The longstanding view that maternal use of sedatives and opioids were safe in breastfeeding changed dramatically after we published a case report of an 11-day-old healthy full-term baby who had died, without postmortem anatomical findings. The baby had a high blood morphine level, and genetic testing determined that the baby’s mother, who had been taking codeine and acetaminophen for episiotomy pain for 2 weeks, was an ultrarapid metabolizer of codeine, as was her own mother, due to a polymorphism of cytochrome P450 2D6 (CYP2D6). A very high level of morphine was also detected in breast milk she had stored (Lancet 2006;368:704).
Dr. Gideon Koren
CYP2D6 metabolizes codeine to morphine, and individuals with the CYP2D6 ultrarapid metabolizer genotype, are also more likely to experience codeine-associated adverse effects, as did this baby’s mother.
In August 2007, a year after the case report was published, the Food and Drug Administration issued an alert, based on this report, warning about the potential increased risk for morphine overdose in nursing infants, whose mothers are taking codeine and are ultrarapid metabolizers of codeine, with the recommendation that manufacturers add this information to the label of their products.
Previous reports describing ultrarapid codeine metabolizers are scant. About 6 months before the coroner contacted me about the baby, an ultrametabolizer adult with pneumonia and renal failure developed respiratory depression, went into a coma, and nearly died during treatment with small codeine doses (N. Engl. J. Med. 2004;351:2827-31).
We followed up the case report with a search of the literature, and identified 35 reports of breastfeeding infants with unexplained symptoms of drowsiness, apnea, bradycardia, and cyanosis, described in three abstracts and two studies, which we attributed to codeine-morphine exposure through breast milk (Paediatr. Drugs 2008;10:399-404). These include an abstract published in 1982 describing four premature babies at Children’s Hospital of Philadelphia with severe apnea that resolved after their mothers, who were breastfeeding, stopped taking codeine.
In another study, we followed 72 women who had called Motherisk with questions about the safety of codeine during breastfeeding and found that about 23% reported symptoms of central nervous system (CNS) depression in their babies while breastfeeding while they were taking codeine, indicating that this adverse event is more common than we had thought (Clin. Pharmacol. Ther. 2009;85:31-5).
We then evaluated whether certain strategies could reduce this rate in a group of 300 women treated with codeine for postpartum pain while breastfeeding at a Toronto birthing unit. The rate of CNS depression in the babies was reduced to 1% – a result of recommendations that mothers not take codeine for more than 4 days (about the time it apparently took morphine levels to accumulate in our case report and in the Children’s Hospital of Philadelphia report), and that babies exhibiting symptoms such as grogginess, not feeding well, not latching on, not waking up for feeding, or appearing limp be immediately examined by a physician.
As a comparison, in a recent study that was submitted for publication, we found that benzodiazepines were not associated with problems in the baby during breastfeeding. If benzodiazepines cross the blood brain barrier of the baby, they do so in small amounts.
Because of these reports, some clinicians have switched from prescribing codeine to oxycodone for postpartum pain, but oxycodone undergoes the same metabolic pattern as codeine. In a follow-up study, we showed that oxycodone, which is metabolized to oxymorphone, is not a safer option: We compared 533 breastfeeding mothers who were taking codeine, oxycodone, or acetaminophen for postpartum pain, and their infants, and found the rates of CNS depression in the infants was similar (20%) among those whose mothers took oxycodone and among those whose mothers took codeine (nearly 17%). In comparison, only one baby (less than 1%) whose mother was taking acetaminophen had such symptoms (J. Pediatr. 2012;160:33-7.e2).
Codeine continues to be widely used as a treatment for postpartum pain, often in combination with acetaminophen, and while it may not be safe for all women and babies during breastfeeding, it should not be contraindicated during breastfeeding. A genetic test for determining CYP2D6 genotype is available, but is expensive and is currently not routine for screening. The ultrarapid metabolizer genotype is more common in some populations: 29% in Ethiopia, 10% in Greece and Portugal, compared with 1% in Finland and Denmark. In Caucasians overall, the range is 1%-10% and in the black population, about 3%, according to the FDA.
Based on the evidence now available, symptoms of CNS depression in a nursing baby or mother who is taking codeine should prompt examination of the baby. If the mother is known to be an ultrametabolizer, she should not receive codeine. These are among the recommendations in the Motherisk guidelines for maternal codeine use during breastfeeding (Can. Fam. Phys. 2009;55:1077-8).