The aim of this observational cohort study was to specify the risk of the vitamin K antagonist (VKA) phenprocoumon during first trimester of pregnancy, in particular to estimate the risk of birth defects and spontaneous fetal loss. Four hundred eight pregnancies with phenprocoumon exposure were compared to 1,642 pregnancies neither exposed to VKA nor to other major teratogens or fetotoxicants. There was no typical warfarin embryopathy in our exposed cohort. However, the overall rate of major birth defects was significantly increased (7.4 % vs 2.3 %; adjusted odds ratio [ORadj] 2.14; 95 % confidence interval [CI] 1.4-3.4). With early cessation until five completed gestational weeks the birth defect risk was similar to the comparison cohort (2.4 % vs 2.3 %; ORadj 1.07; 95 % CI 0.2-3.6). With treatment duration exceeding seven gestational weeks the rate of major birth defects increased up to five-fold (10.8 % vs 2.3 %; ORadj 5.18; 95 % CI 2.0-11.6). The overall risk of spontaneous abortion (SAB) was 38.0 % vs 17.5 % in the comparison cohort (adjusted hazard ratio [HRadj] 2.9; 95 % CI 2.2-3.9). The treatment duration had a significant effect on the hazard of SAB (HRadj 1.12; 95 % CI 1.01-1.25 per each additional exposure week). Phenprocoumon and other VKA carry an embryotoxic risk. This risk seems to be time-dependent with a steep risk increase for birth defects and also for fetal loss after week 5. If maternal disease permits, VKA therapy should be switched to safer alternatives such as heparins immediately after early recognition of pregnancy.
Keywords: Time-dependence; congenital malformation; observational cohort study; spontaneous abortion; thrombosis prophylaxis.