| [Download PDF]
|Year : 2010 | Volume
| Issue : 3 | Page : 280--286
Benign vs malignant soft tissue neoplasms: Limitations of magnetic resonance imaging
J Sen1, S Agarwal1, S Singh2, R Sen2, S Goel1,
1 Department of Radiodiagnosis, Pt. B.D. Sharma Post Graduate Institute of Medical Sciences, Rohtak, Haryana- 124001, India
2 Department of Pathology, Pt. B.D. Sharma Post Graduate Institute of Medical Sciences, Rohtak, Haryana- 124001, India
Department of Radiodiagnosis, Pt. B.D. Sharma Post Graduate Institute of Medical Sciences, Rohtak, Haryana- 124001
Aims: Various features have been described in the literature to differentiate benign from malignant lesions. The aim of the present study was to study the accuracy of each of these features and that of magnetic resonance imaging (MRI) in diagnosing malignant lesions. Materials and Methods: Fifty-five consecutive patients presenting with neoplastic (both benign and malignant) lesions diagnosed clinically and on ultrasound were studied and their MRI features were compared with the findings on surgical exploration and histopathologic examination. Results: There were 32 (58%) benign and 23 (42%) malignant masses. Malignant masses were more common in patients older than 20 years (83%), and these had symptoms of less than 6 months duration (75%), as against benign lesions. The swelling was painful in 8 malignant masses and these were more common in the upper limbs (61%). Various features of malignant lesions were size more than 5 cm in 83%, change in signal intensity from homogenous on T1-weighted images to heterogenous on T2-weighted images in 74%, irregular margins in 74%, and heterogenous contrast enhancement in 91%. The accuracy of these features was 76%, 58%, 78%, and 60%, respectively. Most benign and malignant lesions were intramuscular in location. A significant number (38%) of benign lesions were located in the intermuscular facial plane. Definitive diagnosis was made in 42% of the lesions. Conclusions: MRI is an excellent modality for evaluating soft tissue neoplasms; however, prediction of a specific diagnosis and differentiation of malignant and benign lesions is not always possible.
|How to cite this article:|
Sen J, Agarwal S, Singh S, Sen R, Goel S. Benign vs malignant soft tissue neoplasms: Limitations of magnetic resonance imaging.Indian J Cancer 2010;47:280-286
|How to cite this URL:|
Sen J, Agarwal S, Singh S, Sen R, Goel S. Benign vs malignant soft tissue neoplasms: Limitations of magnetic resonance imaging. Indian J Cancer [serial online] 2010 [cited 2020 Jul 6 ];47:280-286
Available from: http://www.indianjcancer.com/text.asp?2010/47/3/280/64725
The treatment and prognosis of soft tissue tumors depends on histologic diagnosis, tumor localization and spread, and the presence and absence of metastasis. Various radiologic techniques are available to characterize these tumors. One of the earliest musculoskeletal applications of magnetic resonance imaging (MRI) was in the evaluation of bone and soft tissue tumors.  Today, a correct assessment of disorders is provided by MRI because of its high soft tissue contrast resolution and multiplanar imaging capability.  It is particularly well suited to determine the extent of soft tissue, bone, neurovascular, and joint involvement, which are critical factors for staging and treatment planning.  Gadolinium diethylene triamine acetic acid (Gd-DTPA)-enhanced images not only provide information on vascularity of the tumor, but also enhance the demarcation between tumor, muscle, and edematous tissue in some cases. An important limitation of MRI, however, is its relative inability to detect soft tissue calcification. 
Petterson et al. concluded from their study that for soft tissue tumors and bone tumors with soft tissue extension, MRI was significantly better than the other modalities in all variables examined: delineation between tumor and muscle, tumor and vessel, tumor and fat, tumor and joint, tumor and bone, as well as depicting intralesional necrosis and bleeding.
Certain soft tissue tumors have characteristic MR appearance, in which a specific diagnosis may be made or strongly suspected. These include lipomas; liposarcoma; benign vascular lesions, such as hemangioma, arteriovenous malformation; hemosiderin-laden lesions, such as pigmented villonodular synovitis, ganglion cyst, and benign neural tumors.  Moulton et al were able to predict the diagnosis confidently and correctly in 44% of 225 cases of soft tissue tumors.
