- Robert J. Boyle, MD*
After completing this article, readers should be able to:
Describe the short- and long-term effects of moderate-to-severe alcohol use in pregnancy.
List the risks and long-term effects of smoking relative to pulmonary disease, sudden infant death syndrome, and infant intelligence.
Characterize the range of anomalies associated with antiepileptic drug use during pregnancy.
Describe the risk of Ebstein anomaly following lithium exposure.
Describe the effect of angiotensin-converting enzyme inhibitors in the newborn period.
Many pregnant women use multiple prescription, over-the-counter, and recreational drugs during their pregnancies. Questions frequently arise from the mother, the mother’s clinician, and the physician caring for the newborn about the potential risks of these agents for the fetus and newborn. Few of these agents have been evaluated prospectively in pregnant women prior to licensure and marketing. Our knowledge of potential risks to the fetus and newborn, therefore, is based on case reports, epidemiologic surveys of exposed individuals, and animal studies.
Multiple factors may influence whether a specific drug may affect an individual fetus or newborn and the magnitude of that effect (Table 1). Clearly, not every infant exposed to an individual drug is affected. Likewise, it is important to note that many of these agents are critical to the mother’s health, and discontinuing the drug may place the mother and fetus at higher risk than continuing the drug. The physician also must acknowledge that patients often take medications that were not prescribed specifically for them. These may include over-the-counter medications that the patient does not recall taking, does not understand the active ingredients, or even does not consider a “medication” or “drug.” Or they may be taking medications that they are reluctant to disclose or that were prescribed to another individual. Reluctance to disclose exposures may be especially true of drugs that carry negative social stigmas in some circumstances, such as alcohol and tobacco.
This article addresses the potential clinical effects, relative risk, diagnostic suggestions, and management of the fetus or infant exposed to a select group of commonly used drugs. Clearly, this is not a complete list of common agents. Likewise, because the primary focus is prenatal exposure, this article does not discuss postnatal exposure through breastfeeding.
Fetal Alcohol Syndrome and Alcohol-related Neurodevelopmental Disorders
Alcohol and tobacco are the most common chemicals transferred to the fetus transplacentally. It has been estimated that approximately 10% of women of childbearing age are problem drinkers (up to six to seven drinks per day). Women who are acutely intoxicated in the perinatal period may deliver an infant who smells of alcohol, has a blood alcohol level in the cord blood indicating legally intoxication, and has signs and symptoms of intoxication, including central nervous system (CNS) depression and hypoglycemia. Some older studies suggest that infants born to mothers who chronically abuse alcohol also may present with withdrawal symptoms, including tremors, wakefulness and restlessness, excessive crying, hypertonia, excessive startling, and opisthotonos. These symptoms may persist for 72 hours.
The more common problem related to alcohol exposure during pregnancy relates to the agent’s teratogenic effects. Alcohol is probably the most common teratogen ingested during pregnancy and may be one of the most commonly identifiable causes of mental retardation. Fetal alcohol syndrome (FAS) defines a constellation of physical, behavioral, and cognitive abnormalities (Table 2). FAS may be difficult to diagnose, and most experts agree the diagnosis requires three criteria: 1) prenatal or postnatal growth restriction, 2) CNS involvement, and 3) specific craniofacial dysmorphic features. Although many children lack the full complement of FAS criteria, they present with growth deficiency, behavioral problems, and motor and speech delays in conjunction with a clinical history of alcohol exposure in utero. The term “fetal alcohol effects” has been used to describe these conditions. Recently defined terminology clarifies conditions in which there is a history of substantial regular or heavy episodic maternal alcohol exposure and clinical findings associated with alcohol exposure. Alcohol-related neurodevelopmental disorder (ARND) describes infants who have CNS, cognitive, or behavioral abnormalities without the classic facial features or other birth defects. Alcohol-related birth defects (ARBD) describes infants who have birth defects without the classic facial features.
The etiology of FAS/ARND/ARBD is poorly defined and may be multifactorial, but clearly it relates to the effect of alcohol or its metabolites on placental function, causing poor nutrient and oxygen transfer, or on multiple biochemical pathways in the developing fetus. This may include but is not limited to protein synthesis, cellular energy metabolism, cell signaling, cell migration, increased apoptosis, and neurotransmitter production.
