Complex Left Atrial Appendage Morphology and Left Atrial Appendage Thrombus Formation in Patients With Atrial Fibrillation

Malignant left atrial appendage morphology and embolization risk in atrial fibrillation

Context: In patients with rheumatic heart disease (RHD), left atrial appendage (LAA) dilation and thrombus formation is widely known. LAA thrombus formation is a risk factor for stroke even in patients with sinus rhythm. The aim of this study was to determine an association between LAA structure and function with respect to thrombus formation and differences in LAA size and velocity between patients with sinus rhythm and those with atrial fibrillation (AF). Materials and Methods: We prospectively studied LAA structure and function in 120 patients with RHD by transthoracic echocardiography and/or transesophageal echocardiography by measuring left atrial (LA) dimension, LA area, LAAmax/min, LAA ejection fraction (EF), and LAA emptying velocity. Results: Four out of 48 patients with sinus rhythm had LAA thrombus. In 72 patients with AF, 32 had LAA thrombus. Patients with LAA thrombus had lower mean LAA EF and emptying velocity both variables P-value is same (P<0.0001 and P<0.0001) Patients with LAA thrombus had increased mean LAA max and LAA min as compared to LAA max and LAA min in patients without LAA thrombus (P 0.008 for mean LAA max and P<0.0001 for LAA min respectively). Patients with AF with LAA thrombus had greater LAAmax compared to that in patients with AF without LAA thrombus (P < 0.0001). Doppler demonstrated a recognizable sawtooth LAA outflow velocity pattern in 32 of 36 (88.9%) patients with LAA thrombus versus 32 of 84 (38.1%) patients without LAA thrombus. Conclusions: We conclude that LAA contractility is reduced in RHD with LAA thrombus, and loss of both contractility and LAA dilation is associated with increased risk of thrombus formation and hence the risk of stroke.

Circulating Galectin-3 is Associated With Left Atrial Appendage Remodelling and Thrombus Formation in Patients With Atrial Fibrillation

We investigated the characteristics of patients with non-valvular atrial fibrillation (NVAF) and left atrial (LA) appendage (LAA) thrombus who had been given appropriate oral anticoagulation therapy. We studied 737 NVAF patients who were scheduled for catheter ablation or electrical cardioversion. All patients received appropriate oral anticoagulation therapy for at least 3weeks prior to echocardiography in accordance with the guidelines. Whether LAA thrombus was present or absent on transesophageal echocardiography (TEE) was determined by at least three senior echocardiologists. LAA thrombi were observed in 22 patients (3.0%). Multivariate logistic regression analysis showed that LAA flow and LA volume index were both independent predictors of LAA thrombus formation; however, LAA flow (≤ 18cm/s) was indicated as a more powerful predictor. Moreover, the prevalence of LAA thrombus formation in patients with NVAF without LA enlargement (LA volume index ≤ 34mL/m2) was extremely rare (0.4%). LAA thrombus formation in patients with a mildly dilated LA volume index of 34-49.9mL/m2 and paroxysmal AF was also extremely rare (0.0%). LAA flow is strongly associated with LAA thrombus formation, even in NVAF patients treated with appropriate oral anticoagulation therapy. Augmented oral anticoagulation therapy or transcatheter or surgical LAA closure should be considered for such patients, especially for those with an LAA flow < 18cm/s. Furthermore, TEE for evaluating LAA thrombus before catheter ablation or electrical cardioversion may be unnecessary for NVAF patients who are undergoing appropriate oral anticoagulation therapy, depending on LA size.

