Defining Blanking Period Post-Pulmonary Vein Antrum Isolation

JACC: CLINICAL ELECTROPHYSIOLOGY

VOL.

ª 2017 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER

-, NO. -, 2017

ISSN 2405-500X/$36.00 http://dx.doi.org/10.1016/j.jacep.2017.01.006

Defining Blanking Period Post Pulmonary Vein Antrum Isolation Pouria Alipour, BSC,a Zahra Azizi, MD,a Meysam Pirbaglou, MSC,b Paul Ritvo, PHD,b Alfredo Pantano, MD,a Atul Verma, MD,a Yaariv Khaykin, MDa

ABSTRACT OBJECTIVES This study sought to determine the exact period after pulmonary vein antrum isolation (PVI) during which early recurrence of atrial tachyarrhythmia (ERAT) does not predict late arrhythmia recurrence (LR), in order to better define the blanking period. BACKGROUND Recurrence of atrial fibrillation after PVI is not uncommon. The first 3 months after PVI have been commonly treated as a blanking period, during which ERAT is not thought to predict LR after PVI; however, recent studies have shown that ERAT does predict LR. METHODS Baseline and follow-up data for 636 patients (mean age: 61.4
10.6 years; 67.1% male; 59% paroxysmal atrial fibrillation; 31.4% ERAT) who underwent PVI between 2010 and 2014 were included. Recurrences were monitored by electrocardiography and Holter monitoring at 1-, 3-, 6-, 9-, and 12-month intervals post-procedure. Receiveroperating characteristic curve analysis was used to define the blanking period after PVI. RESULTS Overall, 51%, 76%, and 92% of patients who had ERAT in the first, second, and third month post-PVI, respectively, also experienced LR (p ¼ 0.001). Using a logistic regression model, those manifesting ERAT during the first, second, and third month post-PVI were 4.22, 9.03, and 19.43 (p ¼ 0.001) times more likely to experience LR, respectively, compared to those without ERAT. Furthermore, receiver-operating characteristic analysis revealed that 23 days post-PVI is the optimal cutoff date for the blanking period, with area under the curve of 0.7, sensitivity of 69.2%, and specificity of 61.2%. CONCLUSIONS The likelihood of experiencing LR progressively rises with ERAT after the first month post-PVI. Blanking period after PVI should be limited to the first 23 days clinically and in future studies. (J Am Coll Cardiol EP 2017;-:-–-) © 2017 by the American College of Cardiology Foundation.

R

adiofrequency catheter ablation has been

during which any recurrence of AF, atrial flutter

used for pulmonary vein antrum isolation

(AFL), or atrial tachycardia (AT) is not considered a

(PVI) in patients with paroxysmal atrial fibril-

failure of the procedure nor is it suggestive of long-

lation (PAF) and persistent atrial fibrillation (AF)

term AF recurrence (8,9,11,12). According to the 2012

(1–5). Early recurrence of atrial tachyarrhythmia

Consensus Statement on Catheter and Surgical Abla-

(ERAT), common during the initial 90 days after abla-

tion of Atrial Fibrillation, a blanking period of

tion, has been reported by most investigators, ranging

3 months should be used to report efficacy of AF

in incidence from 6.7% to 65% because of the varying

ablation procedures (11). Proposed causes of early

definitions of the post-ablation blanking period

recurrence during the blanking period include post-

(6–10). This blanking period or therapy stabilization

ablation inflammation, short-term autonomic imbal-

period is defined as a period of time post-ablation

ance, and the required time for maturation of lesions

From the aHeart Rhythm Program, Southlake Regional Health Centre, Newmarket, Ontario, Canada; and the bFaculty of Health, School of Kinesiology, York University, Toronto, Ontario, Canada. Dr. Verma has received grants from Biosense Webster, Medtronic, Bayer, and Boehringer; and has served on the advisory boards of Biosense Webster and Bayer. All other authors have reported they have no relationships relevant to the contents of this paper to disclose. Francis Marchlinski, MD, served as Guest Editor for this paper. Manuscript received September 29, 2016; revised manuscript received December 7, 2016, accepted January 11, 2017.

2

Alipour et al.

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017 - 2017:-–-

Blanking Period Post-PVI

ABBREVIATIONS

(7,9,10,12,13). Although these factors may not

3 months post-procedure. Transesophageal echocar-

AND ACRONYMS

predispose patients to late recurrence (LR),

diography was performed on the day of the procedure

studies have shown a relationship between

to ensure the absence of left atrial appendage

these early recurrences and late arrhythmia

thrombus in all patients. In brief, for the ablation

recurrences after 3 months (3,9,12). The

procedure, vascular access was obtained using the

time point at which transient reversible

right and left femoral veins as well as the right in-

AAD = antiarrhythmic drug AF = atrial fibrillation AFL = atrial flutter AT = atrial tachycardia

post-ablation factors yield to reconnection

ternal jugular vein. A coronary sinus catheter was

CI = confidence interval

of the pulmonary veins (PVs) as the cause of

placed via the right internal jugular vein, and an

ERAT = early recurrence of

early arrhythmia recurrence has not been

intracardiac

atrial tachyarrhythmia

precisely established (12). Recent studies

placed in the right heart via the left femoral vein

ICE = intracardiac

have looked at early recurrences at different

(AcuNav, Siemens, Washington, DC). Transseptal ac-

echocardiography

(ICE)

probe

were

time points between 48 hours to 3 months

cess to the left atrium was established under fluoro-

with respect to their predictive ability for

scopic and ICE guidance, and the irrigated ablation

LRs (9,10,12). However, to date, no study

catheter (ThermoCool or Surround Flow, Biosense

has used receiver-operating characteristic

Webster, Diamond Bar, California) along with a

(ROC) analysis of early arrhythmia recur-

decapolar

rences as a predictor of LR, using time during

Webster) were placed in the left atrium via indepen-

the first 3 months as a continuous variable.

