- Email: [email protected]
Malignant fibrous histiocytomas in dogs and cats: an immunohistochemical study
Research in Veterinary Science 1992, 53, 198-204
Malignant fibrous histiocytomas in dogs and cats: an immunohistochemical study R. J. M. M. THOOLEN*,#, J. H. VOS+,*, J. S. VAN DER LINDE-SIPMAN*, R. A. DE WEGER§, J. A. M. VAN UNNIK§, W. MISDORP*, J. E. VAN DIJK*,* *Department of
Veterinary Pathology, Branch Laboratory Animal Pathology and Special Animal Diseases, Faculty of Veterinary Medicine, State University, 3508 TD Utrecht, +Animal Health Service in the Southern Netherlands, Molenwijkseweg 48, PO Box 4, 5280 AA Boxtel, *Department of Veterinary Pathology, Faculty of Veterinary Medicine, State University, 3508 TD Utrecht, §Department of Pathology, University Hospital, PO Box 85500, 3508 GA Utrecht, The Netherlands
Immunohistochemical staining was performed on seven canine and 10 feline soft tissue tumours histologically diagnosed as malignant fibrous histiocytomas (MFHS) or MFH-like tumours, and eight other histologically specified tumours (non-MFH). This was done to determine if commercially available antibodies that are used routinely in human diagnostic pathology for MFHs would express the same immunohistochemical patterns in canine and feline MFHSand MFH-like tumours. The antibodies were directed against human ~l-anti-trypsin (AT), human cO-anti-chymotrypsin (ACT), human lysozyme,bovine S-100 protein and human desmin. AT did not show any immunoreactivityin the tissues investigated. Except for one MFH, all canine MFHs and other soft tissue tumours with a 'histiocytic' character stained for lysozyme and not for S-100. Six out of seven canine MFHsand MFH-liketurnouts stained positive for desmin as did most non-MFH sarcomas. Most of the canine and feline MFHs and MFH-like tumours were positive for ACT. These findings for ACT staining in canine and feline MFHs and MFH-like tumours are in agreement with the findings in human MFHS.The immunohistochemical results of canine MFHs and MFH-like tumours were different from those in cats. Feline MFHSdiffered from canine MFHS for both lysozyme and desmin staining. THE term malignant fibrous histiocytoma (MFH) was introduced to refer to a group of soft tissue #Present address: Solvay-Duphar BV,C. J. van Houtenlaan 36, 1381CP Weesp, The Netherlands
tumours which were histologically cha.racterised by a storiform or 'cartwheel' growth pattern (Ozello et al 1963, O'Brien and Stout 1964). In man, this pleomorphic sarcoma constitutes the largest group within the soft tissue tumours (Weiss and Enzinger 1977, Fletcher 1987, Enzinger and Weiss 1988). The MFH is a pleomorphic sarcoma composed of spindle (fibroblast-like) cells often arranged in a storiform pattern and rounded (histiocytelike) cells. The MFH may contain pleomorphic giant cells with multiple hyperchromatic nuclei (Fu et al 1975, Churg and Kahn 1977, Taxy and Battifora 1977, Weiss and Enzinger 1977, Alquacil-Garcia et al 1978, Burns and Evans 1982, Kempson and Kyriakos 1982, Kindblom et al 1982, Roholl et al 1986, Enzinger and Weiss 1988). The giant cells have a rounded or multipolar shape and show abundant eosinophilic cytoplasm. The mitotic activity appears to be very prominent in this neoplasm, with many typical and atypical mitotic figures (Kyriakos and Kempson 1976, Weiss and Enzinger 1977, Kindblom et al 1982, Enzinger and Weiss 1988). In human MFH, haemorrhage, necrosis and inflammatory cells occur frequently. The light microscopic aspects of several subtypes of MFHs were described in detail by Enzinger and Weiss (1988). These morphological variants of MFH are classified as the storiformpleomorphic, myxoid, inflammatory, angiomatoid and giant cell type. The occurrence of MFH seems to be relatively rare in animals (Pool 1990). They have been
198
199
Immunohistoehemieal study of animal MFHs
reported most frequently in dogs and cats (Gleiser et a11979, Confer et a11981, Renlund and Pritzker 1984, Garma-Avina 1987, Thomas 1988, Allen and Duncan 1988). The giant cell subtype of M F H is mostly seen in cats (Gleiser et al 1979, Confer et al 1981, Renlund and Pritzker 1984, Allen and Duncan 1988). Animal and human MFHS show great histomorphological similarities (Pool 1990). In addition to conventional histomorphological examination, immunohistochemical staining techniques are frequently used in diagnostic pathology, to classify undifferentiated turnouts or tumours of unknown origin (Du Boulay 1982, Gatter et al 1985, Roholl et al 1985, Garter 1989). There are also several reports concerning the immunohistochemical staining of feline and canine neoplasms which indicate that some commercially available antibodies can be used for a more accurate classification of some tumours (Andreasen and Mahaffey 1987, Sandusky et al 1987, Baer et al 1989, Moore et al 1989, Rabanal et al 1989). In this study, the staining pattern of MFH, MFH-like and some other histologically specified tumours (haemangiopericytoma, lymphosarcoma, fibrosarcoma, histiocytoma and neurofibrosarcoma) of both dogs and cats were examined. Commercially available antibodies directed against (~l-anti-trypsin (AT), od-anti-chymotrypsin (ACT), lysozyme, S-100 and desmin were used. The immunohistochemical results in these tumours are discussed in relation to the immunohistochemical characteristics of human MFH.
TABLE 1 : Immunohistochemical staining of canine and feline malignant fibrous histiocytomas f o r ACT, lysozyme, S-100 and desmin
Number
1 23 4 5
Dogs ACT Lys S-100 Desmin
+ + +/-
-
+
+ + + +
-
Cats ACT Lys S-100Desmin
+
+/+ +
+
+/+
+
+/-
+
+ +/-
Over 100 labelled cells per 2.5 mm 2 0 to 100 labelled cells per 2-5 mm 2 No labelled cells ACT o~1-antichymotrypsin Lys Lysozyme S-100 Acidic neural protein
FIG 1 : Canine malignant fibrous histiocytoma (storiform-pleomorphic subtype). Note the distinct storiform arrangement of fibroblast-like cells. Van Gieson x 205
Materials and methods
Five canine and five feline subcutaneous neoplasms which were histologically diagnosed as malignant fibrous histiocytoma were used. Two canine and five feline sarcomas, showing some histological features of MFHs but lacking the overall storiform-pleomorphic pattern as was seen in the MFHS were classified as MFH-like turnouts. The diagnosis of M F H was made according to the criteria of Enzinger and Weiss (1988). All canine M F H S (Table 1: numbers 1 to 5) were histologically diagnosed as the storiform-pleomorphic subtype (Figs 1 and 2). Three MFHS of the cat were storiform-pleomorphic subtypes (Table 1: numbers 2, 3 and 4) and two feline MFHs were classified as the giant cell subtype (Table 1: numbers 1 and 5).
