Comparison of Transthoracic and Transesophageal Three-dimensional Echocardiography in Adult Patients

J Clin Ultrasound 22:381-389, July/August 1994 CCC 0091-2751/94/060381-09 0 1994 by John Wiley & Sons, Inc.

Comparison of Transthoracic and Transesophageal Echocardiography in the Diagnosis of Intracardiac Tumors in Adults Kou-Gi Shyu, MD, Jin-Jer Chen, MD, JunJack Cheng, MD, JueyJen Hwang, MD, Peiliang Kuan, MD, and Wen-Pin Lien, MD

Abstract: To compare transthoracic and transesophageal echocardiographyin the diagnosis of intracardiac tumors, 17 patients (8 men and 9 women, aged 19 years to 67 years) whose intracardiac tumors were detected by echocardiography were studied in a 4-year period. Of these, 14 patients underwent cardiac surgery and 13 were proved to have tumors. There were 4 false-positive and 2 false-negative diagnoses by transthoracic echocardiography,but only one false-positive and no false-negative diagnosis by transesophageal echocardiography.The stalk of a myxoma was detected clearly in 5 of 11 patients by transthoracic echocardiography,whereas in 10 of 11 it was detected by transesophageal echocardiography.The detailed morphologic characteristics of the tumor, such as contour of the tumor, and the presence of cysts and calcification in the tumor, were seen more clearly with transesophageal echocardiographythan with transthoracic echocardiography.0 1994 John Wiley & Sons, Inc. Indexing Words: Transthoracic echocardiography . Transesophageal echocardiography . Cardiac tumor ' Myxoma

Primary intracavitary tumors of the heart are unbut surgical treatment offers a potential cure.4 Early diagnosis is essential so that the tumors can be promptly removed to reduce morbidity and mortality. In the past, antemorten di.~ agnosis of these tumors was i n f r e q ~ e n tToday, they can be detected readily by noninvasive imaging Two-dimensional echocardiography has become the technique of choice for the detection of intracardiac turn or^.'^-^^ However, in some patients, the precordial echo window may be poor because of chest deformity, obesity, or emphysema, and inadequate image quality will be obtained. Thus, no reliable diagnosis can be made. Transesophageal echocardiography, a new window to the heart, provides optimal imaging quality for most cardiac structures because of the From the Division of Cardiology, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan, Republic of China. For reprints contact Peiliang Kuan, MD, Department of Internal Medicine, National Taiwan University Hospital, No. 7, Chung-Shan South Road, Taipei 100, Taiwan, R.O.C. VOL. 22, NO. 6, JULY/AUGUST 1994

close proximity of the esophagus to the heart.l49l5 It has become a well-established tool for evaluating cardiovascular disease^.'^'^' However, reports on the comparison of transthoracic and transesophageal echocardiography in the diagnosis of intracardiac tumors are few because of the rarity of these turn or^.^'^^^ In this study, we compared transthoracic and transesophageal echocardiography in the diagnosis of primary intracardiac tumors in a 4-year period. PATIENTS AND METHODS

Patients Between October 1988 and August 1992, 17 pa-

tients whose intracardiac tumors were detected by echocardiography were studied. Clinically, only 3 patients were suspected to have cardiac tumors because their clinical manifestations were related to body-position change. The other intracardiac tumors were diagnosed incidentally when the patients were referred for echocardiography. The initial diagnosis of the 14 patients before 381

SHYU ET AL.

