Treating mental disorders in pregnant or breast-feeding women is an important clinical issue (American College of Obstetricians and Gynecologists Committee on Practice Bulletins—Obstetrics [ACOG], 2008). In the first article of this series (Howland, 2009b), I described the kinds of studies conducted to evaluate the safety of medications during pregnancy and lactation. In the second article (Howland, 2009a), I described the current and proposed U.S. Food and Drug Administration (FDA) classification system for medication labeling in these situations. In this article, I discuss general principles and guidelines for prescribing psychotropic medications during pregnancy and lactation.
Medication Safety Concerns During Pregnancy and Lactation
The ways in which psychotropic medications might affect pregnancy depend on the timing of exposure (Altshuler et al., 1996). Fetal physical malformations are most likely to arise from drug exposures during the first trimester when organ and limb development occurs. This is an especially important potential problem for unplanned pregnancies, because the patient might not know she is pregnant until well into the first trimester. Drugs might affect intra-uterine fetal growth, contributing to growth retardation (resulting in low birth weight neonates) or weight gain (resulting in large for gestational age neonates).
Perinatal drug effects might adversely affect labor and delivery. Drugs can have immediate postnatal effects, including drug withdrawal effects or toxicity in the neonate (e.g., respiratory, feeding, or neuromuscular disturbances). Some drugs might also affect the fetus in ways that are apparent only during early childhood development or even into adulthood. Finally, infants could be adversely affected by drug exposure during breastfeeding, but this will depend on understanding the extent of drug transfer into breast milk, as well as the consequent effects of breast milk drug exposure in infants.
Since the initial introduction of tricyclic antidepressant (TCA) drugs, antidepressant medications have been widely used in clinical practice for nearly 50 years. The serotonin reuptake inhibitor (SRI) drugs are the most commonly used class of antidepressant agents, and they are one of the most commonly prescribed drugs among all prescription medications. Hence, clinical experience with and research data on antidepressant drug use during pregnancy and lactation is more extensive than for other psychotropic drugs. By and large, there is limited evidence that TCA and SRI drugs are significantly harmful when used during pregnancy or breastfeeding (ACOG, 2008; Freeman, 2009). The best available information pertains to the use of fluoxetine (Prozac®), paroxetine (Paxil®), and sertraline (Zoloft®). Concerns about possible adverse cardiovascular effects with paroxetine in particular have been raised, prompting a change in its FDA pregnancy classification to D. The methodology of the studies identifying a specific cardiovascular risk for paroxetine has been strongly criticized, however, and the totality of information about paroxetine does not consistently support an association (ACOG, 2008; Einarson et al., 2008; Gentile & Bellantuono, 2009).
Newer generation non-SRI atypical antidepressant drugs also have limited evidence of safety concerns, although it should be noted that these drugs have been less well studied and have been used clinically for a much shorter time compared with TCA and SRI drugs (Way, 2007). With the exception of paroxetine (FDA D rating) and bupropion (Wellbutrin®) (FDA B rating), the TCA, SRI, and atypical antidepressant drugs have FDA C ratings.
During the postpartum period, neonatal SRI-discontinuation syndromes have been described but without significant sequelae. Long-term follow up of infants born to mothers taking antidepressant drugs during pregnancy have not consistently revealed significant adverse developmental effects, whereas the long-term effects of antidepressant exposure during lactation have not been well studied. Various psychotherapies are effective for treating depression, and they could be an appropriate alternative to medication (Raudzus & Misri, 2009).
Anti-Anxiety and Sedative-Hypnotic Medications
Benzodiazepine drugs had been associated with a small risk of birth defects in early studies, although more recent case control studies have not confirmed this association (ACOG, 2008). Those benzodiazepine drugs used for anxiety, such as alprazolam (Xanax®), lorazepam (Ativan®), and clonazepam (Klonopin®), all have FDA D ratings, whereas those used for insomnia, such as estazolam (ProSom®), flurazepam (Dalmane®), quazepam (Doral®), temazepam (Restoril®), and triazolam (Halcion®), all have X ratings. By contrast, the non-benzodiazepine drugs buspirone (Buspar®), used for anxiety, and zolpidem (Ambien®), zaleplon (Sonata®), and eszopiclone (Lunesta®), all marketed for insomnia, have not been associated with significant adverse pregnancy outcomes. These drugs all have FDA C ratings. A concern with the use of benzodiazepine drugs during late pregnancy is the risk of postpartum neonatal toxicity (“floppy infant syndrome”) and withdrawal syndromes. Sedation in infants is possible with benzodiazepine use during breastfeeding, but these drugs are less concentrated in breast milk compared with other psychotropic drugs. The long-term effects on infants whose mothers took anti-anxiety or sedative-hypnotic drugs have not been clearly established.
Because SRI drugs are effective for various anxiety disorders, these drugs and buspirone might be preferred over benzodiazepine drugs for the prenatal and postpartum treatment of anxiety disorders. Likewise, low dosages of antidepressant drugs, such as trazodone (Desyrel®), mirtazapine (Remeron®), or sedating TCA drugs, could be considered for the treatment of insomnia in such women. Variations of cognitive-behavioral psychotherapy are effective for treating anxiety disorders and insomnia, and these would be considered clinically appropriate to use as an alternative to medication (Ross & McLean, 2006).
