- Email: [email protected]
Left Atrial Appendage Contractile Function in Atrial Fibrillation
Left Atrial Appendage Contractile Function in Atrial Fibrillation* Influence of Heart Rate and Cardioversion to Sinus Rhythm Kwame 0. Akosah, MD; john T. Funai, MD; Thomas R. Porter, MD; Robert L. jesse, MD, PhD; and Pramod K. Mohanty, MD Background: A high incidence of embolic phenomena is associated with atrial fibrillation (AF) and the left atrial appendage (LAA) is frequently the source of the emboli. Thrombus formation may be due to stasis within the fibrillating and inadequately emptying LAA. Because LAA emptying in AF may be the result of mechanical compression by the adjacent left ventricle, it is possible that left ventricular diastolic filling duration will importantly influence passive emptying of the LAA. We hypothesized that the magnitude of emptying of the LAA in AF is related to the duration of left ventricular diastolic filling which is determined by the ventricular response rate in AF. Objective: The objective of our study was to determine the relationship of ventricular response rate in AF to LAA emptying and to assess the influence of sinus rhythm and heart rate on LAA emptying immediately after direct current cardioversion to sinus rhythm. Methods: To study this, we used transesophageal echocardiography to measure LAA ejection fraction ([LAAmax-LAAmin ]/LAAmaxX l 00%) and evaluated its relationship to left ventricular response rate (VRR) in 26 patients with AF (mean age, 65 ± 7 [l SD] years). Results: There was a strong inverse relationship between LAA ejection fraction and VRR in AF (r= -0.73; p
26± 10%, and immediately after successful cardioversion, it increased to 46± 12% (p<0.001). However, during sinus rhythm there was no relationship between LAA ejection fraction and VRR (r=0.06; p=NS) in the subgroup of patients who were successfully converted to sinus rhythm. There were poor relationships between LAA ejection fraction and peak transmitral flow velocity (r=-0.41; p=NS) or pulmonary venous flow velocity (r=-0.03; p=NS) in AF. Conclusion: These results indicate that the magnitude of LAA emptying in AF is strongly and inversely influenced by ventricular rate. Direct current cardioversion to sinus rhythm is associated with an increase in the magnitude of LAA emptying that is not influenced by heart rate. The magnitude of LAA emptying may be an important factor in the formation of thromboemboli in AF. The extent to which controlling the VRR in chronic AF will prevent stasis and LAA thrombus formation remains to be determined. (Chest 1995; 107:690-96)
The association of systemic emboli with atrial fibrillation (AF) and the formation of intracardiac thrombi in the left atrial appendage (LAA) are well established. The exact mechanism, however, for the preponderance of thrombi in the LAA during AF has not been defined. It is believed that attenuation of LAA contractile activity and stagnation of blood flow within the appendage are important factors. Several observers have proposed 1•2 that the magnitude of LAA emptying during AF is determined, in part, by extrinsic compression by the adjacent left
ventricular lateral free wall during diastolic ventricular filling. Since ventricular diastolic chamber volume is strongly influenced by heart rate and diastolic filling duration, we speculated that heart rate may affect the magnitude of LAA emptying. The magnitude of LAA emptying may then be a function of the ventricular response rate (VRR) in AF . We reasoned that lower ventricular response rates in atrial fibrillation would be expected to increase the ventricular diastolic filling duration, increase ventricular diastolic volume, increase the amplitude of compression of the LAA, and result in greater LAA emptying. On this basis, we hypothesized that the amplitude of LAA emptying is related to VRR during AF. To test this hypothesis, we evaluated therelationship between LAA emptying and VRR in patients with chronic AF . Asecond objective of this study was to examine the effect of sinus rhythm and heart rate on LAA emptying immediately after
*From the Hunter Holmes McGuire Veterans Affairs Medical Center, Medical College of Virginia/ Virginia Commonwealth University, Richmond. This study was supported in part by the funds from Veterans Affairs Research Service. Manuscript received April 5, 1994; revision accepted May 26. Reprint requests: Dr. Akosah, Cardiology Section (111]), McGuire VA Medical Center, 1201 Broad Rock Blvd., Richmond, VA 23249
690
AF=atrial fibrillation; LAA=left atrial appendage; TEE= transesophageal echocardiography; VRR=ventricular re· sponse rate
Key words: atrial fibrillation; cardioversion; left atrial appendage; left atrium; thrombus
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
direct current cardioversion from AF to sinus rhythm. We evaluated the effect of sinus rhythm on the heart rate-LAA emptying relationship and compared this to the VRR-LAA emptying relationship in AF. Transesophageal echocardiographic imaging of the LAA was performed in patients who were undergoing elective direct current cardioversion of AF. The LAA emptying was assessed immediately before and after cardioversion to sinus rhythm. METHODS
Patient Population The study group consisted of 26 patients with chronic AF scheduled for elective direct current electrical cardioversion. The mean age was 65 ± 7 years (range, 50 to 79 years) and duration of AF was 28 ± 22 months (range, 1 week to 7 years). The underlying cardiac disease was dilated cardiomyopathy in four patients, coronary artery disease in five patients, and mitral stenosis in two patients. Hypertension, COPD and diabetes were documented in 10, 3, and 4 patients, respectively. Table 1 lists the clinical characteristics of the study group. Prolongation of the prothrombin time indicative of effective anticoagulation with warfarin therapy was documented in 18 patients (69%). Antiarrhythmic therapy included digoxin in 13, diltiazem in 6, and beta blockers in 4 patients. Type 1 antiarrhythmic agents used included procainamide in 11 patients, quinidine in 2, and propafenone in 2 patients.
Echocardiography Transthoracic echocardiography (TEE) was performed in all patients with a commercially available echocardiograph (HewlettPackard Sonos 1000) using a 2.25- or 3.5-MHz transducer. The left atrial diameter was obtained from two-dimensional directed M-mode echograms using leading edge-to-leading edge technique according to the American Society of Echocardiography criteriaa Transesophageal echocardiography was performed in all 26 patients with a single-plane TEE probe equipped with a 5-MHz transducer (Hewlett-Packard Sonos 1000). Imaging was performed both immediately before and immediately after (less than 2 mins) direct current cardioversion to sinus rhythm. The TEE was carried out in the standard fashion following local anesthesia of the oropharynx with topical lidocaine (Xylocaine) spray and coating of the probe with viscous lidocaine. The TEE probe was inserted with the patient lying in the left lateral position. The LAA was viewed initially on the basal short axis section, and then the probe was flexed to visualize the entire appendage.4-7 Pulsed Doppler echocardiography was used to measure peak flow Table !-Characteristics of the 26 Patients With Atrial
Fibrillation Who Underwent Transesophageal Echocardiography
Age, yr (mean±1SD) Duration of AF, mo Left atrial size, mm Cardiac diagnosis, No. Dilated cardiomyopathy Coronary artery disease Hypertension Mitral stenosis Non-cardiac diagnosis, No. COPD Diabetes mellitus
65±7 18±22 46±7 4
5 10 2 3 4
velocities at the outlet of the LAA, diastolic transmitral flow velocity, and diastolic pulmonary venous flow velocity. All images were recorded with 0.5-inch videotape at 30 frames per second. Left atrial appendage maximum and minimum cavity areas were measured during AF and after direct current cardioversion to normal sinus rhythm by a blinded observer. The cavity area enclosed within the LAA was planimetered manually using an off-line video analysis system. The typical LAA cavity area extended from the top of the limbus between the left superior pulmonary vein along a straight line drawn to the aorta at the shortest point of the base of the LAA. During AF, visually determined maximum and minimum LAA cavity areas during each beat for five consecutive beats (QRS complex) were analyzed and averaged to determine the mean maximum and minimum cavity areas. After direct current cardioversion, during sinus rhythm, LAAmax was measured just before the P wave (end of LAA diastole) and LAAmin just after the QRS complex (end of LAA systole). The LAA emptying was assessed by the LAA ejection fraction which was calculated by the method described by Pollick and Taylor 1 as ([LAAmax-LAAmin]/LAAmaxX100%).
