Oral Anticoagulation Therapy After RFA of AF and Risk
Final study population comprised 4050 patients undergoing first-time RFA. The median age was 59.5 years (IQR: 52.9–65.2); 73.5% were men. According to the CHA2DS2-VASc score, 1275 (31.5%) low-risk, 1268 (31.3%) intermediate-risk, and 1507 (37.2%) high-risk patients were identified. Median follow-up duration was 3.4 years (IQR: 2.0–5.6). Table 1 shows the baseline patient characteristics. After the 3 months blanking period, 2179 (53.8%) of patients had recurrent AF. Patients underwent a total of 5983 and a mean of 1.48 (SD 0.7) RFA procedures.
Oral anticoagulation was initiated in 3707 (91.5%) patients after RFA. Figure 1 depicts long-term persistence of OAC treatment according to thromboembolic risk profile. At Day 90, 1181 (97.3%) of low-risk, 1140 (94.7%) of intermediate-risk, and 1450 (95%) of high-risk patients were on OAC therapy. Corresponding numbers at 1 year were lower; 55.7, 67.4, and 70.4%, respectively. Approximately 70% of high-risk patients and half of the total RFA population received OAC therapy beyond the first year.
(Enlarge Image)
Figure 1.
Persistence in use of oral anticoagulation therapy after first-time radiofrequency ablation according to the CHA2DS2-VASc score: the proportion of patients receiving oral anticoagulation therapy by years.
After discharge from first-time RFA, 103 (2.5%) patients experienced thromboembolism (71 (68.9%) stroke, 27 (26.2%) TIA, and 5 (4.9%) peripheral arterial embolism) corresponding to an incidence rate of 0.80 [95% confidence interval (CI): 0.66–0.98] per 100 person-years. Of note, no patients in this cohort underwent left atrial appendage occlusion.
Figure 2 shows the incidence rates of thromboembolism and serious bleeding according to time intervals after first-time RFA, illustrating high event rates for both outcomes immediately after RFA procedure. Among the 103 thromboembolism cases, 32 (31%) patients developed thromboembolism [incidence rate 3.27 (95% CI: 2.31–4.62) per 100 person-years] during the 3 months blanking period. After 3 months, 71 (69%) patients experienced thromboembolism [incidence rate 0.60 (95% CI: 0.48–0.76) per 100 person-years]; only 37 (52.1%) of thromboembolic events occurred in high-risk patients. In the 111 cases with serious bleeding, events during and after the blanking period occurred in 24 (21.6%) and 87 (78.4%) cases, respectively, corresponding to incidence rates of 2.45 (95% CI: 1.64–3.65) and 0.73 (95% CI: 0.60–0.90) per 100 person-years. The incidence rate of serious bleeding was 0.60 (95 CI: 0.46–0.80) per 100 person-years after the blanking period when patients with aspirin therapy were excluded.
(Enlarge Image)
Figure 2.
Incidence rates of (A) thromboembolism and (B) serious bleeding according to time intervals after first-time radiofrequency ablation.
Table 2 shows the incidence rates of thromboembolism and serious bleeding after the 3 months blanking period according to OAC therapy with results from the uni- and multivariable analyses. Oral anticoagulation discontinuation was not associated with significantly increased thromboembolic risk in low- or high-risk patients. In intermediate-risk patients, thromboembolic risk was significantly increased in the multivariable adjusted analysis. Oral anticoagulation therapy was associated with higher incidence rates of serious bleeding compared with OAC discontinuation.
Figure 3 depicts the 5-year cumulative incidence of thromboembolism and serious bleeding according to OAC therapy. Oral anticoagulation discontinuation was associated with 0.6% higher long-term risk of thromboembolism, while risk of serious bleeding was decreased by 1.8%.
(Enlarge Image)
Figure 3.
The cumulative incidences of thromboembolism and serious bleeding according to oral anticoagulation therapy. (A) Thromboembolism, adjusted for the components of CHA2DS2-VASc score and (B) serious bleeding, adjusted for the components of HAS-BLED score.
The present cohort was compared (1 : 4) with 15 848 non-ablated AF patients treated with rhythm-control therapy, i.e. 6180 (39%) on AADs, 6656 (42%) on DC, and 3012 (19%) receiving AADs+DC. Over 5-year follow-up, 1107 (7%) 'no-RFA' patients had thromboembolism with an overall incidence rate of 1.77 (1.67–1.88)95%CI per 100 person-years, where thromboembolic rates per 100 person-years with and without OAC were 1.34 (1.21–1.49)95%CI and 2.14 (1.98–2.30)95%CI, respectively. Crude incidence rate ratio (RFA vs. no-RFA) favoured RFA and was 0.47 (0.39–0.57)95%CI. Adjusted incidence rate ratio controlling for use of OAC, Aspirin, Clopidogrel, and the components of the CHA2DS2-VASc score was 0.53 (0.43–0.65)95%CI and remained in favour of RFA.
Figure 4 displays individual predictors of thromboembolism and serious bleeding. Recurrent AF was significantly associated with increased thromboembolic risk, and previous history of stroke/TIA was the only and strongest predictor associated with thromboembolism among the components of CHA2DS2-VASc score. Previous history of bleeding was the strongest predictor of serious bleeding, while OAC therapy, aspirin use, and kidney disease were also significantly associated with increased serious bleeding risk. These results were consistent in sensitivity analyses (Supplementary material online, Tables S3 and S4).
(Enlarge Image)
Figure 4.
Individual predictors associated with thromboembolism and serious bleeding after first-time radiofrequency ablation. (A) Thromboembolism, HR (95% CIs) and (B) serious bleeding, HR (95% CIs).
