109th Semi-Annual Seminar
December 3, 2000 Ritz Carlton Hote~ San Francisco, California
Soft Tissue Tumors
Christopher D.M. Fletcher~ M.D. Professor and Director ofSurgical .Pathology Brigham & Women's JlQspital Bt1stan, MA. California Tumor Tissue Registry
l109TH SEMI-ANNUAL CANCER SEMINAR
SURGICAL.PATHOLOGY OF SOFT TISSUE TUMORS
Christopher D. M. Fletcher, M.D., FRCPath
Brigham &. Women's Hospital Boston, Massachusetts
December 3, 2000 Ritz-Carlton Hotel San Francisco, California 8:30a.m.- 4:30p.m.
Dr. Christopher Fletcher is an internationally renowned expert on soft tissue tumors and has published over 300 papers in the field. Formerly the Director of Soft Tissue Tumor Unit at St. Thomas's Hospital in London, UK, Dr. Fletcher is now Professor of Pathology at Harvard Medical School, Boston, Ma.%achusetts, where he is aiso Director of Surgical Pathology, Brigham and Women's Hospital. ln 1996, his textbook, "Diagnostic Histopathology of Tumors" won the Royal Society of Medicine prize for Best Tllusrrated Medical Textbook, and in 1997 be received the prestigious Young Investigator Award from the US-CAP. TABLE OF CONTENTS
Case 1 Desmoplastic Fibroblastoma (Collagenous Fibroma) page 1 Case2 Cellular Angiofibroma page3 Case3 Myofibroblastoma of Breast pageS Case4 Inflammatory Myofibroblastic Tumor (Inflammatory Fibrosarcoma) page 8 Case 5 Pseudos_arcomatous Rt::active Myofibroblastic Proliferation page 11 CaseS Benign Fibrous Histiocytoma with Cellular & Aneurysmal Features page 17 Case? Fibrosarcomatous Variant of Dermatofibrosarcoma Protuberans page 21 Case 8 Solitary Fibrous Tumor page 24 Case9 Monophasic Synovial Sarcoma page27 Case 10 Gastrointestinal Stromal Tumor page 31 Case 11 Spindle Cell & Epithelioid Neoplasm, consistent with page 38 Malignant Peripheral Nerve Sheath Tumor Case 12 Diffuse Giant Cell Tumor (Pigmented Villonodular Tenosynovitis) page 42 Case 13 Extraskeletal Myxoid Chondrosarcoma page44 Case 14 High Grade Myxofibrosarcoma page47 Case 15 Atypical Lipomatous Tumor page49 Case 16 Dedifferentiated Liposarcoma page 54 Case 17 High Grade Spindle Cell Sarcoma with Osteogenic Differentiation page 58 Case 18 Alveolar Rhabdomyosarcoma page 62 Case 19 Intra-abdominal Desmoplastic Small Cell Tumor page67 Case 20 Epithelioid Hemangioendothelioma page 70 Color Plates page 73 Figure Legends page 82
Contacting the CTTR page 85
109TH CTIR SEMI-ANNUAL CANCER SEMINAR SURG/CA.L PATHOLOGY OF SOFT TISSUE TUMORS Case#l
History - Ace. #28587 Over a twelve month period, this 52-year-old male noted a slowly growing nodule on his right thigh. A 6.0 x 4.5 x 4.5 em well-circumscribed firm, gray-white nodule was excised. (Contributed by Kenneth Frankel, M.D.)
Diagnosis: Desmoplastic Fibroblastoma (Collagenous Fibroma)
Desmoplastic fibroblastoma was first described by Harry Evans 1 and has 2 3 subsequently been renamed 'collagenous fibroma' for fairly sound linguistic reasons. · Collagenous fibroma affects mainly adults (most often over the age of 40) with a male predominance and presents usually as a painless, slowly growing subcutaneous mass, most often less than 4 em in diameter (although larger lesions do occur). The anatomical distribution is wide, with some predilection for the upper limb and shoulder regions, and occasional cases are intramuscular. These lesions seem not to recur locally.
Grossly collagenous fib'romas appear fusiform or ovoid, firm a:nd rubbery and well circumscribed but, histologically, most cases have infiltrative margins with entrapment of fat or skeletal muscle. They are often centred on fascial or aponeurotic tissue and are composed of spindle-shaped or stellate fibroblasts, sparsely and haphazardly distributed in a copious, usually hyalinised collagenous or focally myxoid matrix containing very few blood vessels. Tumour cell nuclei are mainly ovoid with small nucleoli and mitoses are generally absent or very hard to find. Necrosis and cytologic atypia are not seen. Immunohistochemistry shows focal actin (SMA or HHF35) positivity in some (but not all) cases and, as with otber lesions having a myofibroblastic component, rare keratin-positive cells may be seen. Recent cytogenetic data suggest. a close relationship with the otherwise clinicopathologically distinct fibroma of tendon sheath.4
Differential diagnosis includes principally nodular fasciitis, which generany affects younger patients, is of shorter duration, is both more cellular and more vascular and often shows microcystic and inflammatory components, fibroma of tendon sheath, which ,generally arises in the hands or feet, is bette~ circumscribed, often more cellular and contains slit-like vascular channels, and desmoid fibromatosis, which is usually deep-seated, more cellular and has a consistently fascicular growth pattern.
Cl!R I 09"' Cancer S.e.mi nar Soft Tissue Tumors References: I. Evans In... Desmoplastic fibroblastoma. A repon of seven cases. Am J Surg Patlwl 1995; 19:1077-1081. 2. Nielsen OP, O'Connell JX, Dickersin OR. Rosenberg AE. Collagenous fibroma (desmoplastic fibroblastoma): A reporlofseven cases. Mod Patholl996; 9:781-785. 3. Miettinen M, Fetsch JF. Collagenous fibroma (desmoplastic fibroblastoma): A clinicopathologic analysis of 63 cases of a distinctive soft tissue lesion with stellate shaped fibroblasiS. Hum Patlwl 1998; 29:676-682. 4. Scior R et al. Collagenous fibroma (desmoplastic fibroblastoma): Genetic link with fibroma of tendon sheath? Mod Patlwll999; 12:565-568.
CTIR I09"' Cancer Seminar Soft Tiss11e Tumors 2 Case#2
History - Ace. #28412 This 49-year-old male develeiped a mass in the right groin area, present for several weeks and felt to be a very large, non-reducible, right inguinal hernia. The excision specimen consisted of a 59 gram, encapsulated mass of soft to rubbery yellow fat, measuring 6.0 x 5.5 x 4.0 em. (Contributed by Douglas Hoffmann, M.D.)
