Anticoagulation in cancer patients with atrial fibrillation and grade 3-4 thrombocytopenia

Introduction: Atrial fibrillation or flutter (AF) is prevalent in cancer patients. Many of these patients have an indication for anticoagulation (AC) but are also at risk for developing chemotherapy-induced thrombocytopenia. There are scarce data regarding management of AC and risk of bleeding and thrombosis in cancer patients with AF and thrombocytopenia. Aim: To assess anticoagulation management and incidence of bleeding and arterial thromboembolism (ATE) in cancer patients with AF and grade 3-4 thrombocytopenia (platelets <50 x 10(9)/L). Methods: A retrospective cohort study included adults with active cancer, grade 3-4 thrombocytopenia and AF with CHA(2)DS(2)-VASc score >= 1. Patients were stratified according to AC discontinuation (No-AC) or continuation (Continue-AC) when platelets dropped below 50 x 10(9)/L and followed for 30 days. The study outcomes were ATE (ischemic stroke, transient ischemic attack or systemic emboli) and major bleeding. Cox proportional hazards model was used to calculate hazard ratios (HR) with death as a competing risk (Fine and Gray model). Results: The cohort included 131 patients; 90 in the No-AC group and 41 in the Continue-AC group. Patient characteristics were balanced between the groups. The 30-day cumulative incidence of ATE was 2 % [95 % CI 0.4 %-7 %] in the No-AC group and 2 % [0.2 %-11 %] in the Continue-AC group (HR 0.92 [95 % CI 0.09-9.88]). The 30-day cumulative incidence of major bleeding was 7.8 % [95 % CI 3.40 %-14.52 %] and 2.44 % [95 % CI 0.18 %-11.22 %] in the No-AC and Continue-AC groups, respectively (HR 3.29 [95 % CI 0.42-26.04]). Conclusions: The high rate of bleeding and low rate of ATE in thrombocytopenic cancer patients with AF suggests that holding AC during time-limited periods may be a reasonable approach.