Refining the Prothrombotic State and Prognosis in Atrial Fibrillation With Left Atrial Appendage 3D Echocardiography

Left atrial (LA) volume is an echocardiographic marker of remodeling, thromboembolic 33 risk, and prognosis in atrial fibrillation (AF); limited data are available on LA appendage (LAA) 34 characterization beyond morphology. We sought to evaluate LAA characteristics in 2-dimensional (2D) 35 and 3-dimensional (3D) transesophageal echocardiography (TEE) and the correlation with LA/LAA 36 prothrombotic state and prognosis. 37

We prospectively studied 206 hospitalized patients with AF using 2D transthoracic 38 echocardiography (TTE) and 2D/3D TEE of the LAA ≤24 hours from admission. Patients were divided 39 according to the presence or absence of LAA sludge and/or thrombus. Data on clinical events were 40 collected for 2 years. 41

Patients with LAA sludge/thrombus (n=35) on admission had higher LA volumes, lower left 42 ventricular ejection fraction, lower LAA emptying and filling flow velocity, larger 2D LAA measurements 43 (2D LAA ostium diameter, 2D LAA area) as well as larger 3D LAA measurements (higher 3D LAA 44 volumes (LAAV), higher 3D end-systolic [ES] LAA ostium area), and more frequently non-chicken wing 45 morphology. On multivariable logistic regression analysis, LAA filling flow velocity and 3D ES LAAV were 46 associated with the presence of LAA sludge/thrombus at admission (P=0.031 and P<0.0001 respectively). 47 Receiver operating characteristic curve analysis revealed the optimal cut-off for 3D ES LAAV to 48 discriminate patients at risk of death within 2 years was 9.3 mL. Kaplan–Meier curves demonstrated a 49 significant difference in survival at 2-year follow-up according to this value: 3 deaths occurred in the group 50 with 3D ES LAAV 9.3mL and 11 in those with volume 9.3 mL (P=0.02). 51

3D characterization of LAAV depicts a degree of LAA remodeling in AF that appears 52 associated with LA/LAA thrombogenicity and mid-term prognosis. 53 54 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 4 CONDENSED ABSTRACT 55 Limited data are available on left atrial appendage (LAA) remodeling in atrial fibrillation (AF). We 56 hypothesized that 3-dimensional (3D) evaluation of the LAA volume in AF could help to refine the 57 prothrombotic state and prognosis in AF. Patients with LAA sludge and/or thrombus exhibited lower LAA 58 filling and emptying flow velocities, and higher 2-dimensional (2D) and 3D LAA measurements. On 59 multivariable analysis, LAA filling flow velocity and 3D end-systolic LAA volume were associated with the 60 presence of LAA sludge/thrombus at admission (respectively, P=0.031 and P<0.0001). Kaplan–Meier 61 curves demonstrated a significant difference in survival at 2 years according to 3D ES LAA volume 62 (P=0.02). Three dimensional LAA volume reflects the degree of LAA remodeling in AF and is associated 63 with prothrombotic state and prognosis. 64 65 Key Words: three-dimensional echocardiography; atrial fibrillation; left atrial appendage 66 67 68 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 5 Non-standard Abbreviations and Acronyms 69 2D 2-dimensional 70 3D 3-dimensional 71 AF atrial fibrillation 72 BNP B-type natriuretic peptide 73 CRP C-reactive protein 74 ED end-diastole 75 ES end-systole 76 I indexed 77 LAA left atrial appendage 78 LAAV left atrial appendage volume 79 TEE transesophageal echocardiography 80 TTE transthoracic echocardiography 81 82 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 6

83 A prothrombotic state in atrial fibrillation (AF) fulfils Virchow’s triad of blood hypercoagulability, endothelial 84 injury and abnormal blood stasis.1 Left atrial (LA) blood stasis (i.e., transesophageal echocardiography 85 [TEE]-detected reduced LA appendage [LAA] blood flow, LAA spontaneous echo contrast, sludge) occurs 86 as a direct consequence of AF and contributes to thrombus formation. The thromboembolism risk 87 assessment is essential in the management of AF.2 88 Cardiac imaging plays a central role in evaluating the AF substrate (i.e., atrial cardiomyopathy).3, 4 89 Two-dimensional (2D)/3-dimensional (3D) transthoracic echocardiography (TTE), TEE, cardiovascular 90 magnetic resonance imaging and computed tomography are useful imaging modalities for the volumetric 91 and functional assessment of atrial remodeling.5 Several cardiac imaging parameters have been 92 described as markers of cardioembolic risk: large LA volumes,6 left ventricular systolic dysfunction,7 TEE-93 detected LA spontaneous echo contrast,8 decreased LAA emptying and filling velocities,9 and presence of 94 aortic atheroma.10 95 Limited data are available on the association between LAA parameters and LAA prothrombotic 96 state in computed tomography, magnetic resonance imaging11-13 and TEE 14-16 (Table S1). While LA 97 volume appears associated with patient prognosis,17 little is known about the relationship between LAA 98 parameters and LAA prothrombotic state and prognosis. 99 The aims of this study in an AF population were threefold: 1) to provide a comprehensive 100 evaluation of LAA using 2D and 3D TEE; 2) to determine which LAA parameter is best associated with a 101 prothrombotic state; and 3) to study the relationship between LAA-derived parameters and prognosis. 102 103