Diagnosis: Cel(ular Angiofibroma
Cellular angiofibroma is a fairly recently described benign vulval neoplasm which clinically is often mistaken for a Bartholin gland cyst. '·2 It occurs mainly in 1niddle aged women but comparable lesions have been described in lhe inguinal region of male patients3 under lhe inappropriate rubric of 'angiomyofibrobla~toma-like tumour'. We have seen rare cases in pelvic retrOperitoneum and putative examples at non-pelvic/perineal soft tissue locations have also been reponed.4 Thus far, jn our experience, no case· has eilher recurred locally or metastasised.
Morphologically, cellular angiofibroma is generally a circumscribed but unencapsulated neoplasm composed of uruform spindle cells arranged in short, irregularly intersecting fascicles with wispy stromal collagen bundles, numerous small to medium sized blood vessels which often show prorrunent hyalinization of their walls, and a varying (usually small) proportion of mature adipocytes. The tumour cells have ill-defined cell borders, pale cytoplasm and oval or fusiform, bland nuclei, thus closely resembling tumour cells of spindle cell lipoma. They stain positively for vimentin and, in some cases, for CD34 as well, whi le SMA, desmin and S-100 protein are negative. Generally, these neoplasms are cellular and show quite frequent mitoses, but pleomorphism and tumour necrosis are absent. StrOmal mast cells are common. Examples in males seem often to be larger and less clearly marginated than those in females, but this seems to have no biologic significance and there is no increased tendency for recurrence.
Because of the brisk miiotic activity, examples of ·cellular angiofibroma may be 1nislabeled as sarcoma, especially as leiomyosarcoma. However, smooth muscle tumours of lhe external genitalia have abundant eosinophilic cytoplasm and cigar-shaped nuclei, and they stain positively for myogenic markers. Because of its location, cellular angiofibroma has also to be distinguished from angiomyofibroblastoma, which is much less cellular and composed ofrather rounded myoid tumour cells, usually oriented around vessels, whicb·stain positively for desmiu. In contrast to cellular angiofibroma, spindle cell lipoma is more conunon in male patients and shows a predilection for the neck an.d upper back region. It typically contains ropey refracille collagen bundles and lacks numerous blood vessels w'ith hyalinised vessel walls; tumour cells in spindle cell lipoma stain consistently for CD34.
CTTR l 09" Cancer Seminar Soft Tissue Tumors 3 References: I. Nucci MR. Granter SR, Aetcher COM. Cellular angiofibroma: A benign neoplasm distinct from angiomyofibroblastoma and spindle cell lipoma. Am J Surg Pathol 1997; 21 :636-644. 2. Nucci MR, Fletcher COM. Vulvovaginal soft tissue tumours: update and review. Histopathology 2000; 36:97-108. 3. Laskin WB, Fetsch JF, Mostofi FK. Angiomyofibroblastoma-like rumor of the male genital tract. Analysis of II cases with comparison to female angiomyofibroblastoma. Am J Surg Patlzoll998; 22:6-16. 4. Garijo MF, Val-Bema! JF. Extravulvar subcutaneous cellular angiofibroma. J Cutan Pathol2000; 25:327-332.
C1TR 109" Cancer Seminar Soft Tissue T umon 4 Case#3
History (Composite) Acc.#28475 A 61 -year-ofd Hispanic male presented with a two-year history of a nontender, freely movable, hard subareolar left breast mass. There was no family history of either benign or malignant breast disease. Tbe patient had a long-standing history of cimetidine use. Upon discontinuing this, be noted a decrease in the size of the mass. The excisional biopsy specimen consisted of a 4.3 gram, 2.0 x 1.5 x 1.5 em fmn white, well circumscribed nodule. (Contributed by J$esa Chase, M.D.)
Acc.#28601 This 85-year-old Black male developed bilateral gynecomastia following trealroent of prostate cancer. A 3 x 2 x J.5 em well-circumscribed nodule was excised from the left breast. (Contributed by Evelyn Blazado, M.D.)
Diagnosis: Myofibroblastoma of Breast . Myofibroblastoma of the breast was first formally described just over ten years ago by Wargotz and colleagues from ·the Armed Forces Institute of Pathology.1 It seems that similar lesions were most often formerly labeled as spindle cell lipoma of the breast. It is a relatively infrequent (although increasingly recognised) breast tumour, which is unusual for showing a predilection for males- in fact approximately 60-70% of cases occur in men and the majority of affected patients are over the age of 50 years. However, the incidence in females appears to be rising as a consequence of mammographic screening.2 Most patients present with a small, painless, well-circumscribed mass, generally measuring Jess than 2-3 em although larger examples do occur and a single lesion measuring 10 em in dia~eter bas peen described. 3 Based on presently avail Surgically aod macroscopically, most cases of mammary myofibroblastoma appear very well circumscribed and even pseudoencapsulated - although this impression is not necessarily borne out histologically, since some cases contain entrapped lobules of breast tissue. Histologically these lesions bear some superficial resemblance to solitary fibrous tumour4 (but tend to be more fascicular) or to spindle cell lipoma (but usually are more cellular and lack the distinctive ropey collagen bundles). They are composed of short bundles and whorls of palely eosinophilic spindle cells with generally uniform tapering nuclei, set in a variably prominent, often rather hyaline or wiry coll!!genous stroma. ClTR 109'h Cancer Seminar Sojr Tiss~te Tllmors Mitoses are scarce and nuclear atypia is generally absent, although I have now seen several cases with focally striking nuclear pleomorphism and even rare atypical mitoses (which has raised concerns with regard to biologic potential). Many cases contain variable numbers of mature adipocytes, which are sufficiently numerous to suggest that they are an intrinsic part of the lesion. Rare ca~es have been described in which there was focal differentiation into more eosinophilic cells with cif-ar-shaped nuclei indicative of true smooth muscle differentiations and I (and others ) have also seen occasional cases containing nodules of mature cartilage. However an intrinsic epithelial component (which might suggest the possibility of a myoepithelial/mixed tumour) has never been described. Immunohistochemically these lesions are consistently CD34 positive, which led some authors to suggest a relationship (or even identity) with solitary fibrous tumour. However this seems highly . unlikely, given that most, if not al l mammary 2 6 7 myofibroblastomas also show convincing desmin positivity • • as well as focal positivity for smooth muscle actin. Furthermore, at the ultrastructural level, these lesions show 18 features of myofibroblastic differentiation. • In terms of differential diagnosis one thinks mainly of benign spindle cell neoplasms such as fibromatosis (more fascicular. more infiltrative. desmin and CD34 usually negative), nodular fasciitis (more inflamed with myxoid matrix, granulation tissue type myofibroblasts, desmin and CD34 usually negative) and perhaps neurofibroma (intralesional nerve fibres, more wavy nuclei, S-100 positive). One might also