Response to ibutilide and the long-term outcome after catheter ablation for non-paroxysmal atrial fibrillation

Summary

Purpose:

This study aimed to assess the relationship between the cardiac rhythm response to ibutilide infusion after pulmonary vein isolation and the recurrence of long-term atrial arrhythmias.

Methods

One hundred and thirty-eight patients with nonparoxysmal atrial fibrillation who had had their first catheter ablation were retrospectively included. All patients whose atrial fibrillation did not terminate after pulmonary vein isolation were administered intravenous ibutilide (1.0 mg). Those with termination of atrial fibrillation after ibutilide administration were defined as responders (n = 86); those without termination of atrial fibrillation, as non-responders (n = 52). The primary endpoint was any documented recurrence of atrial arrhythmia lasting more than 30 seconds after the initial catheter ablation.

Results

Conversion of atrial fibrillation to sinus rhythm, directly or via atrial flutter, with ibutilide administration was achieved in 62.3% of patients. A longer duration of atrial fibrillation was associated with failed termination of atrial fibrillation (odds ratio 1.009, 95% confidence interval 1.002–1.017, p = 0.011). During a median follow-up period of 610 days (interquartile range 475–1 106) post ablation, non-responders (n = 24, 46.2%) had a higher recurrence rate of atrial arrhythmia than the responders (n = 26, 30.2%; log-rank, p = 0.011) after the initial catheter ablation. Multivariate Cox regression analysis revealed that non-responders (hazard ratio 1.994, 95% confidence interval 1.117–3.561, p = 0.020) was significantly correlated with recurrence of atrial arrhythmias.

Conclusion

In patients whose atrial fibrillation persisted after pulmonary vein isolation, the response to ibutilide administration could predict the recurrence of atrial arrhythmias after catheter ablation, which may be useful for risk stratification for recurrence of atrial fibrillation and individualised management of atrial fibrillation.

Atrial fibrillation (AF) is the most common clinical arrhythmia, with a prevalence of 3.0% in persons aged over 21 years.¹ Since over 90% of the triggering factors of AF are found in the pulmonary veins (PVs), electrical isolation of the PVs has been the cornerstone of AF ablation.^2-4 However, AF termination during or upon completion of pulmonary vein isolation (PVI) occurs in only a minority of patients with persistent AF.

During catheter ablation (CA) of persistent AF, antiarrhythmic drug (AAD) administration is a common practice to facilitate the ablation. Ibutilide, which is a class III AAD, can block the rapidly outward delayed rectifier potassium (K+) current to prolong the atrial refractory period, subsequently terminating AF.⁵ In some previous studies, ibutilide showed a 50.5 to 56% cardioversion rate for AF in patients without ablation.^6,7 After PVI completion, the re-entrant wave fronts between the PV and the left atrium are interrupted, which could in turn facilitate a higher AF termination rate after ibutilide administration. However, studies reporting the efficacy of ibutilide in the termination of AF after PVI completion are limited.

In patients with non-paroxysmal AF, there was a 31% recurrence rate of AF after catheter ablation, which was associated with a poor prognosis, such as stroke.⁸ The prediction of AF recurrence could help optimise individualised AF management. The approach employed for AF termination was reported to be associated with AF recurrence after ablation.⁹ As mentioned, ibutilide can be used for AF termination, while patients with different clinical characteristics had different responses to ibutilide. Currently, studies about intraprocedural ibutilide mostly focus on low-dose application and the difference in ablation strategy may influence the judgement of the effect of ibutilide on the prognosis.^10,11

In patients with persistent AF after PVI, it remains to be established whether the heart rhythm response to ibutilide infusion with a standard dose is associated with ablation outcome. The use of ibutilide in non-paroxysmal AF treatment is unknown. This study therefore aimed to assess the relationship between the cardiac rhythm response to ibutilide infusion after PVI and recurrence of long-term atrial arrhythmias (AAs), which may provide a useful tool for prediction of AF recurrence. We hypothesised that AF termination with ibutilide infusion after PVI indicates good rhythm control during a long-term follow up.

Methods

A total of 193 consecutive patients with non-paroxysmal AF who underwent CA for the first time between January 2014 and December 2018 at our institution were retrospectively included. Patients with valvular heart disease, a history of hyperthyroidism or cardiac surgery, or failed intraprocedural cardioversion were excluded. Of the remaining patients, seven had sinus rhythm (SR) before CA, 11 had AF conversion to SR during or upon completion of PVI, and 11 underwent direct-current synchronised cardioversion (DCC) after PVI because the left ventricular ejection fraction (LVEF) was ≤ 35% (n = 6), or based on the operator’s discretion.

