Objectives: The main purpose of the present study was to compare left atrium fibrosis measured with integrated backscatter in atrial fibrillation patients with and without prior thromboembolism defined as cerebral infarction, transient ischemic attack or peripheral embolism.

Method: We performed a cross-sectional study implementing a case-control design. The enrolled patients were allocated to two main categories; Stroke or Non Stroke. Clinical characteristics, standard echocardiographic parameters and integrated backscatter measurements from posterior wall in left atrium were obtained. Integrated backscatter measurements from left atrium posterior wall were calibrated to integrated backscatter measurements in the pericardium.

Results: Of the 90 patients included, 66% were men, and mean age was 71.8 ± 10.3 years. A history of thromboembolic disease was documented in 21 (23%) patients. Mean calibrated integrated backscatter was measured to -10.9 ± 6.3 dB in the Stroke group and -10.1 ± 5.6 dB in the Non Stroke group, with no significant difference between groups (p = 0.57). Calibrated integrated backscatter was significant higher in permanent atrial fibrillation vs persistent atrial fibrillation (p = 0.01) and borderline significantly higher in permanent AF vs recurrent AF (p = 0.05). There was no significant difference between paroxysmal and persistent AF or between paroxysmal and permanent AF. Calibrated integrated backscatter tended to increase with time since AF diagnosis.

Conclusion: Calibrated integrated backscatter did not differ between AF-patients with and without prior stroke.

Atrial fibrillation, Stroke, Echocardiography, Integrated backscatter

Aas TS, Loennechen JP, Salvesen Ø, Midgard R (2020) No Association between Stroke and Left Atrial Fibrosis estimated with integrated backscatter. J Cardiol Clin Res 8(1): 1148

AF: Atrial Fibrillation; IBS: Integrated Backscatter; TIA: Transient Ischemic Attack; LGE-MRI: Late Gadolinium Enhancement Magnetic Resonance Imaging; ACE: Angiotensin Converting Enzyme; LVEF: Left Ventricular Ejection Fraction; S’med: Mitral Annulus Peak Systolic Tissue Velocity Medial; S’lat: Mitral Annulus Peak Systolic Tissue Velocity Lateral; LVEDV: Left Ventricular End-Diastolic Volume; LVESd: Left Ventricular End-Systolic Volume; MAPSEmed: Mitral Annulus Peak Systolic Excursion Medial; MAPSElat: Mitral Annulus Peak Systolic Excursion Lateral; LVEDd: Left Ventricular End-Diastolic Diameter; LVESd: Left Ventricular End-Systolic Diameter; LAd: Left Atrial Maximum Anteroposterior Diameter; LAA1: Left Atrial Area Measured End-Systolic in 4-Chamber; LAA2: Left Atrial Area Measured End-Systolic in 2-Chamber; LAL1: Left Atrial Longitudinal Diameter in 4-Chamber; LAL2: Left Atrial Longitudinal Diameter in 2-Chamber; LAvolume: Left Atrial Volume; E: Peak Diastolic Velocity in Early Mitral Flow; A: Peak Diastolic Velocity in Late Mitral Flow; DT: Deceleration Time in Early Mitral Flow; E’med: Peak Early Diastolic Tissue Velocity in Mitral Annulus Medial; E’lat: Peak Early Diastolic Tissue Velocity in Mitral Annulus Lateral; A’med: Peak Late Diastolic Tissue Velocity in Mitral Annulus Medial; A’lat: Peak Late Diastolic Tissue Velocity in Mitral Annulus Lateral; cIBS: Calibrated Integrated Backscatter; IBSa: Integrated Backscatter Measured in Posterior Wall of Left Atrium; IBSp: Integrated Backscatter Measured in Pericardium

Atrial fibrillation (AF) is the most prevalent sustained arrhythmia with a prevalence of 2-3 % in the adult population [1]. It is associated with significant morbidity and mortality, primarily due to associated comorbidities and increased risk of ischemic stroke [2]. It is suggested that AF is the cause of 17-30 % of ischemic strokes [3]. Strokes caused by AF are larger and more disabling than those of other etiologies [4]. Oral anticoagulants prevent thromboembolism and improve survival, but increase the risk of bleeding [5, 6]. It is therefore of great importance to identify the individual risk of stroke as precisely as possible in patients with AF.