consider a spindle cell (metaplastic) carcinoma but the absence of any intraductal component or any immunopositivity for epithelial markers, as well as the general absence of c9tologic atypia. helps rule out this possibility. Regarding the distinction (if any) from spindle cell lipoma, this ha~ become a more complex issue following the recent recognition that mammary myofibroblastoma shows exactly the same cytogenetic aberrations at l3q and 16q as are seen in spindle cell lipoma.9 While this clearly suggests that these lesions are closely related entities, the subtle cytologic differences and the fact that spindle cell lipoma is only very rarely desmin positive leaves this issue unresolved as yet. An additional point of interest is the fact that lesions histologically identical to mammary-type myofibroblastoma (and showing the same CD34/desmin-positive phenotype} may occur at non-mammary locations. Specifically we have now encountered nearly 10 cases in this category, most of which were located in the inguinal or perineal regions (unpublished data). CTIR I 09'" Cancer Seminar Soft Tlssut Tumors 6 References: I. Wargotz ES et al. Myoflbroblastoma of the breast. Sixteen cases of a distinctive benign mesenchymal tumor. Am J Surg Patho/1987; 11:493-502. 2. Nucci MR. Fletcher CDM. Myofibroblastoma of the breast - a distinctive benign stromhl.tumor. Pathol Case Re.v 1999; 4:214-219. 3. Ali S et al. Giant myofibroblastoma of the male breast. Am J Surg Pathol 1994; 18:1170-1176. 4. Damiani S et al. Solitary fibrous tumour (myofibroblastoma) of the breast. Vir:chow.1· Ar:chiv 1994; 425:89-92. 5. Thoma~ TMM et al. Mammary myofibroblastoma with leiomyomatous differentiation. Am J Clin Patho/1997; 107:52-55. 6. Fukunaga M, Ushigome S. .Myofibroblastoma of the breast with divergent differentiations. Arch Pathol Lab Med 1997; 121:599-603. 7. Lee AHS et al. Myofibroblastoma of brea~t: an immunobistochemi:cal study. Histopathology 1993; 22:75-78. 8'. Begin LR. Myogenic stromal tumour of the male breast (so-called myofibro blastoma). Ultrastruct Pathol 1991; 15:613-622. 9. Pauwels P et al. Myofibroblastoma of the breast: genetic link with spindle cell lipoma. J Patho/2000; 191:282-285. CTTR 109'• Cancer Seminar Soft Tissue Tumors 7 Case#4 History - Ace. 26886 Because of a chronic cough, this 32 year old former smoke underwent a chest x-ray which showed a 5.0 x 4.0 em mass in the superior segment of the right lower lobe of lung. The 185 gram, 12.0 x 10.0 x 7.0 em segment of lung had a 4.7 x 4.5 x 3.5 em well· circumscribed, unencapsulated fmn nodule immediately adjacent to the bronchus. The nodule was composed of homogeneous rubbery firm yellow-tan to gray-tan tissue. No communication was found between the nodule and the bronchial tree. (Contributed by Raymond Lesonsky, M.D.) Diagnosis: Inflammatory Myofibroblastic Tumour (lnf/ilmmatory Fibrosarcoma) The concept of reactive "tumourous" proliferations composed of fibroblasts, myofibroblasts and chronic inflammatory cells grew in popularity following Bahadori and Liebow's seminal paper on plasma cell granuloma of the lung.' Subsequently such lesions were more often referred to as inflammatory pseudotumours or, more recently, as inflammatory myofibroblastic rumours2 to emphasise their major ceUulru:. constituent and, increasingly, to imply that such proliferations may represent benign neoplasms. They have been described in virtually all organs including the mesentery and retroperitoneum. In 1991 Meis and Enzi ngef proposed that, especially in children, at least some of the intra·abdominal lesions previously designated as inflammatory pseudotumour actually represent a true sarcoma which they called inflammatory fibrosarcoma It is my experience that there is sufficient clinicopathologic evidence to justifY such a designation for at least a small minority · of these lesions, but this is a controversial topic. The originally published clinical data in Meis and Enzinger's series described 38 cases. These arose predominantly in the mesentery and retroperitoneum of children (median age 8.5 years), although comparable lesions in adults also occur. Over one third of patients had multiple tumours. Among 27 patients with foUow·up, I0 developed local recurrence (3 repeatedly) and 3 developed metastases. Overall 6 patients died of their rumour, although interestingly most. were adults. Microscopically, these lesions displayed a spectrum of histologic appearances which incorporated 2 major components; plump spindle cells arranged in cellular fascicles or whorls and a prominent infiltrate of plasma cells and some lymphocytes. The spindle cells showed features of fibroblastic and myofibroblastic differentiation and, invariably, at lea~t some showed nuclear irregularity and contained very large nucleoli . In common with intlammatory pseudotumour, some cases showed areas of myxoid change and/or marked hyalinisation of collagen. CTTR 109'" Cancer Seminar Soft Tissue T1m1ors 8 It soon became clear, however, that a significant number of pathologists felt uncomfortable with the designation of sarcoma for these lesions. Such concerns were based upon the fact that some examples were histologically indistinguishable from the reactive process, long known a~ inflammatory pseudotumour, and upon the possibility that so-called metastasis in fact represented multicentricity. This alternative point of view wa~ supported by a large study of 84 cases from Coffin et al.4 These authors again noted the predilection for children and location within the abdomen, although they described cases at a wide variety of sites and occurring up to the age of 46 years. Frequently associated anaemia was noted, as was the occasional presence of other systemic features such as elevated ESR or unexplained · fever. From the histologic point of view, Coffin et al4 described three principal patterns which were commonly intermixed: .a) a myxoid fasciitis-like pattern with mixed inflammatory cells; b) a more cellular pattern resembling fibromatosis or fibrous histiocytoma but with numerous admixed plasma cells and c) a hypocellular fibrous pattern resembling scar tissue.. in which there were scattered chronic inflammatory cells and sometimes foci of calcification. Because some cases were multinodular, hard to eradicate and sometimes recurrent (approximately 25%), Coffin et al4 further reaffirmed their preference for the designation ' inflammatory myofibroblastic. mmour'. They also nofed 4 patients whose tumours regressed spontaneously and 2 patients who died of uncontrollable local disease. Importantly they made no mention of nuclear or cytologic atypia in any of their primary lesions and none of their patients developed metastatic di.sease. They did, however, describe two cases in which local recurrence was associated with the development of areas having histological features of a histiocytoid- malignant neoplasm. This phenomenon. had been described previously in lung lesions by Spencer and conceptually is quite difficult to reconcile with an intrinsically reactive proce-ss. My personal experience of more than 70 cases; based largely on consultation material with no paediatric bias, has revealed the same predilection for ·anatomic site and