Myofibroblasts were first observed in granulation tissue in 1971.1 Myofibroblastic sarcomas (MS) mainly composed of myofibroblasts were described by various terms,2 and used to be a controversial neoplasm. Recently, the clinicopathological features of low-grade myofibroblastic sarcomas occurring in various regions were reported.3,4 Low-grade myofibroblastic sarcoma is classified as a distinct entity in the newly published WHO classification of tumors of soft tissue.5 In this study, we examined the clinicopathological features of 20 cases of MS, and discussed the differential diagnosis from other spindle cell lesions showing fibroblastic and smooth muscle cell differentiation.
Patients and tissue samples
The cases were retrieved from the consultation files and surgical pathology archives of West China Hospital, Sichuan University between 1998 and 2005. All sections of the cases were confirmed by at least two independent pathologists and fulfilled the diagnostic criteria of MS:4 a myxoid to fibrous spindle-shaped cell tumor showing a predominant fascicular growth pattern with a diffusive and strong immunoreactivity for at least one myogenic markers (muscle specific actin, smooth muscle actin, or sometimes desmin), with signs of morphological malignancy such as infiltrating margins, mild to moderate nuclear pleomorphism; and WHO classification of tumors of soft tissue.5 The tumors were studied with respect to growth pattern (circumscribed or infiltrative), cellularity, nuclear atypia, mitotic figures and necrosis. MS was graded according to the modified NCI system,6 in which the presence of necrosis (up to 15%) and six or more mitotic figures per 10 high power field (HPF) was considered to be grade 2. Clinical information and follow-up data were directly obtained from the hospital records and the patients.
In all cases, the hematoxylin and eosin slides were reviewed, and paraffin-embedded blocks of tumors were used for immunohistochemistry by streptavidin-biotin complex method. The primary antibodies (all from the same manufacturer, Dako, Glostrup, Denmark) included: smooth muscle actin (SMA, monoclonal, 1A4, 1:100), desmin (monoclonal, D33,1:100), Ki-67 (monoclonal, MIB-1, 1:100), muscle specific actin (MSA, monoclonal, HHF35, 1:100), type IV collagen (polyclonal, 1:100), fibronectin (polyclonal, 1:300), laminin (monoclonal, 4C7, 1:50), h-caldesmon (monoclonal, h-CD, 1:100), vimentin (monoclonal,V9,1:100), ALK1 (monoclonal, ALK1, 1:100), and CD68 (monoclonal, PG-M1, 1:100). The positive and negative controls were included in each run. The tissues of leiomyoma were served as positive controls for SMA, MSA, desmin, h-caldesmon and vimentin. The tissues of basement membrane were served as positive controls for fibronectin, type IV collagen and laminin. The tissues of inflammatory myofibroblastic tumors (IMT) and malignant fibrous histiocytomas were served as positive controls for ALK1 and CD68 respectively. Primary antibody was substituted by normal mouse serum in a section as a negative control. Four µm-thick sections were deparaffinized, rehydrated and washed in phosphate buffer solution (PBS) for 15 minutes. Endogenous peroxidase activity was blocked by 30-minute incubation in 3% hydrogen peroxide solution. Tissue antigens were retrieved by pressure cooking at 120˚C for 8 minutes in a citrate buffer solution, pH 7.0. The specimens were washed with PBS, pH 7.5. Non- specific binding was blocked by incubating the slides with normal goat serum in PBS for 15 minutes at 37˚C, then incubated overnight at 4˚C with the primary antibodies. After washing with PBS for 3 times, the sections were incubated with secondary antibody, goat anti-mouse biotinylated antibodies for 40 minutes at 37˚C. After washing with PBS for 3 times, the sections were immunostained with avidin-biotin complex for 40 minutes at 37˚C. Visualization of the immunoreaction was conducted with 3, 3'-diaminobenzidine (DAB;Sigma, Co. UK) for 5 minutes. Finally, sections were counterstained with hematoxylin. Ki-67 labeling index was evaluated by counting the positively stained nuclei in at least 500 tumor cells.7 Immunoreactivity for other 10 antigens was evaluated semi-quantitatively by scoring as -, negative; +, <25% positive cells; ++, 25%-50% positive cells; +++, 50%-75% positive cells; ++++ ,>75% positive cells.