echocardiography was congestive heart failure in 5 ; hypertension, palpitation, and mitral insufficiency each in 2; and mitral valve prolapse, mitral stenosis, and peripheral embolism each in l. All 17 patients underwent both transthoracic and transesophageal echocardiography. There were 8 men and 9 women, ranging in age from 19 years to 67 years, with a mean age of 48 years. The case records of all 17 patients were reviewed. Case 11 has been described in a previous case report.20 Echocardiographic Imaging All echocardiograms were recorded using a Toshiba SSH-65A or Aloka SSD-870 commercially available ultrasound system with a 2.5MHz or 3.75-MHz precordial transducer and a 3.75-MHz or 5.0-MHz transesophageal transducer. The review of all echocardiograms for tumor location, size, point of attachment, mobility, and morphologic characteristics was performed separately by two observers without knowledge of the surgical or other confirmatory data. These echocardiographic parameters were checked and measured during the operation if the patient underwent cardiac surgery. Images from both transthoracic and transesophageal examinations were stored on 3/4-in. videotape for playback. All transthoracic examinations included standard parasternal, apical, and subxiphoidal transducer positions. Doppler flow mapping studies were also concomitantly obtained. Continuouswave Doppler was obtained using a 2.5-MHz transducer fitted to the same system. Transesophageal echocardiography was performed using a 3.75-MHz transesophageal probe with a Toshiba SSH-65A or a 5.0-MHz transesophageal probe with an Aloka SSD-870 ultrasound system. The transesophageal echocardiographic procedure was explained, and informed consent was obtained from all the patients. Transesophageal studies were performed after precordial examinations. Local spray anesthesia of the hypopharynx was used with 2%lidocaine to suppress gag reflex. No other premedications were given. The patient was placed in the left lateral recumbent position. Standard transesophageal projections were used as previously described.21 All echocardiographic studies were reviewed separately by two observers unaware of clinical data. There was one disagreement between the two observers in interpretation of the transthoracic echocardiogram in Patient 1. This difference in interpretation was resolved by a third observer 382

who did not know any findings. There was no difference in interpretation of the transesophageal echocardiogram between these two observers in all patients. Statistical Analysis Maximal dimension of the tumor determined by transthoracic echocardiography, transesophageal echocardiography, and surgical pathology was used as the size parameter. Variance analysis was used to compare the size parameter in these three groups. A p < 0.05 was considered to indicate statistically significant difference. RESULTS

Clinical Characteristics The common presenting symptoms in this study were dyspnea on exertion in 7 patients and chest pain in 6 . Other presenting symptoms included palpitation in 4 and extremity cyanosis due to embolism in 1. The majority of the patients had unremarkable physical findings. Pansystolic murmur occurred in 2 patients; diastolic rumbling murmur in 1;and pansystolic and diastolic rumbling murmurs in 1. The electrocardiogram was mostly in sinus rhythm. Atrial fibrillation occurred only in one patient. Detection Rate by Echocardiography Of the 17 patients, 14 underwent cardiac surgery, which proved the presence of a tumor except in 1 patient in whom the operation revealed only Ebstein's anomaly. Transesophageal echocardiography detected a mass-like lesion near the tricuspid annulus (Figure 1) in this patient. Fat accumulation at the tricuspid annulus was found at operation. Of the 13 surgically proved intracardiac tumors, 11 were myxoma, 1 teratoma, and 1 hemangioma (Table 1). Transesophageal echocardiography diagnosed intracardiac tumors in all the 14 patients who underwent surgery, whereas transthoracic echocardiography missed myxoma in 1patient, misdiagnosed myxoma as thrombus in 1 patient, and missed a hemangioma at the right ventricular outflow tract in 1 patient. Transthoracic echocardiography diagnosed intracardiac tumors in the other 3 patients (Figure 2) who did not undergo surgery. Other diagnostic modalities-including cardiac catheterization, computed tomography, and JOURNAL OF CLINICAL ULTRASOUND

TRANSESOPHAGEAL ECHO AND CARDIAC TUMORS

FIGURE 1. Echocardiographic pseudomass in patient 3. Transesophageal transverse plane showed a welldefined tumor (T) at the tricuspid annulus. No tumor was found at cardiac surgery. (RA: right atrium; RV: right ventricle.)

magnetic resonance imaging of the heart-were performed and showed no tumors in these 3 patients. Thus, there were 4 false-positive and 2 false-negative diagnoses with transthoracic echocardiography, whereas there were only 1 falsepositive and no false-negative diagnoses with transesophageal echocardiography. Of the 13 tumors proved by surgery, 8 were located in the left atrium, 2 in the right atrium, 2 in the right ventricle, and 1 in the right ventricular outflow tract.