Direct Current Cardioversion All direct current cardioversion procedures were carried out in the coronary care unit or the electrophysiology laboratory. Digoxin was withheld the morning of the procedure and all patients were fasted for at least 12 h prior to the procedure. Sedation was achieved in all cases with etomidate (Abbott), 0.2 to 0.6 mg / kg, administered intravenously. Direct current shocks (100 to 300 W / s) synchronized to the R wave of the ECG were applied to restore sinus rhythm. A maximum of three counter shocks and energy levels up to 300 W / s was used to convert the rhythm to the sinus pattern. Patients who were clinically unstable were excluded from the study.
Statistical Analysis Differences between groups were assessed by analysis of variance. The difference between before and after direct current cardioversion Doppler flow and ejection fraction was analyzed by Student's t test. All values in the graphs and tables are expressed as mean± 1 SD. A probability value of 0.05 or less was considered significant. RESULTS
Left Atrial Appendage Ejection Fraction vs Heart Rate During AF in the subset of patients with available transesophageal echocardiographic data both immediately before and immediately after cardioversion , mean heart rate (VRR) was 103 ± 31 beats per minute during AF, and mean LAA ejection fraction was 26± 10%. The LAA appeared akinetic with poor to absent excursion of its lateral wall. The LAA ejection fraction during AF did not correlate with either left atrial chamber size (r=-0.014; p=NS) or with the duration of AF (r=-0.04; p=NS). Table 2 is a compilation of these results of the entire study group before and after direct current cardioversion. There was a significant inverse relationship between the LAA ejection fraction and the VRR during AF (r=-0.73; p<0.001). Figure 1 illustrates this strong inverse relationship. Immediately after successful direct current cardioCHEST /107/3/ MARCH, 1995
691
Table
2-Compilation of Results From Study Group* LAA EF
Patient No.
VRR, beats/ min
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
100 89 77 131 138 104 70 96 102 63 152 113 92 76 168 132 109
18 19 20 21 22 23 24 25 26
162 95 87 70 95 80 92 135 45
Pre
TMF, cm / s LA Size, mm
Post
Pre
Subjects With Both Pre- and Post-cardioversion TEE Data 44 54 56 20 27 51 131 24 35 55 88 18 43 44 75 23 48 63 100 20 30 51 131 52 39 36 85 38 68 20 35 43 74 29 34 35 55 38 42 23 48 42 133 21 50 46 54 64 43 74 30 62 41 58 38 23 64 42 72 18 42 44 105 24 37 48 71 Subjects With Only Precardioversion TEE Data 9 39 22 48 38 50 35 25 44 37 45 18 44 17 51
PVF, (cm / s)
Post
Pre
Post
50 136 75 83 95 127 95 51 68 40 119 52 83 70 89 99 75
40 39 50 69 45 19 27 42 64 42 81 52 66 50 40 77 92
42 43 34 81 52 16 59 36 65 34 102 39 68 46 60 72 86
*LA, left atrium; TMF, transmitral flow ; PVF, pulmonary venous flow; EF, ejection fraction; pre, before cardioversion (AF); post, after cardioversion (sinus rhythm). Ellipses=indicate data not available.
version to sinus rhythm, the LAA ejection fraction increased to 46 ± 12% (p
•
50
c
0 ....... _.,.)
40
(.)
ro
....
r:x...
c
0 ....... _.,.)
(.) Q)
........
20 10
.....J
0
0 .......
0
60
60
_.,.)
*T
(.)
....ro
'
~
:1
c
• • •• • • •• • • • • •••• •
30
LAA exhibited visibly organized contractions with filling and ejection that appeared to coincide with ventricular systole and diastole, respectively. Figure 2 illustrates the increase in LAA ejection fraction immediately following cardioversion. Figure 3 is a
•
50 100 150 Heart Rate (beats/min)
0
r:x...
c
0 .......
1
40
40
_.,.)
(.)
Q) .........
T
0
~
<:r: <:r:
J.
20
20
.....J
0
PRE
POST
0
FIG URE L Relationship of LAA ejection fraction to VRR during AF (r=0.73; p<0.001).
FIGURE 2. The LAA ejection fraction in the same group of patients before and after cardioversion. Immediately after cardioversion, LAA ejection fraction is markedly improved (46 ± 12% vs 26±8%). Asterisk=p
692
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
FIGURE 3. Transesophageal echocardiographic images of the LAA in one patient before and immediately after successful ca rdioversion of AF. The top panel shows LAAmax and LAAmin during AF. The lower panel is a similar representation immediately after cardioversion.
representative TEE image of the LAA before and immediately after successful cardioversion, illustrating the increase in LAA ejection fraction. In direct contrast to the strong inverse relationship between LAA ejection fraction and VRR during AF, immediately after direct current cardioversion to sinus rhythm there was no discernible relationship between the LAA ejection fraction and the ventricular rate (r=0.06; p=NS).
Left Atrial Appendage Thrombus and Spontaneous Echo Contrast Seventeen patients had TEE performed immediately before and after cardioversion. After direct current cardioversion, TEE data were not available in three cases. Four patients were found to have a LAA thrombus and were therefore not converted. Spontaneous echocardiographic contrast was present in the left atrium and appendage in six patients including two who were found to have a thrombus. All eight patients with either thrombus or spontaneous echo contrast tended to have higher VRR compared with the group that exhibited neither spontaneous
contrast nor thrombus (131 ± 3 vs 82 ± 6 beats per minute, respectively; p<0.001 ).