FAS is estimated to occur at a rate of 5 to 10 per 10,000 live births in the United States, but it may be higher in certain populations (eg, 30 per 10,000 among Native Americans). The true incidences of FAS and ARBD are probably greater because it is difficult to recognize the signs and symptoms and to obtain a reliable history of alcohol ingestion. Some estimate that 1 in 300 children may be affected. Exposure early in gestation is more likely to cause the classic FAS clinical features with anomalies; exposure limited to later gestation more likely is associated with the developmental and behavioral aspects. Fetal alcohol exposure may account for up to 5% of all congenital anomalies. The overall incidence of FAS among infants born to mothers who have a history of drinking alcohol during pregnancy is only approximately 2% to 3%. However, the risk increases to 85% among women who have prenatal risk factors, including high percentage rates of drinking days, high parity, African-American or Native American race, and positive screening on questionnaires of alcohol use.
The potential for harm to the fetus is stronger with larger amounts of alcohol than with smaller and with binge exposure of excessive amounts compared with chronic smaller amounts. Features compatible with FAS were seen in approximately 10% of infants who were exposed daily to 1 to 2 oz of absolute alcohol. (Most studies define a “drink” as 12 oz of beer, 5 oz of wine, or 1.5 oz of liquor, approximating 0.5 oz of absolute alcohol.) The risk increased to 40% among those exposed to more than 5 oz per day. The risks of ARND are more difficult to define because of the wide spectrum of effects. However, heavy alcohol exposure is associated with decreased intelligence quotient (IQ) and behavioral disturbances. Current studies do not support identifying a “safe level” of consumption by pregnant women.
FAS, ARND, and ARBD have tremendous financial, emotional, and social impacts on the individual, family, and society. The medical, surgical, behavioral, custodial, and judicial services are estimated to cost $75 million to $9.7 billion annually in the United States.
The physician needs to evaluate neonates, infants, and young children carefully and repetitively for evidence of FAS, ARND, or ARBD when there is a history of significant alcohol exposure. Evidence of significant exposure may include frequent episodes of intoxication, development of tolerance or withdrawal, social or legal problems related to drinking, or alcohol-related health problems such as hepatic disease. The clinical features of FAS, ARND, or ARBD often are not obvious in the immediate newborn period, and many of the behavioral and developmental aspects may not be apparent until a later age. Children in whom the diagnosis is suspected should be evaluated by a pediatrician who is knowledgeable and competent in the evaluation of neurodevelopmental and psychosocial problems. In addition, appropriate referral to early intervention services and education agencies is important.
Although the prevalence of smoking in the United States has decreased, recent evidence suggests that up to 20% of pregnant women continue to smoke throughout gestation. There is no documented increased risk of birth defects, but smoking has been associated with adverse pregnancy outcomes both for the mother and the infant (Table 3). The major impact on the neonate is decreased birthweight, although some investigators have reported abnormal neurobehavior in infants of smoking mothers, including deficits in mental development, language comprehension, and expression; hyperactivity; attention deficit disorder; and lower IQ. The long-term effects remain unclear.
Epidemiologic studies identify prenatal (and postnatal) smoking as a risk factor for sudden infant death syndrome (SIDS), possibly due to arousal abnormalities. (See accompanying article in this issue.)
The pathophysiologic effect of smoking is related to maternal and fetal exposure to nicotine and carbon monoxide. Nicotine stimulates catecholamine release and vasoconstriction in the uterine circulation, with resultant decreased uteroplacental blood flow and potentially decreased delivery of oxygen and nutrients to the fetus. Nicotine levels are 10% to 15% higher in the fetus than in the mother. These physiologic effects also have been noted when smokeless tobacco products were studied. Carbon monoxide binds to hemoglobin, producing carboxyhemoglobin, which inhibits oxygen delivery to tissue. Fetal hemoglobin has a higher affinity for carbon monoxide, further reducing oxygen delivery. Placental pathology shows decidual necrosis, microinfarcts, fibrinoid changes, and hypovascular and atrophic villi secondary to vasoconstriction and ischemia. The overall effects of tobacco are related to the number of cigarettes smoked per day, with minimal effects noted when the mother smokes fewer than 10 cigarettes daily.
Smoking usually continues postnatally. The infant then is exposed to secondhand smoke, which has been associated with an increased risk of respiratory disease and SIDS. Human milk from smoking mothers also contains significant concentrations of nicotine.
The perinatal period is an ideal time to discuss the risks of smoking with the infant’s mother and encourage smoking cessation. Multiple behavioral and pharmacologic programs can assist the mother (and father) in this often difficult task.
Seizure disorders affect 0.4% to 0.8% of the population, and many of the drugs used for treatment are established teratogens. The risk of birth defects in offspring of mothers who have seizure disorders that are treated with anticonvulsants is two to three times that of the general population (ie, 5% to 9%). However, the factors causing this increase are not well-defined. Epilepsy itself without drug therapy, genetic predisposition, severity of seizure disorder, type of anticonvulsant therapy, drug levels, and multidrug therapy all have been associated with an increased risk of malformation. Although multiple dysmorphic syndromes have been reported with specific agents (Table 4), there is significant overlap in the anomalies that may, in fact, represent a general fetal antiepileptic drug syndrome.
