Outcomes of Paroxysmal AF ablation Studies are Affected more by Study Design and Patient Mix than Ablation Technique

Short title: Trial design and outcomes in AF ablation

João Ferreira-Martinsc MRCP PhD, James Howardb MB BChir, Becker Al-khayattb MBBS,

Joseph Shalhoubd MBBS PhD, Afzal Sohaibc MBBS PhD, Matthew Shun-Shinb BM BCh,

Paul G Novaka MD, Rick Leathera MD, Laurence D Sternsa MD, Christopher Lanea MD,

Phang Boon Limb,c MBBS PhD, Prapa Kanagaratnamb,c PhD FRCP, Nicholas S Petersb,c MD

PhD FHRS, Darrel P Francisb,c1 MA MD, Markus B Sikkela,b MBBS PhD

aRoyal Jubilee Hospital, Victoria, Canada V8R 1J8

bNational Heart and Lung Institute, Imperial College London, London W12 0NN, UK

cDepartment of Cardiology, Imperial College Healthcare NHS Trust, Hammersmith Hospital, London W12 0HS, UK

dDepartment of Surgery & Cancer, Imperial College London, London SW7 2AZ, UK

1 Corresponding author

Prof Darrel FrancisInternational Centre for Circulatory Health, National Heart and Lung Institute, Imperial College London59 North Wharf RoadLondon, W2 1LAUKEmail: darrel@drfrancis.orgTel: +447973105394Fax: +442080825109

Total word count: 5024

M.B. Sikkel is supported by a National Institute of Health Research Clinical Lectureship award (#2670).ABSTRACT

1

Objective: We tested whether ablation methodology and study design can explain the

varying outcomes in terms of AF-free survival at 1 year.

Background:

There have been numerous paroxysmal AF ablation trials, which are heterogeneous in

their use of different ablation techniques and study design. A useful approach to

understanding how these factors influence outcome is to dismantle the trials into

individual arms and reconstitute them as a large meta-regression.

Methods: Data was collected from 66 studies (6941 patients). With freedom from AF as

the dependent variable, we performed meta-regression using the individual study arm

as the unit.

Results: Success rates did not change regardless of the technique used to produce

pulmonary vein isolation. Neither were adjunctive lesion sets associated with any

improvement in outcome.

Studies that included more males and fewer hypertensive patients were found more

likely to report better outcomes. ECG method selected to assess outcome also plays an

important role. Outcomes were worse in studies that used regular telemonitoring (by

23%, p<0.001) or in patients who had implantable loop recorders (by 21%, p=0.006),

rather than less thorough periodic Holter monitoring.

Conclusions: Outcomes of AF ablation studies involving pulmonary vein isolation are not

affected by the technologies used to produce PVI. Neither do adjunctive lesion sets

change the outcome. Achieving high success rates in these studies appears to be

dependent more on patient mix and on the thoroughness of AF detection protocols.

2

This should be carefully considered when quoting success rates of AF ablation

procedures which are derived from such studies.

Key words: paroxysmal atrial fibrillation, ablation, pulmonary vein isolation

3

Abbreviations list

AF – atrial fibrillation

AT – atrial tachycardia

CFAE – complex fractionated atrial electrograms

GP – ganglionated plexi

HIFU - high intensity focal ultrasound

ILR – internal loop recorder

LA – left atrium

PAF – paroxysmal atrial fibrillation

PV – pulmonary vein

PVI – pulmonary vein isolation

RF – radiofrequency

WACA – wide area circumferential ablation

4

Introduction

Percutaneous catheter ablation is now well established in international guidelines as an

effective therapy for paroxysmal atrial fibrillation (AF) 1,2. However, there is a wide range

of success rates (20%-100% at 12 months)3,4 and this is frequently attributed to technical

differences in the ablation process5-7.

No analysis has ever systematically addressed the possibility that the disparate design of

trials might also contribute to the disparate results. In part this is because the disparities

are so extensive that conventional trial-by-trial analysis is unable to extract the effects.

In this analysis, we perform an arm-by-arm meta-regression in which the multi-level

heterogeneity becomes a strength rather than a weakness. Treating the individual arms

(with their design and corresponding outcome) as the unit of analysis, we quantify:

Whether adjunctive ablation strategies such as CFAE, GP ablation or additional

lines might be additive to PVI in improving success rate of PAF ablation;

Whether different methods of producing PVI (e.g. cryoablation vs ostial RF vs

wide area RF) give rise to differing success rates;

Whether any of these strategies lead to reduction in RF time or procedure time;

and

Whether studies with different characteristics (e.g. longer vs shorter blanking

periods, different methods of AF detection) would alter success rates.

The last point alludes to the possibility that factors that have nothing to do with

technique have the potential to influence study outcomes. In AF ablation trials, the

“success” of a study is usually quoted as the recurrence-free survival at 6-12 months.

How such recurrence is defined is as variable as the techniques used in these studies

with variable definitions of recurrence (AF vs AF plus atrial tachycardia) 8,9, durations of

blanking period 8,10, and the means by which recurrence is detected (e.g. using periodic

Holter monitoring vs using long-term implantable-loop recorder to detect recurrence)

5

8,11. Whether such methodological details are important in defining outcomes is vital to

put results of ablation studies into context.

Methods

Search strategy and data extraction

Two reviewers (JFM and BA) independently searched the Medline and the Cochrane

Central Register of Randomised Controlled Trials databases using the search terms

“paroxysmal atrial fibrillation ablation”. Abstracts were screened by both reviewers,

who independently extracted data from the full texts. A third author (MS) resolved any

conflicts. Inclusion criteria were: (1) randomized and non-randomized trials published in

English; (2) patient population with paroxysmal atrial fibrillation; (3) at least one

intervention arm including some form of left atrial ablation. Study titles and abstracts

were initially screened. Those not excluded at this stage underwent full text screening.

From each study report, the following parameters were extracted: number of patients in

each intervention arm; characteristics of included patients; procedure characteristics,

including specific methodologies and procedure duration; and blanking period and its

duration. Included and excluded reports are shown in the PRISMA diagram (Figure 1).

Outcome selection

The primary outcome was the percentage of patients free from AF in each study arm. In

most cases the 12-month timepoint was chosen. Where this was not available, the

nearest available timepoint was selected. The 12-month timepoint was chosen as it was

commonly available in almost all studies and was felt to be a clinically relevant

timepoint by which the majority of recurrences would have been expected to occur.

Data analysis and synthesis

Where possible, continuous variables were reported as mean (+/- standard deviation)

and categorical variables as proportions (%). A funnel plot (Supplemental Figure 1)

6

showed that a large number of trials fell outside the predicted standard error margins

but without evidence of asymmetry, suggesting no important publication bias. With

percentage success (freedom from AF) as the dependent variable, we performed meta-

regression using the restricted maximum likelihood method, with study-level

heterogeneity factored using a random-effects model. Using this, we examined the

effects of baseline clinical (age, gender, hypertension, LA diameter), procedural

characteristics (CFAE, GP, lines, WACA, cryoballoon, single-shot RF, laser balloon, force-

sensing, HIFU), non-procedural study characteristics (e.g. anti-arrhythmic use), and

methods of endpoint assessment (blanking period; method of assessment of arrhythmia

recurrence such as telemonitoring or implantable loop recoder [ILR]; definition of

recurrence such as inclusion of atrial tachycardias (AT) and minimum duration of AF/AT).

Statistical analyses were performed using R 12 with the metafor 13 package. Plots were

created using ggplot2 14.

Results

Medline and Cochrane database searches yielded 493 and 349 studies, respectively

(total 842). Seven hundred and thirty-four studies were excluded on initial screening

and 119 studies underwent full text review and 11 further studies relating to references

found in these studies which had not been found by the initial search also underwent

full review. Sixty-six manuscripts met criteria for onward analysis after excluding

duplicate studies, conference abstracts, studies with patients affected by

persistent/permanent AF, studies including further ablations after the index procedure,

follow-up period shorter than 3 months and studies which did not perform LA ablation

(please refer to Figure 1 and Supplemental Table 1 for specific details). With regards to

follow-up, 676 (9.7%) of the patients had follow-up between 3 and 6 months. 514 (7.4%)

had follow-up between 6 and 12 months and 5751 (82.9%) of the patients had follow-up

of at least 12 months. Only study arms which included ablation were considered in the

final analysis. In these 111 arms, a total of 6941 patients were studied, with a follow up

of 20.1 ± 15.4 months (3 to 72 months) and total follow-up 11,631 patient-years.

7

For the reader unfamiliar with metaregression, interpretation of the analyses requires

some explanation. Each of figures 2-6 represents a separate meta-regression analysis in

which the outcome (e.g. AF-free survival) is explained by a variety of variables used to

build the statistical model. The effect of each of these variables is then separated from

each of the others in the model. There is no “control” group since all the data is placed

into a single model but the comparison, for example in the first row of Figure 2, is

effectively between “CFAE” vs “no CFAE” regardless of what other lesion sets are

performed.

Adjunctive strategies to PVI do not enhance success rate in PAF ablation

Among the 3 adjunctive strategies to PVI studied (CFAE, GP ablation and Additional

Ablation Lines), none improved success rate (Figure 2). Use of adjuvant anti-arrhythmic

drugs classes I/III post ablation showed a trend to improved success rate which did not

reach significance (95%CI -0.8 to 23.5%, p=0.07, Figure 2).

The success of PVI is independent of the ablation methodology

There are different methodologies for performing PVI. We compared seven different

methods (WACA, cryoballoon, single-shot RF, laser balloon, force-sensing, HIFU and

robot) in this meta-regression to basic ablation, which was defined as antral PVI using

radiofrequency, i.e. an absence of the other techniques included in the metaregression

(Figure 3). Wide area circumferential ablation (WACA) was not associated with a

significantly different outcome in comparison to antral radiofrequency ablation (95% CI -

3.3 to 12%, p=0.267). Similarly, none of the other 6 methodologies studied (cryoballoon,

single-shot RF, laser balloon, force-sensing, HIFU, robotic navigation) were associated

with a better outcome (Figure 3). Together, the results presented in Figure 2 and 3

indicate that, among the paroxysmal AF techniques, the effectiveness of PVI is not

enhanced by the use of any adjunctive techniques nor by the specific approaches used

to achieve it.

