Pregnancy with prosthetic heart valves - 30 years' nationwide experience in Denmark

Abstract

Objective: Pregnancy in women with prosthetic heart valves remains a risk factor for both mother and fetus, but unselected and unbiased outcome and complication data remain scarce. We analyzed nationwide outcome data from 1977 to 2007 for all pregnancies in women with prosthetic valves.

Methods: Cardiac, obstetric, and neonatal data were obtained from obligatory databases and compared with general female population data. Questionnaires were used to corroborate important information. Outcome data were analyzed according to type of anticoagulation used. The data were compared between the two first and the last decades of the study period. In the last decade, patients were compared to an age-adjusted selected population of healthy, pregnant women.

Results: Of 356 women between 15 and 40 years of age, 79 women had 155 pregnancies after valve replacement. Two women died during pregnancy, one from heart failure and one from post-partum bleeding. There were four thrombo-embolic episodes in the early study period in women with mitral prosthesis on unfractionated heparin. Important cardiac complications were otherwise almost absent. There were significantly more early miscarriages and terminations in patients compared with controls (last decade 34%, vs 20% (p=0.0036) and 26% vs 13% (p=0.000019)). Post-partum bleeding was more common in the patient group (p=0.0021). Two late fetal losses (one from intracerebral bleeding) were seen. The remaining pregnancies resulted in 60 live births. Cesarean section was the predominant method of delivery in patients as opposed to controls (55% vs 16%, p=0.000000000097). Premature births were more frequent in patients (49% vs 5.5%, p=0.00000000039) as were congenital malformations (14% vs 5.7%, p=0.044). Two of the six malformations were warfarin embryopathy (8% of all first-trimester warfarin exposures), both seen in high-risk patients on high warfarin dosing. Small for gestational age did not differ significantly from the general population (9.3% vs 6.0%, p=0.39).

Conclusion: Data acquired over 30 years confirm that women with prosthetic heart valves, especially aortic prostheses for congenital lesions, generally tolerate pregnancy well, although cardiac mortality, mortality related to anticoagulation and thrombo-embolic risks are raised. Our data provide further documentation on the significance and importance of the risks associated with predominantly warfarin-based treatment regimens, which still remains optional for a number of patients. Finally, the data also serve as a comparison for recently published series based on low-molecular-weight heparin (LMWH) regimens.