104 Study Design and Population 105 FASTRHAC is a multicenter, prospective study of patients hospitalized for AF (Ethics committee 106 authorization: CPP Ile de France V, number: 2014-A00280-47. NCT 02741349). The FASTRHAC 107

have been described previously.5, 18 All consecutive patients (≥18 years) hospitalized for 108 paroxysmal or persistent AF diagnosed on a 12-lead electrocardiogram who provided written informed 109 consent were included. Exclusion criteria were organic valvular disease defined according to the 110 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 7 guidelines,19-21 presence of a mechanical or biological prosthesis, contraindication to anticoagulant 111 treatment, lack of affiliation to a social security regimen, severe psychiatric history, and subjects 112 considered unlikely to present for follow-up. Comprehensive clinical characteristics, biological variables, 113 and 2-D and 3D TTE and TEE data were collected over 2 years. CHA2DS2-VASc score (Congestive heart 114 failure, Hypertension, Age ≥75 [doubled], Diabetes mellitus, prior Stroke or transient ischemic attack or 115 thromboembolism [doubled], Vascular disease, Age 65 to 74, Sex category [female]) was determined in 116 each patient. 117 This analysis focuses on the first 218 consecutive patients enrolled in the study who underwent 118 3D TEE at admission and had a comprehensive follow-up of 2 years (Figure 1). 119 120 Transthoracic and Transesophageal Echocardiography 121 2D and 3D TTE and TEE were performed within 24 hours of admission by experienced cardiologists using 122 X5-1, X72T and X82T transducers connected to an EPIQ 7 or CVx (Philips Medical Systems, Andover, 123 MA, USA). The data were transferred and analyzed offline using a TOMTEC workstation (Image Arena; 124 TOMTEC, Unterschleissheim, Germany) (L.S.D., T.S., E.C. and C.A.). 2D and 3D measurements were 125 analyzed following US and European Chamber Quantitation Guidelines.22, 23 Volumetric measurements 126 were indexed to body surface area. 127 2D transesophageal evaluation of the LAA included the following parameters (Figure 2): 128 evaluation of spontaneous echo contrast between 0° and 120° according to the Fatkin classification8 129 (grades 0, 1, 2, 3, sludge, thrombus); measurements at 90° of 2D ES and ED LAA ostium diameter, 2D 130 ES and ED LAA area; measurements at 90° of LAA emptying and filling flow velocities; and evaluation of 131 trabeculation severity (mild, moderate, severe). 132 The analysis of LAA with 3D TEE used 3D 1-beat zoom mode. 3D datasets of the LAA were 133 deemed adequate for analysis if all cavity segments were visible in the dynamic dataset. 3D full-volume 134 datasets were analyzed using software specifically designed for 3D volumetric analysis (Tomtec, Generic 135 Volume Software). The 3D TEE evaluation of the LAA included the following parameters (Figure 2): ES 136 and ED LAA volumes (LAAV); ES and ED LAA ostium areas; LAA morphology; and LAA number of lobes. 137 138 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 8 Statistical Analysis 139 The study population was divided into 2 groups according to the presence or absence of sludge and/or 140 thrombus in the LAA. Then, using the non-parametric Kruskal−Wallis rank test for continuous variables 141 and the2 or Fisher’s exact test for categorical variables, the baseline parameters were compared 142 between the 2 groups. First, logistic regression analyses were performed to identify determinants of the 143 presence of sludge and/or thrombus in the LAA. For each continuous variable, the choice between a 144 continuous or categorical classification was based on the lowest value of Akaike’s information criterion for 145 the corresponding univariate Cox model. To avoid collinearity and overfitting problems, each clinical and 146 biological variable was entered into a stepwise backward multivariable model and variables with P<0.20 147 were retained. The same method was repeated for the echocardiographic variables (Table S2). Finally, a 148 multivariable model adjusted for the selected variables was constructed on the population of interest. 149 Second, a receiver operating characteristic curve was depicted to analyze the capacity of the 3D 150 ES LAAV to identify patients at risk of dying within 2 years. A cut-off of 9.3 mL was determined (Figure S1) 151 and was used to build the Kaplan−Meir survival curves and was compared using the log-rank test. Cox 152 regression models analyses were performed to identify variables associated with death within 24 months 153 of follow-up. 154 Biological characterization of the population with a C-reactive protein (CRP) value >10 mg/L and 155 B-type natriuretic peptide (BNP) value ≥400 pg/mL were based on a careful analysis of the literature.24, 25 156 All analyses were performed using STATA V12 (StataCorp, College Station, TX). P<0.05 was 157 considered statistically significant. 158 159

