Left Ventricular Thrombi Identified by Cross-Sectional Echocardiography

Transcription

1 Left Ventricular Thrombi Identified by Cross-Sectional Echocardiography ANTHONY N. DeMARIA, M.D.; WILLIAM BOMMER, M.D.; ALEXANDER NEUMANN, B.S.; TODD GREHL, M.D.; LYNN WEINART, B.S.; SALLY DeNARDO, M.D.; EZRA A. AMSTERDAM, M.D.; and DEAN T. MASON, M.D.; Davis and Sacramento, California We studied 25 patients with anterior myocardial infarction and two with congestive cardiomyopathy to evaluate twodimensional echocardiography in the diagnosis of left ventricular thrombi. Five coronary patients had systemic emboli. Four of these patients manifested apical filling defects on cineangiogram, while a levophase cine was equivocal for clot in the fifth patient. Neither echocardiography nor cineangiography visualized ventricular thrombi in the nonembolus coronary patients. Echoes from a distinct apical mass, however, were visualized in all five patients in the embolus group by two-dimensional echocardiography at the cardiac apex. Apical thrombi were confirmed in all four patients in the embolus group undergoing surgery. The irregular configuration of recent thrombi in the coronary patients differed from the circumscribed appearance of chronic thrombi in the cardiomyopathy patients on two-dimensional echocardiogram. Thus, two-dimensional echocardiography can be used to detect and characterize left ventricular thrombi. LEFT VENTRICULAR THROMBI represent an important, potentially catastrophic complication of acute myocardial infarction (1, 2). Mural thrombi have been detected in 20% to 60% of patients at autopsy after transmural infarction (3-7), and systemic emboli have been seen in 3% to 5% of patients after an episode of myocardial necrosis (7-9). In the past the identification of left ventricular clots in the postinfarction period required invasive diagnostic precedures such as left ventricular cineangiography. The advent of M-mode echocardiography was accompanied by the anticipation that ultrasound would provide a noninvasive diagnostic technique capable of detecting intracardiac thrombi. The ability of M-mode echocardiography to identify cardiac clots with an acceptable sensitivity and specificity, however, was never established (10). The failure of M-mode echocardiography to identify cardiac thrombi was primarily due to the lack of spatial orientation inherent in this technique, which impaired visualization of the cardiac apex, the site of most cardiac clots (11), and prevented the definition of a mass of echoes as a thrombus rather than a technical artifact. Recently, cross-sectional or two-dimensional echocardiographic techniques have been developed that provide spatial orientation and enable the imaging of the entire left ventricle, including the cardiac apex (12-16). Our study represents an evaluation of the ability of two-dimensional echocardiography to detect left ventricular thrombi in a group of patients with transmural myocardial infarction. From the Section of Cardiovascular Medicine, Departments of Medicine and Physiology, University of California School of Medicine, Davis, and University Medical Center, Sacramento; California. 14 Annals of Internal Medicine 90:14-18, 1979 Methods Twenty-five patients, all with transmural myocardial infarction documented by conventional electrocardiographic and enzymatic methods, form the basis of our study. The group consisted of 16 men and nine women ranging in age from 38 to 67 years, with a mean age of 54 years. Five of the 25 patients manifested documented episodes of systemic emboli and were classified as the embolus group. Since all five embolus patients had infarctions of the anterior wall of the left ventricle, the remaining 20 patients represent persons who also had anterior infarction but did not manifest evidence of emboli and who were selected to serve as control subjects. Two-dimensional echocardiograms were also available for examination in two additional patients with congestive cardiomyopathy but without evidence of systemic emboli in whom the presence of left ventricular thrombi was confirmed by cardiac catheterization or autopsy. Table 1 shows important clinical features of patients in the embolus group. The sex and age distribution of these patients was not significantly different from those patients with anterior infarction who did not manifest evidence of systemic emboli. Two of the five patients had evidence of multiple emboli, and the femoral artery was the most frequent site of occlusion in this group. The interval between myocardial infarction and the appearance of vascular obstruction secondary to