The majority of the early case reports and studies, both prospective and retrospective, examined the tricyclic antidepressants (TCAs). Such studies examining the risk of congenital anomalies in over 400 cases of first-trimester exposure to TCAs have not revealed an increased risk of major malformation (44,45,46,47,48). Many reviews suggest that the selective serotonin reuptake inhibitors (SSRIs) are the first-line treatment of depression during the perinatal period (43,49,50). Two large meta-analyses
examining exposures to TCAs and SSRIs during pregnancy suggest that the tricyclic antidepressants and selective serotonin reuptake inhibitors are unlikely to contribute to major congenital anomalies when used in gravid women (45,51). There are substantial data supporting the safety of fluoxetine (52,53,54,55). A large European trial examining the outcomes of 969 infants exposed to citalopram also did not reveal any significant increase in congenital anomalies (56). Other studies have explored the reproductive safety of other SSRIs, including sertraline, fluvoxamine, and the serotonin-norepinephrine reuptake inhibitor venlafaxine (44,56,57,58); these accumulating data are also reassuring and do not suggest increased risk of anomalies. Thus, until recently, a substantial body of research had not linked the SSRIs to a rate of teratogenicity above the 1% to 3% background rate in the population at large.

Given that background, however, a recent preliminary and unpublished retrospective analysis of 3,581 pregnant women in the United States conducted by GlaxoSmithKline (GSK) resulted in release of an advisory (subsequently incorporated into a Health Canada advisory) suggesting that infants exposed to paroxetine during the first trimester had an increased risk of congenital malformations, particularly cardiovascular anomalies. Paroxetine was moved to a pregnancy category D medication at the time of this release. The prevalence of major congenital defects and of cardiovascular defects in paroxetine- exposed pregnancies were 4% and 2%, respectively, as compared with 3% and 1% in one estimate of the overall prevalence in the U.S. general population regardless of drug usage (59). This advisory has been questioned because it is based on non-peer-reviewed and retrospective analyses. It was noted (60) that a large proportion of these defects were ventricular septal defects (of the muscular type), which are minor, resolve spontaneously, and may have been discovered only because of increased use of echocardiograms carried out on depressed and anxious women. Another large prospective sample of 1,170 infants in eight teratogen information centers throughout the world did not link paroxetine to an increased risk of congenital anomalies. In this study, 0.8% of infants in the paroxetine-exposed group were born with cardiovascular anomalies, and 0.7% of infants in the nonexposed group had this complication (61,62).

Data about other antidepressants are somewhat incomplete. GlaxoSmithKline has also sponsored a retrospective analysis of congenital malformations in a study of 7,005 infants with antidepressant exposure during pregnancy, 1,213 of whom were exposed to bupropion. In this study, there was no greater risk of congenital malformations overall (2.3%) or of cardiovascular malformations specifically (1.1%) following first-trimester bupropion exposure compared with exposure to all other antidepressants in the first trimester. The results of the study have not yet been corroborated (63).

Although the vast majority of studies to date do not corroborate major malformations, some studies do suggest that first-trimester exposure to SSRIs may contribute to minor malformations, increase preterm delivery, and restrict fetal growth (64,65,66).

With regard to direct neonatal toxicity, there have been studies suggesting that prenatal exposure to the SSRIs is associated with a neonatal withdrawal syndrome manifesting in irritability, increased tone, increased crying, and sleep and appetite dysregulation (67) and may be associated with lower cord blood 5-hydroxyindoleacetic acid (68). Most physicians note that such difficulties in neonatal adaptation are relatively benign and short lived, and the vast majority of these infants return home with their mother on the second or third day of life. One recent and concerning study suggests that maternal paroxetine use in late pregnancy may be associated with persistent pulmonary hypertension of the newborn (68a). Many authors note that it may be the maternal depression, rather than the pharmacologic treatment, that is affecting neonatal outcomes. One study compared control (depression plus no fluoxetine) and treated (depressed plus fluoxetine) groups and found no effect of fluoxetine exposure on birth weight, gestational age, or neonatal intensive care admissions (69).

Limited longer-term outcome studies exist, although one important study assessed the developmental impact of antidepressants and
demonstrated that infants exposed to tricyclic agents and fluoxetine were similar to sibling controls on measures of IQ and learning at the age of 5 years (70,71).

Although several studies have focused on the fetal safety of antidepressants, there is a paucity of research documenting the effectiveness of antidepressants during pregnancy. Largely because of concerns regarding human subjects, there are no randomized, controlled studies examining antenatal use of antidepressants and no specific dosing guidelines for use in pregnancy (72). There is, however, a body of literature that suggests careful monitoring of women using antidepressants during pregnancy is essential. Earlier studies with the tricyclic antidepressants revealed that as maternal plasma volumes increase, antidepressant blood levels fall, with consequent reemergence of mood symptoms during later pregnancy (73,74). Although the plasma levels of the SSRIs are not frequently monitored, it is a common clinical phenomenon to observe increases in mood symptomatology in the late second and third trimester, which respond to increasing doses of the antidepressant medication.

In summary, a great many studies suggest that undertreatment of depression is problematic for optimal pregnancy outcomes. Most studies do not suggest an increased risk for teratogenicity, but do suggest that SSRIs may contribute to some difficulty in neonatal adaptation, usually within the first 48 hours of life. Recent concerns about the potential contribution of paroxetine to cardiovascular anomalies and persistent pulmonary hypertension of the neonate are being explored. Always, the treatment of depression during pregnancy is guided by clinical experience, while carefully balancing the risks of undertreating the maternal illness with the potential negative consequences of fetal exposure. Typically, the risk/benefit analysis requires an assessment of severity of maternal illness, and this is where the job of the emergency room clinician may be simplified. The vast majority of cases that present urgently in an emergency setting are of sufficient severity that pharmacotherapy may be necessary. Clinicians should work collaboratively with the woman, her partner, and obstetrician to arrive at the safest decision given available evidence. For women who are SSRI naive, fluoxetine and citalopram have been the best studied during pregnancy and should be considered. Paroxetine is currently being more closely examined within the scientific community, and probably should be reserved for women who present already medicated during pregnancy or for those women who have proved refractory to other medications.