Doppler Flow Velocity Profiles in Atrial Fibrillation Doppler evaluation of the LAA outflow tract during AF revealed continuous, multiphasic, low-amplitude and rapid emptying and filling waveforms. Mean peak emptying blood velocity was 20 ± 13 cm / s, and the mean peak filling blood velocity was 25 ± 12 cm / s during AF. However, immediately after cardioversion to sinus rhythm , there was unambiguous biphasic flow through the LAA that was synchronous with the cardiac cycle. Mean peak emptying and filling blood flow velocities increased to 45±15 cm / s and 40±18 cm / s, respectively, immediately after successful cardioversion (both p<0.001 compared with respective values in AF). In contrast, there were no significant changes in the transmitral or pulmonary venous diastolic flow velocities after cardioversion. Mean transmitral flow velocities before and after direct current cardioversion were 84 ± 28 and 83 ± 28 cm / s, respectively CHEST I 107 I 3 I MARCH, 1995
693
(p=NS). Mean pulmonary venous flow velocities before and after direct current cardioversion were 52±20 and 55±22 cm/s, respectively (p=NS) . DISCUSSIO"'
The major findings of our study are twofold: (1) LAA emptying is strongly and inversely related to the VRR during AF but is not related to heart rate following direct current cardioversion to sinus rhythm, and (2) LAA emptying increases significantly immediately after direct current cardioversion from AF to sinus rhythm .
Left Atrial Appendage Emptying vs Ventricular Response Rate Our data show that the magnitude of LAA emptying is inversely related to the VRR during AF (Fig 1). We are not aware of any previous reports of this relationship. The mechanistic basis of this relationship is not clear but may include, in part, the heart rate dependency of the amplitude of left ventricular diastolic filling, systolic ejection, and excursion or translation of the chamber. This heart rate dependency may affect the extent of dynamic mechanical impingement of the LAA by the left ventricle. In their studies of the LAA, Pollick and Taylor 1 and Garda-Fernandez et al 8 found that the appendage ejection fraction in patients with AF was markedly less than that in patients in sinus rhythm. Pozzoli et al 2 observed that during AF the appendage seemed to display only passive filling and emptying due to mechanical compression by the adjacent left ventricle. We propose the following mechanism to explain our findings. Lower VRR in AF prolongs ventricular diastolic filling duration and leads to greater distension of the left ventricular chamber. Because the cardiac chambers share the pericardia] space, the increased left ventricular diastolic volume may cause greater compression of the LAA during atrial systole. This would yield a smaller systolic LAA volume. Subsequently, during ventricular systole the larger left ventricular ejection (via Frank-Starling mechanism) results in greater LAA decompression and a larger LAA diastolic volume. Therefore, lower VRR in AF may result in greater LAA emptying. Higher VRRs would be expected to result in smaller oscillations of ventricular chamber volume, less compression of the appendage, and less appendage emptying. Our observations show that during AF the magnitude of emptying and filling of the fibrillating LAA is determined by the VRR rather than by the fibrillatory mechanical activity of the appendage. Furthermore, our data indicate that immediately after direct current cardioversion to sinus rhythm, LAA emptying increased substantially but that the magnitude of the emptying is not dependent on heart rate. These data 694
offer compelling support to our speculation that (1) the fibrillatory mechanical activity of the LAA during AF does not contribute to emptying and filling but that extrinsic mechanical compression by the left ventricle which is dependent on VRR plays a greater role and (2) immediately after direct current cardiaversion to sinus rhythm the LAA regains substantial coordinated mechanical activity such that the magnitude of emptying and filling is much less a function of heart rate or extrinsic compression but is governed principally by intrinsic appendage contractile capability. There are no published reports that, to our knowledge, have previously established any relationship between the magnitude of LAA emptying and the VRR in AF.
Left Atrial Appendage Emptying in Atrial Fibrillation The ejection function of the LAA was studied by Pollick and Taylor 1 in patients with AF and compared with control patients in sinus rhythm. They found that in AF the mechanical activity of the LAA assessed by the ejection fraction was poor. The presence of a thrombus in the LAA was associated with a dilated and poorly contracting appendage, even in patients with sinus rhythm. This is consistent with the findings of Garda-Fernandez et al. 8 The relationships between atrial rhythm, blood flow pattern, and the presence or absence of spontaneous echocardiographic contrast or thrombus within the LAA have been reported by others. 2·9 Pozzoli et aF studied 25 patients in AF and 57 in sinus rhythm. These investigators described Doppler characteristics which are similar to our findings. A biphasic, clearly defined emptying and filling pattern was observed in patients with sinus rhythm. In the patients with AF, the flow profile was multiphasic. In patients with poor LAA function, including four patients in sinus rhythm, no blood flow pattern was discernible. The authors concluded that an absent or low blood flow velocity within the LAA represents a predisposing factor for thrombus formation . Pozzoli et al 2 observed that during AF the appendage seemed to display only passive filling and emptying due mainly to mechanical compression from the adjacent left ventricle onto the LAA. Our observations are consistent with the observations of Pozzoli et al 2 and support the hypothesis that left atrial filling is importantly influenced by dynamic compression from the adjacent left ventricle. These studies have established that LAA emptying is poor during AF and suggest that appendage emptying during AF is importantly determined by extrinsic mechanical compression by the left ventricle. Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
Restoration of Left Atrium and Appendage Function After Direct Current Cardioversion to Sinus Rhythm The time course of restoration of left atrial function after direct current cardioversion from AF to sinus rhythm has been studied by several workers. Using kinetocardiography, Mahlich et aF 0 detected atrial mechanical function in almost all patients immediately and at 24 h after cardioversion. O'Neill et al 11 used the atrial filling fraction of the transthoracic Doppler transmitral flow velocity as a marker for atrial mechanical activity and evaluated eight patients in AF before cardioversion and again at 5 min and 30 min and at 24 h after successful cardioversion to sinus rhythm. They found no evidence of mechanical activity in two patients immediately after cardioversion, and two had no return of atrial mechanical activity until 24 h later. Manning et aP 2 carried out serial transthoracic two-dimensional, M-mode, and transmitral pulsed Doppler echocardiographic studies in 21 patients undergoing elective cardioversion for AF. These workers noted that at follow-up examination 3 months later there were significant increases in both peak A wave velocity and percentage of atrial contribution to left ventricular filling. When compared with a control population, however, peak A wave velocity and percentage of atrial contribution to left ventricular filling did not attain control levels at 3 months after cardioversion, suggesting ongoing but delayed restoration of atrial contractile function. Of importance with respect to our data, however, was the presence of identifiable atrial filling waves in 95% of their patients within 1 h after cardioversion. These studies seem to indicate that left atrial contractile performance resumes immediately after direct current cardioversion from AF to sinus rhythm but remains at subnormal levels for up to 3 months after direct current cardioversion. In a recent notable study of LAA function, Grim et aF 3 found that electrical cardioversion of AF resulted in a return of organized contractions of the LAA in most patients immediately after cardioversion. These observations are consistent with our findings that immediately after cardioversion, LAA ejection fraction increased and that the ejection fraction was no longer dependent on heart rate as it was during AF. Moreover, the resumption of biphasic pulse wave Doppler flow patterns in most (65%) of their patients after cardioversion is remarkably similar to our pulse Doppler observations after cardioversion, supporting our speculation that organized LAA function returns immediately after cardioversion of AF. This contrasts sharply with the previous observations of delayed and subnormal restoration of left atrial contractile function. Using TEE to more directly visualize the atrial appendage, we have de-
scribed a marked increase in the LAA ejection fraction immediately after cardioversion in contrast to the previously reported studies which have described a delayed return of contractile function of the left atrium . Two important methodologic differences between the present investigation and most of these previous studies should be recognized. First, most of the previous studies employed only transthoracic echocardiography. We believe that in our study TEE permitted us to assess more accurately atrial mechanical activity immediately before and after direct current cardioversion to sinus rhythm. Second, none of the previous studies focused specifically on LAA function in the same group of patients serially before and after cardioversion. These are features unique to the present investigation. It is important to emphasize that no previous information exists regarding the relationship between VRR and appendage ejection fraction, thrombus formation or thromboembolism rates. We emphasize that our investigation has not established a causal link between VRR in AF and LAA thrombus formation but has instead only presented data that suggest a possible linkage.
