METHODS

Subjects

Ten healthy, paid volunteers were recruited through advertisements posted in clinic waiting rooms, and advertisements in parent periodicals. Women over the age of 18 years were eligible to participate if they were currently breast feeding at least three times per day and planned to discontinue breast feeding in the next few weeks. We excluded women who: (1) were currently using any psychoactive medications; (2) had current major depression, schizophrenia, or bipolar disorder; and (3) had a history of a seizure disorder, anorexia or bulimia, or intolerance to bupropion.

Protocol

Subjects were studied as outpatients at the clinical study centre at San Francisco General Hospital. An eligible woman was scheduled for an initial study visit at the time that she intended to stop breast feeding. At this visit eligibility was confirmed, informed consent was obtained, a health history was performed and blood pressure, weight and height were measured. At this visit, women were also instructed on the use of a breast pump, and provided with one if they did not already have one available. Women were instructed to stop breast feeding on that day and to begin pumping, and discarding, their breast milk at least three times per day. The following day, women were instructed to initiate bupropion SR 150 mg orally once a day. Women were instructed to increase the dosage to 300 mg a day after three days. Phone contact was maintained with women during this period to facilitate compliance. On the seventh day of the protocol, women came in for a second study visit. At this visit, a single sample of plasma, saliva, urine, and breast milk were each obtained. The initial 3 ml of breast milk was obtained with the use of electric breast pump. Pill bottles were collected to ensure compliance with the medication dosing. We calculated an approximate time since the last dose ingestion by calculating the time elapsed between the start of the clinic visit and the scheduled time of the last pill ingestion.

Analytical chemistry

Measurement of bupropion and its active metabolites, hydroxybupropion, erythrohydrobupropion, and threohydrobupropion were quantitated in breast milk, plasma, saliva and urine using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The method is similar to the one reported by Hsyu et al, but is capable of measuring the erythrohydrobupropion and threohydrobupropion individually.²⁵ Deuterium labelled internal standards, which were synthesised in our laboratory, were used as internal standards. The lower limit of quantitation of the bupropion assay (LLOQ) is 1 ng/ml. Samples were frozen and assayed in one batch at the completion of the study. There was no significant change in the concentration of bupropion and metabolites stored for one month in a freezer at −20°C. Quality control procedures that are appropriate for pharmacokinetic studies were utilised.

Data analysis

An average daily dose of bupropion and each of its active metabolites in breast milk was calculated. An average infant milk consumption of 0.15 l/kg/day was assumed.²⁶ Since bupropion metabolites are pharmacologically active,²⁴ the exposure of the infant to the sum of bupropion and its metabolites was estimated, and compared (on a molar basis) to the typical maternal weight−normalised exposure to bupropion. The Pearson correlation coefficients (r) was used to calculate a coefficient of determination (r²) to compare the plasma, saliva, and urine concentrations of bupropion with breast milk concentration to determine whether saliva or urine measurement could be used (instead of plasma) to monitor bupropion concentrations in breast milk.

RESULTS

The median age of the 10 women who participated in this protocol was 29 years (range 22–37 years). Their median weight was 59.4 kg (range 54.1–95.5) and their median body mass index was 25.3 kg/m² (range 20.9–38.4). At the time of participation, the average age of the infants of these women was 12.5 months. Half of the women were taking an oral contraceptive pill. No other medications were being used by any of the participants. The median number of hours since the last dose of bupropion was 2.5 hours (range 1–12 hours). The concentrations of bupropion and its active metabolites in breast milk, plasma, urine, and saliva are shown in table 1. The average (SD) breast milk to plasma (M/P) bupropion ratio was 2.8 (3.7). The average M/P ratios of hydroxybupropion, erythrohydrobupropion, and threohydrobupropion were 0.1 (0.03), 0.9 (0.3), and 1.2 (0.3), respectively. Since on average an infant drinks 0.15 l/kg/day,²⁶ the calculated dosage of bupropion in breast milk is 6.75 μg/kg/day (table 2). The estimated daily infant dosages of the metabolites were: hydroxybupropion 15.75 μg/kg/day, erythrohydrobupropion 10.80 μg/kg/day, and threohydrobupropion 68.85 μg/kg/day.

The concentration of bupropion and its active metabolites in breast milk was not associated with age, body mass index, age of infant, use of oral contraceptive pills, or the frequency of breast feeding at the time the study was initiated.