Antipsychotic medications are prescribed less frequently than antidepressant drugs, but they have been used clinically for more than 50 years. Overall, there is limited evidence that antipsychotic drugs pose a significant hazard when they are used during pregnancy or breastfeeding (ACOG, 2008), although much more data are available for first generation (typical) antipsychotic (FGA) drugs compared with second generation (atypical) antipsychotic (SGA) drugs. Because maternal obesity is associated with a small but significantly increased risk of congenital anomalies (Stothard, Tennant, Bell, & Rankin, 2009), there are potential concerns about weight gain and use of antipsychotic medications.
A recent study found that infants exposed to SGA drugs had a significantly higher incidence of being large for gestational age and having a heavier birth weight compared with infants exposed to FGA drugs (Newham, Thomas, MacRitchie, McElhatton, & McAllister-Williams, 2008). Sedation and parkinsonian effects are possible in breastfed infants whose mothers are taking antipsychotic drugs. With the exception of clozapine (Clozaril®), FGA and SGA drugs have FDA C ratings. Clozapine has a B rating but has been associated with several reports of serious adverse effects in breastfed infants (Gentile, 2008). No form of psychotherapy would be considered appropriate to use alone as an alternative to medication for the treatment of schizophrenia.
Lithium (Eskalith®, Lithobid®) is associated with an increased risk of congenital cardiac malformations and has an FDA D rating (ACOG, 2008). The use of lithium during breastfeeding also is associated with significant adverse effects in infants. The anticonvulsant drugs valproic acid (Depakene®), valproate sodium (Depacon®), divalproex sodium (Depakote®), and carbamazepine (Tegretol®, Equetrol®) are each associated with birth defects, although the relative risk is greater with valproic acid (Harden & Sethi, 2008; Meador, 2008; Meador et al., 2008). Both drugs have FDA D ratings, but they do not pose a significant risk to infants during breastfeeding. Unlike the other anticonvulsant drugs, lamotrigine (Lamictal®) has not been associated with a significant risk of birth defects (FDA C rating), but its clinical effects in breastfed infants are not well established (Newport, Pennell, et al., 2008).
Because women with bipolar disorder have a high risk of relapse, especially during the postpartum period, pharmacological treatment is mandatory, and psychotherapy alone is not appropriate (Yonkers et al., 2004). Lithium and anticonvulsant drugs (with the exception of lamotrigine) should be avoided during pregnancy. Antipsychotic drugs, having anti-manic efficacy, are an appropriate alternative if lamotrigine monotherapy is not sufficient (Newport, Stowe, et al., 2008). Lithium can be started immediately postpartum in non-breastfeeding women. Valproic acid or carbamazepine could be initiated postpartum even if breastfeeding is planned.
Medication Management During Pregnancy and Lactation
Considerations for the use of medication during and after pregnancy should begin before pregnancy ever occurs. Is the patient sexually active, using birth control, capable of becoming pregnant, or planning a pregnancy? Would the patient expect to breast-feed? Patients should be advised that pregnancy should be planned but also told what to do if they become pregnant unexpectedly. Discussions should include the patient and partner, primary care provider, obstetrician, pediatrician, and mental health clinician. Untreated maternal psychiatric illness can have adverse effects on pregnancy outcome and infant well-being. Therefore, available treatments and their potential risks when used during pregnancy or lactation should be discussed in the context of the risks of not treating maternal psychiatric illness effectively.
The selection and use of psychotropic medication during pregnancy or lactation should be based on a careful analysis and discussion of various risks and benefits. What is the risk of not using medication? What is known about the relative risks of different psychotropic medications for the condition? Are any nonpharmacological alternative therapies available? Are there other current medical conditions that might affect pregnancy outcome? Is the patient taking other nonpsychotropic medications? Does she use nicotine, alcohol, or any illicit drugs?
If the patient is treatment naïve, selecting a medication with the best-known reproductive and breastfeeding safety profile is the best approach. For patients with a known treatment history, medication that has been previously or is currently effective should be used or continued. Unless the patient has a strong preference, trying or switching to a different medication that does not have a clear safety advantage over previously or currently effective medication is not advisable. The risk of not effectively treating the patient during pregnancy or lactation by trying a new medication is probably greater than any potential safety advantage.
Regardless of what medication is used, mother and fetus should be closely monitored during pregnancy. Because the pharmacokinetics of many drugs are affected during pregnancy, dosage adjustments might be needed to maintain optimal effectiveness (Fotopoulou et al., 2009; Sit, Perel, Helsel, & Wisner, 2008). Using the lowest effective drug dosage is appropriate. However, targeting a lower subtherapeutic dosage to further minimize fetal exposure, resulting in mild or residual symptoms, should be discouraged. This practice may not only expose the fetus to any potential adverse effects of drug exposure but also to the effects of the inadequately treated illness.
Clinical monitoring of mother and infant during breastfeeding is also obviously important. The pharmacokinetics of many drugs may change again in mothers during the postpartum period. Laboratory studies in infants (e.g., monitoring drug concentrations) are not routinely necessary during breastfeeding. Breastfed infants should be closely observed for any behavioral changes that might be associated with the drug.