8

Procedure and fluoroscopy times vary by ablation methodology

For procedures of similar efficacy, procedural and fluoroscopy times are relevant

considerations.

WACA has become the most widely used point-by-point method to produce PVI.

Procedure time did not significantly differ from an antral approach (95% CI -6.4 to 4.8,

p=0.784, Figure 4.1, Supplemental Table 3), but it was associated with significantly

reduced fluoroscopy exposure (-2.08min, 95% CI -3.34 to -0.83, p<0.001, Figure 4.2,

Supplemental Table 4). Although statistically significant, such a difference is unlikely to

be clinically significant.

Of the single-shot technologies, only single-shot RF was associated with shorter times,

both procedure time (by 63 mins, CI -71.6 to -55.3 mins, p<0.001, Figure 4.1) and

fluoroscopy time (by 11 mins, CI -12.9 to -9.0 mins, p<0.001, Figure 4.2).

No other techniques were associated with shorter procedure time. Laser balloon

ablation was associated with increased procedure time (Figure 4.1). In addition to WACA

and single-shot RF, robotic navigation also reduced fluoroscopy time (Figure 4.2).

Freedom from AF varies by the characteristics of patients enrolled

Most study arms contained a substantial proportion of patients with hypertension

(mean 40.6% across all study arms). The higher the proportion of patients with

hypertension, the lower the AF-free survival. The presence of hypertension in a patient

reduces the chance of AF-free survival by 24% in that individual (95% CI -48.3 to 0.5%,

p=0.046, Figure 5).

Gender mix also had a significant affect on outcome. Studies including more males had

better outcomes. Extrapolating to the individual patient level, a male patient stands a

9

42.8% better chance at being AF free after 1 year as compared to a female patient post

ablation (95% CI 11.5 to 74.2%, p=0.007, Figure 5).

Neither age nor LA dimensions had a significant impact on outcomes (Figure 5).

Impact of protocols for detecting recurrence

We tested whether success rates were associated with duration of blanking period, the

inclusion of freedom from AT (rather than just AF) in the primary outcome, whether

patients needed to be free from antiarrhythmic drugs to count as successful, the

thoroughness of ECG monitoring methods, and how many seconds of AF were required

to define recurrence.

For freedom from AF at 6 months, the blanking period was found to be a significant

factor with an improvement in freedom from AF of 5.4% per month blanked (95% CI -1.1

to 9.9%, p=0.013, Supplemental Figure 2).

By the 12-month timepoint, however, blanking period is no longer significant in affecting

the outcomes of AF ablation studies (Figure 6). Thoroughness of ECG sampling in the

search for AF recurrences was important throughout the follow up period. By 12

months, trials whose protocols included frequent ECG sampling rather than Holter

monitoring, reported a poorer arrhythmia-free survival, by 23% for telemonitoring (95%

CI 11.4 to 33.2%, p<0.001) and by 21% for ILR (95% CI 6.7 to 33.5%, p=0.006, Figure 6).

Discussion

In terms of technique, multiple strategies targeting ablation of tissue that either triggers

or sustains AF have been used 15. Pulmonary vein isolation (PVI) is the commonest

employed technique aimed at stopping pulmonary vein ectopy from triggering AF 16.

Other aspects of atrial substrate sometimes targeted include complex fractionated atrial

electrograms (CFAE) and ganglionated plexi (GP) 17-19. Additional linear ablation lesions

10

have been proposed to further enhance the success of the procedure 8,20,21, potentially

by preventing left atrial macro-reentrant tachycardias which can degenerate into AF.

Furthermore, PVI itself has evolved as a technique since its first description, although it

is not clear whether these changes have resulted in improved freedom from AF for the

patient 22-24.

This analysis of 111 arms of 66 trials including 6941 patients shows that different

techniques for achieving PVI or adjunctive strategies on top of PVI do not translate into

better or worse outcomes. Secondly, these techniques differ in terms of fluoroscopy

times and procedure times despite outcomes being similar. Finally, patient selection and

the protocol for defining recurrence have a large effect on outcomes.

Additional Lesion sets do not Improve Outcome Compared to PVI Alone

Additional ablation targets such as CFAE 9,25,26, ganglionated plexus ablation and

additional lines 9,15,27 were not associated with improved outcomes (Figure 2). Although

the results of PVI are not perfect, our analysis suggests that adding these alternatives do

not convincingly improve success in the ablation of paroxysmal AF.

This does not mean that these alternative targets must be irrelevant. It is possible that

PVI might, through an accident of anatomy, be coincidentally ablating them too. This is

particularly a possibility for ganglionated plexi, which may well be involved in the

pathogenesis of PAF 28,29 and be disrupted by PVI 30.

Another possible reason for ineffectiveness of other approaches might be that they

enhance arrhythmia by as much as they reduce it, leaving a neutral effect. For example,

both CFAE and lines can slow conduction enough to enable macro-reentrant AT circuits 31,32.

11

Choice of ablation technology affects procedure and fluoroscopy time, but not success

rate

We found no difference between the ablation technologies in their AF-free survival. In

particular, wider area ablation (WACA) 10 does not, across the data as a whole, seem to

be associated with improved success, in comparison with antral ablation, at least in

paroxysmal AF ablation.

Nor were any of the single-shot ablation techniques, i.e. cryoballoon, single-shot RF,

laser balloon and high frequency ultrasound balloon (HIFU), associated with a better

outcome. This is consistent with the FIRE AND ICE trial, published after our analysis was

completed, which showed no difference in success between cryoballoon and WACA 33.

The fact that FIRE AND ICE shows similar results to our analysis also adds some

credibility to the methodology we use here.

Although success rates are similar across the technologies, procedural times were

notably shorter for single-shot RF (by 63 minutes), and longer with CFAE (by 61 minutes)

and laser-balloon ablation (by 60 minutes). Fluoroscopy times were slightly shorter for

WACA, and this probably relates to these studies being done in the modern setting of

electroanatomical mapping whilst some of the ostial RF studies were performed before

its widespread use. In addition, fluoroscopy time was shorter with single-shot RF and

robotic ablation, but longer where adjunctive lesions such as CFAE and lines were

performed.

Since outcomes are similar with all these techniques, the procedure and fluoroscopy

times may be useful in selecting the right technology for AF ablation if aiming for

maximum efficiency with the least exposure to harmful radiation for the patient and

electrophysiology lab staff.

Patients characteristics alter success rates of AF ablation studies

12

It is now accepted that modifiable or non-modifiable risk factors can increase an

individual’s risk of AF development. Accumulating evidence also suggests that aggressive

modification of the modifiable risk factors can also decrease the risk of PAF recurrence34.

Our meta-regression analysis identified arterial hypertension as having a negative

impact on the relative success of paroxysmal AF ablation, an effect possibly due to

ongoing electrical and structural remodeling of the atria and associated electrical

instability. Consistent with this finding, intensive post-ablation blood pressure treatment

has been shown to markedly decrease the risk of AF recurrence34.

Seemingly at odds with the fact that male gender is a recognized predisposing factor for

AF development, our meta-regression showed that studies containing a high proportion

of males had significantly better outcomes. This suggests that men with AF have more to

gain from PVI than women. This corresponds with recent studies which reported a lower

success rate and a higher complication rate among women post catheter AF ablation35,36.

It has also been suggested that women are referred for catheter ablation not only less

frequently but also at later stages than men, despite being more symptomatic, having a

lower quality of life and being less tolerant of anti-arrhythmic drugs36.

The changes in outcome associated with being male (42% improvement in outcome)

and hypertensive (24% reduction in success) are quoted on a per patient basis in Figure

5. Equally relevant when looking at results of studies, is to think about how this would

affect the results of a study based on percentage of patients with these characteristics

recruited. For example, per 20% increase in the proportion of females versus males in a

study, one can expect an 8.6% reduction in AF-free survival. Per 20% reduction in the

proportion of hypertensive patients, we can expect a 4.9% improvement in the final

outcome of a study arm.

13

Complex effect of blanking period on outcomes

Eighty percent of the trials had blanking periods, for the good clinical rationale that

transient post-procedural inflammation37 could elicit early recurrences which were not

reflective of the long-term treated state.

Unlike other parameters tested in this study, the effect of blanking period on outcomes

depends on when you look. By 12 months, the blanking period makes little difference.

However, earlier on in the follow up, the blanking period can make quite a large

difference. With respect to 6-month outcomes, every extra month of blanking period

yielded 5.4% fewer recurrence events (Supplemental Figure 2). Designing a protocol

with a 3-month rather than a 1-month blanking period will raise success rates at 6

months by 10.8%, an effect that will dissipate by 12 months. This is consistent with data

suggesting that blanking period recurrences, are predictive of later recurrences38 such

that ignoring blanking period recurrences, particularly in months 2 and 3, is unlikely to

improve success rates in the long run. So although these later blanking period

recurrences may not predict further recurrence by the time 6 month outcomes are

measured, by 12 months they almost certainly do.

More thorough ECG follow-up reduces success rates

Studies use a variety of approaches to detect and define AF recurrence, ranging from a

single Holter monitor to continuous monitoring with an ILR. Quite rightly, each

individual trial only made comparisons between its arms of comparable design, but

outcomes are often compared between trials without adjusting for, or even mentioning,

differences in recurrence detection protocols.

Our data indicate that designing a study to rely on a Holter monitor rather than an ILR or

telemonitoring improves apparent success rate by around 20%. This is in stark

comparison to the absence of influence of technical factors such as method of

producing PVI or adjunctive lesion sets.

14

Limitations

This is not a randomized trial, but rather a meta-regression treating each arm of

published studies as a unit of analysis. It is effectively an observational study. However,

it provides a solution to the practical problem that, although one might suppose that

these non-procedural aspects may have some impact on outcome, as clinicians we have

no idea what the likely magnitudes of such outcomes will be. This study provides an

understanding of these effect sizes.

In addition, we are limited by the data available to us. As an example, there may be

limited variation in physiological parameters such as mean LA size in this group of PAF

patients, limiting our ability to detect a difference in outcomes in this metaregression. A

further example is that some technologies that we have grouped together may not be

completely homogeneous, such as first vs second generation cryoballoons.

We could only study variables disclosed by the authors of the manuscripts. There may

be other variables that are important but were not disclosed, or even not documented.

All we can tell is the apparent effect size of the non-procedural variables that happen to

have been presented are much larger than the effect size of the choice of procedure.

