Treatment of patients with venous thromboembolism and malignant disease: should vena cava filter placement be...

It has been proposed that inferior vena cava filter placement should be the initial treatment of deep venous thrombosis (DVT) or pulmonary embolus (PE) in patients with coexisting malignant disease. We have chosen instead to selectively place filters only in patients with either a contraindication to anticoagulation therapy or a subsequent complication from anticoagulation therapy. The treatment efficacy and mortality rates in patients with concomitant malignant disease and venous thromboembolism using this approach was determined.

We retrospectively reviewed all patients at our institution with malignant disease in whom venous thromboembolism developed between August 1991 through August 1996 and identified 166 patients with PE (n = 8), DVT (n = 147), and DVT/PE (n = 11). Of these patients, 138 (83.1%) were initially treated with anticoagulation therapy, and 28 (16.9%) had primary filter placement because of contraindications to anticoagulation therapy (10 for intracranial tumors, 11 for recent or upcoming operations, 6 for recent hemorrhage, and 1 for a malignant bloody pericardial effusion).

Thirty-two (23%) of the 138 patients who initially underwent anticoagulation therapy subsequently required a filter for the following reasons: bleeding (n = 15, 10.9%); recurrent thromboembolism (n = 6, 4.3%); heparin-induced thrombocytopenia (n = 1, 0.7%); and perceived high risk for bleeding with continued anticoagulation therapy (n = 11, 8%). Both bleeding and recurrent thromboembolism developed in 1 patient. Sixty patients (36%) received filters. No major technical complications occurred from filter placement. Major recurrent thromboembolic complications developed in 10 patients: DVT (n = 6, 10%), PE (n = 2, 3.3%), inferior vena cava thrombosis and phlegmasia cerulea dolens (n = 1, 1.7%), superior vena cava thrombosis (n = 1, 1.7%). Venous gangrene developed in 1 patient with DVT. The 1-year actuarial survival rates for patients treated with filter and anticoagulation therapy were 35% and 38%, respectively (P = NS).

In summary, our experience suggests that 64% of patients with malignant disease and venous thromboembolism are effectively treated with anticoagulation alone; 17% require primary filter placement for standard indications, and an additional 19% require subsequent filter placement because of complications (primarily bleeding) or failure of anticoagulation therapy. Although technical complications of filter placement are low, serious life-threatening or limb-threatening thromboembolic complications developed in 17% of patients. Survival was poor in all patients, regardless of treatment. These data support a conservative approach of routine anticoagulation therapy with selective filter placement.