Materials retrieved from paraffin blocks of 5 cases (case No. 1, 2,4,6 and 10) were used for ultrastructural study. The tissues were deparaffinized in xylene and rehydrated in alcohol, then transferred to glutaraldehyde fixative, and then processed and embedded in Epon. Grids were cut and stained using lead citrate and urany1 acetate.8 Sections were observed under H-1600 LV transmission electron microscope, Japan.
The SAS9.0 statistical software was used for statistical analysis. The difference of Ki-67 labeling index between grade 1 and grade 2 MS was evaluated by Student's t test. P<0.05 was considered statistically significant.
The clinicopathological data of 20 MS cases is summarized in Table 1.The patients reported in the study included 12 men and 8 women aged 7-53 years (average, 32 years). Lesions involved the soft tissue of the head and neck region (n=4), femur (n=4), kidney (n=2), truck (n=6), extremity (n=1), lung (n=1) and the base of skull (n=2). Six cases were subcutaneous, 2 cases were submucosal, 3 cases were intramuscular, 6 cases were intraosseous, 2 cases were in the kidneys and 1 case in the lung. Case No. 4 was multicentric lesions, and the other 19 cases were single nodular. The most common clinical symptom was painless enlarging mass. The cases arising in the bone showed an osteolytic bone-destructive lesion without periosteal reaction under radiographic examination. Surgical resection of tumor was performed on 20 patients, with 5 patients (case No. 3, 11, 12, 13, 18) complemented with additional chemotherapy and 2 patients (case No. 15, 20) with radiotherapy. Three patients (case No. 4, 6, 10) were complemented with a combination of chemotherapy and radiotherapy. Clinical follow-up information of 19 patients was available over a mean duration of 27 months (range, 8－73 months): 2 cases (10%) recurred locally and developed distant metastasis 6 months to 4 years after initial operation. Nine cases (45%) recurred locally 17 to 46 months after initial excision,and 9 cases (45%) are alive without evidence of disease.
|view in a new window |
Table 1. Clinicopathological features of 20 cases of myofibroblastic sarcoma
The size of tumors ranged from 2 to 15 cm in diameter (average, 4.7 cm). The tumors were firm and fibrous with pale section surface. The tumors were diffusely infiltrative (10 cases), ill-defined (6 cases), and well-circumscribed but not encapsulated (4 cases). Microscopically, lesions in the muscle and bone showed the tumor cells infiltrated between muscle bundles (Fig.1A).Tumors occurring in the bone always had an infiltrative and destructive growth pattern. The tumor cells infiltrated into the bone trabeculae, and formation of reactive bone rimmed by osteoblasts was seen at the periphery of the tumors. Neoplastic bone and osteoid were absent in the tumors. Hypercellular areas and hypocellular areas, both of which distributed irregularly and transitioned to each other, can be identified in the same tumor. In some hypocellular areas, the feathery or stellate tumor cells with abundant eosinophilic cytoplasm and tapering nuclear showing moderate hyperchromatin and pleomorphism were scattered randomly in the myxoid stroma with infiltration of sparse lymphocytes and plasma cells, resembling tissue culture-like growth pattern (Fig.1B). In other hypocellular areas, the spindle tumor cells with tapering nuclei were arranged in fascicles pattern and deposited in a hyalinized collagenous matrix (Fig.1C). The hypercellular areas of the case occurring in the lung showed the spindle tumor sells with abundant eosinophilic cytoplasm, unclear cell border and minimal hyperchromatin and nuclear pleomorphism arranged in intersecting fascicle or vaguely storiform pattern, with abundant fibrous collagen in the stroma. The metastatic tumor in the brain of the case (Fig.1D) was histologically similar to the primary tumor in the lung. In other hypercellular areas, the moderate atypia tumor cells with indistinct cell boundary were arranged in a intersecting pattern with rich collagen in the stroma (Fig.1E).The mitotic figures, which arranged from 1-7/10 HPF, were easy to find. Coagulative necrosis can be seen in the centric areas of 6 cases in the study. Based on the nuclear atypia, mitotic figure rate and necrosis, 14 cases were grade 1 and 6 cases were grade 2. Tumors of grade 1 showed mild nuclear atypia, few mitotic figures and less cellulality (Fig.1A),and tumors of grade 2 indicated moderate nuclear atypia and hyperchromatin (Fig.1E).