Conclusion

The key practical finding from this analysis is that when we look at the highly varied

results in outcomes of AF ablation studies, study methodology is a bigger determinant

than any modifications to the technique. Clearly performing a pulmonary vein isolation

is important, but how this is done, and the performance of adjunctive lesion sets are

less important to the final outcome than patient characteristics and protocol for

defining AF recurrence. Adjunctive lesion sets and some methods of producing PVI add

15

to procedure and fluoroscopy time without enhancing efficacy across the study arms

assessed.

Non-procedural aspects show much greater potential to increase observed success rate.

Per 20% increase in the proportion of males versus females in a study, one can expect

an 8.6% improvement in final AF-free survival. Per 20% reduction in the proportion of

hypertensive patients, we can expect a 4.9% improvement in success. Finally, using less

thorough methods of follow-up (e.g. Holter monitoring) can result in an apparent 21-

23% improvement in success compared to more thorough methods such as ILR or

telemonitoring.

Acknowledgements

We would like to acknowledge the BRC, BHF, and ElectroCardioMathsProgramme of the

Imperial Centre for Cardiac Engineering.

Author contributions:

João Ferreira-Martins: Concept/design, Data analysis/interpretation, Drafting article,

Critical revision of article, Approval of article.

James Howard: Concept/design, Data analysis/interpretation, Critical revision of article,

Approval of article, Statistics.

Becker Al-khayatt: Concept/design, Data analysis/interpretation, Drafting article, Critical

revision of article, Approval of article.

Joseph Shalhoub: Data analysis/interpretation, Critical revision of article, Approval of

article.

Afzal Sohaib: Data analysis/interpretation, Critical revision of article, Approval of article.

Matthew Shun-Shin: Data analysis/interpretation, Critical revision of article, Approval of

article.

16

Paul G Novak: Data analysis/interpretation, Critical revision of article, Approval of

article.

Rick Leather:Data analysis/interpretation, Critical revision of article, Approval of article.

Laurence D Sterns: Data analysis/interpretation, Critical revision of article, Approval of

article.

Christopher Lane: Data analysis/interpretation, Critical revision of article, Approval of

article.

Phang Boon Lim: Data analysis/interpretation, Critical revision of article, Approval of

article.

Prapa Kanagaratnam: Data analysis/interpretation, Critical revision of article, Approval

of article.

Nicholas S Peters: Data analysis/interpretation, Critical revision of article, Approval of

article.

Darrel P Francis: Concept/design, Data analysis/interpretation, Drafting article, Critical

revision of article, Statistics Approval of article.

Markus B Sikkel: Concept/design, Data analysis/interpretation, Drafting article, Critical

revision of article, Statistics Approval of article.

References

1. Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. European heart journal. 2016;37(38):2893-2962.

2. January CT, Wann LS, Alpert JS, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130(23):2071-2104.

3. Katritsis DG, Giazitzoglou E, Zografos T, Pokushalov E, Po SS, Camm AJ. Rapid pulmonary vein isolation combined with autonomic ganglia modification: a randomized study. Heart rhythm. 2011;8(5):672-678.

4. Providencia R, Marijon E, Combes S, et al. Higher contact-force values associated with better mid-term outcome of paroxysmal atrial fibrillation ablation using the SmartTouch catheter. Europace : European pacing,

17

arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. 2015;17(1):56-63.

5. Lo LW, Chen SA. Hope or Hype- Does Cryoballoon Simplify Atrial Fibrillation Ablation Technique or Just Another Costly Toy? Circulation journal : official journal of the Japanese Circulation Society. 2016;80(8):1695-1696.

6. Proietti R, Santangeli P, Di Biase L, et al. Comparative effectiveness of wide antral versus ostial pulmonary vein isolation: a systematic review and meta-analysis. Circulation Arrhythmia and electrophysiology. 2014;7(1):39-45.

7. Woods CE, Olgin J. Atrial fibrillation therapy now and in the future: drugs, biologicals, and ablation. Circulation research. 2014;114(9):1532-1546.

8. Cosedis Nielsen J, Johannessen A, Raatikainen P, et al. Radiofrequency ablation as initial therapy in paroxysmal atrial fibrillation. The New England journal of medicine. 2012;367(17):1587-1595.

9. Chen M, Yang B, Chen H, et al. Randomized comparison between pulmonary vein antral isolation versus complex fractionated electrogram ablation for paroxysmal atrial fibrillation. Journal of cardiovascular electrophysiology. 2011;22(9):973-981.

10. Oral H, Scharf C, Chugh A, et al. Catheter ablation for paroxysmal atrial fibrillation: segmental pulmonary vein ostial ablation versus left atrial ablation. Circulation. 2003;108(19):2355-2360.

11. Perez-Castellano N, Fernandez-Cavazos R, Moreno J, et al. The COR trial: a randomized study with continuous rhythm monitoring to compare the efficacy of cryoenergy and radiofrequency for pulmonary vein isolation. Heart rhythm. 2014;11(1):8-14.

12. Team RC. R: A language and environment for statistical computing. 2016; https://www.R-project.org/.

13. Viechtbauer W. Conducting meta-analyses in R with the metafor package. Journal of Statistical Software. 2010;36(3):1-48.

14. Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York; 2009.

15. 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, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design: a report of the Heart Rhythm Society (HRS) Task Force on Catheter and Surgical Ablation of Atrial Fibrillation. Developed in partnership with the European Heart Rhythm Association (EHRA), a registered branch of the European Society of Cardiology (ESC) and the European Cardiac Arrhythmia Society (ECAS); and in collaboration with the American College of Cardiology (ACC), American Heart Association (AHA), the Asia Pacific Heart Rhythm Society (APHRS), and the Society of Thoracic Surgeons (STS). Endorsed by the governing bodies of the American College of Cardiology Foundation, the American Heart Association, the European Cardiac Arrhythmia Society, the European Heart Rhythm Association, the Society of Thoracic Surgeons, the

18

Asia Pacific Heart Rhythm Society, and the Heart Rhythm Society. Heart rhythm. 2012;9(4):632-696 e621.

16. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. The New England journal of medicine. 1998;339(10):659-666.

17. Kabra R, Singh JP. Catheter ablation targeting complex fractionated atrial electrograms for the control of atrial fibrillation. Current opinion in cardiology. 2012;27(1):49-54.

18. Linz D, Ukena C, Mahfoud F, Neuberger HR, Bohm M. Atrial autonomic innervation: a target for interventional antiarrhythmic therapy? Journal of the American College of Cardiology. 2014;63(3):215-224.

19. Corradi D, Callegari S, Gelsomino S, Lorusso R, Macchi E. Morphology and pathophysiology of target anatomical sites for ablation procedures in patients with atrial fibrillation: part II: pulmonary veins, caval veins, ganglionated plexi, and ligament of Marshall. International journal of cardiology. 2013;168(3):1769-1778.

20. Neumann T, Wojcik M, Berkowitsch A, et al. Cryoballoon ablation of paroxysmal atrial fibrillation: 5-year outcome after single procedure and predictors of success. Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. 2013;15(8):1143-1149.

21. Gavin AR, Singleton CB, Bowyer J, McGavigan AD. Pulmonary venous isolation versus additional substrate modification as treatment for paroxysmal atrial fibrillation. Journal of interventional cardiac electrophysiology : an international journal of arrhythmias and pacing. 2012;33(1):101-107.

22. Haissaguerre M, Marcus FI, Fischer B, Clementy J. Radiofrequency catheter ablation in unusual mechanisms of atrial fibrillation: report of three cases. Journal of cardiovascular electrophysiology. 1994;5(9):743-751.

23. Haissaguerre M, Shah DC, Jais P, et al. Electrophysiological breakthroughs from the left atrium to the pulmonary veins. Circulation. 2000;102(20):2463-2465.

24. Jais P, Haissaguerre M, Shah DC, et al. A focal source of atrial fibrillation treated by discrete radiofrequency ablation. Circulation. 1997;95(3):572-576.

25. Di Biase L, Elayi CS, Fahmy TS, et al. Atrial fibrillation ablation strategies for paroxysmal patients: randomized comparison between different techniques. Circulation Arrhythmia and electrophysiology. 2009;2(2):113-119.

26. Deisenhofer I, Estner H, Reents T, et al. Does electrogram guided substrate ablation add to the success of pulmonary vein isolation in patients with paroxysmal atrial fibrillation? A prospective, randomized study. Journal of cardiovascular electrophysiology. 2009;20(5):514-521.

27. Link MS, Haissaguerre M, Natale A. Ablation of Atrial Fibrillation: Patient Selection, Periprocedural Anticoagulation, Techniques, and Preventive Measures After Ablation. Circulation. 2016;134(4):339-352.

19

28. Pokushalov E, Romanov A, Shugayev P, et al. Selective ganglionated plexi ablation for paroxysmal atrial fibrillation. Heart rhythm. 2009;6(9):1257-1264.

29. Pokushalov E, Romanov A, Artyomenko S, et al. Ganglionated plexi ablation directed by high-frequency stimulation and complex fractionated atrial electrograms for paroxysmal atrial fibrillation. Pacing and clinical electrophysiology : PACE. 2012;35(7):776-784.

30. Malcolme-Lawes LC, Lim PB, Wright I, et al. Characterization of the left atrial neural network and its impact on autonomic modification procedures. Circulation Arrhythmia and electrophysiology. 2013;6(3):632-640.

31. Wong KC, Paisey JR, Sopher M, et al. No Benefit of Complex Fractionated Atrial Electrogram Ablation in Addition to Circumferential Pulmonary Vein Ablation and Linear Ablation: Benefit of Complex Ablation Study. Circulation Arrhythmia and electrophysiology. 2015;8(6):1316-1324.

32. Sawhney N, Anousheh R, Chen W, Feld GK. Circumferential pulmonary vein ablation with additional linear ablation results in an increased incidence of left atrial flutter compared with segmental pulmonary vein isolation as an initial approach to ablation of paroxysmal atrial fibrillation. Circulation Arrhythmia and electrophysiology. 2010;3(3):243-248.

33. Kuck KH, Brugada J, Furnkranz A, et al. Cryoballoon or Radiofrequency Ablation for Paroxysmal Atrial Fibrillation. The New England journal of medicine. 2016;374(23):2235-2245.

