In 2019, Ayuk et al. 1 published in this journal a proof-of-concept study investigating dabigatran secretion into breast milk following oral administration to two non-breastfeeding postpartum women. Their study demonstrated milk-to-plasma (M/P) ratios of 0.02–0.1, with peak breast milk concentrations reaching 4%–12% of plasma levels. Crucially, the estimated maximum plasma dabigatran concentrations in the neonate were calculated to be 100 000 times below concentrations that would significantly affect coagulation indices. We now report the first case of measured dabigatran concentrations in an actively breastfeeding mother–infant dyad, including the first-ever measured infant plasma dabigatran level during breastfeeding. A 35-year-old woman (gravida 2, para 2; BMI 35 kg/m2; height 176 cm; weight 109 kg) with a family history of deep vein thrombosis was identified as requiring postpartum venous thromboembolism (VTE) prophylaxis, scoring 2 points on the Swedish Society of Obstetrics and Gynecology (SFOG) Hem-ARG VTE risk assessment 2. Standard prophylaxis with tinzaparin 8000 units subcutaneously daily for 7 days was prescribed. However, severe needle phobia, shown by a panic attack during epidural placement, rendered the patient unable to self-administer injections despite topical anesthetics. Following comprehensive counseling regarding the limited available evidence and potential risks of off-label use, the clinical team prescribed dabigatran etexilate 220 mg once daily for 7 days, with planned monitoring of maternal plasma, breast milk, and infant plasma drug concentrations as a safety measure accompanying off-label prescribing. This dosage was selected as it mirrors the only approved indication for dabigatran as primary VTE prophylaxis (short-term thromboprophylaxis following elective hip or knee arthroplasty) for which pharmacokinetic and safety data are available. An alternative twice-daily regimen of 110 mg was not considered, as this dosing has not been evaluated for primary prophylaxis. Both deliveries had been uncomplicated vaginal births, and the newborn weighed 3520 g with Apgar scores of 9–10–10. Following discharge approximately 12 h after delivery, the patient did not fill the prescription immediately and was reminded by telephone on Day 3, after which she took her first dose that afternoon. Subsequent doses were taken on the afternoons of days 4 and 5, with the fourth dose administered on Day 6 at 07:00. Blood samples from both mother and infant, along with expressed breast milk, were collected at approximately 09:00 on Day 6, 2 h after the fourth dose, corresponding to the expected time of peak plasma concentration 1. Maternal and infant plasma dabigatran concentrations were analyzed at Karolinska University Laboratory using validated liquid chromatography–tandem mass spectrometry (LC–MS/MS) methodology 3. Breast milk dabigatran concentrations were analyzed at the ADMEoT unit, Science for Life Laboratory Drug Discovery and Development Platform, Uppsala University, using a purpose-developed LC–MS/MS method that met established bioanalytical criteria for accuracy (0%–7% deviation) and precision (2%–6% RSD) across the calibration range (Table 1). The lower limit of quantification (LLOQ) was 5 nM (2.3 ng/mL). Full analytical methodology is provided in the Supporting Information S1. The maternal plasma dabigatran concentration was 45 ng/mL. The infant plasma concentration was below the detection limit (< 2 ng/mL), and the breast milk concentration was below the LLOQ (< 2.3 ng/mL; < 5 nM). A residual chromatographic signal at the dabigatran retention time in the breast milk sample yielded an estimated concentration of approximately 1.5 ng/mL (3 nM) upon extrapolation below the LLOQ, a value that should be interpreted with caution (Figure 1). The prodrug dabigatran etexilate was not detected in breast milk (Figure 2). No adverse events occurred in either mother or infant; the mother did not develop VTE, and the infant showed no signs of coagulation abnormalities. The maternal plasma concentration of 45 ng/mL is on the lower end of expected values for the 220 mg once-daily prophylactic regimen. Samama et al. 4 reported a mean peak concentration of 60 ng/mL (range 0–320) in orthopedic patients, while Ayuk et al. 1 observed peak concentrations of 205 and 415 ng/mL following a single 220 mg dose. Although published pharmacokinetic data indicate that body weight does not significantly affect dabigatran plasma concentrations 5, the lower observed level may reflect enhanced renal clearance due to postpartum physiological diuresis, individual pharmacokinetic variation, or the fact that steady-state conditions may not have been fully achieved after only four doses with delayed initiation. Maternal renal function was not measured. Importantly, the observed concentration confirms adequate drug exposure and excludes non-adherence as an explanation for undetectable infant levels. The undetectable infant plasma dabigatran concentration (< 2 ng/mL) validates the theoretical predictions of Ayuk et al. 1, who estimated that neonatal exposure through breastfeeding would be far below clinically relevant thresholds. Dabigatran, the active metabolite present in breast milk, is a highly polar zwitterion that is not meaningfully absorbed from the gastrointestinal tract. Even with the prodrug formulation (dabigatran etexilate), absolute oral bioavailability is only 6.5% 5, 6. This suggests that any dabigatran ingested by the infant through breast milk would be poorly absorbed, consistent with the undetectable infant plasma levels observed here. Notably, the breast milk concentration below the LLOQ precluded formal calculation of the M/P ratio and relative infant dose (RID). The World Health Organization Working Group on Drugs and Human Lactation has proposed that drugs with an RID below 10% are generally acceptable during breastfeeding 7. Although RID could not be calculated in this case, the negligible transfer observed (with breast milk concentrations too low to quantify) supports compatibility with breastfeeding