Intramuscular Myxoma: Characteristic MR Imaging Features

Size: px

Start display at page:

Download "Intramuscular Myxoma: Characteristic MR Imaging Features"

Aldous Hensley
6 years ago
Views:

1 Laura W. ancroft 1 Mark J. Kransdorf 1 David M. Menke 2 Mary I. O Connor 3 William C. Foster 4 Received September 17, 2001; accepted after revision October 30, Department of Radiology, Mayo Clinic, 4500 San Pablo Rd., Jacksonville, FL ddress correspondence to M. J. Kransdorf. 2 Department of Pathology, Mayo Clinic, Jacksonville, FL Department of Orthopedic Surgery, Mayo Clinic, Jacksonville, FL Department of Orthopedic Surgery, Medical College of Virginia, Virginia Commonwealth University, 1200 E. road St., P. O. ox , Richmond, V JR 2002;178: X/02/ merican Roentgen Ray Society Intramuscular Myxoma: Characteristic MR Imaging Features OJECTIVE. The purpose of our study was to identify the characteristic MR imaging features of intramuscular myxoma. MTERILS ND MTERILS. We retrospectively reviewed the MR imaging features of 20 patients with intramuscular myxoma. Clinical assessment included the age and sex of the patient and location of the tumor. Radiologic evaluation included the lesion size and shape, border definition, signal on T1- and T2-weighted or fluid-sensitive MR sequences, enhancement pattern, presence or absence of a fat rind, and presence or absence of increased signal in the adjacent muscle on T2-weighted or fluid-sensitive MR sequences. RESULTS. The mean age of patients presenting with intramuscular myxoma was 61 years (range, years; median, 64 years). The mean lesion size was 6.9 cm (range, 3 17 cm; median, 6.3 cm). peritumoral fat rind was present in 13 of the patients (65%) with myxoma, and an increased signal in the adjacent muscle on fluid-sensitive sequences was present in 11 patients (55%). Intramuscular myxomas were homogeneously low in signal intensity on T1-weighted MR sequences in 19 patients (95%), with all lesions showing a high signal intensity on T2- weighted or fluid-sensitive MR sequences. Twelve of the myxomas had well-defined borders, and eight had borders that were partially ill defined. Of the 11 lesions imaged after gadolinium administration, six (55%) showed intense heterogeneous enhancement. CONCLUSION. Findings of a mass that on MR images shows a perilesional fat rind, the signal intensity of fluid, and an increased signal in the adjacent muscle on T2-weighted or fluid-sensitive MR sequences are strongly suggestive of intramuscular myxoma. The degree of lesion enhancement varies but is most frequently intense and heterogeneous. lthough the recognition of these features likely will not obviate biopsy of any individual lesion, it will allow more accurate prebiopsy diagnosis and preoperative planning. T he intramuscular myxoma is a frequently imaged member of a family of lesions characterized by an abundant myxoid matrix and a paucity of spindle-shaped stromal cells. The myxoma was established as a distinct lesion by Stout [1] in a 1948 comprehensive report of 49 cases drawn from the Laboratory of Surgical Pathology of Columbia University and 93 additional cases drawn from the literature (exclusive of those lesions within the heart). lthough Stout in the original study found only 3% of the lesions to be intramuscular, Enzinger [2] in 1965 reported that 17% of approximately 200 myxomas were intramuscular in origin. The intramuscular myxoma is of particular interest to radiologists because it may have imaging features similar to other myxoid lesions, especially the myxoid liposarcoma. If a myxoid liposarcoma is in an intramuscular location and has a predominantly myxoid morphology (the so-called cystic appearance), the two lesions may be strikingly similar in appearance. The similarity of these lesions extends to the gross and histologic appearances as well, and differentiating between the two may be difficult [3]. Making this differentiation has major implications because intramuscular myxoma has a benign clinical course, with no tendency to recur or metastasize [4, 5]. We retrospectively reviewed our experience with intramuscular myxoma to identify the MR imaging characteristics of this lesion. Materials and Methods We reviewed the MR images of 20 patients with intramuscular myxoma. Most of the patients JR:178, May

records of consultation cases. ll cases of intramuscular myxoma were surgically confirmed. The MR imaging planes, sequences, and field strengths varied because of the nature of the study groups.