160 Patient Baseline Characteristics 161 3D LAA evaluation was feasible in 206 of 218 patients undergoing TEE (94.5%) (Figure 1). At admission, 162 LAA spontaneous echo contrast grade 0 was found in 85 (41.3%) patients, grade 1 in 47 (22.8%), grade 2 163 in 26 (12.6%), grade 3 in 15 (7.3%), sludge in 33 (16.0%) and thrombus in 7 (3.4%). One-hundred and 164 twenty-nine (62.6%) patients were male, and the median age was 66.3±11.5 years; 75.7% patients had a 165 CHA2DS2-VASc ≥2, 30.1% had paroxysmal AF and 69.9% had persistent AF at admission (Table 1). 166 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 9 Thirty-five (17.0%) patients had LAA sludge/thrombus at admission (Table 1). Patients with LAA 167 sludge/thrombus had a higher prevalence of heart failure (and hospitalization due to AF associated with 168 heart failure). Patients with sludge and/or thrombus had higher values for BNP and CRP. There were no 169 significant differences between the groups regarding clinical and cardiovascular risk factors, CHA2DS2-170 VASc score, or troponin and D-dimer values. 171 Echocardiographic characteristics are shown in Table 2. On 2D TTE, patients with LAA 172 sludge/thrombus had higher median LA volume and lower median LVEF than those without. On 2D TEE, 173 patients with LAA sludge/thrombus had lower LAA emptying and filling flow velocities and greater LAA 174 measurements (ostium diameter, area). On 3D TEE, patients with LAA sludge/thrombus had a higher 175 median LAAV, greater ED LAA ostium area, and more frequently exhibited a non-chicken wing 176 morphology. 177 178 Determinants of LAA Prothrombotic State 179 On univariable analysis, predictors of LAA sludge/thrombus at admission were: history of heart failure or 180 acute heart failure, persistent AF, hospitalization for AF with heart failure, CRP >10 mg/L, BNP 181 ≥400 pg/mL, decreased LVEF, LAVI ≥45 mL/m2, decreased LAA filling flow velocity, greater 2D LAA 182 measurements (LAA ostium diameter, LAA area), increased 3D LAA measurements (LAAV, ostium area) 183 and non-chicken wing morphology (Table S3). On multivariable analysis, LAA filling flow velocity and 3D 184 ES LAAV were associated with LAA sludge/thrombus at admission (Table 3). 185 186 Clinical, Biological and Echocardiography Characteristics According to 3D ES LAAV 187 Receiver operating characteristic curve analysis revealed that the optimal cut-off for 3D ES LAAV to 188 discriminate patients at risk of death within 2 years was 9.3 mL (area under the curve 0.61±0.05; 95% 189 confidence interval 0.50–0.72) (Figure S1). 190 Patients with LAAV ≥9.3 mL were more frequently male, had a history of heart failure or renal 191 failure and were hospitalized for AF associated with heart failure. They had also higher values for BNP, 192 CRP and glycated hemoglobin A1c (Table S4). Patients with LAAV ≥9.3 mL had a lower median LVEF, 193 greater LA volumes, more severe grade of LAA sludge and/or thrombus, lower LAA emptying flow 194 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 10 velocity, and greater 2D LAA measurements (ostium diameter, area). In 3D TEE, patients with LAAV ≥9.3 195 mL had greater 3D LAA measurements (volume, ostium area) and no significant difference in LAA 196 morphology including number of lobes (Table S5). 197 198 Clinical, Biological and Echocardiography Characteristics According to AF Pattern 199 Patients with persistent AF had a higher body mass index, higher prevalence’s of hypertension and 200 diabetes, history of AF, heart failure, renal failure, hospitalization for AF and heart failure, a higher 201 CHA2DS2-VASc score, higher BNP, CRP, and HbA1c values, and lower glomerular filtration rate (Table 202 S6). 203 Echocardiographic analysis showed that patients with persistent AF had a lower LVEF, higher 204 LAV and pulmonary artery pressures, higher degree of LAA spontaneous echo contrast, higher rates of 205 LAA sludge and/or thrombus, lower LAA emptying and filling flow velocities, greater 2D LAA area and 206 ostium diameter, and greater 3D LAAV and ostium area (Table S7). 