|view in a new window |
Fig. 1. The morphologic spectrum of myofibroblastic sarcomas (HE, original magnification ×200). A: The spindle tumor cells with tapering nuclei infiltrating between the muscle bundles, admixed with sparse lymphocytes (case No.16); B: In the hypocellular areas, the feathery or stellate tumor cells with abundant eosinophilic cytoplasm showing a tissue culture-like pattern in the myxoid matrix accompanied by infiltration of inflammatory cells (case No.15); C: The spindle tumor cells with abundant pale cytoplasm arranging in fascicles, with abundant fibrous collagen in the stroma (case No.18); D: The tumor cells metastasized to the brain showing similar histological appearance with the primary tumor in the lung (case No. 6); E: The neoplastic cells with indistinct cell boundary and moderate nuclear atypia and hyperchromatin arranging in intersecting pattern (case No. 7).
The immunohistochemical findings of MS were summarized in Table 2. Immunohistochemically, the tumor cells strongly and extensively expressed vimentin (20/20), SMA (18/20, Fig. 2A), MSA (16/20, Fig. 2B) and fibronectin (20/20). Only two cases (2/20) focally expressed desmin. Type IV collagen, h-caldesmon, laminin, ALK1 and CD68 were uniformly negative staining. The average Ki-67 labeling index in MS of grade 2 was significantly higher than that in MS of grade 1 ((74.00±16.73) vs (32.67±14.62), t =-3.281, P=0.004).
|view in a new window |
Table 2. Results of immunohistochemistry of myofibroblastic sarcomas
|view in a new window |
Fig. 2. Immunohistochemical staining of MS. A: Diffuse cytoplasmic positive immunostaining for SMA (case No.7) (SABC, original magnification ×200); B: Extensive and strong positive staining for MSA (case No. 3, SABC, original magnification ×400);
Electron microscopic findings
Ultrastructural examination of 5 cases consistently showed that the tumor cells were spindle-shaped, and poorly developed fine filaments with focal densities and rich rough endoplasmic reticulum (rER) were present in the cytoplasm. Some bundles of fine filaments, known as the fibronexus junctions, which were roughly parallel with extracellular fibrils and diverged into the outside collagenous matrix, were seen at the cell surface (Fig.3A and B).
|view in a new window |
Fig. 3. Ultrastructrural features of MS. A: Abundant rough endoplasmic reticulum in the cytoplasm and the fibronexus at the cell surface (arrow, original magnification ×6900, case No.4); B: Spindle tumor cells showing poorly developed fine filaments with focal densities (arrows, original magnification ×5000, case No.2).
MS may occur in the patients of any age but seem to have a predilection for the adult patients with a slight male predominance.3,4 In this study, 16 cases (80%) were adults and 4 cases (20%) were children; 12 cases (60%) were males and 8 cases (40%) females. The current study indicated that MS predominantly involved the adult males. MS had been observed in diverse sites.3,9-12 In this study, the tumors involved the bone (6 cases, 30%), truck (6 cases, 30%), head and neck region (4 cases, 20%), extremity (1 case, 5%), lung (1 case, 5%) and the kidney (2 cases, 10%). According to our study and other literatures,4,9-11 MS involved preferentially the bone. Clinically, all patients presented with painless enlarging mass. A destructive and osteolytic lesion without the reaction of bone membrane was the most striking feature for MS in the bone under X-ray examination, which was demonstrated in this study and other report.4 On the whole, an infiltrative growth pattern is characteristic of MS.
