Biological Therapies in Psychiatry
The use of antidepressant drugs by women who are pregnant or breast-feeding is an important issue in psychopharmacology. Last year, we reviewed the risks of women taking serotonin-selective reuptake inhibitor (SSRI) antidepressants during pregnancy or lactation (BTP 1996; 19: 47-48). The basic questions still remain: Are antidepressant drugs safe for women to take during pregnancy and lactation? Are some antidepressants safer than others? A number of recent papers address these topics.
In our previous article, we reviewed a study by Chambers et al, who found that children of women who took fluoxetine (Prozac and others) during pregnancy had a greater incidence of three or more minor anomalies than did the offspring of women who had taken drugs not considered teratogenic. (1) In addition, infants exposed to fluoxetine in the third trimester were born prematurely more often than those exposed during the first or second trimester and had higher rates of admission to special care nurseries, poor neonatal adaptation, low birth weight (their mothers had not gained much weight), and short birth length. These results were questioned in an editorial by Dr Elizabeth Robert, who pointed out methodologic problems in this study. (2) Additional methodologic concerns were subsequently noted in a letter from Cohen and Rosenbaum, who emphasize that both the risk of relapse following antidepressant discontinuation and the risk of puerperal worsening of mood in women with histories of depression argue against discontinuing antidepressant drugs late in pregnancy "without careful consideration." (3)
Cohen et al assessed birth outcomes following prenatal exposure to fluoxetine in 31 newborns. (4) There was a nonstatistically significant trend towards a higher risk of neonatal complications to be associated with a longer total duration of fluoxetine exposure. However, as the authors point out, a longer duration of treatment could indicate more severe depression or anxiety; both of which have been associated with poor neonatal outcome. Also, this was a retrospective study with a small sample size and no control group.
A report from Mhanna and others attributes jitteriness and bleeding abnormalities in a newborn to fluoxetine exposure. (5) A full-term baby born to a 34-year-old mother who was taking fluoxetine, 60 mg daily, had scattered petechiae, a small subdural hematoma, extreme jitteriness, and hypertonia. On his second postnatal day, serum levels of fluoxetine and norfluoxetine were 129 ng/mL and 227 ng/mL, respectively. (Adults taking therapeutic doses of fluoxetine typically have levels of 91 to 302 ng/mL of fluoxetine and 72 to 258 ng/mL of norfluoxetine.) Discharged home on formula feeding, the infant showed dramatic improvement in jitteriness by 2 weeks and was normal at a 5-month followup.
Goldstein et al at Eli Lilly, a manufacturer of fluoxetine, prospectively identified 123 infants exposed to fluoxetine during all trimesters and failed to find an association between fluoxetine exposure and neonatal complications. (6) In another report, Goldstein and coworkers describe the outcomes of 796 pregnancies in which fluoxetine was taken by the mother during the first trimester. (7) Spontaneous abortions occurred in 110 (13.8%) of the pregnancies. Malformations, deformations, and disruptions were reported in 34 (5.0%) of the remaining 686 pregnancies. These rates are consistent with those found in historic reports of newborn surveys.
Nulman and others acquired subjects through the Motherisk program, a service in Toronto that provides information and consultation concerning the potentially hazardous effects of assorted agents in pregnancy. (8) They studied the children of 80 women who had taken tricyclic antidepressants during pregnancy. Most took the agents for depression, but some used them for migraines, pain, or bladder control. Investigators also studied the children of 55 women who took fluoxetine while pregnant and another 84 children whose mothers had not been exposed to any agent known to affect a fetus adversely. The children's global IQ and language and behavioral development were assessed between 16 and 86 months of postnatal age using specialized rating instruments.
At birth there were no differences among groups in weight, height, head circumference, rates of perinatal complications, or incidents of major malformations. In later tests, IQ scores also were comparable among the three groups, both in younger and older children. More, there were no significant differences in temperament, mood, arousability, activity level, distractibility, behavior problems, or scores on neurobehavioral tests.
Last year, Dr Katherine Wisner and colleagues conducted a literature search on the use of antidepressant drugs in women who were breast-feeding. (9) They only included articles that reported serum drug levels in nursing infants. In these articles, the following drugs were unmeasurable in infant plasma and caused no adverse effects: amitriptyline (Elavil and others), nortriptyline (Pamelor and others), desipramine (Norpramin and others), clomipramine (Anafranil), dothiepin, (*) and sertraline (Zoloft). In infants whose mothers had been treated with doxepin (Sinequan and others) or fluoxetine, there were detectable drug serum levels and some adverse effects in the infants. One of these reports attributed colic in a nursing infant to high serum levels of fluoxetine plus norfluoxetine: 548 ng/mL. (10) However, the serum measurement occurred after only 2 days of nursing following 3 weeks of formula feeding (i.e., no exposure to fluoxetine). Wisner et al suggest that the high levels may have been the result of a laboratory error.
In another study, Wisner and Perel looked at nortriptyline levels in seven nursing mothers and their infants. (11) Mothers' serum nortriptyline levels were between 47 and 164 ng/mL, but there were no detectable blood levels of nortriptyline and no adverse clinical effects in the infants. There were low serum levels of nortriptyline's less active metabolite, 10-hydroxynortriptyline, in the two youngest babies. In a follow-up letter, these authors report on five additional infants whose mothers were taking nortriptyline during breast-feeding. (12) Serum nortriptyline levels in these women ranged from 47 to 201 ng/mL. Again, none of the infants had quantifiable nortriptyline levels, and none had adverse effects. In a more recent report, Wisner and others took serum samples at 4 weeks of age in six full-term infants and one premature baby whose mothers were taking nortriptyline during breast-feeding. (13) One infant had blood nortriptyline levels of 10 ng/mL but negligible traces of metabolite. Two other infants had low levels of metabolite. None of the infants had any ill effects. Mammen and colleagues reported on two nursing infants, aged 7.5 and 20 weeks, whose mothers were taking nortriptyline. (14) An active hydroxy metabolite of nortriptyline was detected in the serum of both babies, but neither showed adverse effects.
