Soft-Tissue Tumors Evaluated by Line-Scan Diffusion-Weighted Imaging: Influence of Myxoid Matrix on the Apparent Diffusion Coefficient

Transcription

1 JOURNAL OF MAGNETIC RESONANCE IMAGING 25: (2007) Original Research Soft-Tissue Tumors Evaluated by Line-Scan Diffusion-Weighted Imaging: Influence of Myxoid Matrix on the Apparent Diffusion Coefficient Masayuki Maeda, MD, 1 * Akihiko Matsumine, MD, 2 Hiroya Kato, MD, 3 Katsuyuki Kusuzaki, MD, 2 Stephan E. Maier, MD, PhD, 4 Atsumasa Uchida, MD, 2 and Kan Takeda, MD 1 Purpose: To compare the apparent diffusion coefficients (ADCs) of myxoid and nonmyxoid soft-tissue tumors using line-scan diffusion-weighted imaging (LSDWI), and to investigate the myxoid matrix influence on ADCs of soft-tissue tumors. Materials and Methods: This study enrolled 44 patients with soft tissue tumors. They were divided into two groups: one with myxoid-containing soft-tissue tumors (N 23) and the other with nonmyxoid soft-tissue tumors (N 21). The 44 patients were also classified histologically into 26 with malignant soft-tissue tumors and 18 with benign softtissue tumors. LSDWI was performed using b values of 5 and 1000 second/mm 2. The ADCs of the tumors were calculated and compared for myxoid and nonmyxoid tumors and for benign and malignant tumors. Results: The ADC (mean SD) was mm 2 /second in myxoid containing tumors, whereas the ADC was mm 2 /second in nonmyxoid tumors. The ADCs of the myxoid and nonmyxoid tumors were significantly different (P 0.01). The ADCs were mm 2 /second in malignant tumors and mm 2 /second in benign tumors. The ADCs of benign and malignant soft-tissue tumors were not significantly different. Conclusion: The ADCs of myxoid-containing soft-tissue tumors were significantly higher than those of nonmyxoid soft-tissue tumors. The myxoid matrix influences ADCs of both benign and malignant soft-tissue tumors. Key Words: magnetic resonance imaging; diffusion; soft tissue tumor; myxoid; apparent diffusion coefficient J. Magn. Reson. Imaging 2007;25: Wiley-Liss, Inc. 1 Department of Radiology, Mie University School of Medicine, Tsu, Japan. 2 Department of Orthopaedic Surgery, Mie University School of Medicine, Tsu, Japan. 3 Department of Pathology, Mie University School of Medicine, Tsu, Japan. 4 Department of Radiology, Brigham and Women s Hospital, Boston, Massachusetts, USA *Address reprint requests to M.M., MD, Department of Radiology, Mie University School of Medicine, Edobashi, Tsu, Mie , Japan. mmaeda@clin.medic.mie-u.ac.jp Received February 3, 2006; Accepted December 21, DOI /jmri Published online in Wiley InterScience ( DIFFUSION-WEIGHTED IMAGING (DWI) has been used for assessment of tumors, particularly those of the central nervous system (1 4). Results of studies comparing apparent diffusion coefficients (ADCs) and histopathological findings have strongly suggested that greater cellularity is associated with more restricted diffusion (1,3,4). It has been reported that DWI has the potential to differentiate benign and malignant soft-tissue tumors because malignant tumors have greater cellularity and therefore have more restricted diffusion than benign tumors (5). On the other hand, some investigators have reported that the ADC values of benign soft-tissue tumors and sarcomas overlap and are therefore not useful to differentiate between bulk benign and malignant tumors (6). The pathological composition of interstitial spaces, in particular soft-tissue tumors, is peculiar compared to that of brain tumors. For example, myxoid matrix, which is rarely seen in brain tumors, is found occasionally in several soft-tissue tumors, whether benign or malignant (7). We surmised that myxoid matrix influences ADC values of soft-tissue tumors. The purpose of this study was to evaluate the ADCs of myxoid and nonmyxoid soft-tissue tumors using line-scan DWI (LSDWI). MATERIALS AND METHODS Patients We enrolled 44 patients (19 women and 25 men; age 6 82 years: mean age 50.5 years) with soft-tissue tumors in the extremities or on the trunk. All patients were examined before initiation of treatment. In all patients, diagnoses were confirmed using histologic biopsy and surgical specimens. Our study included various histologic entities, but excluded ganglionic cyst, synovial cyst, and typical lipoma. Based on pathologic analyses, the patients were divided into two groups: one with myxoid-containing soft-tissue tumors (N 23) and the other with nonmyxoid soft-tissue tumors (N 21). Based on histologic data, these 44 patients were classified into 26 patients with malignant soft-tissue tumors and 18 patients with benign soft-tissue tumors. Details of enrolled tumor histology are summarized in Table Wiley-Liss, Inc. 1199

2 1200 Maeda et al. Table 1 Myxoid and Nonmyxoid Soft-Tissue Tumors and Mean ADC Benign (N 18) ADC ( 10 3 mm 2 /second) a Malignant (N 26) ADC ( 10 3 mm 2 /second) a Myxoid (N 23) Myxoma (N 3) Myxoid liposarcoma (N 9) Schwannoma (N 2) Myxofibrosarcoma (N 3) Neurofibroma (N 1) 2.06 UHGPS (N 3) Spindle cell lipoma (N 1) 1.80 ES myxoid chondrosarcoma (N 1) 2.10 Nonmyxoid (N 21) Epidermal cyst (N 5) Lymphoma (N 2) Hemangioma (N 4) Synovial sarcoma (N 2) Pleomorphic lipoma (N 1) 1.48 ES Ewing (N 2) Desmoid (N 1) 1.42 WD liposarcoma (N 1) 0.71 MPNST (N 1) 0.60 UHGPS (N 1) 0.88 UHGPS with PI (N 1) 1.55 a Values are the mean ADC SD. UHGPS undifferentiated high grade pleomorphic sarcoma, ES extraskeletal, WD well-differentiated, MPNST malignant peripheral nerve sheath tumor, PI prominent inflammation. MRI Sequences We performed MRI studies using a 1.5-T MRI system (Signa; GE Medical Systems, Milwaukee, WI, USA). All images were obtained using a surface coil. The LSDWI studies of patients were conducted within the guidelines of the research committees of our institution. Informed consent was obtained from patients or their authorized representatives. The LSDWI method has been described previously (8 11). Neither cardiac gating nor respiratory triggering was employed in LSDWI; no antisusceptibility devices were used to reduce susceptibility artifacts. LSDWI images were acquired using the following scan parameters: repetition time (TR) msec, echo time (TE) msec, excitations 1, matrix size , and bandwidth khz. The effective section thickness and intersection gap were set from 3 mm/1 mm to 6 mm/2 mm. Two different b values were used, with the maximum b value applied along the three orthogonal directions: one with a low diffusion-weighting (b factor) of 5 second/mm 2 and the other with a high (maximum) b factor of 1000 second/ mm 2. The scan time per slice was seconds; a total of five slices were obtained. Other MRI sequences included T1-weighted imaging (TR/TE 500 msec/9 msec, matrix size , excitations 2), fast spin echo T2-weighted imaging with or without fat suppression (TR/TE 3000 msec/90 msec, matrix size , echo train length 8, excitations 2), and contrast-enhanced T1-weighted imaging with or without fat suppression. Imaging Data Analysis Isotropic diffusion-weighted images with a b factor of 1000 seconds/mm 2 were generated from the three diffusion directions assessed. Trace ADC maps were generated using the equation described by Stejskal and Tanner (12), S S 0 e badc, where b is the diffusion weighting factor, S is the signal intensity of the diffusion trace for b maximum, and S 0 is the signal intensity for b 5 seconds/mm 2. The ADC value measurements were obtained from the trace ADC maps using regions of interest (ROIs) placed over the tumors. In some patients, a water phantom at room temperature was also imaged. The ADC of the water was measured as a reference. For ROI measurement in the tumors, care was taken to include the solid-appearing portions of the tumors and to exclude obviously necrotic or cystic regions, as demonstrated in the corresponding T2- weighted and contrast-enhanced MR images. ADC values were expressed as mean SD. Statistical Analyses The Mann-Whitney U-test was used to detect significant differences in mean ADC values between myxoid and nonmyxoid soft-tissue tumors and between benign and malignant soft-tissue tumors. A P value less than 0.05 was considered