Phenytoin Toxicity

Clinical Features

Diphenylhydantoin is poorly excreted by the kidney. Removal from the body depends on its hydroxylation. Kutt et al. (1964) found a family in which 3 members had reduced ability to hydroxylate diphenylhydantoin. The proband, who developed toxicity on usual doses of the drug, showed accumulation of the drug and much less hydroxylated derivative than normal in the urine. A defect in the hydroxylation of diphenylhydantoin can be produced by simultaneous administration of isoniazid (INH) which inhibits hydroxylation by liver microsomes (Kutt et al., 1968). Patients who show intolerance to diphenylhydantoin when receiving INH at the same time are patients who are the slow acetylators (243400) of INH (Kutt et al., 1970; Brennan et al., 1970). The family reported by Kutt et al. (1964) had a mother and 2 sons with inadequate hydroxylation. The proband was one of the sons, a 24-year-old male without liver disease, who consulted the authors 3 weeks after he had been given a daily dosage of 300 mg diphenylhydantoin and 90 mg phenobarbital for control of seizures after head injury. He showed marked nystagmus, ataxia and mental blunting, which disappeared when diphenylhydantoin was discontinued and reappeared when it was given again. Barbiturates alone produced no toxicity.

Vasko et al. (1977) reported a family with phenytoin hypometabolism. The proband was a 32-year-old epileptic who developed high blood levels and toxicity on a moderate dose. The 24-hour urinary output of 5-(p-hydroxyphenyl)-5-phenylhydantoin was only 50% of the ingested drug. The half-life of the drug was 32 hours. At least one child had a prolonged half-life.

Spielberg et al. (1981) studied individual susceptibility to toxicity from phenytoin metabolites by exposing human lymphocytes to metabolites generated by a murine hepatic microsomal system. Cells from 17 controls showed no toxicity at concentrations of phenytoin from 31 to 125 micromoles. Cells from 3 patients with phenytoin hepatotoxicity manifested dose-dependent toxicity from the metabolites. Phenytoin alone was not toxic to cells. The patients' dose-response curves resembled the response of control cells in which epoxide hydrolase, a detoxification enzyme for arene oxides, was inhibited. Detoxification of non-arene oxide metabolites (e.g., of acetaminophen) was normal in patients' cells. Cells from parents of 2 patients had intermediate responses. Cells from a sib of 1 patient showed no toxicity. A sib of another patient had a response similar to that of the patient.

Strickler et al. (1985) hypothesized a mutant form of microsomal epoxide hydrolase (EPHX1; 132810) as the molecular basis for abnormal reactions to phenytoin and some other drugs. Phenytoin (diphenylhydantoin, dilantin) is metabolized by cytochrome P-450 monooxygenases to several oxidized products, including parahydroxylated and dihydrodiol metabolites (see 124020). Arene oxides, which are reactive electrophilic compounds, are intermediates in these oxidative reactions. If not detoxified, arene oxide metabolites can covalently bind to cell macromolecules, resulting in cell death, mutation, tumors, birth defects, and, by acting as haptens, can lead to secondary immune phenomena. In animals, toxic effects of phenytoin, including gingival hyperplasia and teratogenicity, have been attributed to the arene oxide metabolites. Presumably the defect in hydroxylation of diphenylhydantoin is unrelated to the mephenytoin-metabolizing P450 system (124020) (Spielberg, 1988).

Gennis et al. (1991) described 3 sibs out of 12 who developed hypersensitivity reactions to phenytoin characterized by fever, rash, lymphadenopathy, and anicteric hepatitis. All recovered completely after discontinuation of treatment. One sib tolerated phenobarbital without toxic sequelae. Peripheral blood monocytes from the 3 patients and from 5 additional sibs who had never taken anticonvulsants were exposed to oxidative metabolites of phenytoin, phenobarbital, and carbamazepine. The cells from each of the 3 patients demonstrated increased toxicity from metabolites of phenytoin and carbamazepine, while the cellular response to metabolites of phenobarbital was within normal limits. Cells from 4 of the 5 other sibs showed an abnormal response to phenytoin metabolites, while cells from the fifth sib detoxified phenytoin metabolites normally.

Fetal Hydantoin Syndrome

Phelan et al. (1981) observed dizygotic twins in whom the evidence of diandric origin through superfecundation was strong (about 150 to 1). One suspected father was black, the other white. Throughout pregnancy the mother had taken phenobarbital and dilantin. Only 1 of the twins had signs of the fetal hydantoin syndrome (FHS).