207 208 Clinical Events During 2 Years of Follow-Up 209 Over a mean ± standard deviation follow-up of 22.0±5.4 months, 56 (27.2%) patients had recurrent AF, 23 210 (11.2%) had heart failure, 5 (2.4%) myocardial infarction, 2 (1.0%) stroke, and 14 (6.8%) died (11 211 cardiovascular deaths). Kaplan–Meier curves demonstrated a significant difference in survival at 2 years 212 according to 3D ES LAAV: 3 deaths occurred in the group with 3D ES LAAV 9.3 mL and 11 in the group 213 with a volume 9.3 mL (P=0.02) (Figure 3). 214 On univariable analysis, female sex, history of heart failure, presence of heart failure at admission, 215 CRP >10 mg/L, low LVEF, high left atrial volume index, grade 2 spontaneous echo contrast, low LAA 216 filling and emptying velocities, and LAAV ≥9.3 mL were predictors of death (Table S8). On multivariable 217 analysis, CRP >10 mg/L was associated with death (Table 4). 218 219

220 The findings from our study suggest that increased 3D LAAV is associated with LAA prothrombotic 221 markers and risk of death. On multivariable analysis, LAA filling flow velocity and 3D ES LAAV were the 222 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 11 only determinants associated with LAA sludge/thrombus, independently of clinical, biological, and other 223 echocardiographic parameters. Kaplan–Meier curves demonstrated a significant difference in survival at 2 224 years when using 3D ES LAAV. 225 226 3D LAAV: A New Parameter to Evaluate Prothrombotic State in AF 227 Several parameters have been described using echocardiography to evaluate LAA prothrombotic state in 228 AF. In TEE, LA/LAA spontaneous echo contrast8 and LAA emptying or filling velocities9 are the most 229 commonly validated and used parameters. In the literature, 3D LAA morphology (chicken wing, windsock, 230 cactus, cauliflower) is also associated with thromboembolic history, with a higher thromboembolic risk for 231 non−chicken-wing morphology.13 However, LAA morphology is a subjective evaluation. In our study, 3D 232 ES LAAV was more accurately associated with thromboembolic state (on multivariable analysis) (Table 3) 233 when compared with 3D LAA morphology (on univariable analysis) (Table S3). Thus, we propose to 234 consider 3D LAAV as an additional pertinent parameter to evaluate the LAA thrombogenic state in AF. 235 236 LAA Remodeling in AF 237 In comparison with healthy individuals, patients with AF had higher LAAV.14 The increase in LAAV may 238 correlate with the degree with LAA dysfunction, and be an integral part of the atrial cardiomyopathy 239 concept, defined as “structural, architectural, contractile or electrophysiological changes affecting the atria 240 with the potential to produce clinically-relevant manifestations”.4 In a recent study, LAA fibrosis and 241 endothelial damage were associated with LAA thrombus and stroke, independently of clinical factors.26 It 242 is interesting to note the association between 3D LAAV and clinical patterns of AF; in our study, patients 243 with paroxysmal AF had a lower median LAAV compared to those with persistent AF (Table S7). These 244

are consistent with the literature: patients with paroxysmal AF had fewer thromboembolic events 245 and deaths compared to those with persistent or permanent AF.27 246 Recently reported data indicate that early rhythm control reduces cardiovascular death, stroke, 247 hospitalization for heart failure and acute coronary syndrome across AF clinical patterns.28 We 248 demonstrated that management of AF focused on restoration of sinus rhythm could induce anatomical 249 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 12 and/or functional cardiac cavity reverse remodeling.5 The integration of LAA remodeling could be a key 250 future step in understanding the pathophysiological processes involved in AF. 251 252 Clinical events 253 Over 2 years of follow-up, 14 deaths (11 cardiovascular) and 2 strokes occurred, highlighting the 254 effectiveness of AF management in the prevention of strokes. LAAV is associated with the prothrombotic 255 state (LAA sludge/thrombus) independently of clinical and biological factors. Furthermore, in 256 Kaplan−Meier analysis, 3D LAAV was associated with death in the univariable analysis. Thus, LAAV 257 appears to be a key parameter of atrial remodeling and could be integrated into the evaluation of 258 thromboembolic risk (Figure 4). 259 260

261 In cardiac imaging, the irregular rhythm in AF always presents a challenge for image acquisition and 262 quantitation. However, accurate 3D evaluation of LAAV was possible in 206 (94.0%) examinations. It 263 would have been interesting to use an additional imaging technique to measure LAAV (e.g., computed 264 tomography or cardiac magnetic resonance imaging); however, our study was performed as part of 265 routine care and did not benefit from examinations not indicated in the patient's usual management. In this 266 study, we described LAA morphology in 2 stages (chicken wing versus non-chicken wing). In our practice, 267 this 2-stage evaluation appears to reduce the subjectivity of LAA morphology evaluation in comparison 268 with the multiple morphology assessment (chicken wing, cauliflower, cactus, windsock). In addition to the 269 LAA morphology evaluation, we suggest that LAAV evaluation is a more objective parameter. In our study, 270 the reason for hospitalization was heart failure and AF in 35.4%, and we cannot exclude that the LAAV 271 could have been modified by changes in filling pressure in the heart failure and AF group. Finally, our 272 population included patients with paroxysmal or persistent AF; the absence of permanent AF does not 273 allow us to extend our results to this population. 274 275

276 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 13 3D evaluation of LAAV is a new tool for evaluation of thromboembolic state and prognosis in patients with 277 AF, which appears to be more accurate than LAA morphology for the evaluation of prothrombotic state. 278 LAAV could be integrated into the characterization of LA/LAA remodeling during AF, and into the concept 279 of atrial cardiomyopathy as a pertinent diagnostic tool. Further investigations are necessary to understand 280 the role of LAA remodeling in the thrombogenic function of LAA as well as in the long-term prognosis in 281 patients with AF. 282 283 Acknowledgments 284 Sophie Rushton-Smith, PhD (Medlink Healthcare Communications, London), provided editorial assistance 285 including editing, checking content and language and formatting, and was funded by the authors. 286 287 Sources of Funding 288 This work was partially funded by Bayer and the Fondation de France. Dr. Soulat-Dufour has received a 289 grant from Fédération Française de Cardiologie. 290 291 Disclosures 292 Dr Cohen reports research grants from RESICARD (research nurses) and the companies ARS, Bayer and 293 Boehringer-Ingelheim; and consultant and lecture fees from AstraZeneca, Bayer Pharma, BMS-Pfizer 294 Alliance, Boehringer-Ingelheim, Daiichi Sankyo and Novartis, unrelated to the present work. The other 295 authors declare that they have no competing interest. 296 297 Supplemental Material 298 Tables S1–S8 299 Figure S1 300

1–6 301 302 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 14 Clinical Perspective 303 What Is New? 304 Three-dimensional evaluation of left atrial appendage volume is a useful new tool for evaluating the 305 thromboembolic state and prognosis in patients in atrial fibrillation. 306 307 What Are the Clinical Implications? 308 Further investigations are needed to clarify the integration of left atrial appendage volume into the imaging 309 diagnostic criteria in atrial cardiomyopathy. 310 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 15

311 312 1. Ding WY, Gupta D and Lip GYH. Atrial fibrillation and the prothrombotic state: revisiting Virchow's 313 triad in 2020. Heart. 2020;106:1463-1468. 314 2. Joglar JA, Chung MK, Armbruster AL, Benjamin EJ, Chyou JY, Cronin EM, Deswal A, Eckhardt 315 LL, Goldberger ZD, Gopinathannair R, et al. 2023 ACC/AHA/ACCP/HRS Guideline for the Diagnosis and 316 Management of Atrial Fibrillation: A Report of the American College of Cardiology/American Heart 317 Association Joint Committee on Clinical Practice Guidelines. Circulation. 2024;149:e1-e156. 318 3. Donal E, Lip GY, Galderisi M, Goette A, Shah D, Marwan M, Lederlin M, Mondillo S, Edvardsen T, 319 Sitges M, et al. EACVI/EHRA Expert Consensus Document on the role of multi-modality imaging for the 320 evaluation of patients with atrial fibrillation. Eur Heart J Cardiovasc Imaging. 2016;17:355-383. 321 4. Goette A, Kalman JM, Aguinaga L, Akar J, Cabrera JA, Chen SA, Chugh SS, Corradi D, D'Avila 322 A, Dobrev D, et al. EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: 323 definition, characterization, and clinical implication. Europace. 2016;18:1455-1490. 324 5. Soulat-Dufour L, Lang S, Addetia K, Ederhy S, Adavane-Scheuble S, Chauvet-Droit M, Jean ML, 325 Nhan P, Ben Said R, Kamami I, et al. Restoring Sinus Rhythm Reverses Cardiac Remodeling and 326 Reduces Valvular Regurgitation in Patients With Atrial Fibrillation. J Am Coll Cardiol. 2022;79:951-961. 327 6. Ogata T, Matsuo R, Kiyuna F, Hata J, Ago T, Tsuboi Y, Kitazono T, Kamouchi M and Investigators 328 FSR. Left Atrial Size and Long-Term Risk of Recurrent Stroke After Acute Ischemic Stroke in Patients 329 With Nonvalvular Atrial Fibrillation. J Am Heart Assoc. 2017;6. 330 7. Zabalgoitia M, Halperin JL, Pearce LA, Blackshear JL, Asinger RW and Hart RG. 331 Transesophageal echocardiographic correlates of clinical risk of thromboembolism in nonvalvular atrial 332 fibrillation. Stroke Prevention in Atrial Fibrillation III Investigators. J Am Coll Cardiol. 