Tumor Characteristics Of the 11 patients with myxoma proved by surgery, the stalk of the myxoma was clearly defined with transthoracic echocardiography in 5 patients, whereas it was seen with transesophageal echocardiography in 10. Transesophageal echocardiography also demonstrated the contour of the tumor clearer than did transthoracic echocardiography (Table 1).Transesophageal echocardiography demonstrated a cauliflower-like (Figure 3) or lobulated tumor contour in 3 patients that were considered t o be round by transthoracic echocardiography. Some of the characteristics of these tumors (such as cysts, calcification, or spontaneous echo contrast within cysts) were not seen VOL. 22, NO. 6,JULYYIAUGUST 1994

with transthoracic echocardiography but were clearly seen with transesophageal echocardiography (Figure 4). Further cardiac computed tomography and right atrial angiogram showed enhanced contrast medium in the cystic mass. The content of the cyst was proved to be blood by surgery, and the myxoma was in cyst form. The pathological findings showed homogeneity of the tumor in 10 patients and nonhomogeneity (interspersed cysts and calcification) in 3 patients. The pathological findings were consistent with the characteristics detected by transesophageal echocardiography. The tumors with a solid or gelatinous consistency detected by transesophageal echocardiography were noncompressible, and the tumors with cysts were deformable by palpation during surgery. All the myxomas were mobile. Of these, 9 were prolapsing into the ventricle during diastole and two remained in the atrial cavity. All the myxoma prolapses into the ventricle during diastole were detected both by transthoracic and transesophageal echocardiography, whereas the tumors limited t o the atrium were detected only by transesophageal echocardiography. Although both techniques could detect tumor 383

3.5 x 4.5 x 1.0

Myxoma

RV

Round Round Oval Oval

Thrombi

Oval Round Round

Round Not seen Round Round Round Round Round Nodular Not seen

TTE

Oval

-

Cauliflower like Lobulated

Round Round

-

Round Round Round Round Round Lobulated Round Nodular Round

TEE

T u m o r Contour

-

-

-

-

-

Good

-

Good

Not seen

-

Not seen

Good Good

Not seen Not seen

-

Good

-

Good Good Good Good

-

Good

TEE

Not seen

Good Good Good Good

-

Good

TTE

-

-

IAS

IAS

IAS IAS

-

IAS

IAS IAS IAS IAS

-

-

IAS

Site Attachment

Stalk Definition

3.0 X 2.0 2.0 x 2.0 2.0 x 1.3 3.0 x 4.0

1.1 x 2.1

3.0 x 3.0 5.0 X 5.0 4.0x 4.0

Solid

-

Gel Solid Solid Gel Gel Gel Gel Solid

3.0 X 4.5

Gel

-

Gel Solid Solid Gel

5.0 X 3.0 Gel

-

4.1 x 2.0 Thrombi

-

Cyst with SEC Cyst with SEC, calcification Gel

Gel Solid Solid Gel Gel Gel Gel Solid Gel with cyst and calcification

TEE

Consistency

TTE

5.5 x 5.5 cyst 4.0 x 4.0 Cyst

-

4.0 X 2.5 5.0X 6.9 x 3.3 7.5 x 5.5 x 8.6 6.0X 3.0 x

4.0 3.0 3.5 6.0 5.0 3.5 4.0 9.0 2.0

cm

TEE

cm 6.0 x 3.5 x Uncertain 4.0 x 5.5 x 5.5 5.5 x 5.9 x 4.2 6.0 x

cm c m 5.5 x 3.5

TTE

Tumor Size

TTE: transthoracic echocardiography; TEE: transesophageal echocardiography; LA: left atrium; IAS: interatrial septum; Gel: gelatinous; RVOT: right ventricular outflow tract; RA: right atrium; RV: right ventricle; SEC: spontaneous echo contrast; -: absent. *Patient 3 had TTE and TEE false-positive diagnoses.