to indicate a statistically significant difference. RESULTS ADC Values of Myxoid and Nonmyxoid Soft-Tissue Tumors LSDWI provided excellent diagnostic images and permitted ADC values to be measured without significant motion artifacts and susceptibility artifacts (Figs. 1 4). Details of mean ADC values of each soft tissue tumor are summarized in Table 1. The ADC (mean SD) was mm 2 /second in myxoid-type tumors, whereas the ADC was mm 2 / second in nonmyxoid-type tumors. A significant difference existed in ADC values between myxoid-type and nonmyxoid-type (P 0.01). However, considerable overlap was apparent between the two types of tumors (Fig. 5). The ADC was mm 2 /second in malignant soft-tissue tumors and mm 2 /second in benign soft-tissue tumors. Significant differences did not exist in the ADCs between benign and malignant soft-tissue tumors. The ADC values of

3 Soft-Tissue Tumors Evaluated by LSDWI 1201 Figure 1. A 63-year-old woman with intramuscular myxoma of the right thigh. a: Axial fat-suppressed T2- weighted image shows a homogeneous high signal mass in her right thigh (arrow). b: Axial fat-suppressed contrast-enhanced T1-weighted image shows a faint enhancement of the mass (arrow). c: Axial ADC map shows remarkably high signal intensity of the mass (long arrow). The signal intensity of the mass is almost equal to that of water phantom (small arrow). The respective mean ADC values of the mass and water phantom are mm 2 /second and mm 2 /second. d: Photomicrograph shows abundant myxoid matrix of the tumor. benign and malignant soft-tissue tumors overlapped greatly (Fig. 6). Among all tumors, the highest ADC value was apparent for benign intramuscular myxoma (mean mm 2 /second). The ADC value of intramuscular myxoma was almost equal to that of the water phantom (Fig. 1). Microscopic findings revealed abundant myxoid matrix in intramuscular myxoma (Fig. 1d). Among malignant tumors, extraskeletal myxoid chondrosarcoma (N 1), myxoid liposarcoma (N 4), and myxofibrosarcoma (N 1) exhibited ADC values greater than mm 2 /second. They all had a large quantity of myxoid matrix (Fig. 3). Among nonmyxoid benign tumors, epidermal cyst (N 4) showed ADC values lower than mm 2 /second. Several nonmyxoid malignant tumors showed ADC values lower than mm 2 /second, including extraskeletal Ewing sarcoma (N 2), malignant lymphoma (N 2), synovial sarcoma (N 2), undifferentiated high-grade pleomorphic sarcoma (N 1), well-differentiated liposarcoma (N 1), and malignant peripheral nerve sheath tumor (N 1). Pathological findings showed hypercellularity of those soft-tissue tumors (Figs. 2 and 4). DISCUSSION Because of its high intrinsic contrast resolution, it is expected that MRI has great potential for histologic classification of soft-tissue tumors. Several advanced MRI techniques have been investigated to characterize soft-tissue tumors. Some investigators have analyzed the value of dynamic MRI in grading soft-tissue tumors and have reported that most malignant soft-tissue tumors exhibited early and peripheral enhancement with a steep slope, an early maximum followed by a transition to a stable level, or a slight decrease of signal intensity (13). Static and dynamic contrast-enhanced MRI, when added to nonenhanced MRI, improved differentiation between benign and malignant soft-tissue lesions (14). Recently, van Rijswijk et al (5) characterized soft-tissue tumors using the DWI technique. They reported that true diffusion measurements allowed the differentiation of benign tumors from malignant tumors Figure 2. A 6-year-old boy with extraskeletal Ewing sarcoma of the left shoulder. a: Axial fat-suppressed T2-weighted image shows a high signal mass in his left shoulder (arrow). b: Axial contrast-enhanced T1- weighted image shows moderate enhancement of the tumor (arrow). c: Axial ADC map shows a homogeneously low signal intensity of the mass (arrow). The mean ADC value of the mass is mm 2 / second, whereas that of the water phantom (small arrow) is mm 2 /second. d: Photomicrograph shows hypercellularity of the tumor.