FIG 2: Canine malignant fibrous histiocytoma (storiform-pleomorphic subtype). Both histiocyte-like (short arrows) and fibroblast-like (long arrows) cells are present. Also, a few multinucleated giant cells are present (short arrowheads). Haematoxylin and eosin x 515
200
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
Additionally, five canine and three feline histologically specified soft tissue tumours (non-MFH) were used in this study. These turnouts were classified as haemangiopericytoma, lymphosarcoma, fibrosarcoma, histiocytoma or neurofibrosarcoma. Subcutaneous pre-existent unaltered muscle tissue in the tumour specimens and canine and feline lymph nodes with sinus histiocytosis were used as control tissues. All tissue specimens were surgically removed and sent for histological examination to this department. The tissues were fixed in 10 per cent neutral buffered formalin, paraffin embedded and sectioned at 7 ~ma. Sections were stained with haematoxylin and eosin and used for immunohistochemicalanalysis. Rabbit polyclonal antibodies used were directed against human ~zlanti-trypsin (AT; 1:150; Dakopatts), human c~lanti-chymotrypsin (ACT; 1:100; Boehring Diagnostica), human lysozyme (1:100; Dakopatts), human desmin (1:50; Euro-Diagnostics) or bovine S-100 (1:1000; Dakopatts). Antibody dilutions were made in phosphate buffered saline (PBS)containing 10 per cent normal goat serum. The primary antibodies were incubated for 60 minutes at room temperature. Immunoreactivity was detected by the three step avidin-biotin peroxidase method using biotinylated-anti-rabbit polyclonal antibodies, followed by an avidin-biotin-peroxidase complex incubation (1:100 in PBs containing 1 per cent normal serum; Vectastain, Vector Labs). Peroxidase was localised by diaminobenzidine (DAB; 50 mg 100 ml-a PBS 0"03 per cent hydrogen peroxide [H202] ). Endogenous peroxidase was blocked by incubating the sections for 30 minutes in 0.3 per cent H202 in methanol before incubation with the primary antibodies. The sections were counterstained with haematoxylin and embedded in synthetic resin.
TABLE 2: Immunohistochemieal staining of canine and feline MFH-like sarcomas for ACt, lysozyme, S-100 and desmin
Number 1 2 3 4 5 MFH + +/ACT Lys S-100
ACT + +
Dogs Lys S-100 Desmin + +
-
Cats ACT Lys S-100Desmin
+ +
+ + +
+ +/+
+
+
+
--/+ --
Malignant fibrous histiocytoma Over 100 labelled cells per 2.5 mm 2 0 to 100 labelled cells per 2.5 mm 2 No labelled cells C~1-antichymotrypsin Lysozyme Acidic neural protein
FIG 3: Immunohistochemical demonstration of ~l-antichymotrypsin in canine malignantfibrous histiocytoma. ACT-stained cells (long arrows) and non-stained (short arrows) tumour cells are seen. Note also the ACT-staining of multinucleated giant cells (thick arrows), x 515
Results
AT did not show any immunoreactivity in the lymph nodes, muscle tissue or tumour~s,mvesti, gated in either dogs or cats. The histiocytic cells in the sinus of the lymph nodes of both dogs and cats showed a positive staining for ACT and lysozyme. S-IO0 was present in normal nerve fibres of both dogs and cats. Desmin staining was observed in striated muscle of both species. The results of the immunohistochemical staining of histologically diagnosed canine and feline
FIG 4: Immunohistochemical demonstration of lysozyme in canine malignant fibrous histiocytoma. Stained tumour cells are indicated (long arrows), x 515
Immunohistochemical study of animal MFHs
201
T A B L E 3: Immunohistochemical staining of different canine and feline sarcomas for ACT, LYS, S - 1 0 0
and desmin
Staining
n
ACT
LYS
Haemangiopericytoma Lymphosarcoma Fibrosarcoma Histiocytoma Neurofibrosarcoma
1 2 2 1 2
. + +
+/. + +/-
Dogs S-100 . -
desmin + + . + +
ACT
LYS
+ .
Cats S-100
+
desmin
-
. +
+/-
+
+/-
+ +/-
O v e r 100 labelled cells per 2.5 mm2 0 to 100 labelled cells p e r 2.5 mm 2 No labelled cells Ac~r czl -antichymotrypsin LYS Lysozyme S - 1 0 0 Acidic neural protein
are presented in Table 1. In Table 2, the results in pleomorphic tumours resembling MFHS are summarised. Most of the canine MFHS and MFH-like sarcomas (five of seven) were positive for ACT (Fig 3), and all except one showed staining for lysozyme (Fig 4). All except one were negative for S-100. Four out of five canine MFHS and both MFH-like tumours showed desmin staining. Canine MFHs and MFH-like sarcomas which stained for ACT also showed staining for desmin. The majority of feline MFHs and MFH-like tumours stained for ACT. Only one feline MFH stained for lysozyme, whereas most MFH-like tumours stained for this enzyme. Only one feline MFH and one MFH-like tumour stained for S100. In the feline tumours, only one MFH and three of five MFU-like tumours stained for desmin. The staining results of the non-MFH tumours in dogs and cats are summarised in Table 3. In the canine non-MFH sarcomas both ACT positive tumours (n=2) showed staining for desmin. All canine non-MFH tumours were negative for S100. Four out of five tumours stained for desmin. The fibrosarcomas of both dogs and cats investigated did not stain with any of the tumour markers used in this study. The neurofibrosarcoma of the cats reacted positively with anti-S100 antibodies. MFHS
Discussion
•The immunoreactivity of some commercially available antisera directed against AT, ACT, lysozyme, S-100 and desmin that are used routinely in human diagnostic pathology was examined in canine and feline MFHS, MFH-like sarcomas and non-MFH sarcomas.
AT is a glycoprotein that is mainly synthesised by the liver and, to a lesser extent, by normal histiocytes and peripheral blood monocytes. A significant number of AT-positive cells has been demonstrated in human storiform-pleomorphic MFHS, although one can also obtain negative staining results in MFHS for this proteolytic enzyme (Du Boulay 1982, Nemes and Thomazy 1988). ACT is a proteolytic enzyme that neutralises the effect of chymotrypsin and can be localised in macrophages and within certain cells of the gastrointestinal tract. ACT stains a high proportion of both normal and malignant histiocytes in human neoplasms (Roholl et al 1985). The 'histiocytic' character of human MFHs can be confirmed by staining for AT and, or, ACT (Meister 1984, Roholl et al 1985). Mostly, ACT immunoreactivity is restricted to tumours of fibrohistiocytic phenotype, and Brooks (1986) concluded that 78 per cent of these fibrohistiocytic lesions expressed ACT. Lysozyme is present in histiocytes but can also be detected in serous acinar cells of the bronchial and salivary glands, Paneth cells, breast epithelium and renal tubular cells (Enzinger and Weiss 1988). Positive staining results with antibodies t o AT, ACT and lysozyme have been found in MFHS in man (Permanetter and Meister 1982, Roholl et al 1985, Mir et al 1986). Therefore, some workers endorsed the use of the markers mentioned above in the diagnosis of fibrohistiocytic tumours in man (Du Boulay 1982, Kindblom et al 1982). Desmin, as a member of the family of intermediate filaments, serves as an integral part of the cytoskeleton of cardiac, skeletal and smooth muscle fibres and can be used as a marker, indi-
202
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