Several possible mechanisms of teratogenesis with these agents have been suggested: 1) metabolism of the drugs, leading to formation of toxic unstable epoxides, which may be exacerbated in some individuals by a genetic defect in fetal epoxide hydrolase activity; 2) production of toxic free radicals during drug metabolism; 3) increased hepatic metabolism or decreased absorption of folate, resulting in relative folate deficiency; and 4) inhibition of folinic acid production (valproate). Low serum folate levels have been correlated with increased malformation rates in pregnant women who have epilepsy. Both valproate and carbamazepine increase the risk of neural tube defects (1% to 2%), most of which are lumbar meningomyeloceles. It has not been established that periconceptional folate supplementation reduces the risk of neural tube defects in this population, but the current recommendation is for all women of childbearing age to take a folate supplement (0.4 mg/d). Some experts recommend up to 4 mg/d.
Phenytoin, phenobarbital, primidone, carbamazepine, and ethosuximide have been shown to decrease placental transfer of vitamin K, leading to an increased risk of hemorrhagic disease of the newborn. These agents also may interfere with vitamin D metabolism, resulting in secondary hypocalcemia in the infant and mother.
Very few data exist regarding safety during pregnancy and teratogenic risk for the newer agents, such as lamotrigine, gabapentin, and felbamate.
Seizure control during pregnancy is important. Maternal seizures during pregnancy are associated with an increased risk of miscarriage, preterm labor, intracranial hemorrhage in the fetus, stillbirth, and possible developmental or learning difficulties. Possible etiologies include trauma and hypoxia/ischemia due to decreased placental perfusion or maternal hypoxia. For these reasons, altering or discontinuing anticonvulsant regimens during the pregnancy is not recommended and clearly should be done only after considering all of the potential risks and benefits to the mother and infant.
After birth, the physician should assess the infant carefully for evidence of anomalies and anticonvulsant-associated dysmorphic features. Routine vitamin K administration shortly after birth is especially important for infants who have been exposed to drugs that may cause decreased vitamin-K dependent coagulation factors. Neonates exposed to phenobarbital or primidone should be monitored for evidence of withdrawal. Because infants may not become symptomatic until age 5 to 7 days, physicians should inform parents of the potential signs and symptoms.
Most mothers continue with anticonvulsant therapy postpartum. None of these agents is contraindicated during breastfeeding. However, there are reports of sedation in the infant exposed to phenobarbital and primidone.
Isotretinoin (13-cis-retinoic acid) was introduced in the 1980s for treatment of severe cystic acne. Within the first year of its introduction, multiple reports defined a pattern of malformations (Table 5) that occurred in infants born to mothers who took isotretinoin beyond the 15th day following conception. Approximately 35% of infants exposed to this drug will have malformations. Peroxyl free radicals probably cause cytotoxic changes in the neural crest tissues of the head. The drug had been shown to cause similar malformations in animals. Explicit warnings about drug use during pregnancy did not prevent the continuing exposure of significant numbers of infants due to failure of birth control methods or failure to use birth control. Currently, the Retinoid Pregnancy Prevention Program provides explicit and detailed warnings and requires signed informed consent before initiating therapy. These concerns do not apply to topical products.
Although many exposed infants demonstrate malformations at birth, absence of malformations does not predict a normal intellectual outcome. Therefore, if there has been an exposure, ongoing care of the child should include careful serial neurodevelopmental assessment.
Lithium is an effective therapy for bipolar disorder and mania. Early studies suggested an increased risk of Ebstein anomaly of the tricuspid valve, a rare condition associated with downward displacement of the valve into the right ventricular cavity that produces cyanosis due to valvular incompetence and supraventricular arrhythmias in the newborn period. Initial studies suggested a risk as high as 8% among exposed infants. However, more recent epidemiologic data suggest a significantly lower risk, with no cases noted in more than 100 exposed pregnancies. There also are case reports of fetal goiter, cyanosis, hypotonia, atrial flutter, T-wave inversion on electrocardiography, diabetes insipidus, shock, seizures, and an increased risk of preterm labor associated with lithium. The actual incidence of these symptoms is not known.
Due to the severity of and difficulty in treating bipolar disorder, most experts believe that the risks of stopping the medication outweigh the potential risks of the cardiac disease. They suggest fetal echocardiography if the infant has been exposed to lithium in the first trimester. If the infant is asymptomatic at birth, little is gained by obtaining echocardiography or other diagnostic tests for Ebstein anomaly.
