- Email: [email protected]
Evaluation of transesophageal echocardiography before cardioversion of atrial fibrillation and flutter in nonanticoagulated patients
Evaluation of transesophageal echocardiography before cardioversion of atrial fibrillation and flutter in nonanticoagulated patients This study prospectively evaluated the role of transesophageal echocardiography (TEE) in screening for atrial thrombi before electrical cardioversion in 40 nonanticoagulated patients with nonvalvular atrial fibrillation (n = 33) or atrial flutter (n = 7). Transthoracic echocardiography did not detect atrial thrombus in any patient. TEE detected left atrial appendage thrombi in five patients (12%, p = 0.03), significantly associated with left ventricular systolic dysfunction (p = 0.02) and left atrial spontaneous echo contrast (p = 0.04). Cardioversion was cancelled in the five patients with thrombi and in two patients with spontaneous reversion before planned cardioversion. Cardioversion was successful in 25 (76%) of the 33 remaining patients. Cerebral embolism occurred 24 hours after successful cardioversion in one patient with atrial fibrillation and left ventricular dysfunction, who had left atrial spontaneous echo contrast, but no thrombus was detected by TEE before cardioversion. Repeat TEE after embolism showed a fresh left atrial appendage thrombus and increased left atrial spontaneous echo contrast. These results indicate that TEE improves the detection of left atrial appendage thrombi in candidates for cardioversion, in whom the procedure may be deferred. However, the exclusion by TEE of preexisting atrial thrombi before cardioversion does not eliminate the risk of embolism after cardioversion because of persistent atrial stasis and de novo thrombosis. (AM HEART J 1993;126:375-381.)
Ian W. Black, M B B S , Andrew P. Hopkins, MBBS, Lincoln C. L. Lee, M B B S , and W a r r e n F. Walsh, M B B S Sydney, Australia
Electrical cardioversion of atrial fibrillation and atrial flutter is p e r f o r m e d to relieve symptoms, improve cardiac function, and reduce the potential for subsequent thromboembolism. Cardioversion itself m a y be complicated by cerebral, systemic, and p u l m o n a r y embolism, which occurs in up to 5.6 % of patients. 1-3 Despite the lack of randomized controlled trials, findings in some studies 4-7 suggest a reduction in the incidence of embolism after cardioversion in patients receiving anticoagulant therapy. However, prophylactic anticoagulation does not eliminate the risk of embolism, 4' 8-11 m a y be associated with a delay in cardioversion, additional cost, and potential morbidity, and consequently is not always used.
Fromthe Departmentof CardiovascularMedicine,PrinceHenryHospital. Supportedin part bythe PrinceHenryHospitalCentenaryResearchFund. Receivedfor publicationOct. 14, 1992;acceptedJan. 6, 1993. Reprintrequests:Dr. Ian W.Black,Departmentof CardiologyF-15,Cleveland ClinicFoundation,9500EuclidAve.,ClevelandOH 44195. Copyright© 1993by Mosby-YearBook,Inc. 0002-8703/93/$1.00+ .10 4/1/47113
Embolism following cardioversion is assumed to result from dislodgment of atrial t h r o m b i after the r e t u r n of atrial mechanical activity. 2,3,12 Transesophageal echocardiography ( T E E ) is sensitive and specific for the detection of left atrial thrombi. 13-15 T h e purpose of this s t u d y was to d e t e r m i n e whether screening patients for atrial t h r o m b i by T E E m a y allow cardioversion to be p e r f o r m e d safely in patients not receiving anticoagulation. We prospectively evaluated the echocardiographic findings and clinical outcome in a consecutive series of nonanticoagulated patients screened by T E E before cardioversion of nonvalvular atrial fibrillation or flutter. METHODS Patients. Inclusion and exclusion criteria for the study
patients were as follows: (1) nonvalvular atrial fibrillation or atrial flutter, (2) referral for elective electrical cardioversion, (3) not receiving anticoagulant therapy, and (4) no history of embolism. The decision not to prescribe anticoagulant therapy was determined by the referring physician and was not random. The baseline characteristics of the 40 patients meeting the study criteria are shown in Table I. Protocol. No patient was receiving anticoagulant therR7~
376
August 1993 American Heart Journal
Black et al.
Table I. Patient characteristics Characteristics
Age (yr) (range) M/F Atrial fibrillation/flutter Cause of arrhythmia Systemic hypertension Idiopathic dilated cardiomyopathy Alcohol ingestion Postoperative Lone arrhythmia Ischemic heart disease Hypertrophic cardiomyopathy Other Duration of arrhythmia 0-1 mo 1-12 mo 1-5 yr Previous cardioversion Left ventricular function Normal Mild dysfunction Moderate dysfunction Severe dysfunction
62 + 13 (28-79) 29/11 33/7 7 4 6 6 4 6 1 6 21 16 3 3 28 6 2 4
apy before the study. After the initial transthoracic echocardiography, all patients underwent TEE. Patients in whom atrial thrombi were not detected proceeded to cardioversion. Patients in whom cardioversion was successful were not given anticoagulant therapy after the procedure, whereas patients in whom cardioversion was not successful received anticoagulant therapy subsequently at the discretion of the referring physician. If atrial thrombi were detected by TEE, cardioversion was not performed and the patients were begun on a regimen of anticoagulant therapy. Echocardiography. Two-dimensional and Doppler transthoracic echocardiography, with 2.5 MHz or 3.5 MHz imaging transducers (HP 77020A, Hewlett-Packard Co., Medical Products Group, Andover, Mass.), was performed immediately before TEE. Left atrial dimension and left ventricular diastolic dimension were determined by standard criteria. 16 Left ventricular systolic function was assessed by transthoracic two-dimensional echocardiography 17 and graded qualitatively as normal or mild, moderate, or severe overall impairment. TEE was performed with a 5 MHz single-plane (n = 27) or biplane (n = 13) phasedarray transducer (HP 21362A and HP 21363A). Written informed consent was obtained, and patients were examined in the fasting state, after sedation With intravenous midazolam (1.8 + 1.2 mg) and fentanyl (83 + 19/~g); pharyngeal anesthesia was induced with topical lidocaine. The probe was introduced by means of standard technique, and previously described tomographic planes were obtained, is The presence of atrial thrombus and spontaneous echo contrast was assessed specifically. Atrial thrombus was diagnosed by the presence of a clearly defined intracavitary mass, acoustically distinct from underlying endocardium, and was not due to the pectinate ridges of the atrial ap-
pendage. 19 Spontaneous echo contrast was diagnosed by the presence of dynamic, smokelike echoes with a characteristic swirling motion, distinct from echoes caused by exCessive gain. 2° TEE duration was 10 to 20 minutes, and technically adequate images were obtained from all patients. Echocardiographic findings were recorded immediately and were therefore blinded to the results of cardioversion. There were no complications. Cardioversion. Cardioversion was performed in the coronary care unit after sedation with intravenous thiopentone or propofol. Both the anterolateral and anteroposterior paddle positions were used. Synchronized direct-current shocks were administered with increasing discharge energy until cardioversion was either successful or unsuccessful at a maximum of 360 joules. After the procedure, patients were monitored for at least 4 hours before they were discharged from the coronary care unit and assessed clinically for evidence of embolism. Successful cardioversion was defined as the presence of sinus rhythm at discharge from the coronary care unit. All patients were followed for 1 month after cardioversion. Cardioversionrelated embolism was defined prospectively as any clinically evident acute cerebral or systemic ischemic event or pulmonary embolism during the 1-month period after cardioversion. Data collection and analYsis. All clinical and echocardiographic data were recorded prospectively. Clinical data were determined from patient and physician interviews and chart review. Transthoracic and transesophageal echocardiographic studies were performed and interpreted independently by experienced echocardiographers. Categoric variables were compared by means of chi-square or Fisher's exact test, and continuous variables were compared by means of the unpaire d t test, with statistical significance defined as p < 0.05. Continuous variables were reported as the mean + standard deviation.