34. Pathak RK, Middeldorp ME, Lau DH, et al. Aggressive risk factor reduction study for atrial fibrillation and implications for the outcome of ablation: the ARREST-AF cohort study. Journal of the American College of Cardiology. 2014;64(21):2222-2231.

35. Zylla MM, Brachmann J, Lewalter T, et al. Sex-related outcome of atrial fibrillation ablation: Insights from the German Ablation Registry. Heart rhythm. 2016;13(9):1837-1844.

36. Beck Md H, A BCM. Sex Differences In Outcomes Of Ablation Of Atrial Fibrillation. Journal of atrial fibrillation. 2014;6(6):1024.

37. Arya A, Hindricks G, Sommer P, et al. Long-term results and the predictors of outcome of catheter ablation of atrial fibrillation using steerable sheath catheter navigation after single procedure in 674 patients. Europace : European pacing, arrhythmias, and cardiac electrophysiology : journal of the working groups on cardiac pacing, arrhythmias, and cardiac cellular electrophysiology of the European Society of Cardiology. 2010;12(2):173-180.

38. Lellouche N, Jais P, Nault I, et al. Early recurrences after atrial fibrillation ablation: prognostic value and effect of early reablation. Journal of cardiovascular electrophysiology. 2008;19(6):599-605.

20

Figure 1

21

Figure 2

22

Figure 3

23

Figure 4.1

24

Figure 4.2

25

Figure 5

26

Figure 6

27

Figure Legends

Figure 1. Flow diagram of literature search and study selection. N, number of studies.

Figure 2. The effect of adjunctive strategies in addition to PVI on AF-free survival.

None of the adjunctive strategies resulted in significant improvements over the basic

PVI lesion set. Routine use of anti-arrhythmic drugs class I/III post-ablation came closest

to reaching significance p=0.07 with a trend towards greater success. CFAE, complex

fractionated atrial electrograms; GP, ganglionated plexi; PVI, pulmonary vein isolation.

Figure 3. The effect of PVI methodology on arrhythmia-free survival. The success of PVI

is independent of the specific methodologies available to achieve it. HIFU, high-intensity

focal ultrasound; PVI, pulmonary vein isolation; RF, radiofrequency; WACA, wide area

circumferential ablation.

Figure 4.1. The effect of ablation methodology on procedure time. CFAE and laser

balloon increase procedure time, and single-shot RF reduces procedure time. CFAE,

complex fractionated atrial electrograms; GP, ganglionated plexi; HIFU, high-intensity

focal ultrasound; RF, radiofrequency; WACA, wide area circumferential ablation.

Figure 4.2. The effect of ablation methodology on fluoroscopy time. Performing CFAE

or lines increased fluoroscopy time. Use of single-shot RF or robot reduces fluoroscopy

time. CFAE, complex fractionated atrial electrograms; GP, ganglionated plexi; HIFU, high-

intensity focal ultrasound; RF, radiofrequency; WACA, wide area circumferential

ablation.

Figure 5. The effect of patient characteristics on arrhythmia-free survival. Each

hypertensive patient recruited had a 24.4% greater chance of recurrence. Male gender

conferred greater success. Age and LA diameter did not affect success. LA, left atrium.

28

Figure 6. The effect of follow-up protocols on arrhythmia-free survival. By 12 months,

the duration of a blanking period had little impact on AF free survival (unlike results at 6

months – see text). Utilization of telemonitoring and ILR also significantly reduced AF-

free survival. AT included: whether atrial tachycardia, as well as AF recurrences, counted

towards the primary outcome; Drug-free: whether patients had to be off class I/III

agents to be classed recurrence-free; ILR: implantable loop recorder; Seconds of AF:

whether number of seconds of AF required to be deemed a recurrence made a

difference to the outcome.

29

Supplemental material

Supplement Figure 1. Funnel plot: the symmetric configuration of the Funnel plot

suggests no important publication bias of the trials included in the meta-analysis.

30

Supplement Figure 2. Metaregression showing impact of study characteristics

on freedom from AF at 6 months. Unlike at 12 months, blanking period has a

significant impact of 5.4% per month blanked at this timepoint.

31

Supplement Figure 3. AF-free survival by ablation technique and year of

publication in included studies (A – G). RF, radiofrequency; PVI, pulmonary vein

isolation; HIFU, high-intensity focal ultrasound; WACA, wide area circumferential

ablation.

32

Group Number of study arms

Number of patients

Control: RF antral PVI ablation

16 866

RF antral PVI + CFAE 3 261RF antral PVI + GP 2 116

RF antral PVI + additional lines 15 919RF antral PVI + AAD 5 381

Supplemental Table 1. Number of study arms and patients analysed in manuscript

Figure 2.

Group Number of trial arms

Number of patients

Control:RF antral PVI ablation

16 866

WACA 39 2687Cryoablation 7 478

Single shot RF 7 471Laser balloon 3 297Force sensing 3 222

HIFU 1 22Robot 2 52

Supplemental Table 2. Number of study arms and patients analysed in manuscript

Figure 3.

33

Group Number of study arms

Number of patients

Study year

RF antral PVI 10 535 2003 – 2014WACA 29 1973 2008 – 2016CFAE 3 261 2006 – 2011

GP 1 34 2010Lines 11 919 2006 - 2014

Cryoablation 6 442 2008 – 2014Single shot RF 5 277 2010 – 2015Laser balloon 3 297 2009, 2013Force sensing 3 222 2014

HIFU 1 22 2009Robot 2 53 2009, 2010

Supplemental Table 3. Procedure time: number of study arms, number of patients and study publication year.

Group Number of study arms

Number of patients

Study year

RF antral PVI 10 414 2003 – 2014WACA 30 2081 2008 – 2016CFAE 3 261 2006 – 2011

GP 2 116 2010, 2013Lines 10 548 2006 – 2014

Cryoablation 6 442 2008 – 2014Single shot RF 5 277 2010 – 2015Laser balloon 3 297 2009, 2013Force sensing 3 222 2014

HIFU 1 22 2009Robot 2 53 2009, 2010

Supplemental Table 4. Fluoro time: number of study arms, number of patients and study publication year.

34

Search strategy

A Pubmed search using the terms (MESH) "atrial fibrillation" OR "atrial" AND "fibrillation" OR "atrial fibrillation" OR

"paroxysmal" AND "atrial" AND "fibrillation" OR "paroxysmal atrial fibrillation" AND “ablation” was performed with the filter

“Clinical Trials” and with publication dates between 1st January 1985 and 12th June 2015 (date of literature search). The

search terms paroxysmal atrial fibrillation ablation were used on Cochrane Central Register of Controlled Trials with the filter

“Trials” and with publication dates between 1st January 1985 and 3rd June 2015 (date of literature search).

Supplemental Table 3

Study reference Abstract screen Full text screen reason for exclusion

Dagres 2001excluded       RF ablation of accessory pathways

Cosedis Nielsen 2012   Included   Included  

Nam 2012   Included excluded   randomization after PVI

Han 2014excluded       persistent AF

Cosedis Nielsen 2012excluded       duplicate

Chen 2011   Included   Included  Morillo 2014 (RAAFT-2)   Included   Included  

Looi 2013excluded       focus on QoL

Wang 2011excluded       population with early recurrence subject to reablation

Verma 2015excluded       persistent AF

35

Sawhney 2010   Included   Included  

Fiala 2008excluded       mixed population pAF and persistent AF

Gillis 2009excluded       study design

Morillo 2014 (RAAFT-2)excluded       duplicate

Luik 2010excluded       study design Freeze AF

Oral 2004excluded      

focused on subgroup of pts with inducible AF post initial ablation

Pokushalov 2009   Included   Included  Zhao 2013   Included   Included  Providencia 2014   Included   Included  

Providencia 2014excluded       duplicate

Brignole 2002   Includedexcluded   AV node ablation+ pacemaker for AF

Pokushalov 2013excluded        

Andrade 2014   Includedexcluded   substudy of STOP AF trial

Papone 2006   Included   Included  

Calo 2012   Includedexcluded   Right atrium Ganglionated plexi ablation only

Di Biase 2009   Included   Included  Oral 2003   Included   Included  

Bogachev-Prokophiev 2014excluded       surgical ablation

Nuhrich 2014excluded      

focused on subgroup of patients with intraprocedure sustained AF

Stazi 2014excluded       ischaemic preconditioning

Steinwender 2010excluded      

focused on subgroup of patients screened with CT for anatomical specificities

Stabile 2006excluded       mixed population pAF and persistent AF

Chen 2011 exclude       duplicate

36

d

Duncan 2012excluded       no follow-up of AF recurrence

Brignole 1999   Includedexcluded   AV node ablation

Caponi 2010excluded       mixed population pAF and persistent AF

Bogachev-Prokophiev 2012excluded       conference abstract

Katritsis 2013   Included   Included  

Shim 2014excluded       conference abstract

Vassilikos 2011excluded       no follow-up of AF recurrence

Pokushalov 2013excluded       focused on subgroup of pts with failed AF ablation

Brignole 2003excluded       review

Zhao 2013excluded       duplicate

Kojodjojo 2010excluded       mixed population pAF and persistent AF

Corrado 2010excluded       mixed population pAF and persistent AF

Linhart 2009excluded       case control study

Packer 2013excluded       mixed population pAF and persistent AF

Calo 2006excluded       mixed population pAF and persistent AF

Gu 2011excluded      

subpopulation of T2DM treated or not with pioglitazone

Wang 2011excluded       duplicate

Nori 2009excluded       mixed population pAF and persistent AF

Gaita 2008excluded       mixed population pAF and persistent AF

Lutomsky 2008 exclude       no follow-up of AF recurrence

37

dMcCready 2014   Included   Included  Wilber 2010   Included   Included  

Chin 2014excluded       focused on LA size measurement post ablation

Bogachev-Prokophiev 2014excluded       duplicate (Bogachev-Prokophiev)

Baran 2013excluded       echo for LA thrombus

Deisenhofer 2009   Included   Included  

Duncan 2012excluded        

Kim 2010   Included   Included  

Pokushalov excluded        

Gavin 2012   Included   Included  Bulava 2010   Included   Included  

Lemke 2003excluded       mixed population pAF and persistent AF

Gilis 2000   Includedexcluded   AV node ablation

Pokushalov 2013excluded      

mixed population pAF and persistent AF/conference abstract

Baman 2009excluded       mixed population pAF and persistent AF

Beukema 2012   Included   Included  

Tang 2008   Includedexcluded  

retrospective study assessing the eficacy of PVI in trigered PAF

Padeletti 2003   Includedexcluded   Right atrial ablation +/- right atrial appendage pacing

