Left atrial diameter, CHA2DS2-VASc score and type of atrial brillation predict pulmonary vein isolation outcome

Purpose Pulmonary vein isolation (PVI) is the most effective treatment strategy for atrial brillation (AF). This study evaluated Pulmonary Vein Isolation Outcome Degree (PVIOD) as a new semi-quantitative measure for PVI success after a 7-year follow-up and determined predictors associated with PVIOD. Methods We enrolled 117 patients with symptomatic AF who underwent PVI and after a 7-year follow-up applied PVIOD with 4 possible outcomes. PVIOD 1 group included patients with successful single PVI. PVIOD 2 group included patients with ecacy after ≥ 2 re-PV isolation and/or additional substrate modication (ASM). PVIOD 3 group contained subjects with clinical success after PVI±ASM. Patients with procedural and clinical failure were in PVIOD 4 group. Results In multivariate ordinal logistic regression analysis PVIOD was independently associated with longstanding persistent AF with paroxysmal AF as referent category: odds ratio (OR) 4.1, 95% condence interval (95% CI) 1.3-12.8 (P=0.014), left atrial (LA) diameter: OR 1.2, 95% CI 1.1-1.3 (P<0.001) and CHA 2 DS 2 -VASc score: OR 1.5, 95% CI 1.0-2.2 (P=0.039). LA size, CHA 2 DS 2 -VASc score and AF type predicted 7-year probability for procedural and procedural with clinical failure. LA diameter >41mm (AUC 0.741, 95% CI 0.6-0.8, P<0.001) and CHA 2 DS 2 -VASc score ≥ 2 (AUC 0.718, 95% CI 0.6-0.8, P<0.001) predicted long-term procedural and clinical failure. Conclusion PVIOD is a new classication for PVI success. LA diameter, CHA 2 DS 2 -VASc score and AF type are with PVIOD and predict procedural and procedural with after the 7-year follow-up. CHA DS 2 -VASc score ≥ predict long-term failure.

Numerous trials evaluated the e cacy of PVI and usually used a qualitative measure with 2 possible outcomes, successful or ineffective procedure [2,7]. Our study aimed to evaluate Pulmonary Vein Isolation Outcome Degree (PVIOD) as a new semi-quantitative measure for PVI success after a 7-year follow-up and to determine parameters for predicting the PVIOD.

Methods
This prospective study included 124 consecutive patients with symptomatic AF who underwent PVI at the Institute for Cardiovascular Diseases Dedinje, from January 2012 to December 2013. AF was classi ed according to the 2012 HRS/EHRA/ECAS Consensus Statement on Catheter and Surgical Ablation of AF [10]. Paroxysmal or persistent AF was de ned, respectively, if there was a success or failure to terminate AF spontaneously or with intervention within 7 days of onset. Longstanding persistent AF was diagnosed if arrhythmia lasted more than 1 year. The study excluded those who had permanent AF, pregnancy, acute reversible causes of AF, myocardial infarction that occurred within 3 months, moderate-to-severe valvular stenosis or regurgitation and coronary artery disease indicated for revascularization procedure.
The trial was approved by the Ethics Committee of the Institute for Cardiovascular Diseases Dedinje.
Work was conducted in accordance with the Declaration of Helsinki. All patients signed informed consent prior to inclusion in the study. Initially, all patients were evaluated for age, gender, AF duration, AF type, clinical symptoms and signs, body mass index, CHA 2 DS 2 -VASc score, risk factors for cardiovascular disease (hypertension, hypercholesterolemia and diabetes mellitus), the presence of structural heart disease and usage of antiarrhythmic drugs (AAD) prior to the PVI. All medical documents, 12-lead electrocardiograms (ECG) and 24-h Holter monitorings were reviewed for the presence of AF.
Transthoracic echocardiography measured the end-systolic LA size through antero-posterior diameter from parasternal long-axis view. Transesophageal echocardiography (TEE) was used to exclude the presence of thrombus in left auricula and LA. 64-multislice computed tomography of pulmonary veins (PVs) together with images from TEE provided better visualization of LA and PVs anatomy because of signi cant inter-and intra-patient variability in the number, size and bifurcation of the PVs.