Two patients were lost to follow up after the initial CA. The remaining 138 patients who still had AF after PVI were included in this study (Fig. 1). All patients received 1 mg ibutilide infusion for AF termination. The patients were assigned to two groups according to their response to the infusion: those with successful AF termination (ibutilide responders, n = 86) and those without AF termination (ibutilide non-responders, n = 52). This study was approved by the institutional review board and ethics committee of Peking Union Medical College Hospital (approval number: S-K998).

Fig. 1.

Patient flow diagram. Ibutilide responders were defined as those with successful AF termination after the first infusion of ibutilide (1 mg). Ibutilide non-responders were defined as those with failed AF termination. PVI, pulmonary vein isolation.

Persistent AF was defined as AF lasting more than seven days. Long-standing persistent AF was defined as AF lasting more than one year. AF duration was defined as the duration from the date of AF detection on 12-lead electrocardiogram or 24-hour Holter monitor, to the date of initial CA.

All patients underwent transoesophageal echocardiography to exclude any left atrial thrombi before CA. The target-activated clotting time was 300–350 seconds. Circumferential PVI was performed in all patients. The PVI endpoint was bidirectional electrical conduction block between the left atrium and the PV. After PVI, 1 mg of ibutilide (dissolved in 20 ml of 0.9% sodium chloride) was administered by slow intravenous injection over 10 minutes. The injection was discontinued on conversion to SR, occurrence of ventricular tachycardia, or any adverse reaction. All patients were observed for 20 minutes after ibutilide administration. Patients with AF termination after the ibutilide infusion were defined as responders; those with persistent AF, as non-responders. If AF was converted to atrial flutter (AFL) or atrial tachycardia (AT), the AFL or AT was mapped and ablated. If AF persisted, a second ibutilide infusion (1 mg) or DCC was administered until termination of AF. No ventricular tachycardia, torsade de pointes, or significant hypotensive response was observed during or after the ibutilide infusion. Additional linear ablation was performed at the operator’s discretion.

The endpoint of linear ablation was bidirectional electrical conduction block across the lines. Superior vena cava (SVC) isolation was performed when rapid SVC activity was observed. Ablation was performed using a catheter with an open, irrigated tip (Thermocool SmartTouch; Biosense-Webster Inc, Diamond Bar, CA, USA). Electro-anatomical mapping systems (CARTO XP or CARTO 3; Biosense-Webster Inc, Diamond Bar, CA, USA) were used to guide the procedures.

AADs were prescribed during the first three months after the initial ablation; thereafter, AAD use was at the physician’s discretion. Anticoagulation therapy was employed for a minimum of three months after the ablation. After three months, anticoagulants were administered at the clinician’s discretion.

Patients were assessed every three months to determine AA recurrence by 24-hour Holter monitoring at the out-patient clinic. Patients were followed up for at least 12 months after ablation. In patients with symptoms suggestive of recurrence of arrhythmia, 12-lead electrocardiogram and 24-hour Holter monitor recordings were additionally performed. The primary endpoint was any documented AA recurrence (AF, AFL or AT). AA recurrence was defined as the presence of AA lasting longer than 30 seconds after the blanking period of 90 days.

Statistical analysis

Continuous data are expressed as mean ± standard deviation for normally distributed data or as median (interquartile range, IQR) for non-normally distributed data. Categorical data are presented as absolute values and percentages. Tests for significance were conducted using the t-test or non-parametric test (Mann−Whiney U-test) for continuous variables, or χ2 test or Fisher’s exact test for categorical variables.

Multivariate logistic regression analysis was used to analyse the clinical covariates associated with failed AF termination after ibutilide administration. Univariate and multivariate Cox proportional hazards models were employed to determine the clinical covariates associated with recurrent AA, and the results are presented as hazard ratio (HR) with 95% confidence interval (CI). Comparisons of the Kaplan−Meier curves between ibutilide responders and non-responders were assessed with a two-sided Mantel−Haenszel (log-rank) test. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using the statistical software SPSS (v.21.0; IBM Corporation, Armonk, NY, USA).