Clinical Implications Stroke is the third leading cause of death and long-term disability in the United States. 14 Emboli arising from the heart are estimated to account for 15 to 20% of all ischemic strokes with a reported high prevalence of 34%. 15· 17 Emboli from the left atrium and LAA may account for a large portion of all embolic strokes, although no quantitative data exist. On this basis, the influence of LAA emptying and filling dynamics on thrombus formation and subsequent embolic strokes appears to be important. Our data indicate that the VRR in AF strongly determines the dynamics of emptying and filling of the LAA. The magnitude of appendage emptying increases as the VRR decreases. Higher VRRs are associated with poorer appendage emptying and possibly greater blood flow stasis. We take our findings to suggest that restraint of the VRR may minimize the conditions which favor thrombus formation in the LAA . Further studies are needed to define whether control of the VRR in chronic AF can prevent or reduce the incidence of thrombus formation in the LAA.
Limitations of the Study An important methodologic limitation of this study is the method used to assess LAA emptying. The cavity area of the appendage was assessed quantitatively from transesophageal echocardiographic tomographic sections of the appendage. Translation of the appendage during the cardiac cycle may have CHEST /107/3/ MARCH, 1995
695
been sufficient to cause misalignment of the sections leading to errors in the estimation of cavity volume changes. This is a limitation inherent to quantitative analysis of echocardiographic images that is difficult to eliminate and, therefore, must be acknowledged as potentially confounding. Nevertheless, the transesophageal echocardiographic imaging was performed with care and diligence to minimize this factor. The LAA cavity area was scanned carefully to select image sequences that showed identical luminal landmarks. A second limitation of this study concerns the notion that the LAA possesses an ejection fraction, implying that it possesses an intrinsic contractile capability. Although the normal atrial appendage is known to possess contractile capabilities, whether this is adequately expressed by the ejection fraction is uncertain. It is conceivable that the term "ejection fraction" does not accurately describe the source of the appendage cavity area oscillations within a cardiac cycle and from AF to sinus rhythm. However, we regard our finding of the heart rate independence of the magnitude of LAA emptying during sinus rhythm and the heart rate dependence during AF to be strong evidence that the intrinsic contractile capability of the appendage changed. The systematic changes in blood flow velocity patterns and amplitudes that accompanied the alterations in appendage ejection fraction can also be taken as evidence that the intrinsic contractile capability of the appendage changed.
expert assistance in the preparation of the manuscript.
REFERENCES
2
3
4
5
6
7
8 9
SUMMARY
This study offers evidence that during AF the magnitude of LAA emptying is inversely related to the VRR and that, immediately after direct current cardioversion to sinus rhythm, appendage emptying increases markedly and is not dependent on heart rate. These results support our hypothesis that LAA emptying in AF is dependent on the diastolic volume of the left ventricle which is determined by the VRR. Our data can be taken further to indicate that the increase in LAA emptying immediately after direct current cardioversion to sinus rhythm is due to a change in the intrinsic emptying or contractile capability of the appendage. These findings suggest that higher VRR in AF may be an important predisposing factor for blood flow stasis and thrombus formation in the LAA. Further studies are needed to define the influence of VRR control during AF on the incidence of LAA thrombus formation.
10 11
12 13
14 15 16
Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography: implications for the development of thrombus. Circulation 1991; 84:223-31 Pozzoli M, Febo 0, Torbicki A, et al. Left atrial appendage dysfunction: a cause of thrombosis? Evidence by transesophageal echocardiography-doppler studies. J Am Soc Echocardio 1991; 4:435-41 Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978; 58:1072-83 Aschenberg W, Schluter M, Kremer P, et al. Transesophageal two-dimensional echocardiography for the detection of left atrial appendage thrombus. JAm Coli Cardiol1986; 7:163-66 Seward JB, Khandheria BK, Oh JK, et al. Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988; 63:649-80 Mugge A, Daniel WG, Hausmann D, et al. Diagnosis of left atrial appendage thrombi by transesophageal echocardiography: clinical implications and follow-up. Am J Cardiac Imaging 1990; 4:173-79 Castello R, Pearson AC, Labovitz AJ, et al. Prevalence and clinical implications of atrial spontaneous contrast in patients undergoing transesophageal echocardiography. Am J Cardiol 1990; 65:1149-53 Garda-Fernandez MA, Torrecilla EG, San Roman D, etal. Left atrial appendage Doppler flow patterns: implications on thrombus formation. Am Heart J 1992; 124:955-61 Suetsugu M, Matsuzaki M, Toma Y, et al. Detection of mural thrombi and analysis of blood flow velocities in the left atrial appendage using transesophageal two-dimensional echocardiography and pulsed Doppler flowmetry. J Cardiol 1988; 18: 385-94 Mahlich J, Schweizer W, Burkat F. Atrial function after cardioversion for atrial fibrillation. Br Heart J 1973; 35:24-7 O'Neill PG, Puleo PR, Bolli R, et al. Return of atrial mechanical function following electrical conversion of atrial dysrhythmias. Am Heart J 1990; 120:353-59 Manning WJ, Leeman D, Gotch P, et al. Pulsed Doppler evaluation of atrial function after electrical cardioversion of atrial fibrillation. J Am Coli Cardiol1989; 13:617-23 Grim RA, Stewart WJ, Maloney JD, et al. Impact of electrical cardioversion for atrial fibrillation on left atrial appendage function and spontaneous echo contrast: characterization by simultaneously transesophageal echocardiography. J Am Colt Cardiol 1993; 22:1359-66 Adams RLM , Victor M, eds. Principles of neurology. New York: McGraw Hill, 1985: chap 33 Stroke Prevention in Atrial Fibrillation Study Group. Preliminary report of the Stroke Prevention in Atrial Fibrillation Study. N Engl J Med 1990; 322:863-68 Petersen P, Kastrup J, Helweg-Larsen S, et at. Risk factors for thromboembolic complications in chronic atrial fibrillation: The Copenhagen AFASAK Study. Arch Intern Med 1990; 150:819-21 Cerebral Embolism Task Force. Cardiogenic brain embolism: the second report of the cerebral embolism task force. Arch Neurol1989; 46:727-43
ACKNOWLEDGMENTS: We thank James A. Brownie from Medical Media for assistance with graphics, Dorothy Robinson for assistance with data management, anCI Michelle Rene Martin for
17
696
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah et alj
