The Pregnancy-Breastfeeding Interface

Monoclonal antibodies

One extreme example of drug persistence is that of the immunosuppressant monoclonal antibodies such as adalimumab, infliximab, and rituximab that are used in rheumatoid arthritis and inflammatory bowel disease. Many of them are structurally similar to IgG and can be actively transported across the placenta beginning in the second trimester and increasingly in the third trimester. If the infant is born with one of these drugs in the bloodstream, the drug can persist for months because of their long half-lives and infants' immature clearance mechanisms. Although clinical data are sparse, several adverse reactions to BCG vaccine in such exposed infants have been reported and a few have had a suboptimal immune response to live vaccines. ¹ In infants whose mothers were exposed to these monoclonal antibodies in pregnancy, drug labeling and expert opinion recommend that the infant not receive live or live-attenuated vaccines for at least the first 6 months of life or until the drug is undetectable in the infant's bloodstream. Possible exceptions are certolizumab pegol, an antibody fragment, and etanercept, a fusion protein, which are not subject to active placental transport and, therefore, have very low neonatal blood levels.^2,3

Although the optimal management of pregnant women requiring one of these drugs is patient dependent, it is generally thought that stopping the drug in the second trimester is desirable, usually around weeks 30 to 32, with the possible exception of the poorly transported drugs, which probably do not need to be discontinued. ⁴ However, a recent study in patients with inflammatory bowel disease in remission found that stopping adalimumab or infliximab before week 25 of pregnancy did not result in an increased relapse rate compared to continuing the drug beyond 30 weeks. ⁵ All infants had excellent 1-year outcomes, but those in the delayed discontinuation group breastfed less (31%) than those in the early stop group (86%). However, this difference was not necessarily a drug effect.

Use of monoclonal antibodies during breastfeeding is not of concern because the amounts in breast milk are very low and the infant receives the drug orally, where bioavailability is negligible.^1,6 In addition, case reports and clinical trials have found no evidence of harm to exposed infants.

Fluoxetine

Pharmacokinetic differences exist among the selective serotonin reuptake inhibitor (SSRI) antidepressants. The half-lives of sertraline, paroxetine, citalopram, and escitalopram are in the range of 1 to 1.5 days in adults, and probably longer in neonates. In sharp contrast, the half-lives of fluoxetine and its equally active metabolite are about 5 and 9 days, respectively, in adults. Using the standard five half-lives for near-complete elimination of the metabolite, infants exposed to fluoxetine in utero near term will have detectable serum concentrations of the drug and metabolite for over a month. One implication of this persistence is that additive drug effects and drug interaction can occur in the neonate, even if the mother's drug has been discontinued. This could occur if fluoxetine is switched to a different SSRI near term or postpartum. In this case, the infant would be exposed to multiple drugs simultaneously, rather than just fluoxetine and its metabolite, potentially increasing the risk of an adverse effect. Generally, it is recommended to continue an antidepressant that is working well for the patient and to monitor the infant for side effects, rather than to change medication because of pregnancy or lactation.

In one report, neonatal symptoms were apparently caused by drug-induced serotonin syndrome following in utero fluoxetine exposure. This implies that sufficient fluoxetine was present in the infant to inhibit serotonin reuptake, resulting in an excess of serotonin that caused the symptoms. Some symptoms of serotonin syndrome were similar to withdrawal effects (see the Withdrawal section below), complicating the diagnosis. Discontinuation of breastfeeding was helpful in this case. ⁷ More study is required to clarify the difference between serotonin syndrome and drug withdrawal in breastfed infants.

Benzodiazepines

Long-acting drugs in this class, such as chlordiazepoxide, clonazepam, and diazepam, given near term can persist in the infant postpartum and cause sedation, hypotonia, and other symptoms of central nervous system depression. Short-acting alternatives such as oxazepam or lorazepam are preferable during pregnancy and breastfeeding, although avoidance of benzodiazepines altogether is probably best, if possible.

Cannabis

The psychoactive component of cannabis, tetrahydrocannabinol (THC), is very fat soluble and accumulates in body fat with prolonged use. This could theoretically result in continued excretion into breast milk in women who use cannabis during pregnancy, which in turn could cause the breastfed infant's urine to be positive for THC exposure. Positive results might have legal consequences in some jurisdictions, but the clinical importance of such findings is unknown. The increasingly legal status of cannabis at the state level and greater use of the drug mean that more information on cannabis and THC during pregnancy and breastfeeding is sorely needed.

Cancer chemotherapy

Discovery of cancer during pregnancy presents difficult challenges to the mother and medical team. Nevertheless, a number of normal infants have been born to women who underwent chemotherapy in the third trimester. ⁸ With some long-acting chemotherapeutic drugs, excretion into the breast milk postpartum could occur if they are given too close to delivery. Examples of these drugs are doxorubicin, platinum analogs (e.g., cisplatin), mitoxantrone, paclitaxel, high-dose methotrexate, vinca alkaloids (e.g., vincristine), and retinoids (e.g., oral tretinoin).

Antidepressants

Both the tricyclic (e.g., amitriptyline and clomipramine) and serotonin reuptake inhibitors (e.g., sertraline and venlafaxine) have resulted in withdrawal symptoms in neonates after in utero exposure. An extensive review of published literature and FDA data on the serotonin reuptake inhibitors found that both breastfed and nonbreastfed infants can be affected. ⁹ The most common symptoms, which last for 2 to 4 weeks, resemble colic and include irritability, hypertonia, jitteriness, trouble feeding, tremor, and agitation. In the case of an infant who is breastfeeding, these symptoms could easily be attributed to the drug in breast milk and breastfeeding terminated. Continuing breastfeeding may provide a slow tapering of drug dosage and reduce symptoms. Nevertheless, nondrug treatments (e.g., skin-to-skin contact, reduced stimulation, and demand feeding) appear to be effective in alleviating this relatively short-term syndrome. ⁹

Anticonvulsants

Use of certain older sedating anticonvulsants such as phenobarbital and primidone, as well as benzodiazepines, during pregnancy can also cause neonatal withdrawal symptoms, which may require drug treatment of the infant.

Withdrawal mitigation

In some reported cases, phenobarbital, primidone, and alprazolam seem to have been excreted in milk in sufficient amounts to mitigate neonatal abstinence. Similar cases have been reported with opiates, such as methadone. In these cases, abrupt discontinuation of breastfeeding appears to have precipitated or worsened withdrawal symptoms. In one unusual case, a mother who was using heroin clandestinely continued to breastfeed her daughter until she was 8 years old to prevent the child's withdrawal symptoms and discovery of her own heroin use. ¹⁰

In summary, drug withdrawal is usually not reason to avoid breastfeeding in infants exposed in utero to sedatives or opiates. Additional drug therapy might be required to mitigate abstinence symptoms, but abrupt discontinuation of breastfeeding should be avoided.