Reynolds 2010excluded       focus on QoL

Sebag 2013excluded       persistent AF

Koch 2012 MACPAF studyexcluded       no follow-up of AF recurrence

Hocini 2005excluded       no follow-up of AF recurrence

38

Calo 2012excluded       duplicate

Liu 2006   Includedexcluded  

Stepwise left atrial linear ablation tailored by inducibilty of AF after systemic SPVI

Buer 2006excluded        

Haeusler 2010excluded        

Mallow 2013excluded        

Rizzo 2012excluded        

Marshall 1999   Includedexcluded   AV node ablation

Mun 2012   Included   Included  

Bassiounyexcluded       persistent AF/conference abstract

Roux 2009excluded       only 6 weeks follow-up

Tamborero 2009   Includedexcluded  

mixed population (paroxysmal/persistent/longstanding)

Stabile 2001excluded       mixed population pAF and chronic AF

Chierchia 2011excluded       no follow-up of AF recurrence

Brignole 1997   Includedexcluded   AF inducibility before/after PVI and outcome

Adlbrecht 2013   Includedexcluded   paper not available

Solheim 2012excluded       mixed population pAF and persistent AF

Kriatselis (abst) 2012excluded       conference abstract

Richter 2011excluded      

patient stratification according to healing biomarkers post PVI

Luria 2008excluded       atrial flutter ablation

to ACexcluded       Computed tomography vs TOE

39

Della Bella 2009excluded       mixed population pAF and persistent AF

DiBiase 2009excluded       study of oesophageal fistulas

Cheema 2006 excluded       mixed population pAF and persistent AF

Pokushalov 2013excluded       duplicate

Wang 2007   Includedexcluded   paper not available

Verma 2010excluded       mixed population pAF and persistent AF

Kriatselis 2014   Included   Included  

Maly 2008excluded       paper not available

Mantovan 2013excluded       STAR AF substudy

Koch 2012 MACPAF studyexcluded       duplicate

Koch 2011 MACPAF study abstract

excluded       conference abstract

Crawford 2008excluded       focused on subgroup of pts with failed AF ablation

Wieczorek 2010   Includedexcluded   uncontrolled, limited follow-up data

Lu 2014 excluded       conference abstract

Budera 2012excluded       mixed population pAF and persistent AF

Romanov 2011excluded       conference abstract

Nielson 2011 MANTRA-PAF Study

excluded       conference abstract

Leong-Sit 2011   Includedexcluded  

included patients with previous AF ablation; details aof current AF ablation mnot available

Pappone 2011 APAF study   Includedexcluded   extension of study Papone 2006

Pokushalov 2010excluded       conference abstract

40

Nielsen 2012excluded       conference abstract

Zhao 2012excluded       conference abstract

Theis 2014excluded       conference abstract

Schade 2013excluded      

focused on subgroup of patients with failed AF ablation

Arbelo 2014   Included   Included  

Kim 2012excluded       conference abstract

Lu 2013excluded       conference abstract

DiBiase 2014excluded      

no follow-up of AF recurrence/ only bleeding/thromboembolic complications

Pokushalov 2011excluded       focused on subgroup of pts with failed AF ablation

Bassiouny 2011excluded       conference abstract

Richmond 2008excluded       mixed population pAF and persistent AF

Caponi 2010excluded       mixed population pAF and persistent AF

Kang 2014excluded       conference abstract

DeRuvo 2012excluded       conference abstract

Rostock 2006excluded       mixed population pAF and persistent AF

Proclemer 1999   Includedexcluded   AV node ablation

Bencsik 2009excluded       mixed population pAF and persistent AF

Pokushalov 2012excluded       surgical ablation

Verma 2014excluded       persistent AF

Lan 2009   Included   Included  Nedios 2011 exclude       mixed population pAF and persistent AF

41

d

Sciarra 2013excluded       conference abstract

Jais 2008 (A4 study)   Included   Included  

Turco 2007excluded       conference abstract

Lambiase 2013excluded       conference abstract

Fichtner 2014excluded       conference abstract

Sra 2007excluded       mixed population pAF and persistent AF

Park 2014excluded       conference abstract

Tuohy 2014excluded       conference abstract

Hwang 2009   Included   Included  

Jons 2009excluded       study design

Pokushalov 2009excluded       conference abstract

Verma 2007excluded       mixed population pAF and persistent AF

Sairaku 2012excluded       retrospective study

Heart Rhythm Congress, 2011.excluded       Europace proceedings

Letsas 2014excluded       mixed population pAF and persistent AF

Duncan 2010excluded       conference abstract

Rilling 2013excluded       mixed population pAF and persistent AF

Lim 2013excluded       conference abstract/persistent AF

Romanov 2012excluded       conference abstract

Crawford 2010excluded      

study focused on outcomes of patient subgroups based on AF non-inducibility

42

Tse 2005excluded       focused on platelet activation

Tsao 2005excluded      

patient population who had MRI before and after ablation

chen 1999   Includedexcluded   electrophysiologic study of PVs

Podd 2012excluded       conference abstract

Stabile 2003excluded       mixed population pAF and persistent AF

Martinek 2009excluded       oesophageal damge post ablation

Khaykin 2009excluded       mixed population pAF and persistent AF

Nilsson 2006excluded       mixed population pAF and persistent AF

Kaba 2014excluded       summary of RAAFT-2 trial

Wang 2008   Included   Included  

Fiala 2008excluded       duplicate

Kautzner 2009   Included   Included  

Walfridsson 2015excluded       MANTRA-PAF substudy focusing on QoL

Walfridsson 2014excluded       duplicate

Suleiman 2012excluded       mixed population pAF and persistent AF

Stazi 2013excluded       conference abstract

Kamalvand 1997excluded       atrial tacgyarrythmias and pacing

Malmborg 2013excluded       conference abstract

Long 2006excluded       duplicate

Lin 2001excluded       no follow-up

Shimano 2008 exclude       mixed population pAF and persistent AF

43

dDeftereos 2014   Included   Included  

Wang 2014   Includedexcluded   paper not available

Oral 2004   Includedexcluded  

focused on asymptomatic AF recurrences ~2years after ablation procedure

Biase 2014excluded       mixed population pAF and persistent AF

Romavov 2011excluded       conference abstract

Finlay 2012excluded       mixed population pAF and persistent AF

Forleo 2009excluded       mixed population pAF and persistent AF

Wieczorek 2013excluded       focused on subgroup of pts with failed AF ablation

Koyama 2010   Included   Included  

Martinek 2012excluded       no follow-up

Lim 2012excluded       mixed population pAF and persistent AF

Hunter 2013excluded       no follow-up

Piorkowski 2011excluded       mixed population pAF and persistent AF

Knecht 2010excluded       mixed population pAF and persistent AF

Rillig 2013excluded       study design

McLellan 2013excluded       conference abstract

Nuehrich 2013excluded       conference abstract

Nolker 2012excluded       mixed atrial arrhythmias

Pak 2008excluded      

focused on subgroup of PAF patients with unilateral arrhythmogenic PVs

Wu 2008excluded       not in english

44

Gillis 1999   Includedexcluded  

focused on atrial pacing before ablation - no ablation data on paper

Nilsson 2013excluded       conference abstract

Tang 2009excluded      

focused on patients with previous AF ablation and AF recurrence based on OSA risk

Lin 2012   Included   Included  

Pokushalov 2013excluded       focused on subgroup of pts with failed AF ablation

Di Biase 2013excluded       conference abstract

Sheikh 2006   Included   Included  

Kim 2014excluded       conference abstract

Khaykin 2009excluded       cost comparison study

Gillis 2000excluded       duplicate

De Potter 2010excluded       case-control/mixed paroxysmal/persistent AF

Willems 2006excluded       persistent AF

Yamaji 2013excluded       conference abstract

Schmidt 2009excluded       ablation in pts with therapeutic INR

Liu 2005excluded       not in english

Liu 2005excluded       mixed population pAF and persistent AF

Lickfett 2013excluded       atrial flutter

Doi 2013excluded       no follow-up

Pontoppidan 2009excluded       mixed population pAF and persistent AF

Neuzil 2013excluded       no follow-up

Brunelli 2011 exclude       mixed population pAF and persistent AF

45

d

Hayashi 2014excluded       randomization after PVI

Katritsis 2011   Included   Included  

Atienza (2013) RADAR-AF trialexcluded       conference abstract

Andrade 2012 STAR AF trialexcluded       mixed population pAF and persistent AF

Mantovan 2013 (STAR AF substudy)

excluded       mixed population pAF and persistent AF

Raatikainen 2013 MANTRA-PAFexcluded       conference abstract

Tsao 2010excluded       no follow-up

Sheng 2013excluded       systematic review/conference abstract

Walfridsson 2013excluded       conference abstract

Duncan 2010excluded       conference abstract

Kettering 2008excluded       study on ablation complications (oesophageal fistula)