cating myogenic origin, for rhabdomyosarcoma anti-desmin to sections of canine non-myogenic in both man (Miettinen et al 1982, Denk et al tumours can occur. 1983) and dogs (Meuten et al 1985, Andreasen In the cats, the immunohistochemical results et al 1988, Madewell et al 1988). Also in dogs, were different from those in the dogs. The majordesmin positivity has been found in normal mus- ity of the MFHs was also positive for ACT (three cle tissue and tumours of smooth muscle of five) but did not stain for lysozyme or desmin (Andreasen and Mahaffey 1987, Vos et al 1989, (four of five). Both the giant cell subtype as well Moore et al 1989, Rabanal et al 1989). as the storiform-pleomorphic type of MFH gave S-100, an acidic protein, is widely distributed variable results for ACT and lysozyme staining. in the central and peripheral nervous systems. Most of the MFH-like sarcomas of cats (n=5) Although S-100 was initially considered a brain stained for both ACT, lysozyme (four of five) and specific protein, it has been demonstrated in a for desmin (three of five). In both groups (MFHS variety of non-neural tumours as well (Enzinger and MFH-like tumours), the majority of cases did and Weiss 1988). Although S-100 detection is not stain for S-100 (four of five). Although feline used in the diagnosis of benign nerve sheath MFHs show histological features comparable to tumours and melanoma in man (Stefansson et human MFHs, the immunohistochemical staining al 1982, Kahn et al 1983, Angervall et al 1984, with regard to the antibodies used is clearly difHerrera and Pinto de Moraes 1984) and desmin ferent from canine MFHSand human MFHSa p a r t for myogenic tumours, positive immunoreactivity from the ACT staining. for S-100 or desmin was also found in human The results of this study show that several comMFHs (Lawson et al 1987, Donhuijsen et al 1988, mercially available antibodies to certain proteins Abdelatif et al 1989, Lin 1990). (ACT, lysozyme, S-100 and desmin) can be used The histiocytic markers directed against human for recognition of similar epitopes in certain soft ACT and human lysozyme stained positive in the tissue tumours of both dogs and cats. Anti-trypsin, control sections of lymph nodes of both dogs however, is an exception to this, as it did not and cats. Possibly, these markers could be of show any immunoreactivity for the histiocytes in additional help as histiocytic markers in both lymph nodes or for the tumours investigated. dogs and cats. In this study, the authors found Obviously, the antibodies used in this study that all canine MFHS showed positive staining cannot be used to differentiate conclusively for ACT and, or, lysozyme. The results for ACT between canine or feline MFHs, and other sarcostaining correspond with the results found in mas. In human beings also, MFH cannot always man where a large proportion of MFHs stains be classified unequivocally by immunohistochempositive for ACT (Permanetter and Meister 1982, istry, as quite frequent immunostaining for AT Roholl et al 1985, Mir et al 1986). In this study, and, or, ACT as well as the presence of various one of the canine MFHS (storiform-pleomorphic other antigens, for example, cytokeratins, subtype) showed S-100 positivity which can also vimentin, desmin, smooth muscle actin, IGF-I and be found in human MFH, as mentioned above. factor XIIIa, has been reported (Miettinen et al All ACT positive MFHS and MFH-like tumours 1982, Van Unnik 1986, Schurch et al 1987, in dogs showed positive staining for desmin. Miettinen 1988, Nemes and Thomazy 1988, Desmin positivity has previously been reported Roholl et al 1989, 1990a,b). in both human and murine MFHS (Lawson et al In conclusion, antibodies directed against 1987, Leader et al 1987, Roholl et al 1988, Lin human ACT, lysozyme and desmin gave positive 1990). Co-expression of ACT and desmin has been staining results in most canine MFHS and MFHdemonstrated in human MFHS by Donhuijsen et like tumours taken together. The results for ACTal (1988). Desmin staining of MFHS and MFH- staining in both canine and feline MFHS correlike sarcomas, could be due to differentiation of sponded with the findings in human MFHS. primitive mesenchymal ceils residing in the MFH, towards myofibroblasts. However, except for one Acknowledgements MFH and the fibrosarcoma, which did not stain for desmin, all canine tumours investigated Mrs J. van de Meer is thanked for her excellent showed desmin staining. These results may indi- technical assistance. The authors are also grateful cate that non-specific binding of this commercial to Rob Buijs for his assistance.
I m m u n o h i s t o c h e m i c a l s t u d y o f a n i m a l MFHs
References ABDELATIF, O. M., KHANKHANIAN, N. K., CROSBY, J. H., CHAMBERLAIN, C: R. Jr, SEIGLER, M. M. & TOM, G. D. (1989) Malignant fibrous histiocytoma and malignant melanoma: the role of immtmohistochemistry and electron microscopy in the differential diagnosis. Modern Pathology 2, 477-485 ALLEN, S. W. & DUNCAN, J. R. (1988) Malignant fibrous histiocytoma in a cat. Journal of the American Veterinary Medical Association 192, 90-91 ALQUACIL-GARCIA, A., UNNI, K. K. & GOELNNER, J. R. (1978) Malignant fibrous histiocytoma - an ultrastrnctural study of six cases. American Journal of Clinical Pathology 69, 121-129 ANDREASEN, C. B. & MAHAFFEY, E. A. (1987) Immunohistochemical demonstration ofdesmin in canine smooth muscle tumors. Veterinary Pathology 24, 211-215 ANDREASEN, C. B., WHITE, M. R. & SWAYNE, D. E. (1988) Desmin as a marker for canine botryoid rhabdomyosarcomas. Journal of Comparative Pathology 98, 23-29 ANGERVALL, L., KINDBLOM, L. G. & HAGLID, K. (1984) Dermal nerve sheath myxoma. A light and electron microscopical, histochemical and immuno-histochemical study. Cancer $3, 17521759 BAER, K. E., PATNAIK, A. K., GILBERTSON, S. R. & HURVITZ, A. I. (1989) Cutaneous plasmacytomas in dogs: a morphologic and immunohistochemical study. Veterinary Pathology 26, 216-221 BROOKS, J. J. (1986) The significance of double phenotypic patterns and markers in human sarcomas. A new model of mesenchymal differentiation. Journal of Pathology 125, 113-123 BURNS, B. F. & EVANS, W. K. (1982) Tumours of the mononuclear phagocyte system, a review of clinical and pathological features. American Journal of Haematology 13, 171-184 CHURG, A. M. & KAHN, L. B. (1977) Myofibroblasts and related cells in malignant fibrous histiocytoma and fibrohistiocytic tumours. Human Pathology 8, 205-218 CONFER, A. W., ENRIGHT, F. M. & BEARD, G. B. (1981) Ultrastrncture of a feline extraskeletal giant cell tumor (malignant fibrous histiocytoma). Veterinary Pathology 18, 738-744 DENK, H., KREPLER, R., ARTLIEB, U., GABBIANI, Q., RUNGER-BR~.NDLE, E., LEONCINI, P. & FRANKE, W. W. (1983) Proteins of intermediate filaments. An immunohistochemical and biochemical approach to the classification of soft tissue tumours. American Journal of Pathology 110, 193-208 DONHUIJSEN, K., METZ, K. & LEDER, L. D. (1988) Desmin und S-100-Protein positive maligne fibrose Histiozytome: praktische und theoretische Implikationen. Pathologe 9, 261-267 DU BOULAY, C. E. M. (1982) Demonstration of an alphal-antitrypsin and alpha 1-antichymotrypsin in fibrous histiocytomas using immunoperoxidase technique. American Journal of Surgical Pathology 6, 559-564 ENZINGER, F. M. & WEISS, S. W. (1988) Malignant fibrohistiocytic tumours. In Soft Tissue Tumours, St Louis, C, V. Mosby. pp269300 FLETCHER, D. S. M. (1987) Malignant fibrous histiocytoma? Histopathology 11, 433-437 FU, Y. S., GABBIANI, G., KAYE, G. I. & LATTES, R. (1975) Malignant soft tissue tumors of probable histiocytic origin (malignant fibrous histiocytoma) general considerations and electron microscopic and tissue culture studies. Cancer 3g, 176 GARMA-AVINA, A. (1987) Malignant fibrous histiocytoma of the giant cell type in a cat. Journal of Comparative Pathology 97, 551557 GATTER, K. C. (1989) Diagnostic immunocytochemistry: achievements and challenges. Journal of Pathology 159, 183-190 GATTER, K. C., ALCOCK, C., HERYET, A. & MASON, D. Y. (1985) Clinical importance of analyzing malignant tumoars of uncertain origin with immunohistochemical techniques. Lancet i, 13021305 GLEISER, C. A., RAULSTON, G. L., JARDINE, J. H. & GRAY, K. N. (1979) Malignant fibrous histiocytoma in dogs and cats. Veterinary Pathology 16, 199-208
