Time‐trends and treatment gaps in the antithrombotic management of patients with atrial fibrillation after percutaneous coronary intervention: Insights from the CHUM AF‐STENT Registry

Abstract Background The management of atrial fibrillation and flutter (AF) patients undergoing percutaneous coronary intervention (PCI) has undergone a rapid recent evolution. In 2016, the Canadian Cardiovascular Society (CCS) published expert recommendations to help guide clinicians in balancing bleeding and thrombotic risks in these patients. Hypothesis Antithrombotic regimen prescriptions for AF patients undergoing PCI evolved after the publication of the 2016 CCS AF guidelines. Methods A prospective cohort of AF patients undergoing PCI with placement of a coronary stent from a single tertiary academic center was analyzed for the recommended antithrombotic regimen at discharge. Prescribing behavior was compared between three time periods (Cohort A [2010‐2011]; Cohort B [2014‐2015]; Cohort C [2017]) using the χ 2 test. In addition, antithrombotic management in Cohorts B and C were compared to guideline‐recommended therapy. Results A total of 459 patients with AF undergoing PCI were identified. Clinical and procedural characteristics were similar between cohorts, with the exception of an increase in drug‐eluting stent (DES) use over time (P < .01). Overall, the rate of oral anticoagulation (OAC) increased over time (P < .01), associated with an increase in nonvitamin K OAC prescription (P < .01) and a concomitant decrease in vitamin K antagonist prescription (P < .01). Despite this, the overall rate of anticoagulation remains below what would be predicted with perfect guideline compliance (75% vs 94%, P < .01). Conclusion There has been a dramatic shift in clinical practice for AF patients requiring PCI, with increases in prescription of OAC even in the context of an increase in the use of DES. However, room for further practice optimization still exists.

Conclusion: There has been a dramatic shift in clinical practice for AF patients requiring PCI, with increases in prescription of OAC even in the context of an increase in the use of DES. However, room for further practice optimization still exists.

K E Y W O R D S
acute coronary care, antiplatelet therapy, percutaneous coronary intervention, atrial fibrillation

| INTRODUCTION
Contemporary antithrombotic management of patients with either atrial fibrillation/flutter (AF) or coronary artery disease (CAD) has largely been well defined in clinical guidelines. [1][2][3][4] However, up to 30% of patients with AF also have CAD 5 and the optimal management of AF patients requiring percutaneous coronary intervention (PCI) has, up until recently, been less clear. While oral anticoagulation (OAC) is indicated for the prevention of stroke and systemic embolism in most cases of AF, 6 dual antiplatelet therapy (DAPT) is recommended after PCI in patients without AF. 7,8 Simply combining these two recommendations in patients with AF requiring PCI (so-called triple antithrombotic therapy, TATT) increases the bleeding risk significantly. 9 In 2016, both the Canadian Cardiovascular Society (CCS) and European Society of Cardiology (ESC) published expert recommendations to help guide clinicians in balancing bleeding and thrombotic risks in these patients. 1,3 The landmark PIONEER AF-PCI 10 was also published in 2016, followed closely by REDUAL 11 and then AUGUSTUS, 12 providing further evidence in support of nonvitamin K oral anticoagulation (NOAC)-based antithrombotic regimens that could minimize the bleeding risk in AF patients having benefitted from PCI.
A recent international multicenter analysis demonstrated that the availability of newer antiplatelet and anticoagulant agents was associated with a significant increase in practice variability in the management of AF patients post-PCI, but also that a major shift in clinical practice would be necessary in order to align with AF guidelines. 13 We therefore sought in this analysis to determine whether the publication of the 2016 CCS and ESC guidelines, in conjunction with landmark clinical trials, were associated with such a change in practice patterns and to assess the size any residual treatment gap.

| METHODS
We conducted a single-center retrospective cohort analysis of a prospective registry in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. 14 The need for   All statistical analyses were conducted using SAS 9.3 statistical software (SAS Institute, Cary, North Carolina). A two-tailed P-value <.05 was considered statistically significant without correction for multiple analyses.

