The importance of cancer in the prediction of embolic and hemorrhagic risk in patients with atrial fibrillation

Abstract

Abstract Introduction Cancer is a prevalent disease and is associated with a two-fold greater risk of developing cardiovascular disease. Atrial fibrillation (AF) is a common cardiovascular comorbidity among oncologic patients and the co-occurrence of these two represents several clinical challenges. Many cancers interact with the coagulation system and the representation of this population in clinical trials with cardiovascular drugs is scarce. Thus the impact of cancer regarding clinical outcomes in patients with AF is unclear. Purpose The aim of this study was to assess how cancer influences the prediction ability of embolic and hemorrhagic risk in patients with AF. Methods The study population comprised 16,056 patients from a Spanish health area diagnosed with AF between 2014 and 2018 of whom 1,137 (7.1%) had a history of cancer. During a median follow-up of 4.9 years we assessed the relationship between cancer and bleeding and embolic events by competing risk analysis considering death a competing risk. Results After adjusting for age, sex, CHA2DS2-VASC score, HAS-BLED score and anticoagulation therapy, cancer was not associated with a higher risk of embolic events (sHR 0.73, 95% CI 0.41–1.26; p=0.256). Neither was an association between an increased rate of embolic events and cancer detected, overall (sHR 0.73, 95% CI 0.41–1.26), regarding active cancer, or any subgroup of cancer. However, cancer was associated with an increased risk of bleeding specifically in patients with active cancer (sHR 1.42, 95% CI 1.20–1.67) or prior radiotherapy (sHR 1.40, 95% CI 1.19–1.65). Both CHA2DS2-VASC and HAS-BLED scores showed a suboptimal performance to predict embolic and bleeding risk in non-anticoagulated patients with active cancer (c-statistic <0.50). Discrimination and calibration ability of ATRIA and HEMORR2HAGES bleeding risk scores was also poor in patients with cancer (c-statistic <0.6 and Brier score >0.1). Moreover, the ratio between the increase in bleeding and decrease in embolic events in anticoagulated patients was similar in those with and without cancer (5.6 vs 7.8 bleeding events, p<0.001). Conclusion Cancer was associated with an increased risk of bleeding but not an increased risk of embolic events. Neither the CHA2DS2-VASC nor the HAS-BLED scores showed a good performance for predicting embolic and hemorrhagic events in AF patients with active cancer. The embolic-hemorrhagic ratio profile of anticoagulation was similar in patients with and without cancer, and parallel to cancer-free patients, anticoagulation significantly reduced embolic events while increasing bleeding risk in oncological patients. Funding Acknowledgement Type of funding sources: None.