Four-year incidence of major adverse cardiovascular events in patients with atherosclerosis and atrial fibrillation.

Abstract Background There is a paucity of contemporary data assessing the implications of atrial fibrillation (AF) on major adverse cardiovascular events (MACE) in patients with or at high‐risk for atherosclerotic disease managed in routine practice. Hypothesis We sought to evaluate the 4‐year incidence of MACE in patients with or at risk of atherosclerotic disease in the presence of AF. Methods Using US MarketScan data, we identified AF patients ≥45 years old with billing codes indicating established coronary artery disease, cerebrovascular disease, or peripheral artery disease or the presence of ≥3 risk factors for atherosclerotic disease from January 1, 2013 to December 31, 2013 with a minimum of 4‐years of available follow‐up. We calculated the 4‐year incidence of MACE (cardiovascular death or hospitalization with a primary billing code for myocardial infarction or ischemic stroke). Patients were further stratified by CHA2DS2‐VASc score and oral anticoagulation (OAC) use at baseline. Results We identified 625,951 patients with 4‐years of follow‐up, of which 77,752 (12.4%) had comorbid AF. The median (25%, 75% range) CHA2DS2‐VASc score was 4 (3, 5) and 64% of patients received an OAC at baseline. The incidence of MACE increased as CHA2DS2‐VASc scores increased (P‐interaction<.0001 for all). AF patients receiving an OAC were less likely to experience MACE (8.9% vs 11.6%, P < .0001) including ischemic stroke (5.4% vs 6.7%, P < .0001). Conclusion Comorbid AF carries a substantial risk of MACE in patients with or at risk of atherosclerotic disease. MACE risk increases with higher CHA2DS2‐VASc scores and is more likely in patients without OAC.


| INTRODUCTION
Atherosclerotic disease and atrial fibrillation (AF) share many of the same comorbidities and risk factors 1 and frequently coexist.
Compared to the general population, AF prevalence in patients with established atherosclerotic disease is fivefold higher (2.3% vs 11.7%, respectively). 2 Both are independently associated with elevated major adverse cardiovascular events (MACE) risk and understanding the synergistic increases in incidence rates and comparative risk based on disease characteristics may greatly aid treatment efforts by identifying patient factors associated with high risk. 3 The Reduction of Atherothrombosis for Continued Health (REACH) registry prospectively evaluated patients at risk for or with established atherosclerotic disease. 4 Subanalyses of REACH showed a 4-year MACE incidence of 24.3% in patients with comorbid AF. 3 They also found a linear trend in the correlation between MACE and CHA 2 DS 2 -VASc scores. However, REACH enrolled patients between 2003 and 2004 with follow-up through 2008. Advances in the management of AF in the past decade, such as the approval of non-vitamin K antagonist oral anticoagulants (NOAC) and their recommendation over warfarin in evidence-based guidelines, may have lowered MACE risk. [5][6][7] While the early hazard of MACE in AF patients has been previously evaluated, such studies only focused on incident AF patients and did not require patients to have atherosclerotic disease. 8 It is therefore important to identify MACE incidence in a more contemporary population reflecting modern medical practices. This study aimed to estimate the contemporary 4-year incidence of MACE for United States AF patients with or at risk for established atherosclerotic disease in routine practice, the association between CHA 2 DS 2 -VASc scores and MACE risk, and how oral anticoagulation (OAC) use modifies this risk.

| METHODS
We performed a retrospective claims database analysis using US IBM MarketScan data from January 1, 2013     therapy. [17][18][19] Other studies have also evaluated the association of risk factors and AF on MACE or its components. The GARFIELD-AF registry reported age, heart failure, prior stroke, vascular disease, moderate-tosevere chronic kidney disease, diabetes mellitus, and CHA 2 DS 2 -VASc score of ≥3 to be independent predictors of higher risk of early death. 8 Although this is generally consistent with our findings, several differences between the studies exist. The prospective GARFIELD-AF regis- This study has limitations worth discussing. First, misclassification bias must always be considered in claims database analyses and can detrimentally affect a study's internal validity when present. 26 Namely, our study only assessed OAC use at baseline and does not account for initiation during follow-up. Second, clinical adjudication of events was not possible within our claims database analysis. Of note, the REACH registry, while performed prospectively, also did not independently adjudicate outcomes. Third, antiplatelet agents, such as aspirin, are acquired over the counter and are difficult to track in claims data; however, OAC therapy is superior to antiplatelet therapy and is a mainstay in AF guidelines. Fourth, we were unable to differentiate between paroxysmal and persistent AF because our baseline period in 2013 used ICD-9 codes which do not make this distinction. Fifth, although we utilized a conservative definition of cardiovascular death which resulted in a smaller rate in comparison to that of REACH, survival bias cannot be ruled out.
Finally, we used US commercial and Medicare supplemental plan claims data. As a consequence, our results are most generalizable to an insured US population with established or at high-risk for atherosclerotic disease. 9

| CONCLUSION
In conclusion, this large, contemporary real-world study demonstrates the impact of comorbid AF on MACE rates in patients with or at high-risk for atherosclerotic disease. We showed that increasing CHA 2 DS 2 -VASc scores were associated with higher MACE risk and a significant number of patients will experience a MACE by 4 years. OAC usage remains relatively low, although it may be increasing in recent years. Reductions in MACE risk may be attributed to changes in practice for AF management and the adoption of NOACs. These results emphasize the continued need for optimizing anticoagulant and antiplatelet therapy in AF patients at risk of or with atherosclerotic disease.