Every month seems to bring a new FDA advisory or an alarming research finding about the use of medications in pregnancy. In this article, we update you on what we consider to be the most important developments over the past couple of years.
A new study is reassuring on atypicals. That great engine of teratogen information, Toronto’s Motherisk Program, recently published an informative study of exposure to atypical antipsychotics. Researchers identified 151 pregnant women who were taking atypical antipsychotics and compared them to a matched control group of 151 unexposed pregnant women. Atypicals represented included the following medications: Zyprexa (N = 60), Risperdal (N = 49), Seroquel (N = 36), and clozapine (N = 6). The group reported that rates of major malformations were not different between the two groups (0.9% in the exposed group versus 1.5% in the nonexposed group). There were also no significant differences in premature birth, weight at birth, or obstetrical complications (McKenna et al., J Clin Psychiatry 2005;66:444-449).
Experts have generally considered conventional antipsychotics, particularly Haldol (haloperidol), Prolixin (fluphenazine), and Thorazine (chlorpromazine), to be relatively safe, based largely on older studies of pregnant women who took these meds for the treatment of hyperemesis gravidarum (for a good review of this data, see Psychiatry 2005;2(8):36-44). This new data implies that Zyprexa, Seroquel, and Risperdal are relatively safe. We have less data on Abilify and Geodon.
Neonatal SSRI syndrome. SSRIs are associated with a “neonatal SSRI syndrome” of jitteriness, irritability, increased muscle tone, and respiratory distress. Nobody knows for sure whether this represents SSRI withdrawal or toxicity. A meta-analysis of studies indicated that 30% of infants exposed to SSRIs in the third trimester get this syndrome, vs. about 8% of infants with no third-trimester exposure (Moses-Kolko et al., JAMA 2005;293:2372- 2383). Neonatal syndrome does not appear to be a medically serious problem; none of the studied infants died, and all of them ended up going home with their mothers within a few days of birth, with no apparent long-term problems.
SSRIs and persistent pulmonary hypertension of the newborn. Every ten years, the epidemiologist Christina Chambers publishes bad news about SSRIs and pregnancy in the New England Journal of Medicine. It was her NEJM study in 1996 that first suggested that SSRIs (in that case, fluoxetine) may lead to perinatal complications (Chambers et al., NEJM 1996;335:1010-1015). Her latest salvo is a study linking SSRI exposure to persistent pulmonary hypertension of the newborn (PPHN) (Chambers et al., NEJM 2006;354:579-587). What is PPHN? It’s a condition leading to significant shunting of blood flow away from the lungs; about 10% of infants with this condition die and half end up with major cognitive and neurological problems.
The Chambers PPHN study was a well-done case-control study. She and her colleagues identified 377 women across the United States and Canada whose infants had PPHN and matched them with a comparison group of 836 women with normal infants. They discovered that infants with PPHN were six times more likely than unaffected infants to have been exposed to SSRIs in the second half of the gestation. What does this mean in absolute numbers? The baseline population risk of PPHN is about 1.5 out of 1,000 births. Thus, if Chambers’ estimate of a sixfold relative risk is accurate, about one out of a hundred, or 1%, of SSRI-exposed babies would be expected to develop PPHN. These are not reassuring odds!
However, this is the only study that has shown such an association. While statistically significant, it was based on only 14 infants who developed PPHN after late gestational SSRI exposure. Furthermore, the study did not control for maternal depression, which many authorities believe is a risk factor for birth defects in itself.
Thus, most experts are not advocating a huge change in your prescribing practices based on this data. As always, prudence dictates avoiding any medication in pregnancy if possible. But sometimes, the risk of a serious psychiatric relapse outweighs the risk of medication.
