Danny R. Rose, Jr., MD
Aguiar de Sousa D, Canhão P, Ferro JM. Safety of Pregnancy After Cerebral Venous Thrombosis: A Systematic Review. Stroke. 2016
Cerebral venous thrombosis (CVT) is one of the most common causes of stroke during pregnancy and the puerperium, accounting for between one fourth and one half of all strokes during that period. Risk of CVT and other venous thrombotic events (VTEs) is increased during pregnancy, and women with a previous episode of non-cerebral VTE have a threefold to fourfold higher risk of VTE during subsequent pregnancies. Although thrombosis and thromboembolism is the leading cause of maternal mortality in developed countries, few studies have specifically investigated the safety of pregnancy in women with prior CVT. Aguiar de Sousa et al. performed a systematic review to ascertain the risk of CVT recurrence, occurrence of non-cerebral VTE, risk to the viability of pregnancy and the effect of antithrombotic prophylaxis.
The authors performed a systematic search of Pubmed, Cochrane Database of Systematic Reviews and clinicaltrials.gov for observational studies, excluding case reports or selected case series, that included data suitable to derive the relative risk of CVT recurrence or non-cerebral VTE associated with subsequent pregnancy. Studies were assessed for quality of design and data reporting according to the GRADE approach. The primary outcome was the frequency of CVT and non-cerebral VTEs related to pregnancy in patients with a history of prior CVT. When available, subgroup analysis was performed for patients given antithrombotic prophylaxis. The risk of spontaneous abortion, stillbirth, and preterm birth were assessed as secondary outcomes. Pooled frequencies and 95% confidence intervals were calculated, using relative risk in previously described incidences in the general population, as no case control studies were identified. A simple pooled analysis of individual patient data was utilized, although a meta-analysis of proportions was also conducted. Heterogeneity among studies was assessed with Cochrane’s Q-test.
A total of 13 studies were included from an initial pool of 523. Of these, 9[L1] studies included long term follow-up of patients with CVT and 5 studies evaluated complications during subsequent pregnancies in women with prior cerebrovascular disease. All but one of the studies were scored as having “moderate” methodological quality, with none considered to have unacceptable quality. Out of the 217 pregnancies in which CVT recurrence was assessed, two cases of recurrent CVT were reported (9 per 1,000 pregnancies; 95% CI 3-33 per 1,000). Compared to general population data obtained from the Healthcare Cost and Utilization Project, the relative risk was 85 (95% CI 21-339) after sensitivity analysis. The occurrence of non-cerebral VTE was assessed in 186 pregnancies, with 5 events reported (27 VTEs per 1,000 pregnancies CI 95% 12-61). This resulted in a relative risk of 16 (95% CI 7-39) using sensitivity analysis with the general population data as utilized above.
The authors did not find sufficient data regarding antithrombotic prophylaxis for either of the above outcomes due to incomplete reporting. Among 186 reported pregnancies, 33 resulted in spontaneous abortion, with a calculated overall crude risk of 17.7% (95% CI 13-24). As there were no case control studies, the authors compared this value to the 10-15% rate of spontaneous abortion that has been previously reported in cohort studies. The rate of abortion was slightly higher in patients without antithrombotic prophylaxis (19% versus 11%), although the number of pregnancies that had associated data regarding prophylaxis was very low.
This study represents the first comprehensive systematic review of the risk of recurrent CVT and VTE in pregnancy in patients with a history of prior CVT. Although there was a clear increase in the relative risk of recurrent CVT (80 times) and extracerebral VTE (10 times) in this patient population, the absolute risk remained low. The frequency of spontaneous abortion was slightly higher than published population data, and there was a trend towards lower rates in women treated with heparin prophylaxis. Despite heterogeneity in study design and data collection methods in the included studies, the authors’ choice to use objective outcomes and the lack of statistical heterogeneity strengthened their results.
It is unfortunate that the outcome that would have the potential to influence clinical practice (effect of antithrombotic prophylaxis) could not be adequately assessed due to the method of data collection in the studies reviewed. A larger study evaluating VTE, pregnancy outcomes and prophylaxis complications could be useful in this regard. Inconsistent and incomplete reporting of the etiology of the index CVT represents another gap in data reporting with the potential to influence clinical practice. Given that the two reported cases of recurrent CVT were present in patients with pre-existing prothrombotic states (Protein S deficiency and sickle-cell anemia, respectively), more thorough data collection in a larger study could aid clinicians in risk stratifying patients to identify those who may be suitable for more aggressive prophylactic measures.