Despite the efficacy of catheter ablation in increasing numbers of patients, it is unclear if it impacts the subsequent risk of stroke and whether oral anticoagulation (OAC) can be discontinued. There are a number of issues to consider. First, patients who seem to have be free of atrial fibrillation (AF) after ablation may still have late recurrences of AF (2-3%/year).[1,2] Second, some apparently AF free patients may have intermittent asymptomatic episodes. For example in the DISCERN AF study, ablation reduced AF burden, but 80% of remaining episodes were asymptomatic. Third, we do not know the thromboembolic risk of extensively ablated left atrial tissue. Finally, there is the concern that AF may be an epiphenomenon and not directly related to stoke; for example sub-studies of AFFIRM and other trials also suggested that strokes continued to occur even in patients who were in sinus rhythm for a long period of time.[5–7] The 2017 consensus document on AF ablation recommends "adherence to AF anticoagulation guidelines for patients who have undergone an AF ablation procedure, regardless of the apparent success or failure of the procedure."
In this issue of the Journal, Romero et al present a systematic review and meta-analysis of OAC after catheter ablation of AF. They found that OAC continuation was associated with a significant decrease in the risk of thromboembolic events (TE) in the high-risk cohort (CHA2DS2VASC ≥ 2; risk ratio 0.41, 95% Confidence Interval [CI] 0.21-0.82, P = .01). No significant benefit from continued OAC was observed in the low-risk cohort. Intra cranial hemorrhage was significantly higher in the OAC group (risk ratio 5.78, 95% CI, 1.33-25.08, P = .02).
These conclusions seem relatively straightforward until it becomes clear that there are two previously published meta-analyses, purportedly examining the same clinical question, and finding contrasting results.[10,11] Both Deng et al and Proietti et al found no difference in TE risk between stopping or continuing OAC. The table highlights the key points of the three analyses and gives us some clues as to why there might be different conclusions. Perhaps most importantly, the new study by Romero et al required CHA2DS2VASC to be reported and stratified their analyses by risk category, whereas Deng et al did not perform a sub-analysis stratified by CHADS or CHA2DS2VASC score. Prioetti et al did perform this analysis pooling data from three studies and found a nonsignificant trend in the same direction as Romero's finding, i.e. favouring continued OAC in patients with higher stroke risk. The individual studies also have additional limitations beyond the methodology of the meta-analysis. Most importantly, none of the included studies were randomized. A substantial proportion of patients included had a very low stroke risk and would not have warranted OAC treatment even with ongoing AF. Furthermore, many of the individual studies were either retrospective or used administrative datasets with no independent endpoint adjudication Table 1.
So how should clinicians advise their patients? In our own practice we discuss all these details and then state that the bottom line is that there are three options to consider. The first option is to follow the 2017 guideline, the second is to "flip a coin" and either continue or stop OAC. The final option, and the one that we obviously favor, is to scientifically "flip a coin" and be enrolled in the Optimal Anticoagulation for Higher Risk Patients Post-Catheter Ablation for Atrial Fibrillation (OCEAN) trial (NCT02168829) trial. The OCEAN trial is randomizing 1572 patients more than 1 year after successful ablation with a primary endpoint of overt stroke, covert stroke (assessed by MRI), and systemic embolism and is expected to report in 2023. There is another similar ongoing trial Prevention of Silent Cerebral Thromboembolism by Oral Anticoagulation with Dabigatran After PVI for Atrial Fibrillation (ODIn-AF) trial (NCT02067182). These trials should in time give us the definitive answer to the question of whether to continue or stop oral anti-coagulation in higher risk patients after AF ablation success.
J Cardiovasc Electrophysiol. 2019;30(8):1258-1260. © 2019 Blackwell Publishing