Ablation protocol
In all patients, the rst PVI included only catheter ablation of PV antrum without additional substrate modi cation (ASM). After trans-septal puncture, an intravenous bolus of heparin (5.000 IU) was administrated, followed by additional doses to maintain activated clotting time between 300 and 350s.
During the procedure, the antrum of ipsilateral PVs was widely encircled with an irrigated tip catheter (ThermoCool Navi-Star, Biosense-Webster) with point-by-point ablation lesions using the CARTO system. The RF energy was applied at a target temperature of 43 degrees Celsius with a power limit of 30-35 W for 30-60 seconds. Entrance block, exit block and non-inducibility of AF con rmed the initial success of PVI. When typical atrial utter (AFL) was documented, cavotricuspid isthmus was ablated.
All subjects received AAD for 3 months after initial PVI, a daily oral dose of either 200mg amiodarone or 150mg propafenone 3 times a day or 80mg sotalol twice a day. Oral anticoagulation was administrated for 6 months after catheter ablation in all patients and continued after that in those with a high CHA 2 DS 2 -VASc score. Repeated procedures were performed in case of AF recurrences with re-PV ablation in patients with recovered PV conduction. In those subjects without PV reconnection, ablation of complex fractionated atrial electrograms and/or additional substrate strategies were applied based on observations made during the electrophysiology study. ASM included LA roof line, mitral annular isthmus, anterior line, inferior line, lines around superior vena cava and coronary sinus. 2.2 Follow-up AF type and burden were determined 1, 3 and 6 months after catheter ablation, and later by at least annual clinical review. During visits, ECG and 24-h Holter monitoring were performed for rhythm evaluation. Recurrences of AF, AFL and atrial tachycardia (AT) were de ned as episodes of arrhythmias lasting for more than 30s. Subjects who could not be reached were excluded from the study.
We introduced a new semi-quantitative method for measuring PVI success, i.e. Pulmonary Vein Isolation Outcome Degree (PVIOD) with 4 possible outcomes after a 7-year follow-up: PVIOD 1 group included patients with successful single PVI; PVIOD 2 group included subjects with e cacy after ≥ 2 re-PV isolation and/or ASM; Patients with clinical success after PVI ± ASM regardless of procedure number were in PVIOD 3 group; PVIOD 4 group contained patients with procedural and clinical failure despite PVI ± ASM. Clinical success was de ned as a signi cant reduction (≥ 70%) in the number and duration of AF episodes with or without AAD.

Statistical analysis
Continuous variables are presented as the mean ± standard deviation or median and interquartile range, while categorical variables are showed as absolute values and percentages. The differences between continuous values were assessed using an unpaired Student's t-test when the data were normally distributed, a Mann-Whitney test for skewed variables and λ 2 test or Fisher's exact test for nominal variables. Ordinal logistic regression analysis was used to identify the predictors for PVIOD. Multivariate (MV) logistic regression analysis included variables with a statistical signi cance of 0.05 on the univariate (UV) logistic regression model. Nomogram as a two-dimensional diagram was used for graphical presentation of MV logistic regression model prediction. Receiver operating characteristic (ROC) curves were applied to determine the cut-off points for LA diameter and CHA 2 DS 2 -VASc score.
Signi cance was established at P < 0.05.

