In patients with mitral stenosis(MS) the predilection of the left atrial appendage(LAA) for thrombus formation has been known, while the characteristics and clinical implications of LAA flow have not been clearly analyzed. This prospective study with transesophageal echocardiography(TEE) was done to compare the LAA flow velocities of normal controls and patients with tight MS, to evaluate the correlation between LAA flow and hemodynamic indices and to observe the effects of successful percutaneous mitral valvuloplasty(PMV) on LAA flow.

TEE was performed in 12 normal controls and 50 patients with tight MS using a 5.0 MHz biplane transducer(Hewlett Packard SONOS 1000). Left atrial spontaneous echo contrast(SEC) was semiquantified(Grade 0-III) according to the relative extent of swirling echo movement in left atrial cavity : Grade I denotes SEC confined to LAA and swirling movement over the half of left atrial cavity was graded as III. In patients with MS. TEE was performed one day before and after PMV.Hemodynamic data(left atrial pressure, transmitral diastolic pressure gradient, cardiac output and mitral valve area) were obtained just before and after PMV.

1) In normal controls LAA flow recording showed two pairs of LAA ejection(positive) and filling(negative) waves. One pair of LAA ejection(appendage peak positive flow ; APF) and filling flow(appendage peak negative flow ; ANF) occurred 143±22msec and 289±33msec respectively after the onset of the ECG P wave. In early diastole another small pair of LAA ejection(ACF) and filling(ADF) occurred 550±21msec and 671±50msec respectively after the onset of ECG QRS wave. Mean absolute velocities of APF and ANF were 54±21 and 54±22cm/sec respectively, which were significantly higher than those of ACF(16±6cm/sec) and ADF(16±5cm/sec). 2) In patients with tight MS(mitral valve area of 0.9±0.3cm²) mean velocities of APF(20±12cm/sec) and ANF(23±16cm/sec) were significangly decreased compared with normal controls. There was no significant correlation between hemodynamic indices and absolute velocities of LAA flow. SEC was observed in 62%(31/50) and mean velocities of APF(10±8cm/sec) and ANF(12±11cm/sec) were significantly lower in patients with SEC than those values(25±19, 27±21cm/sec) in patients without SEC. There was strong negative correlation between the absolute values of APF and ANF and the grade of SEC(r_APF=-0.75, pAPF=0.00 ; rANF=-0.70, pANF=0.00). 3) After successful PMV(mitral valve area of 2.0±0.4cm²), APF increased from 20±12cm/sec to 36±24cm/sec and statistically significant increase of ANF was also observed(23±16cm/sec vs 36±22cm/sec, p<0.05). In patients with sinus rhythm, PMV normalized LAA flow(APF=52±14cm/sec, ANF=51±10cm/sec) while still decreased flow velocities were recorded in patients with atrial fibrillation(APF=15±4cm/sec, ANF=16±6cm/sec).

Recording of LAA flow with TEE is an indicator of LAA function and risk of thrombus formation and cardiogenic embolism, which may not be obtained with conventional hemodynamic indices. Improvement of LAA flow and normalization of flow velocities in patients with sinus rhythm immediately after successful PMV raises a question of role of PMV in prevention of thromboembolism. Randomized sudies of the long-term effects of PMV in MS will be required before this important question can be answered.