Stroke and bleed related deaths in newly diagnosed atrial fibrillation patients: insights from the GARFIELD-AF registry

Abstract Background Atrial fibrillation (AF) is associated with a significant increase in stroke risk. Anticoagulation (AC) guidelines recommend stratification of stroke risk to aid AC choice. However, despite evidence supporting AC, the associated bleeding risk often leads to underdosing or omission of AC. Transition from stratification of stroke and bleeding risk to stratification of mortality associated with stroke and bleeding may overcome this therapeutic inertia. Purpose To quantify the risk of stroke- and bleed-related mortality in newly diagnosed AF patients according to different AC strategies. Methods GARFIELD-AF is the largest multinational, prospective AF registry worldwide. Stroke- and bleed-related deaths were defined as a death occurring within 30 days after each event. Predictors of stroke- and bleed-related deaths were identified through least absolute shrinkage and selection operator and were selected from a comprehensive list of demographic, clinical and lifestyle factors. Expected probabilities of stroke- and bleed-related death by AC strategy were extracted from the developed Cox proportional-hazards models. Results Among the 52,018 GARFIELD-AF patients, 195 stroke-related deaths and 172 bleed-related deaths occurred. Patients who suffered stroke- or bleed-related deaths were older (median [Q1; Q3]: 78.0 [72.0; 84.0] and 77.0 [70.5; 83.0]) than those who did not (71.0 [63.0; 78.0]) and had a higher prevalence of comorbidities including heart failure, vascular disease, and prior stroke. Patients who suffered a stroke-related death less frequently received vitamin K antagonists (VKAs) and non-vitamin K Antagonist Oral anticoagulants (NOACs) compared to those who were alive at two years or died of a non-stroke-related death. In contrast, patients who suffered a bleed-related death more often received VKAs compared to those who did not. NOACs and AP monotherapy treatment were less common in patients who had bleed-related death (Figure 1). Predictors for stroke-related death included age, ethnicity, heart failure, prior stroke, AC treatment, pulse, and dementia. Bleed-related mortality predictors were age, ethnicity, chronic kidney disease, AC treatment, vascular disease, and smoking status. VKAs and NOACs were associated with a lower risk of stroke-related death, reducing 2-year risk from 0.73% without AC to 0.41% and 0.36%, respectively. In contrast, bleed-related deaths increased with VKA treatment, but not with NOACs (Figure 2). The overall net benefit versus no AC treatment was greater with NOACs than VKAs. Conclusion Among AF patients at high stroke risk, NOACs and VKAs were associated with a reduced risk of stroke-related death compared to no AC, but the risk of bleed-related death was higher with VKA. This suggests that a new approach to risk stratification based on the net mortality benefits of NOAC use in newly diagnosed AF patients at high risk of stroke should be considered. Funding Acknowledgement Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): This work was supported by the Thrombosis Research Institute (London, UK)