Eleven of the 20 patients with myxoma also had IV gadolinium-enhanced images. Clinical assessment included the age and sex of each patient and the location of the tumor.

defined). Lesions were assessed on T1-weighted MR images relative to skeletal muscle as homogeneously low in signal intensity or low in signal intensity with areas of intermediate signal.

2 ancroft et al. D in our study sample were selected from the archives of our institution chosen from a review of computer-generated patient diagnoses made from January 1996 through December 2000 as well as from records of consultation cases. ll cases of intramuscular myxoma were surgically confirmed. The MR imaging planes, sequences, and field strengths varied because of the nature of the study groups. ll patients included in the study had at least one T1-weighted MR sequence and either a T2-weighted or fluid-sensitive MR sequence. Eleven of the 20 patients with myxoma also had IV gadolinium-enhanced images. Clinical assessment included the age and sex of each patient and the location of the tumor. The radiologic evaluation included lesion size in centimeters (largest diameter), shape (spherical, ovoid, or lobulated), and border definition (well defined, partially ill defined, or wholly ill defined). Lesions were assessed on T1-weighted MR images relative to skeletal muscle as homogeneously low in signal intensity or low in signal intensity with areas of intermediate signal. On T2-weighted MR images, a signal intensity similar to that of fluid was designated as high, whereas a signal intensity similar to that of fat was considered intermediate. Signal intensity on fat-suppressed T2-weighted turbo spin-echo and short tau inversion recovery images was more difficult to grade, but lesions were considered high in signal intensity if their signal intensity was similar to that of fluid. The appearance of lesions was classified as homogeneous with high signal intensity, homogeneous with high signal intensity and low- or intermediate-signal septal components, homogeneous with high signal intensity and low- or intermediate-signal nodular components, or heterogeneous with both septal and nodular components. The presence or absence of gadolinium enhancement was noted. If such enhancement was present, it was characterized as homogeneous or heterogeneous. The margin of the lesion was assessed for the presence or absence of a rind of fat at the interface between the lesion and muscle on T1-weighted MR images, and the presence or absence of increased signal in the skeletal muscle adjacent to the lesion on T2-weighted or fluid-sensitive MR sequences. In addition, we compared the reliability of the last two signs as tools in successfully distinguishing between an intramuscular myxoma and an intramuscular myxoid liposarcoma, the lesion with which a myxoma is most likely to be confused. To assess this reliability, we compared findings of the study group with those of a control group of 12 patients with intramuscular myxoid liposarcomas using the Wilcoxon s rank sum test for continuous variables and the chi-square test for categorized variables. Logistic regression was used to test for the influence of the study variables after adjusting for age and sex differences between the groups. The final multivariate logistic model was built using a stepwise selection method. We used a software package (SS Institute, Cary, NC) for data analysis. p value of 0.05 was considered statistically significant. The institutional review board approved our study. Results Most of the 20 lesions were located in the thigh (n = 12). The remaining eight lesions occurred in the pelvis (n = 3); upper arm (n = 2); and shoulder, popliteal fossa, and elbow (n = 1 each). ll patients had been referred for the evaluation of a soft-tissue mass. The mean patient age was 61 years (range, years; median, 64 years). Patients included 10 males and 10 females. Lesions varied in size from 3 to 17 cm (mean, 6.9 cm; median, 6.3 cm). Twelve of the myxomas had well-defined borders (Fig. 1), and eight had partially illdefined borders (Fig. 2). We found that imaging planes along the longitudinal axis of the involved muscle yielded the most information regarding lesion borders and conspicuity of signal differences between the lesions and the adjacent muscle. Lesions were most commonly ovoid (13/20; 65%) or less commonly lobulated (6/20; 30%); one lesion was spherical (5%). Fig. 1. Intramuscular myxoma in adductor compartment of left thigh in 69-year-old man., Coronal T1-weighted spin-echo MR image (TR/TE, 400/20) shows rind of adipose tissue with varying thickness surrounding ovoid, homogeneously hypointense lesion (asterisk). and C, Comparison of unenhanced () and enhanced (C) corresponding axial T1-weighted spin-echo MR images (400/20) shows prominent heterogeneous enhancement. Rind of adipose tissue is seen to better advantage on long-axis coronal image () rather than on axial image (). D, xial T2-weighted MR image (1800/80) obtained at same level as shows slightly increased signal in muscle adjacent to mass (arrow). C 1256 JR:178, May 2002

MR Imaging of Intramuscular Myxoma ll lesions showed a fluidlike signal intensity. Most of the lesions (19/20; 95%) were of homogeneously low signal intensity on T1-weighted MR sequences (Fig. 1).