an age distribution of 2 months to 79 years, with 40% of cases occurring in patients. older than 18. At least a minority of cases does show the nuclear atypia described by Meis and Enzinger - and when this is combined with high cellularity, this inevitably raises concern for malignancy. This concern has been substantiated by the development of liver metastases (by all usual criteria) in at least 4 patients (all children), as well as a small number of thoracic and abdominal lesions with uncontrolled local disease, and I am ·aware of other convincing clinically malignant cases in other people's practices. Occasional aggressive cases have very bland morphology. Given this premise, are there any reliable histologic features which allow accurate recognition of inflammatory fibrosarcoma from among the group of inflammatory myofibroblastic lesions as a whole? The major "discriminant feature employed by Meis and Enzinger was thl!l of nuclear atypia. In my experience, focally, there may also be very cellular, fascicular areas. Undoubtedly, reliable distinction (in prognostically useful tern1s) is very difficult but consideration of the nuclear detail of such tumours perhaps allows recognition of a more aggressive subset. Available cytogenetic ·data have demonstrated clonal 5 6 aberrations in lesions of this type, · supporting their neoplastic nature, and the original CITR I09"' Cancer Seminar Soft Tissue Tumors 9 protagonists in the dispute concerning the nature of these lesions have moved towards a consensus viewpoint that they should all be regarded as low grade sarcomas.'·8 Further evidence strongly supporting the neoplastic nature of these lesions has been the demonstration of frequent clonal ALK gene rearrangements in these lesions,9 often in the context of chromosomal translocations indistinguishable from tliose seen in anaplastic large cell 9 10 lymphoma. • Such rearrangements generally lead to overexpression of ALK protein, especially in tumours affecting young patients, and this may be a useful diagnostic marker irnmunohistochernically. Principal considerations in the differential diagnosis include fasciitis-like proliferations, fibromatosis, deep benign fibrous histiocytoma and inflammatory leiomyosarcoma. Some of the fasciitis-like lesions reported at visceral locations may in fact belong in the inflammatory myofibroblastic tumour category - but most C References: I. BahadoriM, Liebow AA. Plasma cell granulomas of the lung. Cancer 1973; 31:191- 208. 2. Pettinato G et ·a!. inflammatory myofibroblastic tumor (plasma cell granuloma). Clinicopatholgic study of 20 cases witb immunohistochemical and ultrastructural observations. Am J Clin Pathol \990; 94:538-546. 3. Meis JM, Enzinger FM. Inflammatory fibrosarcoma of the mesentery and retroperitoneum. A tumor closely simulating inflammatory pseudoturnor. Am J Surg Pathol199J; 15:1146-1156. 4. Coffin CM, Watterson J, Priest J, Dehner LP. .Extrapulrnonary inflammatory myofibroblastic tumor: a clinicopathologic and immunohistochemical study of 84 cases. Ami Surg Pathol\995; 19:859-872. 5. Treissman SP et al. Omental-me.~entenc inflammatory pseudotumor. Cytogenetic demonstration of genetic changes' and monoclonality in one tumor. Cancer 1994; 73:1433-1437. 6. Sciot Ret al. Inflammatory myofibroblastic tumor of bone. Report of two cases with evidence of clonal chromosomal changes. Am J Surg Patho/1997; 21:1166-1172. 7. Coffin CM. Dehner LP, Meis-Kindblom JM. Inflammatory myofibroblastic tumor, ·inflammatory fibrosarcoma and related lesions: A historical overview with differential diagnostic considerations. Semin Diagn Patliol 1998; 15:102-1 10. CTTR I 09"' Cancer Seminar Soft Tiss"e T"nwrs I () 8. Meis-K.indblom JM, Kjellstrorn C, K.indblom L-G. fuflarnrnatory fibrosarcoma: an . update, reappraisal and perspective on its place in the spectrum of inflammatory rnyofibroblastic tumors. Semin Diagn Patho/ 1998;15: 133-143. 9. Griffin CA et a!. Recurrent involvement of 2p23 in inflammatory rnyofibro,blastic tumors. Cancer Res 1999; 59:2776-2780. 10. Lawrence B et a!. TPM3-ALK and TPM4-ALK oncogenes in inflammatory myofibroblastic tumoJ;S. Am J Patho/2000; 157:377-384. CITR 109"' Cancer Seminar Sojr Tissue Tumors I I Case#S History - Ace. #26921 At I0 weeks post-partum, this 20~year-o ld female presented with four days of gross hematuria. Delivery had been vaginal, and required the use of forceps. Physical examination revealed a firm mass in the anterior vaginal wall, on the left side. Cystoscopy showed a large fungating mass extending from the left posterolateral bladder wall. A partial cystectomy was performed. The resulting specimen was a soft 3.8 x 3.9 x 3.0 em ovoid portion of tissue with a yellow-tan cut surface. (Contributed by William Herrick, M.D.) Diagnosis: Pseudosarcomatous ReactiveMyofibroblastic Proliferation This case provides an opportunity to discuss the spectrom of what we currently regard as reactive myofibroblastic proliferations; among which the best known is nodular fasciitis. The proportion of cases in this category which are associated with an identifiable preceding cause or event is substantially less than 50% and, increasingly, molecular genetic data suggest that these represent self-limiting clonal lesions. The bone forming lesions in this category, fibro-osseous pseudotumour and myositis ossificans, are not included in this brief overview. Nodular fasciitis 1 1 Nodular fasciilis, " in its prototypical form, presents as a rapidly enlarging, often painful or tender subcutaneous nodule, most often affecting young adults and showing a predilection for the upper limb, especially the forearm. Although duration is variable, most cases have been present for less than 10- 12 weeks. Almost any anatomic site occasionally is affected and the less frequent cases seen in children predominate in the head and neck region.4 The depth of the lesion is also variable; approximately 10% of cases are entirely intramuscular and a small minority of cases arise either from periosteum (parosteal fasciitis) or in the skin (intradermal fasciitis). Parosteal lesions arising on the head, most commonly in children, often show striking erosion of bone and have been referred to as cranial fasciitis. Another infrequent subgroup of cases involves blood vessels, principally veins, in a multinodular manner, usually in the upper limb or head and neck region, and is known as intravascular fasciitis.5 These are all benign processes with a self-limiting clinical course and, if left untreated, they regress spontaneously. Local recurrence (or persistent growth) is very infrequent, occurring in about 2% of cases, and almost invariably is the result of piecemeal, macroscopically incomplete excision during the active growth phase. In the majority of cases, local marginal excision is adequate treatment and any microscopic residual lesional tissue undergoes spontaneous attrition by scarring. ClfR 109'' Cancer Seminar Soft Tis.wt Tumors 12 The majority of lesions are well-circumscribed but unencapsulated and most measure less than 3 em in diameter. The histologic features, irrespective of location, are characteristic ·and reproducible. Contrary to popular belief, most cases are well-marginated and no