4 1202 Maeda et al. Figure 3. A 54-year-old woman with myxofibrosarcoma of the left thigh. a: Axial fat-suppressed T2-weighted image shows a high signal intensity mass in the left thigh (arrow). b: Axial fat-suppressed contrast-enhanced T1-weighted image shows strong but heterogeneous enhancement of the mass (arrow). Necrosis is apparent in the central area of the mass. c: Axial ADC map shows high signal intensity of the mass (arrow). The mean ADC value of the mass is mm 2 / second. d: Photomicrograph shows many myxoid areas with scattered pleomorphic cells. by showing lower true diffusion coefficients for malignant tumors than for benign tumors. That result suggests that diffusion can provide information that allows differentiation between benign and malignant soft-tissue tumors. Inverse correlation between tumor cellularity and ADC values has been reported in brain tumors (1,3,4). Sugahara et al (1) reported that the minimum ADC value showed an inverse correlation with histologic cellularity in gliomas. Guo et al (4) also reported a clear inverse relationship between the ADC value and cellularity of brain tumors, such as lymphomas and highgrade astrocytomas. Their findings suggest that greater cellularity is associated with more restricted diffusion. However, the ADC values might be affected not only by cellularity, but also by the nature of the extracellular matrix. Soft-tissue tumors might have more various extracellular substances than brain tumors. Particularly, myxoid matrix is peculiar in that free water is abundant in extracellular spaces; furthermore, myxoid tissue forms parts of many benign and malignant softtissue tumors (15,16). Our results show that myxoid-containing soft-tissue tumors have significantly higher ADC values than nonmyxoid soft-tissue tumors do. For that reason, we infer that diffusion of soft-tissue tumors is significantly affected by the presence of myxoid matrix. Among myxoid soft-tissue tumors, intramuscular myxoma showed the highest ADC values; they were almost equal to those of the water phantom. These high values directly reflect of the high mucin and low collagen content in the lesion, representing a lesion composed of a large amount of water, as seen histologically. The typical MRI features of intramuscular myxoma consist of a sharp border, and very low signal on the T1-weighted image and a very high signal on the T2-weighted image (7). Consequently, the noncontrast MRI characteristics of this tumor mimic a cyst. However, contrast-enhanced MRI images more accurately reflect the truly solid consistency of the myxoma because it shows internal enhancement. Based on our preliminary results, ADC values were incapable of differentiating benign soft-tissue tumors from malignant soft-tissue tumors; the ADC values of Figure 4. A 75-year-old man with undifferentiated high grade pleomorphic sarcoma of the right thigh. a: Axial fat-suppressed T2-weighted image shows a high signal intensity mass in the right thigh (arrow). b: Axial fatsuppressed contrast-enhanced T1-weighted image shows strong and heterogeneous enhancement of the mass (arrow). c: Axial ADC map shows low signal intensity of the mass (arrow). The ADC value of the mass is mm 2 /second. d: Photomicrograph shows hypercellularity of the tumor.

