Infant Exposure to Methylphenidate and Duloxetine During Lactation

Detailed analytic method of duloxetine and methylphenidate in breast milk

The aim of the method was to quantify simultaneously methylphenidate and duloxetine in human breast milk using a selective and sensitive liquid chromatography coupled to electrospray ionization tandem mass spectrometry (LC-ESI-MS/MS). Briefly, the medications were extracted from milk (80 μL) after the precipitation of milk proteins with a mixture of acetonitrile and methanol containing deuterated internal standards. After centrifugation, the supernatant was diluted with water containing formic acid. Extracted samples (3 μL) were analyzed by LC-ESI-MS/MS in positive ion mode.

The analytes were separated on a nonporous C18 reversed-phase LC column (Phenomenex Kinetex XB-C18, 30 × 3 mm, 2.6 μm with an Ultra C18 guard column) in gradient elution mode using a flow rate of 0.7 mL/min. The mobile phase A consisted of 0.1% formic acid and 5 mM ammonium formate in water, and mobile phase B consisted of 0.1% formic acid in 95/5 in methanol/water. The gradient cycle consisted of a linear increase in mobile phase B from 5% to 100% over a period of 2 minutes, and kept at 100% for 2 minutes, before being equilibrated with 5% mobile phase B for 2 minutes. The total run time for each injection was 6 minutes.

LC-ESI-MS/MS system comprised of an Agilent 1100 LC system coupled to a triple quadrupole QTRAP^® 4000 System (AB SCIEX) fitted with a Turbo V ionization source interface, which was operated in the positive ion mode. Quantification was performed using the multiple reaction monitoring mode with the following transitions (80 ms dwell time): m/z 234 → 84, collision energy of 29 (methylphenidate), and m/z 298 → 44, collision energy of 35 (duloxetine). The source-dependent parameters were as follows: curtain gas (CUR), nitrogen at 25 L/min; collision gas (CAD), medium; ionspray voltage (IS), 5,500 eV; temperature (TEM), 650°C; ion source gas 1 (GS 1) and gas 2 (GS 2) flow rate, 50 and 60 L/min. The LC-MS/MS system was controlled by the Analyst^® software version 1.6.2.

The bioanalytical method was first developed then validated before its use to determine the concentrations of both medications in breast milk samples provided by both cases presented in this article. Accuracy, precision, selectivity, extraction recovery, matrix effect, benchtop stability, and freeze/thaw stability were tested during method validation. The calibration curve and QC samples were prepared in blank breast milk that contained neither duloxetine nor methylphenidate.

The LC-MS/MS method covered a range from 0.2 to 100 ng/mL for methylphenidate and from 1 to 800 ng/mL for duloxetine. Calibration curves were plotted using peak area ratios of analyte/deuterated internal standard versus nominal analyte concentration, using a weighted 1/x quadratic regression fit.

Relative infant dose

We calculated the RID by multiplying the amount found in breast milk to the average ingestion of milk per day by a healthy infant (150 mL/kg/day) then divided by the mother dose in mg/kg/day. As we could not find the mother's actual postpartum weight, we used a standard weight of 60 kg. Then we multiplied it by 100 to obtain a percentage.

Case 1

The patient was a 30-year-old woman (gravida 4, parity 2, abortus 1) who was a light smoker (2–3 cigarettes/day) and had a history of ADHD, generalized anxiety disorder, borderline personality disorder, and depression with multiple suicide attempts at a young age. She was treated with duloxetine 90 mg daily and an extended-release formulation of methylphenidate 36 mg daily. These two medications were continued while pregnant.

During pregnancy, the patient developed gravidic cholestasis, hypothyroidism, and hypertension with a probable preeclampsia diagnosed at 36 weeks of pregnancy. She presented an elevated blood pressure for 3 consecutive days before admission to the hospital. She remained normotensive during her hospitalization, but presented adverse symptoms, such as headaches, epigastric pain, hand and feet edema, and proteinuria. In accordance with this clinical presentation, her physicians settled for a diagnosis of probable preeclampsia.

While pregnant, the patient also took insulin, levothyroxine, omeprazole, ursodiol, ferrous sulfate, sodium docusate, and cyclobenzaprine. While breastfeeding, ursodiol and insulin were stopped and levothyroxine was tapered after delivery. The patient also started domperidone 20 mg two to four times a day to increase her milk production.

The fetus was found to have congenital pulmonary airway malformation (CPAM) during the second trimester, consisting of small uniform cysts in the right inferior pulmonary lobe. At 37 weeks +1/7, the patient was induced due to her cholestasis and the fetus' CPAM. She gave birth vaginally to a 3,164 g and 48 cm female infant (60th percentile for weight and 50th for height). The APGAR scores were 9 at 1, 5, and 10 minutes. The newborn was not exclusively breastfed, as bottle feeding was less exhausting for the baby.

