Magnetic resonance imaging of malignant fibrous histiocytoma

Magnetic Resonance Imaging, Vol. 7, pp. 283-288, Printed in the USA. All rights reserved.

1989 Copyright

0730-725X/89 $3.00 + .oO 0 1989 Pergamon Press plc

l Original Contribution MAGNETIC RESONANCE IMAGING OF MALIGNANT FIBROUS HISTIOCYTOMA HARSH MAHAJAN, * E. EDMUND KIM, * SIDNEY WALLACE, * RICHARD ABELLO, * ROBERT BENJAMIN,? AND HARRY L. EVANS$ Departments of *Diagnostic Radiology, tMedica1 Oncology and SPathology, The University of Texas M. D. Anderson

Cancer Center, Houston,

Texas 77030, USA

The magnetic resonance imaging (MRI) changes in 39 patients with malignant fibrous histiocytoma (MFH) were reviewed retrospectively. Twenty-one sarcomas were in the lower extremity, five each in the upper extremity and trunk, two each in the neck and heart, and one each in the maxillary sinus, sella turcica, tongue, and spermatic cord. The examinations were performed with spin-echo sequences on a 1.5 Tesla Signa Scanner (GE, Milwaukee WI). Twenty-two tumors exhibited intermediate signal intensity on T,-weighted images and 23 were of high signal intensity on T,-weighted images. There was no significant difference in signal intensity of 12 preoperative and 13 recurrent neoplasms. Twelve of 13 patients were correctly diagnosed as having postoperative changes. The MR sensitivity and specificity for detecting a neoplasm were 96% and 83% respectively, but the signal changes were nonspecific for MFH. When compared to CT in 14 patients, MR better defined the extent of the MFH, its relationship to surrounding tissues and vessels, and best differentiated residual or recurrent disease from postoperative changes when examined at least 3 months after surgery. Keywords: Magnetic resonance imaging; Malignant fibrous histiocytoma.

INTRODUCTION

was found in the lower extremity in 21 patients; upper extremity, 5; trunk, 5; neck, 2; heart, 2; and one patient each in the maxillary sinus, sella turcica, tongue, and spermatic cord. MRI was performed on a 1.5 Tesla Signa Scanner (General Electric, Milwaukee WI). Spin-echo pulse sequences were obtained using TR of 600-800 msec and TE of 20-30 msec for T,-weighted images, and a TR of 2000 msec and TE of 20 msec and 80 msec for proton density and &weighted images. Contiguous slices, 5 or 10 mm in thickness, were obtained in the axial, coronal, and sagittal planes. The 39 patients underwent 67 MRI examinations; 18 patients, including 13 with no tumor on the initial MRI, were examined more than once with a maximum of five times in one patient. Twelve patients had an MRI at the time of the initiaI diagnosis, while the other 27 were examined 15 days to 10 years, usually 3-6 months, after surgical resection to define residual or recurrent disease. MRI scans were obtained to search for changes in size of the tumor and character of the signals in seven patients who received chemo-

In 1964, O’Brien and Stout” described malignant fibrous histiocytoma (MFH) as a neoplasm characterized by a proliferation of pleomorphic sarcomatous elements including histiocytes, fibroblasts, giant cells and xanthomatous cells arranged in a storiform pattern.6~‘6~‘8MFH is the commonest soft tissue sarcoma.8 Typically these tumors arise in late adult life in the extremities and are located adjacent to deep fascia or within skeletal muscle.7~8~12Less than 10% are found in the subcutaneous tissues.6 MFH also originates in bone, at times in a preexisting bone infarct.4 The magnetic resonance imaging (MRI) findings are described in 39 patients with biopsy-proved MFH at the time of the initial diagnosis, after surgery to identify residual or recurrent disease and after chemotherapy. MATERIALS

AND METHODS

Of the 39 patients, there were 22 males and 17 females, ages 18 to 74 years (mean: 46 years). MFH RECENED I l/6/88; ACCEPTED12/8/88. Address reprint requests to E. Edmund Kim, M.D., Division of Diagnostic Imaging, Box 057, The University of