Discussion
To the best of our knowledge, this is the rst study that presented and used semi-quantitative assessment for PVI success which we called Pulmonary Vein Isolation Outcome Degree. PVIOD 1 stands for patients with the best result of catheter ablation while PVIOP 4 refers to the worst outcome of patients. Until now the trials usually used a qualitative measure of PVI success with 2 possible outcomes: effective or unsuccessful procedure [2,5,7].
After a 7-year follow-up, a single PVI had a modest success of 32.5% in our study group, with the best result of 40.2% in patients with paroxysmal AF, 20.8% in persistent AF and as low as 12.5% in longstanding persistent AF. Numerous trials detected long-term success after single catheter ablation from 29-57% [2,4,5,7]. Repeated ablations without ASM, after the 7-year follow-up, increased cumulative success in our study to 57.3%. Additional substrate strategies with an e cacy rate of 4.3% made very small bene t to procedural success which increased to 61.6%. In the study by Teunissen et al., the e cacy of additional substrate ablations was 4.2%, which magni ed procedural success to 62.5% [2]. The work of Ouyang et al. showed that patients with paroxysmal AF and normal left ventricular ejection fraction after single PVI (4.8-year follow-up) had a success rate of 46.6%, after the second procedure 73.9% and after the third ablation the e cacy was 79.5% [4]. Our results were satisfactory in patients with paroxysmal AF; after the rst PVI and the 7-year follow-up, the success rate was 40.2%, and after the last ablation success was 69.4%.
In our patients, long-term clinical success was reported in 14.5% of subjects which increased cumulative success to 76.1% with excellent result in paroxysmal AF (85%), good outcome in persistent AF (70.8%) and modest e cacy in longstanding persistent AF (43.7%). These data show the importance of clinical improvement and the role of AAD therapy in AF management. Therefore, long-term clinical success (on top of procedural success) is a very signi cant outcome. Teunissen et al. found that clinical improvement on or off AAD was 25% and increased cumulative success to 87.5% in the total population of 509 study patients [2].
The most challenging issue in managing patients with AF is identifying those predictors for unsuccessful ablation. A large number of studies determined the risk factors for poor outcome following PVI as nonparoxysmal AF (particularly longstanding persistent AF), LA dilatation and higher CHA 2 DS 2 -VASc score, which was con rmed in our trial [2,8,9,12]. In our study, MV ordinal logistic regression analysis found an independent association between PVIOD and LA diameter, CHA 2 DS 2 -VASc score and AF type. LA diameter > 41mm was a high predictor of procedural failure (AUC 0.726) and procedural with clinical failure (AUC 0.741) in our patients. LA enlargement is a well-known risk factor for AF, associated with atrial electroanatomical remodeling, which plays the main role in the perpetuation and progression of AF [11]. The computational model con rmed that a critical LA effective conducting size > 40mm was required for sustained multiple wavelet reentry [11]. A meta-analysis of D'Ascenzo et al. showed that persistent AF, LA diameter > 50mm and arrhythmia recurrences within the rst month after catheter ablation are the most powerful predictors of procedural failure [12]. PVI success is highly dependent on atrial disease stage and enlarged LA is widely accepted as a predictor of AF recidivism.
Our study determined that CHA 2 DS 2 -VASc score ≥ 2 was a predictor of poor PVI outcome. Patients with CHA 2 DS 2 -VASc scores 0 and 1 displayed successful ablations, but those with other scores showed high predictive value for procedural failure (AUC 0.711) and procedural with clinical failure (AUC 0.718). A score of ≥ 2 for both CHADS 2 and CHA 2 DS 2 -VASc scores had the highest predictive value for AF recurrence after single catheter ablation for paroxysmal AF in the study of Letsas et al. [13]. In a relatively small cohort of patients with longstanding persistent AF, it was shown that CHA 2 DS 2 -VASc score ≥ 3 and renal dysfunction were signi cantly associated with ablation failure within 31 months [14]. Other scores, like APPLE, DR-FLASH and MB-LATER predicted electro-anatomical substrate and arrhythmia recurrences in patients with AF who underwent catheter ablation [15]. The identi cation of speci c markers before PVI in patients with AF de nes the best candidates for the procedure and future ablation responders. Prediction of arrhythmia recurrences is complex because of the in uence of disease stage, patients' comorbidities, operators' experience, equipment for ablation, etc.
In the current study, UV ordinal logistic regression analysis showed an association of PVIOD with structural heart disease and diabetes mellitus, but not in the MV model. Structural heart disease is a wellknown predictor of early and very late recurrence of AF after RF ablation [9]. Cardiac risk factors such as hypertension and diabetes mellitus are closely correlated with in ammation and consequently with atrial brosis and remodeling [16]. Although in our MV analysis it does not reach signi cance, the importance of diabetes mellitus prevention and treatment in AF patients is obvious.
In our study nomograms presented a seven-year probability for procedural failure and for procedural with clinical failure, which were predicted with 3 risk factors: LA diameter, CHA 2 DS 2 -VASc score and type of AF.
Determination of AF type, CHA 2 DS 2 -VASc score and measurement of LA diameter could provide an easy and available way of predicting catheter ablation outcome.
PVIOD is a new PV ablation outcome scoring system. This classi cation considers the number and e cacy of PVI, clinical success, optional additional substrate modi cation and antiarrhythmic therapy.
The scoring of PVIOD is a novel way of assessing PVI success since other studies with standard instruments rate PVI e cacy by procedural success or failure. The novelty of our results is an independent association between PVIOD and 3 clinical parameters: LA diameter, CHA 2 DS 2 -VASc score and type of AF, after a 7-year follow-up. These risk factors are very convenient to predict future PVI outcomes and easy to be measured already at baseline with cardiac echocardiography and clinical evaluation of patients. By using nomograms we can calculate both long-term procedural failure and procedural with clinical failure for patients with AF who will be treated with PVI.
This study showed signi cant major ablation complications in 3.3% of procedures, which is acceptable and can compare with the result (2.5-8%) from other eminent electrophysiology laboratories [17,18]. There were no deaths connected to catheter ablations in our patients which are the excellent results.