CHA2 DS2 -VASc score has been established as the tool of choice to estimate the individual risk of stroke in AF patients and is recommended in guidelines from European Society of Cardiology (in collaboration with The European Association for CardioThoracic Surgery) and guidelines from The American College of Cardiology/ American Heart Association/ The Heart Rhythm Society [5,6]. It is a point-based risk stratifying tool that includes the most important clinical risk factors. However, CHA2 DS2 -VASc score has limited abilities to predict individual stroke risk [7], and there is a clear need for improvement.

Existing risk stratifying tools do not take into account individual pathophysiological criteria like structure and function of the left atrium. Left atrium fibrosis is a well-established risk factor for AF and a hallmark for progression of the disease. AF induces electrical, contractile and structural remodeling [8], with fibrosis as a common feature. It has also been shown that AF per se induces atrial fibrosis. Left atrium fibrosis measured by late gadolinium enhancement magnetic resonance imaging (LGE-MRI) is associated with stroke in AF, also after correction for other risk factors. [9]. The mechanism of this is unknown. It might be due to reduced left atrium contractile function or pathological atrial endothelial function [10].

Hypertension, diabetes mellitus, heart failure and increasing age are all associated with atrial fibrosis [10, 11], which might be a common mechanism linking risk factors and stroke in AF patients. Estimation of left atrium fibrosis might therefore improve risk stratifying in AF patients. LGE-MRI is currently the best method to quantify left atrium fibrosis [12]. However, it is an expensive and not easily accessible examination. Echocardiography is generally accessible, without risk and has a low cost. Left atrium fibrosis can be estimated with echocardiography measuring tissue reflectivity on two-dimensional grey-scale images, a method called integrated backscatter (IBS). This method was validated in a study comparing left atrium fibrosis estimated with IBS and histological graded left atrium fibrosis in biopsies from left atrium [13]. IBS is measured in decibels (dB) on a negative logarithmic scale with higher values indicating more fibrosis. Structures with no fibrosis (eg. blood) have a low ultrasound reflectivity and correspondingly low IBS, whereas cardiac structures with high content of fibrosis have high ultrasound reflectivity and IBS near 0 dB. Normal myocardium has an intermediate IBS value, which increases as the content of fibrosis increases. So far no studies have investigated the association between integrated backscatter and risk of thromboembolism. The main purpose of the present study was to examine whether AF patients with a history of thromboembolism had higher IBS values than those without. In addition we wanted to examine the association between IBS measurements and CHA2 DS2 -VASc score, left atrium size, age, duration of AF, and left ventricular diastolic and systolic function.

Study design and population

We performed a cross-sectional study implementing a casecontrol design. Patients with AF were recruited from hospitalized patients and patients admitted to the outpatient clinic at the Department of Internal Medicine, Molde Hospital.

We applied the following inclusion criteria:

1. Age > 18 years

2. Known AF

3. Mandatory informed consent signature

The exclusion criteria comprised:

1. Clinically significant valve disease

2. Prior valve surgery

3. Arterial precerebral stenotic disease

4. Ischemic stroke, transient ischemic attack (TIA), peripheral embolism with probable alternative etiology than AF

5. Prior ablation procedure for arrhythmi

All patients went through an interview and a clinical examination and the comprehensive data including medical history, clinical characteristics, and current medication were extracted from the electronic patient records. Blood pressure, heart rate, body weight and height were measured and electrocardiogram recorded at inclusion. Diagnoses from medical history were critically evaluated, but strict fulfillment of diagnostic criteria was not required because of the non-scientific nature of the electronic patient record. A diagnosis was accepted if it was considered more probable than not.

The study protocol was approved by The Regional Committee for Medical and Health Research Ethics.

Atrial fibrillation

AF had to be documented with a 12-channel electrocardiogram or as AF with duration of at least 30 seconds in any other electrocardiographic registration. Patients without AF at the time of inclusion had to have at least two episodes of AF with at least seven days interval. Atrial flutter alone did not qualify for inclusion.