The myxomas, similar to other myxoid lesions, all showed high signal on T2-weighted or fluid-sensitive MR sequences.

Four lesions (20%) were of homogeneously high signal intensity, and three (15%) showed a nodular septal pattern (Fig. 3). lthough enhancement patterns varied, heterogeneous enhancement (Fig.

In two patients, the fat rind was so extensive that it simulated intralesional fat on MR imaging in at least one plane.

Discussion The typical intramuscular myxoma is a well-defined ovoid lesion with fluidlike signal intensity, a peritumoral fat rind visible on T1weighted MR images, and an increased signal in the

These latter two features have not been previously emphasized and are the most reliable radiologic features for differentiating intramuscular myxoma from other myxoid soft-tissue lesions.

3 MR Imaging of Intramuscular Myxoma ll lesions showed a fluidlike signal intensity. Most of the lesions (19/20; 95%) were of homogeneously low signal intensity on T1-weighted MR sequences (Fig. 1). One lesion (5%) was of heterogeneously low signal intensity with poorly defined areas of intermediate signal intensity. The myxomas, similar to other myxoid lesions, all showed high signal on T2-weighted or fluid-sensitive MR sequences. Most lesions (13/20; 65%) showed associated peripheral linear regions of intermediate signal intensity (designated as a septal pattern). Four lesions (20%) were of homogeneously high signal intensity, and three (15%) showed a nodular septal pattern (Fig. 3). lthough enhancement patterns varied, heterogeneous enhancement (Fig. 1) was seen in 55% (6/11) of the lesions in patients with myxomas who received gadolinium contrast material. peritumoral fat rind was found in 13 of the patients with myxoma (Figs. 1 and 2). In two patients, the fat rind was so extensive that it simulated intralesional fat on MR imaging in at least one plane. Increased signal in the adjacent muscle on T2-weighted or fluid-sensitive sequences was present in 11 of the patients with myxoma (Figs. 2 and 3). Discussion The typical intramuscular myxoma is a well-defined ovoid lesion with fluidlike signal intensity, a peritumoral fat rind visible on T1weighted MR images, and an increased signal in the adjacent muscle on T2-weighted or fluid-sensitive MR sequences. These latter two features have not been previously emphasized and are the most reliable radiologic features for differentiating intramuscular myxoma from other myxoid soft-tissue lesions. D E We speculate that both of these features are the result of the infiltrative pattern of slow growth seen in intramuscular myxoma. Intramuscular myxoma is a paucicellular lesion composed of a mucoid basophilic matrix rich in mucopolysaccharide [4 6]. lthough lesions may appear well defined on imaging studies, these lesions have no capsule and will infiltrate the adjacent atrophic and edematous striated muscle [2, 4, 6] (Fig. 4). Mucoid matrix from the lesion may split the cells or become embedded in them [6]. We found that MR imaging planes along the long axis of the involved muscle yielded the greatest information regarding tumor border definition and differentiation of signal in the lesion and the adjacent muscle. The rind of adipose tissue is likely reactive fat, caused by muscle atrophy associated with the slowly growing mass [4] (Fig. 4). This rind of adipose C F Fig. 2. Intramuscular myxoma in vastus medialis of left thigh of 85-year-old woman., Coronal T1-weighted spin-echo MR image (TR/TE, 500/15) reveals subtle rind of adipose tissue at periphery of heterogeneous ovoid lesion., xial T1-weighted spin-echo MR image (600/15) at mid portion (asterisk) of lesion depicts heterogeneously hypointense and isotense signal intensity of lesion. C, Obtained at more inferior position than, this axial T1-weighted MR image allows easier appreciation of rind (arrow) of adipose tissue at inferior aspect of lesion. However, in axial plane, contour of rind may appear to be intermuscular fat rather than interface between lesion and surrounding muscle. D, Coronal short tau inversion recovery MR image (2000/20; inversion time, 150 msec) shows markedly hyperintense signal in myxoma as well as in adjacent skeletal muscle. E, xial T2-weighted spin-echo MR image (2000/80) obtained 10 mm inferior to level at which was obtained at very inferior margin of lesion (asterisk), shows extensive increased signal in adjacent skeletal muscle. F, xial T1-weighted MR image (600/15) obtained after injection of contrast material shows prominent heterogeneous enhancement (arrows). JR:178, May

ancroft et al. Fig. 3. Intramuscular myxoma in right adductor compartment in 76year-old man.

, Corresponding coronal T2-weighted fast spin-echo MR image (5150/125) exhibits primarily hyperintense lesion, with hypointense septal and nodular foci.

lthough enhancement was variable, intense heterogeneous enhancement was seen in 55% (6/11) of the cases.

The lesion is poorly vascularized, typically with capillarylike vessels [4].

4 ancroft et al. Fig. 3. Intramuscular myxoma in right adductor compartment in 76year-old man., Coronal T1-weighted MR image (TR/TE, 665/15) displays heterogeneous low and intermediate signal intensity of myxoma., Corresponding coronal T2-weighted fast spin-echo MR image (5150/125) exhibits primarily hyperintense lesion, with hypointense septal and nodular foci. tissue may be so extensive that it simulates fat in the lesion on MR imaging, as we found in two patients in our study, although it is typically more subtle. lthough enhancement was variable, intense heterogeneous enhancement was seen in 55% (6/11) of the cases. This pattern of enhancement seemed counterintuitive in that lesions are composed of bland spindle cells with an avascular highly myxoid stroma [6]. The lesion is poorly vascularized, typically with capillarylike vessels [4]. This paucity of vessels can be apparent on arteriography, on which lesions may be identified as defects in the enhanced surrounding muscle [4]. The absence of an elaborate vasculature is in sharp contrast to what is found in myxoid liposarcoma, which tends to show a rich, delicate, plexiform capillary vascular network [7]. nalysis of the enhancement pattern was difficult not only because of the mucoid nature of the lesion but also because of our inability to map the entire myxoma using the histologic samples provided for this retrospective review. lthough contrast enhancement has been previously reported [8 10], we speculate that the heterogeneous enhancement in most myxomas in our series likely reflects focal areas of relative hypervascularity within the lesions. Myxomas are usually described as welldefined lesions with a fluidlike signal intensity [8 14]. Features of perilesional fat rind and increased signal intensity on the adjacent muscle of T2-weighted or fluid-sensitive MR images have been noted but not emphasized as characteristic of intramuscular myxoma [11]. We tested the reliability of these signs as aids to distinguishing intramuscular myxoma from myxoid liposarcoma, the lesion with which it is most frequently confused. In a nonblinded review of 12 patients with myxoid liposarcomas, we found these signs yielded p values of less than for the presence of a fat rind and less than for the presence of the abnormal muscle signal. The identification of these features alone increases the likelihood a lesion is a myxoma rather than a myxoid liposarcoma by 20.4fold and 13.4-fold, respectively. Of the six patients with heterogeneously enhancing lesions, three (50%) had both a fat rind and increased signal intensity in the adjacent muscle in the same lesion. It is difficult to know whether a possible variability in the degree of aggressiveness of myxomas was due to the small sample size of our series. Limitations of our study include the small size of the study sample as well as the limitations inherent to a retrospective study. lthough T1-weighted MR images were available for all our patients, the fluid-sensitive MR sequences available for the patients varied and included conventional T2-weighted, T2weighted fast (turbo) spin-echo, fat-suppressed T2-weighted fast spin-echo, and short tau inversion recovery images. Not all of the studies included images paralleling the long axis of the involved muscle or gadolinium-enhanced images. Finally, no dynamic imaging was done. lthough imaging was performed immediately after injection of contrast material, it is possible that enhancement was caused in part by diffusion of the contrast material into the lesion. MR images that reveal a mass with the signal intensity of fluid, a perilesional fat rind, and an increased signal in the adjacent muscle on the T2-weighted or fluid-sensitive MR sequence are strongly suggestive of intramuscular myxoma. Enhancement of the lesions varies but is most frequently intense and heterogeneous. Fig. 4. Photomicrographs of intramuscular myxoma in 77-year-old woman., Infiltration of myxoma (asterisk) into adjacent skeletal muscle (arrows) is visible. Myxoma is composed of scanty, poorly outlined cells in mucoid basophilic matrix and lacks distinct capsule. (H and E, 100), Myxoma (black asterisk) seen on left lacks distinct capsule and is separated by condensed fibrous matrix (white asterisk) from atrophic and adipose-replaced skeletal muscle on right. (H and E, 40) 1258 JR:178, May 2002