In the study, myxoid stroma, collagenization and infiltration of inflammatory cells were common features for MS. Myxoid degeneration seemed to be more prominent in lesions of the bone than that of soft tissue. In the myxoid areas, the tumors were less cellular by showing spindle or stellate cells with eosinophilic cytoplasm and tapering nuclei scattered haphazardly in the stroma. Focal myxoid stroma of different degrees was present in 17 cases (85%) of the study. The myxoid stroma was frequently accompanied by the infiltration of inflammatory cells, which seemed to be a common feature for MS. A large number of inflammatory cells including lymphocytes, plasma cells and macrophages were noted in the pleomorphical myofibrosarcoma of bone.9 Moreover, inflammatory cells were considered to be occasionally numerous, but seem not to be a consistent feature for MS.13 In this study, the infiltration of inflammatory cells was were in 13 cases (65%), including 6 occurring in the bone, 4 in the head and neck region, 2 in the trunk, 1 in the lung and 1 in the extremity. In addition, MS variably have rich collagenous stroma, at least focally. It was reported that extensive stroma hyalinization was a common feature of myofibroblastic tumors.10 In our study, focal or extensive fibrous hyalinization areas were present in all cases of MS. Thus, we considered that MS frequently exhibits a diverse histological appearance, which includes myxoid areas, hyalinization areas and different degrees of inflammatory infiltration.
Diverse nuclear atypia in the same case or in different cases of MS was also noticeable. Nuclear pleomorphism was mild in the hypocellular areas, and moderate in the hypercellular areas. The mitotic figures were predominantly found in the hypercellular areas. In some areas, especially the hypocellular areas, the mitotic figures were absent. Thus, thorough sampling is crucial in establishing the diagnosis of MS. Several times of biopsy were performed on case No. 6, which was misinterpreted as reactive lesion owing to the large areas of hemorrhagic infarct, mild nuclear pleomorphism and few mitotic figures. Only when metastasis occurred in the brain six months after initial operation, was the tumor confirmed malignant. It was reported that necrosis was not uncommon for MS.2,13 In this study, coagulative necrosis was present in 6 cases, which all belonged to grade 2. Necrosis was not identified in the MS of grade 1. No high grade (grade 3) case were found in the current study. However, some cases of high-grade myofibroblastic sarcomas had been reported in other studies.3,13,14
Immunohistochemically, desmin-negative and actin- positive or desmin-positive and actin-negative were found in different myofibroblastic neoplasms.15 Recently, it was demonstrated that myofibroblasts were positive staining for calponin and negative for h-caldesmon, whereas smooth muscle cells and smooth muscle cell tumors were consistently positive staining for both calponin and h-caldesmon.16 Apart from desmin, α-smooth muscle actin and vimentin, an important marker for myofibroblastic differentiation, is the positive staining for fibronectin in combination with negative staining for laminin and type IV collagen,2 with which the immunohistochemical results of the study were consistent. Desmin, which was focally positive staining in two cases of this study, was a variable marker for MS. Desmin was found to be negative staining in the myofibrosarcoma of the breast,3 the upper jawbones,14 the salivary gland,17 and the pleomorphic myofibrosarcoma of bone.9 In addition, desmin was found to be positive staining in some cases of MS.4,10,12,18 According to our results, desmin was rarely expressed in MS. The immunohistochemical findings further demonstrated the variable immunophenotype of myofibroblastic tumors and confirmed the existence of desmin-negative/ actin-positive and desmin-positive/actin-negative myofibroblastic neoplasms.4 Additionally, we found that the proliferative marker of Ki-67 had a higher average labeling index in MS of grade 2 than that of grade 1. Although the statistical significance from this small group does not allow us to draw stringent conclusion, Ki-67 index may be useful in grading of MS. However, some MS of grade 2 indicated a very low Ki-67 index, while still showed an aggressive behavior by showing metastasis to the brain in a short period (case No. 6). Some grade 1 MS even demonstrated a very low Ki-67 index (case No. 17, 1%). Thus, low Ki-67 index can not exclude the diagnosis of MS.
