Limited dose warfarin throughout pregnancy in patients with mechanical heart valve prosthesis: a meta-analysis

Abstract

The continuation of warfarin throughout pregnancy in patients with a mechanical valve prosthesis is a valid anticoagulation regimen, provided that warfarin dose does not exceed 5 mg/day. Two decades after being introduced, the efficacy and safety of this regimen merit evaluation. We performed a systematic review for cases published between January 1991 and January 2013. We compiled our prospective data on 55 pregnancies and calculated pooled estimates (95% confidence interval) of adverse foetal and maternal outcomes. Events were expressed as proportions of total pregnancies, except embryopathy and maternal death, which were related to the number of live births and number of patients, respectively. There were 494 eligible pregnancies reported in 11 studies. The rate of embryopathy was 0.9% (0.4-2.4%) and most of the 13.4% (8.4-24.7%) foetal losses were due to the 12.8% (7.7-22.7%) rate of spontaneous abortion. No maternal mortality was encountered (0-1.3%) but 0.6% (0.3-2%) prosthetic valve thrombosis, 1.8% (1.1-3.6%) total thromboembolic events and 3.4% (2-5.1%) major maternal bleeding events were recorded. Foetal loss, spontaneous abortions and foetal embryopathy dropped to 8.1% (2.9-13.7%), 7.3% (3.1-11.8%) and 0.6% (0.1-2.1%) among the 344 pregnancies (69.6%) observed in the 6 prospective studies (54.5%). Prosthetic valve thrombosis (0.6%; 01-2%), total thromboembolic (2.3%; 1.2-4.6%) and major bleeding events (2.9%; 1.8-6%) remained comparable with overall results. Foetal embryopathy and prosthetic valve thrombosis were not robust on sensitivity analysis, regardless of the study design. A prospective subgroup of 96 patients (19.4%) received smaller warfarin dose, through targeting a lower international normalized ratio (INR) between 1.5 and 2.5. The associated rate of foetal loss (2.1%; 0.5-6.9%) was significantly lower than that observed in the remaining patients targeting a higher INR between 2.5 and 3.5 (16.1%; 13.1-34.4%). Adverse maternal outcomes were also fewer but rates remained comparable. Limited dose warfarin throughout pregnancy was associated with improved foetal outcomes, without jeopardizing maternal safety. Foetal outcomes were better when patients were followed up prospectively or receiving smaller warfarin doses through targeting a lower INR than recommended (1.5-2.5). Large randomized controlled trials are mandatory to evaluate our findings.

Keywords: Mechanical heart valve prosthesis; Oral anticoagulants; Pregnancy; Systematic review; Warfarin.