Because of its high contrast resolution, it was expected that MRI had great potential for the histologic classification of soft tissue tumors. Unfortunately, the initial enthusiasm has not been entirely confirmed.
Various features have been described in the literature to differentiate benign from malignant lesions. The aim of the present study was to study the accuracy of each of these features and that of MRI in diagnosing malignant lesions.
Materials and Methods
Fifty-five consecutive patients who presented to the outpatient department with swelling and who were suspected to have neoplastic (both benign and malignant) lesions clinically and on ultrasound were included in the study. All patients with contraindications to MRI examination were excluded from the study. MRI was performed on 1.5 Tesla (The Netherlands) using T1-weighted (TR/TE 500/16 [repetition time, ms/echo time, ms]), T2-weighted (TR/TE 3000/100), Short Tau Inversion Recovery (STIR) (TR/TE 2500/100)/fat-suppressed sequences and Gd-DTPA-enhanced T1-weighted scan (0.1 mmol/kg of body weight gadopentetate dimeglumine (Berlin, Germany) was used intravenously). Matrix size of 512 Χ 512 was used. Slice thickness, intersection gap, and field of view varied depending on the size of the tumor. The findings so obtained were compared with the findings on surgical exploration and histopathologic examination.
There were 32 (58%) benign and 23 (42%) malignant lesions. Commonest benign masses were of neurogenic origin (20%) and commonest malignant masses were malignant fibrous histiocytoma (12%). Benign masses were commonly seen in patients younger than 20 years (54%), whereas malignant masses were commonly seen in patients older than 20 years (83%). Sex ratio was nearly the same.
Most of the patients with malignant masses presented with symptoms of less than 6 months duration (75%) as against patients with benign masses (92%). The swelling was painful in 8 malignant masses and in 2 benign masses. The benign masses were slightly more common in the lower limbs 19/32 (59%) and malignant masses were more common in the upper limbs 14/23 (61%). MRI features of all the tumors are shown in [Table 1]. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of size of more than 5 cm size in predicting the malignant potential of a soft tissue mass lesion [[Figure 1]a, b and c ] were 83%, 72%, 68%, 85%, and 76%, respectively. Similarly, sensitivity, specificity, PPV, NPV, and accuracy change in signal intensity from homogenous on T1-weighted images to heterogenous on T2-weighted images [[Figure 1]a and b] for predicting malignancy were 67%, 50%, 58%, 59%, and 58%, respectively. While, irregularity of margins [[Figure 1]a and b] and heterogenous contrast enhancement [Figure 1]c had a sensitivity, specificity, PPV, NPV, and accuracy of 74%, 81%, 74%, 81%, 78%, and 91%, 38%, 51%, 86%, 60%, respectively, for predicting malignancy. Definitive diagnosis could be made in only 23/55 cases (42%) [Figure 2]a, b, [Figure 3]a, and b]. Overall sensitivity, specificity, PPV, NPV, and accuracy of MRI for diagnosing malignant mass lesions were found to be 83%, 81%, 76%, 87%, and 82%, respectively.
Characterization consists of both grading and tissue-specific diagnosis. Whereas grading implies a differentiation between benign and malignant tumors and definition of malignancy grades, tissue-specific diagnosis implies pathologic typing. Although pathologic diagnosis is the gold standard in the diagnosis of soft tissue tumors, prediction of a specific histologic diagnosis remains one of the ultimate goals of each new imaging technique. 
Current guidelines suggest that the most important variables for assessing the risk of malignancy in a soft tissue lesion include size, depth in relation to fascia, increasing size, and pain.  Datir et al. concluded that significant risk factors for malignancy include increasing patient age and lesion size greater than or equal to 5 cm. Tung et al. combined the data from 3 investigations and postulated that a diameter of less than 3 cm is a reasonable indicator that a lesion is benign. Conversely, a diameter of 5 cm predicts a malignant nature of a soft tissue mass with a sensitivity of 74%, specificity of 59%, and an accuracy of 66%. In our series also malignant lesions were seen in the older age group. However, only 8 malignant lesions were painful. Majority of malignant lesions (83%) were more than 5 cm in size, whereas 72% of benign tumors were less than 5 cm in size.