RESULTS Echocardiographic findings before cardioversion. T r a n s t h o r a c i c echocardiography did not detect atrial t h r o m b u s in any patient. In contrast, T E E d e m o n strated atrial t h r 0 m b i i n 5 (12%) of 40 patients (p = 0.03; Fig. 1), all in the left atrial appendage. T h e clinical and echocardiographic characteristics of these five patients are shown in Table II. Left atrial t h r o m b i were f o u n d in 4 (12%) of 33 patients with atrial fibrillation and in one (14%) of seven with atrial flutter (p = NS). None of the left atrial t h r o m b i were mobile. T h e single p a t i e n t with left atrial t h r o m b u s and atrial flutter h a d undergone continuous E C G monitoring for 7 days before T E E and did not have atrial fibrillation recorded at any time. T r a n s t h o r a c i c echocardiography did not detect atrial s p o n t a n e o u s echo contrast in any patient. T E E detected left atrial s p o n t a n e o u s echo contrast in 14 (35 % ) of 40 patients (p < 0.001), including 11 (33 % ) of 33 with atrial fibrillation and three (43%) of seven
Volume 126, Number 2 American Heart Journal
Black et al.
377
Fig. 1. Transesophageal echocardiogram in candidate for cardioversion showing thrombus (T) at apex of left atrial appendage (LAA). AO, Aorta; LA, left atrium; PV, pulmonary vein.
Table II. Findings in five patients with atrial thrombus before cardioversion Age (yr)
Sex
Rhythm
Cause
Duration
28 62 79 48 71
F M M M M
AFib AFib AFlutter AFib AFib
HCM IDC IHD HT Postop
7 days 15 days 14 days 6 wk 7 days
LA (mm)
L V (ram)
L V dysfunction
Thrombus
LA SEC
50 56 49 56 41
36 69 61 58 57
Normal Severe Severe Moderate Normal
LAA LAA LAA LAA LAA
+ + + + -
AFib, Atrial fibrillation; AFlutter, atrial flutter; HCM, hypertrophic cardiomyopathy; IDC, idiopathic dilated cardiomyopathy; IHD, ischemic heart disease; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; Postop, postoperative; SEC, spontaneous echo contrast; HT, hypertension.
with atrial flutter (p = NS). TEE also detected right atrial spontaneous echo contrast in 1 (2 % ) of 40 patients (p = NS). The characteristics of patients with and without left atrial thrombi are shown in Table III. There was no significant difference between the two groups with regard to age, duration or type of arrhythmia, or cardiac dimensions. However, patients with left atrial thrombus were more likely to have moderate or severe left ventricular dysfunction (p = 0.02) and left atrial spontaneous echo contrast (p = 0.04). Cardioversion. Cardioversion was cancelled and anticoagulant therapy was begun in the five patients with atrial thrombi. Two other patients with postoperative atrial fibrillation (mean duration 6 days) in whom atrial thrombi were not detected reverted spontaneously to sinus rhythm without embolic complications within 12 hours after TEE, before scheduled cardioversion. Cardioversion was performed within 1 hour after
TEE in 25 of the 33 remaining patients. Eight patients underwent cardioversion 1 day after TEE. Cardioversion was successful in 25 (76%) of 33 patients, including 19 (70%) of 27 with atrial fibrillation and all six patients with atrial flutter. Fellow-up. Two of the five patients with left atrial thrombi detected by TEE, in whom cardioversion was cancelled, reverted spontaneously to sinus rhythm without embolic complications I and 10 days, respectively, after TEE, while receiving anticoagulant therapy. Repeat TEE in the latter patient 5 days after reversion, and after 3 months in two of the three other patients with thrombi, showed no change in the appearance of the thrombi. One patient had a cerebral embolism after cardioversion. This 75-year-old man with atrial fibrillation of 4 weeks' duration had a history of remote myocardial infarction and congestive heart failure without previous embolism. Transthoracic echocardiography showed left ventricular dilation (diastolic dimension
August 1993
378
Black et al.
AmericanHeartJournal
T a b l e Ill. Characteristics of patients with and without left atrial thrombi Characteristics
L A thrombus (n = 5)
No L A thrombus (n = 35)
p Value
Age (yr) A r r h y t h m i a duration <1 mo Atrial flutter LA dimension (mm) LV diastolic dimension (mm) LA spontaneous echo contrast Moderate-severe LV dysfunction
58 _+ 20 4 (80%) I (20 % ) 50 + 6 56 _+ 12 4 (80 % ) 3 (60 % )
63 _+ 12 17 (49%) 6 (17 % ) 46 _+ 8 52 _ 7 10 (29 % ) 3 (9 % )
NS NS NS NS NS 0.04 0.02
LA, Left atrial; LV, left ventricular; NS, not significant.
62 ram) and severe left ventricular systolic dysfunction without left ventricular thrombus or aneurysm. The left atrial dimension was 54 mm. Biplane TEE clearly visualized the left atrium and appendage and showed mild left atrial spontaneous echo contrast but no thrombus (Fig. 2, A). Cardioversion was successfully performed 1 hour after TEE. Twenty-four hours after cardioversion the patient suddenly had dysphasia and right hemiplegia. ECG showed sinus rhythm, and cerebral computed tomography 2 hours after embolism showed no hemorrhage or infarction. Intravenous heparin was begun, and transthoracic echocardiography and T E E were repeated 4 days later. T E E showed a fresh thrombus attached to the lateral wall of the left atrial appendage (Fig. 2, B), associated with marked left atrial spontaneous echo contrast. There was no evidence of left ventricular thrombus, patent foramen ovale, or significant aortic atheroma. Results of ultrasonographic imaging of the carotid artery were normal. There were no embolic events in the other 39 patients during the month after cardioversion. DISCUSSION
Results in recent clinical trials have accentuated the risk of embolism in patients with nonvalvular atrial fibrillation. 21 These trials have stimulated interest in cardioversion, which is also associated with a risk of embolism from atrial thrombi. 1~ Although nonrandomized case series suggest that the incidence of embolism after cardioversion may be reduced in patients receiving anticoagulant therapy, 4-7 there remains concern about the risk of bleeding, especially in the elderly population. 22 In recent atrial fibrillation trials, 21 the incidence of major bleeding including cerebral hemorrhage was up to 3.5% per year in patients receiving warfarin. Furthermore, these studies excluded patients considered to have an increased risk of bleeding. Therefore there remain a significant number of patients with atrial fibrillation or flutter who do not undergo anticoagulation and in whom
cardioversion without the need for several weeks of anticoagulant therapy would be desirable. This is the first study to evaluate the role of T E E in screening nonanticoagulated patients for atrial thrombi before cardioversion for nonvalvular atrial fibrillation and flutter. T E E improved the identification of patients with left atrial appendage thrombi in whom cardioversion was deferred. However, this study has shown that embolism may still occur after cardioversion in nonanticoagulated patients, despite screening by TEE. Detection
and risk factors for left atrial thrombi.