emboli ranged from 10 to 150 days. One of the patients in the embolus group had recurrent episodes of ventricular fibrillation, while all patients had severe congestive heart failure. Four of the five patients ultimately underwent cardiac surgery, including coronary artery bypass grafting and aneurysmectomy. All 25 postmyocardial infarction patients underwent both M- mode and two-dimensional echocardiography as well as left ventricular cineangiography. Echocardiography was done within 24 h of cineangiography except in one patient in the embolus group in whom these procedures were separated by 4 months. M-mode echocardiographic sweeps from the body to the base of the left ventricle were done in all patients in the standard manner by rocking the ultrasonic transducer along the course of the longitudinal axis of this chamber. M-mode echograms were done with a commercially available echocardiograph (Smith- Kline Instruments, Sunnyvale, California) interfaced to either a fiberoptic (Honeywell Corporation, Denver, Colorado) or a photographic oscilloscope recorder (Irex Corporation, Saddle River, New Jersey). M-mode records were carefully examined for intracavitary echoes indicative of left ventricular clot. Two-dimensional echocardiograms were obtained with either a dynamically focused phased array echograph with a sector arc of 70 deg (Grumman Corporation, Woodbury, New York) or a wide-angle mechanical sector scanner with an arc of 80 deg (Smith-Kline). In all instances cross-sectional echocardiograms providing ultrasonic tomograms of the left ventricle in different projections were obtained from three different transducer placements. The long-axis view was obtained with the transducer positioned at the left sternal border and the ultrasonic arc aligned along the longitudinal axis of the left ventricle extending from the aortic valve leaflets to the apex (Figure 1, top). The short-axis view was obtained by rotating the transducer 90 deg so that the sector arc was aligned along a plane perpendicular to the longitudinal axis. Finally, the four-chamber view of the 1979 American College of Physicians

2 heart was obtained with the ultrasonic transducer rotated as if recording the short-axis view but positioned at the point of maximal cardiac impulse and directed toward the right shoulder so as to provide a tomographic view of all four cardiac chambers simultaneously (Figure 1, bottom). Cross-sectional echocardiograms were subjected to qualitative analysis consisting of a careful search for intracavitary echoes indicative of left ventricular thrombus. In addition, two-dimensional echocardiograms were examined for abnormalities of left ventricular contraction by analyzing wall motion along four ventricular chords obtained from the four-chamber view: a long-axis chord that bisects the chamber from mitral leaflets to apex, and three equidistant perpendicular short-axis chords. Biplane left ventricular cineangiography was done by injecting 0.75 to 1 ml/kg of Hypaque 75 (Winthrop Laboratories, New York, New York) into the left ventricle except in one patient with emboli in whom the ventricle was opacified by levophase flow after pulmonary artery injection of contrast. Cineangiograms were recorded on 35 mm-film using a 22.9-cm image intensification system. Direct visual examination either at the time of surgery or autopsy was done in four of the five patients in the embolus group, as well as the patient with cardiomyopathy in whom cineangiography was not done. Results CINEANGIOGRAPHY Although prominent areas of anterior-apical akinesis or dyskinesis were identified in all control coronary patients on left ventricular cineangiogram, no evidence of intracardiac filling defects was observed in any of these 20 ischemic heart disease patients. Akinesis or dyskinesis of the cardiac apex was also observed in all five patients with systemic emboli. In addition, left ventricular cineangiograms were judged to manifest unequivocal evidence of intracavitary filling defects secondary to cardiac thrombi in four of the five emboli patients. In the fifth patient in the embolus group, the ventricle was opacified on the levophase of a pulmonary artery injection and the ventriculogram was interpreted to demonstrate probable, although not unequivocal, evidence of left ventricular clot. In each individual the thrombus was located at the apex of the left ventricle. Left ventricular cineangiography revealed evidence of an apical filling defect in the one patient with congestive cardiomyopathy in whom this procedure was done. SURGERY AND NECROPSY Neither surgical nor necropsy specimens were