1998;31:1622-1626. 333 8. Fatkin D, Kelly RP and Feneley MP. Relations between left atrial appendage blood flow velocity, 334 spontaneous echocardiographic contrast and thromboembolic risk in vivo. J Am Coll Cardiol. 335 1994;23:961-969. 336 9. Goldman ME, Pearce LA, Hart RG, Zabalgoitia M, Asinger RW, Safford R and Halperin JL. 337 Pathophysiologic correlates of thromboembolism in nonvalvular atrial fibrillation: I. Reduced flow velocity 338 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 16 in the left atrial appendage (The Stroke Prevention in Atrial Fibrillation [SPAF-III] study). J Am Soc 339 Echocardiogr. 1999;12:1080-1087. 340 10. Amarenco P, Cohen A, Tzourio C, Bertrand B, Hommel M, Besson G, Chauvel C, Touboul PJ and 341 Bousser MG. Atherosclerotic disease of the aortic arch and the risk of ischemic stroke. N Engl J Med. 342 1994;331:1474-1479. 343 11. Beinart R, Heist EK, Newell JB, Holmvang G, Ruskin JN and Mansour M. Left atrial appendage 344 dimensions predict the risk of stroke/TIA in patients with atrial fibrillation. J Cardiovasc Electrophysiol. 345 2011;22:10-15. 346 12. Burrell LD, Horne BD, Anderson JL, Muhlestein JB and Whisenant BK. Usefulness of left atrial 347 appendage volume as a predictor of embolic stroke in patients with atrial fibrillation. Am J Cardiol. 348 2013;112:1148-1152. 349 13. Di Biase L, Santangeli P, Anselmino M, Mohanty P, Salvetti I, Gili S, Horton R, Sanchez JE, Bai 350 R, Mohanty S, et al. Does the left atrial appendage morphology correlate with the risk of stroke in patients 351 with atrial fibrillation? Results from a multicenter study. J Am Coll Cardiol. 2012;60:531-538. 352 14. Meltzer SN, Phatak PM, Fazlalizadeh H, Chang I, Bering P, Kenigsberg B, Weissman G, Shah 353 MH, Satler LF, Rogers T, et al. Three-Dimensional Echocardiographic Left Atrial Appendage Volumetric 354 Analysis. J Am Soc Echocardiogr. 2021;34:987-995. 355 15. Yamamoto M, Seo Y, Kawamatsu N, Sato K, Sugano A, Machino-Ohtsuka T, Kawamura R, 356 Nakajima H, Igarashi M, Sekiguchi Y, et al. Complex left atrial appendage morphology and left atrial 357 appendage thrombus formation in patients with atrial fibrillation. Circ Cardiovasc Imaging. 2014;7:337-358 343. 359 16. Chen R, Wu X, Jin H, Wang B, Ma M and Zhao B. Assessment of Left Atrial Appendage 360 Morphology and Function in Patients with Non-valvular Paroxysmal Atrial Fibrillation with Different 361 Rhythms Using Real-Time 3D Transesophageal Echocardiography. Ultrasound Med Biol. 2016;42:118-362 124. 363 17. Hoit BD. Left atrial size and function: role in prognosis. J Am Coll Cardiol. 2014;63:493-505. 364 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 17 18. Soulat-Dufour L, Lang S, Ederhy S, Ancedy Y, Beraud AS, Adavane-Scheuble S, Chauvet-Droit 365 M, Hammoudi N, Scheuble A, Nhan P, et al. Biatrial remodelling in atrial fibrillation: A three-dimensional 366 and strain echocardiography insight. Arch Cardiovasc Dis. 2019;112:585-593. 367 19. Baumgartner H, Falk V, Bax JJ, De Bonis M, Hamm C, Holm PJ, Iung B, Lancellotti P, Lansac E, 368 Rodriguez Munoz D, et al. 2017 ESC/EACTS Guidelines for the management of valvular heart disease. 369 Eur Heart J. 2017;38:2739-2791. 370 20. Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP, 3rd, Gentile F, Jneid H, Krieger EV, 371 Mack M, McLeod C, et al. 2020 ACC/AHA Guideline for the Management of Patients With Valvular Heart 372 Disease: Executive Summary: A Report of the American College of Cardiology/American Heart 373 Association Joint Committee on Clinical Practice Guidelines. Circulation. 2021;143:e35-e71. 374 21. Zoghbi WA, Adams D, Bonow RO, Enriquez-Sarano M, Foster E, Grayburn PA, Hahn RT, Han Y, 375 Hung J, Lang RM, et al. Recommendations for Noninvasive Evaluation of Native Valvular Regurgitation: A 376 Report from the American Society of Echocardiography Developed in Collaboration with the Society for 377 Cardiovascular Magnetic Resonance. J Am Soc Echocardiogr. 2017;30:303-371. 378 22. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, Flachskampf FA, Foster E, 379 Goldstein SA, Kuznetsova T, et al. Recommendations for cardiac chamber quantification by 380 echocardiography in adults: an update from the American Society of Echocardiography and the European 381 Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015;28:1-39 e14. 382 23. Lang RM, Badano LP, Tsang W, Adams DH, Agricola E, Buck T, Faletra FF, Franke A, Hung J, de 383 Isla LP, et al. EAE/ASE recommendations for image acquisition and display using three-dimensional 384 echocardiography. Eur Heart J Cardiovasc Imaging. 2012;13:1-46. 