26± 10%, and immediately after successful cardioversion, it increased to 46± 12% (p<0.001). However, during sinus rhythm there was no relationship between LAA ejection fraction and VRR (r=0.06; p=NS) in the subgroup of patients who were successfully converted to sinus rhythm. There were poor relationships between LAA ejection fraction and peak transmitral flow velocity (r=-0.41; p=NS) or pulmonary venous flow velocity (r=-0.03; p=NS) in AF. Conclusion: These results indicate that the magnitude of LAA emptying in AF is strongly and inversely influenced by ventricular rate. Direct current cardioversion to sinus rhythm is associated with an increase in the magnitude of LAA emptying that is not influenced by heart rate. The magnitude of LAA emptying may be an important factor in the formation of thromboemboli in AF. The extent to which controlling the VRR in chronic AF will prevent stasis and LAA thrombus formation remains to be determined. (Chest 1995; 107:690-96)
The association of systemic emboli with atrial fibrillation (AF) and the formation of intracardiac thrombi in the left atrial appendage (LAA) are well established. The exact mechanism, however, for the preponderance of thrombi in the LAA during AF has not been defined. It is believed that attenuation of LAA contractile activity and stagnation of blood flow within the appendage are important factors. Several observers have proposed 1•2 that the magnitude of LAA emptying during AF is determined, in part, by extrinsic compression by the adjacent left
ventricular lateral free wall during diastolic ventricular filling. Since ventricular diastolic chamber volume is strongly influenced by heart rate and diastolic filling duration, we speculated that heart rate may affect the magnitude of LAA emptying. The magnitude of LAA emptying may then be a function of the ventricular response rate (VRR) in AF . We reasoned that lower ventricular response rates in atrial fibrillation would be expected to increase the ventricular diastolic filling duration, increase ventricular diastolic volume, increase the amplitude of compression of the LAA, and result in greater LAA emptying. On this basis, we hypothesized that the amplitude of LAA emptying is related to VRR during AF. To test this hypothesis, we evaluated therelationship between LAA emptying and VRR in patients with chronic AF . Asecond objective of this study was to examine the effect of sinus rhythm and heart rate on LAA emptying immediately after
*From the Hunter Holmes McGuire Veterans Affairs Medical Center, Medical College of Virginia/ Virginia Commonwealth University, Richmond. This study was supported in part by the funds from Veterans Affairs Research Service. Manuscript received April 5, 1994; revision accepted May 26. Reprint requests: Dr. Akosah, Cardiology Section (111]), McGuire VA Medical Center, 1201 Broad Rock Blvd., Richmond, VA 23249
690
AF=atrial fibrillation; LAA=left atrial appendage; TEE= transesophageal echocardiography; VRR=ventricular re· sponse rate
Key words: atrial fibrillation; cardioversion; left atrial appendage; left atrium; thrombus
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
direct current cardioversion from AF to sinus rhythm. We evaluated the effect of sinus rhythm on the heart rate-LAA emptying relationship and compared this to the VRR-LAA emptying relationship in AF. Transesophageal echocardiographic imaging of the LAA was performed in patients who were undergoing elective direct current cardioversion of AF. The LAA emptying was assessed immediately before and after cardioversion to sinus rhythm. METHODS
Patient Population The study group consisted of 26 patients with chronic AF scheduled for elective direct current electrical cardioversion. The mean age was 65 ± 7 years (range, 50 to 79 years) and duration of AF was 28 ± 22 months (range, 1 week to 7 years). The underlying cardiac disease was dilated cardiomyopathy in four patients, coronary artery disease in five patients, and mitral stenosis in two patients. Hypertension, COPD and diabetes were documented in 10, 3, and 4 patients, respectively. Table 1 lists the clinical characteristics of the study group. Prolongation of the prothrombin time indicative of effective anticoagulation with warfarin therapy was documented in 18 patients (69%). Antiarrhythmic therapy included digoxin in 13, diltiazem in 6, and beta blockers in 4 patients. Type 1 antiarrhythmic agents used included procainamide in 11 patients, quinidine in 2, and propafenone in 2 patients.
Echocardiography Transthoracic echocardiography (TEE) was performed in all patients with a commercially available echocardiograph (HewlettPackard Sonos 1000) using a 2.25- or 3.5-MHz transducer. The left atrial diameter was obtained from two-dimensional directed M-mode echograms using leading edge-to-leading edge technique according to the American Society of Echocardiography criteriaa Transesophageal echocardiography was performed in all 26 patients with a single-plane TEE probe equipped with a 5-MHz transducer (Hewlett-Packard Sonos 1000). Imaging was performed both immediately before and immediately after (less than 2 mins) direct current cardioversion to sinus rhythm. The TEE was carried out in the standard fashion following local anesthesia of the oropharynx with topical lidocaine (Xylocaine) spray and coating of the probe with viscous lidocaine. The TEE probe was inserted with the patient lying in the left lateral position. The LAA was viewed initially on the basal short axis section, and then the probe was flexed to visualize the entire appendage.4-7 Pulsed Doppler echocardiography was used to measure peak flow Table !-Characteristics of the 26 Patients With Atrial
Fibrillation Who Underwent Transesophageal Echocardiography
Age, yr (mean±1SD) Duration of AF, mo Left atrial size, mm Cardiac diagnosis, No. Dilated cardiomyopathy Coronary artery disease Hypertension Mitral stenosis Non-cardiac diagnosis, No. COPD Diabetes mellitus
65±7 18±22 46±7 4
5 10 2 3 4
velocities at the outlet of the LAA, diastolic transmitral flow velocity, and diastolic pulmonary venous flow velocity. All images were recorded with 0.5-inch videotape at 30 frames per second. Left atrial appendage maximum and minimum cavity areas were measured during AF and after direct current cardioversion to normal sinus rhythm by a blinded observer. The cavity area enclosed within the LAA was planimetered manually using an off-line video analysis system. The typical LAA cavity area extended from the top of the limbus between the left superior pulmonary vein along a straight line drawn to the aorta at the shortest point of the base of the LAA. During AF, visually determined maximum and minimum LAA cavity areas during each beat for five consecutive beats (QRS complex) were analyzed and averaged to determine the mean maximum and minimum cavity areas. After direct current cardioversion, during sinus rhythm, LAAmax was measured just before the P wave (end of LAA diastole) and LAAmin just after the QRS complex (end of LAA systole). The LAA emptying was assessed by the LAA ejection fraction which was calculated by the method described by Pollick and Taylor 1 as ([LAAmax-LAAmin]/LAAmaxX100%).