available for examination in any of the 20 control coronary patients. Four of the five embolus patients, including the one whose levophase cineangiogram was equivocal for the presence of clot, underwent aneurysmectomy that Table 1. < Clinical Featui es of Five Patients witt i Systemic Ei mboli Patient Age yrs Sex Site Time After Infarction days Ultimate Therapy 1 54 M Brachial and 150 Surgery femoral 2 52 F Renal 21 Surgery 3 56 M Mesenteric-femoral 10 Surgery 4 67 F Femoral 92 Surgery 5 45 M Femoral 19 Medicine Figure 1 Top. Stop-frame long-axis projection of two-dimensional echocardiogram from a normal subject in our hospital. The apex is not well visualized. RV = right ventricle; LV = left ventricle; LA = left atrium; AO = aorta. Bottom. Four-chamber view of the heart obtained with the ultrasound transducer positioned at the point of maximal cardiac impulse and directed toward the right shoulder of a normal subject. No intracavitary echoes are observed in the cardiac apex, which is narrow and angular. LV = left ventricle; LA = left atrium; RV = right ventricle; RA = right atrium. confirmed the presence of friable, fresh apical left ventricular thrombus in each patient. In addition, a highly organized, immobile mural left ventricular clot was found in the left ventricle at necropsy examination in the patient with cardiomyopathy who did not undergo cineangiography. ECHOCARDIOGRAPHY Technically adequate M-mode echographic sweeps from the body to the base of the left ventricle were obtained in all patients. However, in no patient was the apex of the left ventricle imaged during the standard examination. Accordingly, the presence of clot was not detected in any patient by M-mode echocardiogram. Technically adequate two-dimensional echocardiograms of the left ventricle in each projection were also obtained in all patients. As was true of M-mode echocardiography, the apex of the left ventricle was not completely imaged with the transducer positioned along the left sternal border to obtain either the long- or short-axis projections. Accordingly, the presence of a thrombus was detected by long-axis two-dimensional echogram in only one of the five embolus patients and in no patient in DeMariaetal. Echocardiography for Ventricular Thrombi 15

3 of left ventricular thrombi in each of the five patients in the embolus group (Figure 2, bottom, and Figure 3, top). The echographic abnormality consisted of a dense mass of echoes with a shaggy irregular border. In the four patients who manifested recent evidence of systemic emboli, random unrestricted motion about the apex of the left ventricle of portions of the clot was observed during real time two-dimensional study. Thereby, the shaggy irregular border and random motion observed by the thrombi on real time two-dimensional echocardiography in patients in the embolus group suggested the continuing propagation of friable clot. Evidence of apical cardiac thrombus was also observed on four-chamber view by cross-sectional ultrasound in the two patients with congestive cardiomyopathy documented to have mural left ventricular thrombi (Figure 3, bottom). The configuration of the apical clots in the cardiomyopathy patients, however, differed from those observed in the postmyocardial infarction patients in that the margins of the clot were rounded and well circumscribed (Figure 3, bottom). Further, on real time ultrasonic study, little, if any, motion of these thrombi was noted. Thus the smooth immobile masses in the patients with cardiomyopathy suggested chronic, organized thrombotic lesions. Figure 2 Top. Four-chamber projection of two-dimensional echocardiogram from a patient with coronary artery disease but no systemic embolus. The apex of the left ventricle (LV) bulges abnormally, but no intracavitary echoes indicative of a left ventricular thrombus are observed. RV = right ventricle. Bottom. Four-chamber view of two-dimensional echocardiogram from one of the patients with systemic emboli. The apex of the left ventricle (LV) again bulges abnormally, and an intense mass of intracavitary echoes secondary to a left ventricular thrombus (T) is well visualized. The left ventricular thrombus is shaggy in appearance and has irregular margins. RV = right ventricle. short-axis view. The cardiac apex was well visualized by the four-chamber view in all patients, however, and thereby this ultrasonic projection formed the basis of the detection of left ventricular thrombi in this study. The four-chamber two-dimensional echogram revealed total absence of motion or paradoxical systolic expansion in either the longitudinal chord or