Texas M. D. Anderson Cancer Center, Blvd., Houston, TX 77030, USA. 283

1515 Holcombe

284

Magnetic Resonance Imaging 0 Volume 7, Number 3, 1989

therapy, two systemic and five intra-arterial. Fourteen patients had computed tomography (CT) and seven had angiography within one month of the comparable MRI. RESULTS

Of the 39 patients studied, 31 were subjected to surgery to confirm the MRI findings, while 8 patients were followed for a minimum of 6 months, clinically, and by repeated MRI examinations with no evidence of recurrence. Twenty-two (88%) of the 25 biopsy-proved MFHs were masses of intermediate signal intensity (equal to the surrounding muscles) on T, -weighted images and 23 (92%) were predominantly of high signal intensity (equal to or higher than subcutaneous fat) on T2weighted images (Fig. 1). The tumors were heterogeneous on TZ-weighted images with irregular areas of extremely high or low signals (Fig. 2). Calcification within the neoplasm in 2 patients exhibited areas of decreased signal intensity on T,- and T,-weighted images. There was no significant difference in signal intensity of the 12 preoperative and the 13 residual or recurrent tumors. Postoperative changes were depicted by MRI as areas of decreased or low signal intensity on T2weighted images (Fig. 3). Persistent increased signal intensity on Tz-weighted images suggests residual or recurrent tumor (Fig. 1). Twenty-seven patients were studied after definitive surgery had been performed. In 14 patients MRI identified local recurrence. By the histiopathological evaluation after resection definite recurrence of MFH was confirmed in 13 patients while only cicatricial tissue was found in one. The other 13 patients were diagnosed by MRI as postoperative changes with no evidence of residual or recurrent disease on the initial and follow-up examinations, usually at 3-6 month intervals. In twelve of these 13 patients, the presence of MFH was confirmed but one resected specimen disclosed microscopic foci of residual tumor. Of the 7 patients in whom tumors were treated by chemotherapy and followed by MRI examinations, reduction in size was noted in 4 (Fig. 4). In two of these MFHs postchemotherapy MRI showed a decrease in signal intensity on T,-weighted images with no change on T, (Fig. 5). The third case demonstrated a lesion which appeared better marginated when compared to the prechemotherapy scan. The fourth patient had no significant alteration in the signal intensity. All four patients experienced clinical improvement. In the other 3 patients, the disease progressed despite chemotherapy with no change in the

Fig. 1. Recurrent MFH (arrows). (A) A rounded mass, 2 cm in maximum diameter, in the subcutaneous tissue of the right thigh with intermediate signal intensity on T,-weighted coronal image and (B) high signal intensity on Tz-weighted axial images.

signal character of administration, i.e., no bearing on signal at times in patients

the lesions. The route of drug intravenous or intra-arterial had changes in the tumor. However, treated by intra-arterial chemo-

MRI

of malignant fibrous histiocytoma 0 H. MAHAJAN_m AL.

A

285

B

Fig. 2. Paraspinal MFH (arrows) with involvement of the adjacent vertebrae. (A) A large left paraspinal mass involving adjacent T12 vertebral body with heterogeneous intermediate signal intensity on T,-weighted and (B) high signal intensity on

Tz-weighted axial images.

A

B

Fig. 3. Post-operative changes (arrows). (A) Comparable &-weighted axial images with 1l-week interval show a decreased curvi-linear signal intensity along the anteromedial subcutaneous tissue and muscle of left thigh on the follow-up image (B). (A) was obtained one month following the surgery.

therapy, an increase in signal intensity on T,-weighted images was seen in otherwise normal muscles. This probably resulted from a localized increased concentration of the chemotherapeutic agents due to streaming

or laminar flow directed into small muscular branches as the result of the extremely slow rate of the infusion. The sensitivity and specificity of MRI for detecting a neoplasm were 25/26 (96%) and 5/6 (83%) respec-

Magnetic Resonance Imaging 0 Volume 7, Number 3, 1989

286

A

B

Fig. 4. Post-chemotherapy response of MFH. (A) Comparable T2-weighted axial images with 13-week interval show marked reduction of MFH (arrows) in size along the lateral subcutaneous tissue of the right lower leg on the follow-up image (B). (A) was obtained prior to any therapy, and the diagnosis was established by the percutaneous biopsy.