Ultrastructural findings are the gold standard in defining myofibroblasts, which show the ultrastructural features of both fibroblasts (abundant rough endoplasmic reticulum) and smooth muscle cells (microfilaments with focal densities).19 The smooth-muscle cells are distinguished from the myofibroblasts by the fibronectin fibril and the fibronexus junctions, which had been argued as the most important criteria for defining the myofibroblast.2,11 Fibronexus junctions, which were first described and designated by Singer20 in 1979,is an unit composed of fibronectin and 5 nm microfilaments that converge at the cell surface, and is considered as the most characteristic ultrastructural marker for myofibroblasts.21 But it is not necessarily seen in every myofibroblastic tumor.9 In this study, we found that fibroblasts and smooth muscle cells in the cytoplasm were prominent features for MS. Fibronexus junctions were focally seen at the cell surface in 2 cases, and nuclei with indentation was also noted.
MS should be distinguished from leiomyosarcoma, IMT, fibromatosis and nodular fasciitis. Not infrequently, myofibrosarcomas bear resemblance to leiomyosarcoma under light microscopy.2 Leiomyosarcomas exhibit mainly a well delineated pushing margin and generally lack a diffusely infiltrative growth pattern, and show more eosinophilic and longitudinally fibrillary cytoplasm and cigar-shaped vesicular nuclei with paranuclear vacuolation. In contrast, the spindle-shaped cells of MS exhibit indistinct and paler cytoplasm, which is less fibrillary than that in leiomyosarcoma, and have a tapering rather than blunt-ended nucleus with common infiltration of inflammatory cells in the stroma. Immunohistochemically, previous studies demonstrated the stable expression of h-caldesmom in most smooth muscle tumors; in contrast, h-caldesmon was consistently negative staining in reactive myofibroblasts and myofibroblastic tumors.16 It is reported that h-caldesmon is a specific marker for smooth muscle cells and is valuable in differentiating leiomyosarcoma from myofibrosarcoma.22
Owing to the diverse histological appearance of MS, the hypocellular areas of MS may resemble the IMT morphologically, making the differential diagnosis become very difficult. Macroscopically, IMT is frequently sharply demarcated and composed of myofibroblasts and fibroblasts with more prominent inflammatory cell infiltration, whereas MS shows an invasive growth pattern and is mainly or entirely composed of myofibroblasts. IMT shows lower cellularity and no prominent nuclear atypia and mitotic figure, which are always present with variable degrees in MS. Additionally, more than 50% of IMT are positive staining for ALK1, which was negative staining in the present study. So ALK1 may be helpful in the distinction between MS and IMT. Case No. 10 in the study was initially misdiagnosed as IMT owing to the rare mitotic figures and marked myxoid and inflammatory cells infiltration in the background, but the metastasis occurring four years later excluded the diagnosis of benign tumor. We considered that myofibroblastic tumors with infiltrative growth pattern, high cellulality, and mild to moderate nuclear atypia should be classified as low grade MS instead of IMT, even very few mitotic figures are present.