Most mothers taking lithium during the gestation will continue treatment in the postpartum period. Cyanosis, hypotonia, and electrocardiographic changes have been reported in breastfeeding infants whose mothers were taking lithium. Lithium is concentrated in the milk, and an infant may have serum levels 10% to 15% of the mother’s level.
Angiotensin-converting Enzyme (ACE) Inhibitors
ACE inhibitors (captopril, enalapril, lisinopril, fosinopril, quinapril) are very effective antihypertensive agents that have few side effects and frequently are prescribed to women of reproductive age. The enzyme converts angiotensin I to the potent vasoconstrictor angiotensin II. The angiotensin system has been shown to be important for normal fetal renal development and function.
No association has been documented between ACE inhibitor use during the first trimester and congenital defects. However, there are multiple reports of both malformations and physiologic disturbances in infants exposed to the drug during the second and third trimesters (Table 6). Experts postulate that decreased fetal blood pressure and poor control of renal perfusion lead to fetal oliguria/anuria, renal tubular dysgenesis, and oligohydramnios. Low fetal blood pressure, poor peripheral perfusion of the superficial tissues, and oligohydramnios cause poor perfusion of the skull that, in conjunction with increased forces applied by the uterine wall, cause poor ossification of the calvaria. Affected newborns may present with severe hypotension and anuria. They also may have evidence of chronic fetal compression, with deformations and pulmonary hypoplasia. Interestingly, there is no evidence of intrauterine growth restriction secondary to these drugs.
Initial management should include assessment of arterial blood pressure, urine output, and respiratory status. Blood pressure may require fluid or pharmacologic support. Careful fluid management and monitoring of blood urea nitrogen, creatinine, and serum electrolytes are important for infants who have anuria/oliguria. In most cases, the renal output will recover, but some infants have required dialysis. Renal ultrasonography with Doppler studies may help define both the etiology and renal blood flow status. Skull roentgenograms may help to evaluate cranial calcification.
Anticoagulants are used during pregnancy relatively infrequently. However, the clinical conditions for which they are used pose significant risks to the mother, including venous thrombosis and pulmonary embolism or cardiac valve prosthesis with risk of valve thrombosis and cerebral emboli.
Coumadin is used commonly to both prevent and treat thromboembolic disease. It works primarily by depressing vitamin K-dependent clotting factors (II, VII, IX, and X). It crosses the placenta readily and causes both early teratogenic effects and later fetal effects (Table 7). The warfarin syndrome results from fetal exposure to the drug from the 6th to the 12th week of gestation. The incidence of the syndrome when exposure occurs during this time approaches 25%. After this period, there is some debate about whether there continues to be a real but significantly lower (<5%) risk of other anomalies and effects that probably are due to hemorrhage into fetal tissue with resultant scarring. Finally, there is an increased risk of hemorrhage, especially CNS hemorrhage, when the infant is delivered while the mother is still receiving coumadin.
On the other hand, heparin does not cross the placenta and does not cause teratogenic or other fetal or neonatal effects. This also is true for the newer, low-molecular weight heparins (enoxaparin). Therefore, heparin is usually the drug of choice for anticoagulation during pregnancy, especially during the first trimester and intrapartum.
The infant who has been exposed in the first trimester should be assessed carefully for evidence of embryopathy. If the exposure occurred later in gestation, careful assessment of the CNS, including cranial ultrasonography, is important. If the mother was taking coumadin within 48 to 72 hours of delivery, the infant should be observed carefully for evidence of bleeding.
With a few exceptions (isotretinoin), physicians should expect that the majority of infants born to mothers using these agents will be normal. However, clinicians should keep the potential risks resulting from the exposure in mind during assessments in the newborn period and throughout childhood.
- Briggs GG, Freeman RK, Jaffe S. Drugs in Pregnancy and Lactation Baltimore, Md: Williams and Wilkins; 1998
- Committee on Drugs, American Academy of Pediatrics. Use of psychoactive medication during pregnancy and possible effects on the fetus and newborn. Pediatrics.2000;105:880–887.
- Committee on Substance Abuse and Committee on Children with Disabilities, American Academy of Pediatrics. Fetal alcohol syndrome and alcohol-related neurodevelopmental disorders. Pediatrics.2000;106:358–361.
- Stratton K, Howe C, Battaglia F, eds. Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention and Treatment. Washington, DC: National Academy Press;1996:4–21. Available online from www.iom.edu
- Zahn CA, Morrell MJ, Collins SD, et al. Management issues for women with epilepsy: a review of the literature. Neurology.1998;51:949–956.
- Copyright © 2002 by the American Academy of Pediatrics