The clinical importance of magnetic resonance imaging versus computed tomography in malignant pleural mesothelioma

Lung Cancer 22 (1998) 215 – 225

The clinical importance of magnetic resonance imaging versus computed tomography in malignant pleural mesothelioma Aija Knuuttila a,*, Maija Halme a, Leena Kivisaari b, Arto Kivisaari b, Jarmo Salo c, Karin Mattson a a

Department of Medicine, Di6ision of Pulmonary Medicine, Helsinki Uni6ersity Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland b Department of Radiology, Helsinki Uni6ersity Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland c Department of Thoracic Surgery, Helsinki Uni6ersity Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland Received 22 June 1998; received in revised form 22 September 1998; accepted 25 September 1998

Abstract There is no standard therapy for malignant pleural mesothelioma (MPM), but recent reports have shown that extensive surgery combined with chemo- and radiotherapy prolongs the survival of selected patients with early stage disease. This emphasises the need for accurate staging procedures at diagnosis and reliable imaging methods to assess response to treatment. Computed tomography (CT) of the chest has been the standard imaging method for these purposes for the last decade, but it is limited in its ability to demonstrate accurately the platelike growth pattern of MPM within the thorax due to the partial volume effect on curved surfaces. In order to define the value of magnetic resonance imaging (MRI) in the imaging of MPM, we have compared the findings from 26 parallel paired CT and MRI scans of mesothelioma patients at various stages of the disease. MRI showed tumour spread into the interlobar fissures, tumour invasion of the diaphragm and through the diaphragm, and invasion of bony structures better than CT. Invasion of the chest wall and mediastinal soft tissue and tumour growth into the lung parenchyma were equally well seen on both imaging methods. CT was better for detecting the inactive pleural calcifications. MRI is a sensitive detector of the characteristic growth pattern and extension of MPM and we recommend its use more widely for the clinical management of MPM especially when evaluating tumour resectability and in research protocols when an accurate evaluation of disease extent is essential. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Malignant pleural mesothelioma; Computed tomography; Magnetic resonance imaging; Pleural neoplasms

* Corresponding author. Tel.: +358 9 4711; fax: +358 9 4715791. 0169-5002/98/$ - see front matter © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S0169-5002(98)00083-X

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

216

1. Introduction

2. Patients and methods

Malignant pleural mesothelioma (MPM) is a locally aggressive neoplasm deriving from any mesothelial surface within the thorax. It grows by platelike extension and often causes pleural effusion. It may affect the entire pleura, invade the thoracic wall, the diaphragm, the mediastinum and the lung, and grow along fissures across the mediastinum to the opposite pleura, or through the diaphragm into the peritoneal cavity. This peculiar but characteristic growth pattern makes accurate radiological imaging for diagnosis, staging and treatment difficult. Surgery has earlier had only a palliative role, but recent studies suggest that selected patients with early stage, nodenegative disease might benefit from extensive surgery (extrapleural pneumectomy) combined with chemo- and radiotherapy [1,2]. Imaging techniques are crucial to the management of MPM at diagnosis and throughout the clinical course of the disease. The platelike growth pattern of MPM is often difficult to interpret on axial CT scans, because of the partial volume effects when imaging curved structures at the axial planes [3,4]. In earlier studies using the old 0.5 T MRI techniques, conflicting results were reported on the clinical value of MRI for evaluating pleural disorders. New, faster 1.5 T equipment with fast gradients (over 20 mT) that enable breath-hold sequences and the routine use of contrast agents (gadolinium – DTPA) has become available. The motion artefacts are effectively suppressed and the quality of thoracic MRI has markedly improved. We compared the findings from 26 pairs of CT scans and MR images from patients with malignant pleural mesothelioma, in order to evaluate the clinical utility and limitations of MRI as a primary imaging method for MPM compared to CT.

