Maternal Complications and Pregnancy Outcome in Women With Mechanical Prosthetic Heart Valves Treated With Enoxaparin

Abstract

Anticoagulation therapy is recommended throughout pregnancy for women with mechanical heart valves, to reduce the risk of thromboembolic complications. Although warfarin is the preferred therapeutic agent for such patients when not pregnant, warfarin is teratogenic; during pregnancy, unfractionated or low molecular weight heparin (LMWH) may be considered because it is safer for the fetus, although it is not clear that it provides adequate anticoagulation. A large number of studies have reported a wide variation (3.9%–22%) in the percentage of thrombotic complications reported in pregnant women with mechanical heart valves managed with LMWH (enoxaparin), although most studies were small. When analysis was limited to studies including 5 or more pregnancies, a rate of 6.9% was reported. This retrospective single center study investigated the rate of maternal thrombotic and hemorrhagic complications and other maternal and fetal outcomes in 31 women with mechanical heart valves who were treated with therapeutic dose LMWH (enoxaparin) during pregnancy. Between 1997 and 2008, data were extracted from the medical records of these women and analyzed for the maternal primary endpoints of thromboembolism and hemorrhagic complications. Secondary maternal and fetal outcomes that were analyzed included miscarriage, stillbirth, neonatal or infant death, live birth, small for gestational age, warfarin embryopathy, and warfarin-associated fetal loss. Most patients received therapeutic dose enoxaparin (1 mg/kg BID) and aspirin 100 to 150 mg q day until planned delivery. A total of 47 pregnancies were evaluated in 31 subjects. Thirty-four (72.3%) were managed with enoxaparin as the primary anticoagulant and the remaining 13 (27.7%) received warfarin as the main anticoagulant, with enoxaparin used between 6 and 12 weeks of gestation and/or peridelivery. Thromboembolic complications occurred in 7 (14.9%) pregnancies; of these, 5 (10.6%) were directly related to enoxaparin therapy. Poor compliance with enoxaparin was a major factor in 3 of these events and failure to achieve therapeutic target anti-Xa levels was a major contributing factor in the other 2 cases. The other 2 thrombotic complications occurred because the patients electively stopped taking their anticoagulant (warfarin). Antenatal bleeding occurred in 17% (8/47) of the pregnancies; 5 of the 8 bleeds were associated with enoxaparin, and the other 3 occurred during treatment with intravenous unfractionated heparin. Primary or secondary postnatal hemorrhage complicated 32% (15/47) of the pregnancies; 6 were due to enoxaparin, one occurred on unfractionated heparin, one on unfractionated heparin and warfarin, and one on warfarin alone. Among the 35 pregnancies continued beyond 20 weeks of gestation, 96% (22/23) of the women taking mainly enoxaparin had a surviving infant compared with 75% (9/12) of those taking primarily warfarin (P = 0.11). There were 4 perinatal deaths; 3 due to warfarin. These findings suggest to the investigators that administration of therapeutic dose enoxaparin in combination with low-dose aspirin to pregnant women with mechanical heart valves can achieve a low risk of valve thrombosis and a high rate of infant survival, provided that there is close clinical follow-up and high compliance.