Abstract
Nursing mothers may have to undergo some form of general anesthesia and may be concerned about the effect of the anesthetic on their infant. This column reviews information on general anesthetics on breastfeeding and the breastfed infant. The reader is referred to more complete guidelines for details on performing surgical1,2 or obstetric anesthesia.3,4 Opioids are the drugs of most concern, but will be considered separately in a forthcoming column. More detailed information and references on the use of specific drugs, when available, can be found in the corresponding LactMed® records.
Anesthesia's Effects on Lactation
General anesthesia might interfere with lactation, reducing milk volume postoperatively because of perioperative volume losses and fluid restriction, as well as stress-induced inhibition of milk production. 5 This decrease was demonstrated in a small study of mothers undergoing general anesthesia with midazolam, propofol, and fentanyl who had a much lower milk volume than expected at 5 hours after surgery and to a lesser extent in the cumulative postoperative 24-hour milk volume. 6 A separate randomized study compared the effects of cesarean section using general anesthesia, spinal anesthesia, or epidural anesthesia with normal vaginal delivery on serum prolactin and oxytocin as well as time to initiation of lactation. General anesthesia was performed using propofol and rocuronium for induction, followed by sevoflurane and rocuronium as needed. Fentanyl was administered after delivery in the general anesthesia group. Patients in the general anesthesia group had higher postprocedure prolactin levels and a longer mean time to lactation initiation (25 hours) than in the spinal anesthesia (10.8 hours), epidural anesthesia (11.8 hours) and normal vaginal delivery (10.9 hours) groups. Opioids, including fentanyl, can increase serum prolactin, which might explain the higher prolactin levels after general anesthesia.7–9 Conversely, postpartum oxytocin levels in the nonmedicated vaginal delivery group were higher than in the general and spinal anesthesia groups. 10 This situation is complicated by the fact that intrapartum oxytocin use can decrease maternal postpartum oxytocin excretion.11,12
A randomized, but nonblinded, study in women undergoing cesarean section compared epidural anesthesia with bupivacaine to general anesthesia with intravenous thiopental and succinylcholine for induction followed by nitrous oxide and isoflurane. The time to the first breastfeed was significantly shorter (107 versus 228 minutes) with the epidural anesthesia than with general anesthesia. This difference was probably caused by the anesthetic's effects on the infant, because the Apgar and neurologic and adaptive scores were significantly lower in the general anesthesia group of infants. 13
Two retrospective studies found that mothers who received a nitrous oxide–oxygen mixture during labor for analgesia had higher breastfeeding rates at 2 days, 7 days, 1 month, and 3 months postpartum than women who received no nitrous oxide during labor. A double-blind randomized study and a retrospective study of dexmedetomidine had similar findings. Dexmedetomidine given during and postoperatively after cesarean section deliveries had shorter times to the first milk production, achieved exclusive breastfeeding sooner (8 versus 11 days), and had a greater amount of milk on the second day postpartum. The differences are probably attributable to better analgesia rather than any direct effect of the drugs on milk production.
Lactation's Effect on Anesthesia
In one study of mothers who were 3–7 days postpartum undergoing tubectomy, breastfeeding before general anesthesia induction reduced the requirements of sevoflurane and propofol to maintain a predetermined level of anesthesia compared with the requirements of nursing mothers whose breastfeeding was withheld or non-nursing women. 14 The authors attributed the difference to the higher levels of oxytocin in the serum of women who breastfed, with oxytocin providing an anxiolytic effect. Breastfeeding immediately before receiving anesthesia has since been incorporated into some perioperative lactation management guidelines.1,15
Anesthetics in Milk
Inhalational anesthetics
No measurements of concentrations in breast milk have been made of any currently used inhalational anesthetics, such as desflurane, isoflurane, or sevoflurane. Foremilk samples taken on two occasions from a practicing anesthesiologist after administering halothane anesthesia in the operating room contained from 630 mcg/L to 2.1 mg/L (0.63–2.1 ppm) of halothane, which is no longer available in the United States. These values were higher than air samples from the operating room but may be underestimates because of sampling techniques. All of the currently used gases have very short serum half-lives, so any that reaches breast milk should rapidly diffuse back into the maternal serum and be eliminated after cessation of anesthesia. This model was supported in a study of xenon as a general anesthetic. It could not be detected in milk at any time.
Injectable anesthetics
Several drugs are used by injection during anesthesia. Midazolam can be used to facilitate intubation. The small amounts of midazolam excreted into breast milk would not be expected to cause adverse effects in most breastfed infants. Midazolam is 60–70% metabolized to 1-hydroxymidazolam and 5% metabolized to 4-hydroxymidazolam, which are about equipotent to midazolam. The half-life of 1-hydroxymidazolam is ∼12 hours in adults and can accumulate with prolonged or repeated doses or in renal impairment. This effect has been seen during prolonged sedation in the intensive care unit, resulting in very slow awakening of the patient unlike the short duration of action of single doses. Two expert panels advocate waiting for at least 4 hours after a single intravenous dose of midazolam (e.g., for endoscopy) before resuming nursing. However, there is probably no need to wait to resume breastfeeding, except perhaps with a newborn or preterm infant when a cautious approach would be to wait 6–8 hours before resuming nursing.