Between September 1996 and December 1997 parallel paired CT and MRI scans were obtained from 14 patients with histologically confirmed malignant pleural mesothelioma. All patients were participating in clinical multimodality therapy trials which included the taking of serial images in order to define treatment response. There were 13 male patients and one female, mean age 58 years (44–75). Ten patients had epithelial type tumours, one patient had a mixed type tumour, one patient had a sarcomatoid type tumour and for two patients the subtype was not defined. Tumour stage was assessed using the IUAC staging system. Five patients had stage IV disease, eight had stage III disease and one had stage I disease according to primary CT scans and/or thoracoscopy/thoracotomy. In all cases the time from the diagnostic thoracoscopy/thoracotomy to the first comparative imaging procedure was 1 month or longer, in order to avoid the interpreting difficulties the acute postoperative changes may create. The CT scans were performed using one of two scanners (Siemens Somatom plus S B30B or Siemens Somatom 4 B20B, Germany). Imaging parameters were 140 kV, 170 mA, pitch 1, increment 10. An i.v. bolus injection of iohexol at a dose of 300 mg/l per kg of body weight (Omnipaque®) was used for contrast enhancement. The entire thorax and the upper abdomen was scanned in each patient. All 26 CT scans were of comparable quality and technique. The MR images were obtained using a 1.5 T magnet imager (Magnetom Vision, Siemens). A phased-array surface coil and breath-hold sequences were used. Imaging parameters were T, 2d flush, haste, true fisp (Table 1). Gadolinium– DTPA 0.1 mmol/kg (Dotarem®) was used as a

Table 1 Imaging parameters used for MRI examinations

2d-Flush True fisp Haste

TR (ms)

TE (ms)

Flip (degree)

Slice (mm)

Matrix

Time (s)

85.6 5.0 4.2

4.1 2.5 59.0

50 70 150

10 5 10

140×256 224×256 128×256

24 12 18

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

217

Fig. 1. MPM of the right hemithorax in a 56-year-old man. (a) The contrast-enhanced CT scan (mediastinal window) shows the pleural thickening and pleural fluid posteriorly (arrow). (b) The sagittal T1-weighted MR image depicts the nodular, interlobar pleural thickening and the fissural growth of the tumour (arrow). (c) The CT scan (lung window) of the axial imaging direction, parallel to the interlobar pleura, raises a problem in differentiation of interlobar pleural changes from parenchymal lesions (arrow). (d) The sagittal T1-weighted MR image shows the early invasion of the chest wall (straight arrow) and the tumour extending to the posterobasal pleural space (curved arrow) and also diaphragmatic invasion (arrowhead) is seen. (e) The axial T2-weighted MR image allows differentiation between the tumour (arrowhead), which has intermediate signal intensity, and the pleural fluid (arrow), which has high signal intensity.

contrast agent. Post contrast images were obtained in the axial, coronal and sagittal planes. Two different thoracic radiologists for CT and MRI evaluated the examinations separately and independently without knowledge of each other’s interpretations or the patient’s clinical status. Special attention was given to pleural surfaces, pleural fluid and its differentiation from the tumour, chest wall, diaphragm, mediastinal soft tissue and lymph nodes, and subcutaneous tissue. The mediastinal lymph nodes were considered abnormal when the short axis diameter of a node was more than 1 cm seen on two planes and this normally means a diameter of 1.5 cm maximum. When the lymph nodes had central necrosis and/ or were not homogenously contrast enhanced, they were also considered pathological [5]. The number of definite and specified abnormal findings in all CT scans or MRIs was counted.