more than 10% have an irregularly infiltrative or arborizing growth pattern. The classic case consists of a cellular proliferation of myofibroblasts with plump nuclei and occasional nucleoli set in a loose collagenous stroma showing variable· degrees of myxoid, often microcystic change. Normal mitoses may be numerous but the spindle cell nuclei are never hyperchromatic or pleomorphic. The stroma contains delicate thin-walled capillaries, extravasated red blood cells and scattered inflammatory cells, mainly lymphocytes and neutrophils. Up to 10% of cases contain multinucleate giant cells, most often of osteoclastic type. The spindle cells are arranged in short interweaving fascicles and, when combined with the relatively pale, loose stromal tissue, this ·confers a feathery appearance. Occasional more cellular cases have a storiform pattern. In some cases, especially in children, the stroma shows· marked diffuse myxoid change, while longer standing lesions tend to show progressive stromal hyalinization and even keloidal change. A small proportion of q~ses show reactive new bone formation with plump osteoblast&, akin to the appearances in myositis ossificans; such cases are sometimes termed fasciitis ossificans. Immunohistochemically, as anticipated in any myofibroblastic lesion, the bulk of the cells show positivity for pan-muscle actin and smooth muscle actin (SMA}, while desmin is usually negative. With regard to differential diagnosis, this should only infre Proliferative faseiitis Proliferative fasciitis6 typically affects somewhat older adults than nodular fasciitis, with a peak incidence in the sixth decade. Tt is very uncommon in children.7 Predilection for the upper limb is not so marked as in nodular fasciitis and a sigruficant proportion of cases develop in th~ lower limb. Other than these differences, the clinical attributes including rarity of recurrence, short history aod lesional size ll[e much the same as in the classical nodular type (see above). Being essentially a variant of nodular fasciitis, there are many shared features. The major and distinctive difference, which led to the separate designation of this variant, is the presence in proliferative fasciitis of large, rounded or polygonill cells with amphophilic or rather basophilic cytoplasm and one or two vesicular nuclei containing a prominent nucleolus; these cells are described common! y as ganglion-like. The proportion of these cells within tbe lesion is variable but when they are numerous there is frequently confusion CITR I09"' Cancer Seminar Soj1 Tissue Tumors 13 with sarcoma. Other more minor differences from nOdular fasciitis are that the stroma is often even more myxoid and may contain fibrinous material. The large amphophilic ceUs are regarded generally as fibroblasts and it is interesting that, despite their copious cytoplasm, they are usually actin-negative, in contrast to the spindle-.shaped myofibroblasts with which they are admixed. Probably the only realistic differential diagnosis is pleomorphic rhabdomyosarcoma. However, the latter is rarely subcutaneous and shows far greater cytologic atypia and pleomorphism. frequently with atypical mitoses. Proliferative mvositis Proliferative myositis8 is essentially the intramuscular counterpart of proliferative fasciitis. In common with the latter it affects mainly older adults but shows anatomic predilection for the shoulder girdle and upper trunk. Ca~es. in childhood are rare. 7 The clinical duration qften is Jess than 4 weeks. As with all other reactive myofibroblastic proliferations, the outlook is excellent and recurrence is extremely infrequent. The histologic composition, including the presence of large ganglion-like cells, is the same as proliferative fasciitis. A characteristic and distinctive architectural feature is the manner in which this reactive process extends along fibrous septa and between individual muscle fibers, leaving the latter virtually intact. This produces a so-called checkerboard appearance of numerous individually separated eosinophilic muscle fibers set in a stroma of fasciitis. The growth pattern allows distinction from virtually any sarcoma and also from myositis ossificans. which is more localized and does not preserve muscle fibers. In common with the latter, however, !0-20% of cases of proliferative myositis contain small foci of metaplastic bone or cartilage. Ischemic fasciitis Ischemic fasciitis,9 which was initially reported under the term 'atypical decubital fibroplasia', 10 is a distinctive reactive fibroblastic lesion which OC·curs most ·often over bony prominences particularly (but not invariably) in immobile, usually debilitated, patients. The lesion develops most often in deep subcutis, especially around the limb girdles or sacral region. Hi~tologically such cases are characterized by broad zones of fibrinoid necrosis surrounded by vascular inflamed granulation tissue in which plump mitotically active fibroblasts and myofibroblasts, very similar to those seen in proliferative fasciitis, are encountered. These lesions are thought to have an ischemic, degenerative etiology, akin to decubitus ulcers. Local recurrence after excision is seen in some cases, most likely reflecting persistence of the predisposing cause. Reactive mvofibroblastic proliferations at visceral locations Lesions with histologic features similar to nodular or proliferative fasciitis have been recognized increasing1y in recent years to occur also at visceral locations, especially 11 17 the genitourinary tract and particularly thy bladder. ' Other rare locations in my experience include the larynx and sinonasal tract. Although originally believed to be mainly post-traumatic (usually postoperative), it is now realized that most such cases are apparently spontaneous. These clinically benign lesions occur most often in young adults and sometimes in children. Sex incidence is approximately equal. Lesions in the bladder typically present with haematuria. In my experience, Jess than 50% of cases are clearly associa(ed with a prior history of instrumentation or surgery of some kind - such cases are CTTR 109"' Cancer Seminar Soft Tiss11e T1tmors 14 sometimes known as ' postoperative spindle cell nodule'. These lesions may be large, often measuring around 5 em and sometimes larger and, when ass.ociated with a short bistory and dramatic symptoms, may be clinically very worrisome, making their pathologic interpretation especi11Uy anxiety-provoking. Local recurrence is very uncommon, is non destructive and is usually associated with incomplete excision. Histologically this group of lesions is characterised by fasciitis-like features but, because of mucosal ulceration and frequent involvement of the muscle coat (particularly in the bladder), there is often diagnostic concern for rhabdomyosarcoma, leiomyosarcoma or spindle cell (sarcomatoid) carcinoma. Such concern is enhanced by the frequent presence of elongated eosinophilic strap-like cells or plumper bizarre fibroblasts, similar to those Seen in proliferative fasciitis. Immunohistochemistry may not be so helpful in differential diagnosis since, in addition to consistent SMA positivity, these lesions are keratin positive in around 50% of cases and are also sometimes desmin positive, a phenotype entirely· consistent with myofibroblastic differentiation bm nevertheless unnerving on occasion! Important pointers in the differential diagnosis are clinical context (spindle cell TCC would be very rare under age 40, while vesical rhabdomyosarcoma would be rare over age 10), the discohesive distribution of lesiomil cells in a prominent fibromyxoid stroma, the diffuse stromal inflrurunatory infiltrate (most often neutropbils and lymphocytes) and the absence of significant nuclear pleomorphism or hyperchroinasia. Mitoses are variable in number but have normal morphology. These lesions are distinguished from inflammatory myofibroblastic rumor by clinical context and the fact that the latter tends to have a much more proininent chronic inflammatory component, especially plasma cells, and may show some degree of nuclear atypia Furthermore, these lesions are consistent! y ALK negative. References: 1. Allen PW. Nodular fasciitis. Pathology 1972; 4:9-26. 