203
HERRERA, G. A. & PINTO DE MORAES, J. (1984) Neurogenic sarcomas in patients with neurofibromatosis (yon Recklinghausen's disease). Virchows Archiv far Pathologische Anatomie 404, 361-376 KAHN, H. J., MAEKS, A., THOM, H. & BAUMAL, R. (1983) Role of antibody to S-100 protein in diagnostic pathology. American Journal of Clinical Pathology 79, 341-347 KEMPSON, R. L. & KYRIAKOS, M. (1982) Fibroxanthosarcoma of the soft tissues. A type of malignant fibrous histiocytoma. Cancer 29, 961-976 KINDBLOM, L., JACOBSEN, G. K. & JACOBSEN, M. (1982) Immunohistochemicalinvestigationsof tumours of supposed fibroblastic-histiocytic origin. Human Pathology 13, 835-840 KYRIAKOS, M. & KEMPSON, R. L. (1976) Inflammatory fibrous histiocytoma an aggressive and lethal lesion. Cancer 37, 15841606 LAWSON, C. W., FISCHER, C. & GATTER, K. C. (1987) An immunohistochemical study of differentiation in malignant fibrous histiocytoma. Histopathology 11, 375-383 LEADER, M., COLLINS, M., PATEL, J. & HENRY, K. (1987) Desmin: its value as a marker of muscle derived tumours using a commercial antibody. Virchows Archiv [A] 411, 345-349 LIN, J. S. (1990) Diagnostic application of immunohistochemistry in pleomorphic sarcomas. Chung-hua Ping Li Hsueh Tsa Chi 19, 67-70 MADEWELL, B., LUND, J., MUNN, R. & PINt, M. (1988) Canine laryngeal rhabdomyosarcoma: an immtmohistochemical and electron microscopic study. Japanese Journal of Veterinary Science 50, 1079-1084 MEISTER, P. (1984) Immunohistochemische Methoden bei der Diagnostik von Weichgewebstumoren. Pathologe 5, 90-98 MEUTEN, D. J., CALDERWOOD, M. M. B., DILLMAN, R. C., COOPER, B. J., VALENTINE, B. A., KUHAJDA, F. P., PASS, D. A. & MAYS, M. B. C. (1985) Canine laryngeal rhabdomyoma. Veterinary Pathology 22, 533-539 MIETTINEN, M. (1988) Antibody specific to muscle actins in the diagnosis and classification of soft tissue tumors. American Journal of Pathology 130, 205-215 MIETTINEN, M., LEHTO, V. P., BRADLEY, R. A. & VIRTANEN, I. (1982)Expression of intermediate filamentsin soft tissue sarcomas. International Journal of Cancer 30, 541-546 MIR, R., KAHN, L. B. & SELZR, G. (1986) Immunohistochemistry of primary gastrointestinal lymphomas: a study of 76 cases. Histopathology 10, 391-403 MOORE, A. S., MADEWELL, B. R. & LUND, J. K. (1989) Immunohistochemical evaluation of intermediate filament expression in canine and feline neoplasms. American Journal of Veterinary Research 50, 88-92 NEMES, Z. & THOMAZY, V. (1988) Factor XIIIa and the classic histiocyticmarkers in malignant fibrous histiocytoma: a comparative immunohistochemical study. Human Pathology 19, 822-829 O'BRIEN, J. E. & STOUT, A. P. (1964) Malignant fibrous xanthomas. Cancer 17, 1445-1455 OZELLO, L., STOUT, A. P. & MURRAY, M. R. (1963) Cultural characteristics of malignant histiocytomas and fibrous xanthomas. Cancer 16, 331-344 PERMANETTER, W. & MEISTER, P. (1982) Distribution of lysozyme (muramidase) and alphal-anfichymotrypsin in normal and neoplastic tissues. American Journal of Surgical Pathology 6, 715-727 POOL, J. R. (1990) Tumors of skin and soft tissues. In Tumors of Domestic Animals. Ed Jack Moulton. University of California Press, pp143-145 RABANAL, R. H., FONDEVILA, D. M., MONTANE, V., DOMINGO, M., & FERRER, L. (1989) Immunocytochemical diagnosis of skin tumours of the dog with special reference to undifferentiated types. Research in Veterinary Science 47, 129-133 RENLUND, R. C. & PRITZKER, K. P. H. (1984) Malignant fibrous histiocytoma involving the digit in a cat. Veterinary Pathology 21, 442-444 ROHOLL, P. J. M., ELBERS, H. J. R., PRINSEN, I. & VAN UNNIK, J. A. M. (1990a) Distribution of actin isoforms in sar-
204
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
comas: an immunohistochemical study. Human Pathology21, 12691274 ROHOLL, P. J. M., KLEYNE, J., ELBERS, H., VAN DER VEGT, M. C. D., ALBUS-LUTTER, C. H. & VAN UNNIK, J. A. M. (1985) Characterization of tumour cells in malignant fibrous histiocytomas and other soft tissue turnouts in comparison with malignant histiocytes. I Immunohistochemical study on paraffinsections. Journal of Pathology 147, 87-95 ROHOLL, P. J. M., KLEIJE, J., BLOKLAND VAN, M. E., SPIES, P. H., RUTGERS, D. H., ALBUS-LUTrER, CH.E. & VAN UNNIK, J. A. M. (1986) Characterization of two cell lines, derived from two malignant fibrous histiocytomas. Journal of Pathology 150, 103-112 ROHOLL, P. J. M., PRINSEN, I., RADEMAKERS, L. P. H. M., HSU, S. M. & VAN UNNIK, J. A. M. (1989) Two cell lines with epithelial-like characteristics, established from malignant fibrous histiocytomas. Uttramicroseopy31, 479-480 ROHOLL, P. J., RUTGERS, D. H., RADEMAKERS, L. H., DE WEGER, R. A., ELBERS, J. R. & VAN UNNIK, J. A. (1988) Characterization of human soft tissue sarcomas in nude mice. Evidence for histogenic properties of malignant fibrous histiocytomas. American Journal of Pathology 131, 559-68 ROHOLL, P. J. M., SKOTTNER, A., PRINSEN, I., LIPS, C. J. M., DEN OTTER, W. & VAN UNNIK, J. A. M. (1990b) Expression of insulin-like growth factor I in sarcomas. Histopathology 16, 455-60 SANDUSKY, G. E., CARLTON, W. W. & WIGHTMAN, K. A.
(1987) Diagnostic immunohistochemistry of canine round cell tumors. VeterinaryPathology 24, 495-499 SCHURCH, W., SKALLI, O., SEEMAYER, T. A. & GABBIANI, G. (1987) Intermediate filament proteins and actin isoforms as markers for soft tissue turnout differentiation and origin. I. Smooth muscle tumors. American Journal of Pathology 128, 91-103 STEFANSSON, K., WOLLMANN, R., & JENKOVIC, M. (1982) S-100 protein in soft-tissue tumors derived from Schwann cells and melanocytes. American Journal of Pathology 106, 261-268 TAXY, J. B. & BATTIFORA, H. (1977) Malignant fibrous histiocytoma - An electron microscopic study. Cancer 40, 254-267 THOMAS, J. B. (1988) Malignant fibrous histiocytoma in a dog. Australian VeterinaryJournal 65, 513-517 VAN UNNIK, J. A. M. (1986) Management of Soft Tissue and Bone Sarcomas. Ed Van Oosterom and J. A. M. Van Unnik. New York, Raven Press, p58 VOS, J. H., VAN DEN INGH, T. S. G. A. M., MISDORP, W., RAMAEKERS, F. C. S., VAN MIL, F. N. & DE NEIJS, M. (1989) An immunohistochemical study of canine tissues with vimentin, desmin, glial fibrillar acidic protein, and neurofflament antisera. Journal of Veterinary Medicine 36, 561-75 WEISS, S. W. & ENZINGER, F. M. (1977) Myxoid variant ofmafignant fibrous histiocytoma. Cancer 39, 1672-1685.
