Surveillance, anticoagulation, or filter in calf vein thrombosis

Objective This study compared the efficacy and complication rates of inferior vena cava (IVC) filters for calf vein thrombosis (CVT) vs conservative treatment with or without anticoagulation. Methods Vascular laboratory studies of patients who had an isolated CVT (anterior and posterior tibialis, peroneal, soleal, and gastrocnemius veins) from April 2009 to January 2014 were retrospectively analyzed from a single institution. Of 647 patients with isolated CVT, 285 (44%) received an IVC filter, and 362 (56%) received medical treatment alone (38.9% surveillance, 11.6% prophylactic anticoagulation, and 49.4% therapeutic anticoagulation). Univariate, multivariate, propensity matching, and Kaplan-Meier analyses were performed on abstracted data, which included, but was not limited to, risk factors, treatment modalities, venous thromboembolism (VTE) complications (defined as propagation of deep vein thrombosis [DVT] or pulmonary embolism [PE]), bleeding complications, and IVC filter-related complications (ie, filter tilting >15°, perforation >3 mm, fracture, migration >10 mm). Results The overall incidence of PE in was 2.5% in the IVC filter group and 3.3% in the medical group (P =.27). The overall incidence of VTE complications (propagation of DVT, PE) was 35% for the surveillance group without anticoagulation, 30% in patients treated with prophylactic anticoagulation, and 10% in patients treated with therapeutic anticoagulation (P =.0003). Only a minority of patients underwent duplex ultrasound imaging after filter insertion. In the IVC filter group, the most common reasons that contraindicated anticoagulation were bleeding (35%) or recent surgery (27%). The number of IVC filter-related complications in the IVC filter group was 29 (10%). Because the IVC filter group was older (mean age, 65 vs 61 years, P =.004) and more likely to have a history of thromboembolic events (56% vs 16%, P <.0001), and malignancy (49% vs 28%, P <.0001), propensity analyses were performed yielding a homogenous cohort. The overall complication and thromboembolic rates did not differ for muscular (soleal, gastrocnemius) vs tibial DVTs (anterior, posterior, peroneal veins). Conclusions The use of anticoagulation in patients with CVT significantly decreases the rates of VTE complications. The use of IVC filters in this study was associated with a 10% complication rate and did not significantly reduce the incidence of PE. Nevertheless, given the overall low rates of PE and the higher risk of VTE in patients who receive filters, the decision to insert a filter in patients with calf CVT should be individualized.