Strickler et al. (1985) presented evidence suggesting a genetic predisposition to phenytoin-induced birth defects. Lymphocytes from 24 children exposed to phenytoin throughout gestation and from their families were challenged with phenytoin metabolites generated by a mouse hepatic microsomal drug-metabolizing system. Fourteen of the children had a positive assay result, i.e., a significant increase in cell death associated with phenytoin metabolites. Each of these 14 children had 1 parent whose cells were also positive. A positive in vitro challenge was highly correlated with major birth defects including congenital heart disease, cleft lip/palate, microcephaly, and major genitourinary, eye, and limb defects. There was no difference between children with positive and negative results in the number or distribution of minor birth defects and even features that have been thought to be pathognomonic of the fetal hydantoin syndrome, such as distal digital hypoplasia, were distributed evenly among children with positive and negative assays. Some have questioned whether the epilepsy rather than the drug used in its treatment is responsible for the clinical abnormalities observed in the children of epileptic women treated with hydantoin.

Chodirker et al. (1987) presented instructive observations of the hydantoin effect in a child born of a nonepileptic mother who had been given the drug during pregnancy for seizure prophylaxis after brain surgery.

Goldman et al. (1987) found that children with the fetal hydantoin syndrome had glucocorticoid receptor (138040) levels in circulating lymphocytes significantly higher than those of unaffected children with similar exposure to phenytoin. The receptor level of affected children was also significantly elevated above that of fathers of children with FHS and of fathers and mothers of control children.

Buehler et al. (1990) appeared to have demonstrated that low epoxide hydrolase activity in amniocytes is a risk factor for congenital malformations in the infants of mothers receiving phenytoin. In a random sample of amniocytes from 100 pregnant women, thin-layer chromatography showed an apparently trimodal distribution, suggesting that the level of the enzyme was controlled by a single gene with 2 allelic forms. In a prospective study of 19 pregnancies monitored by amniocentesis, an adverse outcome was predicted for 4 fetuses on the basis of low enzyme activity (less than 30% of the standard). In all 4 cases, the mother was receiving phenytoin monotherapy, and, after birth, the infants had clinical findings compatible with the fetal hydantoin syndrome. The 15 fetuses with enzyme activity above 30% of the standard were not considered to be at risk, and all 15 neonates lacked any characteristics of the fetal hydantoin syndrome.

Sabry and Farag (1996) suggested that hand anomaly in the fetal hydantoin syndrome can be unilateral acheiria at one extreme with nail/phalangeal hypoplasia at the other extreme. They reported the case of a baby born with absence of the right hand with rudimentary tags at the distal end of the right forearm. The infant was born of a nonepileptic mother who had a history of first trimester prophylactic anticonvulsant therapy after surgical excision of a meningioma. The status of the nails and phalanges in the left hand was not stated.

De Smet and Debeer (2002) described 2 children whose mother had been treated with phenylhydantoin for epilepsy that developed after surgery for a brain tumor. The first son had hypoplasia of the terminal phalanx of the fifth finger of the left hand. The second son was born with severe malformation of the right hand consistent with vascular disruption. He had facial dysmorphism with ocular hypertelorism, a small triangular shaped skull, and a depressed nasal bridge.

Inheritance

Dominant inheritance of phenytoin toxicity was proposed by Vesell (1979).

Vasko et al. (1980) observed phenytoin hypometabolism in 4 members of 4 generations of a kindred.

Vermeij et al. (1988) studied the inheritance of deficient phenytoin p-hydroxylation in the family of a patient who had previously suffered from phenytoin intoxication caused by insufficient metabolism of this drug (de Wolff et al., 1983). The rate of phenytoin metabolism was derived from the phenytoin/metabolite ratio in serum 6 hours after an oral test dose of 300 mg phenytoin. The propositus, a brother, and a sister were very slow metabolizers of phenytoin, with a metabolic ratio of approximately 20. All 22 children of these 3 individuals showed a mean metabolic ratio of 6.6 (SD = 1.7), whereas a control group of 37 individuals showed a mean metabolic ratio of 3.7 (SD = 1.8).

The fetal hydantoin syndrome has been observed in multiple sibs (e.g., Hanson et al., 1976).