Received August 7, 1991 Accepted February 4, 1992
Malignant fibrous histiocytomas in dogs and cats: an immunohistochemical study R. J. M. M. THOOLEN*,#, J. H. VOS+,*, J. S. VAN DER LINDE-SIPMAN*, R. A. DE WEGER§, J. A. M. VAN UNNIK§, W. MISDORP*, J. E. VAN DIJK*,* *Department of
Veterinary Pathology, Branch Laboratory Animal Pathology and Special Animal Diseases, Faculty of Veterinary Medicine, State University, 3508 TD Utrecht, +Animal Health Service in the Southern Netherlands, Molenwijkseweg 48, PO Box 4, 5280 AA Boxtel, *Department of Veterinary Pathology, Faculty of Veterinary Medicine, State University, 3508 TD Utrecht, §Department of Pathology, University Hospital, PO Box 85500, 3508 GA Utrecht, The Netherlands
Immunohistochemical staining was performed on seven canine and 10 feline soft tissue tumours histologically diagnosed as malignant fibrous histiocytomas (MFHS) or MFH-like tumours, and eight other histologically specified tumours (non-MFH). This was done to determine if commercially available antibodies that are used routinely in human diagnostic pathology for MFHs would express the same immunohistochemical patterns in canine and feline MFHSand MFH-like tumours. The antibodies were directed against human ~l-anti-trypsin (AT), human cO-anti-chymotrypsin (ACT), human lysozyme,bovine S-100 protein and human desmin. AT did not show any immunoreactivityin the tissues investigated. Except for one MFH, all canine MFHs and other soft tissue tumours with a 'histiocytic' character stained for lysozyme and not for S-100. Six out of seven canine MFHsand MFH-liketurnouts stained positive for desmin as did most non-MFH sarcomas. Most of the canine and feline MFHs and MFH-like tumours were positive for ACT. These findings for ACT staining in canine and feline MFHs and MFH-like tumours are in agreement with the findings in human MFHS.The immunohistochemical results of canine MFHs and MFH-like tumours were different from those in cats. Feline MFHSdiffered from canine MFHS for both lysozyme and desmin staining. THE term malignant fibrous histiocytoma (MFH) was introduced to refer to a group of soft tissue #Present address: Solvay-Duphar BV,C. J. van Houtenlaan 36, 1381CP Weesp, The Netherlands
tumours which were histologically cha.racterised by a storiform or 'cartwheel' growth pattern (Ozello et al 1963, O'Brien and Stout 1964). In man, this pleomorphic sarcoma constitutes the largest group within the soft tissue tumours (Weiss and Enzinger 1977, Fletcher 1987, Enzinger and Weiss 1988). The MFH is a pleomorphic sarcoma composed of spindle (fibroblast-like) cells often arranged in a storiform pattern and rounded (histiocytelike) cells. The MFH may contain pleomorphic giant cells with multiple hyperchromatic nuclei (Fu et al 1975, Churg and Kahn 1977, Taxy and Battifora 1977, Weiss and Enzinger 1977, Alquacil-Garcia et al 1978, Burns and Evans 1982, Kempson and Kyriakos 1982, Kindblom et al 1982, Roholl et al 1986, Enzinger and Weiss 1988). The giant cells have a rounded or multipolar shape and show abundant eosinophilic cytoplasm. The mitotic activity appears to be very prominent in this neoplasm, with many typical and atypical mitotic figures (Kyriakos and Kempson 1976, Weiss and Enzinger 1977, Kindblom et al 1982, Enzinger and Weiss 1988). In human MFH, haemorrhage, necrosis and inflammatory cells occur frequently. The light microscopic aspects of several subtypes of MFHs were described in detail by Enzinger and Weiss (1988). These morphological variants of MFH are classified as the storiformpleomorphic, myxoid, inflammatory, angiomatoid and giant cell type. The occurrence of MFH seems to be relatively rare in animals (Pool 1990). They have been
198
199
Immunohistoehemieal study of animal MFHs
reported most frequently in dogs and cats (Gleiser et a11979, Confer et a11981, Renlund and Pritzker 1984, Garma-Avina 1987, Thomas 1988, Allen and Duncan 1988). The giant cell subtype of M F H is mostly seen in cats (Gleiser et al 1979, Confer et al 1981, Renlund and Pritzker 1984, Allen and Duncan 1988). Animal and human MFHS show great histomorphological similarities (Pool 1990). In addition to conventional histomorphological examination, immunohistochemical staining techniques are frequently used in diagnostic pathology, to classify undifferentiated turnouts or tumours of unknown origin (Du Boulay 1982, Gatter et al 1985, Roholl et al 1985, Garter 1989). There are also several reports concerning the immunohistochemical staining of feline and canine neoplasms which indicate that some commercially available antibodies can be used for a more accurate classification of some tumours (Andreasen and Mahaffey 1987, Sandusky et al 1987, Baer et al 1989, Moore et al 1989, Rabanal et al 1989). In this study, the staining pattern of MFH, MFH-like and some other histologically specified tumours (haemangiopericytoma, lymphosarcoma, fibrosarcoma, histiocytoma and neurofibrosarcoma) of both dogs and cats were examined. Commercially available antibodies directed against (~l-anti-trypsin (AT), od-anti-chymotrypsin (ACT), lysozyme, S-100 and desmin were used. The immunohistochemical results in these tumours are discussed in relation to the immunohistochemical characteristics of human MFH.
TABLE 1 : Immunohistochemical staining of canine and feline malignant fibrous histiocytomas f o r ACT, lysozyme, S-100 and desmin
Number
1 23 4 5
Dogs ACT Lys S-100 Desmin
+ + +/-
-
+
+ + + +
-
Cats ACT Lys S-100Desmin
+
+/+ +
+
+/+
+
+/-
+
+ +/-
Over 100 labelled cells per 2.5 mm 2 0 to 100 labelled cells per 2-5 mm 2 No labelled cells ACT o~1-antichymotrypsin Lys Lysozyme S-100 Acidic neural protein
FIG 1 : Canine malignant fibrous histiocytoma (storiform-pleomorphic subtype). Note the distinct storiform arrangement of fibroblast-like cells. Van Gieson x 205
Materials and methods
Five canine and five feline subcutaneous neoplasms which were histologically diagnosed as malignant fibrous histiocytoma were used. Two canine and five feline sarcomas, showing some histological features of MFHs but lacking the overall storiform-pleomorphic pattern as was seen in the MFHS were classified as MFH-like turnouts. The diagnosis of M F H was made according to the criteria of Enzinger and Weiss (1988). All canine M F H S (Table 1: numbers 1 to 5) were histologically diagnosed as the storiform-pleomorphic subtype (Figs 1 and 2). Three MFHS of the cat were storiform-pleomorphic subtypes (Table 1: numbers 2, 3 and 4) and two feline MFHs were classified as the giant cell subtype (Table 1: numbers 1 and 5).