Study Limitations
This study is limited because of the relatively small sample size and being a single-centre trial. PVIOD as a new semi-quantitative measure for PVI success needs to be tested in future studies with a more homogenous and larger group of patients, especially to con rm the in uence of diabetes mellitus and structural heart diseases in the genesis of atrial brillation.

Conclusion
This study developed Pulmonary Vein Isolation Outcome Degree as a new semi-quantitative measure for PVI success. LA diameter, CHA 2 DS 2 -VASc score and AF type were independently associated with PVIOD and predicted PVI outcome after the 7-year follow-up. LA size > 41mm and CHA 2 DS 2 -VASc score ≥ 2 predict long-term procedural and procedural with clinical failure. Funding This research did not receive any speci c grant from funding agencies in the public, commercial, or not-for-pro t sectors.

Abbreviations
Code availability SPSS statistics 22.0 software (IBM, USA) was used for statistical analysis.
Consent to participate All authors had access to the data used in this study and participated in writing the manuscript. The authors con rm that this manuscript is original and that no portion is under consideration elsewhere in another journal. None of the paper's contents have been previously published. All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
Consent for publication This publication has been read and approved by all authors, as well as by the responsible authorities at the Institute for Cardiovascular Diseases Dedinje, where the work has been carried out. The deidenti ed participant data will be shared on a request basis. Please directly contact the corresponding author to request data sharing. We will share statistical analysis plan, the data base with all patients' data: clinical, echocardiography, radiofrequency ablations characteristics etc. The data will be available immediately after the publication and will be shared on a request basis for anyone. Any type of analysis will be available and we will send the data through e-mail with ne explanations.

Authors' Contributions
Conceptualization: Ruzica Jurcevic is guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Ruzica Jurcevic coordinated study, analyzed the data, wrote the manuscript as rst author and is the corresponding author; Methodology: Ruzica Jurcevic, Lazar Angelkov; Formal analysis and investigation: Ruzica Jurcevic, Velibor Ristic, Dejan Vukajlovic; Writing-original draft preparation: Ruzica Jurcevic;  Receiver operating characteristic (ROC) curves for CHA2DS2-VASc score. blue curve: procedural success compared with procedural failure; red curve: procedural and clinical success compared with procedural and clinical failure