Rivaroxaban Treatment in Two Breastfeeding Mothers: A Case Series

Introduction

Rivaroxaban (Xarelto) is an oral direct factor Xa inhibitor approved for the treatment of deep venous thromboembolism, pulmonary embolism (PE), and the prevention of stroke in nonvalvular atrial fibrillation. ¹ In a retrospective study done in 1997, the incidence of deep vein thrombosis during pregnancy was reported as 0.71 per 1,000 deliveries (95% confidence interval [CI] 0.5–0.9) with 17,000 deliveries. ² The incidence of PE was 0.15 per 1,000 deliveries (95% CI 0.06–0.24). ³ Although the mortality rate in pregnant women has declined in more recent years, PE remains an important diagnosis that requires treatment during and after pregnancy. Empiric anticoagulation therapy is generally recommended using subcutaneous low-molecular weight heparin. Other highly selective medications, such as rivaroxaban, are unique anticoagulants that directly inhibit factor Xa.

Rivaroxaban has high bioavailability and, unlike classic vitamin K antagonists, has no clinically relevant drug–drug interactions. It can be given at fixed doses without requiring routine coagulation monitoring due to its predictable pharmacokinetics and dynamics. ⁴ Rivaroxaban has low tissue penetration, circulates as an unchanged drug, and utilizes multiple elimination pathways. ⁵ Rivaroxaban is reversible with the use of Andexxa (coagulation factor Xa [recombinant], inactivated-zhzo) for emergent use in life-threatening hemorrage. Because PE is the seventh leading cause of maternal mortality, the use of rivaroxaban is potentially useful in many postpartum breastfeeding women with hemorragic disease.

Unfortunately, there is very limited information published on the transfer of this drug into human milk at a confirmed steady state. This study describes the drug concentration–time profile in the milk samples collected from two lactating mothers taking 15 mg twice daily and then after transition to 20 mg once daily.

Case Presentation

Patient 1 is a 35-year-old woman weighing 58.9 kg who gave birth to a female infant after 39 weeks of gestation. At 3 months postpartum, she was diagnosed with a mild stroke. She was initiated therapy at 15 mg twice daily and then after 21 days converted to 20 mg once daily for 6 months. She donated milk samples on day 10 of treatment at 0, 1, 3, 5, 8, 10, and 12 hours after taking 15 mg of rivaroxaban twice daily. Thereafter, she donated another set of milk samples at 0, 1, 2, 4, 8, 10, 12, and 24 hours on day 7 after changing her dose to 20 mg once daily.

Patient 2 is a 33-year-old woman, delivered a healthy male infant at 39 weeks gestation. At 4 weeks postpartum, she was diagnosed with a bilateral PE. She was administered rivaroxaban 15 mg twice daily for 21 days and then continued on 20 mg once daily for 6 months. Patient 2 donated two sets of milk samples at 0, 1, 2, 4, 8, 10, 12, and 24 hours, on day 10, first after taking 15 mg twice daily, and then on day 7 after taking 20 mg of rivaroxaban once daily.

Both patients were advised against breastfeeding once they began rivaroxaban treatment. They continued to pump to keep up their supply and donated milk samples for this study. At the time of the study, they were both at steady state. To preserve the homogeneity of the samples, the participants were advised to pump both the breasts completely, gently mix them and collect 1–2 ounces into a collcetion tube. The samples were frozen at −20°C and mailed to our research facility overnight. When the samples reached our facility, they were stored at −80°C until analyzed.

Materials and Methods

Quantification of rivaroxaban was determined using ABSciex QTRAP 5500 Ultra performance liquid chromatography - tandem mass spectrometer (UPLC - MS/MS). A Phenomenex Luna C-18 column, 50 × 2 mm, 3 μm, was used. Gradient elution was followed using water and acetonitrile with a flow rate of 0.5 mL/minute. Multiple reaction monitoring was m/z 436.1–145 for rivaroxaban and m/z 440.1–145 for internal standard. Milk samples were extracted using protein precipitation. Blank milk was spiked with appropriate concentrations of rivaroxaban and internal standard for determining the calibration curve, range 0.019–6.2 μg/mL. Extraction efficiency observed was 95%. All the samples were processed/analyzed in triplicate and described as a mean of values.

Results

Dose 1 (15 mg twice daily)

The average concentration of rivaroxaban in milk, observed in both the patients taking 15 mg twice daily, was 0.16 ± 0.0004 μg/mL. The maximum concentration was detected at 1 hour, as 0.3 ± 0.02 μg/mL. The levels gradually receded over time as shown in Figure 1A. The infant dose calculated was 0.01 mg/kg per 12 hours and the relative infant dose (RID) calculated for the 15 mg twice daily dose was 5%. The data presented in this study indicate presence of low levels of rivaroxaban in milk.

OPEN IN VIEWER

FIG. 1.

Mean milk concentration–time profile of rivaroxaban in human milk in two mothers (n = 2) after oral administration of (A) 15 mg twice daily and (B) 20 mg daily.

Discussion

Our study reports the presence of rivaroxaban in milk samples collected from two lactating mothers, suggesting only limiting transfer into human milk. The low RID can be explained by the high plasma protein binding of rivaroxaban and its poor penetration into human milk. ⁴ Therefore, it is unlikely to be excreted in clinically relevant amounts. Another factor to consider is the oral bioavailability of the drug in the infant once it has consumed the breast milk. There is an 80–100% bioavailability of a 2.5 mg dose of the drug in adults, and a 66% bioavailability of a 10 mg dose. ⁴ Unfortunately there are no studies on the bioavailability of rivaroxaban in infants although it is probably similar or lower in milk.

There is one previous case report on rivaroxaban administered 5 days postnatally with milk samples collected on day 3 of anticoagulation therapy. ⁷ In that study, milk samples were collected at 0, 3, 6, and 10 hours after administration. Using a similar method as this study, the results suggested an RID of 1.3%, which is below the level of concern hypothesized by Bennett. ⁶ However, the timing of the milk samples being collected on day 3 makes it uncertain whether steady-state pharmacokinetics were reached. The half-life of rivaroxaban was 11.7 hours as observed in a small study of only females with an average age of 33 years, so the steady state would be maximum at 2.4 days. In our study, samples were collected 10 days after consumption and thus quite likely to have reached steady state. An RID of 5% and 4% at 15 mg twice daily and 20 mg once daily, respectively, was calculated from these samples and suggests a low risk of infant toxicity.

Caution must be exercised until further studies are conducted reporting the bioavailability of rivaroxaban in infants and whether the pharmacodynamic effects of rivaroxaban in neonates have age-related differences. ⁸ In vitro studies have observed a dose-dependent response for prothrombim time, and activated partial thromboplastin time, and lag time in neonates, but in vivo studies are required to confirm these findings. ⁹ This report adds important information for breastfeeding women who are advised to take rivaroxaban for anticoagulation therapy.

Conclusions

The results of this study suggest that transfer of rivaroxaban into human milk is quite low. However, caution must be exercised until further studies are conducted that report the bioavailability of rivaroxaban in infants and its anticoagulant effect in breastfeeding infants.