Results

The baseline characteristics of the included patients are summarised in Table 1. The average age of the non-paroxysmal AF cohort was 61 ± 10 years, 73.6% of the patients were male and 32.6% had long-standing persistent AF. The median AF duration was 11 months (interquartile range 3.0–48.8). The left atrial diameter (LAD) was 45 ± 5 mm and the LVEF was 65 ± 9%. AF conversion to SR directly (87.2%) or through AFL (12.8%) was achieved in 62.3% of the patients. Compared with ibutilide responders, non-responders were older and had a longer AF duration. No differences in LAD, LVEF, estimated glomerular filtration rate (eGFR), medical history, pre-ablation medications or post-ablation AAD were found between the two groups. Multivariate logistic regression analysis identified longer AF duration (odds ratio 1.009, 95% CI 1.002–1.017, p = 0.011) as a positive predictor of ibutilide non-responders (Table 2).

Table 1. Baseline characteristics of all included patients.

Baseline characteristics	Total (n = 138)	Ibutilide responders (n = 86)	Ibutilide non- responders (n = 52)	p-value
Age (years)	61 + 10	59 + 9	63 + 10	0.038
Male gender, n (%)	103 (73.6)	67 (77.9)	36 (69.2)	0.256
Body mass index (kg/m²)	26.5 + 3.0		26.7 + 3.2	0.900
Long-standing persistent AF, n (%)	45 (32.6)	21 (24.4)	24 (46.2)	0.008
AF duration (months), median (IQR)	11 (3.0-48.8)	5 (2.0-34.5)	17 (6.0-70.3)	0.004
Left atrial diameter (mm)	45 + 5	44 + 5	46 + 5	0.107
LVEDD (mm)	49 + 4	48 + 4	49 + 5	0.264
Left ventricular ejection frac- tion (%)	65 9	64 + 9	65 + 8	0.501
eGFR < 60 ml/min/1.73 m², n (%)	14 (10.1)	9 (10.5)	5 (9.6)	0.873
Medical history, n (%)
Hypertension	77 (55.8)	48 (55.8)	29 (55.8)	0.996
Diabetes	31 (22.5)	21 (24.4)	10 (19.2)	0.479
Coronary disease	22 (15.9)	10 (11.6)	12 (23.1)	0.075
Cardiomyopathy	3 (2.2)	3 (3.5)	0 (0)	0.290
Heart failure	7 (5.1)	5 (5.8)	2 (3.8)	0.710
Hyperlipaemia	56 (40.6)	38 (44.2)	18 (34.6)	0.267
CHA,DS,-VASC score	1.7 + 1.4	1.6 + 1.4	1.9 + 1.4	0.221
Pre-ablation medications, n (%)
AADs	13 (9.4)	9 (10.5)	4 (7.7)	0.589
Beta-blocker	77 (55.8)	52 (60.5)	25 (48.1)	0.156
ACEI/ARB	41 (29.7)	25 (29.1)	16 (30.8)	0.832
Statins	40 (29.0)	24 (27.9)	16 (30.8)	0.719
Postablation AADs, n (%)
No	14 (10.2)	8 (9.3)	6 (11.5)	0.619
Amiodarone	103 (74.6)	63 (73.3)	40 (77.0)
Propafenone	21 (15.2)	15 (17.4)	6 (11.5)

Reported as mean ± standard deviation, median (IQR), or n (%) AADs, anti-arrhythmic drugs; ACEI/ARB, angiotensin converting enzyme inhibitor or angiotensin receptor blocker; AF, atrial fibrillation; eGFR, estimated glomerular filtration rate; IQR, interquartile range; LVEDD, left ventricular enddiastolic dimension.

Table 2. Predictors of ibutilide non-responders in multivariate logistic regression analysis.

Predictors of ibutilide non-responders	Odds ratio	95% CI	p-value
Age, per year	1.034	0.990-1.080	0.128
Male gender	0.492	0.198-1.226	0.128
Atrial fibrillation duration, per month	1.009	1.002-1.017	0.011
Left atrial diameter, per 1 mm	1.064	0.985-1.150	0.114
eGFR < 60 ml/min/1.73 m²	1.975	0.519-7.512	0.318
Pre-ablation AADs therapy	0.483	0.119-1.968	0.310

AADs, anti-arrhythmic drugs; CI, confidence interval; eGFR, estimated glomerular filtration rate.

PVI alone was performed in 72 (52.2%) patients, and the remaining 66 patients had PVI plus adjunctive linear ablation or SVC isolation. Left atrial roof linear ablation or mitral isthmus linear ablation were performed in 57 (41.3%) and 17 (12.3%) patients, respectively. Linear ablation of the cavotricuspid isthmus was performed in nine (6.5%) patients, while SVC isolation was added in two (1.4%). Compared to ibutilide responders, ibutilide non-responders had more frequent left atrial roof linear ablations (51.9 vs 34.9%, p = 0.049). No differences in the frequency of other types of linear ablation or SVC ablation were found between the two groups.

