Opioid Use in Breastfeeding

Morphine

Morphine is the prototype opiate. It is ∼25% bioavailable and is metabolized to inactive morphine-3-glucuronide (60%) and to active morphine-6-glucuronide (10%). Morphine-6-glucuronide has an oral bioavailability of ∼4%. Glucuronides excreted into milk can probably be converted back to morphine in the neonate's gut by deglucuronidation and reabsorbed as morphine, although this has not been studied.

In general, morphine in milk is tolerated fairly well by infants, depending on the maternal route of administration. Epidural morphine given to mothers for postcesarean section analgesia results in trivial amounts of morphine in their colostrum and milk. Intravenous (IV) or oral doses of maternal morphine in the immediate postpartum period result in higher milk levels than with epidural morphine. Breastfed newborns of mothers using morphine by IV patient-controlled analgesia (PCA) after cesarean section were more alert and better oriented after postpartum day 3 than infants of mothers using PCA meperidine and nonbreastfed control infants. Both epidural and PCA morphine result in low levels in maternal serum, so are advantageous routes of administration. Some experts feel that low-dose morphine might be preferred over other opioids in nursing mothers. ²

Codeine

Codeine is a prodrug with very weak analgesic activity. It is >50% absorbed and metabolized through cytochrome P450 2D6 (CYP2D6) to morphine (5%), norcodeine (15%), and further to codeine-6-glucuronide (80%) and morphine-6-glucuronide by uridine 5′-diphospho-glucuronosyltransferase 2B7 (UGT2B7). The morphine and codeine-6-glucuronide metabolites are responsible for codeine's analgesic properties. Both CYP2D6 and UGT2B7 are subject to genetic variability, which can alter the amount of active narcotic excreted into breast milk.^11,12

Codeine has become infamous as a cause of death in the breastfed infant of a mother with an ultrarapid CYP2D6 variant, presumably causing increased metabolism of codeine to morphine. These findings have been questioned,^13,14 and the original case report, published simultaneously in Canadian medical and pharmaceutical journals, has recently been retracted by the journals' editors. ¹⁵ This case prompted labeling changes by the U.S. Food and Drug Administration (FDA), the European Medicines Agency, Health Canada, and others. Also, many professional organizations recommend against prescribing codeine to nursing mothers. However, others have pointed out that millions of nursing mothers have safely used codeine during breastfeeding over decades.^2,14 The warnings may have shifted postpartum opioid use to more dangerous opioids such as oxycodone. A countervailing prescribing trend seems to have resulted from the U.S. Drug Enforcement Administration's rescheduling of hydrocodone from schedule III to schedule II, which may be shifting prescribing back to codeine. ²

Regardless of the merits of the single case report of a codeine death, side effects such as drowsiness, apnea, bradycardia, and cyanosis have been reported with codeine from breast milk in other case reports and small studies. Several cases of excessive sedation and apnea in breastfed infants were reported as far back as the early 1980s. More recently, six cases of CNS depression in infants breastfed by mothers taking codeine were reported for a 2-year period in Canada. Infants were all 10 days of age or younger and responded favorably to discontinuation of nursing, discontinuation of maternal codeine, or administration of naloxone. One study compared the frequency of drowsiness in breastfed infants whose mothers took acetaminophen plus codeine with infants whose mothers took acetaminophen alone. Infants exposed to codeine had a 17% frequency of drowsiness compared with 0.5% exposed to acetaminophen.

In a telephone follow-up study, 5 of 26 of breastfeeding mothers taking multiple doses of codeine reported drowsiness in their infants. All infants were younger than 1 month. However, even older infants can experience problems with codeine in milk with a sufficiently high dose. A 2-month-old breastfed infant was noted by the mother to be somnolent and sleeping more than usual for 2–3 days. Under observation at an emergency room, the infant slowly recovered. The baby's mother had taken a total of 4–5 tablets of acetaminophen 500 mg plus codeine 30 mg as needed for back pain for the prior 3 days.

Hydrocodone

Hydrocodone is ∼25% absorbed and metabolized to norhydrocodone by CYP3A4 and to active hydromorphone by CYP2D6. A study that measured hydrocodone and hydromorphone in the breast milk of postpartum mothers found relatively low levels of both drugs. One case report has been published of a breastfed infant who experienced excessive sedation with maternal hydrocodone.

Oxycodone

Oxycodone is one of the most orally bioavailable opiates at 60–87% in adults. It is metabolized to the active metabolites, noroxycodone and oxymorphone. Oxycodone elimination is decreased in young infants and much interindividual variability exists. These factors make it a risky drug to use in nursing mothers, in addition to being one of the most prominent drugs involved in the opioid abuse epidemic in the United States.