5 Soft-Tissue Tumors Evaluated by LSDWI 1203 Figure 5. Scatterplot of the ADC values of myxoid (N 23) and non-myxoid (N 21) soft tissue tumors. A significant difference of the ADC values exists between myxoid and nonmyxoid tumors (P 0.01). the benign and malignant soft-tissue tumors overlapped greatly. More recently, Einarsdottir et al (6) used multishot echo-planar DWI sequence for evaluation of soft-tissue tumors. They reported that the ADC values of both soft-tissue tumors overlapped and were not useful to differentiate between the bulk of benign and malignant tumors. Therefore, our results accorded with those of Einarsdottir et al (6). It appears reasonable that ADC values of benign and malignant tumors show substantial overlap because both tumors occasionally contain myxoid matrix. Our results showed that the ADC values were lower in several nonmyxoid malignant tumors such as lymphoma, Ewing sarcoma, and synovial sarcoma. Pathologic findings of such tumors showed hypercellularity of tumors (Figs. 2 and 4). On the other hand, among benign tumors, epidermal cyst showed the lowest ADC value. Epidermal cyst is a benign solid lesion of the cutis and subcutis; such cysts are filled with looselypacked lamellae of keratinous material without cellular components (17,18). However, the ADC values of epidermal cysts were as low as nonmyxoid malignant tumors with hypercellularity. The ADC values of softtissue tumors, therefore, are influenced also by factors other than cellularity. Our preliminary results showed that LSDWI is feasible for the evaluation of soft-tissue tumors. They provided diagnostic images for soft-tissue tumors in the body and extremities without the need for specialized hardware, cardiac gating, or respiratory compensation. In LSDWI, multiple diffusion-weighted spin echo column excitations are used to obtain a two-dimensional image (19). This method is relatively insensitive to motion artifacts because the images are constructed column by column and the acquisition time for an individual column is approximately equal to the echo time. Importantly, in areas near large bone structures or air, the sensitivity of single-shot echo-planar imaging to susceptibility variations can distort images or cause complete signal loss, whereas images obtained with LS- DWI display no such artifacts (8 11,19). For those reasons, it is important for DWI studies of soft-tissue tumors to be insensitive to susceptibility artifacts. Previous reports also resolved the issue of susceptibility related artifacts using spin-echo sequence (5) or multishot spin-echo echo-planar sequence (6). However, the problem of those reports was that the maximum b values used were small, such as those of 701 second/mm 2 and 600 second/mm 2 (5,6). Such maximum b values appear to be too small to evaluate tumor diffusion. In brain tumors, DWI has been performed practically using maximum b values of 1000 seconds/mm 2 or greater (1 4). As for the DWI studies of brain tumors, we presume that the maximum b value of 1000 seconds/mm 2 is appropriate for quantitative evaluation of diffusion of soft-tissue tumors. Van Rijswijk et al (5) used perfusion-corrected DWI for characterizing soft-tissue tumors. They showed a statistically significant difference between true diffusion coefficient values and ADC values in cases with soft-tissue tumors and concluded that the true diffusion measurements might be useful to characterize soft-tissue tumors. However, it remains uncertain whether true diffusion measurements were superior to ADC measurements for soft-tissue tumors. Regarding brain tumors, no reports have described true diffusion measurements, but several reports have shown the usefulness of ADC measurements for tumor differentiation and characterization (1 4). One limitation of this study is that we might have inadvertently included tiny cysts or small necroses in ROI measurements of tumors, even though we carefully avoided apparent cysts or necrosis referred to T2- weighted and contrast-enhanced MR images. Lang et al (20) and Herneth et al (21), using animal tumor models, reported that tumor necroses can cause increased ADC values. However, the relationship between tumor necroses and ADC values is based not only on the amount but also on the size of the necrotic area. Lyng et al (22) reported after an experimental study that the fraction of a massive necrosis can be correlated with ADC values, but not the fraction of small foci of necrosis, because the small foci of necrosis might be smaller than the voxel size of the MR images. Further investigation is Figure 6. Scatterplot of the ADC values of benign (N 18) and malignant (N 26) soft tissue tumors. No significant difference of the ADC values is apparent between the two types of soft tissue tumors.