During a median follow-up period of 610 days (IQR: 475– 1 106) post ablation, AAs recurred in 50 (36.2%) patients in the overall cohort after the initial CA [ibutilide responders, n = 26 (30.2%); ibutilide non-responders, n = 24 (46.2%)]. The HRs relating the time of arrhythmia recurrence to individual demographic and clinical factors after the initial CA are presented in Table 3.

Table 3. Clinical factors related to recurrence of atrial arrhythmia after the first catheter ablation.

	Univariable analysis		Multivariable analysis
Clinical factors	Hazard ratio (95% CI)	p-value	Hazard ratio (95% CI)	p-value
Age, per year	1.003 (0.973-1.035)	0.836
Male gender	0.645 (0.356-1.171)	0.150
Initial BMI, per unit	0.922 (0.840-1.013)	0.092
AF duration, per month	1.003 (1.001-1.006)	0.016	1.002 (1.000-1.005)	0.093
Left atrial diameter, per 1 mm	0.967 (0.921-1.017)	0.191
LVEF, per 1%	1.016 (0.980-1.054)	0.382
eGFR < 60 ml/ min/1.73 m²	2.115 (1.027-4.359)	0.042	2.287 (1.101-4.752)	0.027
Hypertension	0.607 (0.347-1.063)	0.081
CHA,DS,-VASc	1.027 (0.841-1.254)	0.794
Pre-ablation AADs therapy	1.982 (0.929-4.232)	0.077
Left atrial roof linear ablation	0.723 (0.405-1.290)	0.273
Ibutilide non-respond- ers	2.204 (1.159-3.534)	0.013	1.994 (1.117-3.561)	0.020
Post-ablation AADs therapy
Amiodarone VS no	0.639 (0.284-1.437)	0.279
Propafenone VS no	0.488 (0.171-1.393)	0.180
Amiodarone VS propafenone	1.310 (0.582-2.948)	0.515

AADs, anti-arrhythmic drugs; AF, atrial fibrillation; BMI, body mass index;

CI, confidence interval; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction.

Among the factors, AF duration (HR 1.003, 95% CI 1.001– 1.006, p = 0.016), eGFR < 60 ml/min/1.73 m² (HR 2.115, 95% CI 1.027–4.359, p = 0.042) and ibutilide non-responders (HR 2.204, 95% CI 1.159–3.454, p = 0.013) were significantly associated with AA recurrence. In the multivariate Cox proportional hazards analysis, only eGFR < 60 ml/min/1.73 m² (HR 2.287, 95% CI 1.101–4.752, p = 0.027) and ibutilide non-responders (HR 1.994, 95% CI 1.117–3.561, p = 0.020) were independently associated with AA recurrence after CA. Kaplan–Meier analysis showed that the AA-free survival rates after the initial CA were significantly higher in the ibutilide responders than in the ibutilide non-responders (log-rank test, p = 0.011) (Fig. 2A).

Fig. 2.

Kaplan–Meier curves of long-term clinical outcome in patients with non-paroxysmal AF after the first (A) and the second (B) catheter ablation.

The timing for the second procedure was at least 90 days after the initial ablation in patients with AA recurrence. Nineteen of 50 patients (13.8%) with AA recurrence underwent repeat procedures. At the second procedure, 14 patients had a PV–left atrial gap, and the mean number of reconnected PVs was 1.33 ± 0.84 per patient. Of the remaining five patients, one had AF originating from SVC, one had mitral isthmus-dependent AFL, one had left atrial roof-dependent AFL, one had left atrial low-voltage area-associated AF, and one underwent the second procedure at a different medical centre.

After the second CA, one ibutilide responder was lost to follow up and 36/137 (26.3%) patients had AA recurrence. The AA recurrence rate after the second CA was higher in non-responders than in responders [36.5 (19/52) vs 20% (17/85), p = 0.033]. Kaplan–Meier analysis showed that the AA-free survival rate after the second CA remained significantly higher in responders than in non-responders (log-rank test, p = 0.010) (Fig. 2B).

Discussion

The key findings of this study were as follows: (1) in patients with persistent AF after PVI, the rate of AF termination by ibutilide administration was 62.3%; (2) a longer AF duration was associated with failed AF termination by ibutilide infusion; (3) failed AF termination with ibutilide infusion after PVI was associated with an unfavourable arrhythmia-free survival during a long-term follow up.

DCC is frequently used for AF termination and achieves SR in more than 85% of patients.^12,13 However, DCC is typically performed under intravenous procedural anaesthesia, usually resulting in displacement of the patients. By contrast, ibutilide administration is a much simpler procedure.

In patients with persistent AF after PVI completion, ibutilide could prolong the left atrial micro-re-entrant cycle length, which could in turn cause the following fibrillatory wave fronts to enter an effective refractory period and organise the electrical activity, thereby terminating AF. In this study, we found that the rate of AF termination by ibutilide administration was 62.3% after PVI completion in patients with non-paroxysmal AF, which is effective for AF termination. However, in a previous study, only 12 to 26.9% of cases achieved AF termination after ibutilide infusion.^11,14 Such discrepancies may be related to the high dose of ibutilide and the high quality of PVI in our study. Nevertheless, 87.2% of the patients in our cohort had direct conversion to SR.

Although PVI can reduce the interplay between PV and the left atrium, persistent AF may be maintained by pathological changes in electrical remodelling and/or structural remodelling after PVI completion. In our study, we found that ibutilide had a limited effect on AF termination in patients with a longer AF duration. In a previous study, a longer AF duration also showed a significant influence on failed AF termination by ibutilide in patients with sustained AF without PVI.¹⁵ As we know, a longer AF duration contributes to progression of atrial remodelling.¹⁶ Patients with a longer AF duration may have severe atrial remodelling; therefore ibutilide may have a worse effect on AF termination. For patients with longer AF duration, DCC may be the first choice due to the arrhythmic effect of ibutilide.

The response to ibutilide after PVI may be associated with the extent of atrial remodelling.¹⁷ Ibutilide promotes organisation of intracardial electrical activity. Therefore, if AF is terminated and converted to SR or AFL after ibutilide administration, the atrial remodelling may be mild; otherwise, it may be severe. A large anatomical or conductive obstaclebased re-entry may have an excitable gap that cannot be closed by effective refractory period prolongation, and this re-entrant mechanism may be insensitive to ibutilide infusion.¹⁸ Therefore, the response to ibutilide administration after PVI may help guide intraprocedural ablation.

He et al.¹⁴ reported that low-dose ibutilide treatment after PVI could help appropriately select patients with persistent AF who needed to undergo PVI alone. For patients with AF termination after ibutilide administration, PVI is sufficient for rhythm control. For patients with persistent AF, a more aggressive atrial substrate modification may be required.

Available data in the published literature regarding the acute response to AADs for cardioversion after PVI and AA recurrence are limited. Our study demonstrated that the response to ibutilide after PVI could predict AA recurrence in patients with non-paroxysmal AF. As mentioned, the response to ibutilide was associated with the extent of atrial remodelling, which may in turn explain AF recurrence. Although the response to ibutilide was associated with the duration of AF, multivariate Cox analysis showed that the response to ibutilide was associated with AA recurrence, not AF duration. Hence, the response to ibutilide may be more reflective of AF progression than AF duration. Based on the predictive value of AF recurrence, the response to ibutilide during CA may help select patients with high risk of AA recurrence for more frequent post-ablation surveillance, contribute in the prediction of more accurate prognoses, and aid in the decision-making about anti-arrhythmic and anticoagulation therapies.

Limitations

This study has several limitations. First, this was a retrospective observational study in a single centre. All patients had a relatively small left atrial diameter. Therefore, a selection bias exists. Second, ibutilide administration was performed after successful PVI. Hence, the risk stratification after PVI is of limited clinical relevance. However, given its relationship with the long-term outcome, the response to ibutilide may guide intraprocedural ablation and post-ablation management. Third, as left atrial voltage mapping in the SR was not performed, we were unable to assess directly the relationship between the efficacy of ibutilide in AF termination and the size of the atrial low-voltage areas. In future studies, methods for evaluating atrial fibrosis, including intracardiac electro-anatomical mapping and cardiac magnetic resonance, should be combined with those evaluating the efficacy of ibutilide in AF termination. Finally, as some patients with asymptomatic AF may have been ignored during the follow up, we may have underestimated the rate of AA recurrence. Longterm monitoring of an implanted electrocardiogram may help resolve this problem.

Conclusion

Ibutilide (1 mg) for AF termination after PVI was effective in patients with non-paroxysmal AF. The response to intraprocedural ibutilide could predict AA recurrence after CA, which may be useful for risk stratification of AF recurrence and individualised AF management.