the short-axis chord located at the cardiac apex in all 20 control patients with previous anterior wall myocardial infarction (Figure 2, top). Although ultrasonic artifacts were transiently observed in the apex of several nonembolus patients on twodimensional echocardiogram, a consistent mass of echoes indicative of left ventricular thrombus was not found in any patient in this group (Figure 2, top). Absent or paradoxical systolic motion involving the two hemichords traversing the left ventricular apex was also detected on two-dimensional echocardiogram in all five patients manifesting systemic emboli (Figure 2, bottom, and Figure 3, top). In addition, the four-chamber two-dimensional ultrasonic projection revealed evidence Discussion This study documents the ability of cross-sectional or two-dimensional echocardiography to identify left ventricular thrombi in patients with a transmural myocardial infarction. Consistent dense masses of intracavitary echoes were observed in the left ventricular apex by two-dimensional ultrasound in all five patients manifesting systemic emboli (Figure 2, bottom, and Figure 3, top). Further, the presence and apical location of these thrombi were confirmed in all four patients who underwent aneurysmectomy, including one patient in whom evidence of a left ventricular thrombus was equivocal on a levophase left ventricular cineangiogram. Two-dimensional echocardiography also predicted the presence of thrombi in two patients with congestive cardiomyopathy in whom the existence of left ventricular clots was subsequently documented by either cineangiography or necropsy (Figure 3, bottom). In contrast, evidence of left ventricular thrombus was not observed in any of the 20 control patients with anterior infarction without evidence of systemic emboli (Figure 2, top). Although the incidence of mural thrombi after acute myocardial infarction has ranged from 20% to 60% at autopsy examination (3-7), clinical evidence of systemic emboli are observed in less than 5% of patients after transmural myocardial infarction (7-9). Since clots were detected only in postinfarction patients manifesting emboli in our study, cross-sectional echocardiography may be able to detect only the largest of left ventricular thrombi. Nevertheless, this ability would still be of marked clinical importance, inasmuch as thrombi that ultimately result in systemic emboli are presumably those largest in size and propagated farthest into the left ventricle. The determination of the true sensitivity and specific- 16 January 1979 Annals of Internal Medicine Volume 90 Number 1

4 ity of two-dimensional echocardiography in the detection of left ventricular thrombi, however, will await a prospective study in which cross-sectional echocardiograms are obtained and compared with necropsy in a large group of consecutive patients after myocardial infarction. A recent study reported that mural thrombi were detected by contrast cineangiography in only 41% of patients in whom these lesions were subsequently identified at surgery (17). These investigators concluded that angiographic evidence is reasonably specific but somewhat insensitive in the diagnosis of cardiac thrombus. Thus, cineangiography constitutes an imperfect method for the detection of mural thrombi, although, as is true of ultrasound, this technique may be highly sensitive in identifying large intracavitary thrombi. Nevertheless, as compared with cineangiography, echocardiography eliminates the risks of intracardiac catheterization, may be done at bedside or in the coronary care unit, and may be repeated after therapy. Although it was not our intent to characterize the clinical features of systemic emboli after acute myocardial infarction, several important observations were evident. The longest interval between infarction and embolus was 150 days, and no vascular occlusion occurred before 10 days. Thus it appears that emboli are typically late complications of infarction and are related to thrombus propagation secondary to stasis rather than to the acute myocardial injury. All patients with systemic emboli had major infarction manifested by multiple Q waves on electrocardiogram, high levels of cardiac enzymes in the serum, and significant congestive heart failure and arrythmias clinically. Accordingly, such patients seem to constitute a high-risk group for embolic complications of myocardial infarction. Finally, in regard to therapy, a decision to perform cardiac surgery was reached in four of the five embolus patients based on multiple considerations, including the presence of congestive heart failure and