Background: It is well known that the left atrial appendage (LAA) dysfunction plays an important role in the occurrence of the cardioembolic stroke. It has been reported that the atrium is the main source of brain natriuretic peptide (BNP) in atrial fibrillation patients without heart failure. Therefore, we hypothesized that serum BNP levels are sensitive predictor for LAA dysfunction and thrombus formation in patients with acute ischemic stroke. Methods: We performed trans-thoracic and esophageal echocardiography within 7 days after the onset (male, n=64, 69±11 years old; female, n=38, 68±15 years old) and measured serum hemostatic markers and BNP in 103 patients with ischemic stroke. No patients had history of heart failure. Results: LAA thrombus was detected in 8 patients (36%) among 22 patients with atrial fibrillation. Even in 3 patients (4%) among 80 patients with sinus rhythm, LAA thrombus was detected. There were no significant differences in left ventricular end-diastolic dimension (LVDd), and the ratio of the early diastolic trans-mitral inflow velocity (E) to mitral annular velocity (E′) between patients with and without LAA thrombus (LVDd, 49±6 vs. 47±5 mm; E/E′, 8.6±1.3 vs. 7.2±0.9). Serum BNP levels were markedly higher in patients with LAA thrombus formation than in those without it (191 (128 –278) vs. 39 (13–100) pg/ml, P&lt;0.01). Furthermore, in patients with sinus rhythm (n=80), the BNP levels were significantly correlated with LAA peak emptying flow velocity (R=−0.34, P&lt;0.01). The multivariate logistic regression analysis revealed that BNP &gt;120 was an independent predictor of LAA thrombus formation (odds ratio, 13.50; 95% confidence interval, 2.482–73.40; P=0.0026). Conclusions: The BNP levels were correlated with LAA peak emptying flow velocity. The elevated BNP level could be used as a non-invasive surrogate maker to predict LAA dysfunction and thrombus formation in patients with acute ischemic stroke.

Authors' Reply

Background: Left atrial (LA) size is well known to reflect volume overload condition such as mitral regurgitation (MR). However, the effects to left atrial appendage (LAA) have not been well defined. LAA structure has multilobed complicated three-dimensional structures. Recent development of real-time 3-dimensional transesophageal echocardiography (RT3D-TEE) has enabled us to examine morphology and determine absolute volumes of LAA without geometrical assumptions. We investigated the relationship between MR severity and LAA volume. Methods: Seventy-four patients with MR who underwent RT3D-TEE were studied. All measurements were performed off-line software. Maximal LAA orifice area, depth and volume, and the number of LAA lobes were assessed (Figure). We also measured LAA emptying flow velocity (LAAev) as the average of 3 consecutive cardiac cycles in patients in normal sinus rhythm and 5 consecutive cardiac cycles in patients with atrial fibrillation. MR grade was evaluated by vena contracta width (VCW). In 58 patients, we also determined the maximal LA volume by the Simpson method from apical views by 2-dimensional transthoracic echocardiography. Pulsed Doppler E, tissue Doppler E' and E/E' were also measured. Results: Mean LAA orifice area was 5.0±3.8 cm 2 , LAA volume was 11±10 ml, LA volume was 116±57ml, and LAAev was 41±23 cm/sec. Patients with moderate to severe MR had larger LAA orifice area, LAA volume, and LA volume than patients with mild MR (Table). In contrast, there were no relationship between MR severity, the number of LAA lobes and LAAev. 3D derived LAA volume was correlated with VCW significantly (r=0.51, p&lt;0.01). Conclusion: LAA volume was associated with MR severity.

Comparison of Transesophageal Echocardiography Versus Computed Tomography for Detection of Left Atrial Appendage Filling Defect (Thrombus)

