Warfarin

Pregnancy

DailyMed states that warfarin is contraindicated during pregnancy, except for pregnant patients with mechanical heart valves at high risk of thromboembolism when the benefits may outweigh the risks. Its label describes first-trimester warfarin embryopathy with nasal hypoplasia, stippled bone growth, and growth restriction, and reports brain and eye abnormalities, pregnancy loss, fetal bleeding, and fetal death after exposure. The 2010 Wainwright and Beighton autopsy series included 13 fetuses with warfarin embryopathy from 17 to 37 weeks: 11 had nasal hypoplasia, 3 had epiphyseal stippling, and 8 had cerebral hemorrhage. The authors had no maternal dose or exposure-timing data, so this selected series cannot estimate incidence. The 2015 prospective ROPAC registry included 212 pregnancies with mechanical heart valves. First-trimester vitamin K antagonist use, compared with heparin, was associated with miscarriage rates of 28.6% versus 9.2% and late fetal death rates of 7.1% versus 0.7%. This was an observational, high-risk mechanical-valve population and the comparison was not warfarin-only. A 2018 systematic review of 12 cohorts reported a pooled teratogenicity rate of 5.0%, with 2.4% in low-dose and 10.5% in high-dose groups; cohort risk of bias was moderate. DailyMed also reports about 5% of exposed offspring, but the underlying primary denominator was not traced here, so that label figure is marked not verified as a population rate and is omitted from the period summaries. The review’s pooled estimate is retained as a named review result, not a general population estimate. Dose-response findings differ: the 2018 review found lower pooled teratogenicity with low-dose warfarin, while Soma-Pillay et al. (2011) reported that fetal-abnormality risk was not related to dose, although stillbirth increased with dose. No dose threshold has been shown to remove fetal risk. A 2026 single-center KYBELE follow-up assessed 32 live-born children from 30 pregnancies with mechanical valves: 12 had enoxaparin alone, 8 had enoxaparin plus warfarin 2.5 mg/day, and 12 had enoxaparin plus warfarin 4 mg/day. No skeletal dysplasia or features suggestive of warfarin embryopathy were identified, but pregnancy losses were not included and the cohort cannot rule out uncommon outcomes. Pati et al. (1994) discussed brain and eye effects reported after later exposure. Lee et al. (2003) described a fetal subdural hematoma at 24 weeks, and Shan et al. (2023) described fetal intracranial bleeding at 35 weeks. These case reports document possible outcomes but cannot estimate their frequency.

Breastfeeding

LactMed (last revised 15 May 2024) reports that warfarin was not detected in milk at 4 hours after a dose in 13 mothers 3 to 12 days postpartum who received 2 to 12 mg daily; the reported limit was less than 25 mcg/L. LactMed reports no warfarin activity in the plasma of 7 tested infants and normal anticoagulation measures in tested infants. Orme et al. (1977) measured warfarin in 13 mothers; in 7 breastfed infants, warfarin was not detected in plasma. DailyMed summarizes published data in 15 nursing mothers and 15 full-term newborns: prothrombin times were documented within the expected range for 6 infants, were not obtained for the other 9, and premature infants were not evaluated. LactMed reports no adverse reactions in breastfed infants and no special precautions; it also describes two infants breastfed for 56 and 131 days without reduced vitamin K-dependent clotting factors or bleeding. DailyMed recommends monitoring breastfed infants for bruising or bleeding, and NHS SPS calls monitoring a precaution while describing warfarin as the preferred oral anticoagulant during breastfeeding. No infant harm was reported in these small human data, but the studies cannot exclude rare effects and provide little information about premature infants.

What was tested

The pregnancy evidence includes a prospective registry of pregnancies with mechanical valves, a systematic review of 12 cohorts, an autopsy series, and individual case reports. The registry compares first-trimester vitamin K antagonists with heparin in a high-risk group, while the label combines second- and third-trimester outcomes without separate estimates. The breastfeeding evidence includes small milk-level studies, infant blood measurements, and clinical follow-up. DailyMed reports that premature infants were not evaluated.