Suleiman 2012excluded       duplicate

Shu 2014excluded       conference abstract

Tada 2002excluded       no follow-up

Khan 2008excluded       no follow-up

Fichtner 2013   Included   Included  

Pokushalov 2014excluded       conference abstract

Gutleben 2013excluded       conference abstract

Pokushalov 2014excluded       conference abstract

Steven 2013   Included   Included  

46

Duff 2003excluded       permanent AF

Steven 2010   Included   Included  

Mont 2014 (SARA study)excluded       persistent AF

Ahmed 2013excluded       study design

Kowal 2011excluded       conference abstract

Oral 2003excluded       review

Duncan 2010excluded       conference abstract

Knecht 2008   Included   Included  

Lin 2012excluded        

Di Biase 2011   Included   Included  

Mohanty 2013excluded       conference abstract

Lee 2000excluded       mixed population pAF and chronic AF

Dixit 2008excluded       mixed population pAF and chronic AF

Fichtner 2011excluded       conference abstract

Fitts 1998excluded       study design

Schmidt 2010   Included   Included  

Gillis 2003excluded       AV node ablation

Nalliah 2013excluded       conference abstract

Podd 2012excluded       conference abstract

Arentz 2007excluded       mixed population pAF and persistent AF

Haeusler 2013excluded       focused on ischaemic brain lesions post ablation

47

Bittner 2011excluded       mixed population pAF and persistent AF

Duncan 2011excluded       conference abstract

Oral 2008excluded       inducibility of AF by isoproterenol

Bailin 2001excluded       no ablation

Dixit 2006excluded       mixed population pAF and persistent AF

Miyanaga 2009excluded       no follow-up of AF recurrence

Dong 2009excluded       no follow-up of AF recurrence

Tamborero 2010excluded       mixed population pAF and persistent AF

Kriatselis 2014excluded       conference abstract

Liu 2014excluded       conference abstract

Di Biase 2013excluded       conference abstract

Shurrab 2013excluded       conference abstract

Fitts 2000   Includedexcluded   AV node ablation

Steven 2013excluded       conference abstract

Kumagai 2005 Included  excluded   comparison of mapping strategies

Pratola 2011excluded       mixed population pAF and persistent AF

Giannopoulos 2015excluded       conference abstract

Takigawa 2013excluded       focused on effect of non-isolation of PV carina

Fassini 2005excluded       mixed population pAF and persistent AF

Atarashi 2007 exclude       mixed population pAF and persistent AF

48

d

Epstein 2002excluded       no follow-up

Pokushalov 2012excluded       conference abstract

Pokushalov 2012excluded       duplicate

Nalliah 2014excluded       conference abstract

Brunelli 2013excluded       conference abstract

Ko 2013excluded       conference abstract

Pontone 2014excluded       conference abstract

Pontone 2014excluded       duplicate

Al-Azawy 2013excluded       conference abstract

Stavrakis 2013excluded       conference abstract

Pokushalov 2012excluded       mixed population pAF and persistent AF

Bertaglia 2013   Included   Included  

Katritsis 2014   Includedexcluded   paper not available

Di Biase 2013excluded       conference abstract

Bänsch 2013excluded       mixed population pAF and persistent AF

Liakishev 2013excluded       not in english

Deftereos 2012excluded       paper not available

Simpson 2001excluded       mixed population pAF and chronic AF

Naccarelli 2014excluded       design study

Pokushalov 2015 exclude       conference abstract

49

d

Drewirz 2007excluded       not in english

Nalliah 2014excluded       conference abstract

Nalliah 2014excluded       duplicate

Nalliah 2014excluded       duplicate

Scaglione 2012excluded       focused on cerebral lesions post ablation

Tondo 2005excluded       mixed population pAF and persistent AF

Metzner 2012excluded       no follow-up

Podd 2012excluded       conference abstract

van Breugel 2014excluded       no follow-up of AF recurrence

Yamane 2007excluded       mixed population pAF and persistent AF

Yamada 2009   Includedexcluded   paper not available

Katritsis 2004   Includedexcluded   Multiple ablations

Stabile 2014excluded       no follow-up of AF recurrence

Chierchia 2012excluded       no follow-up of AF recurrence

Jaïs 2002excluded       no ablation

Bertaglia 2013   Includedexcluded   paper not available

Lampe 2012excluded      

effects of AV ablation + pacing vs DDD PPM for CHB in HF progression

Aras 2013excluded       conference abstract

Fitts 2000excluded      

focused on rate of atrial tachyarrhythmia detection by PPM

50

Di Biase 2012excluded       conference abstract

Waldo 1999excluded       dicussion paper

Tse 2001   Includedexcluded   AV node ablation

Knecht 2010excluded       conference abstract

Perez-Castellano 2014   Included   Included  

Jiang 2009excluded       no follow-up of AF recurrence

Lau 1995excluded       mixed atrial tachyaarhythmias (SVTs)

Stavrakis 2014excluded       conference abstract

Marshall 1999excluded       overlapp with paper 89

Gao 2007excluded       not in english

Gordon 2014excluded       paper not available

Ullah 2014excluded       conference abstract

Macle 2012excluded       study design

Kimura 2014excluded       mixed population pAF and persistent AF

Herrera 2012excluded       no follow-up

Bittkau 2012excluded       conference abstract

Mulder 2013   Included   Included  

Dorian 1996excluded       no ablation

Chilukuri 2011excluded       mixed population pAF and persistent AF

Herrera 2012excluded       duplicate

Derval 2010 exclude       conference abstract

51

d

Haeusler 2011excluded       no ablation

Deftereos 2012excluded       review

Lyan 2013excluded       conference abstract

Kay 1989excluded       mixed population pAF and persistent AF

RAAFT-2excluded       erratum

Zhang 2007excluded       not in english

Bauer 2006excluded       no follow-up of AF recurrence

Navistar® 2004excluded       no abstract

Manolis 1998excluded       no abstract

Fiala 2008   Included   Included  

Verma 2015excluded       duplicate

Stavrakis 2015excluded       no follow-up for AF recurrence

Scherr 2015excluded       persistent AF

Atienza 2014excluded       mixed population pAF and persistent AF

Giannopoulos 2014excluded      

subpopulation of hypertensive pts undergoing AF ablation +/- minoxidil

Rolf 2014excluded       mixed population pAF and persistent AF

Straube 2014excluded       maximum 7 days follow-up

Zellerhoff 2014   Included   Included  

Nuhrich 2014excluded       duplicate

Miller 2014excluded       mixed population pAF and persistent AF

52

Wutzler 2014excluded       no follow-up for AF recurrence

Uhm 2014excluded       no follow-up for AF recurrence

Natale 2014   Included   Included  

McCready 2014excluded       duplicate

Providencia 2014excluded       mixed population pAF and persistent AF

Wang 2014excluded       mixed population pAF and persistent AF

Oza 2014excluded       no follow-up for AF recurrence

Kim 2014excluded       no follow-up for AF recurrence

Kimura 2014excluded       mixed population pAF and persistent AF

Pokushalov 2014excluded       subpopulation of hypertensive pts

Adachi 2014excluded       no ablation

DiBiase 2014excluded       duplicate

Arbelo 2014excluded       mixed population pAF and persistent AF

De Greef 2014excluded       mixed population pAF and persistent AF

Bisbal 2014excluded       mixed population pAF and persistent AF

Kimura 2014excluded       duplicate

Park 2014excluded       mixed population pAF and persistent AF

De Ville 2014excluded       no follow-up for AF recurrence

Efremidis 2014excluded       focused on QoL

Manganiello 2014excluded       mixed population pAF and persistent AF

53

Yorgun 2014excluded       mixed population pAF and persistent AF

Schmidt 2014excluded       mixed population pAF and persistent AF

Krul 2014excluded       patient population with previously failed PVI

Sairaku 2014excluded       no follow-up for AF recurrence/focus on INR

Kriatselis 2014excluded       duplicate

Song 2014excluded       non-PAF

Kawakami 2014excluded       no follow-up of AF recurrence

kaitani 2014   Includedexcluded  

role of ATP in late re-conduction in pts undergoing second AF ablation

Deftereos 2014excluded       duplicate

Seitz 2014excluded       mixed population pAF and persistent AF

Lakkireddy 2014excluded       focused on bleeding complications

Stabile 2014excluded       duplicate

Takigawa 2014excluded       impact of haemodysis in AF ablation

van Breugel 2014excluded       duplicate

Morillo 2014 (RAAFT-2)excluded       duplicate

Marrouche 2014excluded       mixed population pAF and persistent AF

Andrade 2014excluded       duplicate

Verma 2014excluded       duplicate

Mont 2014 (SARA study)excluded       duplicate

Bogachev-Prokophiev 2014 exclude       duplicate

54

d

Jang 2014excluded       chronic AF

Han 2014excluded       duplicate

Stazi 2014excluded       duplicate

Wang 2014excluded       duplicate

Ejima 2014   Includedexcluded   atrial remodeling and AF recurrence

Takigawa 2014excluded       mixed population pAF and persistent AF

Kirchhof 2014excluded       mixed population pAF and persistent AF

De Maat 2014excluded       mixed population pAF and persistent AF

Perez-Castellano 2014excluded       duplicate

Schmidt 2014excluded       no follow-up for AF recurrence

Loghin 2014   Included   Included  

Katritsis 2013excluded       duplicate

Pokushalov 2013excluded       duplicate

Aytemir 2013excluded       mixed population pAF and persistent AF

Miyazaki 2013excluded       no follow-up for AF recurrence

Malmborg 2013excluded       mixed population pAF and persistent AF

Linhart 2013excluded       mixed population pAF and persistent AF

Ichiki 2013excluded      

focus on cerebral microthromboembolism after catheter AF ablation

Brunelli 2013excluded       no follow-up for AF recurrence

Efremidis 2013 exclude       duplicate

55

d

Brooks 2013   Includedexcluded   mixed population pAF and persistent AF

Sebag 2013excluded       duplicate

Mantovan 2013excluded       duplicate

Verma 2013excluded      

focus on asymptomatic cerebral embolism after catheter AF ablation

Hong 2013excluded       focus on left atrial remodeling in lone AF

De Bortoli 2013excluded       non-PAF

Kobza 2013excluded       mixed population pAF and persistent AF

Haines 2013excluded      

retrospective study focused on dabigatran vs warfarin complications post AF ablation

Reddy 2013excluded      

mixed population pAF and persistent AF/pts with GORD/IBS vs those without

Pokushalov 2013excluded       duplicate

Al-Khatib 2013excluded      

mixed population pAF and persistent AF/focus on apixaban

Sohns 2013excluded       mixed population pAF and persistent AF

Neumann 2013   Included   Included  

Malcolme-Lawes 2013   Includedexcluded   cardiac MRI

Doi 2013excluded       duplicate

Ferrero-de Loma-Osorio 2013excluded       mixed population pAF and persistent AF

Hussein 2013excluded      

look at Spontaneous dissociated firing from the pulmonary veins during ablation

Steven 2013excluded       duplicate

Fichtner 2013excluded       duplicate

Dukkipati 2013   Included   Included  

56

Chun 2013excluded       pts grouped by age. Mixed PAF and persistent

Ahmed 2013excluded       duplicate

Packer 2013excluded       duplicate

Adlbrecht 2013excluded       duplicate

Neuzil 2013excluded       duplicate

Zhao 2013excluded       duplicate

Schernthaner 2013excluded       mixed population pAF and persistent AF

Hunter 2013excluded       duplicate

Schade 2013excluded       duplicate

Pokushalov 2013excluded       duplicate

Bertaglia 2013excluded       duplicate

Mardigyan 2013excluded       survey

Narayan 2013excluded       mixed population pAF and persistent AF

Blanche 2013excluded       mixed population pAF and persistent AF

Bänsch 2013excluded       duplicate

Wieczorek 2013excluded       no follow-up for AF recurrence

Lickfett 2013excluded       duplicate

Uchiyama 2013   Included   Included  

Wójcik 2013excluded       mixed population pAF and persistent AF

Wójcik 2013excluded      

mixed population pAF and persistent AF (not the same as 109)

