Predictors of venous thromboembolism recurrence and bleeding among active cancer patients: a population-based cohort study.

Active cancer is the major predictor of venous thromboembolism (VTE) recurrence, but further stratification of recurrence risk is uncertain. In a population-based cohort study of all Olmsted County, Minnesota, residents with active cancer-related incident VTE during the 35-year period from 1966 to 2000 who survived 1 day or longer, we estimated VTE recurrence, bleeding on anticoagulant therapy, and survival and tested cancer and noncancer characteristics and secondary prophylaxis as predictors of VTE recurrence and bleeding, using Cox proportional hazards modeling. Of 477 patients, 139 developed recurrent VTE over the course of 1533 person-years of follow-up. The adjusted 10-year cumulative VTE recurrence rate was 28.6%. The adjusted 90-day cumulative incidence of major bleeding on anticoagulation was 1.9%. Survival was significantly worse for patients with cancer who had recurrent VTE (particularly pulmonary embolism) and with bleeding on anticoagulation. In a multivariable model, brain, lung, and ovarian cancer; myeloproliferative or myelodysplastic disorders; stage IV pancreatic cancer; other stage IV cancer; cancer stage progression; and leg paresis were associated with an increased hazard, and warfarin therapy was associated with a reduced hazard, of recurrent VTE. Recurrence rates were significantly higher for cancer patients with 1 or more vs no predictors of recurrence, suggesting these predictors may be useful for stratifying recurrence risk.