FIG 2: Canine malignant fibrous histiocytoma (storiform-pleomorphic subtype). Both histiocyte-like (short arrows) and fibroblast-like (long arrows) cells are present. Also, a few multinucleated giant cells are present (short arrowheads). Haematoxylin and eosin x 515
200
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
Additionally, five canine and three feline histologically specified soft tissue tumours (non-MFH) were used in this study. These turnouts were classified as haemangiopericytoma, lymphosarcoma, fibrosarcoma, histiocytoma or neurofibrosarcoma. Subcutaneous pre-existent unaltered muscle tissue in the tumour specimens and canine and feline lymph nodes with sinus histiocytosis were used as control tissues. All tissue specimens were surgically removed and sent for histological examination to this department. The tissues were fixed in 10 per cent neutral buffered formalin, paraffin embedded and sectioned at 7 ~ma. Sections were stained with haematoxylin and eosin and used for immunohistochemicalanalysis. Rabbit polyclonal antibodies used were directed against human ~zlanti-trypsin (AT; 1:150; Dakopatts), human c~lanti-chymotrypsin (ACT; 1:100; Boehring Diagnostica), human lysozyme (1:100; Dakopatts), human desmin (1:50; Euro-Diagnostics) or bovine S-100 (1:1000; Dakopatts). Antibody dilutions were made in phosphate buffered saline (PBS)containing 10 per cent normal goat serum. The primary antibodies were incubated for 60 minutes at room temperature. Immunoreactivity was detected by the three step avidin-biotin peroxidase method using biotinylated-anti-rabbit polyclonal antibodies, followed by an avidin-biotin-peroxidase complex incubation (1:100 in PBs containing 1 per cent normal serum; Vectastain, Vector Labs). Peroxidase was localised by diaminobenzidine (DAB; 50 mg 100 ml-a PBS 0"03 per cent hydrogen peroxide [H202] ). Endogenous peroxidase was blocked by incubating the sections for 30 minutes in 0.3 per cent H202 in methanol before incubation with the primary antibodies. The sections were counterstained with haematoxylin and embedded in synthetic resin.
TABLE 2: Immunohistochemieal staining of canine and feline MFH-like sarcomas for ACt, lysozyme, S-100 and desmin
Number 1 2 3 4 5 MFH + +/ACT Lys S-100
ACT + +
Dogs Lys S-100 Desmin + +
-
Cats ACT Lys S-100Desmin
+ +
+ + +
+ +/+
+
+
+
--/+ --
Malignant fibrous histiocytoma Over 100 labelled cells per 2.5 mm 2 0 to 100 labelled cells per 2.5 mm 2 No labelled cells C~1-antichymotrypsin Lysozyme Acidic neural protein
FIG 3: Immunohistochemical demonstration of ~l-antichymotrypsin in canine malignantfibrous histiocytoma. ACT-stained cells (long arrows) and non-stained (short arrows) tumour cells are seen. Note also the ACT-staining of multinucleated giant cells (thick arrows), x 515
Results
AT did not show any immunoreactivity in the lymph nodes, muscle tissue or tumour~s,mvesti, gated in either dogs or cats. The histiocytic cells in the sinus of the lymph nodes of both dogs and cats showed a positive staining for ACT and lysozyme. S-IO0 was present in normal nerve fibres of both dogs and cats. Desmin staining was observed in striated muscle of both species. The results of the immunohistochemical staining of histologically diagnosed canine and feline
FIG 4: Immunohistochemical demonstration of lysozyme in canine malignant fibrous histiocytoma. Stained tumour cells are indicated (long arrows), x 515
Immunohistochemical study of animal MFHs
201
T A B L E 3: Immunohistochemical staining of different canine and feline sarcomas for ACT, LYS, S - 1 0 0
and desmin
Staining
n
ACT
LYS
Haemangiopericytoma Lymphosarcoma Fibrosarcoma Histiocytoma Neurofibrosarcoma
1 2 2 1 2
. + +
+/. + +/-
Dogs S-100 . -
desmin + + . + +
ACT
LYS
+ .
Cats S-100
+
desmin
-
. +
+/-
+
+/-
+ +/-
O v e r 100 labelled cells per 2.5 mm2 0 to 100 labelled cells p e r 2.5 mm 2 No labelled cells Ac~r czl -antichymotrypsin LYS Lysozyme S - 1 0 0 Acidic neural protein
are presented in Table 1. In Table 2, the results in pleomorphic tumours resembling MFHS are summarised. Most of the canine MFHS and MFH-like sarcomas (five of seven) were positive for ACT (Fig 3), and all except one showed staining for lysozyme (Fig 4). All except one were negative for S-100. Four out of five canine MFHS and both MFH-like tumours showed desmin staining. Canine MFHs and MFH-like sarcomas which stained for ACT also showed staining for desmin. The majority of feline MFHs and MFH-like tumours stained for ACT. Only one feline MFH stained for lysozyme, whereas most MFH-like tumours stained for this enzyme. Only one feline MFH and one MFH-like tumour stained for S100. In the feline tumours, only one MFH and three of five MFU-like tumours stained for desmin. The staining results of the non-MFH tumours in dogs and cats are summarised in Table 3. In the canine non-MFH sarcomas both ACT positive tumours (n=2) showed staining for desmin. All canine non-MFH tumours were negative for S100. Four out of five tumours stained for desmin. The fibrosarcomas of both dogs and cats investigated did not stain with any of the tumour markers used in this study. The neurofibrosarcoma of the cats reacted positively with anti-S100 antibodies. MFHS
Discussion
•The immunoreactivity of some commercially available antisera directed against AT, ACT, lysozyme, S-100 and desmin that are used routinely in human diagnostic pathology was examined in canine and feline MFHS, MFH-like sarcomas and non-MFH sarcomas.
AT is a glycoprotein that is mainly synthesised by the liver and, to a lesser extent, by normal histiocytes and peripheral blood monocytes. A significant number of AT-positive cells has been demonstrated in human storiform-pleomorphic MFHS, although one can also obtain negative staining results in MFHS for this proteolytic enzyme (Du Boulay 1982, Nemes and Thomazy 1988). ACT is a proteolytic enzyme that neutralises the effect of chymotrypsin and can be localised in macrophages and within certain cells of the gastrointestinal tract. ACT stains a high proportion of both normal and malignant histiocytes in human neoplasms (Roholl et al 1985). The 'histiocytic' character of human MFHs can be confirmed by staining for AT and, or, ACT (Meister 1984, Roholl et al 1985). Mostly, ACT immunoreactivity is restricted to tumours of fibrohistiocytic phenotype, and Brooks (1986) concluded that 78 per cent of these fibrohistiocytic lesions expressed ACT. Lysozyme is present in histiocytes but can also be detected in serous acinar cells of the bronchial and salivary glands, Paneth cells, breast epithelium and renal tubular cells (Enzinger and Weiss 1988). Positive staining results with antibodies t o AT, ACT and lysozyme have been found in MFHS in man (Permanetter and Meister 1982, Roholl et al 1985, Mir et al 1986). Therefore, some workers endorsed the use of the markers mentioned above in the diagnosis of fibrohistiocytic tumours in man (Du Boulay 1982, Kindblom et al 1982). Desmin, as a member of the family of intermediate filaments, serves as an integral part of the cytoskeleton of cardiac, skeletal and smooth muscle fibres and can be used as a marker, indi-
202
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