Patients were classified as having paroxysmal, persistent or permanent AF after evaluation of medical records and the actual interview of the patient. Paroxysmal AF was defined as recurrent AF with no episodes of more than seven days duration. Persistent AF was defined as recurrent AF with at least one episode of more than seven days duration. AF in need of cardioversion (electrical or medical) was classified as persistent AF, except when cardioversion was done after less than 48 hours duration. Permanent AF was defined as sustained AF accepted by doctor and patient.

Thromboembolic disease

Cerebral infarction was defined as rapidly developing clinical signs of focal or global disturbance of cerebral function, lasting more than 24 hours caused by acute occlusion of cerebral arterial blood supply. Patients were also categorized as having experienced cerebral infarction when signs of cerebral infarction were documented on any cerebral imaging, irrespective of symptoms and signs. TIA was defined as an episode of temporary and focal dysfunction of vascular origin, commonly lasting from 2 to 15 minutes, but occasionally lasting as long as 24 hours, which do not leave any persistent neurological deficit. Peripheral embolism was defined as acute, impaired arterial perfusion of an extremity.

The enrolled patients were allocated to two main categories; Stroke or Non Stroke. Patients, who had experienced TIA or peripheral embolism, were also included in the Stroke category.

After categorizing the patients in Stroke and Non Stroke groups, a neurologist confirmed the probable diagnosis of prior TIA or ischemic stroke using the electronic patient record for review of medical history and results from cerebral imaging.

Echocardiography

All echocardiographic exams were done with a commercially available ultrasound system (E9 4D Expert BT11, General Electric Vingmed, Milwaukee, Wisconsin, USA) and post processing was done in EchoPac BT 112, General Electric Medical Systems, Horten, Norway. Echocardiography was performed using standard views and harmonic imaging. Acquisition of echocardiography images and cine-loops were done unblinded to patient characteristics. In the apical 4-chamber and 2-chamber view left ventricular end-diastolic and end-systolic volumes (LVEDV and LVESV) were measured and left ventricular ejection fraction (LVEF) calculated by Simpson’s method. In the parasternal long-axis view left atrium maximum anteroposterior diameter (LAd), left ventricular end-diastolic and end-systolic diameter (LVEDd and LVESd) were measured. Left atrium area was measured endsystolic in 4-chamber (LAA1) and 2-chamber (LAA2) apical view. Left atrium longitudinal diameter (from left atrium basal wall to mitral annulus plane) was measured in 4-chamber (LAL1) and 2-chamber (LAL2) apical view. Left atrium volume (LAvolume) was calculated by biplane area-length method (Volume = (8xA1xA2)/ (3xπxLAL )), using the shortest of the two longitudinal left atrium diameters (LAL ). Left atrium volume was indexed to body surface area (LAvolume index). Mitral annulus peak systolic excursion was measured medial and lateral (MAPSEmed and MAPSElat) to mitral annulus in 4-chamber apical view. Pulsed wave Doppler at tips of mitral valve leaflets was used to measure early (E) and late (A) diastolic filling velocities (A only in those with sinus rhythm), E/A-ratio and E deceleration time (DT). Left ventricular early and late diastolic tissue velocities were measured by tissue Doppler imaging of the medial (E’med and A’med) and lateral (E’lat and A’lat) mitral annulus (A’med and A’lat only in those with sinus rhythm). E/ E’ was calculated from the ratio of mean of E over mean of E’med and E’lat. Pulsed wave Doppler measurements in patients in AF were averaged over 3 cardiac cycles.

All IBS measurements were done blinded to patient characteristics. Three cardiac cycles were stored in cine-loop format for offline IBS analysis. IBS was measured in the posterior wall of the left atrium (IBSa) and in pericardium (IBSp) outside left ventricular inferior wall in the parasternal long-axis view. IBSp provided the reference value of ultrasound reflectivity to calculate cIBS. Sample volume for all IBS measurements was set to 2 x 3mm with the frame rate adjusted to 80 - 120 s-1. High resolution zoom was used to improve positioning of the sample volume for IBS measurements. Measured values were calibrated subtracting IBSp from IBSa, resulting in calibrated IBS (cIBS). This method of calibration has been reported from other investigators studying AF and IBS [14]. Four IBSa and two IBSp was measured end-diastolic in three consecutive heart cycles. The mean of the six IBSp measurements subtracted from the median of these 12 IBSa measurements defined cIBS for each patient. Median was used to diminish the risk of one single deviant IBS measurement to influence the result.