arrythmias. Three of the four patients who underwent operation survived surgery and are alive at 6 months, although congestive heart failure continues as a problem in two. Thus, although the relative roles of anticoagulation and surgical therapy for ventricular thrombi remain to be defined, it appears that surgery can be successfully done in selected patients. Except for isolated case reports (17-20), previous attempts to use M-mode echocardiography in the detection of cardiac thrombi have been unsuccessful (10). However, cross-sectional echocardiography, with its ability to provide spatial orientation, has enabled examination of the entire left ventricular circumference and provided the ability to distinguish ultrasonic artifacts from true intracavitary masses. Conglomerations or layers of echoes of uncertain nature on M-mode recordings can be identified as thrombi by virtue of the definition of form and consistency of motion observed on two-dimensional echograms. The ability to diagnose reliably intracardiac thrombi represents a new application of cardiac ultrasound made possible by cross-sectional techniques. The data obtained in this study indicate the ability of cross-sectional echocardiography to detect abnormalities of cardiac contrac- Figure 3 Top. Four-chamber view of two-dimensional echocardiogram from an additional patient in the embolus group. An apical aneurysm and left ventricular thrombus are easily visualized. The echocardiogram was obtained with a wide-angle mechanical sector scanner. LV = left ventricle; RV = right ventricle; LA = left atrium. Bottom. Four-chamber view of two-dimensional echocardiogram from a patient with congestive cardiomyopathy in whom the presence of a left ventricular thrombus (7) was documented by left ventricular cineangiography. The thrombus is well circumscribed and has smooth edges. LV = left ventricle; LA = left atrium. tion after acute myocardial infarction as well. Previous studies have shown that the presence of left ventricular thrombus after an acute myocardial infarction is related to the size of the infarction; that is, the larger the area of myocardial necrosis the greater the incidence of mural clot (4). In addition, the presence of mural clots is substantially increased in the setting of left ventricular aneurysms (5, 6). Anterior wall infarctions typically result in a greater area of necrotic myocardium than do infarctions in other areas. In addition, the opportunity for hemostasis provided by the dyskinetic apex, which frequently occurs secondary to anterior infarction, also favors both the appearance and propagation of left ventricular thrombi. In our study systemic emboli were observed only in patients with anterior wall myocardial infarctions, and thrombi were confined to the left ventricular apex in each person. These data emphasize the value of imaging the left ventricular apex in the detection of postinfarction thrombi. In this regard, the four-chamber projection of the two-dimensional echocardiogram per- DeMaria etal. Echocardiography for Ventricular Thrombi 17

5 formed with the transducer at the cardiac apex has proved to be an invaluable addition to echocardiography and has enabled the identification of ventricular thrombi in four patients in whom the standard long and short-axis projections obtained from the left sternal border did not yield evidence of these lesions. The four-chamber view of the heart has now become a standard part of the ultrasound examination of all patients in our laboratory. Ventricular thrombi in the postmyocardial infarction patients were observed to have shaggy irregular borders and marked mobility during cardiac contraction on twodimensional echocardiogram (Figure 2, bottom, and Figure 3, top). This configuration suggested that there was recent propagation of the thrombus rendering it friable and likely to embolize, a fact confirmed by clinical and surgical observations. In contrast, the patients with congestive cardiomyopathy had no evidence of systemic emboli, and left ventricular clots were observed as incidental findings during echocardiographic examination. In both patients with cardiomyopathy the thrombi appeared well circumscribed and relatively immobile on echocardiogram, a situation subsequently confirmed in the one patient who had necropsy examination (Figure 3, bottom). Thus, although our observations are based on small numbers of patients, our data do suggest that the configuration of thrombi recorded by two-dimensional echocardiography may be potentially valuable in indicating the degree of organization and therefore the risk of subsequent emboli from