dent transseptal punctures under ICE guidance. The

Therefore, we conducted this study to find

patients were systemically anticoagulated, main-

the time point post-PVI during which tran-

taining an activated clotting time between 300 and

sient factors responsible for ERAT reflect an

350

established substrate for arrhythmia recur-

representation of the left atrium and the PVs was

rence, using ROC analysis to determine the scientific

created using the CARTO 3 (Biosense Webster) three-

basis for the blanking period.

dimensional electroanatomic mapping system. Power

LR = late recurrence NPAF = nonparoxysmal atrial fibrillation

OAC = oral anticoagulant OR = odds ratio PAF = paroxysmal atrial fibrillation

PV = pulmonary vein PVI = pulmonary vein antrum isolation

echocardiography

circular

seconds.

mapping

catheter

Three-dimensional

(Biosense

electroanatomic

was titrated between 30 and 40 W. Esophageal tem-

METHODS

perature was not routinely monitored during the procedure. Radiofrequency energy was delivered,

PATIENTS CHARACTERISTICS. Baseline and follow-up

guided by the circular mapping catheter in order to

data for 636 consecutive patients (mean age: 61.4


eliminate local electrograms across the PV antra and

10.6 years; 67.1% male; 59% PAF, 31.4% (n ¼ 200)

the posterior wall from the ridge between the left PVs

ERAT) who underwent their first PVI between 2010

and the appendage, across the entire roof and pos-

and 2014 were prospectively collected and included

terior wall of the left atrium to the septal aspect of the

in the analysis. Patients were stratified into two

right PVs as previously described (14). Using this

groups based on the presence of ERAT post-

approach, entrance block was achieved; however, PV

procedure. Additionally, patients were divided into

exit block was not routinely assessed, nor did the

those with PAF and those with persistent atrial

patients receive adenosine to assess for dormant PV

fibrillation (NPAF). All patients had follow-up infor-

conduction after ablation. In all patients the entire

mation available for at least 12 months post-PVI. Any

lesion set was verified during sinus rhythm (remap-

documentation of atrial tachyarrhythmia (AT) during

ped in those who required cardioversion or converted

follow-up was noted and included for consideration

from AF with ablation after establishment of sinus

in the study. Patients who had at least one episode of

rhythm). Each vein was rechecked at least 20 min

atrial arrhythmia (AF, AT, or AFL) in the first 90 days

after it was initially isolated. Isoproterenol challenge

were included in the study. All data collection was

was not used during the index ablation, nor were

reviewed and approved by the institution’s research

non-PV foci targeted beyond the lesion set as

ethics board. This study did not interfere with any

described. All patients who underwent ablation were

patient’s standard of care and was strictly a pro-

treated using the same approach. After 24 h of close

spective cohort study.

monitoring for post-procedural complications, pa-

PVI PROCEDURE. Amiodarone was discontinued at

least 3 months before the procedure; all other anti-

tients were discharged home. Follow-up visits were scheduled at 3, 6, 9, and 12 months post-procedure.

arrhythmic drugs (AADs) were discontinued at least

FOLLOW-UP AND REPORTING. Post-procedure, pa-

five half-lives before the ablation. All patients were

tients continued taking OACs for at least 3 months.

prescribed oral anticoagulants (OACs) at least 1 month

Further use of OACs was determined using the Ca-

before the procedure and continued for at least

nadian guidelines (11). Patients who were taking

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017

Alipour et al.

- 2017:-–-

AADs before the procedure resumed taking them for

Blanking Period Post-PVI

RESULTS

the first 3 months after the procedure. The drugs were discontinued at the 3-month follow-up if no docu-

DEMOGRAPHIC AND CLINICAL CHARACTERISTICS. Base-

mented recurrence of atrial tachyarrhythmia was

line demographic and clinical characteristics of the

observed. AADs were offered to patients with symp-

study participants are given in Table 1. Study partic-

tomatic recurrences of AF, but for patients who were

ipants included a total of 636 AF patients, 200 (31.4%)

refractory to or intolerant of medical therapy, a redo

of whom experienced ERAT within the 90-day

procedure was offered.

blanking period.

Patients were monitored closely post-procedure

Among the 200 patients with ERAT, 118 (59%) had

for recurrences of symptomatic and asymptomatic

PAF and 82 (41%) manifested NPAF. Mean left atrial

AF. An ambulatory ECG was performed 1 and 4 weeks

size was 41.1
8.5 mm in patients with ERAT, and

post-ablation, and 14-day Holter monitoring was

nearly 86% had a mean ejection fraction of 60% or

performed 4 to 6 weeks post-procedure. Further

higher. Hypertension was present in 61% of the pa-

ambulatory monitoring (48-h Holter) and ECG were

tients, along with congestive heart failure (11%),

performed at the 3-, 6-, 9-, and 12-month follow-up

structural heart disease (11%), and type 2 diabetes

visits and every 6 months thereafter, as well as dur-

mellitus (9%). There were statistically significant

ing any unscheduled ambulatory visits related to

differences

arrhythmia recurrence. For any patient with an

including body mass index (28.6 kg/m 2 in PAF, 30.2

implantable cardiac rhythm management device, the

kg/m 2 in NPAF; p ¼ 0.05) and frequency of amiodar-

device was interrogated for arrhythmia burden at

one failure before ablation (22% in PAF, 35.4% in

between

PAF

and

NPAF

patients,

each clinic visit. Any manifestation of atrial tachyar-

NPAF; p ¼ 0.03). In addition, a significantly greater

rhythmia was noted and recorded in the study data-

proportion of NPAF patients had a history of

base and subsequently used in the data analysis. For

congestive heart failure (6.0% in PAF, 17.2% in NPAF;

those patients who underwent repeat ablation, their

p ¼ 0.01) and structural heart disease (4.2% in PAF,

procedural data were collected and subsequent visits

19.5% in NPAF; p ¼ 0.001) (Table 1). Prevalence of

documented in the database. Recurrence was defined

ERAT differed between PAF and NPAF patients (31.5%

as any atrial tachyarrhythmia (AF or any organized

[27.2% PAF, 40.3% NPAF; p ¼ 0.011]).

tachyarrhythmia such as AFL) documented and lasting at least 30 seconds.