-

-

LA

5.0 x 2.5 x 1.3

Myxoma

14 15 16 17

-

LA

4.0 x 2.0 x 2.0

Mvxoma

13

-

-

RA LA

5.5 x 5.0 x 4.5 3.5x 3.5 x 1.5

6.0 x 2.0 5.0 x 4.0 4.0 x 4.0

LA RVOT None LA LA LA RA RV LA

Location

Myxoma Myxoma

5.5 x 6.0x 6.5x 7.5 X

None

-

cm

5.0 X 3.0 6.0x 9.5 x 4.0 3.0 X 1.6 X 1.2

cm

6.0 x 4.1 x 2.0 3.5 x 4.0 x 3.0

crn

T u m o r Size

Myxoma Hemangioma None Myxorna Myxoma Myxoma Myxoma Teratoma Myxorna

Tumor Type

10 11 12

9

8

7

6

2 31 4 5

1

Patient Number

Surgical Pathology

TABLE 1 Comparison of Transthoracic and Transesophageal Echocardiography in the Detection of lntracardiac Tumors

TRANSESOPHAGEAL ECHO AND CARDIAC TUMORS

FIGURE 2. Echocardiographic ghost tumor in patient 16. Transthoracic echocardiographic-modified parasternal long-axis view showed a small mass (arrow) at left atrial posterior wall, and M-mode echo also showed the mass (arrow) floating in the left atrium (LA). (LV: left ventricle.)

mobility, transesophageal echocardiography was superior t o transthoracic echocardiography in detecting the extent of tumor mobility, especially when it was restricted to only one cardiac chamber. Although transesophageal echocardiography tended to be superior to transthoracic echocardiography in determining tumor size (Table 11, this difference did not reach statistical significance (F value = 1 . 3 8 4 6 , ~ > 0.05).

both diagnostic techniques cannot be calculated because no control patients were studied. Transesophageal echocardiography using the unobstructed view from the esophagus to the heart provides optimal imaging for most cardiac structures in almost all patients when transthoracic echocardiography cannot obtain adequate imaging due to chest-wall deformity, obesity, or emphysema. Furthermore, the close proximity of the esophagus t o the heart allows the use of higher-transducer frequencies, which increase DISCUSSION image resolution. Thus, it is not surprising that transesophageal echocardiography is superior to During the past decade, transesophageal echocartransthoracic echocardiography in the diagnosis diography has become a well-established tool for of intracardiac tumors, as indicated by previous evaluating cardiovascular diseases,l63l7and the st~dies.~'~~~ frequency of use of this new tool in investigations is increasing. However, only a few s t u d i e ~ ~ ' ' ~ ~ Mugge et all8 reported that all right and left atrial myxomas were equally detected by both used this useful tool to improve diagnostic potentechniques. In our study, myxomas were missed tial in patients with suspected intracardiac by transthoracic echocardiography in 1 patient tumors. In this study, transesophageal echocarand misdiagnosed as a thrombus in another padiography was superior to transthoracic echocartient. Both tumors were located in the left atrium diography in the detection of (1) intracardiac tuand attached to the interatrial septum with limmors, (2) the stalk attachment of myxoma, (3) ited mobility. Furthermore, the size of both tuthe detailed morphologic characteristics of the tumors was relatively small (3.0cm and 4.0 cm, mor, and (4) the size of the tumor. The morphorespectively). Tumor mobility and size were not logic characteristics of the tumor, such as cyst described in the study of Mugge et a1.l' The acformation, calcification, and the presence of sponcuracy of transthoracic echocardiography in diagtaneous echo contrast, were proved by surgical nosing cardiac myxomas may be influenced by tupathology and could be seen clearly with transmor size and tumor mobility as shown in our esophageal echocardiography. The patients tolerstudy. ated this procedure well and there were no comEcholucency in atrial myxomas is rarely found plications during the procedure. The specificity of VOL. 22, NO. 6, JULY/AUGUST 1994

385

SHYU ET AL.

FIGURE 3. (A) Transthoracic apical four-chamber view showing an ill-defined mass (arrowheads) attached t o interatrial septum in Patient 13. The echocardiographic diagnosis was suspected left atrial thrombus. There is echo dropout in the lateral wall of LA due t o slight angulation of the probe. (B) Transesophageal echocardiographic view in this patient showed a cauliflower-like mass (M)attached t o interatrial septum. The cauliflowerlike mass was proved t o be a rnyxoma at surgery. (RA: right atrium: LA: left atrium; LV: left ventricle.)