385 24. Mueller C, McDonald K, de Boer RA, Maisel A, Cleland JGF, Kozhuharov N, Coats AJS, Metra M, 386 Mebazaa A, Ruschitzka F, et al. Heart Failure Association of the European Society of Cardiology practical 387 guidance on the use of natriuretic peptide concentrations. Eur J Heart Fail. 2019;21:715-731. 388 25. Ridker PM and Cook N. Clinical usefulness of very high and very low levels of C-reactive protein 389 across the full range of Framingham Risk Scores. Circulation. 2004;109:1955-1959. 390 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 18 26. Miyauchi S, Tokuyama T, Takahashi S, Hiyama T, Okubo Y, Okamura S, Miyamoto S, Oguri N, 391 Takasaki T, Katayama K, et al. Relationship Between Fibrosis, Endocardial Endothelial Damage, and 392 Thrombosis of Left Atrial Appendage in Atrial Fibrillation. JACC Clin Electrophysiol. 2023;9:1158-1168. 393 27. Link MS, Giugliano RP, Ruff CT, Scirica BM, Huikuri H, Oto A, Crompton AE, Murphy SA, Lanz H, 394 Mercuri MF, et al. Stroke and Mortality Risk in Patients With Various Patterns of Atrial Fibrillation: Results 395 From the ENGAGE AF-TIMI 48 Trial (Effective Anticoagulation With Factor Xa Next Generation in Atrial 396 Fibrillation-Thrombolysis in Myocardial Infarction 48). Circ Arrhythm Electrophysiol. 2017;10. 397 28. Goette A, Borof K, Breithardt G, Camm AJ, Crijns H, Kuck KH, Wegscheider K, Kirchhof P and 398 Investigators E-A. Presenting Pattern of Atrial Fibrillation and Outcomes of Early Rhythm Control Therapy. 399 J Am Coll Cardiol. 2022;80:283-295. 400 401 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 19 Table 1. Clinical and Biological Characteristics in the Global Population and According to the 402 presence of LAA Sludge and/or Thrombus at admission 403 Variable All (n=206) No LAA sludge and/or thrombus (n=171) LAA sludge and/or thrombus (n=35) P Clinical characteristics and cardiovascular risk factors Sex, male 129 (62.6) 105 (61.4) 24 (68.6) 0.43 Age, years 66.3±11.5 66.3±11.7 66.1±10.1 0.88 Body mass index, kg/m2 27.0 (23.6−30.9) 26.9 (23.2−31.0) 27.1 (23.9−29.5) 0.66 Hypertension 110 (53.4) 92 (53.8) 18 (51.4) 0.80 Diabetes 39 (18.9) 34 (19.9) 5 (14.3) 0.44 Hypercholesterolemia 69 (33.5) 57 (33.3) 12 (34.3) 0.91 Current smoker 37 (18.0) 32 (18.7) 5 (14.3) 0.53 Medical history AF 72 (35.0) 58 (33.9) 14 (40.0) 0.49 Heart failure 32 (15.5) 21 (12.3) 11 (31.4) 0.004 Stroke 18 (8.7) 14 (8.2) 4 (11.4) 0.54 Myocardial infarction 30 (14.6) 23 (13.5) 7 (20.0) 0.32 Renal failure 17 (8.3) 15 (8.8) 2 (5.7) 0.55 Reason for hospitalization AF without heart failure 133 (64.6) 117 (68.4) 16 (45.7) 0.011 AF with heart failure 73 (35.4) 54 (31.6) 19 (54.3) AF classification Paroxysmal 62 (30.1) 58 (33.9) 4 (11.4) 0.008 Persistent 144 (69.9) 113 (66.1) 31 (88.6) CHA2DS2-VASc score 0 15 (7.3) 13 (7.6) 2 (5.7) 0.12 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 20 Variable All (n=206) No LAA sludge and/or thrombus (n=171) LAA sludge and/or thrombus (n=35) P 1 35 (17.0) 33 (19.3) 2 (5.7) ≥2 156 (75.7) 125 (73.1) 31 (88.6) Biology B-type natriuretic peptide, pg/mL n=205 238 (123−378) n=170 203 (104−317) 446 (171−759) 0.0001 Troponin, ng/mL n=205 0.04 (0.04−0.04) n=170 0.04 (0.04−0.04) 0.04 (0.4−0.04) 0.77 D dimers, ng/mL n=195 337 (270−543) n=162 325 (270−551) n=33 424 (283−690) 0.18 Glomerular filtration rate*, mL/min/1.73 m2 75.7 (63.4−89.2) 75.5 (63.1−90.2) 76.0 (65.1−87.3) 0.99 C-reactive protein, mg/L 4.0 (3.0−10.0) 3.3 (3.0−7.2) 6.2 (3.0−15.0) 0.025 Low-density lipoprotein cholesterol, g/L n=204 1.04 (0.78−1.23) n=169 1.03 (0.78−1.20) 1.08 (0.76−1.32) 0.52 Glycated hemoglobin A1C, % n=194 5.8 (5.5−6.2) n=160 5.7 (5.5−6.1) n=34 6.1 (5.6−6.6) 0.08 Values presented as count (%) or median (interquartile range). 404 *Calculated with the Modification of Diet in Renal Disease equation. 405 AF indicates atrial fibrillation; and LAA, left atrial appendage. 406 407 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 21 Table 2. Echocardiography Characteristics in the Global Population and According to the 408 Presence or Absence of LAA Sludge and/or Thrombus at Admission 409 Variable All (n=206) No LAA sludge and/or thrombus (n=171) LAA sludge and/or thrombus (n=35) P 2D TTE LVEF*, % 52.00 (38.4−60.0) 54.9 (44.8−60.0) 37.0 (21.0−50.0) 0.0001 ES LAV, mL (n=204) 86.1 (69.2−104.5) 80.8 (68.0−99.8) 99.4 (83.6−115.0) 0.0005 ES LAVi, mL/m2 (n=204) 45.0 (36.2−53.9) 42.8 (35.3−51.4) 51.5 (45.1−56.4) 0.0004 PAP, mmHg (n=168) 29 (24−36) 29 (24−35) 28 (24−36) 0.97 2D TEE LAA emptying flow velocity, cm/s (n=199) 32 (23−49) 35 (26−52) 21 (17−28) 0.0001 LAA filling flow velocity, cm/s (n=199) 35 (24−50) 38 (29−53) 22 (19−31) 0.0001 ES LAA ostium diameter, mm 20.82 (17.78−23.65) 20.2 (17.5−23.6) 23.1 (20.9−25.5) 0.004 ES LAA area, cm2 5.98 (4.46−7.94) 5.9 (4.2−7.7) 7.2 (5.4−9.2) 0.02 ED LAA ostium diameter, mm 18.70 (15.10−22.50) 17.7 (14.7−22.0) 21.6 (18.3−24.4) 0.002 ED