Transfer of Olanzapine Into Breast Milk, Calculation of Infant Drug Dose, and Effect on Breast-Fed Infants

Seven breast-feeding women and their infants were enrolled in the study. The study design was approved by the research and ethics committee of King Edward Memorial and Princess Margaret Hospitals, Western Australia, and by the Canterbury ethics committee, Christchurch, New Zealand. Written informed consent was obtained from all participants.

Venous blood samples (5–10 ml, heparinized) were collected from six of the women either with a forearm catheter or by venipuncture five to seven different times across a dose interval of 12 or 24 hours. At approximately the same time intervals, milk samples were obtained with an electric breast pump, and aliquots (10–15 ml) were retained for drug assay. Milk was collected by one of two methods that attempted to provide a mean estimate of olanzapine concentrations in milk over an expression/feeding period. Subjects A–C emptied their breasts of milk, the milk was mixed, the volume was recorded, and a sample from each expression was retained for analysis. Subjects D–F provided milk samples pre- and postfeeding; equal volumes were mixed together in the clinic, and this sample was retained for analysis. For three subjects, the entire study was conducted in the hospital, while plasma and milk sampling after 8 hours occurred in the homes of the remaining three. For the seventh woman, a single set of pre- and postfeeding milk samples was collected, together with a plasma sample, 18 hours after her morning drug dose. Six of the women gave consent for a venous blood sample (0.5–2.0 ml, heparinized) to be taken from their infants. Infant health and well-being were evaluated by inquiry of all the mothers, together with a clinical examination by a specialist neonatologist or general practitioner (patients B, C, D, E, F, and G) using developmental screening and assessment instruments (19), the Denver Development Screening Test (20), or a Griffiths developmental assessment (21) in five cases. Infant body weight for age was checked against gender-specific population percentile graphs (22).

Olanzapine standard material was donated by Eli Lilly Australia, Pty. Ltd., West Ryde, Australia, and mexiletine (internal standard material) by Boehringer Ingelheim, Pty. Ltd., North Ryde, Australia. All solvents and chemicals were of analytical or high-performance liquid chromatography grade.

High-performance liquid chromatography was used to quantify olanzapine in plasma (23) and milk (16). The intra- and interday coefficients of variation for the plasma assay were 4.6% and 2.4%, and 8.2% and 3.9% at 6 μg/liter and 120 μg/liter, respectively. For milk, the intra- and interday coefficients of variation were 17.3%, 6.3%, and 3.4%, and 19.2%, 6.3%, and 3.3% at 1 μg/liter, 6 μg/liter, and 120 μg/liter, respectively. The limit of quantitation was approximately 1 μg/liter for both plasma and milk. However, for infant plasma, the limit of quantitation was sometimes as high as 5 μg/liter because less than a 1-ml sample was available. Data were summarized as means (ranges) or medians (25th and 75th percentile ranges), as appropriate.

The area under the concentration time curves for a dose interval (12 or 24 hours) was calculated with Topfit 2.0 (log trapezoidal rule) (24) for plasma and rectangular areas for milk (Σ concentration by collection time) (25). Milk-to-plasma ratios were calculated from data for area under the curve for six patients and single milk and plasma concentration measurements for one patient. An infant milk intake of 0.15 liter/kg/day was assumed (26); this value was multiplied by the average milk concentration (area under the curve_0–τ/dose interval, in which τ=dose interval; patients A–F) or a single-point milk concentration (patient G) to give the absolute infant dose of olanzapine in μg/kg/day. The infant dose was then expressed as a percentage of the weight-normalized maternal dose (μg/kg/day).

The mean age of the women was 30 years (range=23–39), and their mean body weight was 71 kg (range=59–110). Their infants consisted of three boys and four girls with a mean age of 2.4 months (range=0.1–4.3) and a mean weight of 5.3 kg (range=4.2–6.3). The median dose of olanzapine ingested by the women was 7.5 mg/day (range=5–20) or 127 μg/kg/day (range=75–286). Pharmacotherapy with olanzapine had begun a mean of 114 days (range=19–395) before the study day, and all participants were considered to be at steady state at the time of study. Patients A and F took olanzapine during both pregnancy and lactation, while all of the other patients began taking the drug in the postpartum period because of the development of acute psychosis.