cating myogenic origin, for rhabdomyosarcoma anti-desmin to sections of canine non-myogenic in both man (Miettinen et al 1982, Denk et al tumours can occur. 1983) and dogs (Meuten et al 1985, Andreasen In the cats, the immunohistochemical results et al 1988, Madewell et al 1988). Also in dogs, were different from those in the dogs. The majordesmin positivity has been found in normal mus- ity of the MFHs was also positive for ACT (three cle tissue and tumours of smooth muscle of five) but did not stain for lysozyme or desmin (Andreasen and Mahaffey 1987, Vos et al 1989, (four of five). Both the giant cell subtype as well Moore et al 1989, Rabanal et al 1989). as the storiform-pleomorphic type of MFH gave S-100, an acidic protein, is widely distributed variable results for ACT and lysozyme staining. in the central and peripheral nervous systems. Most of the MFH-like sarcomas of cats (n=5) Although S-100 was initially considered a brain stained for both ACT, lysozyme (four of five) and specific protein, it has been demonstrated in a for desmin (three of five). In both groups (MFHS variety of non-neural tumours as well (Enzinger and MFH-like tumours), the majority of cases did and Weiss 1988). Although S-100 detection is not stain for S-100 (four of five). Although feline used in the diagnosis of benign nerve sheath MFHs show histological features comparable to tumours and melanoma in man (Stefansson et human MFHs, the immunohistochemical staining al 1982, Kahn et al 1983, Angervall et al 1984, with regard to the antibodies used is clearly difHerrera and Pinto de Moraes 1984) and desmin ferent from canine MFHSand human MFHSa p a r t for myogenic tumours, positive immunoreactivity from the ACT staining. for S-100 or desmin was also found in human The results of this study show that several comMFHs (Lawson et al 1987, Donhuijsen et al 1988, mercially available antibodies to certain proteins Abdelatif et al 1989, Lin 1990). (ACT, lysozyme, S-100 and desmin) can be used The histiocytic markers directed against human for recognition of similar epitopes in certain soft ACT and human lysozyme stained positive in the tissue tumours of both dogs and cats. Anti-trypsin, control sections of lymph nodes of both dogs however, is an exception to this, as it did not and cats. Possibly, these markers could be of show any immunoreactivity for the histiocytes in additional help as histiocytic markers in both lymph nodes or for the tumours investigated. dogs and cats. In this study, the authors found Obviously, the antibodies used in this study that all canine MFHS showed positive staining cannot be used to differentiate conclusively for ACT and, or, lysozyme. The results for ACT between canine or feline MFHs, and other sarcostaining correspond with the results found in mas. In human beings also, MFH cannot always man where a large proportion of MFHs stains be classified unequivocally by immunohistochempositive for ACT (Permanetter and Meister 1982, istry, as quite frequent immunostaining for AT Roholl et al 1985, Mir et al 1986). In this study, and, or, ACT as well as the presence of various one of the canine MFHS (storiform-pleomorphic other antigens, for example, cytokeratins, subtype) showed S-100 positivity which can also vimentin, desmin, smooth muscle actin, IGF-I and be found in human MFH, as mentioned above. factor XIIIa, has been reported (Miettinen et al All ACT positive MFHS and MFH-like tumours 1982, Van Unnik 1986, Schurch et al 1987, in dogs showed positive staining for desmin. Miettinen 1988, Nemes and Thomazy 1988, Desmin positivity has previously been reported Roholl et al 1989, 1990a,b). in both human and murine MFHS (Lawson et al In conclusion, antibodies directed against 1987, Leader et al 1987, Roholl et al 1988, Lin human ACT, lysozyme and desmin gave positive 1990). Co-expression of ACT and desmin has been staining results in most canine MFHS and MFHdemonstrated in human MFHS by Donhuijsen et like tumours taken together. The results for ACTal (1988). Desmin staining of MFHS and MFH- staining in both canine and feline MFHS correlike sarcomas, could be due to differentiation of sponded with the findings in human MFHS. primitive mesenchymal ceils residing in the MFH, towards myofibroblasts. However, except for one Acknowledgements MFH and the fibrosarcoma, which did not stain for desmin, all canine tumours investigated Mrs J. van de Meer is thanked for her excellent showed desmin staining. These results may indi- technical assistance. The authors are also grateful cate that non-specific binding of this commercial to Rob Buijs for his assistance.
I m m u n o h i s t o c h e m i c a l s t u d y o f a n i m a l MFHs
References ABDELATIF, O. M., KHANKHANIAN, N. K., CROSBY, J. H., CHAMBERLAIN, C: R. Jr, SEIGLER, M. M. & TOM, G. D. (1989) Malignant fibrous histiocytoma and malignant melanoma: the role of immtmohistochemistry and electron microscopy in the differential diagnosis. Modern Pathology 2, 477-485 ALLEN, S. W. & DUNCAN, J. R. (1988) Malignant fibrous histiocytoma in a cat. Journal of the American Veterinary Medical Association 192, 90-91 ALQUACIL-GARCIA, A., UNNI, K. K. & GOELNNER, J. R. (1978) Malignant fibrous histiocytoma - an ultrastrnctural study of six cases. American Journal of Clinical Pathology 69, 121-129 ANDREASEN, C. B. & MAHAFFEY, E. A. (1987) Immunohistochemical demonstration ofdesmin in canine smooth muscle tumors. Veterinary Pathology 24, 211-215 ANDREASEN, C. B., WHITE, M. R. & SWAYNE, D. E. (1988) Desmin as a marker for canine botryoid rhabdomyosarcomas. Journal of Comparative Pathology 98, 23-29 ANGERVALL, L., KINDBLOM, L. G. & HAGLID, K. (1984) Dermal nerve sheath myxoma. A light and electron microscopical, histochemical and immuno-histochemical study. Cancer $3, 17521759 BAER, K. E., PATNAIK, A. K., GILBERTSON, S. R. & HURVITZ, A. I. (1989) Cutaneous plasmacytomas in dogs: a morphologic and immunohistochemical study. Veterinary Pathology 26, 216-221 BROOKS, J. J. (1986) The significance of double phenotypic patterns and markers in human sarcomas. A new model of mesenchymal differentiation. Journal of Pathology 125, 113-123 BURNS, B. F. & EVANS, W. K. (1982) Tumours of the mononuclear phagocyte system, a review of clinical and pathological features. American Journal of Haematology 13, 171-184 CHURG, A. M. & KAHN, L. B. (1977) Myofibroblasts and related cells in malignant fibrous histiocytoma and fibrohistiocytic tumours. Human Pathology 8, 205-218 CONFER, A. W., ENRIGHT, F. M. & BEARD, G. B. (1981) Ultrastrncture of a feline extraskeletal giant cell tumor (malignant fibrous histiocytoma). Veterinary Pathology 18, 738-744 DENK, H., KREPLER, R., ARTLIEB, U., GABBIANI, Q., RUNGER-BR~.NDLE, E., LEONCINI, P. & FRANKE, W. W. (1983) Proteins of intermediate filaments. An immunohistochemical and biochemical approach to the classification of soft tissue tumours. American Journal of Pathology 110, 193-208 DONHUIJSEN, K., METZ, K. & LEDER, L. D. (1988) Desmin und S-100-Protein positive maligne fibrose Histiozytome: praktische und theoretische Implikationen. Pathologe 9, 261-267 DU BOULAY, C. E. M. (1982) Demonstration of an alphal-antitrypsin and alpha 1-antichymotrypsin in fibrous histiocytomas using immunoperoxidase technique. American Journal of Surgical Pathology 6, 559-564 ENZINGER, F. M. & WEISS, S. W. (1988) Malignant fibrohistiocytic tumours. In Soft Tissue Tumours, St Louis, C, V. Mosby. pp269300 FLETCHER, D. S. M. (1987) Malignant fibrous histiocytoma? Histopathology 11, 433-437 FU, Y. S., GABBIANI, G., KAYE, G. I. & LATTES, R. (1975) Malignant soft tissue tumors of probable histiocytic origin (malignant fibrous histiocytoma) general considerations and electron microscopic and tissue culture studies. Cancer 3g, 176 GARMA-AVINA, A. (1987) Malignant fibrous histiocytoma of the giant cell type in a cat. Journal of Comparative Pathology 97, 551557 GATTER, K. C. (1989) Diagnostic immunocytochemistry: achievements and challenges. Journal of Pathology 159, 183-190 GATTER, K. C., ALCOCK, C., HERYET, A. & MASON, D. Y. (1985) Clinical importance of analyzing malignant tumoars of uncertain origin with immunohistochemical techniques. Lancet i, 13021305 GLEISER, C. A., RAULSTON, G. L., JARDINE, J. H. & GRAY, K. N. (1979) Malignant fibrous histiocytoma in dogs and cats. Veterinary Pathology 16, 199-208