Diazepam is a long-acting benzodiazepine and also has a long-acting metabolite. Cases of infant sedation have been reported in mothers who were taking diazepam, and in a longitudinal study some mothers who were taking diazepam reported discontinuing breastfeeding because of drowsiness in their breastfed infants. Infant drowsiness has not been reported after a single dose of diazepam.
Propofol is often used in general anesthesia by continuous infusion and has been measured in the milk of nursing mothers. Amounts in milk are very small and are not expected to be absorbed orally by the infant. Although one expert panel recommends withholding nursing for an unspecified time after propofol administration, most recommend that breastfeeding can be resumed as soon as the mother has recovered sufficiently from general anesthesia to nurse and that discarding milk is unnecessary. Interestingly, several case reports have noted blue–green or green discoloration of breast milk in mothers who received propofol.
Data from 14 women undergoing cesarean section deliveries with dexmedetomidine indicate that very small amounts of dexmedetomidine are excreted into breast milk over the 6 hours after the end of an infusion. The drug is absent from breast milk by 24 hours. Dexmedetomidine is not expected to cause adverse effects in breastfed infants or neonates, although data are lacking.
Minimal information is available on the use of etomidate during breastfeeding, but amounts of etomidate in milk were very small in 20 women who received etomidate for induction. Milk levels decreased rapidly after ending the infusion and no waiting period appears to be required before resuming breastfeeding after etomidate use in anesthesia.
No information is available on the amount of ketamine in milk. Most sources recommend use of only a low dose with careful monitoring for sedation or a different agent until more data become available.1,2
Barbiturates are sometimes used for anesthesia induction. Nine women who were at least 1 month postpartum received between 120 and 150 mg of methohexital intravenously for induction of general anesthesia for bilateral tubal ligation. The authors estimated that the typical breastfed infant would receive a maximum single dose of 0.04 mg of methohexital in a 100 mL feeding 1 hour after the dose or between 0.1% and 0.8% of the maternal weight-adjusted dosage. Likewise, amounts of thiopental in milk are very small. Seven mothers received intravenous thiopental for anesthesia induction before cesarean section. The four breastfed infants excreted an average of 0.0067% of the maternal dose in their urine while the three nonbreastfed infants excreted an average of 0.008% in their urine, indicating that the transfer was placental rather than through milk. The half-life of thiopental in the infants was not different in the two groups. These findings indicate that trivial amounts of thiopental are received in breast milk by infants in the first 2 days of life after administration to their mothers during delivery.
Neuromuscular Blocking and Reversal Agents
Neuromuscular blocking agents are used for muscular relaxation during surgery. No information is available on the use of depolarizing (succinylcholine) or most nondepolarizing (atracurium, cisatracurium, pancuronium, and vecuronium) agents during breastfeeding. One study found no adverse effects in breastfed infants of mothers who received rocuronium to facilitate intubation at the start of surgery. All neuromuscular blockers are highly polar and poorly absorbed orally, so they are not likely to reach the breast milk in high concentration or to reach the bloodstream of the infant. Atracurium and cisatracurium are especially short acting because they are hydrolyzed rapidly in plasma.
The traditional drugs used to reverse neuromuscular blockade are the cholinesterase inhibitors, neostigmine and physostigmine. An old article on mothers taking neostigmine for myasthenia gravis found no detectable neostigmine in the milk of six mothers, although one infant seemed to have abdominal cramps after each nursing. Single doses of neostigmine to reverse neuromuscular blockade are unlikely to adversely affect the breastfed infant more than transiently. No information is available on physostigmine and nursing.
Sugammadex is a newer reversal agent with a different mechanism of action from the cholinergic drugs. No information is available on the use of sugammadex during breastfeeding. Because sugammadex is a large highly polar molecule with a molecular weight of 2,002 Da., the amount in milk is likely to be very low and oral absorption by the infant is unlikely.
Summary
Information on general anesthesia and breastfeeding is incomplete, and often of poor quality, but some generalizations can be made. General anesthesia might delay lactation postoperatively somewhat, but nursing right before anesthesia might reduce anesthetic requirements. Most drugs currently used in general anesthesia do not adversely affect the infant, although opioids are drugs of concern and prolonged use of benzodiazepines in mothers could cause sedation in their breastfed infants. Mothers of healthy full-term infants can resume nursing as soon as they are awake, stable, and alert. A delay of a few hours might be advisable in mothers of preterm or sick infants. Milk expression is not necessary during this time except as needed for maternal comfort.
Footnotes
Disclosure Statement
No competing financial interests exist.
Funding Information
No funding support was provided for this study.