under the dosing conditions evaluated. An important clinical consideration is the reversibility of dabigatran with idarucizumab, providing a safety advantage should urgent surgical intervention become necessary postpartum, such as for hemorrhage or retained placental tissue. This offers an additional safety margin compared to LMWH, for which reversal options are more limited. Furthermore, epidemiological evidence indicates that VTE risk remains elevated beyond the traditional 6-week postpartum period. Kamel et al. 8 demonstrated a 10.8-fold odds ratio in the first 6 weeks and a 2.2-fold increased risk during weeks 7–12, while Sultan et al. 9 found peak risk in the first 3 weeks with elevated risk continuing thereafter. The availability of oral anticoagulation could improve adherence to extended prophylaxis regimens by eliminating the daily burden of injections. At a 6-month telephone follow-up, the patient expressed satisfaction with the treatment and indicated she would strongly prefer oral tablets over injections for any future VTE prophylaxis. She described the postpartum experience as follows: “After delivery you feel really small, and you don't fully grasp what is going on. Just thinking of injecting myself with those needles gave me anxiety. I'm not normally scared of needles, but with my first pregnancy I had to take multiple blood tests, and this time the epidural I got was really painful. Now, it's like whenever I'm pregnant I associate needles with all that anxiety that something would be wrong with me or my baby. I know it is not rational, but with all the turmoil in the body going on after delivery I just couldn't make myself take those jabs.” This single case report has inherent limitations, including a single sampling time point, delayed treatment initiation, a lower-than-expected maternal plasma concentration without renal function data, and the inability to establish steady-state pharmacokinetics. Both blood samples and breast milk required analysis at external laboratories, although dabigatran plasma concentrations have been shown to be stable across different storage conditions 10. Nevertheless, this case provides the first direct evidence of undetectable plasma dabigatran in a breastfed infant during active maternal treatment, extending the theoretical framework established by Ayuk et al. to a real-world clinical scenario. These findings support the feasibility of pharmacokinetic dose-titration studies to establish optimal oral dosing for breastfeeding mothers, which could subsequently enable randomized non-inferiority trials comparing LMWH versus dabigatran for postpartum VTE prophylaxis in high-risk patients. Rasmus W. Green: conceptualization, patient recruitment and consent, coordination of specimen collection and transport, project administration, writing (original draft and revisions). Aljona Saleh: breast milk dabigatran analysis (investigation and methodology), preparation of figures and tables, writing (methods). Pawel Baranczewski: resources, methodological supervision, writing (review). Plasma concentrations were analyzed by routine clinical laboratory services. All authors approved the final manuscript. Claude Opus 4.6 (Anthropic) was used to assist in drafting the manuscript from author-prepared drafts and outlines to improve readability and compliance with journal formatting guidelines. All clinical data, analytical results, figures, tables, scientific interpretations, and conclusions are solely the work of the authors. The AI tool was not used for data analysis or generation of results. The authors have nothing to report. Under Swedish law (Act on Ethical Review of Research Involving Humans, SFS 2003:460), this case report does not constitute research subject to mandatory ethics review, as it describes clinical management of an individual patient. This position is consistent with precedents established by the Swedish Ethics Review Appeals Board (Överklagandenämnden för etikprövning, ÖNEP decisions Ö 34-2019 and Ö 60-2020), which have held that publication of individual patient cases managed within ordinary clinical care does not in itself render the work subject to the Ethics Review Act. Accordingly, no ethics board approval was sought or required. Although the prescription of dabigatran was off-label, withholding anticoagulation would have exposed the patient to an unacceptable risk of postpartum VTE. Off-label prescribing was therefore the only clinically tenable course of action. Given this circumstance, the clinical team took explicit responsibility for patient and infant safety by planning prospective monitoring of maternal plasma, breast milk, and infant plasma drug concentrations—measures undertaken as a direct consequence of, and safeguard against, the uncertainties inherent in off-label use, rather than as a primary research objective. The treatment decision was made through shared clinical decision-making following comprehensive counseling regarding risks and benefits. The head of department reviewed and approved the off-label use and the allocation of institutional resources for specimen collection. The patient provided informed oral consent for treatment. At 6-month telephone follow-up, the patient gave oral and written consent for publication after having read a manuscript draft and expressed satisfaction with her participation. The patient is fully anonymized in this report. Data collection and reporting were conducted in accordance with institutional protocols. The authors declare no conflicts of interest. All data generated during this study are included in this published article and its Supporting Information. Data S1: Method development and bioanalysis of dabigatran in human breast milk—Full analytical methodology and validation data sufficient for replication of the breast milk LC–MS/MS assay, including sample preparation, chromatographic and mass spectrometric conditions, method validation, representative chromatograms, and prodrug screening. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Green et al. (Thu,) studied this question.