Background: Establishment of optimal risk stratification is the best hope for decreasing the burden of atrial fibrillation (AF)-related thromboembolism. Left atrial appendage (LAA) thrombus is a high risk factor of embolism even in AF with warfarin. LAA contractility evaluated by LAA emptying function (EF), strain and radial velocity (RV) of LAA wall motion could contribute to blood stasis and be an important determinant of thrombus in chronic AF, especially in patients with low CHADS2 or CHA2DS2-VASc score. Methods: Consecutive 260 chronic AF patients with warfarin were examined and divided into 2 groups according to the presence of thrombus; AF with thrombus (n=43, 69±10) and AF without thrombus (n=217, age 66±11). Each group was divided into subgroup according to CHADS2 or CHA2DS2-VASc score≤2 (low CHADS2 or CHA2DS2-VASc subgroup). LAA EF, peak strain and peak RV of LAA wall were measured using speckle tracking echocardiography. LAA volume, spontaneous echo contrast (SEC) and LAA flow velocity were measured. Results: Seventy one % of patients with thrombus and 61% without thrombus met the target INR recommended by Japanese Circulation Society. AF with thrombus had more increased CHADS2 and CHA2DS2-VASc score and reduced LAA EF and strain compared to AF without thrombus (Table). LAA EF and strain were also reduced in low CHADS2 and CHA2DS2-VASc subgroup with thrombus. In multivariate analysis, LAA EF and strain were independent determinants of LAA thrombus. Using 20% of LAA EF as a cutoff, sensitivity and specificity for thrombus were 84% and 90%, using 5 of LAA peak strain, those values were 88% and 76%. Conclusion: LAA EF and peak strain were more useful predictors of thrombus in chronic AF than LAA flow velocity, SEC and CHADS2 and CHA2DS2-VASc score. LAA thrombus formation may depend on LAA contractility. Thus, AF patients with reduced LAA contractility should have a strict anticoagulant therapy regardless of CHADS2 or CHA2DS2-VASc score.

We previously reported a higher left atrial volume index (LAVI) was independently associated with left atrial (LA) appendage (LAA) thrombus formation in 737 patients with non-valvular atrial fibrillation (NVAF) receiving appropriate oral anticoagulation therapy. Since our previous study was a retrospective single-center study, we designed and conducted a prospective multi-center study to verify our findings for LAVI as a predictor of LAA thrombus in patients with NVAF receiving appropriate oral anticoagulation therapy. This prospective multi-center study comprised 746 consecutive patients with NVAF recruited between December 2021 and March 2023 from eight institutions in Japan, who were receiving appropriate oral anticoagulation therapy, had undergone transthoracic echocardiography and transesophageal echocardiography (TEE). LAA thrombi were observed in 21 patients (2.8%). The prevalence of LAA thrombus formation in patients with paroxysmal AF (PAF) was significantly lower than that in patients with non-PAF (0.7% vs. 4.1%, p=.006). LAA thrombus formation was detected in none (0/171) of the patients with normal size LA (LAVI≤34mL/m2 ). The prevalence of LAA thrombus formation in patients with mildly dilated LA (LAVI: 34-49.9mL/m2 ) was 2.1% (6/283), but that in PAF patients was low at 1.0% (1/104). Furthermore, this prevalence in patients with severely dilated LA (LAVI ≥ 50mL/m2 ) was high at 5.1% (15/292). The findings of this prospective multi-center study are consistent with those of our previous study. Thus, the need for TEE prior to catheter ablation or electrical cardioversion can be determined by the level of LAVI.

Background: Clinical and echocardiographic predictors for detecting left atrial appendage (LAA) thrombus are well characterized. However, whether the size of LAA measured by two-dimensional transesophageal echocardiography (2D-TEE) predicts the risk of thrombus formation in LAA has been the subject of much controversy. Purpose: The aim of this study was to investigate the ability of LAA volume derived from real-time three-dimensional transesophageal echocardiography (RT3D-TEE) to predict LAA thrombus and/or sludge in patients with nonvalvular atrial fibrillation (NVAF). Methods and Results: A total of 71 patients with NVAF lasting ≥ 1 month and without severe valvular regurgitation, who underwent 2D-TEE and RT3D-TEE, were included in the study. They were divided into two groups according to the presence of LAA thrombus and/or sludge. LAA thrombus and/or sludge were detected in 28 patients (39%). In these patients, the prevalence of LAA thrombus and that of LAA sludge were 54% and 100%, respectively. All 2D and 3D measurements of LAA were performed at ventricular end-systole.A trend was only seen for a significant correlation between LAA volume measured by RT3D-TEE and LAA area in the longitudinal section imaged (usually between 60 and 90 dgrees) by 2D-TEE (r = 0.31, p = 0.08). On multivariable analysis, LAA volume and the ratio of early transmitral flow velocity to peak early diastolic mitral annular velocity (E/e') were found to be significantly predictive of LAA thrombus and/or sludge. Conversely, LAA emptying flow velocity predicted freedom from them (Table 1). Receiver operating characteristic analysis yielded an optimal cutoff value of 14.7 cm 3 for LAA volume to detect LAA thrombus and/or sludge (area under the curve 0.89). The sensitivity and specificity of this model were 79% and 89%, respectively. Conclusion: Our data suggest that increased LAA volume measurements by RT3D-TEE can assist in the detection of LAA thrombus and/or sludge in the patients with NVAF.

The predilection of the left atrial appendage (LAA) for thrombus formation has long been known. We prospectively studied the two-dimensional echocardiographic and Doppler patterns of LAA function in 82 patients by transesophageal echocardiography. In the 63 patients in sinus rhythm, LAA area was measured during LAA diastole at the onset of the electrocardiographic (ECG) P wave (LAAmax) and after LAA systole at the ECG R wave (LAAmin) and LAA ejection fraction was calculated as (LAAmax-LAAmin)/LAAmax; peak Doppler velocity was recorded from the LAA outlet. The 58 patients in sinus rhythm without LAA thrombus were grouped according to left atrial size on transthoracic echocardiography; 39 patients had a left atrial size of less than 40 mm (group 1) and 19 had a left atrial size of 40 mm or greater (group 2). Five patients in sinus rhythm had LAA thrombus. In the 19 patients with atrial fibrillation or flutter LAAmax was measured independent of the ECG; three of these patients had LAA spontaneous contrast, four had thrombus, and one had both. Patients in sinus rhythm without LAA thrombus demonstrated a characteristic pattern of a contractile LAA apex and a noncontractile base with color flow and pulsed Doppler evidence of LAA emptying that coincided with the P wave. Patients in sinus rhythm with LAA thrombus had a mean +/- SD LAAmax (8.0 +/- 1.5 cm2) larger than that in group 1 (5.0 +/- 1.9 cm2) (p less than 0.01) but not group 2 (6.7 +/- 3.1 cm2), LAAmin (6.5 +/- 1.0 cm2) larger than that in both group 1 (2.3 +/- 1.5 cm2) and group 2 (4.2 +/- 2.7 cm2) (p less than 0.01), and LAA ejection fraction (17 +/- 11%) and LAA velocity (0.24 +/- 0.10 m/sec) less than those in both group 1 (55 +/- 21% and 0.48 +/- 0.24 m/sec, respectively) and group 2 (45 +/- 27% and 0.46 +/- 0.24 m/sec, respectively) (p less than 0.01). Patients with atrial fibrillation or flutter with LAA spontaneous contrast and/or thrombus had LAAmax (10.4 +/- 6.6 cm2) greater than that in patients with atrial fibrillation or flutter without LAA contrast and/or thrombus (6.8 +/- 3.0 cm2) (p less than 0.05). The LAA appeared as a static pouch in seven of eight of the former compared with in two of 11 of the latter. When attempted, Doppler demonstrated a recognizable fibrillatory LAA outflow velocity pattern in none of three in the former versus four of seven in the latter group. We conclude that the LAA has a characteristic pattern of emptying in sinus rhythm. LAA thrombus formation in sinus rhythm and atrial fibrillation is associated with both poor LAA contraction and LAA dilation.

Objective To investigate the application value of the transesophageal echocardiography (TEE) in patients with nonvalvular atrial fibrillation about the size, lobes, morphology and function of left atrial appendage (LAA). Methods One hundred and forty-two patients underwent TEE were divided into nonvalvular atrial fibrillation group (98 cases) and non atrial fibrillation group (44 cases). The orifice diameter, depth, volume, peak emptying velocity (PEV) of the LAA and the 1eft atrial dimension (LAD) were measured. The form and lobes of LAA, thrombus and spontaneous echo contrast (SEC) in LAA were observed. Results The LAA orifice diameter, depth, volume and LAD of patients with atrial fibrillation were significantly higher than those in the group without atrial fibrillation, which showed statistical significance (P<0.05). Forty-one cases in atrial fibrillation group were found with the SEC, and the number with thrombus in LAA was 22. The differences of PEV between chicken wings and non-chicken wings were statistically significant (P<0.05). The SEC in LAA and the lobe number of LAA had no relevance. Conclusions It was reliable to analyze the size, morphologies, lobes and hemodynamic parameters of LAA in patients with atrial fibrillation by TEE, which provided reference for percutaneous LAA occlusion and anticoagulation therapy for the patients with atrial fibrillation. Key words: Echocardiography, transesophageal; Atrial fibrillation; Left atrial appendage