2. Bernstein KE, Lartes R Nodular (pseudpsarcomatous) -fasciitis; a non-recurrent lesion: Clinicopathologic study of 134 cases. Cancer 1982; 49:1668-1678. 3. Shimizu S, Hashimoto H, Enjoji M. Nodular fasciitis: An analysis of 250 patients. Pclthology 1984; 16:161-166. 4. Sarangarajan R, Dehner LP. Cranial and extracranial fa~ciitis of childhood: A clinicopathologic and immunohistochemical study. Hum Patlw/1999; 30:87-92. 5. Patcbefsky AS, Enzinger FM. Intravascular fasciitis. A report of 17 cases. Am J Surg Patho/ 1981; 5:29-36. 6. Chung EB, Enzinger F M. Proliferative fasciitis. Cancer 1975; 36:1.450- 1458. 7. Meis JM, Enzinger FM. Proliferative fasciitis and myositis of childhood. Am J Surg P.atho/1992; 16:364-372. 8. Enziilger FM, Dulcey F. Proliferative myositis. Report of thirty three cases. Cancer 1967; 20:2213-2223. 9. Perosio PM, Weiss SW. Ischemic fasciitis: A juxta-skeletal fibroblastic proliferation with a predilection for elderly patients. Mod Patho/1993; 6:69-72. 10. Montgomery EA, Meis JM, Mitchell MS et aL Atypical decubital fibroplasia. A distinctive fibroblastic pseudotumor occurring in debilitated patients. Am J Surg Patho/1992; 16:708-715. C1TR I 09"' Cancer Seminar Soft Tissue Tumors 15 11. Proppe KH; Scully RE, Rosai J. Postoperative spindle cell nodules of genitourinary tract resembling sarcomas. A report of eight cases. Am J Surg Patho/1984; 8:101-108. 12. Young RH, Scully RE. Pseudosarcomatous lesions of the urinary bladder, prostate gland and urethra. Arch Pathol Lab Med 1987; 111:354-358. 13. Albores-Saavedra J, Manivel JC, Essenfeld H et al . Pseudo-sarcomatous myofibroblastic proliferations in the urinary bladder of children. Cancer 1990; 66: 1234-1241. 14. Lundgren L, Aldenborg F, Angervall L et al. Pseudomalignant spindle cell proliferations of the urinary bladder. Hum Patholl994; 25:181-191. 15. Hollowood K, Pletcher COM. Pseudosarcomatous myofibrob1astic proliferations of the spermatic cord ("proliferative funiculitis"). Am J Surg Pathol 1992; 16:448- 454. 16. Ro JY et al. Pseudosarcomatous tibromyxoid tumor of the urinary bladder and prostate: immunohistochemical, ultrastructural and DNA flow cytometric analyses of nine cases. Hum Patho/1993; 24:1203- 1.210. . 17. Hojo H et al. Pseudosarcomatous myofibroblasric rumor of the urinary bladder in children: A ~tudy of II cases with review of the literature. Am J Surg Pathol 1995; 19:1224-1236. CTIR 109"' Cancer Seminar Soft Tissue TtutUJ rs 16 is that metastasising lesions have often seemed to recur locally more than once. prior to distant spread. Histologically, the most striking feature is the high cellularity of these tumors. often with a relatively more fascicular but partly storiform growth pattern. There is frequent involvement of the superficial subcutaneous tissue (although this is less diffusely infiltrative than that in DFSP) and. rarely, tumors are multinodular. Tumor cells typically have relatively abundant eosinophilic cytoplasm with variably p lump or tapering nuclei. Normal mitotic figures are common and may number up to I 0 per I 0 HPF. In almost all lesions focal cytologic polymorphism (inflammatory cells, foam cells and giant cells) is seen but this may be limited in extent Overlying epidermal hyperplasia is a frequent finding. Up to 12% of cases show central necrosis or infarction and, less commonly, ulceration is seen. Despite the unusual histologic features, identification of the lesion as a tibrous histiocytoma is not difficult if attention is paid to the epidermal changes, the presence of focal polymorphism and the somewhat hyalinized collagen bundles surrounded by tumor cells at the edge of the lesion. Occasionall y there is overlap with the atypical and aneurysmal variants (as in the present case). Up to 60% of cases show focal positivity for alpha smooth-muscle actin in a minority of spindle-shaped cells, but staining for desmin is generally negative and CD34 stains only adjacent or entrapped dermal fibroblaslS. Differential Diagnosis Cellular benign FH is confused often with leiomyosarcoma or dermatofibrosarcoma protuberans. The former, however, has plumper eosinophilic spindle cells with cigar shaped nuclei, a more uniform fascicular pattern, at least focal cytologic atypia and (when occurring in the skin) shows consistent, usually diffuse, positivity for desmin and smooth muscle actin. The latter lacks epidermal changes or cytological polymorphism and has a prominent storiforrn pattern; the neoplastic spindle cells have paler, poorly defined, cytoplasm and usually show both characteristically infiltrative involvement of subcutaneous tissue and diffuse positivity for CD-34. 9 10 Aneurysmal Benign Fibrous Histiocvtoma • Aneurysmal tibmus histiocytoma represents less than 2% of all FHs and, histologically, is often confused with a vascular neoplasm. Tt presents as a solitary, slowly growing, blue-brown nodule most often in the limbs of middle aged adults with slight predilection for females. Lesions tend to be larger than common FH and sometimes appear cystic due to extensive hemorrhage. The local recurrence rate is about 20%. As with 8 0 cellular FH, rare cases give rise to lymph node or pulmonary metastases. ·' Histologically, lesions can be polypoid or dome-shaped and the most characteristic low-power feature is the presenc·e, usually towards the center of the tumor. of blood-filled spaces. These spaces vary from irregular artifact-like clefts to cystic areas mimicking cavemous vascular channels and either can be focal or occupy most of the lesion. No endothelial cells can be seen lining these cystic spaces but there are numerous capillaries in the surrounding stroma, which usually shows foci of hemorrhage, abundant hemosiderin deposition, mono- and multinucleated cells with intraCytoplasmic hemosiderin, foamy macropbages and inflammatory cells. Away from the haemorrhagic areas, the CITR 109'' Cancer Seminar Soft Tis.