Moulton et al,  in a study of 225 soft tissue masses could detect 59% malignant tumors in intramuscular or mixed (intramuscular + subcutaneous) location, whereas 88% benign tumors were in the subcutaneous tissue. In the present study, majority of malignant lesions detected on MRI were deep in intramuscular location (83%). Two (9%) malignant tumors were found in the subcutaneous tissue predominantly, whereas 53% benign tumors were centered within the muscle and 12 cases (38%) were in the intermuscular fascial planes. However, Datir et al. found that relationship to fascia is less important as a predictor of malignant potential.
Benign tumors are well delineated and, malignant tumors have rather ill-defined margins, Bongartz et al,  however, reported that aggressive sarcomas may have a pseudocapsule, whereas benign lesions, such as desmoid tumors may invade neighboring tissues. They concluded that the margin (well-defined vs infiltrating) of soft tissue mass on MRI was of no statistical relevance in the prediction of malignancy. In our study, majority of benign tumors (81%) had well-defined margins. However, 6 benign tumors had ill-defined margins.
More recently, Fernebro et al. suggested that MRI evaluation of soft tissue sarcomas should focus on the peripheral growth pattern because it adds prognostic information of value for decisions on neoadjuvant therapy. Diffusely infiltrative growth on MRI gave 2.5 times increased risk of metastases and a 3.7 times higher risk of local recurrence.
Commonly used individual parameters for predicting malignancy are intensity and homogeneity of the MR signal with different pulse sequences. High signal intensity on T2-weighted images is a sensitive parameter but has an unacceptably low specificity.  In our study, 50% benign tumors were hypointense and 25% were hyperintense on T1-weighted images. On T2-weigted images, 84% benign tumors were hyperintense. The malignant masses were hyperintense in 22% cases on T1-weighted and in 100% cases on T2-weighted images. Similarly Harmann et al. reported that 17% benign tumors were hypointense and 58% were hyperintense on T1-weighted images and 85% benign tumors were hyperintense on T2-weighted images. Forty percent of the malignant tumors were hyperintense on T1-weighted images and 100% were hyperintense on T2-weighted images.
Although smaller lesions tend to be more homogenous, whether they are benign or malignant, 90% of malignant lesions are inhomogenous (disorganized or hectic). Absence of heterogeneity is a reliable negative predictive indicator for the presence of malignancy.  Chen et al. found that positive predictive value of necrosis for malignancy was 84.8% and specificity was 90.9%.
Hermann et al. reported that changing homogeneity (from homogenous on T1-weighted images to heterogenous on T2-weighted images) and the presence of lobular morphology with intervening low signal intratumoral septations had a sensitivity of 72% and 80%, respectively, and a specificity of 87% and 91%, respectively, in predicting malignancy.
In the present study, the sensitivity and specificity were 67% and 50%, respectively.
Schepper et al. reported that although malignant tumors show increased vascularity and have large extracellular spaces, depending on tumoral activity or aggressiveness, there was no correlation between the degree of and pattern of enhancement and malignancy grade.
The results of study conducted by Ma et al. suggest that their rim-to-centre differential enhancement ratio has potential as an additional parameter for the MRI differentiation of indeterminate musculoskeletal masses. Similar results were obtained in our series with most of the malignant tumors (91%) showing heterogenous contrast enhancement.
Schepper et al. reported that the involvement of an adjacent bone, extracompartmental distribution, and encasement of the neurovascular bundle are relatively uncommon findings that are specific but are insensitive signs of malignancy. Crim et al. reported neurovascular bundle involvement in 4% benign and 18% malignant tumors. Berquist et al. found neurovascular bundle involvement in 78% malignant tumors. This was also observed with desmoid tumor (benign) in their series. Similarly, in our series 15 benign (47%) and 10 malignant tumors (43%) showed neurovascular bundle involvement. Bone was involved in 2 malignant tumors (9%) in our series. Crim et al. reported similar findings of bone involvement in 2% benign and 6% malignant masses.
