SAFETY AND EFFECTIVENESS OF ANTICOAGULATION BASED ON AN INSTITUTIONAL RECOMMENDATION ON THE MANAGEMENT OF CANCER ASSOCIATED VENOUS THROMBOSIS

Abstract

Cancer associated venous thrombosis (CAT) can affect 20% of oncologic patients. Its management is challenging due to coexistent risk of bleeding and thrombosis (VTE) recurrence, that are 6 and approximately 4 fold higher compared to non-cancer patients, respectively. Anticoagulant choice needs to be careful and can be guided by a recommendation to avoid those complications. To evaluate effectiveness and safety of anticoagulation according to a public cancer institution recommendation on CAT management. This is a retrospective cohort study carried out from October 2018 to January 2020. The recommendation on CAT management included defined criteria to select the proper anticoagulant for each patient, to make adjustments and to handle anticoagulation in challenging situations, like thrombocytopenia, in the perioperative period and facing bleeding and thrombosis recurrence. A 3-month ambulatory was scheduled to evaluate eligibility of outpatients on enoxaparin to switch to rivaroxaban (new at the institution), based on criteria defined by the recommendation. All consecutive adult oncologic outpatients on therapeutic enoxaparin in the institution were considered eligible for the ambulatory. Retrospective data on cancer, comorbidities, bleeding, VTE recurrence and death during 1 year since the first visit were captured from an electronic recording. Cumulative incidence of bleeding and VTE recurrence were calculated by Kaplan-Meier method and log-rank test. 326 patients were referred to the ambulatory and 268 were included in the analyses, with a median age of 63 years (IQR 53-71). Sixty-one percent were female and 86% had solid tumours (53% metastatic). The main sites were the gastrointestinal tract (27%), breast (12%) and lungs (10%). Forty-six percent of the patients switched to rivaroxaban and 6% were able to stop anticoagulation. From the 85 patients who remained on enoxaparin, 51.7% had some dosing adjustment (to 1.5 mg/kg/day or to prophylactic dose). The main reason for not switching to rivaroxaban was drug-drug interaction (27%). One-year cumulative incidence of VTE recurrence was 11.2% (95% CI 7.7-16.2) and of overall bleeding was 13.4% (95% CI 9.4-18.7). Major bleeding occurred in 8 out of 268 patients (2.9%). Death rate was 32%, with no one due to bleeding. No significant difference in outcomes was observed between patients who had switched or not. Severity of cancer status was considered relevant for the high rate of deaths. Cumulative incidence of bleeding as well as thrombosis recurrence were similar to the described in the literature, although many clinical trials on direct oral anticoagulants in cancer patients had a shorter period of follow-up. Recommendation-based decision on anticoagulant choice and management showed similar effectiveness and safety independently of the chosen anticoagulant. No relevant hazards due to anticoagulation were observed during the follow-up period.