Transthoracic echocardiography did not detect left atrial thrombus in any patient in the present study. There are no reports of left atrial thrombus detected by transthoracic echocardiography in candidates for cardioversion, despite use of the technique in several studies.5,11 In contrast, left atrial appendage thrombi were detected by TEE in 5 (12%) of 40 patients, which is consistent with previous reports. ~3-15 Careful retrospective review of TEE findings in the patient with cerebral embolism, in whom the left atrium and appendage were clearly visualized, showed no evidence of preexisting thrombus. Nevertheless, the left atrial appendage is not optimally visualized in all patients, and it may be difficult to distinguish the pectinate ridges of the appendage from small or recently formed thrombi. T E E may also be used to monitor patients in whom left atrial thrombi have been detected. Although no change in the appearance of the thrombus after anticoagulation was observed in the present study, the usefulness of TEE in demonstrating resolution of left atrial appendage thrombi has been described previously. 23 Left ventricular dysfunction and left atrial spontaneous echo contrast detected by TEE were significantly associated with the presence of left atrial thrombus in this series. The patient in whom thromboembolism occurred after cardioversion had both of these risk factors. Although left ventricular systolic
Volume 126, Number 2 American Heart Journal
Black et al.
379
Fig. 2. Transesophageal echocardiograms in patient with cerebral embolism after cardioversion. A, Before cardioversion. Left atrial appendage is clear of thrombus. B, After embolism. Thrombus (arrows) is present on lateral wall of left atrial appendage. Abbreviations as in Fig. 1.
dysfunction did not emerge as a risk factor for embolism in previous cardioversion series, 1-3 left ventricular dysfunction in patients with atrial fibrillation was associated with increased embolic risk in the Stroke Prevention in Atrial Fibrillation (SPAF) study. 24 As in the SPAF study, left ventricular dysfunction in the present study was assessed qualitatively. The single patient in this series with hypertrophic cardiomyopathy was also found to have a left atrial thrombus. The increased embolic risk of cardioversion in these patients has been noted previously. 1° The criteria for this study excluded patients with rheumatic heart disease or previous embolism, which was also associated with increased risk of thromboembolism.24, 25 The primary purpose of TEE before cardioversion is the detection of formed left atrial thrombi. However, TEE also improved the detection of left atrial spontaneous echo contrast in the present series. We have previously reported an association between left
atrial spontaneous echo contrast and thromboembolism in patients with nonvalvular atrial fibrillation. 2° We have also shown 26 that left atrial spontaneous echo contrast in patients with nonvalvular atrial fibrillation is associated with altered hematologic parameters, indicating a hypercoagulable state in addition to stasis. Thus the detection of left atrial thrombi and/or spontaneous echo contrast by TEE may identify patients at increased risk of embolism after cardioversion. Embolism after cardioversion. Embolism after car-
dioversion of atrial fibrillation and flutter is usually attributed to dislodgment of preexisting atrial thrombus following the return of atrial mechanical activity.2, 3, 12 However, the return of mechanical function in the atrial cavity may be delayed several days to weeks after restoration of electrical sinus rhythm. 27, 2s Recently Grimm et al. 29 reported that left atrial spontaneous echo contrast may be generated or intensified immediately after successful cardiover-
August 1993
380
Black et al.
sion. Grimm et al. 3° also reported that left atrial appendage flow velocities are decreased immediately after cardioversion, despite conversion from atrial fibrillation to sinus rhythm. These findings suggest that persistent or increased atrial stasis early after cardioversion may favor de novo thrombosis and potential embolism. The relevance of this mechanism of thromboembolism is supported by the occurrence of embolism several days after successful cardioversion 4, 7 and by the increased risk of embolism attributed to recently formed atrial thrombi. 12 This mechanism may account for the cerebral embolism in one patient after cardioversion despite screening by T E E in the present series. No source of embolism other than de novo left atrial appendage thrombus after cardioversion was detected in this patient. The thrombus was attached to the lateral wall of the appendage, a frequent location for left atrial appendage thrombi caused by relative immobility. 31 The increased intensity of left atrial spontaneous echo contrast after successful cardioversion in this patient is also consistent with increased stasis and risk of thrombosis. The results of the present study and those of Grimm et al.29, 30 therefore suggest that the exclusion by T E E of preexisting atrial thrombus before cardioversion does not preclude embolism after cardioversion, because of persistent atrial stasis and de novo thrombosis. Further studies will be required to determine whether anticoagulation after cardioversion can prevent left atrial thrombosis during this period. It is of interest that embolism did not occur in two patients with documented atrial thrombi in the present series, who reverted to sinus rhythm 1 and 10 days, respectively, after beginning anticoagulant therapy. The latter case is consistent with Goldman's 12 hypothesis that anticoagulant therapy given for 2 weeks before cardioversion may stabilize a preexisting thrombus and thereby reduce the likelihood of embolism at the time of reversion. The incidence of embolism after cardioversion is 1% to 5 %, and the prevalence of left atrial thrombi in nonvalvular atrial fibrillation is approximately 10 %. Therefore the majority of left atrial thrombi, as in the former case, do not embolize after cardioversion. Risk factors for embolization of left atrial thrombi, such as mobility, site, duration and morphology, require further study. Embolism after cardioversion of atrial flutter is well recognized, 32 although it is considered a lower risk than atrial fibrillation. Three (43 % ) of seven patients with atrial flutter in the present series had left atrial spontaneous echo contrast, including one patient with left ventricular dysfunction in whom a left atrial thrombus was found. We have previously
American Heart Journal
