other than epilepsy. A total of 606 other women reported a history of seizures but had not taken an anticonvulsant drug during the pregnancy. We identified 1186 women who had not been exposed to anticonvulsant drugs during pregnancy and had no history of seizures.
|
|
We examined singleton infants born to 223 of the 386 women who had taken one anticonvulsant drug, 93 of the 123 women who had taken two or more anticonvulsant drugs, 98 of the 98 women with a history of seizures who had not taken anticonvulsant drugs, and 508 of the 1186 women in the control group (Table 4). Among the women who had taken one anticonvulsant drug, 87 had taken phenytoin, 64 phenobarbital, 58 carbamazepine, 6 valproic acid, 6 clonazepam, 1 diazepam, and 1 lorazepam.
|
Most of the major malformations identified were types of abnormalities that also occur in infants whose mothers have not taken an anticonvulsant drug (Table 5). However, two of the major malformations are known to be more common in infants exposed to anticonvulsant drugs: marked hypoplasia of the nails plus stiff joints, which is much more common in infants exposed to phenytoin with or without phenobarbital than in unexposed infants,21,22 and lumbosacral spina bifida, which is most common in infants exposed to either carbamazepine or valproic acid.23
|
Outcomes in Unenrolled Infants
Since not all of the mothers who were approached about enrolling in this study chose to participate, the medical records (i.e., the results of the pediatricians' examinations) of the infants who were eligible but not enrolled were reviewed to determine whether they had a major malformation or growth retardation. This analysis showed that there were no significant differences between infants who were enrolled and examined by the study investigators and infants who were eligible but were not enrolled and examined by study personnel. The 114 eligible, unexamined infants who were exposed to anticonvulsant drugs were somewhat less likely to have a major malformation than the 316 examined infants who were exposed to anticonvulsant drugs (1.8 percent vs. 5.7 percent, P=0.10) and slightly but not significantly more likely to have microcephaly (6.1 percent vs. 3.6 percent, P=0.30) or growth retardation (5.3 percent vs. 4.8 percent, P=0.90). Among the eligible control infants, there were no significant differences between the 508 infants who were enrolled and examined as part of the study and 536 infants who were not enrolled in the frequency of major malformations (1.8 percent vs. 1.7 percent, P=0.90), microcephaly (1.6 percent vs. 2.6 percent, P=0.30), or growth retardation (1.2 percent vs. 1.7 percent, P=5.00). Data were not analyzed for unenrolled infants born to mothers with a history of epilepsy who had not taken an anticonvulsant drug during pregnancy, because we could not confirm the maternal history of epilepsy.
Discussion
We found that infants exposed to a single anticonvulsant drug taken by the mother during pregnancy had a significantly higher frequency of associated abnormalities than control infants, and that infants whose mothers had a history of epilepsy but took no anticonvulsant drugs during pregnancy did not have an increased rate of these abnormalities. This study had several strengths: the examiner, who was almost always unaware of the infants' status regarding exposure to drugs, looked systematically for all of the features of the embryopathy associated with exposure to anticonvulsant drugs in the three groups of infants; the findings were objective12,24,25,26,27; and explicit criteria for the inclusion and exclusion of major malformations were used. Since teratogens cause distinctive patterns of abnormalities, the statistical analysis took into account the interrelation of several outcomes.
This study addressed the conflicting interpretations of two previous analyses of the same 104 infants exposed to anticonvulsant drugs.5,28 In those studies, epidemiologists concluded that the mother's epilepsy, not the anticonvulsant drug, was the teratogen.28 By contrast, the clinicians concluded that these infants had the physical features of embryopathy associated with exposure to anticonvulsant drugs.5 The authors of both reports recommended that future studies enroll a group of infants whose mothers had previously had epilepsy but who had taken no anticonvulsant drugs during pregnancy and that these infants be examined for the features of this embryopathy.5,28,29
The identification and recruitment of mothers with a history of epilepsy who had taken no anticonvulsant drugs during the pregnancy was another strength of our study. Such women were recruited in two other recent studies,27,30 which also found no increase in the risk of embryopathy in infants whose mothers had epilepsy but took no anticonvulsant drugs during pregnancy. Other studies12,28,31,32 have reached different conclusions, but there was a risk of misclassification, because the process of classifying the mother's reported epilepsy did not include a personal interview or a review of her medical records.