LAA area, cm2 4.95 (3.57−7.22) 4.7 (3.4−7.1) 6.4 (4.6−8.7) 0.002 Trabeculations (n=203) 0.16 Mild 102 (50.3) 89 (53.0) 13 (37.1) Moderate 81 (39.9) 62 (36.9) 19 (54.3) Severe 20 (9.9) 17 (10.1) 3 (8.6) Aortic atheroma (n=205) 80 (39.0) 67 (39.2) 13 (38.2) 0.92 3D TEE ES LAAV, mL 9.02 (6.96−12.15) 8.7 (6.8−11.9) 10.9 (8.7−14.8) 0.003 ES LAA ostium area, mm2 (n=205) 3.90 (3.13−4.98) 3.9 (3.1−4.9) 4.5 (3.2−6.1) 0.05 ED LAAV, mL 7.93 (5.98−10.87) 7.8 (5.6−10.6) 10.0 (7.6−13.0) 0.002 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 22 Variable All (n=206) No LAA sludge and/or thrombus (n=171) LAA sludge and/or thrombus (n=35) P ED LAA ostium area, mm2 (n=205) 3.48 (2.69−4.53) 3.4 (2.6−4.3) 4.1 (3.1−6.0) 0.003 LAA morphology (n=205) 0.007 Chicken wing 122 (59.5) 109 (63.7) 13 (38.2) Non-chicken wing 83 (40.5) 62 (36.3) 21 (61.8) Number of lobes (n=205) 0.56 1 94 (45.9) 79 (46.2) 15 (44.1) 2 78 (38.1) 62 (36.3) 16 (47.1) 3 25 (12.2) 23 (13.5) 2 (5.9) 4 8 (3.9) 7 (4.1) 1 (2.9) Values presented as count (%) or median (interquartile range). 410 *Calculated using the Simpson method. 411 AF indicates atrial fibrillation; ED, end diastolic; ES, end systolic; LAA, left atrial appendage; 412 LAAV, left atrial appendage volume; LVEF, left ventricular ejection fraction; TEE, transesophageal 413 echocardiography; and TTE, transthoracic echocardiography. 414 415 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 23 Table 3. Determinants of Prothrombotic State on Multivariable Logistic Regression Analysis 416 (n=198) 417 Variable Odds ratio (95% confidence interval) P value History of heart failure or acute heart failure (vs. no history) 2.07 (0.69−6.24) 0.20 Left ventricular ejection fraction, per 5% decrease 1.08 (0.91−1.28) 0.39 LAA filling flow velocity, per 5 cm/s decrease 1.65 (1.27−2.15) <0.0001 No chicken-wing morphology (vs. chicken-wing morphology) 1.97 (0.80−4.85) 0.14 3D ES LAAV, per 1 mL increase 1.10 (1.01−1.20) 0.044 ES indicates end systolic; LAA, left atrial appendage; and LAAV, left atrial appendage volume. 418 419 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 24 Table 4. Multivariable Cox Model for Prediction of Death at 2-year Follow-Up (n=199) 420 Variable Multivariable Cox model HR (95% CI) P Female sex (vs. male sex) 0.15 (0.02−1.17) 0.07 AF with heart failure (vs. no heart failure) 4.79 (0.86−26.72) 0.07 Renal failure (vs. no renal failure) 3.53 (0.64−19.38) 0.15 CRP >10 mg/L (vs. >10 mg/L) 3.43 (1.06−11.16) 0.04 LVEF, per 5 % decrease 1.16 (0.93−1.44) 0.18 LAA emptying flow velocity <25 cm/s (vs. ≥25 cm/s) 2.74 (0.86−8.72) 0.09 3D ES LAAV ≥9.3 mL (vs. <9.3 mL) 1.45 (0.38−5.59) 0.59 AF indicates atrial fibrillation; CI, confidence interval; CRP, C-reactive protein; ES, end systolic; HR, 421 hazard ratio; LAA, left atrial appendage; LAAV, left atrial appendage volume; and LVEF, left ventricular 422 ejection fraction. 423 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 25 424 Figure 1. Study flow chart. 425 3D indicates 3-dimensional; AF, atrial fibrillation; LAA, left atrial appendage; TEE, 426 transesophageal echocardiography; and TTE, transthoracic echocardiography. 427 428 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 26 429 Figure 2. Evaluation of LAA in 2D (A, B, C) and 3D (D, E, F) TEE. 430 A: examples of LA/LAA spontaneous echo contrast grade 0 (left), sludge (middle), thrombus 431 (right) according to Fatkin classification8; B: evaluation of LAA flow velocities; C: 2D measurements of the 432 LAA area at 92° (end-systolic left atrial appendage area and ostium diameter). Examples of different LAA 433 morphologies evaluated with Zoom 3D mode 1 beat; D: left atrial appendage with chicken-wing 434 morphology and a volume of 9.76 mL; E: left atrial appendage with no chicken-wing morphology and a 435 volume of 22.61 mL; and F: left atrial appendage with 2 lobes and a volume of 7.71 mL. 436 2D indicates 2-dimensional; 3D, 3-dimensional; AF, atrial fibrillation; LA, left atrial; LAA, left atrial 437 appendage; and TEE, transesophageal echocardiography. 438 439 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 27 440 Figure 3. Kaplan–Meier Survival Estimates According to 3D ES LAA Volume at 2 Years. 441 Kaplan–Meier curves showing event-free survival according to the 3D ES LAA Volume 9.3 mL (blue) 442 or 9.3 mL (red). 443 3D indicates 3-dimensional; ES, end-systolic; and LAA, left atrial appendage. 444 445 446 447 448 449 450 451 452 453 454 455 456 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint 28 457 Figure 4. Role of 3D of LAA in AF: Association with Prothrombotic State and Prognosis. 458 3D indicates 3-dimensional; ES, end-systolic; LAA, left atrial appendage; and OR, odds ratio. 459 460 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.09.24301079doi: medRxiv preprint