The plasma and milk concentration time profile for subject D is shown in Figure 1, and the pharmacokinetic descriptors for subjects A–F appear in Table 1. The median maximum olanzapine plasma concentration of 29 μg/liter occurred at a median time of 2.5 hours after the drug dose was taken by the mothers. In the milk, the median maximum concentration of olanzapine was 16 μg/liter at a median of 5.2 hours after the dose was taken by the mothers, while the median of the mean olanzapine concentrations in milk was 8.8 μg/liter. The median milk-to-plasma area-under-the-curve ratio for olanzapine was 0.38. The single milk-to-plasma ratio for subject G, measured 6 hours after the olanzapine dose was taken by the mother, was 0.75.

The median absolute infant dose (mean olanzapine concentration × 0.15 liter/kg/day) for the six patients with complete dose interval data was 1.32 μg/kg/day (25th–75th percentile range=0.74–1.71). The median relative infant olanzapine dose (weight-adjusted comparison with the maternal dose) in these six patients was 1.02%, with a 25th–75th percentile range of 0.93–1.12. For patient G, the absolute and relative infant doses, calculated from a measured milk concentration of 6 μg/liter, were 0.9 μg/kg/day and 1.13%, respectively.

Overall, olanzapine concentrations were below the limit of detection (varying from 1 μg/liter to 5 μg/liter) in the six infants from whom a plasma sample was taken at a median time of 6 hours (25th–75th percentile range=5.5–13.4 hours) after the last maternal dose was taken (Table 1).

Detailed medical examinations of infants B, C, E, and F showed no adverse effects. A Denver developmental examination of infant E indicated no developmental delay. A Griffiths developmental assessment conducted on infant D indicated a developmental age of 78% of her chronological age, but causality was difficult to assess because of recent maternal treatment with clonazepam, sertraline, valproate, thioridazine, and droperidol, as well as olanzapine. For logistic reasons, infant A was not assessed. Infant G was referred to this study because of drowsiness but was unaffected at the time of study after a halving of the maternal olanzapine dose to 5 mg/day 3 weeks earlier. Interviews with the infants’ mothers did not elicit any further problems. As a measure of infant progress, body weights on the study day were assessed with U.S. National Center for Health Statistics growth curves for children (22). On the study day, infants A, B, and E had reached a higher percentile curve than that indicated by their birth weights, but the weights of infants C and D, who were at the 97th percentile at birth, had dropped to the 90th and 50th percentiles, respectively. Infant F (who was 4 days old on the study day) had a weight that was above the 90th percentile at birth, despite olanzapine exposure during pregnancy. The weight of infant G was not available.

The median and maximum relative infant doses of approximately 1.0% and 1.2% indicate that infant exposure to olanzapine through milk is much less than the notional cutoff of 10.0% that has been used to guide drug safety during breast-feeding (26). This low infant exposure was confirmed by failure to detect olanzapine in the plasma of six suckling infants and the lack of observed side effects in seven infants.

Of interest, our median milk-to-plasma ratio for olanzapine of 0.38 is identical to the theoretical milk-to-plasma ratio of 0.38 calculated by the method of Begg et al. (27). It is also similar to the mean of 0.46 measured by Croke et al. (16) from nine single-point milk and plasma concentrations measured in five patients.

The time at which maximum olanzapine concentrations occurred in milk was slightly later in the dosing interval than the corresponding maximum concentration in plasma. Mothers trying to minimize infant exposure may wish to avoid breast-feeding at peak milk concentrations of the drug (i.e., 5 hours postdose).

These data indicate that olanzapine may be considered relatively safe when mothers are breast-feeding healthy term infants. However, as always, risks and benefits should be considered before using any drug in breast-feeding mothers.

Received Sept. 23, 2002; revision received Jan. 8, 2003; accepted Jan. 21, 2003. From the Department of Clinical Pharmacology, Christchurch Hospital; the Departments of Pharmacy, Neonatal Services, and Psychological Medicine, King Edward Memorial and Princess Margaret Hospitals, Subiaco, Western Australia; the Clinical Pharmacology and Toxicology Laboratory, Western Australian Centre for Pathology and Medical Research, Nedlands, Western Australia; the Mothers and Babies Service, Princess Margaret Hospital, Christchurch, New Zealand; and the Department of Pharmacology, University of Western Australia, Crawley, Western Australia. Address reprint requests to Ms. Gardiner, Department of Clinical Pharmacology, Christchurch Hospital, PB 4710 Christchurch, New Zealand; [email protected] (e-mail). Funded in part by the Women and Infants Research Foundation, Subiaco, Western Australia. The authors thank Hector Faulkiner, Michael Paech, Neville Butcher, Nigel Ford, and Nigel Orr for assistance with sample collection.