June 15, 2004
June 15, 2004 from ObGynNews By Lee S. Cohen, M.D.
In April, the National Toxicology Program’s Center for the Evaluation of Risks to Human Reproduction, established by the NTP and the National Institute of Environmental Health Sciences, issued a final report on the reproductive and developmental toxicity of fluoxetine. The report concluded that “third-trimester exposure to therapeutic doses of fluoxetine … is associated with an increased incidence of poor neonatal adaptation,” which includes jitteriness, tachypnea, poor tone, and other symptoms, “as well as increased admissions to special care nurseries.”
Having reviewed the report in draft and final form and having testified at the meeting of the expert panel convened to write the report, my greatest concern is what patients and some clinicians may do with the panel’s conclusions. Information in the report, while comprehensive and technically correct in most cases, might easily be misconstrued by women and their families.
The report provides a summary and review of existing data, with a thorough review of the animal and human literature on reproductive safety of fluoxetine. It does not adequately address the clinical context in which fluoxetine or other selective serotonin reuptake inhibitors (SSRIs) are used. While this may not be the aim of the project, failure to address this issue limits the report’s value with respect to its ability to inform clinical care; the absence of a clinical context with which to interpret the report may lead to incorrect conclusions and clinical treatment decisions, putting women at risk for the sequelae of untreated or relapsing depressive illness.
The report criticizes much of the literature regarding reproductive safety of fluoxetine, which is understandable because controlled studies of exposures to any medication during pregnancy are not done for ethical reasons. Conclusions regarding reproductive safety of medications come from various sources, such as case series, postmarketing surveillance registries, and teratovigilance programs. These sources can sometimes provide large enough numbers of drug exposures to allow for useful conclusions regarding reproductive safety.
The panel’s conclusions regarding the risk for major congenital malformations associated with prenatal exposure to fluoxetine are consistent with the literature and suggest an absence of increased risk with first-trimester exposure to the medicine. The report also addresses the risk for “perinatal toxicity,” which typically includes symptoms of jitteriness and autonomic reactivity in the newborn.
Enough literature has accumulated suggesting that third-trimester exposure to SSRIs may be tied to an increased risk of transient symptoms as noted above. Most reports have not associated such exposure with adverse longer-term sequelae. Fluoxetine is the only SSRI for which we have long-term neurobehavioral data, including follow-up of exposed children through ages 4-7 years. No differences in long-term neurobehavioral outcome between exposed and unexposed children were noted.
One of the greatest failures of the NTP report is that an important confounding factor with regard to outcome of SSRI use in pregnancy is neglected: maternal mood. In the recent literature, one can find the same “toxicity,” such as lower Apgar scores or obstetric complications, in children of mothers who have untreated depression during pregnancy. Failure to address this adequately in the report is a significant omission.
Fluoxetine is used to treat a serious illness; it is not a potential environmental toxin, such as those reviewed by other NTP panels. The report does not indicate that decisions about whether to use fluoxetine during pregnancy are clinical choices made by patients in the context of some risk-benefit analysis made collaboratively between the patient, her family, and the physician. My colleagues and I have described high rates of relapse in women with a history of recurrent major depression who discontinue antidepressants in pregnancy. Depression during pregnancy is associated with compromised fetal and neonatal outcomes-risks that are not reflected in the report. Discontinuation of medication near the end of pregnancy appears to increase the risk for postpartum depression.
The panel notes in the report that it recognizes that any risks of fluoxetine need to be weighed against the risks of
untreated disease. But this brief statement embedded in a lengthy document that describes fluoxetine as “a reproductive toxin” is inadequate. One has to wonder how this report will impact what actually goes on as patients make decisions about using these compounds.
The full NTP report is available at: http://www.ncbi.nlm.nih.gov/pubmed/15334524.
DR. LEE COHEN directs the perinatal psychiatry program at Massachusetts General Hospital, Boston. He is a consultant to fluoxetine manufacturer Eli Lilly, and to manufacturers of other SSRIs.