Direct Current Cardioversion All direct current cardioversion procedures were carried out in the coronary care unit or the electrophysiology laboratory. Digoxin was withheld the morning of the procedure and all patients were fasted for at least 12 h prior to the procedure. Sedation was achieved in all cases with etomidate (Abbott), 0.2 to 0.6 mg / kg, administered intravenously. Direct current shocks (100 to 300 W / s) synchronized to the R wave of the ECG were applied to restore sinus rhythm. A maximum of three counter shocks and energy levels up to 300 W / s was used to convert the rhythm to the sinus pattern. Patients who were clinically unstable were excluded from the study.
Statistical Analysis Differences between groups were assessed by analysis of variance. The difference between before and after direct current cardioversion Doppler flow and ejection fraction was analyzed by Student's t test. All values in the graphs and tables are expressed as mean± 1 SD. A probability value of 0.05 or less was considered significant. RESULTS
Left Atrial Appendage Ejection Fraction vs Heart Rate During AF in the subset of patients with available transesophageal echocardiographic data both immediately before and immediately after cardioversion , mean heart rate (VRR) was 103 ± 31 beats per minute during AF, and mean LAA ejection fraction was 26± 10%. The LAA appeared akinetic with poor to absent excursion of its lateral wall. The LAA ejection fraction during AF did not correlate with either left atrial chamber size (r=-0.014; p=NS) or with the duration of AF (r=-0.04; p=NS). Table 2 is a compilation of these results of the entire study group before and after direct current cardioversion. There was a significant inverse relationship between the LAA ejection fraction and the VRR during AF (r=-0.73; p<0.001). Figure 1 illustrates this strong inverse relationship. Immediately after successful direct current cardioCHEST /107/3/ MARCH, 1995
691
Table
2-Compilation of Results From Study Group* LAA EF
Patient No.
VRR, beats/ min
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
100 89 77 131 138 104 70 96 102 63 152 113 92 76 168 132 109
18 19 20 21 22 23 24 25 26
162 95 87 70 95 80 92 135 45
Pre
TMF, cm / s LA Size, mm
Post
Pre
Subjects With Both Pre- and Post-cardioversion TEE Data 44 54 56 20 27 51 131 24 35 55 88 18 43 44 75 23 48 63 100 20 30 51 131 52 39 36 85 38 68 20 35 43 74 29 34 35 55 38 42 23 48 42 133 21 50 46 54 64 43 74 30 62 41 58 38 23 64 42 72 18 42 44 105 24 37 48 71 Subjects With Only Precardioversion TEE Data 9 39 22 48 38 50 35 25 44 37 45 18 44 17 51
PVF, (cm / s)
Post
Pre
Post
50 136 75 83 95 127 95 51 68 40 119 52 83 70 89 99 75
40 39 50 69 45 19 27 42 64 42 81 52 66 50 40 77 92
42 43 34 81 52 16 59 36 65 34 102 39 68 46 60 72 86
*LA, left atrium; TMF, transmitral flow ; PVF, pulmonary venous flow; EF, ejection fraction; pre, before cardioversion (AF); post, after cardioversion (sinus rhythm). Ellipses=indicate data not available.
version to sinus rhythm, the LAA ejection fraction increased to 46 ± 12% (p
•
50
c
0 ....... _.,.)
40
(.)
ro
....
r:x...
c
0 ....... _.,.)
(.) Q)
........
20 10
.....J
0
0 .......
0
60
60
_.,.)
*T
(.)
....ro
'
~
:1
c
• • •• • • •• • • • • •••• •
30
LAA exhibited visibly organized contractions with filling and ejection that appeared to coincide with ventricular systole and diastole, respectively. Figure 2 illustrates the increase in LAA ejection fraction immediately following cardioversion. Figure 3 is a
•
50 100 150 Heart Rate (beats/min)
0
r:x...
c
0 .......
1
40
40
_.,.)
(.)
Q) .........
T
0
~
<:r: <:r:
J.
20
20
.....J
0
PRE
POST
0
FIG URE L Relationship of LAA ejection fraction to VRR during AF (r=0.73; p<0.001).
FIGURE 2. The LAA ejection fraction in the same group of patients before and after cardioversion. Immediately after cardioversion, LAA ejection fraction is markedly improved (46 ± 12% vs 26±8%). Asterisk=p
692
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
FIGURE 3. Transesophageal echocardiographic images of the LAA in one patient before and immediately after successful ca rdioversion of AF. The top panel shows LAAmax and LAAmin during AF. The lower panel is a similar representation immediately after cardioversion.
representative TEE image of the LAA before and immediately after successful cardioversion, illustrating the increase in LAA ejection fraction. In direct contrast to the strong inverse relationship between LAA ejection fraction and VRR during AF, immediately after direct current cardioversion to sinus rhythm there was no discernible relationship between the LAA ejection fraction and the ventricular rate (r=0.06; p=NS).
Left Atrial Appendage Thrombus and Spontaneous Echo Contrast Seventeen patients had TEE performed immediately before and after cardioversion. After direct current cardioversion, TEE data were not available in three cases. Four patients were found to have a LAA thrombus and were therefore not converted. Spontaneous echocardiographic contrast was present in the left atrium and appendage in six patients including two who were found to have a thrombus. All eight patients with either thrombus or spontaneous echo contrast tended to have higher VRR compared with the group that exhibited neither spontaneous
contrast nor thrombus (131 ± 3 vs 82 ± 6 beats per minute, respectively; p<0.001 ).
Doppler Flow Velocity Profiles in Atrial Fibrillation Doppler evaluation of the LAA outflow tract during AF revealed continuous, multiphasic, low-amplitude and rapid emptying and filling waveforms. Mean peak emptying blood velocity was 20 ± 13 cm / s, and the mean peak filling blood velocity was 25 ± 12 cm / s during AF. However, immediately after cardioversion to sinus rhythm , there was unambiguous biphasic flow through the LAA that was synchronous with the cardiac cycle. Mean peak emptying and filling blood flow velocities increased to 45±15 cm / s and 40±18 cm / s, respectively, immediately after successful cardioversion (both p<0.001 compared with respective values in AF). In contrast, there were no significant changes in the transmitral or pulmonary venous diastolic flow velocities after cardioversion. Mean transmitral flow velocities before and after direct current cardioversion were 84 ± 28 and 83 ± 28 cm / s, respectively CHEST I 107 I 3 I MARCH, 1995
693
(p=NS). Mean pulmonary venous flow velocities before and after direct current cardioversion were 52±20 and 55±22 cm/s, respectively (p=NS) . DISCUSSIO"'
The major findings of our study are twofold: (1) LAA emptying is strongly and inversely related to the VRR during AF but is not related to heart rate following direct current cardioversion to sinus rhythm, and (2) LAA emptying increases significantly immediately after direct current cardioversion from AF to sinus rhythm .
