Pregnancy and delivery represent a major hemostatic challenge in von Willebrand disease (VWD). In addition to maintaining adequate hemostasis, von Willebrand factor (VWF) is an important regulator of angiogenesis, with angiodysplasia contributing to morbidity in patients with VWD. Placental development is complex, with vascular development important for fetal growth and viability. It has been hypothesized that antepartum bleeding or abnormal placental angiogenesis may lead to pregnancy loss in women with VWD. Large studies have reported an increased risk of postpartum bleeding in patients with VWD, however, little information is available for rates of pregnancy loss. We conducted a cross-sectional study to determine the rate of pregnancy loss in women with VWD, when compared to a control population.

We invited all women with VWD followed by the Southern Alberta Rare Blood and Bleeding Disorders Comprehensive Care Program to participate in an online or paper questionnaire, along with a control group of women without VWD identified through a low-risk obstetrical clinic in Calgary, Alberta. We included patients who were 18 years of age or older with a diagnosis of VWD, defined as VWF antigen and activity levels less than 50 percent or a historical diagnosis, in combination with a bleeding score of 4 or more (condensed MCMDM-1 bleeding questionnaire). Patients were excluded if they had no past pregnancies, no mailing address, or had an alternative bleeding disorder. Our 20-item questionnaire was reviewed and pre-tested by hematologists, obstetricians and laypeople. With patient consent, data was supplemented from clinical and hospital records to verify and expand on the information collected in the questionnaire. The primary outcome was the incidence of spontaneous abortion (fetal loss <20 weeks) or late pregnancy loss (≥20 weeks) in women with VWD. The secondary outcomes were antepartum and postpartum bleeding. Data on co-morbidities including placental abnormalities were also collected. Confidence intervals for proportions were calculated using the Wilson’s score method. Groups were compared using χ2 testing, with p values < 0.05 considered significant.

Of the 98 women with VWD who met the study inclusion criteria, 31 (32%) completed the questionnaire. The mean bleeding score was 10.3 in VWD patients (SD 3.6). The majority (81%) of participants were diagnosed with Type 1 VWD, with 16% and 3% diagnosed as Type 2 and 3 VWD, respectively. Twenty-two women in the control group completed the questionnaire. The mean age at recruitment was 46.1 (SD 13.7) and 32.6 (SD 4.2) in the VWD and control group, respectively. There were 83 and 48 pregnancies in the VWD and control group, respectively. Two women in the VWD group and 1 woman in the control group had an elective pregnancy termination.

In women with VWD, the rate of pregnancy loss was 24.7% out of 81 pregnancies (95% confidence interval 16.6-35.1), with 17 (85%) losses occurring less than 20 weeks, and 3 (15%) with unknown timing. In contrast, the rate of pregnancy loss in the control group was 12.8% (95% CI 5.96-25.17) (p=0.106), all occurring less than 20 weeks. In 83 pregnancies, the number of antepartum bleeding episodes in VWD women was 45 (29 first trimester, 9 second trimester, 7 third trimester), compared to the control group (9 first trimester, 3 second trimester, 1 third trimester) in 48 pregnancies. More women with VWD reported excessive postpartum bleeding compared to controls (56.7% versus 13.6%, p=0.002), with more delayed (>24 hour) bleeding post-delivery in women with VWD.

In VWD women with and without pregnancy loss, there was no difference in mean bleeding score (9.7 vs. 10.9). There was also no difference in cases of pre-eclampsia, placental abruption or gestational hypertension in VWD women with and without pregnancy loss. Two VWD patients with previous pregnancy loss developed gastrointestinal angiodysplasia later in life.

There was no significant increase in the rate of pregnancy loss in women with VWD, however, a larger cohort study is needed to confirm the rate of pregnancy loss before further conclusions can be made.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract