Pathophysiologic correlates of thromboembolism in nonvalvular atrial fibrillation: I. Reduced flow velocity in the left atrial appendage (The Stroke Prevention in Atrial Fibrillation [SPAF-III] study)

Citation
Me. Goldman et al., Pathophysiologic correlates of thromboembolism in nonvalvular atrial fibrillation: I. Reduced flow velocity in the left atrial appendage (The Stroke Prevention in Atrial Fibrillation [SPAF-III] study), J AM S ECHO, 12(12), 1999, pp. 1080-1087
Citations number
22
Categorie Soggetti
Cardiovascular & Respiratory Systems
Journal title
JOURNAL OF THE AMERICAN SOCIETY OF ECHOCARDIOGRAPHY
ISSN journal
08947317 → ACNP
Volume
12
Issue
12
Year of publication
1999
Pages
1080 - 1087
Database
ISI
SICI code
0894-7317(199912)12:12<1080:PCOTIN>2.0.ZU;2-M
Abstract
Stroke associated with atrial fibrillation (AF) is mainly due to embolism o f thrombus formed during stasis of blood in the left atrial appendage (LAA) . Pathophysiologic correlates of appendage flow velocity as assessed by tra nsesophageal echocardiography (TEE) in patients with AF have not been defin ed. To evaluate the hypothesis that reduced velocity is associated with spo ntaneous echocardiographic contrast and thrombus in the LAA and with clinic al embolic events, we measured LAA flow velocity by TEE in 721 patients wit h nonvalvular AF entering the Stroke Prevention in Atrial Fibrillation (SPA F-III) study. Patient features, TEE findings, and subsequent cardioembolic events were correlated with velocity by multivariate analysis. Patients in AF during TEE displayed lower peak antegrade (emptying) flow velocity (A nu (p)) than those with intermittent AF in sinus rhythm during TEE (33 cm/s vs 61 cm/s, respectively, P < .0001). A nu(p) < 20 cm/s was associated with d ense spontaneous echocardiographic contrast (P < .001), appendage thrombus (P < .01); and subsequent cardioembolic events (P < .01). Independent predi ctors of A nu(p) < 20 cm/s included age (P = .009), systolic blood pressure (P < .001), sustained AF (P = .01), ischemic heart disease (P = .01), and left atrial area (P = .04). Multivariate analysis found both A nu(p) <20 cm /s (relative risk 2.6, P = .02) and clinical risk factors (relative risk 3. 3, P = .002) independently associated with LAA thrombus. LAA A nu(p) is red uced in AF and associated with spontaneous echocardiographic contrast, appe ndage thrombus, and cardioembolic stroke. Systolic hypertension and aortic atherosclerosis, independent clinical predictors of stroke in patients with AF, also correlated with LAA A nu(p). Our results support the role of redu ced LAA A nu(p) in the generation of stasis, thrombus formation, and emboli sm in patients with AF, although other mechanisms also contribute to stroke .