Statistics

Sample size was calculated based on a previous study [14] using 5dB as a clinical relevant difference. According to this, a two-sided independent sample t-test with a 5% level of significance, a standard deviation of 5 and a power of 90%, gave a target study population of 13 patients in the Stroke group and 75 in the Non stroke group. We planned to include 15 patients in the Stroke group and 75 patients in the Non stroke group. Increasing ratio of patients with stroke will increase power.

Continuous variables are presented as mean ± standard deviation (SD) and dichotomous data are presented as numbers and percentages. Variables of the study groups were compared using the independent samples t-test. Associations between echocardiographic variables and cIBS, AF type and cIBS and CHA2 DS2 -VASC score and cIBS were analyzed with Pearson correlation test. Comparison of mean cIBS between the two main groups was corrected for differences in age and the variables in baseline characteristics which differed significantly, by linear regression analysis. An association was considered statistically significant for a two-sided p-value of < 0.05. From the study population, 30 patients where included in an intraand interobserver analysis. Agreement was measured using intraclass correlation coefficient. All statistical analysis was performed using IBM SPSS Statistics Version 23.0.0.2.

Patients were recruited from October 19th 2012 until October 12th 2015. Patients were included and examined nonconsecutively by one of the investigators during routine work as a clinical cardiologist. All patients asked, accepted to participate in the study. A total of 90 patients were included, 76 (84%) after hospitalization and 14 (16%) from the outpatient clinic. Mean age was 71.8 ± 10.3 years, and 66 % were men (Table 1).

Table 1: Baseline characteristics of study patients.

During echocardiography 40 patients (44%) were in AF. Mean time since first AF diagnoses was 53.9 ± 58.3 months. A history of thromboembolic disease was documented in 21 (23%) patients, including 14 (15%) with cerebral infarction, 5 (6%) with TIA and 2 (2%) with peripheral embolism. Mean time since last thromboembolic episode was 63.1 ± 54.8 months. Patients in the Stroke group had significantly smaller body surface area, higher systolic and diastolic blood pressure and more frequently a diagnosis of hypertension. Other baseline characteristics did not differ significantly between the two groups. No significant valvular disease was seen in any patient. Of the echocardiographic measurements, only MAPSEmed and E’med differed significantly between groups (Table 2).

Table 2: Standard echocardiographic variables of patients.

Of 90 patients, 89 had successfully measured and documented IBS parameters. The IBS measurements for one patient were lost due to technical reasons.

There was no significant difference between cIBS in the two groups with mean cIBS -10.9 ± 6.3 dB in the Stroke group and -10.1 ± 5.6 dB in the Non Stroke group, p = 0.57 (Table 3).

Table 3: IBS measurements.

Mean difference in cIBS between groups was 0.82 ± 1.4 dB (95 % confidence interval -2.0 to 3.7 dB). There were still no differences between the groups after correction for age, body surface, and hypertension, systolic and diastolic blood pressure using linear regression analysis. There was no significant difference in IBSa between the two groups (p = 0.07), but significant higher IBSp in the Non Stroke group (p = 0.002).

Correlation studies between CHA2 DS2 -VASc score and cIBS showed no significant association (p = 0.97) also after excluding stroke from the score (p = 0.97). There was no significant association between parameters of systolic function and cIBS, left ventricular size and cIBS or left atrium size and cIBS. cIBS showed a negative, significant correlation with E and with E/ E’ (Table 4).

Table 4: Correlation analysis between cIBs and echocardiographic variables.

Other parameters for left ventricular filling and diastolic function showed no significant association.

cIBS was -10.2 dB in paroxysmal AF, -12.0 dB in persistent AF, -8.3 dB in permanent AF and -11.0 dB in recurrent AF (paroxysmal AF or persistent AF). cIBS was significant higher in permanent atrial fibrillation vs persistent atrial fibrillation (p = 0.01) and borderline significantly higher in permanent AF vs recurrent AF (p = 0.05).There was no significant difference between paroxysmal and persistent AF or between paroxysmal and permanent AF.

cIBS tended to increase with time since AF diagnosis (r = 0.26, p = 0.01) with cIBS -11.0 dB in the tertile with shortest, -11.9 dB in the tertile with intermediate and -8.5 dB in the tertile with longest time since diagnosis of AF.

