1991, The British Journal of Radiology, 64, 330-333

Ultrasound in the diagnosis of the juxta-pleural lesion By M. J . Bradley, M B ChB, FRCR and C. Metreweli, FRCP, FRCR Department of Diagnostic Radiology and Organ Imaging, Prince of Wales Hospital, Hong Kong {Received May 1990 and in revised form August 1990) Keywords: Ultrasound, Juxta-pleural lesions

Abstract. Any opacity on a chest radiograph has a wide differential diagnosis. Plain radiography and computed tomography may help to differentiate whether bone, intercostal soft tissue or mediastinum are involved. Bronchoscopy is often negative with peripheral lesions. This leaves the physician with a diagnostic problem. We examined ultrasonically 30 patients with juxta-pleural lesions and performed cutting biopsy in 27. Twenty-four of these produced positive histology (90%). The three that were not biopsied were anechoic and had pulsatile lesions due to vascular abnormalities. There were four benign lesions all with hypoechoic appearances and the 20 malignant cases showed a wide spectrum of echogenicity. Of the malignant lesions, 90% showed pleural line disruption with reduced respiratory movement suggesting chest wall invasion. There were no complications, despite using cutting biopsy.

We were stimulated to conduct this study because of our high local incidence of tuberculosis and primary lung tumours. Clinicians need to know whether the patient has dual pathology, or whether the disease is benign or malignant so that he can initiate appropriate treatment. In order for ultrasound to be of help in this assessment, the lesion must be juxta-pleural. Yang et al (1985) have described different echogenicities in the lesions but we wanted to compare this with the histology in order to establish a pattern. The visceral and parietal pleurae can be seen ultrasonically (Saito et al, 1988) and we have used this sign, together with reduced movement, as an indicator of chest wall invasion. This can be compared with loss of the fat plane which can be clearly seen on computed tomography (CT).

All patients had a post-procedure erect expiratory chest radiograph to exclude a pneumothorax. Results

Thirty patients were examined sonographically. The male to female ratio was 2:1 with a mean age of 59 years (range 3-83 years). With only one exception, the chest radiographs showed solitary lesions, twice as many occurring on the right as the left. The patient with multiple lesions was thought to have pleurally based metastases on the chest radiograph, which proved to be due to thymic carcinoma on biopsy. Only five patients had concurrent pleural effusions. The PA chest radiograph showed the majority of the lesions lay peripherally in the midPatients and methods axillary line (14) with a further six in the apices, three Thirty consecutive patients who had a juxta-pleural basal and six central. The lateral film showed four lesion on postero-anterior (PA) and lateral chest radio- central lesions to be adjacent to the posterior chest wall graphs with negative bronchoscopy underwent and two anteriorly. The remaining patient with thymic sonography. carcinoma had pleurally based lesions scattered through The study was carried out on an Aloka 650 using 3.5 the left hemithorax. and 5 mHz convex linear probes with occasional use of The lesions were identified sonographically and they a 5 mHz sector probe to look at smaller lesions between varied from 1 cm to 10 cm in size. We could see several the ribs. Most examinations were performed with the lesions in the patient with the thymic carcinoma and the patient lying down, but use of the erect position largest and most accessible was biopsied. occasionally enabled the use of improved intercostal Three lesions were anechoic and pulsatile and so were windows. The echotexture, size and shape of the lesion, considered to be vascular abnormalities and therefore together with its movement and disruption of the did not undergo biopsy (Fig. 1). Contrast-enhanced CT pleural line, were assessed. confirmed aneurysms of the brachiocephalic artery, arch Using Trucut needles and an aseptic technique, 27 of aorta and an arteriovenous malformation of the lesions were biopsied; three biopsies were contrain- pulmonary vein. At surgery, the last lesion was due to a dicated on the ultrasound appearances. Ultrasound tumour thrombus dilating the pulmonary vein. resulted in an accurate needle placement in each case. The remaining lesions had varying echogenicity compared with the intercostal muscles and this is Address for correspondence: Dr M. J. Bradley, c/o Department summarized in Table I. All the benign lesions were of Radiodiagnosis, Royal Liverpool Hospital, P.O. Box 147, hypo-echoic while malignancy demonstrated any pattern (Fig. 2). Liverpool L69 3BX. 330

The British Journal of Radiology, April 1991

Ultrasound in the diagnosis of the juxta-pleural lesion Table I. Spectrum of echogenicities of the 27 biopsied lesions Malignant (n = 20)

Benign (n = 4)

Non-specific

Hypoechoic Isoechoic Hyperechoic Complex

Figure 1. Ultrasound of this central mass shows it to anechoic and pulsatile, typical of a vascular structure. Surgery proved this to be a pulmonary vein dilated by tumour thrombus.

