In addition, since subjects were enrolled in a consultation clinic, surveillance bias might have detected a higher number of CV malformations. “I don’t have a lot of confidence in this study because there were enormous sources of risk that were not controlled for,” said Dr. Meyer, who is also a psychopharmacology consultant for the California State Department of Hospitals.
In a separate study, researchers led by Lee S. Cohen, MD, assessed data from 487 women in the National Pregnancy Registry for Atypical Antipsychotics based at the Massachusetts General Hospital Center for Women’s Health, Boston. Of the 487 women, 353 were on atypical antipsychotics and 134 served as controls (Am J Psychiatry. 2016;73[3]:263-70). Medical records were obtained at baseline, month 7, and postpartum for 82% of subjects, which left 303 women in the final analysis. Covariates included demographic characteristics, medication use and dosage changes, social habits including smoking, use of alcohol and illicit drugs, medical and psychiatric history, and family history of birth defects.
Of 214 live births with first-trimester exposure to atypical antipsychotics, three major malformations were confirmed, while among the control group of 89 women, one major malformation was confirmed. The absolute risk of major malformations was 1.4% for exposed infants and 1.1% for unexposed infants. Meanwhile, the OR for major malformations comparing exposed infants was 1.25, which did not reach statistical significance (95% CI, 0.13-12.19). Limitations of the study include the small sample size of live births and the low overall rate of malformations, said Dr. Meyer. “This group of women living in the Boston area might not be representative of the general population based on the extremely low rates of congenital malformations for both cohorts in this study,” he commented.
In what Dr. Meyer said was the most robust study of its kind to date, researchers led by Krista F. Huybrechts, MS, PhD, drew from Medicaid data from 2000-2010 and included only women who were enrolled from 3 months before their last menstrual period through at least 1 month after delivery (JAMA Psychiatry. 2016;73[9]:938-46). The sample consisted of 1,341,715 pregnancies. Among those pregnancies, 9,258 women filled at least one prescription for an atypical antipsychotic, and 733 filled at least one prescription for a typical antipsychotic, for a total of 9,991 pregnancies. The researchers used propensity score matching to match for risk of antipsychotic exposure. They also balanced the antipsychotic-exposed and nonexposed groups for covariates that might be related to the outcome of interest (major congenital malformations), including (but not limited to) calendar year, age, race, smoking history, multiple gestation, indications for antipsychotic use, other maternal morbidity, concomitant medication use, and general markers of illness burden in the 3 months prior to pregnancy.
The atypical antipsychotics used included quetiapine (n = 4,221), followed by aripiprazole (n = 1,756), risperidone (n = 1,566), olanzapine (n = 1,394), and ziprasidone (n = 697). The absolute risks for congenital malformations per 1,000 live-born infants was 38.2 (95% CI, 26.6-54.7) for those treated with typicals and 44.5 (95% CI, 40.5-48.9) for those treated with atypicals versus 32.7 (95% CI, 32.4-33.0) for untreated women. In the fully adjusted analysis, the risk ratio was not statistically different for those exposed to atypical antipsychotics, compared with the control group, for malformations overall (relative risk, 1.05; 95% CI, 0.96-1.16) nor for cardiac malformations (RR, 1.06; 95% CI, 0.90-1.24). However, the risk remained elevated for risperidone for overall malformations (RR, 1.26; 1.02-1.56) and cardiac malformations (RR, 1.26; 95% CI, 0.88-1.81).
When Dr. Huybrechts and her colleagues redefined exposure as having filled two more prescriptions or having at least a 1-day supply in the first trimester, the results did not meaningfully change. However, the association appeared to strengthen somewhat for risperidone when filling two or more prescriptions (RR, 1.46 for any malformation; 95% CI, 1.01-2.10; RR, 1.87 for cardiac malformations; 95% CI, 1.09-3.19). The researchers observed no evidence of a dose-response relationship for any of the individual antipsychotics except risperidone. Risperidone dosages of 2 mg/day or more were associated with an increased risk for cardiac malformation (RR, 2.08; 95% CI, 1.32-3.28).
“The small increase in absolute risk and RR for malformations observed with risperidone should be interpreted with caution because no apparent biological mechanism can readily explain this outcome and the possibility of a chance finding cannot be ruled out,” the authors wrote in their study. “This finding should therefore be interpreted as a potential safety signal that will require follow-up in other studies.”
If the finding in this study is replicated, Dr. Meyer said, a number needed to harm analysis suggests that compared with no antipsychotic use. “Given the risks of bad outcomes that might occur related to medication switching or nonadherence, the risk-benefit ratio may tilt towards continuing risperidone, especially when a long-acting injectable [LAI] antipsychotic is needed in someone who responds to risperidone and either doesn’t respond to or tolerate the medications available in other LAI preparations such aripiprazole, haloperidol, or fluphenazine,” he said. “This need to balance the risks of exacerbating the mental disorder and the extremely small chance of an adverse pregnancy outcome is part of a clinical discussion one should have with the patient and her family.”
Dr. Meyer reported having received speaking or advising fees from Acadia, Alkermes, Allergan, Intracellular Therapies, Merck, Neurocrine, Otsuka America, Sunovion, and Teva.