Paxil is now the SSRI to avoid. Recently, two large databases reported risks of congenital heart defects with Paxil (paroxetine). In a Swedish birth registry, Paxil use was associated with double the risk of an atrial or septal cardiac birth defect. In a U.S. insurance claims database, Paxil incurred 1.5 times the risk of cardiac defects and 1.8 times the risk of overall birth defects. In response to these data, an advisory committee of the American College of Obstetricians and Gynecologists advised against use of Paxil in pregnancy (Committee Opinion No. 354, Obstetrics & Gynecology 2006; 108:1601-1604), and the FDA downgraded Paxil’s pregnancy risk category to category D, the most risky category short of X (the FDA advisory, which summarizes this data, can be found by visiting: http://www.fda.gov/cder/drug/advisory/ paroxetine200512.htm).
Lithium may be safer than previously thought. The most serious risk of first-trimester lithium exposure is a cardiac valve defect called Ebstein’s anomaly, a condition that carries a mortality rate of up to 50%. The baseline frequency of Ebstein’s anomaly is one in 20,000. The most recent good review of all relevant studies pertaining to lithium in pregnancy was published a full 13 years ago (Cohen et al., JAMA 1994;271:146-150), but no further relevant exposure data has been generated since then. That review estimated that first trimester lithium use increases the risk of Ebstein’s anomaly about tenfold, meaning that the absolute risk of lithium causing the defect is about one in 2,000. This is a very low risk, so most experts in the field of perinatal psychiatry will counsel patients to stay on lithium if a serious relapse is likely to occur off the drug. As usual, this is a judgment call.
Valproic acid is even more hazardous than previously thought. A recent study indicates that valproic acid (VPA; Depakote and others) is an even worse teratogen than previously thought. Researchers determined the Antiepileptic Drug Pregnancy Registry and compared 149 first-trimester VPA-exposed infants with two control groups: infants exposed to antiepileptic drugs (AEDs) other than VPA, and infants exposed to no teratogens at all. Overall, 10.7% of VPA-exposed infants had major malformations, versus 2.9% of the AED-exposed group, and 1.6 % of the unexposed group. There were three cases of neural tube defects in the VPA group which is a well_known risk, but there were various other anomalies, including cardiac and pulmonary defects (Wyszynski, D F, Neurology 2005; 64(6): 961-965). By the way, most of these women were taking prenatal folic acid supplements, which did not prevent the occurrence of neural tube defects. Bottom line: Avoid valproic acid in the first trimester!
Lamictal may cause cleft palate. Lamictal (lamotrigine) has been thought to be one of the safer treatments for bipolar disorder in pregnant women. However, the FDA recently reported data from the North American Antiepileptic Drug Registry indicating that, out of 564 there were five cases of cleft palate, a prevalence of 0.9%. This is about 10 times higher than the baseline prevalence of 0.1% (http://www.fda.gov/cder/ drug/InfoSheets/HCP/lamotrigineHCP.pdf). However, the FDA also noted that other pregnancy registries have not replicated this finding, so it’s not clear how real it is. Even if the finding is real, Lamictal is far safer than Depakote and somewhat safer than lithium, which causes a very low prevalence of a far more serious condition, Ebstein’s anomaly.
There is nothing new to report on benzos. Many studies have been conducted to examine possible teratogenic effects. While the results are conflicting, most authorities appear to have settled on the concept that benzodiazepines may double the risk of cleft palate from the baseline rate of one in 1,000 to about two in 1,000 (see a recent Medscape review at http://www.medscape.com/viewarticle/ 512650). Using relatively low doses of shorter halflife agents (e.g., 0.5 mg lorazepam or alprazolam) is likely to minimize this risk.
Omega-3 consumption and smarter babies. A recent article in The Lancet showed that pregnant women who consumed large amounts of seafood (more than three servings per week) gave birth to babies with higher IQs than women who ate less seafood (fewer than 3 servings per week). Because the researchers asked respondents about specific types of fish consumed (only certain fatty fish have lots of omega-3), they were able to ascertain that higher levels of dietary omega-3 were specifically associated with more intelligent kids. The researchers controlled for the possible confounding effects of maternal IQ in a number of ways, so these findings appear legitimate (Hibbeln JR et al., The Lancet 2007;369:578-585).