The milk sample was collected on day 29 postpartum (week 4), 6.5 hours after the patient took 90 mg of duloxetine and 36 mg of an extended-release form of methylphenidate. The concentration found in breast milk was 32.8 ng/mL of duloxetine and 7.9 ng/mL of methylphenidate, which corresponds to a calculated RID of 0.3% and 0.2%, respectively (cf. Table 1). The patient stopped breastfeeding a few days after the milk sample was taken as breastfeeding was challenging considering the baby's poor sucking and the mother's low milk production. At the 6-month follow-up, the child's development was considered normal, except for recurrent pneumonia caused by the pulmonary cysts. The mother's condition was stable. She reported no symptoms of depression.

Case 2

The patient was a 41-year-old woman (gravida 2, parity 0, abortus 1), who used to smoke 1 cigarette/day and quit smoking before pregnancy. She was diagnosed with endometriosis and depression 1 and 2 years before pregnancy, respectively. Her regular medications included duloxetine (60 mg daily), tramadol, acemetacin (NSAID), ibuprofen, and esomeprazole. At 16 weeks of gestational age, when the mother found that she was pregnant, she stopped her medications, but continued duloxetine throughout pregnancy and lactation. The patient developed gestational diabetes during her pregnancy, which required insulin therapy. She had a classic C-section at 38 weeks due to a suspected fetopelvic disproportion. She gave birth to a 3,530 g and 53 cm healthy female infant (80th percentile for weight and 97th for height). The APGAR scores at 1, 5, and 10 minutes were 1, 6, and 8, respectively. There were no malformations at birth or any other neonatal complication, and the infant did not present withdrawal symptoms to duloxetine. The breastfeeding was nonexclusive, but no information on the daily amount of milk intake was available. At the 6-week follow-up, no adverse events were observed in the exposed infant. The mother reported no symptoms of depression.

A cord sample (16.1 ng/mL) and a maternal plasma sample (39.7 ng/mL) were collected at birth, corresponding to a cord to maternal plasma (C/M) ratio of 0.4 (Table 2). At day 6 postpartum, a sample of foremilk and of hindmilk was collected 21.6 and 23.3 hours postdose, respectively. The concentration was 23.6 ng/mL of duloxetine in the foremilk and 14.3 ng/mL in the hindmilk, which corresponds to a RID of 0.4% and 0.2%, respectively, for an exclusively breastfed infant (cf. Table 1). A sample of maternal plasma was collected 21.5 hours postdose, and the concentration found was 40.1 ng/mL, respectively. At week 6 postpartum, an additional paired sample of foremilk and of hindmilk was collected 5.2 and 5.6 hours postdose, respectively. The concentration found in foremilk was 25.2 and 29.3 ng/mL in the hindmilk. Based on these data, a RID of 0.4% and 0.4%, respectively, was estimated for an exclusively breastfed infant (cf. Table 1). On the same day, the concentration found in the maternal plasma was 60.6 ng/mL taken 5.4 hours postdose, respectively.

Methylphenidate exposure through breast milk

Hackett et al. reported a mean milk concentration over 24 hours of 19 ng/mL in three mothers taking an average of 52 mg/day of an immediate release formulation of methylphenidate while breastfeeding. These results correspond to an exposure of 0.0029 mg/kg/day (RID of 0.7%). ⁶ In another publication, the same authors described a mother taking methylphenidate 40 mg twice daily (5 days/week). The mean milk concentration over 24 hours was 15.4 ng/mL and the calculated infant exposure was 0.0023 mg/kg/day (RID of 0.2%). ⁷ Spigset et al. estimated that a fully breastfed infant would receive a methylphenidate dose of 0.00038 mg/kg/day (RID of 0.2%), based on a mother who was taking 5 mg of immediate-release methylphenidate in the morning and 10 mg at noon. ⁸ In these three studies, the estimated infant drug exposure through breast milk was less than a RID of 1%.^6–8,10

In case 1, we estimated that a fully breastfed infant would receive a dose of 0.0012 mg/kg/day (RID of 0.2%). Our results are therefore in agreement with the previously cited publications.

Duloxetine exposure through breast milk

A limited number of studies have documented infant exposure to duloxetine through breast milk.^1–3 A pharmacokinetic study reported six lactating women who took duloxetine 40 mg twice daily for a total of seven doses. ³ Breastfeeding was suspended for the duration of the experiment to avoid infant drug exposure. At steady state, the amount of duloxetine in the breast milk was 0.007 mg/day and the normalized milk excretion corresponded to an infant dosage of 0.002 mg/kg/day. ³ These authors used a different calculation method, making it difficult to compare results between studies. Two other cases were published in the literature. The first was described by Briggs et al., in which the mother was taking duloxetine 60 mg/day. ² The total infant dose, 32 days after birth, was 0.0071 mg/kg/day (RID of 0.8%). ² No adverse events, developmental issues, or signs of toxicity were observed in the child. ² Boyce et al. reported the case of a 31-year-old mother, who was taking duloxetine 60 mg daily. ¹ She was breastfeeding a healthy 18-day-old infant with normal weight gain. The calculated absolute dose in the maternal milk was 7.6 ng/mL (RID of 0.8%). ¹

In the present case series, three of the five milk samples were taken ∼6 hours postdose, around the expected moment of peak duloxetine concentration in the milk. The other samples were taken just before the next dose, corresponding to trough duloxetine concentrations (cf. Table 1). These multiple time points provide a more realistic evaluation of the infant's exposure, avoiding overestimation of the drug concentration.^4,11,12

In case 1, the concentration measured in the milk at week 4, 6.5 hours postdose, was 32.8 ng/mL (RID of 0.3%), which is in line with previous publications.