It is evident that malignant fibrous histiocytoma (MFH) frequently contains myofibroblastic cells in varying degrees.21 MS was easy to be confused with so-called pleomorphical malignant fibrous histiocytoma of bone.23,24 Case No. 3 was initially misdiagnosed as MFH, but its immunostaining extensively positive for SMA, but negative for CD68,excluded the diagnosis of MFH. Histologically, MFH frequently shows anaplasia and multinucleated giant cells and marked cytological pleomorphism and nuclear atypia, often with bizarre tumor giant cells, admixed with spindle-shaped cells and round histiocyte-like cells. Immunohistochemically, MFH frequently expresses CD68 and lysozyme, and occasionally exhibits myoid phenotype focally, but extensive expression of the mark can exclude the diagnosis of MFH.25 When MS occurs in abdominal wall, the desmoid-type fibromatosis enters the differential diagnosis. MS and fibromatosis bear great similarity in that both of tumors indicate infiltrative growth pattern and contain spindle myofibroblasts. Case No. 1 in our study was initially misdiagnosed as desmoid-type fibromatosis, but its clinical aggressive behavior did not support the diagnosis of fibromatosis. Histologically, apart from myofibroblasts, fibromatosis contains some proportion of fibroblasts. Additionally, fibromatosis differentiates from MS in that the cells in fibromatosis do not touch one another, but are frequently separated by collagen owing to its richer hyalinization collagen. Furthermore, fibromatosis exhibits lower cellularity with absence of nuclear atypia, atypical mitotic figure and necrosis.
Nodular fasciitis shares a great resemblance with MS in that both kinds of tumors are composed of myofibroblasts with myxoid matrix and infiltration of inflammatory cells. MS can easily be misdiagnosed as nodular fasciitis when it is subcutaneous. Case No. 5 had initially been diagnosed as nodular fasciitis. In contrast, MS is more cellular, fascicle, and infiltrates into the muscle than nodular fasciitis, which exhibits no nuclear hyperchromatin, pleomorphism and atypical mitoses. Additionally, nodular fasciitis is usually characterized by short duration and smaller size.
MS were demonstrated to have a relatively indolent course with frequent recurrence and rare metastasis.3,4 In this study, distant metastasis occurred in 2 cases (all belonged to grade 2), focal recurrence was seen in 9 cases (45%), 7 of which were located in the bone. The result indicated that MS had higher recurrent rate in the bone than that in other locations. The reason may be that the lesions in the bone are difficult to be excised completely. According to our observation, MS of grade 1 are local aggressive lesions with frequent recurrence, so management by wide excision of the tumors with long-term follow-up is suggested. MS of grade 2 exhibit high recurrent rate and frequent metastasis, and should be managed by excision with a wide margin of normal tissue and adjuvant radiation therapy or systemic chemotherapy. Owing to the rare reports of MS, the cytogenetic and molecular genetic alterations of the tumors are presently obscure and need to be investigated in further study.
1. Gabbiani G, Ryan GB, Majne G. Presence of modified fibroblasts in granulation tissue and their possible role in wound contraction. Experientia 1971; 27: 549-550.
2. Eyden BP, Christensen L. Leiomyosarcoma versus myofibrosarcoma: observations and terminology. Ultrastruct Pathol 1993; 17: 231-239.
3. Taccagni G, Rovere E, Masullo M,Christensen L,Eyden B. Myofibrosarcoma of the breast: review of the literature on myofibroblastic tumors and criteria for defining myofibroblastic differentiation. Am J Surg Pathol 1997; 21: 489-496.
4. Mentzel T, Dry S, Katenkamp D,Fletcher CD. Low-grade myofibroblastic sarcoma: analysis of 18 cases in the spectrum of myofibroblastic tumors. Am J Surg Pathol 1998; 22: 1228-1238.
5. Mentzel T, Fletcher JA. Low grade myofibroblastic sarcoma. In: Fletcher CD, Unni KK, Mertens F, eds.World Health Organization Classification of Tumors: pathology and genetics of soft tissue and bone. Lyon, France: IARC Press; 2002: 94-95.
6. Guillou L, Coindre JM,Bonichon F, NguyenBB,Terrier P, Collin F, et al. Comparative study of the national cancer institute and french federation of cancer centers sarcoma group grading systems in a population of 410 adult patients with tissue sarcoma. J Clin Oncol 1997; 15: 350-362.
