Detection of pericardial heart disease by computed tomography

ClinicalRadiology (1984) 35, 397-400 © 1984 Royal College of Radiologists

0009-9260/84/283397502.00

Detection of Pericardial Heart Disease by Computed Tomography* A. G O U L I A M O S , J. A N D R E O U , J. STERIOTIS~', A. K A L O V I D O U R I S , L. V L A H O S and C. P A P A V A S S I L I O U

Department of Radiology, University Aretaieion Hospital, Athens, Greece, and ? Department of Cardiology, Evagelismos Hospital, Athens, Greece The findings on computed tomography (CT) in 10 cases of pericardial disease, nine of which were surgically confirmed, are presented. The 10th patient (Case 10), whose CT diagnosis was 'pericarditis', was found to have a myxosarcoma at surgery. In some of these cases, echocardiographic and conventional radiographic investigation gave inconclusive information regarding the underlying pathology. Our results support the conclusion that CT complements echoeardiography: (a) in the quantitative and qualitative assessment o f pericardial effusions, (b) in the localisation of space-occupying lesions involving the pericardium and (c) in diagnosis of constrictive pericarditis. As echocardiography is so very much operator-dependent, CT may well be a more reliable method of obtaining a result which indicates both the diagnosis and the extent of the disease.

Since the advent of echocardiography, fluoroscopy, chest radiology and angiocardiography are no longer routinely employed in the diagnosis of pericardial heart disease. The sensitivity and specificity of echocardiography in detecting pericardial effusions is high (Feigenbaum et al., 1965; Pate et al., 1967), with some limitations (Lin et al., 1978). False positive and false negative echocardiographic results are mainly due to faulty technique. Third and fourth generation computed tomography (CT) scanners, with exposure times of 5 s or less, can demonstrate the heart and pericardium and provide additional information in pericardial disease (Houang et al., 1979; Isner, et al.,1982; Moncada, et al., 1982). The purpose of this study is to outline the use of CT in diagnosing pathological conditions of the pericardium.

Fig. 1- NormalCT scan at the levelof the heart. White arrowsindicate line of pericardium. LV=left ventricle, RV-right ventricle, RA=right atrium, LA=left atrium (L +39, W +256, -256).

PATIENTS AND METHODS The 10 cases studied included patients with constrictive pericarditis, pericardial effusion, neoplastic involvement of the pericardium (sarcoma, lymphoma), echinococcal cysts and a benign pericardial cyst. Echocardiography was the first method of investigation, followed by CT as a complementary diagnostic test. A Siemens Somatom 2 CT scanner was used with a 4-5 s scanning time and an 8 mm slice thickness; E C G *Paper presented at the Vth European Congress of Radiology, Bordeaux, France, 5-10 September 1983. Address for correspondence: A. Gouliamos, MD, Department of Radiology, UniversityAretaieion Hospital, 76 Vas. Sophias Avenue, Athens 115 28, Greece.

Fig. 2- Normal CT scan at the level of the aortic root. AoR=aortic root (black arrows). White arrows indicate pericardium (L +62, W -I-318, -194). gating was not available. Hounsfield numbers ( H U : 1.000 scale) to determine tissue type were used. Rapid intravenous drip infusion of contrast material, combined with a bolus technique, was used when needed, with cuts at the following levels of interest.

398

CLINICAL RADIOLOGY

Level A: cardiac chambers (Fig. 1) Level B: aortic root (FIR. 2) Level C: aorto-pulmonary window (Fig. 3) Pre-operative biopsies were not performed under computed tomographic control and angiography was not considered necessary for further evaluation of the patients studied. Surgery verified thc CT findings with the exception of the 'pericarditis' case, which proved to be a myxosarcoma.

RESULTS

Fig. 3 - Normal CT scan at the level of the aorto-pulmonary window. AAo=ascending aorta, DAo=descending aorta, SV=superior vena cava, PA=pulmonary artery. Arrows indicate upper portion of pericardium (L +57, W +313, -199).

A total of 10 patients, six with neoplastic and four with non-neoplastic conditions involving the pericardium, were investigated. The neoplastic cases consisted of three lymphomas, two thymomas and a primary myxosarcoma of the pericardium. Two lymphomas (one Hodgkin's or one non-Hodgkin's) showed direct extension from the anterior mediastinum into the pericardium

(a)

(b)

Fig. 4 - Hodgkin's lymphoma: (a) CT section at level of the carina shows a soft-tissue mass (black arrows) occupying the anterior mediastinum due to lymphoma infiltration. Pleural effusion is demonstrated on the left side (white arrows). (b) CT section at the level of the aortic root reveals diffuse thickening of the pericardium (black arrows) with adhesions to the pleural surface. Pleural effusion is again visible on the left (white arrows). AoR=aortic root (L+37, W+293, -219).