386

JOURNAL OF CLINICAL ULTRASOUND

TRANSESOPHAGEAL ECHO AND CARDIAC TUMORS

FIGURE 4. (A) Transthoracic parasternal long-axis view showing an empty cyst (C) prolapsing into the left ventricle during diastole in Patient 12. (B) Transesophageal echocardiographic four-chamber view showing a cystic mass (arrow) with its solid part attached to the interatrial septum in the same patient. There was spontaneous echo contrast in the cyst. The echo contrast was swirling in the cyst during the real-time examination. (LV: left ventricle; RV: right ventricle; LA: left atrium.)

by transthoracic echocardiography. The echolucent area may indicate tumor hemorrhage or c y ~ t . ~Pure ' , ~ ~cyst formation in atrial myxoma is uncommon." Of the 11 myxomas in our study, two were cystic masses containing blood. SpontaVOL. 22, NO. 6,JULY/AUGUST 1994

neous echo contrast, described as smoke-like, discrete reflectance in intracardiac blood, is most commonly seen in clinical settings resulting in low flow states, such as severe left ventricular dysfunction or in the presence of mitral dis387

SHYU ET AL.

eases.24The cysts in our two patients were large, and the content of the cysts was heavy blood. Thus, the heavy blood content in a large cyst made the aggregation of red blood cells into clumps possible producing this unusual echo contrast pattern. The far superior resolution of transesophageal echocardiography compared with transthoracic echocardiography resulted in the echo contrast in the cyst being seen with transesophageal echocardiography but not with transthoracic echocardiography. There were several potential explanations for the lack of transthoracic echocardiographic visualization of intracardiac tumors in this study. The relatively small size of the tumor and the lack of tumor mobility made it difficult to detect some tumors clearly. Other reasons that contributed to the false-negatives included the acoustic characteristics of the tumor, which may not differ sufficiently from those of blood or endocardium to permit adequate reflection of ultrasonography for its d e t e ~ t i o n Occasionally, .~~ it was difficult to distinguish echo signals reflected by the tumor from those of artifacts and background noise. There were 4 false-positive diagnoses by transthoracic echocardiography in our study. These diagnoses were caused by the following: fat accumulation near the tricuspid annulus in Patient 3, hypertrophic moderator band with acoustic interference at the right ventricular apex in Patient 10, an artifact near the mitral annulus due to probe angulation in Patient 15, and a possible calcified mitral annulus with probe angulation in Patient 16. Thus, false-positive diagnoses of intracardiac tumors by transthoracic echocardiography was seen not uncommonly in clinical practice. In 5 of the 11 patients with myxomas, the tumor and stalk attachments were identified by transthoracic echocardiography. The tumor size measured by transthoracic echocardiography was 24 cm and its actual size was 2 5 . 5 cm. Transesophageal echocardiographic findings did not add additional important information in these patients. When the size of the tumor is large and the information provided by transthoracic echocardiography is adequate, transesophageal echocardiography is not needed in these patients. However, when the size of the tumor is small or the information provided by transthoracic echocardiography is insufficient, transesophageal echocardiography should be considered t o confirm or exclude the presence of a tumor. Because of the new tomographic images obtainable with transesophageal echocardiography, there were pitfalls resulting from misinterpreta388

tion of normal and abnormal anatomy.26The tricuspid annulus normally is filled with variable amounts of fat that can produce a mass effect. The pseudomass diagnosed by transesophageal echocardiography is an example of the pitfalls with which echocardiographers should be familiar. ACKNOWLEDGMENTS

