Outcome of Noac Exposure during Pregnancy (...and the problem of event reporting...)

Background: Women of child-bearing potential (WOCBP) in need of oral anticoagulation should apply effective contraceptive measures, since exposure to vitamin-K antagonists (VKA) carries a relevant risk for embroypathy. Currently, non-VKA oral anticoagulants (NOACs) are replacing VKA. NOACs are small molecules that have been shown to cross the placenta but the clinical risk of embryopathy with NOACs is unknown. Since exposure in pregnant women can occur, we aimed to evaluate case numbers and pregnancy outcomes using data from various sources.

Patients and methods:  Cases of NOAC exposure during pregnancy were collected from our institution and by sending out questionnaires to German gynecologists, obstetricians and hematologists. Furthermore, all NOAC manufacturers were asked to provide cases that were registered in the pharmacovigilance system. Finally, German drug authority (BfArM) and the European Medicines Agencies (EMA) were contacted for cases from their databases. All reports were screened by two independent physicians using a standardized case report form.

Results: In this ongoing project, a total of 169 NOAC exposures in pregnancy were collected: 5 cases from our institution, 7 cases reported by physician questionnaires, 25 cases extracted from Boehringer Ingelheim pharmacovigilance, 10 cases extracted from BfArM pharmacovigilance and 122 cases were provided by EMA based on the safety reports from Bayer (export of further reports pending). NOAC exposure consisted of rivaroxaban (n=143), dabigatran (25) and apixaban (1) and venous thromboembolism was the indication for NOAC in >95% of cases with available information. Duration of NOAC exposure ranged between single dose to complete 2^nd and 3^rd trimester.

The most complete data were found in cases collected from physicians: All 12 cases presented details on duration of NOAC exposure (median 6 weeks; range 3-19 weeks) and on pregnancy outcome. In contrast, data provided by manufacturers and authorities had significant data gaps (see table).

Information of pregnancy outcome was available in 85/169 cases (50.3%) and consisted of 31 live births (36.5%); 21 miscarriages/abortions (24.7%); 27 elective pregnancy interruptions (31.8%), while 6 pregnancies (7.1%) were still ongoing at the time of our assessment.

Of the 20 live births, 3 newborns demonstrated abnormalities: renal pelvis dilatation and facial dysmorphismin in 1 case (rivaroxaban exposure in 1^st trimester) and low birth weight in 2 cases (girl of 2570g born in week 37 with dabigatran exposure in 1^st trimester; girl of 1175 g delivered by caesarean section for preeclampsia in week 30 with rivaroxaban exposure in 1^st trimester).

Of the 21 miscarriages, details were available for 5 cases (4 miscarriages occurred in 1^st trimester and one case in week 20 with preceding growth restriction).

Of the 27 elective pregnancy interruptions, details were available for 7 cases and decision was based on social reasons (3), fear of NOAC embroypathy (3) and medical reasons (maternal thyreotoxicosis with heart failure).

Conclusion: Currently available data do not suggest a high risk for NOAC embryopathy but case numbers are limited and demonstrate significant data gaps. Therefore, risk assessment needs to continue and pregnancy has to be further avoided in NOAC patients by effective contraceptive measures and education of WOCBP. However, currently available data do not justify pregnancy termination just because of NOAC exposure and non-directive counselling as well as close pregnancy surveillance is recommended.

Due to the strict anonymization of data provided from NOAC manufacturers and authorities, cases of double-reporting cannot be completely ruled out but, so far, NOAC exposure during pregnancy seems not uncommon. Further initiatives are needed to collect more data on the risk of NOAC embryopathy. However, the quality of data provided by manufacturers and drug authorities on request is inferior to the data quality obtained from direct physicians contact, which indicates the need to improve exposure and outcome assessment in this important medical scenario.

Further data exports from EMA are pending and updated results will be presented at ASH.