Overall, 242 patients (137 [68.5%] ERAT, 105 [24.1%] no ERAT) had LR, of whom 153 underwent a

In our study, any recurrences of AF, AFL, or AT

redo procedure (96 [48%] ERAT, 57 [13%] no ERAT).

within 90 days post-ablation were defined as ERAT,

Prevalence of LR differed between PAF and NPAF

and any recurrences of AF, AFL, or AT from 3 to 12

patients (35.8% PAF vs. 42.9% NPAF; p ¼ 0.001). All

months or more post-ablation were defined as LR.

patients who underwent reablation had PV reconnection of at least one PV.

STATISTICAL ANALYSIS. All continuous variables

Of the 436 patients (68.6%) without ERAT, 315

are expressed as mean
SD, and categorical variables

(72.2%) were diagnosed with PAF and 121 (27.8%)

are expressed as frequency and percentage. Between-

were diagnosed with NPAF before the ablation. In

group differences in demographic and clinical char-

patients without ERAT, mean left atrial size was

acteristics were evaluated using independent-sample

39.9
8.6 mm, and 89.4% of patients had a mean

Student t test for continuous variables and chi-square

ejection fraction of 60% or higher. Similarly, 61.5%

tests of independence for categorical variables. Using

were diagnosed with hypertension, 10.1% with

ROC, we evaluated an ideal time point with highest

congestive heart failure, 13% with structural heart

sensitivity and specificity to empirically define a

disease, and nearly 7.3% with type 2 diabetes mel-

blanking period. Subsequently, separate multivariate

litus. As with participants who experienced ERAT,

logistic regression models were used to assess the

there were statistically significant differences be-

likelihood of late arrhythmia recurrence based on

tween PAF and NPAF patients, specifically with re-

time to ERAT within the 90-day post-ablation win-

gard to gender distribution (68.7% men in PAF and

dow stratified according to the conventional (i.e.,

31.3% men in NPAF; p ¼ 0.02) and amiodarone

monthly classification) time period or the time period

failure rate before ablation (16.1% in PAF and 42%

derived using ROC analysis. Both regression models

in NPAF; p ¼ 0.001). In addition, a significantly

considered patients without ERAT during the blank-

greater proportion of NPAF patients had structural

ing period as the reference group. Statistical analyses

heart disease (8.8% in PAF and 24% in NPAF; p ¼

were performed using SPSS software, version 22 (IBM

0.001), congestive heart failure (6.7% in PAF and

Corp., Armonk, New York).

19% in NPAF; p ¼ 0.01), type 2 diabetes mellitus

3

Alipour et al.

4

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017 - 2017:-–-

Blanking Period Post-PVI

T A B L E 1 Demographic and Clinical Characteristics

With ERAT

Without ERAT

All (n ¼ 636)

All (n ¼ 200)

PAF (n ¼ 118)

NPAF (n ¼ 82)

PAF/NPAF (p Value)

All (n ¼ 436)

PAF (n ¼ 315)

NPAF (n ¼ 121)

PAF/NPAF (p Value)

61.4
10.6

61.9
9.3

61.6
9.6

62.3
9.0

0.62

61.2
11.1

60.6
11.5

62.6
10.0

0.07

Male

427 (67.1)

136 (68)

76 (64.4)

60 (73.1)

0.19

291 (66.7)

200 (63.4)

91 (75.2)

0.02

BMI (kg/m2)

29.6
8.2

29.3
5.5

28.6
5.8

30.2
5.3

0.05

29.7
9.1

29.5
10.2

30.5
5.1

0.29

ERAT

200 (31.4)

200 (100)

118 (100)

82 (100)

—

—

Age (yrs)

100 (15.7)

100 (50)

60 (50.8)

40 (48.8)

0.77

—

Second month

62 (9.7)

62 (31)

38 (32.2)

24 (29.3)

0.65

—

Third month

38 (6.0)

38 (19)

20 (16.9)

18 (22)

0.37

—

LR post-ablation

242 (38%)

137 (68.5)

77 (65.3)

60 (73.1)

0.09

105 (24.1)

78 (24.7)

27 (22.3)

0.60

On AAD post-ablation

566 (89%)

186 (92.5)

108 (91.5)

78 (94)

0.51

380 (87.2)

273 (86.7)

107 (89.2)

0.48

No. of failed AADs

1.14
1.0

1.4
1.04

1.32
1.0

1.48
1.1

0.03

1.03
0.97

1.02
0.94

1.05
1.03

0.81

Failed amiodarone

143 (22.5)

55 (27.5)

26 (22)

29 (35.4)

0.03

88 (20.2)

51 (16.1)

37 (30.6)

0.001

LA size (mm)

40.2
8.7

41.1
8.5

39.2
8.7

44
7.35

0.0001

39.9
8.6

38.9
9.1

42.5
6.3

0.0001

0.001

0.05

First month

Ejection fraction $60%

564 (88.7)

173 (86.5)

112 (94.9)

61 (74.3)

390 (89.4)

288 (91.4)

102 (84.2)

50%

39 (6.3)

15 (7.5)

4 (3.5)

11 (13.6)

24 (5.5)