(a)

(b)

Fig. 5 - Malignant thymoma: (a) CT at the level of the manubrium shows a round solid mass (arrows) projecting to the right of the midlinc. ST-sternum, AAo: ascending aorta. (b) Lower section of the same patient illustrates a low-density area involving the pericardium lateral to the right atrium (arrows) due to neoplastic pericardial invasion (L+80, W +208, -48).

399

CT 1N PERICARDIAL HEART DISEASE

Table 1 - The distribution of C T and e c h o c a r d i o g r a p h i c findings in the conditions described

Case

Age Sex (years)

CT

Echocardiography

Final diagnosis

l

51

F

+

-

2 3 4 5 6 7 8 9 10

66 49 56 21 50 50 38 41 45

F M M F F F F F M

+ + + + + + + + -

+ + +

Constrictive pericarditis Viral pericarditis Echinococcal cyst Pericardial cyst Lymphoma Lymphoma Lymphoma Thymoma Thymoma Myxosarcoma

ND ND +

N D = n o t done.

Fig. 6 - Myxosarcoma of the pericardium: CT section at the level of the ventricles shows a low-density area expanding the entire pericardium and compressing the cardiac chambers (arrows) (L +38, W +166, -90).

Fig. 7 - Echinococcus of the pericardium: CT section at the level of the xyphoid demonstrates multiple cystic structures between the pericardium and the upper surface of the diaphragm. Larger cyst is noted compressing the gastro-oesophagea/junction (arrows) in the subdiaphragmatic space (L +54, W +180, - 7 6 ) .

(a)

(Fig. 4). Two invasive thymomas presented as anterior mediastinal masses extending inferiorly towards the pericardium (Fig. 5). Pericardial involvement was predicted by CT scanning and verified at surgery. The myxosarcoma of the pericardium (Case 10) was falsely interpreted as pericarditis (Fig. 6). In this case echocardiography correctly diagnosed the presence of a solid mass. The non-neoplastic lesions included echinococcal cyst (Case 3) (Fig. 7), a simple pericardial cyst (Case 4), calcified constrictive pericarditis (Case 1) and viral pericarditis (Case 2). In these cases, CT provided valuable information relating to the pre-operative assessment of the pericardial disease. For instance, in Case l the extent of calcification and pericardial involvement by the constrictive process was well demonstrated (Fig. 8). All diagnoses were established by CT, except the myxosarcoma of the pericardium which was correctly diagnosed by echocardiography. On the other hand, echocardiography was negative in the case of constrictive pericarditis and failed to predict pericardial involvement in the cases of lymphoma and thymoma (Cases 5, 6 and 8) (Table 1).

(b)

Fig. 8-Constrictive pericarditis: (a) CT section o1 the aortic root. There is pericardial calcification surrounding the right ventricular outflow tract. (b) CT section at the level of the cardiac chambers. Pericardial thickening and calcification are seen anterior to the right ventricle and lateral to the left ventricle (arrows) (L +36, W +140, -116).

400

CLINICALRADIOLOGY

DISCUSSION Although echocardiography is the recognised method of investigating the pericardium, our results have shown that useful information can be obtained from computed tomography. In almost all patients the normal pericardium can be identified by CT. The thickness of the normal parietal pericardium is generally 1 - 2 m m , except the caudal portion at the insertion of the central tendon of the diaphragm where its thickness is doubled (Moncada et al., 1982). Occasionally, the transmitted motion from the heart chambers can masquerade as pericardial thickening due to myocardial movement; E C G gating may help in overcoming this limitation. The pericardial space is not usually detected by CT unless an effusion is present. In the present study, CT was undertaken for the following reasons. (a) Inconclusive data obtained from echocardiography. (b) The need to indicate a tissue type in a suspected pericardial mass. (c) Visualisation of pericardial involvement by a disease process. (d) Demonstration of the extent of calcification in constrictive pericarditis. The information provided pre-operatively often influenced the management of these patients, for example, indicating that resection, rather than a biopsy, would be possible. In the case of the pericardial cyst, however, the clinical history of hydatid disease of the liver led the surgeons to believe that a solitary cyst in the right costophrenic angle represented an echinococcal cyst. The CT interpretation, however, strongly suggested the diagnosis of a pericardial cyst and echocardiography supported the diagnosis of loculated fluid adjacent to the right side of the heart. While our previous experience has shown that primary tumours of the heart are seen more often than primary tumours of the pericardium (Gouliamos et al., 1983) and that metastatic involvement of the pericardium is seen more frequently than a myocardial invasion, the tumours described in the present series were all primary, either arising in the pericardium or directly invading it. Although CT easily identifies nodular solid masses within the pericardium, density measurements can be misleading. In the case of myxomatous tumours, the values obtained may give a false impression of pericarditis (Gouliamos et al., 1984) or cardiomyopathy (Isner et al., 1982). Whilst carcinomas of the lung and breast are the primary neoplasms most likely to invade or metastasise to the heart and pericardium, numerous primary malignant tumours have produced myocardial and pericardial secondary deposits. These include the stomach, urinary bladder and uterus. Lymphoproliferative malignancies and thymomas are two common causes of neoplastic pericardial invasion. This is supported by the cases presented in this study. Computed tomography cannot always differentiate between a benign and a malignant thymoma unless there is