reported 3u that both left atrial spontaneous echo contrast and left atrial thrombi may occur in patients with atrial flutter, with a lower incidence than in patients with atrial fibrillation. Prolonged atrial dysfunction after cardioversion of atrial flutter has also been reported. 34 These data suggest that patients with atrial flutter may have impaired atrial function and also merit screening for atrial thrombi before cardioversion. Limitations. The present study comprised nonanticoagulated patients only and did not directly address the risks and benefits of anticoagulant therapy in patients undergoing cardioversion. Since the risk of embolism is small, a much larger study would be required to determine precisely the risk of embolism in nonanticoagulated patients screened by TEE. Left atrial appendage flow velocities, which provide an additional assessment of atrial mechanical function, ul were not measured in the present study. Conclusions. This study has shown that TEE improves the detection of left atrial appendage thrombi in candidates for cardioversion, in whom the procedure may be deferred. TEE also enables serial monitoring of the effect of anticoagulant therapy on left atrial thrombi and improves detection of left atrial spontaneous echo contrast. However, the study shows that there remains a risk of embolism in nonanticoagulated patients despite screening for atrial thrombi by TEE. These findings and those of other recent studies suggest that the absence of detectable atrial thrombus before cardioversion does not preclude de novo thrombosis resulting from persistent or increased atrial stasis following a return to sinus rhythm. A randomized, controlled clinical trial is required to determine the relative risk of embolism after cardioversion in patients conventionally managed with anticoagulant therapy compared with patients screened by TEE, with or without anticoagulant therapy. REFERENCES
1. De Silva RA, Graboys TB, Podrid PJ, Lown B. Cardioversion and defibrillation. AM HEART J 1980;100:881-95. 2. Mancini GBJ, Goldberger AL. Cardioversion of atrial fibrillation: consideration of embolization, anticoagulation, prophylactic pacemaker, and long-term success. AM HEART J 1982; 104:617-21. 3. Stein B, Halperin JL, Fuster V. Should patients with atrial fibrillation be anticoagulated prior to and chronically following cardioversion? Cardiovasc Clin 1990;21:231-49. 4. Bjerkelund CJ, Orning OM. The efficacy of anticoagulant therapy in preventing embolism related to DC electrical conversion of atrial fibrillation. Am J Cardiol 1969;23:208-15. 5. Weinberg DM, Mancini GBJ. Anticoagulation for cardioversion of atrial fibrillation. Am J Cardiol 1989;63:745-6. 6. Lown B. Electrical reversion of cardiac arrhythmias. Br Heart J 1967;29:469-89. 7. Arnold AZ, Mick MJ, Mazurek RP, Loop FD, Trohman RG. Role of prophylactic anticoagulation for direct current car-
Volume 126, Number 2 American Heart Journal
dioversion in patients with atrial fibrillation or flutter. J Am Coll Cardiol 1992;19:851-5. 8. Resnekov L, McDonald L. Complications in 220 patients with cardiac dysrhythmias treated by phased direct current shock, and indications for electroconversion. Br Heart J 1967;29:92636. 9. Aberg H, Cullhed I. Direct current countershock complications. Acta Med Scand 1968;183:415-21. 10. Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, Epstein SE. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;53:273-9. 11. Lesser MF. Safety and efficacy of in-office cardioversion for treatment of supraventricular arrhythmias. Am J Cardiol 1990;66:1267-8. 12. Goldman MJ. The management of chronic atrial fibrillation: indications for and method of conversion to sinus rhythm. Prog Cardiovasc Dis 1960;2:465-79. 13. Matsumura M, Shash P, Kyo S, Omoto R. Advantages of transesophageal echo for correct diagnosis on small left atrial thrombiin mitral stanosis [Abstract]. Circulation 1989;80(suppl II):II-678. 14. Acar J, Cormier B, Grimberg D, Kawthekar G, Iung B, Scheuer B, Farah E: Diagnosis of left atrial thrombi in mitral stenosis--usefulness of ultrasound techniques compared with other methods. Eur Heart J 1991;12(suppl B):70-6. 15. Lin SL, Hsu TL, Liou Jy, Chert CH, Chang MS, Chiang HT, Chen CY. Usefulness of transesophageal echocardiography for the detection of left atrial thrombi in patients with rheumatic heart disease. Echocardiography 1992;6:161-8. 16. Sahn DJ, De Maria A, Kisslo J, Weyman A. Recommendations regarding quantification in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58:1072-82. 17. Feigenbaum H, Echocardiography. 4th ed. Philadelphia: Lea & Febiger, 1986;144-8. 18. Seward JB, Khandheria BK, OH JK, Abel MD; Hughes RW Jr, Edwards WD, Nichols BA, Freeman WK, Tajik AJ. Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988; 63:649-80. 19. Beppu S, Park YD, Sakakibara H, Nagata S, Nimura Y. Clinical features of intracardiac thrombosis based on echocardiographic observation. Jpn Circ J 1984;48:75-82. 20. Black IW, Hopkins AP, Lee LCL, Walsh WF. Left atrial spontaneous echo contrast: a clinical and echoeardiographic analysis. J Am Coll Cardiol 1991;18:398-404. 21. Cairns JA, Connolly SJ. Nonrheumatic atrial fibrillation. Risk of stroke and role of antithrombotic therapy. Circulation 1991; 84:469-81.
Black et al.
381
22. Kutner M, Nixon G, Silverstone F. Physicians' attitudes toward oral anticoagulants and antiplatelet agents for stroke prevention in elderly patients with atrial fibrillation. Arch Intern Med 1991;151:1950-3. 23. Tsai LM, Hung JS, Chen JH, Lin LJ, Fu M. Resolution of left atrial appendage thrombus in mitral stenosis after warfarin therapy. AM HEARTJ 1991;121:1232-4. 24. The Stroke Prevention in Atrial Fibrillation Investigators. Predictors of thromboembolism in atrial fibrillation. II. Echocardiographic features of patients at risk. Ann Intern Med 1992;116:6-12. 25. Wolf PA, Dawber TR, Thomas HE, Kannel WB. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: the Framingham study. Neurology 1978;28:973-7. 26. Black I T, Chesterman CN, Hopkins AP, Lee LCL, Chong BH, Walsh WF. Hematologic correlates of left atrial spontaneous echo contrast and thromboembolism in non-valvular atrial fibrillation. J Am Coll Cardiol 1993;21:451-7. 27. Shapiro EP, Effron MB, Lima S, Ouyang P, Siu CO, Bush D. Transient atrial dysfunction after conversion of chronic atrial fibrillation to sinus rhythm. Am J Cardiol 1988;62:1202-7. 28. Manning WJ, Leeman DE, Gotch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation. J Am Coll Cardiol 1989; 13:617-23. 29. Grimm RA, Klein AL, Cohen GI, Maloney JD, Sgarbossa E, Pearce GL, Stewart WJ, Salcedo EE. Concurrent transesophageal echo and electrical cardioversion in patients with atrial fibrillation: effect of cardioversion on left atrial smoke [Abstract]. J Am Coll Cardiol 1992;19:155A. 30. Grimm RA, Klein AL, Cohen GI, Maloney JD, Stewart WJ, Salcedo EE. Return of left atrial appendage function post electrical cardioversion of atrial fibrillation by transesophageal echo [Abstract]. J Am Coll Cardiol 1992;19:156A. 31. Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography. Implications for the development of thrombus. Circulation 1991;84:223-31. 32. Roy D, Marchand E, Gagne P, Chabot M, Cartier R. Usefulness of anticoagulant therapy in the prevention of embolic complications of atrial fibrillation. AMHEARTJ 1986; 112:103943. 33. Black IW, Hopkins AP, Lee LCL, Walsh WF. Thromboembolic risk of atrial flutter [Abstract]. J Am Coll Cardiol 1992;19:314A. 34. Jordaens L, Germonpre E, Vandenbogaerde J. Atrial stunning lasts more than one week after conversion of atrial flutter [Abstract]. J Am Coll Cardiol 1991;17:325A.