Left Atrial Appendage Emptying vs Ventricular Response Rate Our data show that the magnitude of LAA emptying is inversely related to the VRR during AF (Fig 1). We are not aware of any previous reports of this relationship. The mechanistic basis of this relationship is not clear but may include, in part, the heart rate dependency of the amplitude of left ventricular diastolic filling, systolic ejection, and excursion or translation of the chamber. This heart rate dependency may affect the extent of dynamic mechanical impingement of the LAA by the left ventricle. In their studies of the LAA, Pollick and Taylor 1 and Garda-Fernandez et al 8 found that the appendage ejection fraction in patients with AF was markedly less than that in patients in sinus rhythm. Pozzoli et al 2 observed that during AF the appendage seemed to display only passive filling and emptying due to mechanical compression by the adjacent left ventricle. We propose the following mechanism to explain our findings. Lower VRR in AF prolongs ventricular diastolic filling duration and leads to greater distension of the left ventricular chamber. Because the cardiac chambers share the pericardia] space, the increased left ventricular diastolic volume may cause greater compression of the LAA during atrial systole. This would yield a smaller systolic LAA volume. Subsequently, during ventricular systole the larger left ventricular ejection (via Frank-Starling mechanism) results in greater LAA decompression and a larger LAA diastolic volume. Therefore, lower VRR in AF may result in greater LAA emptying. Higher VRRs would be expected to result in smaller oscillations of ventricular chamber volume, less compression of the appendage, and less appendage emptying. Our observations show that during AF the magnitude of emptying and filling of the fibrillating LAA is determined by the VRR rather than by the fibrillatory mechanical activity of the appendage. Furthermore, our data indicate that immediately after direct current cardioversion to sinus rhythm, LAA emptying increased substantially but that the magnitude of the emptying is not dependent on heart rate. These data 694
offer compelling support to our speculation that (1) the fibrillatory mechanical activity of the LAA during AF does not contribute to emptying and filling but that extrinsic mechanical compression by the left ventricle which is dependent on VRR plays a greater role and (2) immediately after direct current cardiaversion to sinus rhythm the LAA regains substantial coordinated mechanical activity such that the magnitude of emptying and filling is much less a function of heart rate or extrinsic compression but is governed principally by intrinsic appendage contractile capability. There are no published reports that, to our knowledge, have previously established any relationship between the magnitude of LAA emptying and the VRR in AF.
Left Atrial Appendage Emptying in Atrial Fibrillation The ejection function of the LAA was studied by Pollick and Taylor 1 in patients with AF and compared with control patients in sinus rhythm. They found that in AF the mechanical activity of the LAA assessed by the ejection fraction was poor. The presence of a thrombus in the LAA was associated with a dilated and poorly contracting appendage, even in patients with sinus rhythm. This is consistent with the findings of Garda-Fernandez et al. 8 The relationships between atrial rhythm, blood flow pattern, and the presence or absence of spontaneous echocardiographic contrast or thrombus within the LAA have been reported by others. 2·9 Pozzoli et aF studied 25 patients in AF and 57 in sinus rhythm. These investigators described Doppler characteristics which are similar to our findings. A biphasic, clearly defined emptying and filling pattern was observed in patients with sinus rhythm. In the patients with AF, the flow profile was multiphasic. In patients with poor LAA function, including four patients in sinus rhythm, no blood flow pattern was discernible. The authors concluded that an absent or low blood flow velocity within the LAA represents a predisposing factor for thrombus formation . Pozzoli et al 2 observed that during AF the appendage seemed to display only passive filling and emptying due mainly to mechanical compression from the adjacent left ventricle onto the LAA. Our observations are consistent with the observations of Pozzoli et al 2 and support the hypothesis that left atrial filling is importantly influenced by dynamic compression from the adjacent left ventricle. These studies have established that LAA emptying is poor during AF and suggest that appendage emptying during AF is importantly determined by extrinsic mechanical compression by the left ventricle. Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah eta/)
Restoration of Left Atrium and Appendage Function After Direct Current Cardioversion to Sinus Rhythm The time course of restoration of left atrial function after direct current cardioversion from AF to sinus rhythm has been studied by several workers. Using kinetocardiography, Mahlich et aF 0 detected atrial mechanical function in almost all patients immediately and at 24 h after cardioversion. O'Neill et al 11 used the atrial filling fraction of the transthoracic Doppler transmitral flow velocity as a marker for atrial mechanical activity and evaluated eight patients in AF before cardioversion and again at 5 min and 30 min and at 24 h after successful cardioversion to sinus rhythm. They found no evidence of mechanical activity in two patients immediately after cardioversion, and two had no return of atrial mechanical activity until 24 h later. Manning et aP 2 carried out serial transthoracic two-dimensional, M-mode, and transmitral pulsed Doppler echocardiographic studies in 21 patients undergoing elective cardioversion for AF. These workers noted that at follow-up examination 3 months later there were significant increases in both peak A wave velocity and percentage of atrial contribution to left ventricular filling. When compared with a control population, however, peak A wave velocity and percentage of atrial contribution to left ventricular filling did not attain control levels at 3 months after cardioversion, suggesting ongoing but delayed restoration of atrial contractile function. Of importance with respect to our data, however, was the presence of identifiable atrial filling waves in 95% of their patients within 1 h after cardioversion. These studies seem to indicate that left atrial contractile performance resumes immediately after direct current cardioversion from AF to sinus rhythm but remains at subnormal levels for up to 3 months after direct current cardioversion. In a recent notable study of LAA function, Grim et aF 3 found that electrical cardioversion of AF resulted in a return of organized contractions of the LAA in most patients immediately after cardioversion. These observations are consistent with our findings that immediately after cardioversion, LAA ejection fraction increased and that the ejection fraction was no longer dependent on heart rate as it was during AF. Moreover, the resumption of biphasic pulse wave Doppler flow patterns in most (65%) of their patients after cardioversion is remarkably similar to our pulse Doppler observations after cardioversion, supporting our speculation that organized LAA function returns immediately after cardioversion of AF. This contrasts sharply with the previous observations of delayed and subnormal restoration of left atrial contractile function. Using TEE to more directly visualize the atrial appendage, we have de-
scribed a marked increase in the LAA ejection fraction immediately after cardioversion in contrast to the previously reported studies which have described a delayed return of contractile function of the left atrium . Two important methodologic differences between the present investigation and most of these previous studies should be recognized. First, most of the previous studies employed only transthoracic echocardiography. We believe that in our study TEE permitted us to assess more accurately atrial mechanical activity immediately before and after direct current cardioversion to sinus rhythm. Second, none of the previous studies focused specifically on LAA function in the same group of patients serially before and after cardioversion. These are features unique to the present investigation. It is important to emphasize that no previous information exists regarding the relationship between VRR and appendage ejection fraction, thrombus formation or thromboembolism rates. We emphasize that our investigation has not established a causal link between VRR in AF and LAA thrombus formation but has instead only presented data that suggest a possible linkage.
