Pregnancy outcomes in women with mechanical heart valves

Abstract

Abstract Background Mechanical heart valves (MHV) and their absolute need for adequate anticoagulation poses a challenge for pregnancy, either due to warfarin fetotoxicity or an increased risk of maternal thromboembolism. This represents a unique patient (P) group where data is scarce and maternal and fetal risks and benefits must be carefully weighed. Purpose To assess maternal and perinatal outcomes in women with MHV on different anticoagulant regimens and compare them with patients with other types of valvular heart disease (VHD). Methods A retrospective analysis of 131 pregnancies in 83 women with VHD (mean age 26.5±5.6 years) was carried out in a tertiary referral centre from 2000 to 2019. 92 pregnancies with VHD, including 11 with biological prosthetic valves, and 39 pregnancies in 22 P with MHV were identified. The main outcome measures were major maternal complications and perinatal outcome. Results MHV implanted were in mitral position (89.7%), aortic (2.6%), or both (7.7%). History of rheumatic heart disease was identified in 16 P (72.7%) and a congenital etiology was present in 2 P (9.1%). 9 P (40.9%) were on warfarine and 13 P (59.1%) on acenocumarol. Regarding anticoagulation strategy, 21 P (65.6%) remained on oral anticoagulation and 10 P (31.3%) had been switched to some form of heparin during part or the entire pregnancy. Mechanical valve thrombosis complicated pregnancy in 4 patients (10.2%), all cases on heparin, and resulted in maternal death in 1 P. MHV P had more hemorrhagic complications (15.4 vs 2.2%, p=0.004) requiring transfusion or surgical revision. MHV P tended to experience more NHYA class worsening demanding initiation or intensification of cardiac medication (17.9 vs 5.4%, p=0.023). Also in the MHV group there was a higher incidence of miscarriage (46.2 vs 12.0%, p≤0.0005), comprising spontaneous abortion (31.6 vs 7.6%, p<0.0005) and fetal malformations (18.4 vs 5.4%, p=0.028), including warfarin embryopathy (10.3 vs 1.1%, p=0.012). The live birth rate was higher in women on heparin compared with those on warfarin (85.9 vs 79.2%, p=0.002). The presence of multivalve disease (p=0.04), mechanical protheses (p<0.001), ACO (p<0.001) and previous impaired LVEF (p=0.02) were related to miscarriage. In multivariate analysis, ACO was the unique independent predictor of unsuccessful pregnancy (p=0.01). Only 29% of the patients with an MHV had a pregnancy free of serious adverse events compared with other types of VHD (81.5%, p<0.0005). Conclusions MHV remains a challenging condition for pregnancy with only 29% chance of experiencing an uncomplicated pregnancy with a live birth. The increased morbimortality warrant extensive prepregnancy counseling with prosthesis type discussion,centralization of care and further larger studies to come up with evidence-based recommendations. Funding Acknowledgement Type of funding source: None