Three biopsies revealed non-contributory tissue on each of two occasions. These patients were followed up and showed no change in their condition or chest radiograph features at 6 months. We therefore presume these lesions to be of little clinical significance. Our biopsy technique yielded positive histology in 24 patients (90%). The range of malignant histology is summarized in Table II. Three of the four benign lesions were due to tuberculosis (TB) and the fourth was a plasma cell granuloma (inflammatory pseudotumour). Only one patient had

dual pathology: one man with acid fast bacilli in his sputum and an apical mass which proved to be a squamous cell carcinoma on biopsy. The radiographic shape may lead the radiologist to infer that the lesion is pulmonary or pleurally based. Similarly, ultrasound of intra-pulmonary lesions usually shows a round lesion with acute angles formed at its wall and pleural lines (Fig. 3). Tapered lesions with obtuse pleural/lesion angles are usually pleurally based (Fig. 4). This correlated well with the tissue type as all the round lesions comprised primary lung tissue tumours and only one of the tapered lesions contained lung tissue. The pleural lines were identified in each case and were either intact or disrupted by the mass. Of the 20 malignant cases, 18 had pleural disruption ultrasonographically (90%) whilst only one benign lesion, due to TB showed the same sign. Suspended respiration aids detail of the pleural line disruption, but it is important to assess the movement of the lesion in relation to the chest wall. Round, mobile lesions are intrapulmonary, whilst if tapered and fixed, these are pleurally based.

Figure 2. (a) Complex mass proven to be a squamous cell carcinoma. Parietal pleura disrupted with intact visceral pleura (r = rib, v = visceral pleura), (b) Hyperechoic mass seen between the ribs due to an adenocarcinoma (c = clavicle). Vol. 64, No. 760

331

M. J. Bradley and C. Metreweli

Figure 4. Pleurally based tapered complex mass due to B cell non-Hodgkin's lymphoma (p = parietal pleura, v = visceral pleura).

Discussion

Figure 3. The ultrasound demonstrates a round hypo-echoic lesion with disrupted parietal pleural line indicating chest wall invasion confirmed at surgery (p = parietal pleura, v = visceral pleura).

Round, fixed lesions with pleural line disruption indicates chest wall invasion. We observed reduced movement in each of our 18 malignant cases with pleural line disruption. We observed no immediate or delayed complications and we attributed this to the lesions being peripheral and to the high degree of confidence that the needle was solely within the lesion as a result of ultrasound guidance. Table II. Histology range in 23 malignant biopsies out of a total 27 biopsies Adenocarcinoma Adenosquamous Squamous Small cell Spindle cell Large cell Pleural secondary deposit from malignant thymoma B-cell non-Hodgkin's lymphoma Infantile fibrosarcoma Total

332

6 1 7 1 1 1 1 1 1 20

The clinical question as to whether the lung lesions in our cases were benign or malignant tumours or tuberculosis was answered in 90% of cases by ultrasound imaging and guided biopsy. Singh et al (1987) also described similar positive histology results with juxtapleural lesions. The echotexture of juxta-pleural lesions showed a wide range of echopatterns. Twenty malignant tumours demonstrated all types of echogenicity as compared to that of the intercostal muscle. Whereas all seven benign lesions were echopoor including, not surprisingly, the vascular abnormalities which were anechoic and pulsatile. Seventy-five per cent of the malignant lesions were isoechoic, hyperechoic or complex. Thus, from our results, it would appear that, if the lesion is echopoor, it is difficult to predict benignity or malignancy. However, other echogenicities are seen then 94% of these can be expected to be malignant. Saito et al (1988) described the visceral and parietal pleural lines sonographically as thin echogenic lines. If the parietal pleura is disrupted around the lesion and there is reduced movement of the lesion in relation to the chest wall, then we may conclude that there is chest wall involvement. However, only six patients with malignancy were felt to have surgically resectable lesions. We accurately predicted chest wall involvement in these cases but are unable to correlate the remainder. Our one case of TB demonstrating pleural line disruption leading to a false positive impression of malignancy may be explained by the inflammatory nature of the lesion. Our results suggest that when there is pleural line disruption with reduced respiratory movement of the lesion then 90% will be malignant. The British Journal of Radiology, April 1991