The coefficient of determination between the concentration of bupropion in breast milk and in urine was 0.77 (p < 0.01) (table 3). There was no significant association between saliva or plasma bupropion concentrations and the concentration of bupropion in breast milk, although there were significant correlations between plasma and saliva bupropion. No side effects were noted by any of the mothers.

DISCUSSION

Bupropion and its active metabolites are present in the breast milk of lactating women. These results indicate that the daily dose of bupropion and metabolites that would be delivered to an infant of a woman taking a therapeutic dose of bupropion is small. The estimated dose of bupropion taken up by the infant milk was less than 0.2 % of the weight normalised dose of an adult woman. The major metabolites of bupropion are active, producing equal or weaker inhibition of neurotransmitter uptake compared to bupropion.²⁴ Since the metabolites of bupropion have pharmacologic activity, the exposures of the metabolites should be considered as part of the infant drug exposure as well. If one conservatively assumes equimolar pharmacologic potency of the hydroxybupropion, erythrohydrobupropion, and threohydrobupropion metabolites to the parent bupropion, and assumes complete oral bioavailability in the infant, the total exposure to bupropion and metabolites would be 2% of the weight normalised maternal bupropion dose. This dose in the infant is lower than 10% of maternal dose that has been used as a guideline for drug safety during breast feeding.²⁶ The failure to detect bupropion in three other breast fed infants whose mothers were taking bupropion in doses of 150–300 mg daily doses supports the idea that infants are exposed to a very low concentration of the drug.^18,19

The coefficients of determination for breast milk with urine, saliva, and plasma concentrations of bupropion and its active metabolites were not high (with the exception of breast milk and urine bupropion concentrations). This suggests that future studies should not rely on these other body fluids as surrogate markers for breast milk.

The concentration of bupropion and its metabolites were measured in breast milk and in other body fluids at one point in time. Sampling at multiple times was not performed. Although the half life of bupropion is relatively long, there is an absorption peak after ingestion of the sustained release preparation, peaking at 1–4 hours.²⁷ Most of our subjects had blood and milk samples taken around the time of the peak. Therefore our estimates of exposure to bupropion are likely to be high; the actual exposure to bupropion would be even less than what we have estimated. Since the metabolites have longer half lives than bupropion and are generated slowly from bupropion, these concentrations are expected to be at or near steady state. Thus, the estimates of exposure to metabolites based on a single sample should be representative of the exposure over the day.

The women who participated had been breast feeding for an average of 12.5 months. It is possible that the milk composition after this length of breast feeding may be different than that in the immediate post-partum period. Although our results suggest that bupropion taken by a breast feeding mother should not present a concern for most infants, exposure may be greater for some infants—for example, premature infants.

Medications are typically avoided in lactating women because of the potential exposure of their infants. Few studies have quantified the true exposure of infants to medications consumed by their breast feeding mothers.^14–17 This practice leads women and their physicians to avoid medications, perhaps unnecessarily, or encourages women not to breast feed.

Pregnant women have been extensively targeted for interventions to reduce prenatal tobacco use, primarily to protect the health of their unborn child. These efforts have been successful. Women are three times more likely to quit smoking during their first pregnancy and the year preceding this pregnancy than at any other time.⁴ Estimates suggest that approximately 30–50% of women who smoke spontaneously quit either immediately before or during pregnancy.^5,28,29 Prenatal educational and behavioural interventions have been shown to increase cessation rates during pregnancy up to 50%.^29–32 However, despite a high likelihood of quitting around the time of pregnancy, many women relapse following delivery. Several studies have demonstrated high rates of smoking relapse post-partum, ranging from 40–90% within six months.^5–8,28 Bupropion has been shown to delay relapse to smoking in non-pregnant adults.²² Bupropion is generally safe and well tolerated, with relatively rare adverse effect.²¹ Our results suggest that bupropion taken by a breast feeding mother should not present a concern for most infants. Particularly given the lack of success of behavioural therapies in promoting post-partum relapse,^11,33–37 the effectiveness of bupropion to prevent post-partum smoking relapse should be evaluated in clinical trials without excluding women who are or may be breast feeding. During such studies, however, it would be advisable to measure bupropion and metabolites in infants, to confirm that the exposure is as low as expected.