There is also concern that the mother's seizures themselves could have a harmful effect on the fetus, as suggested in case reports.33 This issue was addressed to a limited extent in our study. The frequency of major malformations was the same in the infants of women taking anticonvulsant drugs who had loss of consciousness during seizures in the first trimester of pregnancy and in infants of women taking anticonvulsant drugs who had other types of seizures. We also found that infants of women who had taken anticonvulsant drugs as treatment for mood disorders, migraine, or pain had an increase in the frequency of embryopathy that was similar to that among infants of women with epilepsy. A limitation of our study is that we did not include pregnancies that were terminated electively after fetal abnormalities associated with anticonvulsant drugs were diagnosed by prenatal screening. We identified additional cases in which the fetus was exposed to anticonvulsant drugs among pregnancies terminated electively at the largest participating hospital, through a separate surveillance program for malformations,34 but we did not include these pregnancies because we could not enroll a comparison group of unexposed fetuses among the other elective terminations.
In previous studies,7,26,27 infants exposed to carbamazepine were considered by clinical inspection to have an increased frequency of hypoplasia of the face and fingers characteristic of the embryopathy associated with exposure to anticonvulsant drugs. This feature was not found in the infants exposed to carbamazepine whom we examined. The difference in the findings can be addressed with more objective methods, such as cephalometric radiography35,36 and dermatoglyphy and radiography of the hands.22 This difference is important, because the hypoplasia of the midface35,36 associated with hypoplasia of the facial bones could be a marker for cognitive dysfunction.37
One would predict that some infants exposed to anticonvulsant drugs have a greater risk of harmful effects than others because of an underlying genetic susceptibility. Such an interrelation between genetic factors and environmental exposure has been suggested in studies of the teratogenicity of maternal cigarette smoking38 and alcohol use.39 In the case of anticonvulsant drugs, a deficiency of the detoxifying enzyme epoxide hydrolase40,41 and an increase in free radicals formed by the anticonvulsant drug42 are two theories of the reason for increased susceptibility. We predict that the correlations identified in this study will be much stronger in the more susceptible subgroup of children exposed to anticonvulsant drugs. Phenytoin, phenobarbital, and carbamazepine are folic acid antagonists, and one postulated mechanism for their teratogenicity has been the induction of folic acid deficiency.43 However, Hernández-Díaz and her associates44 recently reported that when pregnant women taking these anticonvulsant drugs also took a multivitamin supplement that included folic acid, it did not reduce the incidence of cardiovascular or urinary tract abnormalities or oral clefts in their infants.
We conclude that exposure in utero to anticonvulsant drugs is associated with a distinctive pattern of physical abnormalities in infants that are best identified by objective examination. The physical features of infants exposed to various anticonvulsant drugs are not the same. We found no evidence that infants born to women with a history of epilepsy who took no anticonvulsant drugs during pregnancy have an increased risk of the pattern of physical abnormalities associated with exposure to anticonvulsant drugs. The occurrence of such embryopathy was correlated with exposure to anticonvulsant drugs, regardless of the underlying maternal illness being treated.
Supported by a grant (NS 24125) from the National Institutes of Health.
We are indebted to the many families, nurses, and physicians whose cooperation was essential to the success of this project; to Lynn Rosenberg, Sc.D., Barbara R. Pober, M.D., M.P.H., and Martha Werler, Sc.D., for their suggestions about study design; to the research assistants — Jennifer Greene, Catherine Rooks, Meredith Miller, Nancy Goodman, Susan Tan Torres, M.D., and Joan Drury — for their diligence and diplomacy in carrying out this study; and to Ellice Lieberman, M.D., Dr.P.H., for her comments and suggestions on the manuscript.
Source Information
From the Genetics and Teratology Unit, Pediatric Service, Massachusetts General Hospital (L.B.H., E.A.H., K.B.H., A.M.H.); the Department of Biostatistics, Harvard School of Public Health (B.A.C., L.M.R.); and the Department of Neurology, Brigham and Women's Hospital (S.K.) — all in Boston.
Address reprint requests to Dr. Holmes at the Genetics and Teratology Unit, Warren 801, Massachusetts General Hospital, 55 Fruit St., Boston, MA 02114-2696, or at holmes.lewis{at}mgh.harvard.edu.
References
) TaqI polymorphism and oral clefts: indication of gene-environment interaction in a population-based sample of infants with birth defects. Am J Epidemiol 1995;141:629-636.
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