It is well known that transesophageal echocardiography (TEE) is useful for evaluating left atrial appendage (LAA) function, but TEE is a semi-invasive procedure. Many clinical studies have shown a close relationship between LAA thrombus formation and left atrial (LA) mechanical remodeling. Left atrial (LA) strain assessed by two-dimensional speckle tracking echocardiography is a new tool to evaluate LA function. However, the association of LA strain and LAA function remains to be fully determined. We performed transthoracic and transesophageal echocardiography in 58 consecutive patients with acute ischemic stroke within 7 days of onset and sinus rhythm at TEE performed (40 males, mean age 71±11 years). Longitudinal LA strain was obtained using two-dimensional speckle tracking imaging at each LA segment, and peak systolic strain (S-LAs) was calculated by averaging the results for each segment. Eight patients had LAA thrombus and/or severe LAA spontaneous echo contrast (LAA-SEC). S-LAs was significantly lower in patients with LAA dysfunction than in those without (34.5±12.5 vs. 15.9±7.2%, P<0.01). S-LAs was significantly correlated with LAA eV (r=0.466, p<0.01). The optimal cut off value of S-LAs for predicting LAA dysfunction was determined for 21% (Figure). In patients with LAA dysfunction, 7 patients had low S-LAs. In those 6 patients (86%) had history of paroxysmal atrial fibrillation before stroke onset. Multivariate logistic regression analysis showed that S-LAs was an independent predictor of LAA thrombus and/or LAA-SEC (odds ratio 1.758, p<0.01). In conclusion, decreased S-LAs was independently associated with LAA dysfunction and thrombus formation in patients with acute ischemic stroke and sinus rhythm at echocardiography ![Figure][1] Figure 1. Receiver operating characteristics (ROC) curve analysis of LA peak systolic strain for predicting LAA dysfunction. [1]: pending:yes

PurposeWe aimed to correlate left atrial appendage (LAA) morphology with thrombus formation in patients with severe aortic valve stenosis and atrial fibrillation.MethodsWe analyzed LAA morphology and the prevalence of a thrombus in 231 patients with atrial fibrillation and severe aortic valve stenosis that were referred for pre-interventional CT scan before trans-catheter aortic valve implantation (TAVI) between 2016 and 2018. In addition, we documented neuro-embolic events depending on the presence of LAA thrombus within a follow-up of 18 months.ResultsThe overall distribution of different LAA morphologies was chicken-wing 25.5%, windsock 51.5%, cactus 15.6% and cauliflower 7.4%. Compared to chicken-wing morphology, patients with non-chicken-wing morphology showed a significantly higher thrombus rate (OR: 2.48, 95%; CI 1.05 to 5.86, p = 0.043). Within the 50 patients with a LAA thrombus, we observed chicken-wing (14.0%), windsock (62.0%), cactus (16.0%) and cauliflower (8.0%) configuration. In patients with LAA thrombus those with chicken-wing configuration have a higher risk (42.9%) to develop neuro-embolic events compared to non-chicken-wing configuration (20.9%).ConclusionWe found a lower LAA thrombus rate in patients with chicken-wing morphology compared to patients with non-chicken-wing configuration. However, in the presence of thrombus, those patients with chicken-wing morphology showed a doubled risk for neuro-embolic events compared to patients with non-chicken-wing morphology.These results must be confirmed in larger trials but underline the importance of LAA evaluation in thoracic CT scans and could have an impact on the anticoagulation management.