Drugs in Pregnancy and Lactation: Tenth Edition



PREGNANCY RECOMMENDATION: Limited Human Data—Animal Data Suggest High Risk

BREASTFEEDING RECOMMENDATION: Limited Human Data—Potential Toxicity


Zonisamide is teratogenic in three animal species and embryolethal in a fourth at doses or exposures very near or less than doses and systemic concentrations used or obtained clinically. Human pregnancy experience is limited. Two infants with congenital anomalies have been observed, but in both cases, other anticonvulsants known to be human teratogens also were used. If treatment with zonisamide in pregnancy is required, monotherapy using the lowest effective dose is preferred, but because of its status as adjunctive therapy, this may not be possible.


Zonisamide, a sulfonamide derivative, is indicated as adjunctive therapy in the treatment of partial seizures. Its anticonvulsant mechanism of action is unknown. The drug is bound extensively to erythrocytes with concentrations in red blood cells eight times higher than in plasma. Approximately 40%–60% is bound to plasma proteins, mainly albumin (1,2). After single oral doses, the elimination half-lives from erythrocytes and from plasma are 105 and 50–68 hours, respectively (1,2). The plasma elimination half-life is decreased to about 25–35 hours if the patient is receiving concurrent treatment with hepatic enzyme-inducing anticonvulsants (carbamazepine, phenytoin, phenobarbital, or primidone) (2).

Reproduction studies with zonisamide have been conducted during organogenesis in mice, rats, dogs, and cynomolgus monkeys. In pregnant mice, doses approximately 1.5–6.0 times the maximum recommended human dose of 400 mg/day based on BSA (MRHD-BSA) were associated with increased rates of fetal malformations (skeletal and/or craniofacial defects). In rats, increased frequencies of malformations (cardiovascular defects) and variations (persistent cords of thymic tissue and decreased skeletal ossification) were observed at all doses tested (0.5–5.0 times the MRHD-BSA). Perinatal deaths resulted when a dose 1.4 times the MRHD-BSA was given in the latter part of gestation up to weaning. The no-effect level for this toxicity was 0.7 times the MRHD-BSA (1).

In pregnant dogs, doses that produced peak maternal plasma levels (25 mcg/mL) about 0.5 times the highest plasma levels measured in humans receiving the maximum recommended human dose of 400 mg/day (MRHD-PL) resulted in an increased incidence of fetal cardiovascular defects (ventricular septal defects, cardiomegaly, and valvular and arterial anomalies). At doses that produced plasma levels (44 mcg/mL) approximately equal to the MRHD-PL, about 50% of the fetuses had cardiovascular malformations as well as increased incidences of skeletal malformations. Fetal growth restriction and increased rates of skeletal variations were observed with peak maternal dog plasma levels produced by all doses tested (0.25–1.0 times the MRHD-PL). When zonisamide was given to pregnant monkeys at doses approximately 0.1 times the MRHD-PL or higher, embryo–fetal deaths were observed. The cause of the deaths was unknown, but the possibility that they were due to malformations could not be excluded (1).

Zonisamide was not carcinogenic in mice and rats after 2 years of dietary administration. The agent was mutagenic in one test but not mutagenic or clastogenic in other tests (1).

No reports describing the placental crossing of zonisamide early in gestation have been located. At term delivery of one newborn, the cord blood and maternal blood concentrations were 14.4 and 15.7 mcg/mL (ratio 0.92), respectively (3). A comparable amount of drug was found in a second full-term newborn, but specific data were not given. Zonisamide probably also crosses the placenta early in gestation because of its low molecular weight (about 212). In addition, the long elimination half-life from red blood cells and plasma will result in prolonged concentrations of the drug at the maternal blood:placenta interface, thus increasing the opportunity for embryo/fetal exposure.

A 1996 report described the outcomes of 26 offspring whose mothers had been treated with zonisamide during pregnancy (22 women; 25 pregnancies; 1 set of twins). The offspring were exposed to zonisamide either alone (4 cases) or combined with other anticonvulsants (22 cases) (4). Doses ranged from 100 to 600 mg/day. There were 2 elective abortions and 24 liveborn infants (1 set of twins). Two offspring (7.7%), both exposed to combination anticonvulsants, had major malformations. Neither of the mothers experienced seizures during pregnancy. The first case, electively terminated at 16 weeks’ gestation, involved an anencephalic fetus (sex unknown) exposed to zonisamide (100 mg/day) and phenytoin (275 mg/day) throughout gestation. The second infant, a 2022-g female, was delivered by cesarean section at 37 weeks’ gestation. She had an atrial septal defect. Combination therapy throughout gestation included zonisamide (200 mg/day), phenytoin (200 mg/day), and valproic acid (400 mg/day) (4).

