IMPROVE bleeding score predicts major bleeding in advanced gastrointestinal cancer patients with venous thromboembolism

Abstract

Abstract Background The incidence of venous thromboembolism has been reported as 20% in cancer patients. Anticoagulation therapy is the standard treatment for venous thromboembolism. On the other hand, bleeding should be carefully managed, because advanced cancer, particularly gastrointestinal cancer, carries a high risk of bleeding. However, the optimal management for cancer-associated thromboembolism remains to be clarified. Methods We retrospectively examined patients with advanced gastrointestinal cancer, including gastric cancer and colorectal cancer, who were treated with chemotherapy between 2014 and 2018 for the incidence and characteristics of venous thromboembolism and bleeding. Results In total, 194 patients (120 men, 74 women) were enrolled in this study. The underlying pathology was gastric cancer in 74 cases and colorectal cancer in 120 cases. Of the 194 patients, 40 patients (20.6%) were diagnosed with venous thromboembolism and 10 patients (5.2%) were diagnosed with concomitant pulmonary thromboembolism. Conversely, bleeding was observed in 29 patients (15%). The location of bleeding was the primary tumor in 17 cases, metastatic tumor in 9 and hemorrhagic gastric ulcer in 3. Within the venous thromboembolism group (n = 40), bleeding was observed in 10 patients (25%). Multivariate analysis showed that International Medical Prevention Registry on Venous Thromboembolism (IMPROVE) bleeding score ≥7 correlated significantly with major bleeding (P = 0.01). In patients with a low risk of bleeding, major bleeding was observed in only three patients. Conclusions IMPROVE bleeding score may predict the risk for bleeding in gastrointestinal cancer patients with venous thromboembolism. Selecting patients with a low risk of bleeding using with IMPROVE bleeding score is expected to contribute to the safer management of anticoagulation therapy for cancer-associated thromboembolism.