tively, but MRI was not specific for characterizing the

malignancy as a malignant fibrous histiocytoma. DISCUSSION

MR imaging is being increasingly used in the radiological evaluation of musculoskeletal neoplasms. Many reports have demonstrated the superiority of MR over CT in displaying soft tissue tumors including patients with MFH. The specific signal characteristics of this neoplasm have not been discussed.3*13,14The reason for not approaching this issue is that these studies concentrated on comparison of MR and CT to determine which was the superior modality for demonstrating the extent of disease. We have attempted to (a) characterize the MR signal changes in biopsy-proven MFH, (b) determine tumor response to chemotherapy, and (c) evaluate MR patterns in the postoperative region. Eighty-eight per cent of the tumors showed primarily intermediate signal intensity (equal to surrounding muscles) on T,-weighted images and 92% showed predominantly high signal intensity (equal to or higher than subcutaneous fat) on Tz-weighted images. The tumors were heterogeneous on T,-weighted images representing areas of hemorrhage and necrosis. Locally recurrent tumors showed basically the same signal character as tumors imaged at the time of initial diagnosis. Deep-seated tumors usually have a worse prognosis, but there was no difference in MR signal

and deep tumors. The signal characteristics were nonspecific for MFH, as most malignant tumors, except those which contain fat, atypical lipomatous tumor, or well-differentiated liposarcoma, i.e., bright T, and T,. Though MR imaging was unable to give the specific diagnosis of MFH, it played a very important role in defining the extent of disease. Two out of the four patients who showed decrease in size of tumor following chemotherapy, demonstrated lower signal intensity on Tz-weighted images. Changes in MR signal have also been noted by Aisen et al.’ They also demonstrated the development of a sharp boundary between the lesion and normal tissue, as was noted in one case in our series. Hence, MR may also prove to be useful in gauging tumor response to therapy. MR proved very useful in patients following surgery. Twelve of 13 patients (92%) were correctly diagnosed as having only postoperative changes with 6 biopsy and 7 clinical correlations. In the one instance where MR was incorrect, microscopic foci of tumor were demonstrated in the scar tissue. Vane1 et al.” have previously utilized MR imaging of musculoskeleta1 tumors after surgery. A lesion of high signal intensity on T,-weighted images denoted active tumor, while the lesion of low signal intensity was not associated with active disease. However, the time interval after surgery was not mentioned, which has a great bearing on the signal changes. Most of our patients underwent MR imaging within one month of surgery, between superficial

MRI of malignant fibrous histiocytoma 0 H.

MAIUJAN ET AL.

287

Fig. 5. Post-chemotherapy response of bone MFH. Comparable (A and B) Tr-weighted sagittal and (C and D) T2-weighted axial images with U-week interval show marked decrease in heterogeneous high signal intensity of MFH (asterisk) involving distal femur on (D) T,-weighted image. (A) and (C) were obtained prior to any therapy, and the diagnosis was made by the percutaneous biopsy.

and were then followed by serial scans. The early postoperative (within 3-6 months) MR changes ineluded irregularity of the fascial and muscle planes, and the presence of edema and inflammation as evi-

dented by low signal intensity on Tr- and high signal intensity on T2-weighted images. During this time interval it is very difficult to differentiate residual tumor from postoperative changes. By 3-6 months

288

Magnetic Resonance Imaging 0 Volume 7, Number 3, 1989

after surgery, most of the edema subsides and residual tumor may be identified as a high intensity lesion on T,-weighted images, especially if it exerts a mass effect on the surrounding tissues. On further followup of these postoperative changes there is usually some local fatty replacement and mild atrophy of the adjacent muscles. MFH has an incidence of recurrence as high as 51% with radical complete excision,’ and 100% with local excision.” MR correctly identified recurrence in 13 of 14 cases (93%). Since a high percentage of MFH can recur, it is extremely important to monitor the lesions after surgery by MR imaging. A baseline MR examination should be done approximately two months after surgery, when most of the postoperative changes have subsided. This would give a good estimate of the anatomy of the region, as altered by surgery. The patients can then be followed by regular MR examinations at three month intervals for comparison for the detection of local tumor recurrence. In the early cases CT scanning was also performed along with MR, and 14 of our patients had comparable CT scans. But as MR, with its ability to display direct coronal and sagittal scans with absence of bone induced artifacts, proved superior to CT for assessment of the extent of tumor and visualization of surrounding vascular structures, most of the later patients underwent only MR examination. However, it was noted that a primary MFH of bone and calcification within a soft tissue MFH were better evaluated by CT. Hence, we now perform CT along with MR in primary MFH of bone and where there is suspected calcification within a tumor. Angiography revealed hypervascular tumors in five and hypovascular tumors in two cases. Similar features have been previously described2~5~‘5and are probably due to the wide spectrum of morphologic patterns of this tumor. There was no significant correlation between vascularity and heterogenous MRI findings in this series.

