Many epidemiologic studies have suggested that phenytoin (PHT) is a human teratogen, but the risk of damage and the specific harmful effects have not been determined. In addition, some investigators, but not all, have proposed that the mother with a history of epilepsy, but who is not taking anticonvulsants, has an increased risk of having a child with major malformations. Both exposures are common, each occurring in 0.4% of postpartum women interviewed at Brigham and Women's Hospital. A study is proposed in which three groups of infants will be examined: infants exposed to phenytoin with or without other anticonvulsants, infants whose mothers have had epilepsy in the past but did not take anticonvulsants during the pregnancy and unexposed infants whose mothers did not take anticonvulsants and have not had epilepsy. Each exposed infant and two unexposed infants matched for race and sex will be examined for the features of a Phenytoin Profile which includes evidence of growth retardation and/or microcephaly, cleft lip and/or cleft palate, heart defects, digit hypoplasia with symphalangism, mild craniofacial anomalies and fingernail hypoplasia. Unlike previous studies, the examiners will be unaware of the exposure status of each infant. In a pilot study of similar design of 59 PHT-exposed infants, 42% had a positive Phenytoin Profile in comparison to 9.9% of unexposed infants examined in the same way. The findings in the pilot study and the proposed study will be combined for data analysis. Power calculations show that the number of exposed infants that will be indentified at seven hospitals in Boston during a three year period will be sufficient to identify with 77 to 99% confidence a two to three-fold increase in the occurrence of the Phenytoin Profile among PHT-exposed infants and epilepsy-exposed infants in comparison to unexposed infants. Categorical analyses will be done to detect crude associations between the occurrence of the Phenytoin Profile and either maternal epilepsy treated with PHT or maternal epilepsy without. Step-wise logistic regression will be used to assess the exposure-outcome relationship while accounting for confounding variables, such as socioeconomic status, the drugs used, period of exposure, type of epilepsy and maternal and paternal factors.
| Harvey, Elizabeth A; Coull, Brent A; Holmes, Lewis B (2003) Anticonvulsant teratogenesis 5: observer bias in a cohort study. Birth Defects Res A Clin Mol Teratol 67:452-6 |
| Holmes, L B (2002) The teratogenicity of anticonvulsant drugs: a progress report. J Med Genet 39:245-7 |
| Holmes, L B; Harvey, E A; Coull, B A et al. (2001) The teratogenicity of anticonvulsant drugs. N Engl J Med 344:1132-8 |
| Carlin, J B; Ryan, L M; Harvey, E A et al. (2000) Anticonvulsant teratogenesis 4: inter-rater agreement in assessing minor physical features related to anticonvulsant therapy. Teratology 62:406-12 |
| Holmes, L B; Harvey, E A; Brown, K S et al. (1994) Anticonvulsant teratogenesis: I. A study design for newborn infants. Teratology 49:202-7 |
| Raymond, G V; Holmes, L B (1994) Head circumferences standards in neonates. J Child Neurol 9:63-6 |