| RESULTS
A total of 459 patients with AF undergoing PCI were included across all three cohorts. Clinical and procedural characteristics of patients in the Cohort A (n = 109), Cohort B (n = 246), and Cohort C (n = 104) are detailed in Table 1 and were by and large similar between cohorts, with the exception of an increase in the use of drug-eluted stent (DES) over time (37% vs 61% vs 94%, P < .01). The in-hospital mortality rate was 3% overall, and in-hospital major bleeding was 2% (Table 1).
Antithrombotic prescriptions at both admission and discharge in each cohort are shown in Table 1. There was a significant increase in baseline use of OAC between the pre-and postguidelines cohorts (P < .01) despite a decrease in VKA use (P < .01) due to a marked rise in NOAC use over time (P < .01). A significant increase in P2Y12 inhibitor use at baseline was also observed (P < .01).
Discharge antithrombotic prescriptions also evolved significantly over time. The rate of OAC use at discharge was significantly higher in Cohort C compared to the preguidelines cohorts (P < .01), driven by a significant increase in use of NOAC (P < .01) at the expense of postprocedure VKA prescription (P < .01). Consequently, TATT prescription increased significantly (P < .01), whereas DAPT prescription at discharge was reduced (P < .01). The emergence of a dual pathway (anticoagulant plus a P2Y12-inhibitor) prescription pattern was also observed in the most recent (postguidelines) cohort (Cohort C).
"Real-world" and corresponding theoretical CCS guidelinerecommended OAC rates (based on the patient characteristics in each of Cohorts B and C) are presented in   prescription rate is in agreement with the CCS 2016 AF guidelines recommendation of TATT for 3 to 6 months in these patients with CHADS2 score ≥ 2, placing greater weight on reduction of thromboembolic events and comparatively lesser weight on risk of major bleeding. 3 A course of TATT of a duration of up to 6 months in patients at high risk of thrombosis was also advocated subsequently in the 2018 update of the CCS antiplatelet guidelines. 16 The emergence of dual pathway antithrombotic therapy (anticoagulant plus a single antiplatelet agent) in clinical practice, on the other hand, represents an integration of randomized trial data from PIONEER AF-PCI (rivaroxaban) and REDUAL (dabigatran) that showed that such a regimen could minimize bleeding risk without a signal for increase in clinical ischemic events. 10,11 A shift to dual pathway antithrombotic management is also advocated in the 2018 updates of the CCS antiplatelet and atrial fibrillation guidelines. 16,17 The recently published AUGUSTUS trial (apixaban), that also included medically managed ACS patients, also supports the safety advantage of dual pathway therapy over triple therapy. 12 The ENTRUST-PCI AF trial (edoxaban) furtherly reinforced the safety and anti-ischemic efficacy of dual pathway regimens of dual pathway over triple therapy, with no significant difference in ischemic events between the two groups. 18 Interventional cardiologists at our center no longer appear to be avoiding DES in AF patients when anticoagulation is indicated.  24,25 Avoidance of restenosis with the use of DES may also help reduce the risk of bleeding complications by avoiding repeat procedures in a typically fragile AF population. 26

| LIMITATIONS
Certain limitations must be acknowledged given the retrospective nature of this analysis. First, while the registry is prospective and ongoing, data were abstracted from patients' medical records, giving rise to the possibility of ascertainment bias. Secondly, there is the potential for some "noise" around the ACS presentation variable due to the likely inclusion of some cases of crescendo angina as "unstable." Nevertheless, we believe the impact of this variability to be minimal.
Furthermore, the type of presentation (ACS vs non-ACS) would not affect the recommended antithrombotic therapy prescribed at discharge according to the 2016 AF guidelines (though it would impact the duration). Finally, as this study was conducted in a single tertiary academic center, these results are not necessarily indicative of clinical practice in community centers or other Canadian academic centers.

| CONCLUSION
While the impact of the availability of novel antithrombotic agents without clinical guidelines lead to increased practice variability, the combination of the 2016 CCS AF guidelines and landmark clinical trials appears to have had a major impact on antithrombotic regimen prescriptions for AF patients undergoing PCI at our center, with significantly higher rates of TATT and dual pathway regimen prescription in the most recent cohort. Guideline adherence was high overall, but room for improvement still exists, particularly in light of the most recent guidelines updates.