THE ASSOCIATION OF MALIGNANCY WITH STROKE AND BLEEDING IN PATIENTS WITH ATRIAL FIBRILLATION: A POPULATION-BASED STUDY

Background It is undetermined if malignancy independently increases stroke risk in atrial fibrillation (AF). Objectives The study sought to determine the association of malignancy with stroke and bleeding in AF. Methods Population-based cohort study using administrative datasets of people aged ≥66 years with newly diagnosed AF. People diagnosed with malignancy within 5 years before AF diagnosis were matched to cancer-free control subjects on age, sex, AF diagnosis details, CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, sex category) score, and ATRIA bleeding score. Outcomes included hospitalizations for stroke and hospitalization/emergency visits for bleeding. Cause-specific regression was used to determine the HR for malignancy after adjusting for time-varying anticoagulation status. Analyses were repeated for specific subgroups of cancer patients (with matched control subjects). Results Among 199,710 AF patients, 24,991 (12.5%) people had prior malignancy. Malignancy was associated with more inpatient diagnoses of AF (vs outpatient) and less anticoagulation. We matched 43,802 people with AF (21,901 with malignancy, mean age 78.1 years; 59.5% male). After adjusting for anticoagulation status, malignancy had a similar hazard of stroke (HR: 1.01; 95% CI: 0.88-1.15) but higher hazard of bleeding (HR: 1.45; 95% CI: 1.37-1.53) compared with cancer-free control subjects in the matched sample. Analyses of cancer subgroups with comparison to matched control subjects mostly showed consistent results, except for: 1) increased hazard of stroke in lung cancer; and 2) lack of increased bleeding hazard in breast cancer and lymphoma. Conclusions People with AF and malignancy generally had similar hazards of stroke but higher hazards of bleeding compared with cancer-free control subjects, suggesting that malignancy should not lower the threshold for anticoagulation in AF.