Risk of Stroke and Recurrence After AF Ablation in Patients With an Initial Event-Free Period of 12 Months

Simon Kochhäuser, M.D.; Pouria Alipour, B.Sc.; Tanjah Haig-Carter, B.Sc.; Kathleen Trought, B.Sc.; Philip Hache, B.Sc.; Yaariv Khaykin, M.D.; Zaev Wulffhart, M.D.; Alfredo Pantano, M.D.; Bernice Tsang, M.D.; David Birnie, M.D.; Atul Verma, M.D.


J Cardiovasc Electrophysiol. 2017;28(3):273-279. 

In This Article


This study selectively evaluates the risk of stroke, AF/AT recurrence and discontinuation of anticoagulation specifically in patients that were free of AF/AT for at least 12 months after their last ablation. We found a low risk of stroke and TIA in these patients coupled with a low risk of AF/AT recurrence. In 3 of the 4 patients, the stroke occurred while still on OAC. However, in 1 patient, the stroke occurred at the time of the first AF recurrence while off OAC.

So far, several studies have retrospectively investigated the risk of stroke in patients after AF ablation,[2–11] suggesting that AF ablation might have a positive prognostic impact.

Reynolds et al. investigated 1,602 propensity-matched patients that underwent AF-ablation or were treated with AAD.[10] The rate of stroke or TIA was significantly lower in patients that underwent ablation. Age over 65 years and a history of prior stroke or TIA were independent, significant predictors of stroke in their patients.

In another large retrospective analysis of 766 patients after AF ablation with a very long median FU of over 5 years,[11] the overall rate of thromboembolic events was low and was in fact lower than in the general AF population. This is especially interesting because OAC was discontinued in 65% of the patients in this study indicating that this strategy could be safe in patients after AF ablation.

In a multicenter registry with over 1,200 patients after AF ablation, the risk for stroke and death were lower than in a comparable general AF population when patients were free of AF.[3] Again, OAC had been stopped in a significant portion of patients (64%). Freedom of AF was also significantly associated with a lower risk for stroke after ablation in another retrospective study by Yagishita et al.[9]

Overall, these results suggest that the risk of stroke after AF ablation is lower than in the general AF population,[3,11] but we also know that AF ablation is not 100% effective and episodes of AF can persist postablation.[13] A possible explanation could be that ablation reduces AF burden sufficiently to reduce the risk of stroke. However, it is also possible that patients selected for ablation have unidentifiable confounding factors of better vitality that make them good candidates for ablation but also reduce their overall stroke risk. Furthermore, other factors like the structure and geometry of the left atrium[15] and the anatomy of the left atrial appendage[14] could contribute to a patient's risk of stroke. In these studies, care providers are not blinded to any data and are having individual discussions with patients who have their own values. Therefore, the decision to stop or continue OAC may be made for confounding reasons that we cannot detect. It is always difficult to generalize the findings from retrospective studies. It becomes even more difficult because patients selected for AF ablation are usually younger, have less co-morbidities when compared with the general AF population and have relatively low CHADS2- and CHA2DS2-VASc-scores. As a result the expected frequency of stroke and TIA in this population is rather low and makes it even more difficult to evaluate the value of these studies for general care. Only prospective trials including larger numbers of patients over a significant FU time can reliably confirm whether it is safe to discontinue OAC after AF ablation. Although the number of stroke in our study was low, patients with a higher CHA2DS2-VASc-score seem to be at a higher risk of stroke when compared to low-risk patients. Therefore, until results from prospective trials are available, this decision to discontinue OAC will always have to be made in conjunction with the patient after careful discussion of possible risks and consideration of the individual stroke risk.

Our study results imply that the risk of stroke after AF ablation is lowered, particularly in patients that are actually free from AF after ablation. In our group of patients, the rate of recurrence was quite low when taking into account that it included a significant number of patients with persistent AF. Of note, the median FU of patients with AF/AT recurrence was significantly longer than of those without recurrence. The most probable reason is that patients who have already had a recurrence tend to be followed more frequently and for a longer duration than patients without suspected recurrence after ablation. Furthermore, since we have a regimented AF screening process with ECGs and Holters performed at regular intervals, those who were followed-up longer had more screening and therefore likely had more AF detected.

Like other studies, we found a low overall rate of stroke or TIA although OAC was discontinued in nearly 70% of the patients. In fact, 3 of the 4 strokes in our cohort occurred while the patients were still on OAC and there was no evidence for AF/AT recurrence prior to the event. However, one of the patients was diagnosed with her first recurrence of AF at the time she was admitted to hospital for stroke. Since this episode took place over 4 years after her ablation, it highlights the need for continued vigilance for AF/AT recurrence and stroke risk in all patients that have underwent AF-ablation and may discourage the routine discontinuation of OAC.

Studies by Chao et al.[7] and Kornej et al.[23] have suggested the CHADS2 and CHA2DS2-VASc-scores as predictors of stroke after AF ablation. In our cohort, both scores were not significantly higher in patients experiencing stroke during FU. However, the low number of patients with stroke limits the statistical impact of our data and its meaningfulness in the evaluation of the predictive value of the CHADS2- and CHA2DS2-VASc-score. Interestingly, coronary artery disease was the only factor with a strong trend toward a positive correlation in our study group. Stroke and CAD share several risk factors and are indeed risk factors for each other.[24,25] Therefore, the high prevalence of CAD in patients with stroke in our study might indicate that atherosclerotic, not AF-related causes of stroke also play an important role in the long-term FU of patients after AF ablation.

Persistent AF was a strong, independent predictor of AF/AT recurrence in our study with a 2-fold increase in the risk of recurrence. Taking the case of stroke patient #3 into account, persistent AF should therefore result in an even more rigorous screening for AF recurrence.