There was no difference in cIBS between the groups with higher and lower than median LAvolume index (47.0 ml/ m2 ) with cIBS of -10.4 dB in both groups.

There was no significant correlation between cIBS and age (r = 0.15, p = 0.69). Distributing patients in tertiles by age showed a non-significant trend for higher cIBS in older patients; cIBS -11,2 dB in youngest tertile, -10,1 dB in intermediate tertile and -9,9 dB in oldest tertile.

Intra observer agreement was very good with intra class correlation coefficient 0.9238 with a narrow 95% confidence interval (0.8476-0.9628). Inter observer agreement was also very good with intra class correlation coefficient 0.7722, although with a much broader 95% confidence interval (0.5776-0.8842).

Figure 1 :Example of assessment of end-diastolic IBS variables in one cardiac cycle. Positioning of four sample volumes (colored rings) in the posterior wall of left atrium and two sample volumes in pericardium. Red line marks end of diastole. High resolution zoom was used to improve sample volume positioning

Figure 2: A. cIBS in small and large left atriums (LA). Study population divided by median (47ml/m2). B. cIBS and time since AF diagnosis. Patients grouped in tertiles according to time since AF diagnosis. C. cIBS and age. Patients grouped in age tertiles. D. cIBS and AF type. Mean ± standard deviation. y = years. dB = decibel. m = months.* Significant trend for increasing cIBS with increasing time since AF diagnosis, p = 0.01.† Persistent AF vs permanent AF, p = 0.01. Recurrent AF (persistent and paroxysmal AF) vs permanent AF, p = 0.05

The present study showed no differences in cIBS between the groups with and without stroke. The absence of association between IBS and stroke could be caused by absence of association between IBS and left atrium fibrosis, between left atrium fibrosis and thromboembolism or both.

Several previous studies have shown an association between left atrium fibrosis and stroke or risk factors for stroke. There is a mechanistic rationale for a causative correlation between left atrium fibrosis and stroke. High levels of left atrium fibrosis, measured by LGE-MRI, in patients with AF are associated with reduced atrial mechanical function [15], which is supposed to be an important mechanism of thrombosis. Furthermore, left atrium fibrosis, measured by LGE-MRI, is independently associated with appendage thrombus and spontaneous contrast [16]. These are echocardiographic signs accepted as risk factors for stroke [17]. Data also supports a role of endothelial dysfunction and changed hemostatic function associated with left atrium fibrosis [10]. Moreover, risk factors for stroke in AF such as increasing age, heart failure and diabetes mellitus are all associated with left atrium fibrosis [10,11]. In one study left atrium fibrosis, measured by LGE-MRI, was associated to prior stroke [9]. All in all there is rationale to suspect atrial fibrosis as a risk factor for stroke in patients with AF and the documentation of an association between left atrium fibrosis and stroke is relatively robust.

To our experience, IBS measurements are challenging. Quantification of left atrium fibrosis by IBS is limited by the spatial resolution of echocardiography and cannot give information on trans-morality of fibrosis. Correct positioning of sample volumes in the atrial wall is crucial, but difficult. The atrial myocardium is thin and often not clearly delineated from atrial lumen and pericardium. Reduced image quality hampers correct positioning of sample volume. Structures with high fibrotic content such as descending aorta and posterior mitral leaflet are adjacent to posterior left atrium wall and may influence IBS measurements. Despite these obvious difficulties using IBS to estimate left atrium fibrosis, there are several studies reporting significant associations between IBS and various characteristics and outcomes from AF. These include outcome after ablation [14], prediction of postoperative AF [13] and progression from paroxysmal to persistent AF [18]. There is only one study assessing the association between IBS and histological graded left atrium fibrosis [13]. This study was done on patients undergoing coronary bypass grafting. The associations were highly significant, with low intra- and inter observer variability, like we find in our study. The IBS measured in the left atrium wall was calibrated with IBS measured in the pericardium by a ratio (IBSa/ IBSp). We chose to calibrate IBSa to IBSp by subtraction, a mathematical more reasonable method since IBS measurements are values on a logarithmic scale. Other investigators have used calibration by subtraction successfully with good reproducibility [14,19]. We also calculated the IBSa/ IBSp-ratio and there were still no significant difference between patients with and without stroke. In the present study we used multiple IBS measurements (12 IBSa and 6 IBSp) for each patient to overcome the challenges of correct IBS measurements. Limiting measurements to left atrium posterior wall simplifies application of the method, but disparages the heterogenic distribution of fibrosis in left atrium wall. Most other studies using integrated backscatter to estimate left atrium fibrosis have also restricted IBS measurements to left atrial posterior wall, usually using one single or only a few IBS measurements [13, 14,19,20].