Peritumoral edema, shown on T2-weighted images as an ill-defined area of high signal intensity, can indicate infiltrative tumor, reparative inflammation, or both, and as a consequence is not helpful as a grading parameter. 
Intratumoral hemorrhage is a rare finding, which can be observed in both benign and malignant lesions, and is difficult to differentiate from nontumoral soft tissue hematoma. In a study by Moulton et al,  intratumoral hemorrhage was observed in 23 benign and in 5 malignant tumors among a total of 225 masses. Our study detected intratumoral hemorrhage in 02 benign (6%) and 4 malignant tumors (17%) among a total of 55 cases.
Kransdorf et al in a review of 112 soft tissue tumors (85 benign, 27 malignant) showed that MRI could not reliably differentiate between benign and malignant soft tissue tumors. They also found that MR images were sufficiently characteristic to enable specific diagnosis only in 24% of cases. Specific diagnosis could be made in 42% of the cases in our study.
Berquist et al. reported sensitivity and specificity of 88% and 90%, respectively, for the diagnosis of benign tumors and 94% and 90%, respectively, for the diagnosis of malignant tumors. For the diagnosis of malignant lesions, Moulton et al. reported a sensitivity of 78%, specificity of 89%, PPV of 65%, and NPV of 94%. When the diagnostic benign tumors are excluded, the specificity and NPV decreased to 76% and 86%, respectively. In a prospective analysis of 36 consecutive cases of soft tissue tumors, Ma et al. found that MRI was 100% sensitive but had only 17% specificity and 58% accuracy in predicting malignancy. They also found a wide variability in the appearance of benign and malignant lesions on MR images, with poor correlation between "benign characteristics" and the benignity of the lesion.
In our study, all 19/23 were correctly diagnosed as malignant lesions. These were hyperintense and heterogenous on T2-weighted images. Margins were partially or completely irregular in all except 05. The size was >5 cm in all except 02. Neurovascular bundle was involved in 09 patients with malignant lesions, bone involvement was seen in 02, and intratumoral hemorrhage was seen in 04. Heterogenous contrast enhancement was seen in all.
No single MRI feature has been found diagnostic to reliably distinguish benign from malignant lesions in any of the studies. Schepper et al. found that the signs that had the greatest specificity for malignancy included the presence of tumor necrosis, bone or neurovascular involvement, and a mean diameter of more than 66 mm. Kalayanarooj et al. found heterogenous signal on T2-weighted, perilesional edema or invasion, and necrosis in the masses to be statistically significant for differentiation between benign and malignant soft tissue masses. While, Pang et al. found that significant risk factors for predicting malignancy include increasing patient age and lesion size greater than or equal to 5 cm, they also found that the relationship to the fascia was less important as a predictor for malignant potential. Chen et al. on the other hand found that the parameters favoring malignancy were large lesion size, peritumoral edema, necrosis, and absence of calcification, absence of fibrosis, and lack of fat rim.
MRI is an excellent modality for evaluating the size, extent, intensity of characteristics, and involvement of surrounding structures. However, in our study the overall sensitivity, specificity, PPV, NPV, and accuracy for diagnosing malignancy were found to be 83%, 81%, 76%, 87%, and 82%, respectively. Also specific diagnosis could be made only in 42% of the cases. These findings are similar to the findings seen in other studies and hence we conclude that it is of limited value for this purpose.
