Pregnancy after mechanical mitral valve replacement

Background and aim of the study: Forty-six pregnancies among 32 patients after mitral valve replacement (MVR) were reviewed. Prosthetic valve-related complications, and outcome of pregnancy including fetomaternal mortality and morbidity for different anticoagulation regimens are presented and discussed. Methods: Among 521 women of fertile age undergoing MVR with a St. Jude Medical mechanical prosthesis, 32 patients developed a total of 46 pregnancies. Average patient follow up was 5 4 years (range: 10 months-17 years); total follow up was 155 patient years (pt-yr). Follow up commenced at onset of the first pregnancy. Results: Ten-year Kaplan-Meier survival estimate was 94 +/- 6%; 10-year freedom from valve-related events was 33 +/- 14%. Rates for embolism, anticoagulation-related bleeding and mechanical valve thrombosis were 4.5%, 3.2% and 2.6% per pt-yr, respectively. Among 30 patients receiving uninterrupted low-dose oral warfarin plus aspirin throughout pregnancy, three had normal deliveries, two had premature births, one had a low birth weight, seven had spontaneous abortions, and 17 had therapeutic abortions. By contrast, among eight patients who discontinued anticoagulation despite medical advice, seven had normal-term deliveries without thromboembolic complications, and spontaneous abortion occurred in one patient. Of the five women taking low molecular-weight heparin regimen, three had normal deliveries, one had a premature birth, and one an abortion. Two patients taking warfarin replaced by heparin in the first trimester and in the last two weeks, had term deliveries. One of these women developed left atrial thrombus in the third trimester while receiving heparin; after switching back to warfarin, the thrombus dissolved spontaneously. Another patient on heparin throughout the gestation had an uneventful gestation period that resulted in term delivery. There were four cases of prosthetic valve thrombosis during the postpartum period; all of these developed in women who ceased anticoagulation during pregnancy. Conclusion: There were no congenital malformations or maternal mortality/morbidity during pregnancy in this series of 20 live births, probably due to the low-dose anticoagulation regimen used. However, anticoagulation cessation was associated with a high prosthetic valve thrombosis rate in the postpartum period, even when a new-generation prosthetic valve of unique design and expected low thrombogenicity was implanted.