Ian W. Black, M B B S , Andrew P. Hopkins, MBBS, Lincoln C. L. Lee, M B B S , and W a r r e n F. Walsh, M B B S Sydney, Australia
Electrical cardioversion of atrial fibrillation and atrial flutter is p e r f o r m e d to relieve symptoms, improve cardiac function, and reduce the potential for subsequent thromboembolism. Cardioversion itself m a y be complicated by cerebral, systemic, and p u l m o n a r y embolism, which occurs in up to 5.6 % of patients. 1-3 Despite the lack of randomized controlled trials, findings in some studies 4-7 suggest a reduction in the incidence of embolism after cardioversion in patients receiving anticoagulant therapy. However, prophylactic anticoagulation does not eliminate the risk of embolism, 4' 8-11 m a y be associated with a delay in cardioversion, additional cost, and potential morbidity, and consequently is not always used.
Fromthe Departmentof CardiovascularMedicine,PrinceHenryHospital. Supportedin part bythe PrinceHenryHospitalCentenaryResearchFund. Receivedfor publicationOct. 14, 1992;acceptedJan. 6, 1993. Reprintrequests:Dr. Ian W.Black,Departmentof CardiologyF-15,Cleveland ClinicFoundation,9500EuclidAve.,ClevelandOH 44195. Copyright© 1993by Mosby-YearBook,Inc. 0002-8703/93/$1.00+ .10 4/1/47113
Embolism following cardioversion is assumed to result from dislodgment of atrial t h r o m b i after the r e t u r n of atrial mechanical activity. 2,3,12 Transesophageal echocardiography ( T E E ) is sensitive and specific for the detection of left atrial thrombi. 13-15 T h e purpose of this s t u d y was to d e t e r m i n e whether screening patients for atrial t h r o m b i by T E E m a y allow cardioversion to be p e r f o r m e d safely in patients not receiving anticoagulation. We prospectively evaluated the echocardiographic findings and clinical outcome in a consecutive series of nonanticoagulated patients screened by T E E before cardioversion of nonvalvular atrial fibrillation or flutter. METHODS Patients. Inclusion and exclusion criteria for the study
patients were as follows: (1) nonvalvular atrial fibrillation or atrial flutter, (2) referral for elective electrical cardioversion, (3) not receiving anticoagulant therapy, and (4) no history of embolism. The decision not to prescribe anticoagulant therapy was determined by the referring physician and was not random. The baseline characteristics of the 40 patients meeting the study criteria are shown in Table I. Protocol. No patient was receiving anticoagulant therR7~
376
August 1993 American Heart Journal
Black et al.
Table I. Patient characteristics Characteristics
Age (yr) (range) M/F Atrial fibrillation/flutter Cause of arrhythmia Systemic hypertension Idiopathic dilated cardiomyopathy Alcohol ingestion Postoperative Lone arrhythmia Ischemic heart disease Hypertrophic cardiomyopathy Other Duration of arrhythmia 0-1 mo 1-12 mo 1-5 yr Previous cardioversion Left ventricular function Normal Mild dysfunction Moderate dysfunction Severe dysfunction
62 + 13 (28-79) 29/11 33/7 7 4 6 6 4 6 1 6 21 16 3 3 28 6 2 4
apy before the study. After the initial transthoracic echocardiography, all patients underwent TEE. Patients in whom atrial thrombi were not detected proceeded to cardioversion. Patients in whom cardioversion was successful were not given anticoagulant therapy after the procedure, whereas patients in whom cardioversion was not successful received anticoagulant therapy subsequently at the discretion of the referring physician. If atrial thrombi were detected by TEE, cardioversion was not performed and the patients were begun on a regimen of anticoagulant therapy. Echocardiography. Two-dimensional and Doppler transthoracic echocardiography, with 2.5 MHz or 3.5 MHz imaging transducers (HP 77020A, Hewlett-Packard Co., Medical Products Group, Andover, Mass.), was performed immediately before TEE. Left atrial dimension and left ventricular diastolic dimension were determined by standard criteria. 16 Left ventricular systolic function was assessed by transthoracic two-dimensional echocardiography 17 and graded qualitatively as normal or mild, moderate, or severe overall impairment. TEE was performed with a 5 MHz single-plane (n = 27) or biplane (n = 13) phasedarray transducer (HP 21362A and HP 21363A). Written informed consent was obtained, and patients were examined in the fasting state, after sedation With intravenous midazolam (1.8 + 1.2 mg) and fentanyl (83 + 19/~g); pharyngeal anesthesia was induced with topical lidocaine. The probe was introduced by means of standard technique, and previously described tomographic planes were obtained, is The presence of atrial thrombus and spontaneous echo contrast was assessed specifically. Atrial thrombus was diagnosed by the presence of a clearly defined intracavitary mass, acoustically distinct from underlying endocardium, and was not due to the pectinate ridges of the atrial ap-
pendage. 19 Spontaneous echo contrast was diagnosed by the presence of dynamic, smokelike echoes with a characteristic swirling motion, distinct from echoes caused by exCessive gain. 2° TEE duration was 10 to 20 minutes, and technically adequate images were obtained from all patients. Echocardiographic findings were recorded immediately and were therefore blinded to the results of cardioversion. There were no complications. Cardioversion. Cardioversion was performed in the coronary care unit after sedation with intravenous thiopentone or propofol. Both the anterolateral and anteroposterior paddle positions were used. Synchronized direct-current shocks were administered with increasing discharge energy until cardioversion was either successful or unsuccessful at a maximum of 360 joules. After the procedure, patients were monitored for at least 4 hours before they were discharged from the coronary care unit and assessed clinically for evidence of embolism. Successful cardioversion was defined as the presence of sinus rhythm at discharge from the coronary care unit. All patients were followed for 1 month after cardioversion. Cardioversionrelated embolism was defined prospectively as any clinically evident acute cerebral or systemic ischemic event or pulmonary embolism during the 1-month period after cardioversion. Data collection and analYsis. All clinical and echocardiographic data were recorded prospectively. Clinical data were determined from patient and physician interviews and chart review. Transthoracic and transesophageal echocardiographic studies were performed and interpreted independently by experienced echocardiographers. Categoric variables were compared by means of chi-square or Fisher's exact test, and continuous variables were compared by means of the unpaire d t test, with statistical significance defined as p < 0.05. Continuous variables were reported as the mean + standard deviation.