Ultrasound in the diagnosis of the juxta-pleural lesion

The lesion in the child with infantile fibrosarcoma was so large that we felt that it was not possible to assess pleural disruption completely and we assume that this is the reason why we failed to show pleural involvement in this case. We conclude, therefore, that appraisal of the echogenicity of the lesion and pleural line disruption is an accurate indicator of malignancy. Saito et al (1988) and Hirsch et al (1987) have described the shapes of the lesions. Sonography reflects that shape seen on the chest radiograph and may help to determine whether the lesion is that of pulmonary, chest wall or pleural in origin. Seventeen of 20 malignant lesions appeared round and histology showed primary intrapulmonary tumours. Tapered lesions suggest a chest wall or pleural lesion, and does not necessarily equate with chest wall malignant invasion as most of our benign lesions had this shape. It is more important to identify the intact visceral and parietal pleural lines. Haller et al (1980) have previously reported "through transmission" with various lesions. We have found this an impossible sign to identify as the ribs cast an acoustic shadow around the lesion and even in the plane between the ribs it was difficult to visualize transmission due to the high lung reflectivity. Negative tissue biopsies always raise the possibility of a geographical miss in the sampling or an inappropriate part of the lesion having been sampled. With direct visualization during ultrasound guided biopsy, the needle is demonstrably within the lesion. Heterogeneous lesions can be sampled from the most solid area or different areas of echogenicity. This was certainly the case in our tumour with a necrotic fluid centre in which we were able to sample the solid wall, which might have been missed using other imaging modalities. Several authors including Cheng et al (1986), Yang et al (1985) and Pederson (1986) have used fine needle aspiration techniques with ultrasound or fluoroscopy imaging. This is considered to be less hazardous but is dependent on a good cytological service, which is not available in all hospitals. Also, fine needles are more difficult to follow with ultrasound than the thicker cutting needles. This is because the thin needles do not generate as large a signal and are more likely to be deflected from the target path. Being able to make a "direct hit" makes the use of thick needles for juxta-

Vol. 64, No. 760

pleural lesions as safe as the use of fine needles, as our complication rate suggests. Also, FNA may not be able to describe histology of the tumour which is available on a tissue core (Pang et al, 1987). Conclusion

Sonographic evaluation of the juxta-pleural mass has proved extremely useful in the diagnosis of such a lesion. Its shape and movement help to determine pulmonary from pleurally based lesions and the echogenicity may aid benign malignant differentiation. We believe that pleural line disruption with poor respiratory movement is a sign of chest wall invasion. Ultrasound is a safe and accurate imaging medium for biopsy guidance. References CHENG, R. C , LEE, L. I., Kuo, S. H. & LUH, K. T., 1986.

Ultrasonically guided percutaneous mediastinal cystology and biopsy. Journal of the Formosan Medical Association, 85, 561-566. HALLER, J., SCHNEIDER, M., KASSNER, E., FREDMAN, A. &

WALDCROUP, L., 1980. Sonographic evaluation of the chest in infants and children. American Journal of Roentgenology, 134, 1019-1027. HIRSCH, J., CARTER, S., CHILKOS, P. & CULACURCIO, C , 1987.

Ultrasonic examination of radiographic opacities of the chest. American Journal of Roentgenology, 130, 1153-1156. PANG, J. A., TSANG, V., HOM, B. L. & METREWELI, C , 1987.

Ultrasound-guided tissue core biopsy of thoracic lesions with Trucut and Surecut needles. Chest, 91, 823-828. PEDERSON, C , AASEN, T. B. & GULSVIK, A., 1986. Fine needle

aspiration biopsy of mediastinal and peripheral pulmonary masses guided by real-time sonography. Chest, 89, 504-508. SAITO, T., 1988. Ultrasonically guided needle biopsy in the diagnosis of mediastinal masses. American Review of Respiratory Diseases, 138, 679-684. SAITO,

T.,

KOBAYASHI,

H.

&

KITAMURA,

Ultrasonography approach to diagnosing tumours. Chest, 94, 1271-1275.

S.,

1988.

chest

wall

SINGH, M., GUPTA, R. K., DAS, D. K. & PANT, C. S., 1987.

Ultrasonically guided fine needle aspiration biopsy of intrathoracic lesions. Indian Journal of Chest Disease, 29, 82-89. YANG, P. C , LUH, K. T., SHEU, J. C , Kuo, S. H. & YANG, S.

P., 1985. Peripheral pulmonary lesions: Ultrasonography and ultrasonically guided aspiration biopsy. Radiology, 155, 451-456.

333

Ultrasound in the diagnosis of the juxta-pleural lesion.

Any opacity on a chest radiograph has a wide differential diagnosis. Plain radiography and computed tomography may help to differentiate whether bone,...
2MB Sizes 0 Downloads 0 Views