6 1204 Maeda et al. necessary to determine other possible factors that might considerably influence the ADC values of softtissue tumors. This study carries another limitation concerning tumor types and number of cases. We excluded many frequently found lesions such as ganglionic cyst, synovial cyst, and typical lipoma. In addition, myxoid liposarcoma (N 9) was considerably a larger number compared to other tumor types, which might be a true limitation of the current study. For that reason, further studies should be performed of more numerous cases and tumor types to verify the results of our preliminary study. In conclusion, ADC values were significantly higher in myxoid-containing soft-tissue tumors than those in nonmyxoid soft-tissue tumors. Therefore, it is suggested that myxoid matrix influences ADC values of soft-tissue tumors. The ADC values overlap greatly between benign and malignant soft-tissue tumors and might not be useful for their differentiation. REFERENCES 1. Sugahara T, Korogi Y, Kochi M, et al. Usefulness of diffusionweighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 1999;9: Filippi CG, Edgar MA, Ulug AM, Prowda JC, Heier LA, Zimmerman RD. Appearance of meningiomas on diffusion-weighted images: correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 2001;22: Kono K, Inoue Y, Nakayama K, et al. The role of diffusion-weighted imaging in patients with brain tumors. AJNR Am J Neuroradiol 2001;22: Guo AC, Cummings TJ, Dash RC, Provenzale JM. Lymphomas and high-grade astrocytomas: comparison of water diffusibility and histologic characteristics. Radiology 2002;224: van Rijswijk CSP, Kunz P, Hogendoorn PCW, Taminiau AHM, Doornbos J, Bloem JL. Diffusion-weighted MRI in the characterization of soft-tissue tumors. J Magn Reson Imaging 2002;15: Einarsdottir H, Karlsson M, Wejde J, Bauer HCF. Diffusionweighted MRI of soft tissue tumours. Eur Radiol 2004;14: Peterson KK, Renfrew DL, Feddersen RM, Buckwalter JA, El- Khoury GY. Magnetic resonance imaging of myxoid containing tumors. Skeletal Radiol 1991;20: Maier SE, Gudbjartsson H, Patz S, et al. Line scan diffusion imaging: characterization in healthy subjects and stroke patients. AJR Am J Roentgenol 1998;171: Maeda M, Sakuma H, Maier SE, Takeda K. Quantitative assessment of diffusion abnormalities in benign and malignant vertebral compression fractures by line scan diffusion-weighted imaging. AJR Am J Roentgenol 2003;181: Nagata M, Maeda M, Tsukahara H, Maier SE, Takeda K. Brain stem hypertensive encephalopathy evaluated by line scan diffusionweighted imaging. AJNR Am J Neuroradiol 2004;25: Maeda M, Kato H, Sakuma H, Maier SE, Takeda K. Usefulness of the apparent diffusion coefficient in line scan diffusion-weighted imaging for distinguishing between squamous cell carcinomas and malignant lymphomas of the head and neck. AJNR Am J Neuroradiol 2005;26: Stejskal EO, Tanner JE. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Phys 1965;42: van der Woude HJ, Verstraete KL, Hogendoorn PC, Taminiau AH, Hermans J, Bloem JL. Musculoskeletal tumors: does fast dynamic contrast-enhanced subtraction MR imaging contribute to the characterization? Radiology 1998;208: van Rijswijk CSP, Geirnaerdt MJA, Hogendoorn PCW, et al. Softtissue tumors: value of static and dynamic gadopentetate dimeglumine-enhanced MR imaging in prediction of malignancy. Radiology 2004;233: Enzinger FM. Intramuscular myxoma: a review and follow-up study of 34 cases. Am J Clin Pathol 1965;43: Mackenzie DH. The myxoid tumors of somatic soft tissues. Am J Surg Pathol 1981;5: Misner SC, Mariash SA, Alvarez G. Ruptured plantar epidermal inclusion cyst with foreign body giant cell reaction. J Foot Surg 1991;30: Ma LD, McCarthy EF, Bluemke DA, Frassica FJ. Differentiation of benign from malignant musculoskeletal lesions using MR imaging: pitfalls in MR evaluation of lesions with a cystic appearance. AJR Am J Roentgenol 1998;170: Gudbjartsson H, Maier SE, Mulkern RV, et al. Line scan diffusion imaging. Magn Reson Med 1996;36: Lang P, Wendland MF, Saeed M, et al. Osteogenic sarcoma: noninvasive in vivo assessment of tumor necrosis with diffusionweighted MR imaging. Radiology 1998;206: Herneth AM, Guccione S, Bednarski M. Apparent diffusion coefficient: a quantitative parameter for in vivo tumor characterization. Eur J Radiol 2003;45: Lyng H, Haraldseth O, Rofstad EK. Measurement of cell density and necrotic fraction in human melanoma xenografts by diffusion weighted magnetic resonance imaging. Magn Reson Med 2000;43: