This is a retrospective study that enrolled patients with a history of surgical intervention for VHD who underwent AFL catheter ablation. Two high-volume institutions in Taiwan were included in this study, namely, Taipei Veterans General Hospital and Cheng Hsin General Hospital. Eighty-two patients with a history of surgical intervention for VHD were included, all of whom received AFL catheter ablation between July 2008 and April 2019. Surgical intervention for VHD was defined as mitral valve, tricuspid valve, and aortic valve replacement or repair. Patients who received a concomitant CABG were also included. Clinical AF was defined as the diagnosis of AF before AFL catheter ablation. Concomitant AF ablation was not performed due to absence of clinical symptoms. The exclusion criteria included patients with AF who received pulmonary vein isolation (PVI) (n = 5), patients with focal AT who received catheter ablation (n = 3), and patients who received an incomplete procedure due to complications or unstable hemodynamics (n = 2). Finally, 72 patients were enrolled in the study. The patients were categorized into two groups according to the type of AFL catheter ablation they received: the typical AFL group (patients with typical AFL only; n = 45) and the atypical AFL group (patients who had an atypical AFL for catheter ablation, with or without concomitant typical AFL; n = 27). Baseline and electrophysiological characteristics were collected and analyzed.

This retrospective cohort study obtained ethical approval from the institutional review board of the Taipei Veterans General Hospital and was conducted in full compliance with national ethical and regulatory guidelines (11).

On surface 12-lead electrocardiography, a typical or atypical AFL could be identified by the flutter wave (12). Detailed electrophysiologic study and mapping has been described in our previous studies (13–15). Three-dimensional electroanatomical mapping systems were used in 26 (96.3%) of 27 patients in the atypical AFL group and 4 (8.9%) of 45 patients in the typical AFL group. The EnSite mapping system (St. Jude Medical, Minneapolis, MN, USA), Carto mapping system (Biosense-Webster Inc., Diamond Bar, CA, USA), Rhythmia (Boston Scientific Corporation, Middlesex County, MA, USA), and ablation catheters were used at the discretion of the treating electrophysiologist. The AFL circuit was confirmed using electroanatomical mapping and entrainment maneuvers.

After discharge following the index ablation procedure, the patients were followed up at 2 weeks and then regularly every 1–3 months at our cardiology outpatient clinic. Antiarrhythmic medicines were prescribed for 4–8 weeks after the procedure to prevent early recurrence of AF/AFL within 3 months. The blanking period was defined as within 3 months after ablation (20). Follow-up with 24-h Holter monitoring or 1-week cardiac event monitoring was performed 3 months after the ablation procedure and at any time if the patients experienced symptoms suggestive of tachyarrhythmia. Long-term efficacy was assessed based on resting surface 12-lead electrocardiograms (ECG), 24-h Holter monitoring records, and/or 1-week cardiac event monitoring records. Any atrial tachyarrhythmia such as AF, AFL, and AT lasting ≥30 s was defined as a recurrence. The follow-up period was up to September 30, 2019.

Continuous variables are expressed as mean ± standard deviation (SD). Categorical variables were compared using the chi-square test. Hazard ratio was calculated using univariate and multivariate logistic regression analyses with a stepwise analysis and confirmation. In the multivariate regression analysis, factors with P < 0.1 in the univariate analysis were selected for adjustment. The variables entered in the multivariate logistic regression analysis included chronic kidney disease (CKD) and left atrial diameter (LAD). A two-sided P < 0.05 was considered significant for all statistical determinations. All analyses were performed using IBM Statistical Product and Service Solutions (SPSS) Version 20.

The baseline characteristics are shown in Table 1. The flowsheet of the inclusion, exclusion, classification of subgroups, recurrence number, and patterns is provided in Figure 1. There were 45 patients in the typical AFL group and 27 patients in the atypical AFL group. No significant differences in baseline characteristics were found between these two groups, except for the higher incidence of hyperlipidemia in the atypical AFL group.

The circuit number and the location of the isthmus of AFL in our study cohort are shown in Table 2. The mean number of reentrant circuits was 1.39 ± 0.72 in the total study population, while it was 2.04 ± 0.85 in the atypical AFL group. CTI-dependent AFL occurred in 17 patients in the atypical AFL group (17/27, 63.0%) and in 62 patients out of our total study cohort (62/72, 86.1%).

An example of circuits in typical and atypical AFL demonstrated by the 3D mapping system is shown in Figure 2. In comparison of the typical and atypical AFL group, the mean follow-up period was 24.3 ± 26.0 and 18.0 ± 13.3 months, respectively (P = 0.18). No significant differences in the prevalence of clinical AF, recurrence rate of atrial tachyarrhythmia, and type of recurrent atrial tachyarrhythmias (including AFL, atypical AFL, AF, and focal AT) were found between the two groups (Table 3). The Kaplan–Meier analysis revealed no significant difference in the recurrence-free survival of atrial tachyarrhythmia between the typical and atypical AFL groups (Figure 3A).

As shown in Table 4, among 72 patients, 23 had concomitant AF (31.9%) before enrollment. Patients with concomitant AF had a trend toward a higher recurrence rate of atrial tachyarrhythmia compared to those without concomitant AF (P = 0.074). The recurrence rate of typical AFL was higher in patients with concomitant AF than those without concomitant AF (P = 0.012).

In the typical AFL group, 17 out of 45 (37.8%) cases had a concomitant AF, and these patients had a significantly higher recurrence rate of atrial tachyarrhythmia than those without concomitant AF (P = 0.006; Table 4 and Figure 3B). In the atypical AFL group, we found no significant difference in the recurrence rate of atrial tachyarrhythmia between the patients with and without concomitant AF (P = 0.331; Table 4 and Figure 3C).

In patients without concomitant AF (n = 48), the atypical AFL group had a higher recurrence rate of atrial tachyarrhythmia than the typical AFL group (P = 0.043; Table 4 and Figure 3D).

Among the 72 patients, 40 patients received single-valve surgery and 32 received multiple-valve surgery. Ten patients received concomitant on-pump CABG. The total number of valves treated by surgery in our study population was 110. The surgical detail is shown in Table 5.

In the comparison of the single-valve and multiple-valve groups, the occurrence rate of atypical AFL was the same (15/40 vs. 12/32, P = 1). In the single-valve surgery group, there was a trend toward an increased occurrence of atypical AFL in patients who underwent mitral valve (MV) surgery, when compared to aortic valve (AV) surgery (44.4 and 16.7%, P = 0.07, respectively). Only one patient received lone tricuspid valve (TV) surgery and developed atypical AFL.

LA flutter occurred in 10 patients in our study cohort (10/72, 13.9%), whereas the remaining 62 patients had RA flutter only (62/72, 86.1%). LA flutter was highly associated with prior MV surgery (9/10, 90%).

Table 6 shows the results of the univariate and multivariate Cox regression analyses of each risk factor. In the univariate analysis, CKD and LAD were associated with the recurrence of atrial tachyarrhythmia (P = 0.017 and 0.029, respectively). In the multivariate analysis, CKD and LAD were both identified as independent predictors of recurrent atrial tachyarrhythmia (P = 0.02, hazard ratio 2.73 and P = 0.04, hazard ratio 1.06, respectively).

We found no significant difference in the recurrence rate of atrial tachyarrhythmia between the typical and atypical AFL groups in our cohort. Patients with concomitant AF had a higher incidence of recurrent typical AFL than those without it. In the typical AFL group, patients with concomitant AF had a higher recurrence rate of atrial tachyarrhythmia than those without it. Among patients without concomitant AF, the atypical AFL group had a higher recurrence rate than the typical AFL group. LA flutter was less common than RA flutter and was highly associated with prior MV surgery. CKD and LAD were identified as predictors of recurrence of atrial tachyarrhythmia in patients who received AFL ablation after undergoing surgery for VHD.

Previous studies showed that typical AFL was the most common atrial arrhythmia and demonstrated that patients who had undergone cardiac surgery had higher recurrence rates of AFL (12 vs. 1%) and AF (28 vs. 16%) after AFL catheter ablation than those who had not undergone cardiac surgery (9, 10). Similarly, our present study showed that majority of the total study population (86%) had a CTI-dependent AFL. In the atypical AFL group, CTI-dependent AFL also occurred in 63.0% of the patients. The overall recurrence rate of atrial tachyarrhythmia in our study population was 31.9%. Surgical incisions, either left atriotomy or a trans-septal approach, provided a substrate for atrial arrhythmias, which may have contributed to the perpetuation of AFL (1–5, 23). A higher recurrence rate of atrial tachyarrhythmia after AFL ablation could be explained by the arrhythmogenic substrate in patients after cardiac surgery for VHD (24). As a result, these patients should be closely followed up even after successful elimination of the AFL during the ablation procedure.

Several studies have reported that successful AFL ablation decreased the AF recurrence in 50–75% of patients with a history of AF (25–33). One meta-analysis also demonstrated that the recurrence of AF was 52.7% in patients which a history of AF after a mean of 16 months following period (34). In our present study, the recurrence rate of AF was 12.5% after AFL ablation in patients with concomitant AF; however, the recurrence rate of AFL was 29.2% (12.5% typical AFL, 16.7% atypical AFL). In agreement with previous studies, AFL catheter ablation might be associated with decreased AF recurrence in our study cohort. The higher recurrence rate of AFL might be due to difference in arrhythmic substrate between patients with or without prior cardiac surgery. The underlying mechanism of the effects of AFL ablation on the reduction of AF recurrence rate remains unknown. We hypothesized that the presence of AFL could be a trigger to initiate other types of atrial tachyarrhythmias, and AFL could transform into AF over time. In contrast, in the typical AFL group, patients with concomitant AF had a higher recurrence rate of atrial tachyarrhythmia than those without it (52.9 vs. 14.3%). The presence of clinical AF may represent the diseased substrate and led to perpetuation of atrial tachyarrhythmia (35, 36).

Among patients without AF, a higher recurrence rate was found in the atypical AFL group than the typical AFL group in our study. This finding might be explained by atrial remodeling which is associated with a higher recurrence rate of atrial tachyarrhythmia. Patients with atypical AFL or multiple reentrant circuits might have a more complex and abnormal substrate than those with typical AFL only (14, 21).

Cardiac surgery was associated with a high occurrence of atrial arrhythmias. AF and AFL have been reported in up to 15–40% of patients in the postoperative period following CABG, and a higher occurrence was found in valve surgery (37–50%) (37). RA incision and atriotomy were very common in valve surgery, and surgical scar contributed to more incidence of atypical AFL when compared with RA appendage cannulation which was performed in CABG. Since the location of the AFL circuit was highly associated with different surgical incision, RA flutter was much more common than LA flutter (38, 39). The occurrence of LA flutter was highly associated with LA surgical approach for MV surgery (23, 38–40). In our study cohort, LA flutter occurred in only 13% (10/72) of patients, and 90% (9/10) of these patients had MV surgery. This result was compatible with previous studies. Our study also demonstrated that the number of valves receiving surgery, concomitant CABG, and surgical AF ablation did not affect the occurrence of atypical AFL or recurrence after AFL ablation. Notably, in our cohort of single-valve surgery, there was a trend toward a higher occurrence of atypical AFL in MV surgery than AV surgery (P = 0.07).

CKD and LAD were predictors of recurrent atrial tachyarrhythmia in our present study. The correlation between CKD and AF has been demonstrated. The prevalence of AF in patients with CKD was found to be between 18 and 21.2%, and concomitant AF and CKD had a negative effect on prognosis (41–44). In patients with CKD without a preexisting AF, AF was detected in 7.7% of the patients after 5 years of follow-up (44). Several risk factors for developing atrial arrhythmias in patients with CKD have been reported: increased inflammation, lower serum calcium level, increased platelet count, and higher ferritin level (42, 45, 46). These factors could enhance the arrhythmogenic substrate contributing to the recurrence of atrial tachyarrhythmia after AFL ablation.

LA enlargement is involved in the mechanism of atrial arrhythmia. Several score systems including LA diameter have been used to evaluate the progression of AF (47). LA size was also an independent predictor of atrial tachyarrhythmia recurrence after AF ablation. One meta-analysis study showed three important predictors of recurrence after AF ablation: valvular AF, LAD longer than 50 mm, and recurrence within 30 days after ablation (48). LA volume seemed to be a more accurate predictor of recurrence than the type of AF (paroxysmal or persistent) (49–51). In agreement with previous evidence, our study suggested that LAD played an important role to predict the recurrence of atrial tachyarrhythmia in patients with cardiac surgery for VHD.