Risk of Pregnancy-Associated Venous Thromboembolism in Sickle Cell Trait

Background: Sickle cell trait (SCT), which affects an estimated 3 million people in the United States, is a known risk factor for venous thromboembolism (VTE); however, its association with VTE in pregnancy has been significantly understudied (Austin et al, 2007). Leveraging data from 23andMe, Inc, a leading consumer genetics and research company, we aimed to estimate the risk of pregnancy-associated VTE, pulmonary embolism (PE), and isolated deep vein thrombosis (DVT) in individuals of African ancestry with SCT compared to those without SCT. Methods: Women of African ancestry with a history of pregnancy were included in this analysis. Exclusion criteria included women with sickle cell disease, specifically those with HbSS, HbSC, and compound heterozygotes of HbS and the available genotyped β-thalassemia mutations. VTE was defined by patient self-report on the 23andMe survey with a “yes” in response to “have you ever been diagnosed with, or treated for, a blood clot or condition causing repeated blood clots?” Pregnancy-associated VTE specifically was defined as selecting the response “pregnancy” to the question of “Did your doctor tell you that your blood clotting condition was mainly caused by any of the following?” Additional questions in the survey elicited the location of VTE as “clot in your arms or legs (deep vein thrombosis, DVT)” and “clot in your lungs (pulmonary embolism, PE).” Using these questions, VTE was subcategorized as PE +/- DVT and isolated DVT. We used multiple logistic regression to calculate adjusted odds ratios and 95% confidence intervals (CI) for the association of carrier status to pregnancy-associated VTE, isolated DVT, and PE +/- DVT adjusted for age, number of pregnancies, and principal components of genetic ancestry. Results: We included 73,442 women of African ancestry (based on statistically inferred genetic similarity) with a history of pregnancy in our analysis, of which 5219 women (7.1%) had SCT, which is consistent with prior prevalence estimates of persons of African ancestry with SCT (Pecker and Naik, 2018). The mean age of women with and without SCT was similar: 50.59 (SD 14.04) and 50.38 (SD 14.22) respectively. The mean number of prior pregnancies was also similar: 2.92 (SD 1.57) in women with SCT and 2.93 (SD 1.57) in women without SCT. Among the cohort, there were a total of 171 (0.23%) pregnancy-related VTE events, which is similar to previously published estimates of 1-2 VTE events per 1,000 deliveries (Heit et al, 2005). The prevalence rate of VTE among pregnant women with SCT was 0.44%, consistent with 2.04-fold increased odds of pregnancy-associated VTE (CI: 1.31-3.18, p value 0.0016). On subset analysis, women with SCT were also more likely to have a pregnancy-associated PE than women without SCT (OR 2.3, CI 1.21-4.39, p value 0.011) and more likely to have a pregnancy-associated isolated DVT than women without SCT (OR 1.85, CI 1.01-3.4, p value 0.046). Conclusion: In this largest study to date, we found a twofold increased risk of pregnancy-associated VTE among women with SCT as compared to those without. Given the rarity of pregnancy-associated VTE, our study is the first to have sufficient power to detect a statistically significant increased risk of VTE in pregnant women with SCT. Given that the absolute risk of pregnancy-associated VTE in those with SCT is still quite small at 0.44%, this data likely does not support routine thromboprophylaxis for SCT-positive individuals during pregnancy or postpartum in the absence of other strong risk factors for thrombosis. Nonetheless, our analysis provides important information regarding how SCT contributes to the risk of VTE in pregnancy. Limitations include the use of self-report via survey. Further analysis of an entire cohort of all ancestries is ongoing.