7. Oda Y, Takahira T, Kawaguchi K, Yamamoto H, Tamiya S, Matsuda S, et al. Low-grade fibromyxoid sarcoma versus low-grade myxofibrosarcoma in the extremities and trunk. A comparison of clinicopathological and immunohistochemical features. Histopathology 2004; 45: 29-38.
8. Ryska A, Ludvikova M, Szepe P, Boor A. Epithelioid haemangiosarcoma of the thyroid gland. Report of six cases from a non-alpine region. Histopathology 2004; 44: 40-46.
9. Watanabe K, Ogura G. Fibronexus in “malignant fibrous histiocytoma” of the bone: a case report of pleomorphic myofibrosarcoma. Ultrastruct pathol 2002; 26: 47-51.
10. Watanabe K, Ogura G, Tajino T, Hoshi N, Suzuki T. Myofibrosarcoma of the bone. A clinicopathologic study. Am J Surg Pathol 2001; 25: 1501-1507.
11. Eyden BP. Brief review of the fibronexus and its significance for myofibroblastic differentiation and tumor diagnosis. Ultrastruct Pathol 1993; 17: 611-622.
12. Kondo S, Yoshizaki T, Minato H, Horikawa I, Tatsumi A, Furukawa M. Myofibrosarcoma of the nasal cavity and paranasal sinus. Histopathology 2001; 39: 216-217.
13. Eyden BP, Banerjee SS, Harris M, Mene A. A study of spindle cell sarcomas showing myofibroblastic differentiation. Ultrastruct Pathol 1991; 15: 367-378.
14. Bisceglia M,Tricarico N,Minenna P, Magro G, Pasquinelli G. Myofibrosarcoma of the upper jawbones: a clinicopathologic and ultrastructural study of two cases. Ultrastruct Pathol 2001; 25: 385-397.
15. Fletcher CD, Tsang WY, Fisher C, Lee KC, Chan JK. Angiomyofibroblastoma of the vulva: a benign neoplasm distinct form aggressive angiomyxoma. Am J Surg Pathol 1992; 16: 373-382.
16. Watanabe K, Kusakabe T, Hoshi N, Saito A, Suzuki T. H-caldesmon in leiomyosarcoma and tumors with smooth muscle cell-like differentiation: its specific expression in the smooth muscle cell tumor. Hum Pathol 1999; 30: 392-396.
17. Bisceglia M, Magro G. Low-grade myofibroblastic sarcoma of the salivary gland. Am J Surg Pathol 1999; 23: 1435-1436.
18. Montgomery E, Goldblum JR, Fisher C. Myofibrosarcoma: a clinicopathologic study. Am J Surg Pathol 2001; 25: 219-228.
19. Schurch W, Seemayer TA, Gabbiani G. The myofibroblast: a quarter century after its discovery. Am J Surg Pathol 1998; 22: 141-147.
20. Singer II, Kazazis DM, Kawka DM. Localization of the fibronexus double-label immunogold electron microscopy on ultrathin frozen sections. Eur J Cell Biol 1985; 38: 94-101.
21. Lagace R, Schurch W, Seemayer TA. Myofibroblasts in soft tissue sarcomas. Virchows Arch A Pathol Anat Histol 1980; 389: 1-11.
22. Watanabe K. Leiomyosarcoma versus myofibrosarcoma. Am J Surg Pathol 2002; 26: 393-394.
23. Capanna R, Bertoni F, Bacchini P, Bacci G, Guerra A, Campanacci M. Malignant fibrous histiocytoma of bone. The experience at the Rizzoli Institute: report of 90 cases. Cancer 1984; 54: 177-187.
24. Montgomery E, Fisher C. Myofibroblastic differentiation in malignant fibrous histiocytoma (pleomorphic myofibrosarcoma): a clinicopathological study. Histopathology 2001; 38: 499-509.
25. Rangdaeng S, Truong LD. Comparative immunohistochemical staining for desmin and muscle-specific actin: a study of 576 cases. Am J Clin Pathol 1991; 96: 32-45.