RESULTS Echocardiographic findings before cardioversion. T r a n s t h o r a c i c echocardiography did not detect atrial t h r o m b u s in any patient. In contrast, T E E d e m o n strated atrial t h r 0 m b i i n 5 (12%) of 40 patients (p = 0.03; Fig. 1), all in the left atrial appendage. T h e clinical and echocardiographic characteristics of these five patients are shown in Table II. Left atrial t h r o m b i were f o u n d in 4 (12%) of 33 patients with atrial fibrillation and in one (14%) of seven with atrial flutter (p = NS). None of the left atrial t h r o m b i were mobile. T h e single p a t i e n t with left atrial t h r o m b u s and atrial flutter h a d undergone continuous E C G monitoring for 7 days before T E E and did not have atrial fibrillation recorded at any time. T r a n s t h o r a c i c echocardiography did not detect atrial s p o n t a n e o u s echo contrast in any patient. T E E detected left atrial s p o n t a n e o u s echo contrast in 14 (35 % ) of 40 patients (p < 0.001), including 11 (33 % ) of 33 with atrial fibrillation and three (43%) of seven
Volume 126, Number 2 American Heart Journal
Black et al.
377
Fig. 1. Transesophageal echocardiogram in candidate for cardioversion showing thrombus (T) at apex of left atrial appendage (LAA). AO, Aorta; LA, left atrium; PV, pulmonary vein.
Table II. Findings in five patients with atrial thrombus before cardioversion Age (yr)
Sex
Rhythm
Cause
Duration
28 62 79 48 71
F M M M M
AFib AFib AFlutter AFib AFib
HCM IDC IHD HT Postop
7 days 15 days 14 days 6 wk 7 days
LA (mm)
L V (ram)
L V dysfunction
Thrombus
LA SEC
50 56 49 56 41
36 69 61 58 57
Normal Severe Severe Moderate Normal
LAA LAA LAA LAA LAA
+ + + + -
AFib, Atrial fibrillation; AFlutter, atrial flutter; HCM, hypertrophic cardiomyopathy; IDC, idiopathic dilated cardiomyopathy; IHD, ischemic heart disease; LA, left atrium; LAA, left atrial appendage; LV, left ventricle; Postop, postoperative; SEC, spontaneous echo contrast; HT, hypertension.
with atrial flutter (p = NS). TEE also detected right atrial spontaneous echo contrast in 1 (2 % ) of 40 patients (p = NS). The characteristics of patients with and without left atrial thrombi are shown in Table III. There was no significant difference between the two groups with regard to age, duration or type of arrhythmia, or cardiac dimensions. However, patients with left atrial thrombus were more likely to have moderate or severe left ventricular dysfunction (p = 0.02) and left atrial spontaneous echo contrast (p = 0.04). Cardioversion. Cardioversion was cancelled and anticoagulant therapy was begun in the five patients with atrial thrombi. Two other patients with postoperative atrial fibrillation (mean duration 6 days) in whom atrial thrombi were not detected reverted spontaneously to sinus rhythm without embolic complications within 12 hours after TEE, before scheduled cardioversion. Cardioversion was performed within 1 hour after
TEE in 25 of the 33 remaining patients. Eight patients underwent cardioversion 1 day after TEE. Cardioversion was successful in 25 (76%) of 33 patients, including 19 (70%) of 27 with atrial fibrillation and all six patients with atrial flutter. Fellow-up. Two of the five patients with left atrial thrombi detected by TEE, in whom cardioversion was cancelled, reverted spontaneously to sinus rhythm without embolic complications I and 10 days, respectively, after TEE, while receiving anticoagulant therapy. Repeat TEE in the latter patient 5 days after reversion, and after 3 months in two of the three other patients with thrombi, showed no change in the appearance of the thrombi. One patient had a cerebral embolism after cardioversion. This 75-year-old man with atrial fibrillation of 4 weeks' duration had a history of remote myocardial infarction and congestive heart failure without previous embolism. Transthoracic echocardiography showed left ventricular dilation (diastolic dimension
August 1993
378
Black et al.
AmericanHeartJournal
T a b l e Ill. Characteristics of patients with and without left atrial thrombi Characteristics
L A thrombus (n = 5)
No L A thrombus (n = 35)
p Value
Age (yr) A r r h y t h m i a duration <1 mo Atrial flutter LA dimension (mm) LV diastolic dimension (mm) LA spontaneous echo contrast Moderate-severe LV dysfunction
58 _+ 20 4 (80%) I (20 % ) 50 + 6 56 _+ 12 4 (80 % ) 3 (60 % )
63 _+ 12 17 (49%) 6 (17 % ) 46 _+ 8 52 _ 7 10 (29 % ) 3 (9 % )
NS NS NS NS NS 0.04 0.02
LA, Left atrial; LV, left ventricular; NS, not significant.
62 ram) and severe left ventricular systolic dysfunction without left ventricular thrombus or aneurysm. The left atrial dimension was 54 mm. Biplane TEE clearly visualized the left atrium and appendage and showed mild left atrial spontaneous echo contrast but no thrombus (Fig. 2, A). Cardioversion was successfully performed 1 hour after TEE. Twenty-four hours after cardioversion the patient suddenly had dysphasia and right hemiplegia. ECG showed sinus rhythm, and cerebral computed tomography 2 hours after embolism showed no hemorrhage or infarction. Intravenous heparin was begun, and transthoracic echocardiography and T E E were repeated 4 days later. T E E showed a fresh thrombus attached to the lateral wall of the left atrial appendage (Fig. 2, B), associated with marked left atrial spontaneous echo contrast. There was no evidence of left ventricular thrombus, patent foramen ovale, or significant aortic atheroma. Results of ultrasonographic imaging of the carotid artery were normal. There were no embolic events in the other 39 patients during the month after cardioversion. DISCUSSION
Results in recent clinical trials have accentuated the risk of embolism in patients with nonvalvular atrial fibrillation. 21 These trials have stimulated interest in cardioversion, which is also associated with a risk of embolism from atrial thrombi. 1~ Although nonrandomized case series suggest that the incidence of embolism after cardioversion may be reduced in patients receiving anticoagulant therapy, 4-7 there remains concern about the risk of bleeding, especially in the elderly population. 22 In recent atrial fibrillation trials, 21 the incidence of major bleeding including cerebral hemorrhage was up to 3.5% per year in patients receiving warfarin. Furthermore, these studies excluded patients considered to have an increased risk of bleeding. Therefore there remain a significant number of patients with atrial fibrillation or flutter who do not undergo anticoagulation and in whom
cardioversion without the need for several weeks of anticoagulant therapy would be desirable. This is the first study to evaluate the role of T E E in screening nonanticoagulated patients for atrial thrombi before cardioversion for nonvalvular atrial fibrillation and flutter. T E E improved the identification of patients with left atrial appendage thrombi in whom cardioversion was deferred. However, this study has shown that embolism may still occur after cardioversion in nonanticoagulated patients, despite screening by TEE. Detection
and risk factors for left atrial thrombi.