evidence of metastatic disease in the mediastinum, p l e u r a - o r lungs (Baron et al., 1982). Neoplasms involving the thymus, such as Hodgkin's disease, can cause diffuse infiltration and present with an appearance with may be either cystic or solid. Computed tomography, however, does appear to be reliable in differentiating benign thymic cysts from solid tumours (Gouliamos et al., 1982). Computed tomography is the most effective imaging method for neoplastic pericardial disease. Tumour invasion may cause a plaque-like thickening of the pericardium. Occasionally, nodular masses may be seen. Our results suggest that CT is more reliable than echocardiography in illustrating pericardial involvement. Echinococcal cysts of the pericardium are well demonstrated by CT. Daughter cysts, the clinical history of the patient and the presence of other echinococcal cysts in the liver or lungs may help to confirm the diagnosis. Benign pericardial cysts are also well demonstrated by CT (Moncada et al., 1982). In conclusion, while both echocardiography and CT gave the correct diagnosis in three out of 10 cases, a combination of the two techniques gave diagnostic information in the remaining seven cases. This justifies the complementary use of the two methods of investigation. In addition, CT often yielded valuable information with regard to the extent of pericardial involvement, thus influencing patient management.

REFERENCES

Baron, R. L., Lee, T. K. T., Sagcl, S. S. & Levitt, R. G. (1982). Computed tomography of the abnormal thymus. Radiology, 142, 12%134. Feigenbaum, H., Waldhausen, J. A. & Hyde, L. P. (1965). Ultrasound diagnosis of pericardial cffusion. Journal of the American Medical Association, 191, 711-714. Gouliamos, A., Andreou, J., Kalovidouris, A., Steriotis, J., Aravanis, Ch., Bredakis, J., Levett, J., Vlahos, L., Papavassiliou, C. & Pontifex, Gr. (1983). Space-occupyinglesions of the heart: a computed tomography approach. Letter to the editor. European Journal of Cancer and Clinical Oncology, 19, 691-694. Gouliamos, A., Steriotis, J., Kalovidouris, A., Andreou, J. & Papavassiliou, C. (1984). Water-like densities mimicking pericarditis. Journal of Computer Assisted Tomography, 8, 343-344. Gouliamos, A., Striggaris, K., Lolas, C., Deligeorgi-Politi, H., Vlahos, L. & Pontifex, Gr. (1982). Thymic cyst. Journal of Computer Assisted Tomography, 6, 172-174. Houang, M. I. W., Arozema, X. & Shaw, D. G. (1979). Demonstration of the pericardium and pericardial effusion by computed tomography. Journal of Computer Assisted Tomography, 3, 601-603. Isner, J. M., Carter, B. L., Bankoff, M. S., Konstam, M. A. & Salem, D. N. (1982). Computed tomography in the diagnosis of pericardial heart disease. Annals of Internal Medicine, 97,473-479. Lin, T. K., Stech, J. M., Eckert, W. G., Lin, J. J., Farha, S. T. & Hagan, C. T. (1978). PericardiaI angiosarcoma simulating pericardial effusion by echocardiography. Chest, 73, 881-883. Moncada, R., Baker, M., Salinas, M., Demos, T. C., Churchill, R., Love, L., Reynes, C., Hale, D., Cardoso, M., Pifarre, R. & Gunnar, R. M. (1982). Diagnosticrole of computedtomographyin pericardial heart disease. American Heart Journal, 100, 263-282. Pate, J. W., Gardner, H. C. & Norman, R. S. (1967). Diagnosis of pericardial effusion by echocardiography.Annals of Surgery, 165, 826-829.