In case 2, the concentration of duloxetine in the foremilk and the hindmilk at day 6 was measured in two samples taken 21.6 and 23.3 hours postdose, respectively. The concentration of duloxetine was 23.6 ng/mL in the foremilk and 14.3 ng/mL in the hindmilk (estimated RID of 0.4% and 0.2%, respectively). These results show a lower concentration in the hindmilk than in the foremilk. However, it is known that hindmilk fat composition is usually higher compared with foremilk; duloxetine is therefore expected to have a higher concentration in hindmilk due to its lipophilic properties. Indeed, Boyce et al., found that the concentration of duloxetine was up to twofold higher in the hindmilk than in the foremilk. ¹ Hale and Rowe suggest that the molecule undergoes lipid cotransport across the alveolar epithelium of the breast cells. ¹³ Composition analysis of our samples revealed indeed a lower fat content in the hindmilk than in foremilk (1.9 g/100 mL versus 4.4 g/100 mL, respectively), which would confirm the influence of lipids on duloxetine excretion. A possible explanation for our results could be the natural variations in milk composition over time or a mislabeling of the tubes.

At week 6, the concentration of duloxetine in the foremilk and the hindmilk was measured in two samples taken at 5.2 and 5.6 hours postdose, respectively. The concentration found was 25.2 ng/mL in the foremilk and 29.3 ng/mL in the hindmilk (estimated RID of 0.4% and 0.4%, respectively). These results corroborate the findings of Boyce et al. regarding the higher concentrations of duloxetine usually found in the hindmilk. ¹ In accordance with the previously published cases, our results indicate a RID inferior to 1% and suggest a minimal transfer of duloxetine into breast milk.

Cord to maternal plasma (C/M) ratio of duloxetine

In case 2, the measured cord and maternal plasma concentrations yielded a C/M ratio of 0.4 (Table 2). This result suggests a more significant transfer across the placenta than previously suggested by Boyce et al. (C/M = 0.1), but remains inferior to other authors' observations.^1,2 It is difficult to properly interpret these conflicting results due to missing information; however, these results are likely due to interpatient variability and the limited amount of expositions described in literature. Despite the abovementioned differences in C/M ratios, none of the infants presented symptoms of neonatal adaptation syndrome.^1,2

Neurodevelopment of the exposed infants

The duration of follow-up was 6 months in case 1 and 6 weeks in case 2. In the first case, the infant showed no adverse effects associated with methylphenidate, such as irritability, sleep disturbances (insomnia), or delayed growth; but did demonstrate signs of drowsiness, likely linked to the child's congenital pulmonary condition. The two children in this case series did not present any adverse event or sign of toxicity related to duloxetine exposure, namely irritability, behavioral impairment, feeding difficulties, sedation, or sleeping disturbances. Nevertheless, both infants were partially breastfed, thus limiting the exposure to the medications.

This case series adds additional data to the sparse scientific literature on infant exposure to duloxetine and methylphenidate during breastfeeding. Yet, some caution is warranted in interpreting these results and in generalizing them to a broader population, since the sample size remains very small and the technique used may differ from other centers. In addition, results may not apply to infants exposed to a combination of psychotropic medications. However, based on our findings and other published cases, it is likely that the simultaneous exposure to both duloxetine and methylphenidate will remain low, even with the expected interpatient variability.

Considering the negative impacts, both on the mother and infant, of poorly controlled psychiatric and neurological disorders, the decision to continue maternal treatment during lactation appears acceptable in most cases. This decision is supported by the overall stability presented in this study, of both mothers' psychiatric or neurological condition under uninterrupted treatment. Another reassuring fact is that duloxetine, unstable in acid environment, is degraded in the stomach. ³ The absorbed amount of the medication from the gastrointestinal tract, and therefore the child's exposure, might be lower than estimated in infants aged 1 month or older. ¹⁴ One of the main limitations of this case series was the lack of measurements of duloxetine and methylphenidate in the infant's blood during lactation. Another limitation was the absence of a long-term follow-up of neurobehavioral development of exposed children until preschool aged. Finally, when comparing RIDs from different articles, one should be aware that the method of calculation may differ slightly between articles. We therefore made sure to include our method of RID calculation to facilitate future comparisons and to recalculate all the RIDs presented above with the cited method of calculation.