The authors express appreciation to Misses HweiMing Kuo and Huey-Yann Chang for the preparation of this manuscript. REFERENCES 1. Straus R, Merliss R. Primary tumor of the heart. Arch Pathol Lab Med 39:74-78, 1945. 2. Smith C: Tumors of the heart. Arch Pathol Lab Med 110:371-374, 1986. 3. Wilding G, Gren HL, Long DL, et al: Tumors of the heart and pericardium. Cancer Treat Rev 15:165181, 1988. 4. Dein JR, Frist WH, Stinson EB, et al: Primary cardiac neoplasms: early and late results of surgical treatment in 42 patients. J Thorac Cardiovasc Surg 93502-511, 1987. 5 . Pahdi RK, Kelly HG, Lynn RB: Intra-atrial myxoma: review of literature and report of a right atrial myxoma diagnosed preoperatively and successfully treated. Can J Surg 2:414-420, 1959. 6. Godwin JD, Axel L, Adama JR, et al: Computed tomography: a new method for diagnosing tumor of the heart. Circulation 63:448-451, 1981. 7. Gross BH, Glazer GM, Francis IR: CT of intracardiac and intrapericardial massess. AJR A m J Roentgen01 140:903- 907, 1983. 8. Pizzarello RA, Goldberg SM, Goldman MA, et al: Tumor of the heart diagnosed by magnetic resonance imaging. J A m Coll Cardiol 5:989-991, 1985. 9. Winkler M, Higgins CB: Suspected intracardiac masses: evaluation with MR imaging. Radiology 165117-122, 1987. 10. Fyke FE 111, Seward J B , Edwards WD, et al: Primary cardiac tumors: experience with 30 consecutive p a t i e n t s since t h e introduction of twodimensional echocardiography. J A m Coll Cardiol 51465-1473, 1985. 11. Lappe DL, Bulkley BH, Weiss JL: Two-dimensional echocardiographic diagnosis of left atrial myxoma. Chest 7455-58, 1978. 12. Come PC, Kurland GS, Vine HS: Two-dimensional echocardiography in differentiating right atrial and tricuspid valve mass lesions. A m J Cardiol44: 1207-1212,1979. 13. Liu HY, Panidis I, Soffer J , et al: Echocardiographic diagnosis of intracardiac myxomas: present status. Chest 84:62-67, 1983. 14. Mitchell MM, Sutherland GR, Gussenhoven EJ, et JOURNAL OF CLINICAL ULTRASOUND

TRANSESOPHAGEAL ECHO AND CARDIAC TUMORS al: Transesophageal echocardiography. J Am Soc Echocardiogr 1:362-377, 1988. 15. Fisher EA, Goldman ME: Transesophageal echocardiography: a new view of the heart. Ann Zntern Med 11:91-93,1990. 16. Matsuzaki M, Toma Y, Kusukawa R: Clinical applications of transesophageal echocardiography. Circulation 82:709-722, 1990. 17. Kuecherer HF, Lee E, Schiller NB: Role of transesophageal echocardiography in diagnosis and management of cardiovascular disease. Cardiol Clin 8:377-388, 1990. 18. Mugge A, Daniel WG, Haverich A, et al: Diagnosis of noninfective cardiac mass lesions by twodimensional echocardiography: comparison of the transthoracic and transesophageal approaches. Circulation 83:70-78, 1991. 19. Obeid AI, Marvasti M, Parker F, e t al: Comparison of transthoracic and transesophageal echocardiography in diagnosis of left atrial myxoma. Am J Cardiol 63:1006-1008, 1989. 20. Hwang J J , Lien WP, Kuan P, et al: Atypical myxoma. Chest 100:550-551, 1991.

VOL. 22, NO. 6, JULY/AUGUST 1994

21. Seward JB, Khandheria BK, Oh JK, et al: Transesophageal echocardiography: technique, anatomic correlations, implementation, and clinical applications. Mayo Clin Proc 63:649-680, 1990. 22. Rahilly GT, Nanda NC: Two-dimensional echographic identification of tumor hemorrhages in atrial myxomas. Am Heart J 101:237-239, 1981. 23. Thier W, Schluter M, Krebber HJ, et a1 Cysts in left atrial myxomas identified by transesophageal cross-sectional echocardiography. Am J Cardiol51: 1793-1795, 1983. 24. Beppu S, Nimura Y, Sakakibara H, et al: Smokelike echo in the left atrial cavity in mitral valve disease: features and significance. J Am Coll Cardiol 6:744-749, 1985. 25. Come PC, Riley MF, Markis JE, et al: Limitations of echocardiographic techniques in evaluation of left atrial masses. A m J Cardiol48:947-953,1981. 26. Seward JB, Khandheria BK, Oh JK, et al: Critical appraisal of transesophageal echocardiography: limitations, pitfalls, and complications. J Am Soc Echocardiogr 5:288-305,1992.

389