malignant

1. Aisen, A.M.; Mattel, W.; Braunstein, E.M.; McMillin, K.I.; Phillips, W.A.; Kling, T.F. MRI and CT evaluation of primary bone and soft tissue tumors. AJR 146: 749-756;

1986.

2. Burgener, F.A.; Landman, S. Angiographic features of

Radiology

121:581-

3. Demas, B.E.; Heelan,

4.

5.

6.

7.

8.

9.

10. 11. 12.

13.

R.T.; Lane, J.; Marcove, R.; Hajdu, S.; Brennan, M.F. Soft tissue sarcomas of the extremities: Comparison of MR and CT in determining the extent of disease. AJR 150:615-620; 1988. Enzinger, F.M. Recent developments in the classification of soft tissue sarcomas. In: Management of Primary Bone and Soft Tissue Sarcomas. Chicago: Year Book Medical Publishers; 1977. Fischer, H-J.; Lois, J.F.; Comes, A.S.; Mirra, J.M.; Deutsch, L.S. Radiology and pathology of malignant fibrous histiocytomas of the soft tissues: a report of ten cases. Skeletal Radiol. 13:202-206; 1985. Fu, Y .S. ; Gabbiani, G.; Kaye, G.I.; Lattes, R. Malignant soft tissue tumors of probably histiocytic origin (malignant fibrous histiocytomas). General considerations and electron microscopic and tissue culture studies. Cancer 35:176-198; 1975. Kauffman, S.L.; Stout, A.P. Histiocytic tumors (fibrous xanthoma and histiocytoma) in children. Cancer 14: 469-482; 1961. Kearney, M.M.; Soule, E.H.; Irvins, J.C. Malignant fibrous histiocytoma. A retrospective study of 167 cases. Cancer 45:167-178; 1980. Kempson, R.L.; Kyriakos, M. Fibroxanthosarcoma of the soft tissues. A type of malignant fibrous histiocytoma. Cancer 29:961-976; 1972. Lattes, R. Malignant fibrous histiocytoma. A review article. Am. J. Surg. Pathol. 6:761-772; 1982. O’Brien, J.E.; Stout, A.P. Malignant fibrous xanthomas. Cancer 17:1445-1458; 1964. Ozzello, L.; Stout, A.P.; Murray, M.R. Cultural characteristics of malignant histiocytomas and fibrous xanthomas. Cancer 16:331-344; 1963. Petasnick, J.P.; Turner, D.A.; Charters, J.R.; Gitelis, S.; Zacharias, C.E. Soft tissue masses of the locomotor system: comparison of MR imaging with CT. Radiology 160:125-133;

1986.

14. Pettersson, H.; Gillespy, T.; Hamlin, D.J., et al. Primary musculoskeletal tumors: Examination with MR imaging compared with conventional modalities. Rudiology 1987; 164:237-241.

15. Ros, P.R.; Viamonte, M.; Rywlin, A.M. Review. Malignant fibrous histiocytoma: mesenchymal tumor of ubiquitous origin. AJR 142:753-759; 1984. 16. Taxy, J.; Battifora, H. Malignant fibrous histiocytoma: A clinicopathologic and ultrastructural study. Cancer 40:254-267;

REFERENCES

fibrous histiocytomas.

583; 1976.

1977.

17. Vanel, D.; Lacombe, M.J.; Spielmann, M.; Genin, J. follow-up with MR imaging gery and radiation therapy.

Couanet, D.; Kalifa, C.; Musculoskeletal tumors: after treatment with surRadiology

164:243-245;

1987.

18. Weiss, SW.; Enzinger, F.M. Malignant fibrous histiocytoma. An analysis of 200 cases. Cancer 41:2250-2266; 1978.