3. Results Altogether there were 26 sets of paired CT and MRI scans available for comparative analysis. The mean time between the two procedures was 10 days (range 0–20 days). Calcifications of the pleura were seen on 13 CT scans, but only on three T2-weighted MR images. Tumour spread into the interlobar fissures was seen more often on contrast-enhanced T1-weighted images than on CT scans (Fig. 1). Thickening of pericardium was seen equally well using either method. Pericardial fluid was suspected from the CT scan of three patients, but any of the T2-weighted MR images from these patients did not confirm the same finding (Table 2). Invasion of the diaphragm was seen on the CT scans from 11 patients (18 scans), but was identified on contrast-enhanced T1-weighted images

218

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

Fig. 1. (Continued)

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

Fig. 1. (Continued)

219

220

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

Table 2 Pleural changes Site (n= abnormalities identified)

Pleural fluid Calcifications of the pleura Enhancement (tumour) of interlobar fissures Thickening of pericardium

CT contrast enhanced

22 13 15 16

from 13 patients (24 scans). The extension of the disease through the diaphragm to the peritoneum and/or adjacent organs was easily seen on contrast-enhanced T1-weighted MR images from eight patients (13 scans), but was only seen on the CT scans of three of these patients (Fig. 2). The tumour was seen to have invaded the lung parenchyma on the CT scans of 13 patients (21 scans); 12 of these lung parenchyma invasions were also identified on contrast-enhanced T1weighted images (Fig. 3). In five MRI scans the tumour was seen to have invaded bony structures, but the comparable CT scans did not show these changes (Table 3). Mediastinal lymph node involvement, N1 disease, was seen in three patients on both the CT and MRI scans. N2 lymph nodes were identified in nine patients using CT (15 scans) and in six patients on the contrast-enhanced MRI scans (Table 4).

4. Discussion CT is now routine clinical practice for assessing stage and tumour progression in mesothelioma. Previous studies have shown that for 88% of patients, the pathological staging established at thoracotomy corresponds to the clinical stage established using CT [6]. However it is not always an accurate method for detecting the early stages of pleural pathology, chest wall involvement, invasion of the diaphragm and through the diaphragm or mediastinal lymph node metastases [7,8]. Since it is now suggested that selected patients might benefit from extensive surgery, accurate clinical staging is crucial for decision making

MRI T1 contrast enhanced

T1

T2

19 2 25 14

15 1

20 3

8

10

or when evaluating tumour resectability. Differentiation of thin pleural tumour and compromising pleural fluid can also be difficult using CT, which may lead to erroneous tumour response assessments in treatment trials. CT is faster than MRI and less costly at the moment. CT contrast agents may, however, cause hypersensitivity problems and are nephrotoxic and, thus, contraindicated for patients with impaired renal function. Multiplanar images reconstructed from axial spiral CT scans can now be obtained using the most modern equipment. However, the image quality is not equal to the direct multiplanar MR images, because image reprocessing always loses some of the data [9]. Staging established at thoracotomy has its limitations and cannot always be considered the ‘gold standard’, because N2 involvement, destruction of bony structures and interlobar tumour invasion may be difficult to identify. For our patients it was not possible to perform parallel CT and MRI scans preoperatively in referring hospitals and before the diagnosis was established. Therefore our study material comprises sequential images obtained at follow-up. Earlier studies have reported that MRI with its capacity to produce multiplanar images and excellent contrast resolution can be used to accurately stage patients with primary thoracic tumours [7] and demonstrate tumour extension and associated mediastinal or pleural disease [4,10]. The T2-weighted MR images in true fisp sequence without fat suppression show fluid with increased signal intensity (white) in comparison to the lower signal intensity of tumour tissue (grey). These images have high sensitivity for detecting fluid and superiorly differentiate tumour from the

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

221

Fig. 2. Chest wall and diaphragmatic invasion of MPM in the right hemithorax of a 51-year-old man. (a) The contrast-enhanced CT scan shows the reduction in the size of the right hemithorax and the tumour extends through the right anterior chest wall (arrow). (b) The sagittal T1-weighted MR image clearly visualises the invasion of the diaphragm (straight arrow), tumour growth into the interlobar fissures (curved arrow) and the chest wall invasion anteriorly (arrowhead). (c) The coronal T1-weighted MR image demonstrates the tumour extending down to the insertion site of the diaphragm (arrow) and the tumour is also invading the lung parenchyma (arrowhead). (d) The tumour growth to the insertion site of the diaphragm seen in Fig. 1c is visible by axial CT scans only if imaged caudally enough to the level of L 1–2 (arrow).

