From Pages to Practice
Maria is a 27-year-old woman who you see in the primary care clinic. She was first diagnosed with bipolar disorder 3 years ago and is now taking 600 mg of lithium daily. Her symptoms are well-controlled and she has not experienced mood symptoms in the last few months. During the consultation, she happily tells you about her recent marriage and her desire to become pregnant. However, she is worried about the potential effect lithium could have on her unborn child. She has read about the potential risk for heart defects and seeks your advice on how she should continue treatment for her bipolar disorder during pregnancy.
Lithium is a classic mood stabilizer that has been employed for medical purposes since the 19th century. It was approved by the FDA in 1970 and is widely used for the treatment of bipolar disorder and other psychiatric conditions, such as schizophrenia. Lithium requires close monitoring of serum levels to avoid toxicity. Previous cohort studies have suggested an increased risk of congenital heart malformations among newborns exposed to lithium during pregnancy, specifically Ebstein’s anomaly, a right ventricular outflow obstruction (RVOTO) defect caused by the displacement of tricuspid valve leaflets. The first report of this association originated from registry data in lithium-exposed pregnant women and lacked a control group.
In this week’s issue of NEJM, Patorno and colleagues compared the risk of cardiac malformations among infants exposed and unexposed to lithium during the first trimester in a large cohort of Medicaid-enrolled women who delivered a live-born infant between 2000 and 2010. The primary outcome was the presence of cardiac malformation in the infant. Pregnant women exposed to lamotrigine, an anticonvulsant drug not considered to be teratogenic, were also evaluated. The unexposed reference group included women who had not been dispensed lithium or lamotrigine during the 3 months before the start of the pregnancy or during the first trimester
Among more than one million pregnancies, 663 (0.05%) were in women who were exposed to lithium during the first trimester. The prevalence of cardiac malformations was 2.41 per 100 infants exposed to lithium, 1.15 per 100 unexposed infants, and 1.39 per 100 infants exposed to lamotrigine. This represented a 65% increase in the risk of cardiac malformations in infants exposed to lithium during the first trimester, compared with those who were not (adjusted relative risk, 1.65; 95% CI, 1.02–2.68). Exposure to lamotrigine was not associated with an increased risk of cardiac malformations, compared with no exposure (RR, 0.89; 95% CI, 0.61–1.30).
Among secondary outcomes, lithium was associated with greater risk for RVOTO defects than other heart defects. Furthermore, results suggested that the effects of lithium might be dose-dependent, such that exposure to higher doses is associated with increased risk of cardiac malformations; the relative risk associated with daily doses of lithium >900 mg was 3.22 (95% CI, 1.47– 7.02). Although these observational findings are consistent with previous studies that showed an association between lithium exposure in pregnancy and cardiac malformations in infants, they imply that the magnitude of the risk is lower than previously suggested.
To answer Maria’s question, it is important to balance the modest increase in the risk of cardiac malformation with the clinical benefit of lithium therapy. Despite the small absolute risk described in this study, lithium exposure remains a risk to this patient and her unborn child. Therefore, to fully evaluate the risks and benefits of drug treatment, a multidisciplinary approach — involving a psychiatrist, obstetrician, and the primary care team — will be important.