Transthoracic echocardiography did not detect left atrial thrombus in any patient in the present study. There are no reports of left atrial thrombus detected by transthoracic echocardiography in candidates for cardioversion, despite use of the technique in several studies.5,11 In contrast, left atrial appendage thrombi were detected by TEE in 5 (12%) of 40 patients, which is consistent with previous reports. ~3-15 Careful retrospective review of TEE findings in the patient with cerebral embolism, in whom the left atrium and appendage were clearly visualized, showed no evidence of preexisting thrombus. Nevertheless, the left atrial appendage is not optimally visualized in all patients, and it may be difficult to distinguish the pectinate ridges of the appendage from small or recently formed thrombi. T E E may also be used to monitor patients in whom left atrial thrombi have been detected. Although no change in the appearance of the thrombus after anticoagulation was observed in the present study, the usefulness of TEE in demonstrating resolution of left atrial appendage thrombi has been described previously. 23 Left ventricular dysfunction and left atrial spontaneous echo contrast detected by TEE were significantly associated with the presence of left atrial thrombus in this series. The patient in whom thromboembolism occurred after cardioversion had both of these risk factors. Although left ventricular systolic
Volume 126, Number 2 American Heart Journal
Black et al.
379
Fig. 2. Transesophageal echocardiograms in patient with cerebral embolism after cardioversion. A, Before cardioversion. Left atrial appendage is clear of thrombus. B, After embolism. Thrombus (arrows) is present on lateral wall of left atrial appendage. Abbreviations as in Fig. 1.
dysfunction did not emerge as a risk factor for embolism in previous cardioversion series, 1-3 left ventricular dysfunction in patients with atrial fibrillation was associated with increased embolic risk in the Stroke Prevention in Atrial Fibrillation (SPAF) study. 24 As in the SPAF study, left ventricular dysfunction in the present study was assessed qualitatively. The single patient in this series with hypertrophic cardiomyopathy was also found to have a left atrial thrombus. The increased embolic risk of cardioversion in these patients has been noted previously. 1° The criteria for this study excluded patients with rheumatic heart disease or previous embolism, which was also associated with increased risk of thromboembolism.24, 25 The primary purpose of TEE before cardioversion is the detection of formed left atrial thrombi. However, TEE also improved the detection of left atrial spontaneous echo contrast in the present series. We have previously reported an association between left
atrial spontaneous echo contrast and thromboembolism in patients with nonvalvular atrial fibrillation. 2° We have also shown 26 that left atrial spontaneous echo contrast in patients with nonvalvular atrial fibrillation is associated with altered hematologic parameters, indicating a hypercoagulable state in addition to stasis. Thus the detection of left atrial thrombi and/or spontaneous echo contrast by TEE may identify patients at increased risk of embolism after cardioversion. Embolism after cardioversion. Embolism after car-
dioversion of atrial fibrillation and flutter is usually attributed to dislodgment of preexisting atrial thrombus following the return of atrial mechanical activity.2, 3, 12 However, the return of mechanical function in the atrial cavity may be delayed several days to weeks after restoration of electrical sinus rhythm. 27, 2s Recently Grimm et al. 29 reported that left atrial spontaneous echo contrast may be generated or intensified immediately after successful cardiover-
August 1993
380
Black et al.
sion. Grimm et al. 3° also reported that left atrial appendage flow velocities are decreased immediately after cardioversion, despite conversion from atrial fibrillation to sinus rhythm. These findings suggest that persistent or increased atrial stasis early after cardioversion may favor de novo thrombosis and potential embolism. The relevance of this mechanism of thromboembolism is supported by the occurrence of embolism several days after successful cardioversion 4, 7 and by the increased risk of embolism attributed to recently formed atrial thrombi. 12 This mechanism may account for the cerebral embolism in one patient after cardioversion despite screening by T E E in the present series. No source of embolism other than de novo left atrial appendage thrombus after cardioversion was detected in this patient. The thrombus was attached to the lateral wall of the appendage, a frequent location for left atrial appendage thrombi caused by relative immobility. 31 The increased intensity of left atrial spontaneous echo contrast after successful cardioversion in this patient is also consistent with increased stasis and risk of thrombosis. The results of the present study and those of Grimm et al.29, 30 therefore suggest that the exclusion by T E E of preexisting atrial thrombus before cardioversion does not preclude embolism after cardioversion, because of persistent atrial stasis and de novo thrombosis. Further studies will be required to determine whether anticoagulation after cardioversion can prevent left atrial thrombosis during this period. It is of interest that embolism did not occur in two patients with documented atrial thrombi in the present series, who reverted to sinus rhythm 1 and 10 days, respectively, after beginning anticoagulant therapy. The latter case is consistent with Goldman's 12 hypothesis that anticoagulant therapy given for 2 weeks before cardioversion may stabilize a preexisting thrombus and thereby reduce the likelihood of embolism at the time of reversion. The incidence of embolism after cardioversion is 1% to 5 %, and the prevalence of left atrial thrombi in nonvalvular atrial fibrillation is approximately 10 %. Therefore the majority of left atrial thrombi, as in the former case, do not embolize after cardioversion. Risk factors for embolization of left atrial thrombi, such as mobility, site, duration and morphology, require further study. Embolism after cardioversion of atrial flutter is well recognized, 32 although it is considered a lower risk than atrial fibrillation. Three (43 % ) of seven patients with atrial flutter in the present series had left atrial spontaneous echo contrast, including one patient with left ventricular dysfunction in whom a left atrial thrombus was found. We have previously
American Heart Journal
