Left atrial thrombus and spontaneous echo-contrast in nonanticoagulated mitral stenosis. A transesophageal echocardiographic study.

OBJECTIVE: The aim of the study was to investigate if evidence at transthoracic echocardiography (TTE) of left atrial (LA) thrombus and LA spontaneous echo-contrast (LA SEC), which are potential precursors of embolization, can be predicted by clinical and TTE variables in nonanticoagulated mitral valve stenosis (MS).

DESIGN: Clinical (age, NYHA class, rhythm, previous embolization) and TTE variables were related to transesophageal echocardiography (TEE) evidence of LA thrombus and/or LA SEC.

SETTING: Nonanticoagulated MS was the setting.

PATIENTS: Fifty-nine patients had MS, and they were not receiving anticoagulant or antiplatelet therapy (24 in sinus rhythm and 35 in atrial fibrillation). Previous arterial embolization had occurred in 12 patients (20.3 percent).

MEASUREMENTS: The following TTE variables were analyzed: mitral orifice area (pressure half-time method), mitral gradient (Bernouilli's equation), LA end-systolic area, and mitral regurgitation (color Doppler grading). LA thrombus and LA SEC were analyzed by monoplane TEE.

RESULTS: LA thrombus was found by TEE in 12 patients (20.3 percent). Of these 12, 11 (91.6 percent) were in atrial fibrillation. LA SEC was found by TTE in 2 patients (3.5 percent) and by TEE in 40 (67.8 percent) (p < 0.001). Previous embolization had occurred only in patients with LA SEC, of whom 5 had and 7 did not have LA thrombus. Patients with LA SEC, compared with those without LA SEC, were characterized by more frequent advanced NYHA class, atrial fibrillation, smaller mitral valve area, and larger LA size. By multivariate regression analysis, atrial fibrillation and LA end-systolic area were factors related to both LA thrombus and LA SEC, whereas mitral area was related only to LA SEC. However, whereas LA SEC was accurately predicted by the presence of atrial fibrillation (sensitivity: 87.5 percent; specificity: 100 percent) and a LA area > or = 30 cm2 (sensitivity: 72.5 percent; specificity: 89.5 percent), among patients with LA SEC no clinical or TTE variable accurately identified those with actual LA thrombus.

CONCLUSIONS: TEE is not necessary in many patients with MS in order to recognize LA SEC. However, when actual LA thrombus detection is necessary for clinical decision making, TEE should be performed.