12 (3.8)

12 (9.9)

40%

18 (2.7)

8 (4.0)

0 (0)

8 (9.9)

10 (2.3)

6 (1.9)

4 (3.3)

#30%

3 (0.5)

1 (0.5)

0 (0)

1 (1.2)

2 (0.5)

1 (0.3)

1 (0.8)

Structural heart disease Dilated cardiomyopathy

39 (6.1)

14 (7)

4 (3.5)

10 (12.2)

0.01

25 (5.7)

8 (2.5)

17 (14)

0.001

Ischemic cardiomyopathy

31 (4.9)

10 (5)

4 (3.5)

6 (7.3)

0.21

21 (4.8)

11 (3.5)

10 (8.3)

0.03

19 (3)

8 (4)

4 (3.5)

4 (4.8)

0.60

11 (2.5)

9 (2.85)

2 (1.6)

0.47

57 (9)

26 (13)

15 (12.7)

11 (13.4)

0.08

32 (7.3)

15 (4.8)

17 (14)

0.001

Hypertrophic cardiomyopathy Type 2 diabetes mellitus Congestive heart failure

65 (10.2)

21 (10.5)

5 (4.2)

16 (19.5)

0.001

44 (10.1)

21 (6.7)

23 (19)

0.001

Hypertension

389 (61.2)

121 (60.5)

66 (56)

55 (67.1)

0.11

268 (61.5)

180 (57.1)

88 (72.7)

0.003

Values are mean
SD or n (%) unless otherwise noted. AAD ¼ antiarrhythmic drug; BMI ¼ body mass index; ERAT ¼ early recurrence atrial tachyarrhythmia; LA ¼ left atrium; LR ¼ late recurrence; NPAF ¼ persistent atrial fibrillation; PAF ¼ paroxysmal atrial fibrillation.

(4.8% in PAF and 14% in NPAF; p ¼ 0.001), and

AF RECURRENCE WITHIN THE BLANKING PERIOD

hypertension (57.1% in PAF and 72.7% in NPAF;

AND THE LIKELIHOOD OF LR. Among the patients

p ¼ 0.001).

with ERAT, early recurrence of AF occurred during

Overall, 566 patients (89%) (n ¼ 186 [92.5%] with

the first month post-ablation in 50% of the patients

ERAT [PAF: 108 (91.5%), NPAF: 78 (94%); p ¼ 0.51]

(60% PAF and 40% NPAF; p ¼ 0.77), with an addi-

and n ¼ 380 [87.2%] without ERAT [PAF: 273 (86.7%),

tional 31% of patients (32.3% PAF and 29.3% NPAF;

NPAF:107 (89.2%); p ¼ 0.48]) were taking at least one

p ¼ 0.65) experiencing an ERAT within the second

AAD for the first 3 months after the procedure. Based

month post-ablation. Rates of ERAT progressively

on logistic regression, use of AAD post-procedure did

declined over the 3-month blanking period, with

not show any significant influence on the likelihood

nearly 80% of ERAT (83% PAF and 78% NPAF)

of LR (p ¼ 0.11).

occurring within the first 2 months post-ablation (Figure 2). In patients with ERAT, LR of AF symp-

ROC CURVE FOR DEFINING THE IDEAL CUTOFF

toms occurred in 137 patients (68.5%), with a slightly

POINT FOR THE BLANKING PERIOD POST-PVI. An

higher rate in NPAF patients (65.3% PAF and 73.1%

ROC curve was used to determine the accurate

NPAF; p ¼ 0.09). In patients without ERAT, LR

cutoff time point based on sensitivity and specificity

occurred in 105 patients (24%), with similar rates in

of ERAT for LR to better define the blanking period

both PAF and NPAF patients (24.7% PAF and 22.3%

post-PVI. Our analysis determined 23 days to be the

NPAF; p ¼ 0.60). Rates of LR were significantly

ideal cutoff point for the blanking period with

different between those experiencing ERAT and those

area under the curve of 0.70 (95% confidence interval

without ERAT, regardless of AF type (73.1% vs. 24%;

[CI]: 0.633 to 0.778; p < 0.001), along with sensitivity

p ¼ 0.0001). Timing of ER during the blanking period

and specificity of 69.2% and 61.2%, respectively

was significantly associated with the rate of LR, with

(Figure 1).

the likelihood of LR progressively increasing in

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017

- 2017:-–-

F I G U R E 1 ROC Curve to Determine Ideal Window for Blanking

Period

Alipour et al. Blanking Period Post-PVI

association with the delay between ablation and the ERAT event (51% in the first month, 75.8% in second month, and 92.1% in third month; p ¼ 0.0001) in both patients with PAF (46% in the first month, 68% in the second month, and 95% in third month; p ¼ 0.0001) and those with NPAF (57% in the first month, 87% in the second month, and 88% in third month; p ¼ 0.008) (Figure 3). ERAT was stratified by type of arrhythmia (AF vs. AFL vs. AT). Overall, 157 patients (78.5%) manifested AF, whereas 28 patients (14%) and 15 patients (7.5%) manifested AFL and AT, respectively. Of the 242 patients with LR, 188 manifested AF (77.7%), 32 AFL (13.2%), and 22 AT (9.1%). Furthermore, there was no relationship between the type of ERAT manifestation and the likelihood of LR (p ¼ 0.11). Arrhythmia documented at the time of LR was consistent with that documented during ERAT. Using univariate logistic regression, potential factors contributing to the manifestation of LR were

ROC curve showing sensitivity and specificity for early recurrence of atrial tachyarrhythmia (ERAT) in the first 90 days after pulmonary vein isolation. Area under the curve: 0.701 (95% CI:

analyzed. This model revealed age (odds ratio [OR]: 1.02; 95% CI: 1.01 to 1.04; p ¼ 0.001), ERAT (OR: 6.73;