|1||Kransdorf MJ. MRI and CT evaluation of primary bone and soft-tissue tumors. AJR Am J Roentgenol 2006;187:16-7.|
|2||Kransdorf MJ, Jelinek JS, Moser RP Jr. Imaging of soft tissue tumors. Radiol Clin North Am 1993;31:359-72.|
|3||Siegel MJ. Magnetic resonance imaging of musculoskeletal soft tissue masses. Radiol Clin North Am 2001;39:701-20.|
|4||Pettersson H, Gillespy T 3rd, Hamlin DJ, Enneking WF, Springfield DS, Andrew ER et al. Primary musculoskeletal tumors: Examination with MR imaging compared with conventional modalities. Radiology 1987;164:237-41.|
|5||Moulton JS, Blebea JS, Dunco DM, Braley SE, Bisset GS 3rd, Emery KH. MR imaging of soft tissue masses: Diagnostic efficacy and value of distinguishing between benign and malignant lesions. Am J Roentgenol 1995;164:1191-9.|
|6||Datir A, James SL, Ali K, Lee J, Ahmad M, Saifuddin A. MRI of soft tissue masses: The relationship between lesion size, depth, and diagnosis. Clin Radiol 2008;63:373-8.|
|7||Tung GA, Davis LM. The role of magnetic resonance imaging in the evaluation of soft tissue mass. Crit Rev Diagn Imaging 1993;34:239-8.|
|8||Bongartz G, Vestring T, Peters PE. Magnetic resonance tomography of soft tissue tumors. Radiologe 1992;32:584-90.|
|9||Fernebro J, Wiklund M, Jonsson K, Bendahl PO, Rydholm A, Nilbert M, et al. Focus on the tumour periphery in MRI evaluation of soft tissue sarcoma: Infiltrative growth signifies poor prognosis. Sarcoma 2006;2006:21251.|
|10||Weatherall PT. Benign and malignant masses. MR imaging differentiation. Magn Reson Imaging Clin N Am 1995;3:669-94.|
|11||Hermann G, Abdelwahab IF, Miller TT, Klien MJ, Lewis MM. Tumour and tumour -like conditions of the soft tissue: Magnetic resonance imaging features differentiating benign from malignant masses. Br J Radiol 1992;65:14-20.|
|12||Schepper AM De. Grading and characteristics of soft tissue. 2 nd ed. Imaging of soft tissue tumours. In: Schepper AM De, Parizel PM, Buckelaer L De, editors. Berlin: Springer; 2001. p. 123-41.|
|13||Chen CK, Wu HT, Chiou HJ, Wei CJ, Yen CH, Chang CY, et al. Differentiating benign and malignant soft tissue masses by magnetic resonance imaging: Role of tissue component analysis. J Chin Med Assoc 2009;72:194-201.|
|14||Ma LD, Frassica FJ, Scott WW Jr, Fishman EK, Zerhouni EA. Differentiation of benign and malignant musculoskeletal tumors: Potential pitfalls with MR imaging. Radiographics 1995;15:349-66. |
|15||Crim JR, Seeger LL, Yao L, Chandnani V, Eckardt J. Diagnosis of soft tissue masses with MR imaging: Can benign masses be differentiated from malignant ones? Radiology 1992;185:581-6.|
|16||Berquist TH, Ehman RL, King BF, Hodgman CG, Ilstrup DM. Value of MR imaging in differentiating benign from malignant soft tissue masses: Study of 95 lesions. Am J Roentgenol 1990;155:1251-5.|
|17||Hanna SL, Fletcher BD. MR imaging of malignant soft-tissue tumors. Magn Reson Imaging Clin N Am 1995;3:629-50.|
|18||Kransdorf MJ, Jelinek JS, Moser RP Jr, Utz JA, Brower AC, Hudson TM, et al. Soft-tissue masses: Diagnosis using MR imaging. AJR Am J Roentgenol 1989;153:541-7.|
|19||Ma LD, Frassica FJ, McCarthy EF, Bluemke DA, Zerhouni EA. Benign and malignant musculoskeletal masses: MR imaging differentiation with rim-to-centre differential enhancement ratios. Radiology 1997;202:739-44.|
|20||De Schepper AM, Ramon FA, Degryse H. Statistical analysis of MRI parameters predicting malignancy in 141 soft tissue masses. Rofo 1992;156:587-91.|
|21||Kalayanarooj S. Benign and malignant soft tissue mass: Magnetic resonance imaging criteria for discrimination. J Med Assoc Thai 2008;91:74-81.|
|22||Pang KK, Hughes T. MR imaging of the musculoskeletal soft tissue mass: Is heterogeneity a sign of malignancy? J Chin Med Assoc 2003;66:655-61.|