222

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

Fig. 2. (Continued)

compromising pleural fluid, which can be very helpful in the early stages of MPM. Whereas the T2-weighted fat suppressed images in haste sequence show high signal intensity (white) for both

tumour and fluid and are therefore less useful for this purpose. Pericardial fluid was suspected from the CT in three patients, but we consider these findings to be false positives: in CT the pericardial

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

223

Fig. 3. MPM of the left hemithorax in a 53-year-old female. (a) The contrast-enhanced CT scan obtained at the level of the sternoclavicular joints shows large pleural effusion. Tumorous pleural thickening is seen only paravertebrally (arrow), but not in the dorsal or lateral pleural space. (b) The coronal contrast-enhanced T1-weighted MR image visualises well the circumferential tumour growth (curved arrow) with moderate loss of lung volume and loculated pleural fluid (arrow).

fat may visualise as oedematic and hazy and thus difficult to distinguish from fluid and T2-weighted images (true fisp sequence) did not show high

signal intensity (white) suggesting pericardial fluid in these cases. Contrast-enhanced T1-weighted images are the most reliable for showing the

224

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

Table 3 Tumour invasion Site (n= abnormalities identified)

CT contrast enhanced

Invasion of diaphragm Extension through diaphragm to peritoneum Invasion through pleura to lung parenchyma Chest wall invasion Destruction of rib or spine

MRI

18 6 21 17 0

active enhancement of the tissues. Especially in pleural disease and in chest wall changes contrast-enhanced T1-weighted MR is advantageous because of the improved ability to delineate tumorous lesions from normal structures [11,12]. Plain T1-weighted images and T2-weighted images are not able to provide this information. However, plain T1-weighted images are needed for comparison with contrast-enhanced T1 images when viewing tissues which actively enhance. Sagittal and coronal MR images enable better visualization of tumour spread into the interlobar fissures and invasion of the diaphragm and through the diaphragm than axial CT scans. No bone metastases were detected by CT, but tumour invading the ipsilateral ribs was seen in five contrast-enhanced T1-weighted MRIs. MRI is capable of showing oedema in the bone marrow and is therefore accurate in demonstrating the invasion of bony structures by the tumour [13,14]. Chest wall involvement and mediastinal soft tissue invasion were equally well demonstrated by contrast enhanced CT and MRI. We found that the use of a contrast agent also with CT to

T1 contrast enhanced

T1

T2

24 13 20 19 5

10 6 12 6 3

16 8 13 11 2

be essential, because the contrast agent shows the dynamic processes. We found that CT seemed better able to detect enlarged lymph nodes suspected to be pathological as defined by the criteria mentioned earlier. However, neither CT nor MRI scans can be used to accurately assess lymph node staging because of low sensitivity (normal-sized nodes may contain microscopic metastases) and low specificity (enlarged lymph nodes may be reactive). Nodal size is still the only useful criterion for evaluating lymph node metastases: the contrast enhancement of the lymph nodes is less marked than the enhancement of the tumour tissue and the use of contrast agents does not improve accuracy with respect to metastatic lymph nodes [7]. MRI has become quicker, but the imaging time is still about 20 min and the capacity for 10–20 s breath-holds is required. The MRI contrast agent (gadolinium–DTPA) is markedly safer than iodine-based agents; it is neither nephro-, neuro- or cardiotoxic and allergic reactions are extremely rare. There is no irradiation involved in MRI and this aspect will become more important if imaging methods are used repeatedly in wider screening programmes.

