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


Patient Population and AF/AT Recurrence

In total, 398 patients were included into this study. Further details on the source population are shown in Figure 1. The baseline characteristics are shown in Table 1 according to recurrence of AF/AT. Median FU was 528.5 (373, 111,3.5) days postablation. The detailed distribution of the CHADS2- and CHA2DS2-VASc-scores is shown in Table 2.

Figure 1.

Flow diagram describing the source population. AF = atrial fibrillation; FU = follow-up.

In total, 92.7% were taken off AAD. Recurrence of AF/AT after 1 year of FU occurred in 55 patients (13.8%). The number of patients with persistent AF was significantly higher among those with recurrence during FU (49.1% vs. 26.8%; P < 0.001) and there was a trend toward more diabetes in patients with recurrence (20% vs. 11.1%; P = 0.06). Detailed characteristics comparing patients with and without recurrence are shown in Table 1.

Univariable and multivariable binary regression was performed to identify possible predictors of recurrence in our patient cohort. Persistent AF was the only significant predictor of AF/AT recurrence in univariable, binary analysis (OR 2.63, P < 0.001; Table 3). Analysis of the ROC curve revealed that the CHADS2- (AUC 0.5 ± 0.043; P = 0.97) and CHA2DS2-VASc-scores (AUC 0.52 ± 0.044; P = 0.66) were poor diagnostic tools to predict recurrence of AF/AT. In a multivariate model (including persistent AF, age, hypertension, diabetes, discontinuation of antiarrhythmic therapy and time in FU) persistent AF stayed an independent predictor of AF/AT recurrence.

Incidence of Stroke

Anticoagulation was discontinued in 69.3% of the patients. Patients in whom OAC was discontinued had significantly lower CHADS2- (0.6 [±0.66] vs. 1.5 [±1]; P < 0.001) and CHA2DS2-VASc-scores (1 [±1] vs. 2.5 [±1.3]; P < 0.001) and were less likely to have persistent AF (24.6% vs. 41.8%; P = 0.001). Of the 398 patients in this study, 262 (65.8%) would have had a recommendation for continued OAC according to the 2016 CCS recommendations for anticoagulation[20] (Table 4). OAC was discontinued in the majority (58%) of these patients. When using recommendations from the current guidelines of the American Heart Association[21] and the European Society of Cardiology,[22] 290 (72.9%) and 264 (66.3%) patients would have an indication for continued OAC. Again, OAC was discontinued in the majority of these patients (Table 4).

During the FU period, 4 patients (1%) suffered from stroke resulting in a total number of 0.45 strokes per 100 patient years. There were no significant differences in baseline characteristics between the groups. There were no significant differences in the baseline characteristics of patients that did or did not suffer from stroke. However, there was a trend toward a higher percentage of CAD among patients with stroke (50% vs. 10.2%; P = 0.057). Two of the strokes occurred in patients with a CHA2DS2-VASc score of 3 (total n = 58 with CHA2DS2-VASc of 3), giving a stroke rate of 3.4% in this specific group. The stroke rate in patients with a CHA2DS2-VASc score of 3 or more is 2/85 or 2.4%.

Detailed information for the individual patients who had a stroke and/or TIA is shown in Table 5. Of these 4 patients, 3 were still on anticoagulation when they experienced stroke or TIA. Patient #1 developed an isolated weakness of his right hand and arm that completely resolved within 10–12 hours. His OAC was subtherapeutic at admission but cerebral imaging did not reveal significant pathologies. After the TIA, his OAC was changed from Coumadin to dabigatran. Patient #2 was admitted to the hospital with right-sided upper extremity motor and sensory disturbances. At this time he was on OAC with warfarin and had therapeutic INR levels on admission. A CT scan showed no significant abnormalities, especially no bleeding. In patient #3, anticoagulation had been discontinued because of the long duration of AF-free survival. However, more than 4 years after her last ablation, she presented to the hospital with left-sided weakness and palpitations. On admission, AF was documented and cerebral imaging revealed a thrombotic occlusion of the M2 segment of the right middle cerebral artery that was successfully treated with thrombolysis. OAC was re-initiated after the event. Patient #4 underwent a coronary angiogram and felt dizziness and a loss of speech for several minutes shortly after the procedure. She was on OAC with dabigatran and a head CT remained without pathologies. Doppler ultrasound of the carotid arteries did not reveal a significant stenosis in any patient that experienced stroke.