Clinical Implications Stroke is the third leading cause of death and long-term disability in the United States. 14 Emboli arising from the heart are estimated to account for 15 to 20% of all ischemic strokes with a reported high prevalence of 34%. 15· 17 Emboli from the left atrium and LAA may account for a large portion of all embolic strokes, although no quantitative data exist. On this basis, the influence of LAA emptying and filling dynamics on thrombus formation and subsequent embolic strokes appears to be important. Our data indicate that the VRR in AF strongly determines the dynamics of emptying and filling of the LAA. The magnitude of appendage emptying increases as the VRR decreases. Higher VRRs are associated with poorer appendage emptying and possibly greater blood flow stasis. We take our findings to suggest that restraint of the VRR may minimize the conditions which favor thrombus formation in the LAA . Further studies are needed to define whether control of the VRR in chronic AF can prevent or reduce the incidence of thrombus formation in the LAA.
Limitations of the Study An important methodologic limitation of this study is the method used to assess LAA emptying. The cavity area of the appendage was assessed quantitatively from transesophageal echocardiographic tomographic sections of the appendage. Translation of the appendage during the cardiac cycle may have CHEST /107/3/ MARCH, 1995
695
been sufficient to cause misalignment of the sections leading to errors in the estimation of cavity volume changes. This is a limitation inherent to quantitative analysis of echocardiographic images that is difficult to eliminate and, therefore, must be acknowledged as potentially confounding. Nevertheless, the transesophageal echocardiographic imaging was performed with care and diligence to minimize this factor. The LAA cavity area was scanned carefully to select image sequences that showed identical luminal landmarks. A second limitation of this study concerns the notion that the LAA possesses an ejection fraction, implying that it possesses an intrinsic contractile capability. Although the normal atrial appendage is known to possess contractile capabilities, whether this is adequately expressed by the ejection fraction is uncertain. It is conceivable that the term "ejection fraction" does not accurately describe the source of the appendage cavity area oscillations within a cardiac cycle and from AF to sinus rhythm. However, we regard our finding of the heart rate independence of the magnitude of LAA emptying during sinus rhythm and the heart rate dependence during AF to be strong evidence that the intrinsic contractile capability of the appendage changed. The systematic changes in blood flow velocity patterns and amplitudes that accompanied the alterations in appendage ejection fraction can also be taken as evidence that the intrinsic contractile capability of the appendage changed.
expert assistance in the preparation of the manuscript.
REFERENCES
2
3
4
5
6
7
8 9
SUMMARY
This study offers evidence that during AF the magnitude of LAA emptying is inversely related to the VRR and that, immediately after direct current cardioversion to sinus rhythm, appendage emptying increases markedly and is not dependent on heart rate. These results support our hypothesis that LAA emptying in AF is dependent on the diastolic volume of the left ventricle which is determined by the VRR. Our data can be taken further to indicate that the increase in LAA emptying immediately after direct current cardioversion to sinus rhythm is due to a change in the intrinsic emptying or contractile capability of the appendage. These findings suggest that higher VRR in AF may be an important predisposing factor for blood flow stasis and thrombus formation in the LAA. Further studies are needed to define the influence of VRR control during AF on the incidence of LAA thrombus formation.
10 11
12 13
14 15 16
Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography: implications for the development of thrombus. Circulation 1991; 84:223-31 Pozzoli M, Febo 0, Torbicki A, et al. Left atrial appendage dysfunction: a cause of thrombosis? Evidence by transesophageal echocardiography-doppler studies. J Am Soc Echocardio 1991; 4:435-41 Sahn DJ, DeMaria A, Kisslo J, et al. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978; 58:1072-83 Aschenberg W, Schluter M, Kremer P, et al. Transesophageal two-dimensional echocardiography for the detection of left atrial appendage thrombus. JAm Coli Cardiol1986; 7:163-66 Seward JB, Khandheria BK, Oh JK, et al. Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988; 63:649-80 Mugge A, Daniel WG, Hausmann D, et al. Diagnosis of left atrial appendage thrombi by transesophageal echocardiography: clinical implications and follow-up. Am J Cardiac Imaging 1990; 4:173-79 Castello R, Pearson AC, Labovitz AJ, et al. Prevalence and clinical implications of atrial spontaneous contrast in patients undergoing transesophageal echocardiography. Am J Cardiol 1990; 65:1149-53 Garda-Fernandez MA, Torrecilla EG, San Roman D, etal. Left atrial appendage Doppler flow patterns: implications on thrombus formation. Am Heart J 1992; 124:955-61 Suetsugu M, Matsuzaki M, Toma Y, et al. Detection of mural thrombi and analysis of blood flow velocities in the left atrial appendage using transesophageal two-dimensional echocardiography and pulsed Doppler flowmetry. J Cardiol 1988; 18: 385-94 Mahlich J, Schweizer W, Burkat F. Atrial function after cardioversion for atrial fibrillation. Br Heart J 1973; 35:24-7 O'Neill PG, Puleo PR, Bolli R, et al. Return of atrial mechanical function following electrical conversion of atrial dysrhythmias. Am Heart J 1990; 120:353-59 Manning WJ, Leeman D, Gotch P, et al. Pulsed Doppler evaluation of atrial function after electrical cardioversion of atrial fibrillation. J Am Coli Cardiol1989; 13:617-23 Grim RA, Stewart WJ, Maloney JD, et al. Impact of electrical cardioversion for atrial fibrillation on left atrial appendage function and spontaneous echo contrast: characterization by simultaneously transesophageal echocardiography. J Am Colt Cardiol 1993; 22:1359-66 Adams RLM , Victor M, eds. Principles of neurology. New York: McGraw Hill, 1985: chap 33 Stroke Prevention in Atrial Fibrillation Study Group. Preliminary report of the Stroke Prevention in Atrial Fibrillation Study. N Engl J Med 1990; 322:863-68 Petersen P, Kastrup J, Helweg-Larsen S, et at. Risk factors for thromboembolic complications in chronic atrial fibrillation: The Copenhagen AFASAK Study. Arch Intern Med 1990; 150:819-21 Cerebral Embolism Task Force. Cardiogenic brain embolism: the second report of the cerebral embolism task force. Arch Neurol1989; 46:727-43
ACKNOWLEDGMENTS: We thank James A. Brownie from Medical Media for assistance with graphics, Dorothy Robinson for assistance with data management, anCI Michelle Rene Martin for
17
696
Left Atrial Appendage Contractile Function in Atrial Fibrillation (Akosah et alj