Results
Population
Final study population comprised 4050 patients undergoing first-time RFA. The median age was 59.5 years (IQR: 52.9–65.2); 73.5% were men. According to the CHA2DS2-VASc score, 1275 (31.5%) low-risk, 1268 (31.3%) intermediate-risk, and 1507 (37.2%) high-risk patients were identified. Median follow-up duration was 3.4 years (IQR: 2.0–5.6). Table 1 shows the baseline patient characteristics. After the 3 months blanking period, 2179 (53.8%) of patients had recurrent AF. Patients underwent a total of 5983 and a mean of 1.48 (SD 0.7) RFA procedures.
Persistence of Oral Anticoagulation Treatment After Radiofrequency Ablation
Oral anticoagulation was initiated in 3707 (91.5%) patients after RFA. Figure 1 depicts long-term persistence of OAC treatment according to thromboembolic risk profile. At Day 90, 1181 (97.3%) of low-risk, 1140 (94.7%) of intermediate-risk, and 1450 (95%) of high-risk patients were on OAC therapy. Corresponding numbers at 1 year were lower; 55.7, 67.4, and 70.4%, respectively. Approximately 70% of high-risk patients and half of the total RFA population received OAC therapy beyond the first year.
(Enlarge Image)
Figure 1.
Persistence in use of oral anticoagulation therapy after first-time radiofrequency ablation according to the CHA2DS2-VASc score: the proportion of patients receiving oral anticoagulation therapy by years.
Thromboembolism and Bleeding Events
After discharge from first-time RFA, 103 (2.5%) patients experienced thromboembolism (71 (68.9%) stroke, 27 (26.2%) TIA, and 5 (4.9%) peripheral arterial embolism) corresponding to an incidence rate of 0.80 [95% confidence interval (CI): 0.66–0.98] per 100 person-years. Of note, no patients in this cohort underwent left atrial appendage occlusion.
Figure 2 shows the incidence rates of thromboembolism and serious bleeding according to time intervals after first-time RFA, illustrating high event rates for both outcomes immediately after RFA procedure. Among the 103 thromboembolism cases, 32 (31%) patients developed thromboembolism [incidence rate 3.27 (95% CI: 2.31–4.62) per 100 person-years] during the 3 months blanking period. After 3 months, 71 (69%) patients experienced thromboembolism [incidence rate 0.60 (95% CI: 0.48–0.76) per 100 person-years]; only 37 (52.1%) of thromboembolic events occurred in high-risk patients. In the 111 cases with serious bleeding, events during and after the blanking period occurred in 24 (21.6%) and 87 (78.4%) cases, respectively, corresponding to incidence rates of 2.45 (95% CI: 1.64–3.65) and 0.73 (95% CI: 0.60–0.90) per 100 person-years. The incidence rate of serious bleeding was 0.60 (95 CI: 0.46–0.80) per 100 person-years after the blanking period when patients with aspirin therapy were excluded.
(Enlarge Image)
Figure 2.
Incidence rates of (A) thromboembolism and (B) serious bleeding according to time intervals after first-time radiofrequency ablation.
Effect of Oral Anticoagulation Therapy
Table 2 shows the incidence rates of thromboembolism and serious bleeding after the 3 months blanking period according to OAC therapy with results from the uni- and multivariable analyses. Oral anticoagulation discontinuation was not associated with significantly increased thromboembolic risk in low- or high-risk patients. In intermediate-risk patients, thromboembolic risk was significantly increased in the multivariable adjusted analysis. Oral anticoagulation therapy was associated with higher incidence rates of serious bleeding compared with OAC discontinuation.
Figure 3 depicts the 5-year cumulative incidence of thromboembolism and serious bleeding according to OAC therapy. Oral anticoagulation discontinuation was associated with 0.6% higher long-term risk of thromboembolism, while risk of serious bleeding was decreased by 1.8%.
(Enlarge Image)
Figure 3.
The cumulative incidences of thromboembolism and serious bleeding according to oral anticoagulation therapy. (A) Thromboembolism, adjusted for the components of CHA2DS2-VASc score and (B) serious bleeding, adjusted for the components of HAS-BLED score.
Matched Analyses
The present cohort was compared (1 : 4) with 15 848 non-ablated AF patients treated with rhythm-control therapy, i.e. 6180 (39%) on AADs, 6656 (42%) on DC, and 3012 (19%) receiving AADs+DC. Over 5-year follow-up, 1107 (7%) 'no-RFA' patients had thromboembolism with an overall incidence rate of 1.77 (1.67–1.88)95%CI per 100 person-years, where thromboembolic rates per 100 person-years with and without OAC were 1.34 (1.21–1.49)95%CI and 2.14 (1.98–2.30)95%CI, respectively. Crude incidence rate ratio (RFA vs. no-RFA) favoured RFA and was 0.47 (0.39–0.57)95%CI. Adjusted incidence rate ratio controlling for use of OAC, Aspirin, Clopidogrel, and the components of the CHA2DS2-VASc score was 0.53 (0.43–0.65)95%CI and remained in favour of RFA.
Patient-specific Risk Factors for Thromboembolism and Serious Bleeding
Figure 4 displays individual predictors of thromboembolism and serious bleeding. Recurrent AF was significantly associated with increased thromboembolic risk, and previous history of stroke/TIA was the only and strongest predictor associated with thromboembolism among the components of CHA2DS2-VASc score. Previous history of bleeding was the strongest predictor of serious bleeding, while OAC therapy, aspirin use, and kidney disease were also significantly associated with increased serious bleeding risk. These results were consistent in sensitivity analyses (Supplementary material online, Tables S3 and S4).
(Enlarge Image)
Figure 4.
Individual predictors associated with thromboembolism and serious bleeding after first-time radiofrequency ablation. (A) Thromboembolism, HR (95% CIs) and (B) serious bleeding, HR (95% CIs).
SHARE
