MADRID – Neonatal outcomes were generally worse among pregnant women who take antipsychotics than among those in the general population, but the rates of birth defects were not significantly altered, with an incidence of 5% compared with 4% in the general population.
Outcomes varied with the type of medication, Heather Gilbert, R.N., said at the World Conference on Women’s Mental Health. * The infants of women taking clozapine were most at risk, with two of the six babies born to these mothers (33%) exhibiting a congenital malformation. These infants also had a higher rate of neonatal complications, including premature birth, respiratory distress, and admission to the neonatal intensive care unit, according to Australia’s National Register of Antipsychotic Medications in Pregnancy. The incidence of gestational diabetes also was higher than among the general population (16% vs. 4%).
Dr. Jayashri Kulkarni
It was not surprising that clozapine proved particularly troublesome for babies exposed in utero, primary investigator Dr. Jayashri Kulkarni said in an interview. "The clozapine babies did have the worst outcomes, but this was not unexpected, because that is the most difficult medication to manage. But even in that worst group, the adverse issues were not catastrophic."
Thus, she said, the registry is "essentially a good-news story. "Most of the babies born were pretty healthy. We saw that the use of these medications did not lead to miscarriages and stillbirths," said Dr. Kulkarni, director of the Monash Alfred Psychiatry Research Centre in Melbourne. "What is important, from our standpoint, is that if a patient is doing well on a medication, it’s important to keep that medication going during pregnancy and not let her mental health deteriorate."
In fact, the tragic story of a patient who did stop her antipsychotic drugs during pregnancy inspired Dr. Kulkarni to begin the project .
The National Registry of Psychotic Medications in Pregnancy (NRAMP) obtained ethical approval from 42 boards across Australia before beginning to recruit patients. All women who register undergo a psychiatric evaluation at baseline, which includes the PANSS (Positive and Negative Syndrome Scale). Information about the pregnancy and outcomes is supplied by the clinicians attending the mother. The mother/infant pair is followed for 1 year, with six to eight postnatal interviews using the maternal assessment scales. The baby is assessed for developmental milestones at 6 and 12 weeks and at 6 and 12 months.
Because the registry is not a randomized controlled trial, Dr. Kulkarni has met with some skepticism from clinicians, who say that the data are not a reliable method of judging drug safety. "Ethically, a randomized trial is not something that we could do," she said. "We have to cope with people saying the data are messy," at least until the registry accumulates a much larger number of mother/infant pairs.
The participants’ mean age at recruitment was 32 years; the mean baseline PANSS score was 37. Diagnoses included bipolar affective disorder (40 patients), schizophrenia (28), schizoaffective disorder (8), unspecified psychotic disorder (8), depression (6), anxiety disorder (3), schizophreniform disorder (2), and borderline personality disorder (1). Diagnostic data were missing in four participants.
The largest proportion of the 100 women in the registry thus far took quetiapine during their pregnancy (43); the next most commonly used drug was olanzapine (14). Other medications were risperidone (6), clozapine (6), aripiprazole (4), haloperidol (4), long-acting injectable risperidone (Consta; 3), flupenthixol (2) and ziprasidone (1). In all, 12 women registered but did not receive antipsychotic treatment, and there are missing data on 5.
Pregnancy outcomes were somewhat better than those seen in the general population, with a 2% rate of stillbirth and miscarriage, compared with a 7% rate in the general population. There were two miscarriages (one exposed to quetiapine and one to clozapine) and one stillbirth (quetiapine).
Premature births occurred in 15% of the registry infants, compared with 8% of the general population. This occurred in nine infants exposed to quetiapine (21%), one exposed to risperidone (17%), one to flupenthixol (50%) and in the one exposed to ziprasidone.
Respiratory distress occurred in 27% of the registry infants, compared with 7% of the general population. It was seen in 14 babies exposed to quetiapine (32%), 5 exposed to olanzapine (36%), 1 exposed to risperidone (17%), 2 exposed to clozapine (33%), 1 exposed to haloperidol (24%), 2 exposed to Consta (66%) and in the single baby exposed to ziprasidone.
Medication withdrawal occurred in seven babies who were exposed to quetiapine (16%), and one exposed to olanzapine (7%).
Admission to the neonatal intensive care unit was highest among those exposed to clozapine (three patients; 50%) and flupenthixol (one; 50%). Birth weight below the 10th percentile occurred in babies exposed to quetiapine (four; 9%), olanzapine (one; 7%), risperidone (two; 33%), clozapine (one; 17%), and flupenthixol (one; 50%)
NICU admission also occurred in seven babies who were exposed to quetiapine (16%), one exposed to olanzapine (19%), three exposed to clozapine (50%), two exposed to Consta (66%), and one exposed to flupenthixol (50%).