0.633 to 0.778) (p ¼ 0.001). Ideal point: 23 days; sensitivity:

95% CI: 4.65 to 9.73; p ¼ 0.0001; ERAT in first month:

0.692; specificity: 0.612 (p ¼ 0.0001). ROC ¼ receiver-operating

OR: 4.35; 95% CI: 2.76 to 6.85; p ¼ 0.0001; ERAT in

characteristic.

second month: OR: 9.06; 95% CI: 4.92 to 16.67; p ¼ 0.0001; ERAT in third month: OR: 20.80; 95% CI: 7.92 to 54.65; p ¼ 0.0001), number of failed AADs (OR: 1.33; 95% CI: 1.14 to 1.57; p ¼ .001), and history of

F I G U R E 2 Relationship Between ERAT and LR Post-PVI

Distribution of ERAT post-PVI within initial 30, 60, and 90 days, and relationship between ERAT and likelihood of LR manifestation stratified by type of atrial fibrillation (PAF vs. NPAF). ERAT ¼ early recurrence of atrial tachyarrhythmia; LR ¼ late recurrence; NPAF ¼ persistent atrial fibrillation; PAF ¼ paroxysmal atrial fibrillation; PVI ¼ pulmonary vein isolation.

5

6

Alipour et al.

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017 - 2017:-–-

Blanking Period Post-PVI

F I G U R E 3 Likelihood of LR Based on ERAT Time

T A B L E 2 Univariate Logistic Regression Model of the Likelihood

of Late Recurrence of AF Odds Ratio (95% Confidence Interval)

p Value

Age

1.02 (1.01–1.04)

0.001

Male

0.98 (0.70–1.39)

0.93

BMI

1.00 (0.98–1.02)

0.68

6.73 (4.65–9.73)

0.0001

ERAT First month Second month Third month

4.35 (2.76–6.85)

0.0001

9.06 (4.92–16.67)

0.0001

20.80 (7.92–54.65)

0.0001

AAD post-ablation

1.61 (0.93–2.78)

0.14

No. of failed AADs

1.33 (1.14–1.57)

0.001

Failed amiodarone

1.66 (1.14–2.41)

0.009

LA size

1.01 (0.99–1.03)

0.20

Ejection fraction $60%

Likelihood of LR based on days to first episode of ERAT. Past 23 days, any manifestation of ERAT increases the odds of LR. ER ¼ early recurrence; other abbreviations as in Figure 2.

1.00

-

50%

1.17 (0.10–13.06)

0.90

40%

1.63 (0.14–19.54)

0.71

#30%

1.77 (0.13–25.52)

0.66

Structural heart disease Dilated cardiomyopathy

0.93 (0.57–1.50)

0.77

0.98 (0.50–1.91)

0.96

Ischemic cardiomyopathy

1.02 (0.49–2.15)

0.95

Hypertrophic cardiomyopathy

0.95 (0.37–2.44)

0.92

the patient having failed amiodarone (OR: 1.66; 95%

Type 2 diabetes mellitus

1.61 (0.93–2.80)

0.09

CI: 1.14 to 2.41; p ¼ 0.009) as significant predictors of

Congestive heart failure

1.55 (0.93–2.60)

0.09

LR (Table 2). Gender, body mass index, ejection

Hypertension

0.94 (0.68–1.31)

0.74

fraction, AAD use after ablation, and structural heart

AF ¼ atrial fibrillation; other abbreviations as in Table 1.

disease were not significant predictors of LR. Factors deemed as significant univariate predictors were then entered into the multivariate logistic regression

Although more than 90% of these arrhythmias take

(Tables 3 and 4).

place within the first 10 days post-ablation, ERAT may

Subsequently, two multivariate logistic regression

also occur up to 3 months post-ablation (7,15–17). The

models were built using timing of ERAT classified

rate of early recurrence in our study was 50% in

according to the month after ablation and stratified

the first month, 31% in the second month, and 19% in

around the 23-day cutoff. These models revealed

the third month, and cumulatively 80% of all ERAT

progressively greater odds for LR based on ERAT in

occurred in the first 2 months. Our study showed a

the first (OR ¼ 4.22; p ¼ 0.0001), second (OR ¼ 9.03,

significant decline in the incidence of AF over the

p ¼ 0.0001), and third months (OR ¼ 19.43;

3-month period post-PVI, whereas the likelihood of

p ¼ 0.0001) after the procedure compared to patients

LR progressively increased with timing to ERAT dur-

without ERAT during the conventional blanking

ing the first 3 months after ablation (OR: 4.22, 9.03,

period (Table 2). Greater odds for LR depending on

and 19.43 for the first, second, and third month post-

ERAT stratified around the 23-day cutoff (OR ¼ 4.41;

ablation, respectively; p < 0.001). Of note, all patients

p ¼ 0.0001 for ERAT before 23 days vs. OR ¼ 9.27 for

who were taking an AAD before the procedure

ERAT after 23 days; p ¼ 0.0001) compared to patients

continued taking their medication for at least

without ERAT during the 3-month post-ablation

3 months after ablation. Patients who did not mani-

period (Table 3). Both models included adjustments

fest any ERAT in the initial 90 days discontinued

for the impact of other covariates, including age, AF

taking their AADs, whereas those who experienced at

type (PAF vs. NPAF), number of failed AADs, and

least one episode of ERAT lasting at least 30 seconds

history of the patient having failed amiodarone.

within the same time period continued taking the medication. To address this issue further, we ran a

DISCUSSION

logistic regression model to determine the effect of ongoing AAD therapy on ERAT. Based on this anal-

By considering the blanking period of 3 months,

ysis, ongoing AAD therapy did not significantly affect

studies have reported ERAT rates ranging from 15.9%

the likelihood of LR (p ¼ 0.11). Our findings are

to 65%, with a pooled estimate of 37.8% (7–9).

consistent with a study by Leong-Sit et al. (18), which

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017

Alipour et al.