57

Lo 2013excluded       mixed population AF and paroxysmal SVT

Wi 2013excluded       mixed population pAF and persistent AF

Takigawa 2013excluded       duplicate

Rilling 2013excluded       duplicate

Metzner 2013   Included   Included  

Haeusler 2013excluded       duplicate

Wieczorek 2013excluded       duplicate

Derejko 2013excluded       mixed population pAF and persistent AF/HOCM pts

Wang 2013   Included   Included  

Scharf 2012excluded       mixed population pAF and persistent AF

Chierchia 2012excluded       duplicate

Nolker 2012excluded       duplicate

Wang 2012excluded       mixed population pAF and persistent AF

Shivkumar 2012excluded       mixed population pAF and persistent AF

Sairaku 2012excluded       duplicate

Andrade 2012 STAR AF trialexcluded       duplicate

Budera 2012excluded       duplicate

Reddy 2012   Includedexcluded   some patients had more than 1 ablation

Lin 2012excluded       duplicate

Martinek 2012excluded       duplicate

Deftereos 2012 exclude       duplicate

58

d

Cosedis Nielsen 2012excluded       duplicate

Lim 2012excluded       duplicate

Lampe 2012excluded       duplicate

Koch 2012 MACPAF studyexcluded       duplicate

Pokushalov 2012excluded       focused on pts with resistant hypertension

Narayan 2012excluded       mixed population pAF and persistent AF

Arbelo 2012excluded      

mixed population pAF and lone AF/no follow-up AF recurrence

Scaglione 2012excluded       duplicate

Beukema 2012excluded       duplicate

von Bary 2012   Includedexcluded  

does not distinguish outcomes od 2 different ablation techniques. Results based on LA dimensions

Solheim 2012excluded       duplicate

Liu 2012   Includedexcluded   focus on inducibility of AF after ablation

Pozzoli 2012excluded       3 weeks follow-up only

Pison 2012excluded       survey

Chierchia 2012   Included   Included  

Rivard 2012excluded       persistent AF

Pokushalov 2012   Included   Included  

Dukkipati 2012   Includedexcluded   2 ablations

Santini 2012excluded       mixed population pAF and persistent AF

Duncan 2012excluded       duplicate

59

Kasirajan 2012excluded       mixed population pAF and persistent AF

Miller 2012excluded       mixed population pAF and persistent AF

Bayrak 2012excluded       mixed population pAF and persistent AF

Cheung 2012excluded       mixed population pAF and persistent AF

Berkowitsch 2012excluded       mixed population pAF and persistent AF

Golden 2012excluded       mixed population pAF and persistent AF

Metzner 2012excluded       duplicate

Lakkireddy 2012excluded      

mixed population pAF and persistent AF/focus in anticoagulation

Erdei 2012   Included   Included  

Macle 2012excluded       duplicate

Pokushalov 2012excluded       duplicate

Mun 2012excluded       duplicate

Finlay 2012excluded       duplicate

Calo 2012excluded       duplicate

Herrera 2012excluded       duplicate

Suleiman 2012excluded       duplicate

Mulder 2012   Included   Included  

Tang 2012   Includedexcluded   paper not available

Hunter 2012excluded       mixed population pAF and persistent AF

Gavin 2012excluded       duplicate

Pappone 2011 APAF study exclude       duplicate

60

d

Pokushalov 2011excluded       duplicate

Nedios 2011excluded       duplicate

Chilukuri 2011excluded       duplicate

Meinertz 2011excluded       AF management in primary care in Germany

Gu 2011excluded       duplicate

Bittner 2011excluded       duplicate

to ACexcluded       duplicate

Chen 2011excluded       duplicate

Narducci 2011excluded      

mixed population pAF and persistent AF/no follow-up/focus on tissue inflammation

Wang 2011excluded       duplicate

Bonnemeier 2011excluded       mixed population pAF and persistent AF

Katritsis 2011excluded       duplicate

Kidouchi 2011excluded       multiple atrialarrhythmias

Piorkowski 2011excluded       duplicate

Pratola 2011excluded       duplicate

Haeusler 2011excluded       duplicate

Di Biase 2011excluded       duplicate

Chierchia 2011excluded       duplicate

Spertus 2011excluded       QoL questionnaire

61

Leong-Sit 2011excluded       duplicate

Nagashima 2011excluded       mixed population pAF and persistent AF

Chao 2011   Includedexcluded  

focus on renal function and recurrence of AF after ablation

Kettering 2011excluded       persistent AF

Osmancik 2011excluded       mixed population pAF and persistent AF

Reynolds 2010excluded       duplicate

Reddy 2010excluded       no follow-up/acute safety study

Schmidt 2010excluded       duplicate

Gaita 2010excluded       mixed population pAF and persistent AF

Koyama 2010excluded       duplicate

Patel 2010excluded       mixed population pAF and persistent AF

Edgerton 2010excluded       non-randomized

Kojodjojo 2010excluded       duplicate

Kim 2010excluded       duplicate

Bulava 2010excluded       duplicate

Park 2010excluded       mixed population pAF and persistent AF

Caponi 2010excluded       duplicate

Haeusler 2010excluded       duplicate

Di Biase 2010excluded       mixed population pAF and persistent AF

Veasey 2010 exclude       mixed population pAF and persistent AF

62

d

Sawhney 2010excluded       duplicate

Yamaguchi 2010excluded       mixed population pAF and persistent AF

Verma 2010excluded       duplicate

Tamborero 2010excluded       duplicate

Luik 2010excluded       duplicate

Beukema 2010excluded       mixed population pAF and persistent AF

Tuan 2010excluded       mixed population pAF and persistent AF

Knecht 2010excluded       duplicate

Wieczorek 2010excluded       duplicate

Nam 2010excluded       mixed population pAF and persistent AF

Kirch 2010excluded       QoL study/no ablation

Di Donna 2010excluded       subpopulation of pts with HOCM

Pokushalov 2010excluded       persistent AF

Rillig 2010excluded       mixed population pAF and persistent AF

Chierchia 2010excluded       no follow-up/study to assess effusion post ablation

Tsao 2010excluded       duplicate

Neumann 2010excluded       mixed population pAF and chronic AF

Steinwender 2010excluded       duplicate

Bertaglia 2010excluded       mixed population pAF and persistent AF

63

Wilber 2010excluded       duplicate

Schrickel 2010excluded       no follow-up/study on acute silent cerebral embolism

