Risks and outcomes of gastrointestinal malignancies in anticoagulated atrial fibrillation patients experiencing gastrointestinal bleeding
Abstract Background Oral anticoagulants (OACs) may serve as a type of “screening test” for the diagnosis of occult gastrointestinal (GI) tract malignancies through a clinical presentation with bleeding. Objective We aimed to investigate the 1-year incidence and predictors of GI cancers after GI bleeeding events among patients with atrial fibrillation (AF) treated with warfarin or NOACs. Second, we aimed to compare the risk of mortality after GI cancers between patients treated with warfarin or NOACs. Methods A total of 10,845 anticoagulated AF patients who experienced hospitalizations due to GI bleeding without prior history of GI cancers were identified from the Taiwan National Health Insurance Research Database. Patients were followed up for incident GI cancers for up to 1 year. Results Within 1 year after GI bleeding, 290 (2.67%) patients were diagnosed to have GI tract cancers. More patients treated with NOACs were diagnosed to have GI cancers than those receiving warfarin (68/1,759; 3.87% [NOACs] versus 222/9,086; 2.44% [warfarin], p<0.001) with an odds ratio (OR) 1.606 (95% CI: 1.208–2.117, p<0.001). Age (OR 1.025 [95% CI: 1.012–1.037] per 1 year increment) and male sex (1.356 [95% CI: 1.050–1.700]) were independently associated with the diagnosis of GI cancers within 1 year after GI bleeding. Among 290 patients diagnosed to have GI cancers, 131 (45.2%) experienced mortality within 1 year. The risk of mortality was lower for patients treated with NOACs compared to those receiving warfarin (23.5% versus 51.8%) with an adjusted hazard ratio (aHR) 0.441 (95% CI: 0.262–0.744, p<0.001) (Figure). Conclusions Incident GI cancers were diagnosed in 1 in 37 AF patients at 1 year after OAC-related GI bleeding, which were more common among patients treated with NOACs (1 in 26) compared to warfarin (1 in 41). Detailed surveys for occult GI cancers were necessary for these patients, especially for the elderly males. Survival curves Funding Acknowledgement Type of funding source: None