In a 2002 case report, two women took multiple anticonvulsants throughout gestation and delivered apparently normal infants at about 40 weeks’ gestation (3). One of the women took zonisamide (400 mg/day), carbamazepine (1000 mg/day) and clonazepam (1 mg/day), whereas the other took zonisamide (400 mg/day) and carbamazepine (800 mg/day). Both newborns had Apgar scores at 1 and 5 minutes of 8 and 9, respectively. Birth weights were 3094 g and 3164 g, respectively (3).

The Lamotrigine Pregnancy Registry, an ongoing project conducted by the manufacturer, was first published in January 1997. The final report was published in July 2010 (5). The Registry is now closed. There were 12 prospectively enrolled pregnancies exposed to zonisamide and lamotrigine, with or without other anticonvulsants in the 1st trimester, resulting in 12 live births without birth defects (5).

The effect of zonisamide on folic acid levels and metabolism is unknown. A 2003 review recommended that to reduce the risk of birth defects from anticonvulsants, women should start multivitamins with folic acid before conception (6). Although the recommendation did not specify the amount of folic acid, a recent study found that multivitamin supplements with folic acid (typically 0.4 mg) did not reduce the risk of congenital malformations from four first-generation anticonvulsants (see Carbamazepine, Phenytoin, Phenobarbital, or Primidone). Therefore, until further information is available, the best course is to start folic acid supplementation before conception. Although a specific dosage has not been recommended for patients receiving anticonvulsants, 4 mg/day appears to be reasonable.


Zonisamide is excreted into breast milk. A woman took 300 mg/day (100 mg three times daily) during pregnancy and continued the anticonvulsant while breastfeeding (7). The cord plasma drug concentration at delivery was 6.72 mcg/mL, 2.5 hours after a dose. Milk and maternal plasma levels were determined on postpartum days 3, 6, 14, and 30 (1.5–2.5 hours after a dose). Milk concentrations ranged from 8.25 (day 3) to 10.5 mcg/mL (day 30), whereas maternal plasma levels ranged from 9.52 (day 14) to 10.6 mcg/mL (day 3). The milk:plasma (M:P) ratio consistently increased at each sampling. The average milk:plasma ratio was 0.93 (range 0.81–1.03). No neonatal behavior problems were observed (7).

In another report, two mothers receiving zonisamide (400 mg/day) and other anticonvulsants (mother no. 1: carbamazepine and clonazepam; mother no. 2: carbamazepine) throughout gestation as well as postpartum breastfed their apparently normal infants (see above) (3). Zonisamide concentrations were determined in the breast milk of mother no. 1. From delivery to postpartum day 9, the unbound maternal plasma zonisamide concentration ranged from 10.7 to 13.3 mcg/mL, whereas the total zonisamide level ranged from 17.5 to 25.2 mcg/mL. During the same interval, the drug concentrations in the watery portion of the milk (i.e., whey) ranged from 8.9 to 10.9 mcg/mL. The breast milk transfer rate was 41%–57%. The plasma zonisamide level in the infant on day 24 was 3.9 mcg/mL (15% of the mother’s level on postpartum day 9). No adverse effects in either nursing infant were mentioned (3).

No adverse effects were noted in four nursing infants, but the milk concentrations were high enough that clinically significant doses appear to have been ingested. The relatively high plasma level found in one infant supports this assessment. Therefore, if a woman under treatment with zonisamide chooses to breastfeed, close clinical monitoring of her infant is recommended as well as the measurement of infant plasma drug levels. The long-term effects on nursing infants from this exposure are unknown but warrant investigation.


1.Product information. Zonegran. Elan Biopharmaceuticals, 2003.

2.Percucca E, Bialer M. The clinical pharmacokinetics of the newer antiepileptic drugs. Clin Pharmacokinet 1996;31:29–46.

3.Kawada K, Itoh S, Kusaka T, Isobe K, Ishii M. Pharmacokinetics of zonisamide in perinatal period. Brain Dev 2002;24:95–7.

4.Kondo T, Kaneko S, Amano Y, Egawa I. Preliminary report on teratogenic effects of zonisamide in the offspring of treated women with epilepsy. Epilepsia 1996;37:1242–4.

5.The Lamotrigine Pregnancy Registry. Final Report. 1 September 1992 through 31 March 2010. GlaxcoSmithKline, July 2010.

6.Yerby MS. Clinical care of pregnant women with epilepsy: neural tube defects and folic acid supplementation. Epilepsia 2003;44 (Suppl 3):33–40.

7.Shimoyama R, Ohkubo T, Sugawara K. Monitoring of zonisamide in human breast milk and maternal plasma by solid-phase extraction HPLC method. Biomed Chromatogr 1999;13:370–2.