There have been two case reports of withdrawal syndromes in infants associated with SSRIs. In one, the mother took fluoxetine throughout pregnancy. (15) In the other, the baby developed a withdrawal syndrome when the mother discontinued sertraline after 3 weeks of breast-feeding (BTP 1996; 19: 2). (16) Contrary to these reports, Ratan and Friedman recently described a 32-year-old woman who took sertraline, 150 mg daily, from the 20th week of her pregnancy through birth, then breast-fed her baby while continuing to take the same dose of the antidepressant. (17) Eleven days postpartum, she weaned the child, who failed to show any withdrawal signs. In an earlier report, Altshuler et al. failed to find detectable plasma sertraline levels in a breast-fed infant whose mother was taking sertraline, 100 mg/day (BTP 1995; 18: 9). (18) The baby was breast-fed for 8 months with no apparent effects on growth or behavior.
Mammen and coworkers reported on three breast-fed infants whose mothers were taking sertraline, 50 to 100 mg/day. (19) In each infant, low levels (<2 ng/mL) of sertraline and the active metabolite norsertraline were detected. There were no ill effects or withdrawal reactions in the newborns. Wisner and colleagues also studied infants whose mothers were taking sertraline during breast-feeding (K.L. Wisner, MD, et al, unpublished data, 1997). These eight women were taking 50 to 200 mg/day and had sertraline serum concentrations of 27.9 to 285 ng/mL. Sertraline was detected in two of the eight babies and N-desmethylsertraline in six. One infant had substantial serum levels of both sertraline and N-desmethylsertraline, but none of the babies had any adverse effects. Stowe et al found detectable serum levels of sertraline in 3 and desmethylsertraline in 6 out of 11 nursing infants whose mothers were taking sertraline. (20) No adverse effects were observed in the infants. The serum concentrations detected were below the detection limit of most commercial laboratories.
Birnbaum and colleagues collected serum samples from infants of 28 women who took an antidepressant during the third trimester and/or during lactation (C.S. Birnbaum, MD, et al, unpublished data, 1997). There were no detectable drug levels in the 10 infants whose mothers were taking tricyclic antidepressants -- desipramine, imipramine (Tofranil and others), nortriptyline, or clomipramine -- either during the third trimester and during lactation or only during lactation. Eight infants whose mothers were taking fluoxetine during the third trimester and during lactation had detectable levels of fluoxetine and/or norfluoxetine. (One woman was also taking clonazepam [Klonopin] and valproic acid [Depakene]). In the children of 3 women taking fluoxetine, 1 taking paroxetine (Paxil), and 1 taking fluoxetine and clonazepam during the third trimester and during lactation, antidepressant levels were not detectable in infant plasma. Antidepressant levels were also below detectable levels in the infants of women taking fluoxetine or sertraline only during lactation None of the babies in any category showed toxic effects.
Spigset and collaborators found modest levels of paroxetine in a milk sample from a 39-year-old woman treated with that SSRI. (21) Her baby experienced no adverse effects. These authors note that the daily antidepressant dose taken by a nursing infant would be lower if the mother was taking paroxetine or fluvoxamine (Luvox) rather than fluoxetine, but given the very small amounts ingested, they doubt that the difference would be of clinical significance. Yoshida and others reported detectable concentrations of fluvoxamine in the breast milk of a woman taking this SSRI, but they did not measure concentrations in infant plasma. (22) There were no acute toxic effects of fluvoxamine in the baby, and developmental assessments at 4 and 21 months were normal.
Nulman et al observe that depression in a mother can adversely affect the development of cognition, language, and behavior in infants and young children. (8) When a pregnant woman is depressed, perinatal risks also rise. Current evidence suggests that in most cases, treating depression is better for both mother and offspring than allowing maternal depression to remain untreated or to recur because an antidepressant was withdrawn. Of course, psychotherapy and electroconvulsive therapy are nonpharmacologic options. After a baby is born, the mother can take an antidepressant without exposing her infant, if she elects not to breast-feed. Alternatively, Dr Mammen's group recommends that if a woman needs an antidepressant but wishes to breast-feed, it may be safer to use a drug with a longer record of safety -- namely, a tricyclic antidepressant rather than a newer agent. (19)
Accumulating data suggest that for many antidepressants, use during breast-feeding exposes infants to negligible concentrations and no detectable adverse consequences. On the other hand, there could be idiosyncratic reasons why an occasional neonate might develop unusually high serum drug levels. The decision has to be left to the clinician and the parents, but a baby should be very carefully observed whether the mother is taking the drug in pregnancy or when she is nursing. If there are adverse effects in the baby, the doctor might consider checking the infant's serum antidepressant level. We agree with Wisner et al, (9) who express the need for systematic behavioral assessments and longitudinal developmental follow-up of breast-fed babies and an expanded database on mother and infant serum levels.
* Not available in the United States.
Our thanks to Katherine L. Wisner, M.D., for her review of this article and helpful suggestions.
This topic is indexed under the following subjects:
Dr. Bob is Robert Hsiung, MD, email@example.com
Alan J. Gelenberg, M.D.
Original article copyright 1997 Biological Therapies in Psychiatry.
Web page copyright 1998 Robert Hsiung.