Patterns of Anticoagulation Use and All-Cause of Mortality in Cancer Patients with Atrial Fibrillation

Background: There is uncertainty regarding the clinical benefit of anticoagulant (AC) treatment in patients with cancer with atrial fibrillation (AF). We aimed to evaluate the distribution and patterns of AC use and its impact on all-cause mortality in cancer patients with AF. Methods: A total of 1,653 patients with cancer diagnosed with AF were included in this retrospective cohort analysis. Multivariable logistic regression was applied to identify the clinical predictors of anticoagulant prescription. Kaplan-Meier curve with a log-rank test was used to compare the probability of survival between the AC and non-AC groups. Multivariate Cox proportional hazard regression models were implemented to evaluate the influences of various variables on all-cause death. Results: Of 1,653 patients with cancer with AF, 971 (58.7%) did not receive a prescription for AC. Among patients with CHA2DS2-VASC >= 2 in men or >= 3 in women and HAS-BLED <3, 56.5% were not prescribed AC. Rivaroxaban and warfarin are more frequently used than dabigatran, mainly in patients with lung and breast cancer. Over a median follow-up of 36 months, 776 deaths were identified. The Kaplan-Meier curve suggested that patients with AC prescriptions had better clinical outcomes. On multivariate Cox proportional risk regression analysis, AC prescription was significantly associated with a lower risk of all-cause mortality (hazard ratio, 0.78; 95% confidence interval, 0.66-0.91; P = 0.002). Conclusions: The prescription of anticoagulants in patients with cancer with AF was suboptimal. AC prescription at discharge was associated with a decreased risk of all-cause mortality. Impact: This study highlights the importance of AC prescriptions in patients with cancer with AF.