Two case reports have been published of infants whose mothers were taking oxycodone and developed severe respiratory depression, requiring emergency care. One neonate was found to be well by his physician at 2 days of age. At 3 days postpartum, the infant began to be sedated, became difficult to arouse and did not feed from either breast. The mother reported that her milk had come in the previous evening, illustrating the importance of the volume of milk in determining the dose of drug that the breastfed infant receives.

In a study of 50 mothers taking oxycodone postcesarean section, their 50 neonates were evaluated for sedation for 48 hours after birth. None was severely sedated and <4% had sedation of 3 on a 1 (fully alert) to 5 (difficult to rouse) scale and none more sedated than 3 on the scale. Because these infants were in the first 3 days postpartum, their oxycodone dose was probably limited by the small volumes of colostrum they were ingesting.

Hydromorphone

Hydromorphone has an oral bioavailability of 62% in adults and is metabolized to inactive metabolites. In one case of a nursing mother receiving oral hydromorphone for pain after a cesarean section, her 6-day-old infant developed intermittent bradycardia and had an apneic event requiring bag-valve-mask intervention. The infant received naloxone and within 30 seconds spontaneous respirations and alertness returned, and heart rate increased to 165 beats per minute. Fifteen minutes later, he had another apneic episode that resolved rapidly with another dose of naloxone.

Dihydrocodeine

Dihydrocodeine is metabolized through CYP2D6 to dihydromorphine, which has a potency similar to morphine. A woman began taking dihydrocodeine drops for cough twice daily beginning on the first day postpartum. One day later, her breastfed infant was difficult to arouse and was not breastfeeding well. The infant had bradycardia, hypoglycemia, and an oxygen saturation of 85%. After 24 hours in the hospital, all symptoms resolved.

Tramadol

In adults, tramadol is 70–100% orally absorbed but undergoes considerable first-pass metabolism to the active O-desmethyltramadol (M1). Tramadol has a more potent monoamine reuptake inhibitory effect whereas M1 has a more potent opioid mu-agonist effect. M1 is ∼10% as potent as morphine in mu-opiate binding. Women who are extensive CYP2D6 metabolizers of tramadol may have higher than expected serum levels of M1, potentially leading to higher levels of M1 in breast milk. Apparently using an analogy with codeine's CYP2D6 metabolism, the FDA recommended against the use of tramadol during breastfeeding. However, with usual maternal dosage, the amount excreted into breast milk is much less than the dose that has been given to newborn infants for analgesia, and no adverse reactions in breastfed infants have been published, so the proscription seems unwarranted. ¹⁶

Meperidine

Meperidine is a synthetic opioid that has an oral bioavailability of ∼50% in adults. It is metabolized to the active metabolite normeperidine, which has half the analgesic activity and at least twice the CNS excitatory activity of meperidine. Newborns have impaired meperidine metabolism and possibly a higher oral bioavailability than adults. Newborns may also experience more CNS depression from normeperidine than adults. Meperidine use during labor has a negative impact on both breastfeeding and the infant. Epidural meperidine PCA postpartum is less problematic than IV, but other opioids are preferred.

Fentanyl

Fentanyl and its derivatives are used primarily in hospitalized patients, including during labor and delivery. As such, there have not been reports of outpatient adverse events in breastfed infants. Fentanyl is often used epidurally during labor and postpartum. Maternal plasma levels are low when it is used during labor or for a short time immediately postpartum. The oral bioavailability of fentanyl is 33% in adults. Amounts of fentanyl ingested orally by the neonate are usually small and are not expected to cause any adverse effects in breastfed infants in a controlled setting.

Alfentanil

Alfentanil is highly protein bound, which should result in less transfer to breast milk than other opioids. In one study, alfentanil colostrum levels in nine women given IV alfentanil during tubal ligation surgery were low. Similar to fentanyl, when used IV or epidurally during labor or for a short time immediately postpartum, amounts of alfentanil ingested by the neonate are small and are not expected to cause any adverse effects in breastfed infants.

Remifentanil

Because the half-life of remifentanil is extremely short, it is unlikely to cause any adverse effects in the breastfed newborn if it is given to the mother for labor analgesia or a surgical procedure. A double-blind randomized study compared PCA analgesia with remifentanil to a continuous meperidine infusion for labor analgesia. Breastfeeding difficulties were experienced in 6.3% of the infants of mothers who received remifentanil and 12.8% of infants whose mothers received meperidine; however, this difference was not statistically significant.

Sufentanil

Controlled use of sufentanil during labor or for a short time immediately postpartum results in only small amounts of sufentanil ingested by the neonate that would not be expected to cause any adverse effects in breastfed infants. But, because of sufentanil's long half-life, milk levels might be expected to increase if used for an extended period postpartum by continuous IV infusion or repeated IV administration.