Large-for-gestational-age babies (greater than the 90% percentile) occurred in eight who were exposed to quetiapine (19%), one exposed to olanzapine (7%), one exposed to aripiprazole (25%), one exposed to haloperidol (25%), and in the single baby exposed to ziprasidone.
Five infants were born with congenital anomalies. One infant who was exposed to clozapine had craniosynostosis and hypertelorism; the mother took 350-750 m/day throughout the pregnancy. The baby’s skull deformity had a partial surgical correction and resulted in a developmental delay at 12 months.
The second infant had gastroschisis and a horseshoe kidney. The mother took 400 mg of clozapine daily throughout pregnancy. The gastroschisis had a surgical correction, whereas the kidney is being managed conservatively and the baby is progressing well.
Congenital malformations occurred in 2 of the 14 babies who were exposed to quetiapine. One baby had an atrial-septal defect; the mother took 1,110 mg/day of quetiapine throughout pregnancy as well as 1,000 mg/day of zuclopenthixol. The defect was resolving spontaneously by 12 months with conservative management.
The second baby who was exposed to quetiapine had a cleft lip and palate, as well as hydrocephalus. The mother took 600 mg/day well throughout the pregnancy. The clefts were surgically repaired and excess cephalic fluid was drained. The baby is progressing and being monitored.
[Sidebar: Removing Antipsychotic Medications in Pregnancy: A Case in Point ]
One anomaly occurred in a baby exposed to risperidone: This infant had an abnormal renal collecting tubule and bilateral talipes. The mother took 2 mg/day risperidone up to 36 weeks’ gestation. The abnormal collecting tubule is functioning well, and the talipes underwent surgical repair and braces, with a good result.
Gestational diabetes accounted for most of the maternal complications. "We are seeing a lot of gestational diabetes in the moms – higher than the general population. Maybe this is because they were already overweight, or the drugs contributed to weight gain and insulin intolerance. But clinicians should be aware of this heightened risk, look diligently for signs of it, and give this information to their patients."
The disorder occurred in 16 of the 100 women, including 6 who were taking quetiapine (14%), 4 taking olanzapine (28%), 1 taking risperidone (17%), 4 taking clozapine (67%), and 1 taking Consta (33%). Preeclampsia occurred in one patient who was taking quetiapine (2%).
The registry also tracked smoking, as well as alcohol and substance abuse, but did not control for these factors in any multivariate analysis. Antenatally, smoking was noted in 31% of registry women, compared with 16% of the general population. But antenatal alcohol consumption among registry participants was lower than that in the general population (12% vs. 60%). Substance abuse incidence was about 12% in the registry participants, compared with 7% in the general population. "We have lots of stories from women who smoked a lot of cigarettes and drank a lot, and decreased this during pregnancy. That protective instinct of the mother for the fetus is alive and well, and kicked in among these women."
Managing antipsychotic medications during pregnancy is a new concept to most clinicians in Australia, who continue to assume that the drugs will tip the risk:benefit ratio to the pit of endangering the developing baby, Dr. Kulkarni said. This has led to many recommendations that women with serious mental disorders avoid ever bearing children. She has even received calls from physicians who are angry that the registry seems to support childbearing in this group, "asking me why I’m allowing the gene pool to be polluted."
In her daily work, Dr. Kulkarni deals with significant regrets from women who were told that their mental illness and medication should preclude any childbearing. "I see a number of women in their 50s, approaching menopause, who were given the advice many years back never to have children because of the drug dangers," she said. "Now, they are grieving and experiencing depression over what could have been. The medical profession does not play God. We can only provide information and advice, but not take the role of driving the decision about having children or not."
Dr. Kulkarni said that the registry’s results are interesting, but until many more women are enrolled, it’s difficult to use it as a clinical guideline about what drug is safest for both mother and baby. "We need to accumulate a lot more data, and in the future, we hope to expand the program to sites all over the world."
The NRAMP project data are online; clinicians can register to participate or get more information about setting up similar programs . Physicians interested in obtaining more information about the registry can also e-mail Dr. Kulkarni or Gilbert, the project’s research nurse .
The project is funded by Astra-Zeneca, Janssen-Cilag, Mayne Pharmaceuticals, and the Australian Rotary Health Research Fund. Dr. Kulkarni had no financial disclosures.
* CORRECTION, 5/10/2011: The original version of this article misspelled Heather Gilbert's last name. This version has been corrected.