Table 4 Lymph nodes Site (n= abnormalities identified)

Metastatic lymph nodes N1 Metastatic lymph nodes N2

CT contrast enhanced

4 15

MRI T1 contrast enhanced

T1

T2

4 8

1 5

0 3

A. Knuuttila et al. / Lung Cancer 22 (1998) 215–225

5. Conclusion We conclude that MRI with its multiplanar images is more accurate than CT in showing diaphragmatic tumour invasion and extension of the disease through the diaphragm into the peritoneum. MRI also visualised tumour extension into the interlobar fissures and the destruction of bony structures better than CT. MRI does not show the inactive pleural plaques, but these are not usually of interest when assessing MPM. We found that MRI is better for evaluating the growth pattern and extent of MPM and should be more widely used, especially when evaluating tumour resectability and in research protocols when an accurate evaluation of disease extent is essential. Hopefully this will also lead to a greater understanding of the biology of this tumour, and to more successful treatment strategies for this devastating disease.

References [1] Rusch VW, Piantadosi S, Holmes EC. The role of extrapleural pneumonectomy in malignant pleural mesothelioma. J Thorac Cardiovasc Surg 1991;102(1):1–9. [2] Sugarbaker DJ, Strauss GM, Lynch TJ, et al. Node status has prognostic significance in the multimodality therapy of diffuse, malignant mesothelioma. J Clin Oncol 1993;11(6):1172– 8. [3] Herman SJ, Winton TL, Weinsbrod GL, Towers MJ, Mentzer SJ. Mediastinal invasion by broncogenic carcinoma: CT signs. Radiology 1994;190:841–6.

.

225

[4] Kuhlman JE, Bouchardy L, Fishman EK, Zerhouni EA. CT and MR imaging evaluation of chest wall disorders. Radiographics 1994;14:571 – 95. [5] Pass HI, Mitchell JB, Johnson DH, Turrisi AT. Lung Cancer: Principles and Practice. Philadelphia: LippincottRaven, 1996:445 – 448. [6] Maasilta P, Vehmas T, Kivisaari L. Correlations between findings at computed tomography and at thoracoscopy/ thoracotomy/autopsy in pleural mesothelioma. Eur Respir J 1991;4:952 – 4. [7] Bonomo L, Ciccotosto C, Guidotti A, Storto ML. Lung cancer staging: the role of computed tomography and magnetic resonance imaging. Eur J Radiol 1996;23:35 – 45. [8] Fortier M, Mayo JR, Swensen SJ, Munk PL, Vellet DA, Muller NL. MR imaging of chest wall lesions. Radiographics 1994;14:597 – 606. [9] Remy J, Remy-Jardin M, Artaud D, Fribourg M. Multiplanar and three-dimensional reconstruction techniques in CT: impact on chest diseases. Eur Radiol 1998;8:335 – 51. [10] Patz EF Jr., Shaffer K, Piwnica-Worms DR, et al. Malignant pleural mesothelioma: value of CT an MRI in predicting resectability. Am J Roentgenol 1992;159:961 – 6. [11] Kono M, Adachi S, Kusumoto M, Sakai E. Clinical utility of Gd – DTPA-enhanced magnetic resonance imaging in lung cancer. J Thorac Imaging 1993;8:18 – 26. [12] Low RN, Sigeti JS, Song SYT, Shimakawa A, Pelc NJ. Dynamic contrast-enhanced breath-hold MR imaging of thoracic malignancy using cardiac compensation. J Magn Reson Imaging 1996;6:625 – 31. [13] Hochstenbag MH, Snoep G, Cobben NA, et al. Detection of bone marrow metastases in SCLC. Comparison of MRI with standard methods. Eur J Cancer 1996;32A(5):779– 82. [14] White PG, Adams H, Crane MD, Butchart EG. Preoperative staging of carcinoma of the bronchus: can computed tomography scanning reliably indentify stage III tumours? Thorax 1994;49:951 – 7.