Clinical history and antithrombotic treatment of incidentally detected splanchnic vein thrombosis: a multicentre, international prospective registry

Background Little information is available about the clinical history of patients with incidentally detected splanchnic vein thrombosis and its therapeutic management remains controversial. The aim of this study was to assess the risk factors, therapeutic strategies, and long-term outcomes of incidentally detected splanchnic vein thrombosis. Methods We analysed data from patients with incidentally detected splanchnic vein thrombosis who were enrolled in an international, multicentre, prospective cohort study of splanchnic vein thrombosis between 2008 and 2012. The study was done at 31 centres in 11 countries (Italy, South Korea, Germany, Canada, Belgium, the Netherlands, Brazil, USA, France, Israel, UK). Information about demographic characteristics, risk factors, and treatment was collected. The study outcomes during the 2-year follow-up were major bleeding (International Society on Thrombosis and Haemostasis definition plus the need for hospital admission), thrombotic events (venous or arterial thromboses), and mortality. The primary analysis period was from the diagnosis of incidentally detected splanchnic vein thrombosis until the first adjudicated clinical outcome or the end of follow-up. Findings Between May 2, 2008, and Jan 30, 2012, we enrolled 177 patients with incidentally detected splanchnic vein thrombosis (median age 57 years [IQR 49–66], 118 [67%] men, 138 [78%] patients with portal vein thrombosis). The most common underlying diseases were liver cirrhosis (82 [46%] patients) and solid cancer (62 [35%] patients). Anticoagulant treatment was prescribed to 109 (62%) patients. Median duration of anticoagulation was 6 months (IQR 5–12) for patients who received parenteral anticoagulants alone and 24 months (IQR 12–24) for patients treated with vitamin K antagonists. During a median follow-up of 2 years (IQR 1–2), the incidence of major bleeding was 3·3 events (95% CI 1·7–6·3) per 100 patient-years and the incidence of thrombotic events was 8·0 events (95% CI 5·2–12·1) per 100 patient-years. On-treatment incidence was 3·2 events (95% CI 1·2–8·4) per 100 patient-years for major bleeding and 3·9 events (95% CI 1·6–9·5) per 100 patient-years for thrombotic events. In multivariate analysis, anticoagulant treatment as a time-dependent variable reduced the incidence of thrombotic events (hazard ratio 0·85, 95% CI 0·76–0·96) without increasing the risk of major bleeding (p>0·05). In patients with clinically suspected splanchnic vein thrombosis, the incidence of major bleeding was 3·9 events (95% CI 2·6–5·7) per 100 patient-years and the incidence of thrombotic events was 7·0 events (95% CI 5·2–9·3) per 100 patient-years. Interpretation Our results show that the prognosis of incidentally detected splanchnic vein thrombosis is similar to that of clinically suspected splanchnic vein thrombosis and suggest that similar treatment strategies should be applied. Funding Pfizer Canada research grant.