- 2017:-–-

Blanking Period Post-PVI

T A B L E 3 Multivariate Logistic Regression Model of the

T A B L E 4 Multivariate Logistic Regression Model of the

Likelihood of Late Recurrence of AF Using the Empirically Derived

Likelihood of Late Recurrence of AF Using the Conventional

Criterion of ERAT Time

Criterion of ERAT Time

Odds Ratio (95% Confidence Interval)

Standard Error

Odds Ratio (95% Confidence Interval)

Standard Error

Age

1.03 (1.01–1.05)

0.01

0.005

p Value

Age

1.03 (1.00–1.05)

0.01

NPAF

1.00 (0.67–1.47)

0.20

0.008

0.97

NPAF

1.02 (0.69–1.52)

0.20

No. of failed AADs

1.13 (0.93–1.40)

0.90

0.10

0.23

No. of failed AADs

1.13 (0.93–1.40)

0.10

Failed amiodarone

1.22 (0.75–1.95)

0.21

0.24

0.42

Failed amiodarone

1.23 (0.77–2.0)

0.24

0.42

p Value

ERAT time

ERAT time

No ERAT

1

No ERAT

1

#23 days

4.41 (2.67–7.28)

0.26

0.0001

First month

>23 days

9.27 (5.72–15.03)

0.25

0.0001

4.22 (2.65–6.71)

0.24

0.0001

Second month

9.03 (4.85–16.83)

0.32

0.0001

Third month

19.43 (7.33–51.52)

0.50

0.0001

Abbreviations as in Tables 1 and 2. Abbreviations as in Table 1.

investigated the effects of AADs on LR of AF.

ablation that was associated with early AF recurrence

Although use of AADs had some effect on reducing

during the same time frame. Additionally, the degree

early recurrences, there was no significant differences

of inflammatory response was also associated with

between the control group and an interventional

early AF recurrence during this time. Their study

group with respect to the risk of LRs. They are also

demonstrated that half of the patients who experi-

consistent with a study by Joshi et al. (19), who

enced early recurrence within 30 days had LR,

showed the highest ERAT prevalence rate of 54% in

whereas all patients who had episodes between 30

the first 2 weeks and a subsequent decline to a min-

days to 3 months after ablation experienced LR as

imum of 22% ERAT in the third month.

well; therefore, they concluded that inflammatory

Several factors have been identified as predictors of

responses after ablation usually are limited to the first

late arrhythmia recurrence post-PVI. ERAT has been

month post-ablation (10). According to these findings,

described as an independent predictor of LR in several

there has been some discussion among investigators

studies (2,5,9,19,20). Our group previously described

regarding the modifying effect of steroids and other

early recurrence as the strongest predictor of LR, with

anti-inflammatory substances after ablation and

a hazard ratio of 4.87 (95% CI: 3.88 to 6.12; p ¼ 0.0001)

findings supporting its efficacy in ERAT reduction as

(14,21). This result was comparable to the study by

well as improving mid-term outcomes (9,10,23). A

Maroto et al. (15) of 106 patients post-ablation, which

study by Koyama et al. (24) of 186 patients post-PVI,

showed early recurrence as a risk factor for LR with a

assessing proinflammatory factors and their rela-

hazard ratio of 3.45. Several other prior studies

tionship with AF recurrence, revealed that patients

(3,7,9,12) have demonstrated that a majority of pa-

with immediate recurrence (occurring within 3 days

tients experiencing ERAT during the first 3 months

after ablation) had a higher AF-free rate of 76% than

post-PVI are more vulnerable to LR (53.7% patients

did those with early AF recurrence (occurring be-

with ERAT vs. 6.9% patients without ERAT) (12). This

tween 4 and 30 days post-ablation) of 30% during

study supports a significant association between

6-month follow-up. The exact time at which the

timing of ERAT and the rate of LR. Rate of LR was 51%

predictive power of early recurrences for LRs rises,

in those who had ERAT in the first month, and 75.8%

reflecting when the early transient factors responsible

and 92.1% among those who had ERAT in the second

for arrhythmia recurrences yield to PV reconnection,

and third months, respectively (p < 0.001). Similarly,

was not firmly established (2,5,9,19,20,25–28).

Bertaglia et al. (22) reported LR rates of 56.7% and 80%

In a study by Hsieh et al. (29), 1 month was

with ERAT in the first month and 2 subsequent

described as the time required for recovery of auto-

months, respectively.

nomic function (including changes in heart rate and

Transient factors (post-procedure inflammation,

heart rate variability). Maturation of the ablation le-

autonomic imbalance, and lesion maturation time)

sions in other studies has been estimated to occur

likely play an important role in some early episodes of

within 1 to 2 weeks after ablation (3,12).

ERAT (3,7,9,12). A study by Lim et al. (10) of 90 pa-

A recent study by Das et al. (12) of 40 patients

tients with AF assessing proinflammatory responses

with paroxysmal AF, setting the blanking period at

after radiofrequency ablation showed an inflamma-

1 month, showed that any ERAT beyond 4 weeks

tory response and myocardial injury 3 days after

post-PVI

was

significantly

associated

with

PV

7

8

Alipour et al.