these lesions. In addition to documenting the ability of two-dimensional echocardiography to detect left ventricular thrombi, our study has several potential clinical implications regarding the management of patients with acute myocardial infarction. Systemic emboli are observed in 3% to 5% of infarction patients, and the prevention of these events would constitute a major advance in the management of this disorder. That systemic emboli occurred after an interval of 10 days or more in all patients in our study seems to indicate that sufficient time is available to initiate medical therapy if mural thrombi are identified early in the postinfarction period. Two-dimensional echocardiography may enable the early identification of mural thrombi. Further, ultrasound may provide an ideal method by which to evaluate the response of left ventricular clots to anticoagulant therapy. Since patients with large anterior wall myocardial infarctions and congestive heart failure constitute a high-risk group for systemic emboli, an important topic for future research would be whether two-dimensional echocardiography can identify postinfarction patients in whom the administration of anticoagulants will prevent embolic events. ACKNOWLEDGMENTS: The authors thank Thomas Barstow for technical expertise and Betty Paro for secretarial assistance. Grant support: in part by Research Program Project Grant HL from the National Heart, Lung, and Blood Institute and Research Grants from the California Golden Empire Chapter of the American Heart Association. Requests for reprints should be addressed to Anthony N. DeMaria, M.D.; Section of Cardiovascular Medicine, University of California School of Medicine; Davis, CA Received 3 April 1978; revision accepted 3 October References 1. HURST JW, LOGUE RB: The Heart, 3rd ed. New York, McGraw-Hill, MELTZER LE, DUNNING AJ: Textbook of Coronary Care. Bowie, Maryland, Charles Press Publications, YATER WM, WELSH PP, STAPLETON JF, CLARK ML: Comparison of clinical and pathologic aspects of coronary artery disease in men of various age groups: a study of 950 autopsied cases from the Armed Forces Institute of Pathology. Ann Intern Med 34: , JORDAN RA, MILLER RD, EDWARDS JE, PARKER RL: Thromboembolism in acute and in healed myocardial infarction. I. Intracardiac mural thrombosis. Circulation 6:1-6, PHARES WS, EDWARDS JE, BURCHELL HB: Cardiac aneurysms: clinicopathologic studies. Proc Staff Meet Mayo Clin 28: , SCHLICHTER J, HELLERSTEIN HK, KATZ LN: Aneurysm of the heart: a correlative study of 102 proved cases. Medicine (Baltimore) 33:43-78, HILDEN T, IVERSEN K, RAASCHOU F, SCHWARTZ M: Anticoagulants in acute myocardial infarction. Lancet 2: , Report of the Working Party on Anticoagulant Therapy in Coronary Thrombosis to the Medical Research Council: assessment of short-term anticoagulant administration after cardiac infarction. Br Med J 1: , VETERANS ADMINISTRATION HOSPITAL INVESTIGATORS: Anticoagulants in acute myocardial infarction. Results of a cooperative clinical trial. JAMA 225: , FEIGENBAUM H: Echocardiography, 2nd ed. Philadelphia, Lea & Febiger, HAMBY RI, WISOFF BG, DAVISON ET, HARTSTEIN ML: Coronary artery disease and left ventricular mural thrombi: clinical, hemodynamic and angiocardiographic aspects. Chest 66: , ROELANDT J, KLOSTER FE, TEN CATE FJ, VAN DORP WG, HONKOOP J, BOM N, HUGENHOLTZ PG: Multidimensional echocardiography. An appraisal of its clinical usefulness. Br Heart J 36:29-43, GRIFFITH JM, HENRY WL: A sector scanner for real time two-dimensional echocardiography. Circulation 49: , KISSLO J, VONRAMM OT, THURSTONE FL: Cardiac imaging using a phased array ultrasound system. II. Clinical technique and application. Circulation 53: , WEYMAN AE, FEIGENBAUM H, HURWITZ RA, GIROD DA, DILLON JC, CHANG S: Localization of left ventricular outflow obstruction by cross-sectional echocardiography. Am J Med 60:33-38, SCHILLER N, SILVERMAN N: Apex echocardiography: a new method of imaging the heart using a phased array real time two-dimensional 80 sector scanner (abstract). Am J Cardiol 39:266, SWAN HJC, MAGNUSSON PT, BUCHBINDER NA, MATLOFF JM, GRAY RJ: Aneurysm of the cardiac ventricle. Its management by medical and surgical intervention. West J Med 129:26-40, HORGAN JH, SHIEL FO, GOODMAN AC: Demonstration of left ventricular thrombus by conventional echocardiography. / Clin Ultrasound 4: , TALLURY K, DEPASQUALE ND: Ultrasound cardiography in the diagnosis of left atrial thrombus. Chest 59: , SPANGLER RD, OKIN JT: Echocardiographic demonstration of a left atrial thrombus. Chest 67: , January 1979 Annals of Internal Medicine Volume 90 Number 1