5 MR Imaging of Intramuscular Myxoma lthough the recognition of these features will likely not obviate biopsy of any individual lesion, it will allow more accurate prebiopsy diagnosis and preoperative planning. cknowledgment We thank Elizabeth J. tkinson in the Department of iostatistics, Mayo Clinic, Rochester, MN, for her assistance in the statistical analysis. References 1. Stout P. Myxoma, the tumor of primitive mesenchyme. nn Surg 1948;127: Enzinger FM. Intramuscular myxoma, a review and follow-up study of 34 cases. m J Clin Pathol 1965;43: Weiss SW, Goldblum JR. Enzinger and Weiss s soft tissue tumors, 4th ed. St. Louis: Mosby, 2001: Kindblom LG, Stener, ngervall L. Intramuscular myxoma. Cancer 1974;34: Hashimoto H, Tsuneyoshi M, Daimaru Y, Enjoji M, Shinohara N. Intramuscular myxoma: a clinicopathologic, immunohistochemical, and electron microscopic study. Cancer 1986;58: Kransdorf MJ, Murphey MD. Imaging of soft tissue tumors. Philadelphia: Saunders, 1997: Weiss SW, Goldblum JR. Enzinger and Weiss s soft tissue tumors, 4th ed. St. Louis: Mosby, 2001: Court-Payen M, Ingemann Jensen L, jerregaard, Schwarz Lausten G, Skjoldbye. Intramuscular myxoma and fibrous dysplasia of bone: Mazabraud s syndrome a case report. cta Radiol 1997;38: Peterson KK, Renfrew DL, Feddersen RM, uckwalter J, El-Koury GY. Magnetic resonance imaging of myxoid containing tumors. Skeletal Radiol 1991;20: Schwartz HS, Walker R. Recognizable magnetic resonance imaging characteristic of intramuscular myxoma. Orthopedics 1997;20: Kransdorf MJ, Moser RP, Jelinek JS, Weiss SW, uetow PC, errey H. Intramuscular myxoma: MR features. J Comput ssist Tomogr 1989;13: May D, Good R, Smith DK, Parsons TW. MR imaging of musculoskeletal tumors and tumor mimickers with intravenous gadolinium: experience with 242 patients. Skeletal Radiol 1997;26: Cabral CEL, Guedes P, Fonseca T, Rezende J, Cruz LC, Smith J. Polyostotic fibrous dysplasia associated with intramuscular myxomas: Mazabraud s syndrome. Skeletal Radiol 1998;27: bdelwahab F, Kenan S, Hermann G, Lewis MM, Klein MJ. Intramuscular myxoma: magnetic resonance imaging. r J Radiol 1992;65: The 2003 RRS nnual Meeting Categorical Course will focus on oncologic imaging and will include sessions on CT, MR imaging, and FDG positron emission tomography. JR:178, May

Takayuki Ohguri 1 Takatoshi Aoki 1 Masanori Hisaoka 2 Hideyuki Watanabe 1 Katsumi Nakamura 1 Hiroshi Hashimoto 2 Toshitaka Nakamura 3 Hajime Nakata 1

Takayuki Ohguri 1 Takatoshi Aoki 1 Masanori Hisaoka 2 Hideyuki Watanabe 1 Katsumi Nakamura 1 Hiroshi Hashimoto 2 Toshitaka Nakamura 3 Hajime Nakata 1 Received July 1, 2002; accepted after revision November