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017 - 2017:-–-

Blanking Period Post-PVI

electrophysiological

ROC curve yielded the best combination of sensi-

study. In their study, patients were monitored using a

tivity and specificity for ERAT predicting LR at 23

30-second electrocardiogram daily as well as for

days post-ablation. Any ERAT beyond 23 days yielded

reconnections

at

follow-up

symptoms. Regardless of their symptom status,

an insignificant increase in specificity for a significant

the patients underwent electrophysiological study

tradeoff in sensitivity. The rate of LR for patients with

2 months after PVI to evaluate the number of recon-

ERAT did not differ between PAF and NPAF groups.

nected PVs and to assess the correlation between the

The novelty of this approach compared to the previ-

first and second month’s ERAT with PV reconnection.

ous studies involves use of ROC analysis to establish

The study demonstrated that ERAT occurring in the

the ideal cutoff time after which any early recurrence

first month was not correlated with PV reconnection,

has a significantly higher likelihood of predicting late

whereas ERAT occurring in or persisting into the

arrhythmia recurrences. Our results confirm 23 days

second month was strongly correlated with PV

post-ablation as the time point beyond which any

reconnections. On the contrary, a study using cardiac

ERAT heralds an established substrate for arrhythmia

magnetic resonance imaging to evaluate the lesions

recurrence and should be treated as such rather than

reported 3 months as the time required for formation

“blanked out.” Although our findings are in general

of left atrial scar tissue after ablation (30). A study by

agreement with previous studies, they narrow down

Themistoclakis et al. (20) investigated 1,298 patients

the duration of the “blanking period” after PVI to

who were followed for 41
10 months at 1, 3, 6, 9, and

23 days.

12 months post-ablation and then every 6 months thereafter using a monitoring protocol similar to that used in our study. Patients were further followed using transtelephonic rhythm transmissions and 48-hour Holter monitoring as symptoms dictated, and their ERATs were classified according to the time of their first occurrence. According to their study, the incidence of ERAT was higher in the first month, specifically in the first week (incidence rate of 81%), with a decline in the second and third months (incidences of 10% and 9%, respectively). LR was more frequent in patients who experienced ERAT in the third (98%) and second months (69%) than in those who experienced ERAT in the first month (44%). Additionally, the investigators showed that the first recurrence in the second and third months after ablation has a stronger association with LR than ERAT during the first month (first month OR: 20; second month OR: 54; third month OR: 1052; p < 0.001 for all findings). Our findings demonstrate that after controlling for the effects of age, AF type, left atrial size, and other comorbidities, ERAT within the second and third months of the blanking period is a significant predictor of LR, with OR of 9.03 (p ¼ 0.0001) and 19.43 (p ¼ 0.0001), respectively. However, no relationship was found between ERAT type (AF vs. AFL vs. AT) manifestation and the likelihood of LR (p ¼ 0.11). The likelihood of LR in patients with ERAT during

STUDY LIMITATIONS. This was an observational,

nonrandomized prospective cohort study conducted at single large tertiary electrophysiology center. All ablation procedures were performed using radiofrequency energy, and the findings may not be applicable to ablation procedures performed using other commonly used forms of energy delivery such as cryoablation. Finally, we did not use intensive monitoring

by

transtelephonic

monitoring

or

implantable loop recorder to rule out asymptomatic recurrences during the first 90 days post-PVI and instead followed a more feasible clinical protocol.

CONCLUSIONS Based on the findings of this study, 23 days or about 3 weeks (rather than 3 months) post-AF ablation appears to be an appropriate cutoff for the blanking period. Any recurrence of arrhythmia after the initial 23 days post-ablation should be considered clinically significant. Further studies should be directed to evaluate the ideal timing of a repeat ablation procedure for patients experiencing early recurrences, stratified into early intervention for any recurrence following the 3-week cutoff versus a strategy of waiting to intervene after recurrences occur following the current conventional 3-month blanking period.

the initial 23 days post-PVI was 4.41. In contrast, patients with recurrent episodes after 23 days had an LR

ADDRESS

likelihood of 9.27. In addition, comparing each month

Khaykin, Heart Rhythm Program, Southlake Regional

FOR

separately revealed a likelihood of 4.22, 9.03, and

Health Centre, #602-581 Davis Drive, Newmarket,

19.43 for LR during the first, second, and third

Ontario

months post-ablation, respectively.

utoronto.ca.

L3Y

CORRESPONDENCE:

2P6,

Canada.

E-mail:

Dr.

Yaariv

y.khaykin@

JACC: CLINICAL ELECTROPHYSIOLOGY VOL.

-, NO. -, 2017

Alipour et al.

- 2017:-–-

Blanking Period Post-PVI

PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE: This

TRANSLATIONAL OUTLOOK: By narrowing the

study demonstrates that the traditional 3-month definition

blanking windows to 23 days, clinicians can better inform

of the blanking period post-PVI is not accurate because

and set more realistic expectations for patients and their

recurrences happening past the first month post-ablation

families.

are a significant predictor of subsequent episodes of AF.

REFERENCES 1. Liu J, Fang P-H, Hou Y, et al. The value of transtelephonic electrocardiogram monitoring system during the “blanking period” after ablation of atrial fibrillation. J Electrocardiol 2010;43:667–72. 2. Mulder AA, Wijffels MC, Wever EF, Boersma LV. Early recurrence of atrial fibrillation as a predictor for 1-year efficacy after successful phased RF pulmonary vein isolation: evaluation of complaints and multiple Holter recordings. Int J Cardiol 2013; 165:56–60. 3. Oral H, Knight BP, Özaydın M, et al. Clinical significance of early recurrences of atrial fibrillation after pulmonary vein isolation. J Am Coll Cardiol 2002;40:100–4. 4. Andrade JG, Macle L, Khairy P, et al. Incidence and significance of early recurrences associated with different ablation strategies for AF: a STARAF substudy. J Cardiovasc Electrophysiol 2012; 23:1295–301. 5. Kawasaki S, Tanno K, Ochi A, et al. Recurrence of atrial fibrillation within three months after pulmonary vein isolation for patients with paroxysmal atrial fibrillation: analysis using external loop recorder with auto-trigger function. J Arrhythm 2015;31:88–93. 6. Roux J-F, Zado E, Callans DJ, et al. Antiarrhythmics after ablation of atrial fibrillation (5A Study). Circulation 2009;120:1036–40. 7. Altman RK, Proietti R, Barrett CD, et al. Management of refractory atrial fibrillation post surgical ablation. Ann Cardiothorac Surg 2014;3:91. 8. Eitel C, Piorkowski C, Hindricks G. Clinical relevance and management of early recurrences after catheter ablation of atrial fibrillation. Exp Rev Cardiovasc Ther 2011;9:849–52. 9. Andrade JG, Khairy P, Verma A, et al. Early recurrence of atrial tachyarrhythmias following radiofrequency catheter ablation of atrial fibrillation. Pacing Clin Electrophysiol 2012;35:106–16.

procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design. Europace 2012;14:528–606. 12. Das M, Wynn GJ, Morgan M, et al. Recurrence of atrial tachyarrhythmia during the second month of the blanking period is associated with more extensive pulmonary vein reconnection at repeat electrophysiology study. Circ Arrhythm Electrophysiol 2015;8:846–52. 13. Phillips KP. Role of inflammation in initiation and perpetuation of atrial fibrillation: a systematic review of the published data. J Atr Fibrillation 2013;6:62–7. 14. Khaykin Y, Oosthuizen R, Zarnett L, et al. Clinical predictors of arrhythmia recurrences following pulmonary vein antrum isolation for atrial fibrillation. J Cardiovasc Electrophysiol 2009;22:1206–14. 15. Maroto LC, Carnero M, Silva JA, et al. Early recurrence is a predictor of late failure in surgical ablation of atrial fibrillation. Interact Cardiovasc Thorac Surg 2011;12:681–6. 16. Damiano RJ, Schwartz FH, Bailey MS, et al. The Cox maze IV procedure: predictors of late recurrence. J Thorac Cardiovasc Surg 2011;141:113–21. 17. Ishii Y, Gleva MJ, Gamache MC, et al. Atrial tachyarrhythmias after the maze procedure: incidence and prognosis. Circulation 2004;110:II164–8. 18. Leong-Sit P, Roux JF, Zado E, et al. Antiarrhythmics after ablation of atrial fibrillation (5A Study). Circ Arrhythm Electrophysiol 2011;4:11–4. 19. Joshi S, Choi AD, Kamath GS, et al. Prevalence, predictors, and prognosis of atrial fibrillation early after pulmonary vein isolation: findings from 3 months of continuous automatic ECG loop recordings. J Cardiovasc Electrophysiol 2009;20: 1089–94.

10. Lim HS, Schultz C, Dang J, et al. Time course of inflammation, myocardial injury, and prothrombotic response after radiofrequency catheter ablation for atrial fibrillation. Circ Arrhythm Electrophysiol 2014;7:83–9.

20. Themistoclakis S, Schweikert RA, Saliba WI, et al. Clinical predictors and relationship between early and late atrial tachyarrhythmias after pulmonary vein antrum isolation. Heart Rhythm

11. Calkins H, Kuck KH, Cappato R, et al. 2012 HRS/ EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation: recommendations for patient selection,

21. Alipour P, Khaykin Y, Pirbaglou M, et al. Predictors of arrhythmia recurrence following pulmonary vein antrum isolation. Can J Cardiol 2015; 31:S223–4.

22. Bertaglia E, Stabile G, Senatore G, et al. Predictive value of early atrial tachyarrhythmias recurrence after circumferential anatomical pulmonary vein ablation. Pacing Clin Electrophysiol 2005;28:366–71. 23. Koyama T, Tada H, Sekiguchi Y, et al. Prevention of atrial fibrillation recurrence with corticosteroids after radiofrequency catheter ablation: a randomized controlled trial. J Am Coll Cardiol 2010;56:1463–72. 24. Koyama T, Sekiguchi Y, Tada H, et al. Comparison of characteristics and significance of immediate versus early versus no recurrence of atrial fibrillation after catheter ablation. Am J Cardiol 2009;103:1249–54. 25. Mugnai G, de Asmundis C, Hünük B, et al. Second generation cryoballoon ablation for paroxysmal atrial fibrillation: predictive role of atrial arrhythmias occurring in the blanking period on the incidence of late recurrences. Heart Rhythm 2016;13:845–51. 26. Andrade JG, Khairy P, Macle L, et al. Incidence and significance of early recurrences of atrial fibrillation after cryoballoon ablation: insights from the Multicenter STOP AF Trial. Circ Arrhythm Electrophysiol 2014;7:69–75. 27. Gaztañaga L, Frankel DS, Kohari M, Kondapalli L, Zado ES, Marchlinski FE. Time to recurrence of atrial fibrillation influences outcome following catheter ablation. Heart Rhythm 2013; 10:2–9. 28. Liang JJ, Elafros MA, Chik WW, et al. Early recurrence of atrial arrhythmias following pulmonary vein antral isolation: timing and frequency of early recurrences predicts long-term ablation success. Heart Rhythm 2015;12:2461–8. 29. Hsieh M-H, Chiou C-W, Wen Z-C, et al. Alterations of heart rate variability after radiofrequency catheter ablation of focal atrial fibrillation originating from pulmonary veins. Circulation 1999; 100:2237–43. 30. Badger TJ, Oakes RS, Daccarett M, et al. Temporal left atrial lesion formation after ablation of atrial fibrillation. Heart Rhythm 2009;6:161–8.

2008;5:679–85. KEY WORDS atrial fibrillation, blanking period, early recurrence, early recurrence of atrial tachyarrhythmia, late recurrence, pulmonary vein isolation

9