203
HERRERA, G. A. & PINTO DE MORAES, J. (1984) Neurogenic sarcomas in patients with neurofibromatosis (yon Recklinghausen's disease). Virchows Archiv far Pathologische Anatomie 404, 361-376 KAHN, H. J., MAEKS, A., THOM, H. & BAUMAL, R. (1983) Role of antibody to S-100 protein in diagnostic pathology. American Journal of Clinical Pathology 79, 341-347 KEMPSON, R. L. & KYRIAKOS, M. (1982) Fibroxanthosarcoma of the soft tissues. A type of malignant fibrous histiocytoma. Cancer 29, 961-976 KINDBLOM, L., JACOBSEN, G. K. & JACOBSEN, M. (1982) Immunohistochemicalinvestigationsof tumours of supposed fibroblastic-histiocytic origin. Human Pathology 13, 835-840 KYRIAKOS, M. & KEMPSON, R. L. (1976) Inflammatory fibrous histiocytoma an aggressive and lethal lesion. Cancer 37, 15841606 LAWSON, C. W., FISCHER, C. & GATTER, K. C. (1987) An immunohistochemical study of differentiation in malignant fibrous histiocytoma. Histopathology 11, 375-383 LEADER, M., COLLINS, M., PATEL, J. & HENRY, K. (1987) Desmin: its value as a marker of muscle derived tumours using a commercial antibody. Virchows Archiv [A] 411, 345-349 LIN, J. S. (1990) Diagnostic application of immunohistochemistry in pleomorphic sarcomas. Chung-hua Ping Li Hsueh Tsa Chi 19, 67-70 MADEWELL, B., LUND, J., MUNN, R. & PINt, M. (1988) Canine laryngeal rhabdomyosarcoma: an immtmohistochemical and electron microscopic study. Japanese Journal of Veterinary Science 50, 1079-1084 MEISTER, P. (1984) Immunohistochemische Methoden bei der Diagnostik von Weichgewebstumoren. Pathologe 5, 90-98 MEUTEN, D. J., CALDERWOOD, M. M. B., DILLMAN, R. C., COOPER, B. J., VALENTINE, B. A., KUHAJDA, F. P., PASS, D. A. & MAYS, M. B. C. (1985) Canine laryngeal rhabdomyoma. Veterinary Pathology 22, 533-539 MIETTINEN, M. (1988) Antibody specific to muscle actins in the diagnosis and classification of soft tissue tumors. American Journal of Pathology 130, 205-215 MIETTINEN, M., LEHTO, V. P., BRADLEY, R. A. & VIRTANEN, I. (1982)Expression of intermediate filamentsin soft tissue sarcomas. International Journal of Cancer 30, 541-546 MIR, R., KAHN, L. B. & SELZR, G. (1986) Immunohistochemistry of primary gastrointestinal lymphomas: a study of 76 cases. Histopathology 10, 391-403 MOORE, A. S., MADEWELL, B. R. & LUND, J. K. (1989) Immunohistochemical evaluation of intermediate filament expression in canine and feline neoplasms. American Journal of Veterinary Research 50, 88-92 NEMES, Z. & THOMAZY, V. (1988) Factor XIIIa and the classic histiocyticmarkers in malignant fibrous histiocytoma: a comparative immunohistochemical study. Human Pathology 19, 822-829 O'BRIEN, J. E. & STOUT, A. P. (1964) Malignant fibrous xanthomas. Cancer 17, 1445-1455 OZELLO, L., STOUT, A. P. & MURRAY, M. R. (1963) Cultural characteristics of malignant histiocytomas and fibrous xanthomas. Cancer 16, 331-344 PERMANETTER, W. & MEISTER, P. (1982) Distribution of lysozyme (muramidase) and alphal-anfichymotrypsin in normal and neoplastic tissues. American Journal of Surgical Pathology 6, 715-727 POOL, J. R. (1990) Tumors of skin and soft tissues. In Tumors of Domestic Animals. Ed Jack Moulton. University of California Press, pp143-145 RABANAL, R. H., FONDEVILA, D. M., MONTANE, V., DOMINGO, M., & FERRER, L. (1989) Immunocytochemical diagnosis of skin tumours of the dog with special reference to undifferentiated types. Research in Veterinary Science 47, 129-133 RENLUND, R. C. & PRITZKER, K. P. H. (1984) Malignant fibrous histiocytoma involving the digit in a cat. Veterinary Pathology 21, 442-444 ROHOLL, P. J. M., ELBERS, H. J. R., PRINSEN, I. & VAN UNNIK, J. A. M. (1990a) Distribution of actin isoforms in sar-
204
Thoolen, Vos, van der Linde-Sipman, de Weger, van Unnik, Misdorp, van Dijk
comas: an immunohistochemical study. Human Pathology21, 12691274 ROHOLL, P. J. M., KLEYNE, J., ELBERS, H., VAN DER VEGT, M. C. D., ALBUS-LUTTER, C. H. & VAN UNNIK, J. A. M. (1985) Characterization of tumour cells in malignant fibrous histiocytomas and other soft tissue turnouts in comparison with malignant histiocytes. I Immunohistochemical study on paraffinsections. Journal of Pathology 147, 87-95 ROHOLL, P. J. M., KLEIJE, J., BLOKLAND VAN, M. E., SPIES, P. H., RUTGERS, D. H., ALBUS-LUTrER, CH.E. & VAN UNNIK, J. A. M. (1986) Characterization of two cell lines, derived from two malignant fibrous histiocytomas. Journal of Pathology 150, 103-112 ROHOLL, P. J. M., PRINSEN, I., RADEMAKERS, L. P. H. M., HSU, S. M. & VAN UNNIK, J. A. M. (1989) Two cell lines with epithelial-like characteristics, established from malignant fibrous histiocytomas. Uttramicroseopy31, 479-480 ROHOLL, P. J., RUTGERS, D. H., RADEMAKERS, L. H., DE WEGER, R. A., ELBERS, J. R. & VAN UNNIK, J. A. (1988) Characterization of human soft tissue sarcomas in nude mice. Evidence for histogenic properties of malignant fibrous histiocytomas. American Journal of Pathology 131, 559-68 ROHOLL, P. J. M., SKOTTNER, A., PRINSEN, I., LIPS, C. J. M., DEN OTTER, W. & VAN UNNIK, J. A. M. (1990b) Expression of insulin-like growth factor I in sarcomas. Histopathology 16, 455-60 SANDUSKY, G. E., CARLTON, W. W. & WIGHTMAN, K. A.
(1987) Diagnostic immunohistochemistry of canine round cell tumors. VeterinaryPathology 24, 495-499 SCHURCH, W., SKALLI, O., SEEMAYER, T. A. & GABBIANI, G. (1987) Intermediate filament proteins and actin isoforms as markers for soft tissue turnout differentiation and origin. I. Smooth muscle tumors. American Journal of Pathology 128, 91-103 STEFANSSON, K., WOLLMANN, R., & JENKOVIC, M. (1982) S-100 protein in soft-tissue tumors derived from Schwann cells and melanocytes. American Journal of Pathology 106, 261-268 TAXY, J. B. & BATTIFORA, H. (1977) Malignant fibrous histiocytoma - An electron microscopic study. Cancer 40, 254-267 THOMAS, J. B. (1988) Malignant fibrous histiocytoma in a dog. Australian VeterinaryJournal 65, 513-517 VAN UNNIK, J. A. M. (1986) Management of Soft Tissue and Bone Sarcomas. Ed Van Oosterom and J. A. M. Van Unnik. New York, Raven Press, p58 VOS, J. H., VAN DEN INGH, T. S. G. A. M., MISDORP, W., RAMAEKERS, F. C. S., VAN MIL, F. N. & DE NEIJS, M. (1989) An immunohistochemical study of canine tissues with vimentin, desmin, glial fibrillar acidic protein, and neurofflament antisera. Journal of Veterinary Medicine 36, 561-75 WEISS, S. W. & ENZINGER, F. M. (1977) Myxoid variant ofmafignant fibrous histiocytoma. Cancer 39, 1672-1685.
Received August 7, 1991 Accepted February 4, 1992