De Potter 2010excluded       duplicate

Patel 2010excluded      

mixed population pAF and persistent AF/females only study on complications

Steven 2010excluded       duplicate

Corrado 2010excluded       duplicate

Crawford 2010excluded       duplicate

Wieczorek 2010   Included   Included  

Chierchia 2009excluded       duplicate

Yamada 2009excluded       duplicate

Klinkenberg 2009excluded       no follow-up/effects of adenosin after ablation

Linhart 2009excluded       duplicate

Jiang 2009excluded       mixed population pAF and persistent AF

Baman 2009excluded       duplicate

Cagli 2009excluded       duplicate

Shamiss 2009excluded       duplicate

Jensen-Urstad 2009excluded       AF ablation in WPW syndrome patients

Nori 2009excluded       mixed population pAF and persistent AF

Hwang 2009excluded       atrial tachycardia

Khaykin 2009excluded       duplicate

64

Schmidt 2009   Included   Included  

Abecasis 2009excluded       mixed population pAF and persistent AF

Bhargava 2009excluded       mixed population pAF and persistent AF

Joshi 2009excluded       mixed population pAF and persistent AF

Fredersdorf 2009excluded       mixed population pAF and persistent AF

Ninomiya 2009excluded      

no follow-up/acute effect of ATP in reconduction post ablation

Roux 2009excluded       duplicate

Pokushalov 2009excluded       duplicate

Hof 2009excluded       mixed population pAF and persistent AF

Meissner 2009excluded       mixed population pAF and persistent AF

Han 2009excluded       mixed population pAF and persistent AF

Chun 2009excluded       no follow-up/effect of rapid pacing of RV

Bencsik 2009excluded       duplicate

Pratola 2009excluded       no follow-up

Schmidt 2009excluded       duplicate

Reddy 2009   Included   Included  

Jons 2009excluded       duplicate

Chilukuri 2009excluded      

questionnaire on obstructive sleep apnoea and outcomes of AF ablation

Martinek 2009excluded       duplicate

Edgerton 2009excluded       mixed population pAF and persistent AF

Sohara 2009 exclude       mixed population pAF and persistent AF

65

d

Pontoppidan 2009excluded       duplicate

Yokoyama 2009excluded       no follow-up

Van belle 2009excluded       no follow-up

Kettering 2009excluded       mixed population pAF and persistent AF

Miyanaga 2009excluded       duplicate

Deisenhofer 2009excluded       duplicate

Schmidt 2009excluded       mixed population pAF and persistent AF

Di Biase 2009excluded       duplicate

DiBiase 2009excluded       duplicate

Yoshida 2009   Includedexcluded   multiple ablations

Della Bella 2009excluded       duplicate

Park 2009excluded       mixed population pAF and persistent AF

Dong 2009excluded       duplicate

Laurent 2009excluded       ablation of atrial flutter

Beyer 2009excluded       mixed population pAF and persistent AF

Kautzner 2009excluded       duplicate

Tamborero 2009excluded       duplicate

Kumagai 2009excluded       mixed population pAF and persistent AF

Kriatselis 2009excluded       mixed population pAF and persistent AF

66

Tang 2009excluded       duplicate

Khaykin 2009excluded       duplicate

Forleo 2009excluded       duplicate

Jais 2008 (A4 study)excluded       duplicate

Knecht 2008excluded       duplicate

Malmborg 2008excluded       mixed population pAF and persistent AF

Van Belle 2008   Included   Included  

Luria 2008excluded       duplicate

Gaita 2008excluded       duplicate

Yoshida 2008excluded       mixed population pAF and persistent AF

Muller 2008excluded       mixed population pAF and chronic AF

Piorkowski 2008excluded       mixed population pAF and persistent AF

Shimano 2008excluded       duplicate

Satomi 2008   Included   Included  

Richmond 2008excluded       duplicate

Corrado 2008excluded       mixed population pAF and persistent AF

Fiala 2008excluded       duplicate

Neumann 2008excluded       mixed population pAF and persistent AF

Khadjooi 2008excluded       effects of CRT in AF and SR

Jongnarangsin 2008excluded       mixed population pAF and chronic AF

Shah 2008 exclude       mixed population pAF and persistent AF

67

d

Wu 2008excluded       duplicate

Fiala 2008excluded       duplicate

Kettering 2008excluded       duplicate

Wang 2008excluded       duplicate

Lutomsky 2008excluded       duplicate

Pak 2008excluded       duplicate

Oral 2008excluded       duplicate

Khan 2008excluded       duplicate

Mortada 2008excluded       anticoagulation for AF ablation

Perea 2008excluded       mixed population pAF and persistent AF

Chang 2008excluded       SVT and AF/ no follow-up

Tang 2008excluded       duplicate

Chen 2008excluded       mixed population pAF and persistent AF

Yamada 2008excluded       AF ablation in Brugada syndrome patients

Verma 2008excluded       mixed population pAF and persistent AF

Dixit 2008excluded       duplicate

Maly 2008excluded       duplicate

Phillips 2008excluded       no follow-up

Crawford 2008excluded       duplicate

68

Liakishev 2013excluded       duplicate

Moreira 2008excluded       mixed population pAF and atrial flutter

Yao 2007excluded       mixed population pAF and persistent AF

McClelland 2007excluded       mixed population pAF and persistent AF

Pruitt 2007excluded       mixed population pAF and persistent AF

Narayan 2007excluded       mixed population pAF and persistent AF

Knecht 2007excluded       mixed population pAF and flutter

Issa 2007excluded       no follow-up/AV node ablation + PPM

Wang 2007excluded       duplicate

Arentz 2007excluded       duplicate

Gaita 2007excluded       mixed population pAF and persistent AF/HOCM pts

Chang 2007excluded       no follow-up

Yamane 2007excluded       duplicate

Haïssaguerre 2007excluded       mixed population pAF and persistent AF

Sra 2007excluded       mixed population pAF and persistent AF

Suwalski 2007   Includedexcluded   surgical PVI with VATS

Gao 2007excluded       duplicate

Kurosaki 2007excluded       mixed population pAF and persistent AF

Atarashi 2007excluded       duplicate

Verma 2007 exclude       duplicate

69

d

Nakagawa 2007excluded       mixed population pAF and persistent AF

Turco 2007excluded       mixed population pAF and persistent AF

Solheim 2007excluded       mixed population pAF and persistent AF

Papone 2006excluded       duplicate

Willems 2006excluded       duplicate

Bauer 2006excluded       duplicate

Liu 2006excluded       duplicate

Richter 2006excluded       mixed population pAF and persistent AF

Sheikh 2006excluded       duplicate

Kistler 2006excluded       mixed population pAF and persistent AF

Cheema 2006excluded       mixed population pAF and persistent AF

Di Biase 2013excluded       duplicate

Dixit 2006excluded       duplicate

Heist 2006excluded       no follow-up

Nilsson 2006excluded       duplicate

Wongcharoen 2006excluded       no follow-up

Scanavacca 2006   Includedexcluded   no PVI

Estner 2006excluded       mixed population pAF and persistent AF

Yu 2006excluded       not in english

70

Calo 2006excluded       duplicate

Obergassel 2006excluded       pts with HOCM

Oral 2006   Included   Included  

Akpinar 2006excluded       mixed population pAF and persistent AF

Cheema 2006 excluded       duplicate

Risius 2006excluded       no follow-up

Stabile 2006excluded       mixed population pAF and persistent AF

Hocini 2005excluded       duplicate

Tojo 2005excluded       no follow-up

Kumagai 2005excluded       duplicate

Mantovan 2005excluded       mixed population pAF and persistent AF

Jiang 2005   Includedexcluded   CT guided AF ablation

Reant 2005excluded       mixed population pAF and chronic AF

Fassini 2005excluded       duplicate

Liu 2005excluded       duplicate

Tse 2005excluded       duplicate

Lickfett 2005excluded       ablation of atrial flutter

Ninet 2005excluded       mixed population pAF and persistent AF

Mack 2005excluded       mixed population pAF and persistent AF

Rao 2005excluded       persistent and permanent AF

71

Liu 2005excluded       duplicate

Molloy 2005excluded       mixed population pAF and persistent AF

Kocheril 2005   Includedexcluded   right atrial ablation

Cappato 2005excluded       survey

Tsao 2005excluded       duplicate

Tondo 2005excluded       duplicate

Oral 2004excluded       duplicate

Mokadam 2004excluded       mixed population pAF and persistent AF

Sacher 2004excluded       not in english

Tanner 2004excluded       ablation of atrial tachycardias

Brembilla-Perrot 2004excluded      

effect of transoesophageal pacing in the diagnostic evaluation of patient with unexplained syncope

Oral 2004excluded       duplicate

Katritsis 2014excluded       duplicate

Schwartzman 2004excluded      

study of pulmonry veins as source of arrhythmogenic atrial ectopy

Gillinov 2004excluded       mixed population pAF and persistent AF

Pappone 2004   Includedexcluded   effect of vagal denervation + PVI

Jansens 2004   Includedexcluded   paper not available

Calo 2004excluded       mixed population pAF and persistent AF

Todd 2003excluded       longstanding AF

Gillis 2003 exclude       duplicate

72

d

Oral 2003excluded       no follow-up

Lemke 2003excluded       duplicate

Arentz 2003excluded       mixed population pAF and persistent AF

Oral 2003excluded       duplicate

Weerasooriya 2003excluded       no follow-up

Duff 2003excluded       duplicate

Raman 2003excluded       mixed population pAF and persistent AF

Wang 2003excluded       paper not available

Tada 2003   Includedexcluded   focused on QoL

Berkowitsch 2013   Includedexcluded   focused on QoL

Hocini 2003excluded       AF pts with sinus pauses (>=3sec) after fast AF

Oral 2003excluded       duplicate

Tse 2003excluded       mixed population pAF and persistent AF

Stabile 2003excluded       mixed population pAF and persistent AF

Sanchez 2003excluded       no follow-up

Lin 2003   Includedexcluded   ablation of non PV ectopies

Padeletti 2003excluded       duplicate

Patel 2003excluded       no follow-up

Brignole 2003excluded       duplicate

73

Marchlinski 2003excluded       mixed population pAF and persistent AF

Dill 2003excluded       focus on PV stenosis post ablation

Pürerfellner 2003excluded       focus on PV stenosis post ablation

Camm 2003excluded       no ablation

Stabile 2003excluded       duplicate

Jaïs 2002excluded       duplicate

Brembilla-Perrot 2003excluded       not in english

Goya 2002excluded       PAF post AF ablation

Tada 2002excluded       duplicate

Epstein 2002excluded       duplicate

Takahashi 2002excluded       no follow-up

Brignole 2002excluded       duplicate

Katritsis 2002excluded       no follow-up

Oral 2002excluded       mixed population pAF and persistent AF

Brembilla-Perrot 2002excluded       not in english

Brembilla-Perrot 2002excluded       not in english

Pappone 2001excluded       mixed population pAF and persistent AF

Tse 2001excluded       duplicate

Simpson 2001excluded       duplicate

Bailin 2001 exclude       duplicate

74

d

Hindricks 2001excluded       no follow-up

Stabile 2001excluded       duplicate

Sueda 2001excluded       chronic AF

Dagres 2001excluded       duplicate

Ashar 2000excluded       no follow-up

Padeletti 2000excluded       chronic AF

Pappone 2000excluded       mixed population pAF and persistent AF

Gasparini 2000excluded       mixed population pAF and chronic AF

Chan 2000excluded       focus in coagulum formation during ablation

Gasparini 2000excluded       mixed population pAF and persistent AF

Hocini 2000excluded       no follow-up

Tai 2000excluded       no ablation

Gilis 2000excluded       duplicate

Tsai 2000   Includedexcluded   ablation of SVC ectopies

Fitts 2000excluded       duplicate

Lee 2000excluded       duplicate

Jais 2000   Includedexcluded   paper not available

chen 1999excluded       duplicate

Gasparini 1999excluded       focus on thromboembolic events

75

Waldo 1999excluded       duplicate

Gillis 1999excluded       duplicate

Proclemer 1999excluded       duplicate

Marshall 1999excluded       duplicate

Hsieh 1999   Includedexcluded   paper not available

Gillis 2009excluded       duplicate

Marshall 1999excluded       duplicate

Brignole 1999excluded       duplicate

Furlanello 1999excluded       AF in athletes

Kuck 1998excluded       no ablation

Fitts 1998excluded       duplicate

Herz 1998excluded       not in english

Chen 1998excluded       no follow-up

Kalman 1997excluded       ablation of atrial flutter

Kim 1997excluded       mixed population pAF and atrial flutter

Brignole 1997excluded       duplicate

Leitch 1997excluded       EP study

Kamalvand 1997excluded       duplicate

Schuchert 1997excluded       not in english

Chen 1996 exclude       AV ablation

76

d

Dorian 1996excluded       duplicate

Tai 1995excluded       ablation of AV accessory pathways

Lau 1995excluded       duplicate

Xang 1995excluded       ablation of atrial flutter

Iesaka 1994excluded       ablation of AV accessory pathways

Kay 1989excluded       duplicate

Pokushalov 2010excluded       mixed population pAF and persistent AF

Eitel 2011   Included   Included  Podd 2015   Included   Included mixed population pAF and persistent AFPood 2016   Included   Included  Bjorkenheim 2016   Included   Included  Kuck et al 2016   Included   Included  Verma et al 2010   Included   Included  

Martinek 2007excluded       mixed population pAF and persistent AF

Kapa 2013excluded       mixed population pAF and persistent AF

Veasey et al 2010excluded       mixed population pAF and persistent AF

Yang et al 2016excluded       mixed population pAF and persistent AF

77

78