Absolute values for IBS and cIBS were higher in our study than in other studies, indicating higher degree of left atrium fibrosis. A large proportion of our patients were hospitalized, time since AF diagnoses was long and mean age was high. It might be that our population represents more advanced AF, in which progression of structural changes, including atrial fibrosis, are more pronounced. A difference in cIBS between groups may diminish as the AF disease approaches an end stage situation with extensive left atrium fibrosis. When many of the patients have high degree of left atrium fibrosis, other risk factors may have become more important. No patients in our study were excluded because of reduced image quality. Selective inclusion of patients with highest quality images may give other results.

IBSp was significant higher in the Non stroke group. This was an unexpected result and difficult to interpret. Variation in image quality between groups may explain a difference. Patients in the Non Stroke group had non-significant higher weight and significant higher body surface area. Higher body weight is wellknown related to poor echo image quality. If image quality was systematically better in one group, this mechanism could have biased the results. Other undefined factors influencing image acquisition and analysis may have resulted in differing IBSp. Such factors should also affect IBSa, and therefore be corrected for. This is the motive for calculating cIBS.

Several findings in the present study support that cIBS truly measures a degree of fibrosis. cIBS was significantly higher in permanent AF and tended to be higher with longer duration since AF diagnosis. These results are matching results from earlier studies [18]. cIBS correlated significantly to E and E/ E’, two parameters central to assessment of diastolic function. It is reasonable to suggest that left atrium fibrosis influence diastolic function. cIBS increased with increasing age, a result also supported by earlier studies [12]. The trends in cIBS according to type of AF, time since AF diagnosis and age are as expected but the differences in cIBS are small. It might be that limitations of the present method to quantify fibrosis weaken the ability to detect an association between cIBS and stroke. It might also be that there is no strong association between atrial fibrosis and thromboembolic stroke.

We could not find any association between CHA2 DS2 -VASc score and cIBS. Consequently, this study does not support the hypothesis that left atrium fibrosis is the link between known clinical risk factors for thromboembolism and stroke.

It is strength that this is a real life study on IBS and left atrium fibrosis in a daily clinical practice. We used broad inclusion criteria and all patients asked, accepted to participate. Also, patients with reduced image quality were included. We have only one case of missing IBS data.

There are several limitations to the study. The primary endpoint was studied retrospectively and biopsies or LGE-MRI were not used to verify left atrium fibrosis. If many of the strokes are not due to cardiac embolism, it could have weakened the ability to show differences between groups. Also examinations were done at one center, and one investigator performed all echocardiographic examinations.

In patients with AF cIBS did not differ between those with and without prior stroke. No association between cIBS and known clinical risk factors was documented. In a clinical setting IBS is not a proper method to detect stroke risk, possibly because it is not a precise tool to estimate left atrium fibrosis. Further studies are needed to validate the detection of left atrium fibrosis by other echocardiographic methods and LG-MRI.

1. Wilke T, Groth A, Mueller S, Pfannkuche M, Verheyen F, Linder R, et al. Incidence and prevalence of atrial fibrillation: an analysis based on 8.3 million 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. 2013; 15: 486-493.

2. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke; a journal of cerebral circulation. 1991; 22: 983-988.

3. Kishore A, Vail A, Majid A, Dawson J, Lees KR, Tyrrell PJ, et al. Detection of atrial fibrillation after ischemic stroke or transient ischemic attack: a systematic review and meta-analysis. Stroke; a journal of cerebral circulation. 2014; 45: 520-526.

4. Dulli DA, Stanko H, Levine RL. Atrial fibrillation is associated with severe acute ischemic stroke. Neuroepidemiology. 2003; 22: 118-123.

5. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS: The Task Force for the management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESCEndorsed by the European Stroke Organisation (ESO). 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. 2016; 37: 2893-2962.

6. January CT, Wann LS, Alpert JS, Michael EF, Karen LF, Paul AH, et al. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. Journal of the American College of Cardiology. 2014; 130: 199-267.

7. van den Ham HA, Klungel OH, Singer DE, Leufkens HG, van Staa TP. Comparative Performance of ATRIA, CHADS2, and CHA2DS2-VASc Risk Scores Predicting Stroke in Patients With Atrial Fibrillation: Results From a National Primary Care Database. Journal of the American College of Cardiology. 2015; 66: 1851-1859.

8. Allessie M, Ausma J, Schotten U. Electrical, contractile and structural remodeling during atrial fibrillation. Cardiovascular research. 2002; 54: 230-246.

9. Daccarett M, Badger TJ, Akoum N, Burgon NS, Mahnkopf C, Vergara G, et al. Association of left atrial fibrosis detected by delayed-enhancement magnetic resonance imaging and the risk of stroke in patients with atrial fibrillation. Journal of the American College of Cardiology. 2011; 57: 831-838.

10. Boixel C, Fontaine V, Rucker-Martin C, Paul Milliez, Liliane Louedec, Jean-Baptiste Michel, et al. Fibrosis of the left atria during progression of heart failure is associated with increased matrix metallo proteinases in the rat. Journal of the American College of Cardiology. 2003; 42: 336-344.

11. Everett TH, Olgin JE. Atrial Fibrosis and the Mechanisms of Atrial Fibrillation. Heart rhythm: the official journal of the Heart Rhythm Society. 2007; 4: 24-27.

12. Dzeshka MS, Lip GY, Snezhitskiy V, Shantsila E. Cardiac Fibrosis in Patients with Atrial Fibrillation: Mechanisms and Clinical Implications. Journal of the American College of Cardiology. 2015; 66: 943-959.

13. Wang GD, Shen LH, Wang L, Li HW, Zhang YC, Chen H. Relationship between integrated backscatter and atrial fibrosis in patients with and without atrial fibrillation who are undergoing coronary bypass surgery. Clinical cardiology. 2009; 32: 56-61.

14. den Uijl DW, Delgado V, Bertini M, Tops LF, Trines SA, van de Veire NR, et al. Impact of left atrial fibrosis and left atrial size on the outcome of catheter ablation for atrial fibrillation. Heart. 2011; 97: 1847-1851.

15. Savelieva I, Kakouros N, Kourliouros A, Camm AJ. Upstream therapies for management of atrial fibrillation: review of clinical evidence and implications for European Society of Cardiology guidelines. Part I: primary prevention. 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. 2011; 13: 308-328.

16. Akoum N, Fernandez G, Wilson B, Mcgann C, Kholmovski E, Marrouche N. Association of atrial fibrosis quantified using LGE-MRI with a|trial appendage thrombus and spontaneous contrast on trances ophageal echocardiography in patients with atrial fibrillation. Journal of cardiovascular electrophysiology. 2013; 24: 1104-1109.

17. Fagan SM, Chan KL. Transesophageal echocardiography risk factors for stroke in nonvalvular atrial fibrillation. Echocardiography. 2000; 17: 365-372.

18. Kubota T, Kawasaki M, Takasugi N, Imai H, Ishihara Y, Okubo M, et al. Left atrial pathological degeneration assessed by integrated backscatter transesophageal echocardiography as a predictor of progression to persistent atrial fibrillation: Results from a prospective study of three-years follow-up. Cardiovascular Ultrasound. 2012; 10: 28.

19. Sasaki N, Okumura Y, Watanabe I, Koichi N, Kazumasa S, Rikitake K, et al. Transthoracic echocardiographic backscatter-based assessment of left atrial remodeling involving left atrial and ventricular fibrosis in patients with atrial fibrillation. International journal of cardiology. 2014; 176: 1064-1066.

20. Zhu H, Zhang W, Zhong M, Zhang G, Zhang Y. Myocardial ultrasonic integrated backscatter analysis in patients with chronic atrial fibrillation. The international journal of cardiovascular imaging. 2010; 26: 861-865.