reported 3u that both left atrial spontaneous echo contrast and left atrial thrombi may occur in patients with atrial flutter, with a lower incidence than in patients with atrial fibrillation. Prolonged atrial dysfunction after cardioversion of atrial flutter has also been reported. 34 These data suggest that patients with atrial flutter may have impaired atrial function and also merit screening for atrial thrombi before cardioversion. Limitations. The present study comprised nonanticoagulated patients only and did not directly address the risks and benefits of anticoagulant therapy in patients undergoing cardioversion. Since the risk of embolism is small, a much larger study would be required to determine precisely the risk of embolism in nonanticoagulated patients screened by TEE. Left atrial appendage flow velocities, which provide an additional assessment of atrial mechanical function, ul were not measured in the present study. Conclusions. This study has shown that TEE improves the detection of left atrial appendage thrombi in candidates for cardioversion, in whom the procedure may be deferred. TEE also enables serial monitoring of the effect of anticoagulant therapy on left atrial thrombi and improves detection of left atrial spontaneous echo contrast. However, the study shows that there remains a risk of embolism in nonanticoagulated patients despite screening for atrial thrombi by TEE. These findings and those of other recent studies suggest that the absence of detectable atrial thrombus before cardioversion does not preclude de novo thrombosis resulting from persistent or increased atrial stasis following a return to sinus rhythm. A randomized, controlled clinical trial is required to determine the relative risk of embolism after cardioversion in patients conventionally managed with anticoagulant therapy compared with patients screened by TEE, with or without anticoagulant therapy. REFERENCES
1. De Silva RA, Graboys TB, Podrid PJ, Lown B. Cardioversion and defibrillation. AM HEART J 1980;100:881-95. 2. Mancini GBJ, Goldberger AL. Cardioversion of atrial fibrillation: consideration of embolization, anticoagulation, prophylactic pacemaker, and long-term success. AM HEART J 1982; 104:617-21. 3. Stein B, Halperin JL, Fuster V. Should patients with atrial fibrillation be anticoagulated prior to and chronically following cardioversion? Cardiovasc Clin 1990;21:231-49. 4. Bjerkelund CJ, Orning OM. The efficacy of anticoagulant therapy in preventing embolism related to DC electrical conversion of atrial fibrillation. Am J Cardiol 1969;23:208-15. 5. Weinberg DM, Mancini GBJ. Anticoagulation for cardioversion of atrial fibrillation. Am J Cardiol 1989;63:745-6. 6. Lown B. Electrical reversion of cardiac arrhythmias. Br Heart J 1967;29:469-89. 7. Arnold AZ, Mick MJ, Mazurek RP, Loop FD, Trohman RG. Role of prophylactic anticoagulation for direct current car-
Volume 126, Number 2 American Heart Journal
dioversion in patients with atrial fibrillation or flutter. J Am Coll Cardiol 1992;19:851-5. 8. Resnekov L, McDonald L. Complications in 220 patients with cardiac dysrhythmias treated by phased direct current shock, and indications for electroconversion. Br Heart J 1967;29:92636. 9. Aberg H, Cullhed I. Direct current countershock complications. Acta Med Scand 1968;183:415-21. 10. Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, Epstein SE. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;53:273-9. 11. Lesser MF. Safety and efficacy of in-office cardioversion for treatment of supraventricular arrhythmias. Am J Cardiol 1990;66:1267-8. 12. Goldman MJ. The management of chronic atrial fibrillation: indications for and method of conversion to sinus rhythm. Prog Cardiovasc Dis 1960;2:465-79. 13. Matsumura M, Shash P, Kyo S, Omoto R. Advantages of transesophageal echo for correct diagnosis on small left atrial thrombiin mitral stanosis [Abstract]. Circulation 1989;80(suppl II):II-678. 14. Acar J, Cormier B, Grimberg D, Kawthekar G, Iung B, Scheuer B, Farah E: Diagnosis of left atrial thrombi in mitral stenosis--usefulness of ultrasound techniques compared with other methods. Eur Heart J 1991;12(suppl B):70-6. 15. Lin SL, Hsu TL, Liou Jy, Chert CH, Chang MS, Chiang HT, Chen CY. Usefulness of transesophageal echocardiography for the detection of left atrial thrombi in patients with rheumatic heart disease. Echocardiography 1992;6:161-8. 16. Sahn DJ, De Maria A, Kisslo J, Weyman A. Recommendations regarding quantification in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 1978;58:1072-82. 17. Feigenbaum H, Echocardiography. 4th ed. Philadelphia: Lea & Febiger, 1986;144-8. 18. Seward JB, Khandheria BK, OH JK, Abel MD; Hughes RW Jr, Edwards WD, Nichols BA, Freeman WK, Tajik AJ. Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 1988; 63:649-80. 19. Beppu S, Park YD, Sakakibara H, Nagata S, Nimura Y. Clinical features of intracardiac thrombosis based on echocardiographic observation. Jpn Circ J 1984;48:75-82. 20. Black IW, Hopkins AP, Lee LCL, Walsh WF. Left atrial spontaneous echo contrast: a clinical and echoeardiographic analysis. J Am Coll Cardiol 1991;18:398-404. 21. Cairns JA, Connolly SJ. Nonrheumatic atrial fibrillation. Risk of stroke and role of antithrombotic therapy. Circulation 1991; 84:469-81.
Black et al.
381
22. Kutner M, Nixon G, Silverstone F. Physicians' attitudes toward oral anticoagulants and antiplatelet agents for stroke prevention in elderly patients with atrial fibrillation. Arch Intern Med 1991;151:1950-3. 23. Tsai LM, Hung JS, Chen JH, Lin LJ, Fu M. Resolution of left atrial appendage thrombus in mitral stenosis after warfarin therapy. AM HEARTJ 1991;121:1232-4. 24. The Stroke Prevention in Atrial Fibrillation Investigators. Predictors of thromboembolism in atrial fibrillation. II. Echocardiographic features of patients at risk. Ann Intern Med 1992;116:6-12. 25. Wolf PA, Dawber TR, Thomas HE, Kannel WB. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: the Framingham study. Neurology 1978;28:973-7. 26. Black I T, Chesterman CN, Hopkins AP, Lee LCL, Chong BH, Walsh WF. Hematologic correlates of left atrial spontaneous echo contrast and thromboembolism in non-valvular atrial fibrillation. J Am Coll Cardiol 1993;21:451-7. 27. Shapiro EP, Effron MB, Lima S, Ouyang P, Siu CO, Bush D. Transient atrial dysfunction after conversion of chronic atrial fibrillation to sinus rhythm. Am J Cardiol 1988;62:1202-7. 28. Manning WJ, Leeman DE, Gotch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation. J Am Coll Cardiol 1989; 13:617-23. 29. Grimm RA, Klein AL, Cohen GI, Maloney JD, Sgarbossa E, Pearce GL, Stewart WJ, Salcedo EE. Concurrent transesophageal echo and electrical cardioversion in patients with atrial fibrillation: effect of cardioversion on left atrial smoke [Abstract]. J Am Coll Cardiol 1992;19:155A. 30. Grimm RA, Klein AL, Cohen GI, Maloney JD, Stewart WJ, Salcedo EE. Return of left atrial appendage function post electrical cardioversion of atrial fibrillation by transesophageal echo [Abstract]. J Am Coll Cardiol 1992;19:156A. 31. Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography. Implications for the development of thrombus. Circulation 1991;84:223-31. 32. Roy D, Marchand E, Gagne P, Chabot M, Cartier R. Usefulness of anticoagulant therapy in the prevention of embolic complications of atrial fibrillation. AMHEARTJ 1986; 112:103943. 33. Black IW, Hopkins AP, Lee LCL, Walsh WF. Thromboembolic risk of atrial flutter [Abstract]. J Am Coll Cardiol 1992;19:314A. 34. Jordaens L, Germonpre E, Vandenbogaerde J. Atrial stunning lasts more than one week after conversion of atrial flutter [Abstract]. J Am Coll Cardiol 1991;17:325A.