•ut Tumors 18 cytomorphology closely resembles that of cellular FH. Useful diagnostic clues (e.g. cytologic polymorphism, entrapment of hyaline dermal collagen bundles at the lateral margins), usually present in all lesions, may be overlooked due to the extensive secondary changes described. Cellular atypia is not a common feature but mitotic counts vary up to 10 mitoses per 10 HPF and mitotic figures tend to be especially numerous adjacent to areas of hemorrhage. Epidermal hyperplasia usually is present and ulceration tends to be more common than in other variants of FH. Although FHs with abundant hemosiderin deposition ("hemosiderotic FH" or "sclerosing hemangioma"), but without blood-filled spaces, likely represent part of a morphologic continuum that incorporates aneurysmal FH, we do not include them in this category. Differential Diagnosis Aneurysmal FH lias to be distinguished from benign and malignant vascular tumors and from so-called angiomatoid 'malignant fibrous histiocytoma'. Spindle cell hemangioendothelioma is composed of true cavernous vascular spaces, papillary structures and more solid spindle cell areas with focally vacuolated epithelioid cells. Nodular Kaposi's sarcoma is composed of a .monomorphic, CD34 positive spindle cell population with slit-like spaces containing red blood cells. Although aneurysmal FH shares with angiomatoid 'malignant tibrous histiocytoma', the presence of cystic blood-filled spaces, the latter usually occurs . at a younger age, might present with systemic symptoms, is usually subcutaneous and shows monomorphic, ovoid-to-spindled myoid-looking cells with a prominent lymphopla$mocytic infiltrate and even germinal center formation. Around 40% of angiomatoid 'MFH's are desmin positive and, interestingly, 30-40% are also EMA positive in my expedence. References: 1. Vanni R, Fletcher CDM, Sciot R et a!. Cytogenetic evidence of clonality in cutaneous benign fibrous histiocytomas: a report of the CHAMP Study Group. Histopathology 2000; 37:2~2-217. 2. Chen T, Kuo T, Chan H. Dermatofibroma is a clonal proliferative disease. J Curmt Patho/2000; 27:36-39. 3. Calonje E, Mentzel T, Fletcher CDM. Cellular benign fibrous histiocytoma: Clinicopathologic analysis of 74 cases of a distinctive variant of cutaneous fibrous histiocytoma with frequent recurrence. Am J Surg Patho/1994; 18:668-676 .. 4. Fr CITR 109'b Canc.er Seminar Soft Tissue Tumors 19 9. Santa Cruz DJ, Kyriakos M. Aneurysmal ("'angiomatoid") fibrous histiocytoma of the skin. Cancer 1981; 47:2053-2061. 10. Calonje E, Fletcher COM. Aneurysmal benign fibrous histiocytoma: Clinico pathological analysis of 40 cases of a tumor frequently misdiagnosed as a vascular neoplasm. Histopathology 1995; 26:323-331. CITR I 09"' Cancer Seminar Soft Tissue Tumors 20 Case#7 History - Ace. #27885 About one year after first being noticed by this 37-year-old male, a small lump on the back began growing rapidly. The excision specimen consisted of a 13.3 x 10.4 ellipse of skin with a central raised, partially ulcerated mass. Sectioning showed a 5.6 em diameter well-circumscribed oval mass with a smooth, glistening appearance. (Contributed by Robert E. James, m, M.D., Ph.D.) Diagnosis: Fibrosarcomatous (Higher Grade) Variant of Dermatofibrosarcoma Protuberans ln its classical and relatively common form, dermatofibrosarcoma protuberans (DFSP) is well known to most pathologists. It is perhaps only the data concerning the variants of DFSP and its relationship to giant cell fibroblastoma which merit mention in a seminar of this type. Clinical Features 1 3 DFSP ' presents most often in adults of either sex between tbe ages of 30 and 50, but the lesion frequently has been present for many years and may only have recently caused clinical anxiety. This is because DFSP has a prolonged plaque (flat) phase of very slow growth in most cases and only becomes nodular or protuberant at a later smge. Perhaps for these reasons DFSP in childhood is uncommon, although it does occur• and I have even encountered congenital examples. The trunk (especially the abdominal wall) is by far the commonest location, while smaller percentages of cases arise in the limbs or head or neck region. Lesional size infrequently exceeds 5 em. ln its classical form DFSP has a 30-50% local recurrence rate un less widely excised, but the risk of metastasis is no more than 0.5%. This figure rises substantially in cases showing high grade Lransformation (see below). Histologic Appearances DFSP in most cases is characterized by monomorphic spindle cells, with pale ill defined cytoplasm (hence the lesion looks basophilic) and tapering nuclei, arranged in a consistent storiform (rush mat) pattern. Dermal involvement is invariable. Focal stromal hyalinization is quite frequent and may blur the storiform arrangement. Epidermal hyperplasia is usually absent and, generally speaking. DFSP Jacks the inflammatory component, histiocytes and multinucleate cells of fi brous histiocytoma. An important and distinctive feature is diffuse infiltration of subcutaneous fat, either in a honeycomb pattern or else in complex linear array resembling the roots of a tree. Stromal myxoid change is infrequent and prominent in no more !han 2% of cases (myxoid DFSpS): such cases may CITR I09"' Cancer Seminar Soft Tissue TumtJrs 21 show increased vascularity and relative loss of the storiform pattern. Occasional cases contain localized aggregates of somewhat bizarre or floret-like giant cells, usually close to the deep aspect of the lesion, which appear to represent the subtlest forn1 of differentiation towards giant eel! fibroblastoma. Rare cases of DFSP recur as giant cell fibroblastoma6 (see below). Immunohistochemically tumor cells in DFSP are consistently CD34 pOSlti ve (whereas cellular FH is negative) and stain negatively for S-100, actin and EMA. Ultrastructurally lesional cells most closely resemble fibroblasts. By cytogenetic analysis most cases of DFSP have ring chromosomes which are composed of material derived from chromosomes 17 and 22, arranged in a tumor-specific complex translocation,' the gene fusion product of which has been cloned.8 This is exactly the same molecular genetic aberration that characterises giant cell fibroblastoma,9 a lesion more often seen in childhood. Combining this fact with clear evidence that giant cell fibroblastoma may recur as DFSP and vice versa and that there exist lesions with hybrid features of both 'entities', it is now understood that these are very close! y related lesions. Variant~ 10 Pigmented DFSP (Bednar tumor) - between 5 ahd 10% of cases of DFSP contain variably prominent dendritic pigmented melanocytes. These dendritic cells are S- 100 positive if a chromogen which allows distinction from melanin pigment is used. Such lesions are commoner in black patients and, while this variant was held previously to support the concept of neuroectodermal differentiation in DFSP, it seems most likely that this phenomenon represents lesional colonization by ntelanocytes from the epidermis or skin adnexae. Aside from this finding there are no other clinicopathological differences from usual DFSP. 11 13 Fibrosarcomatous DFSP ' - tbis term refers to the approximately 5% of DFSPs which show transition to higher grade morphology with more frequent mitoses. Such transition may be gradual or abrupt and the higher grade component, while most often having the herringbone fascicular pattern of classic fibrosarcoma, may also be more eosinophilic (myofibroblastic) or more pleomorphic ('MFH' -jjke). Some cases are notable for the presence: of whorled nocjules of myoid eosinophilic cells,14 a feature whi.ch may also be seen (albeit less often) in conventional DFSP. Tbis phenomenon of tumour progression is more common in long-standing de novo lesions than in recurrences of ordinary DFSP and affects the same age group and anatomic locations as usual DFSP. There has been considerable controversy regarding the biologic significance of this variant, particularly with regard to its impact on metastatic rate. However cumulative data from the numerous small published series combined with our own recent study of 40 cases 12 reveals a significant metastatic rate of at least 15-2p%. Clearly, h<;>wever, this prognosis is better thao that for the (nowadays rare) conventional fibrosarcoma, hence whenever one encounters a fibrosarcomatous lesion in dermis or superficial subcutis, it is important to sample thoroughly for areas of conventional DFSP. ClTR 109'" Cancer Seminar Soft Tissue Tumors 22 References: I. Taylor HB, Helwig EB. Dermatofibrosarcoma protuberans. A study of 115 cases. Cancer 1962; 15:717-725. 2. Fletcher CDM et a!. Dermatofibrosarcoma protuberans: A clinicopathological and immunohistochemical study with a review of the literature. Histopathology 1985; 9:921-938. 3, Bowne WB, Antonescu CR, Leung DH et a!. Dermatofibrosarcoma protuberans: A clinicopathologic analysis of patients treated and followed at a single institution. Cancer 2000; 88:2711-2720. 4. McKee PH, Fletcher CDM. Dermatofibrosarcoma protuberans presenting in infancy·and childhood. J Cutan Pathol1991 ; 18:241-246. 5. Frierson HF, Cooper PH. Myxoid variant of dermatofibrosarcoma protuberans. Am J Surg eatholl983; 7:445-450. 6. Coyne J et al. Dermatofibrosarcoma protuberans recurring as a giant cell fibrob1astoma. Histopathology 1992; 21:184-187. 7. Naeem R et a!. Ring chromosomes in dermatofibrosarcoma protuberans are composed of interspersed sequences from chromosomes 17 and 22. Am J Pathol 1995; 147:1553-1558. 8. Simon MP et al. DeregJJlation of the platelet-derived growth factor B-chain gene via fusion with collagen gene COLlA! in dermatofibrosarcoma protuberans and giant cell fibroblastoma. Nature Genetics 1997; 15:95-98. 9. Rubin BP, Fletcher JA, Fletcher CDM. The histologic, genetic and biological relationships between dermatofibrosarcoma protuberans and giant cell fibroblastoma. An unexpected story. Adv Anat Pathol1997; 4:336-341. 10. Fletcher CDM et al. Pigmented dermatofibrosarcoma protuberans (Bednar tumour): Melanocytic colonisation or neuroectodermal differentiation? Histopathology 1988; 13:631-643. II. Ding J, Hashimoto H, Enjoji M. Dermatofibrosarcoma protuberans with fibrosarcomatous areas. Cancer 1989; 64:721-729. 12. Connelly JH, Evans HL. Dermatofibrosarcoma protuberans. A clinicopathologic review with emphasis on fibrosarcomatous areas. Am J Surg Pathol 1992; 16:921 - 925. 13. Mentzel T et al. Fibrosarcomatous ("high grade") dermatofibrosarcoma protuberans: clinicopathologic and immunohistochemiCal study of a series of 41 cases with emphasis on prognostic significance. Am J Surg Pathol 1998; 22:576- 587. 14. Calonje E, Fletcher CDM. Myoid differentiation . in dermatofibrosarcoma protuberans and its fibrosarcomatous variant: Clinicopathologic analysis of 5 cases. J Cutan Pathol 1996; Z:U0-36. CTTR I 09"' Cancer Seminar Soft Tissue Tru11ors 23 Case #8 History - Ace .. #28677 A 52-year-old female noticed a mass in the right submandibular region. The excised 15 gram submandibular gland contained a 2.0 x 1.5 x 1.5 em well-demarcated, firm, ovoid, grdy-white mass. CUt surface of the mass was uniform tan-gray, without hemorrhage or cystic foci. (Contributed by Bernard J. Poletti, M.D.) Diagnosis: Solitary Fibrous T.umour Solitary fibrous tumour, originally known as pleural fibroma or fibrous mesothelioma, 1 is now recognized increasingly at a very wide variety of extrapleural locations, among which the best documented are soft tissue, the peritoneal cavity, retroperitoneum, mediastinum, upper respiratory tract and the orbiL 2-s Other more recently recognized locations also include the meninges,9 liver10 and distal urogenital tract11 and I have also seen cases in the skin, adrenal and pancreas. It seems likely that extrapleural cases are much more common \han the better-known pleural lesions. The majority of published extrapleural cases have seemed to pursue a benign clinical course, but this probably reflects relatively short follow-up of these recently identified cases, since our experience has shown, comparable to pleural lesions, local recurrence or metastasis in approximately 10% of patients.12 In contra~t to intrathoracic cases, systemic features such as hypoglycemia or finger clubbing are much more infrequent. Cases in soft tissue affect mainly middle-aged aduhs, are generally slowly growing and deep-seated and affect the limbs, trunk and retroperitOneum equally. They may often reach a considerable size(> 10 em). Solitary fibrous tumours ansmg in sot't tissue are characteristically well circumscribed and apparently encapsulated. The cut surface tends to be firm, fibrous and variegated in colour. Histologically these lesions are entirely comparable to their pleuropulmonary counterparts and are characterized by a patternless spindle cell proliferation showing marked variation in cellularity. Hypocellular collagenous and small myxoid foci are common and, typically, areas of the collagen tend to be very hyalinized, if not keloidal. Cases with extensively myxoid stroma are infrequent and may be hard to recognise. 13 Prominent, often hemangiopericytoma-like vessels are a common finding and there may b~ perivascular hyalinisation. Tumour cells are usually monomorphic with hyperchromatic tapering nuclei and pale, poorly defined cytoplasm. Mitoses are generally scarce and necrosis in extrapleural lesions is uncommon. Occasional cases show (usually focally) increased cellularity, more frequent mitoses and scattered pleomorphic cells. Given the relative infrequency of metastasis, 1 presently designate such lesions as 'atypical solitary fibrous tumour' and advise careful follow-up. If mitoses> 4 per 10 HPF co-exist with nuclear pleomorphism or necrosis then 1 favour designation as malignant SFr.12 CTIR I09"' Cancer Seminar Soft Ti.rsue 1imwrs 24 However, since some cases desig,nated 'atypical SFf' and even occasional entirely bland lesions may metastasise and pursue a fatal course, it is my personal preference never to regard these tumours as definitively benign ·and I believe that they are all best regarded (and treated) as potentially malignant. In more than 90% of cases, tumour cells are CD34, CD99 and bcl-2 positive but negative for actin, S-100 and keratin. Occasional cases show scattered cells positive for EMA or desmin. There exists a group of probably closely related lesions that snow, in addition to features reminiscent of solitary fibrous tumour, variably prominent admixed adipocytes and usually more evenly distributed cellularity, sometimes with myx.oid foci . The precise nosologic status of the latter group remains unresolved, bur a small number of such cases we.re designated 'lipomatous haemangiopericytoma' by Nielsen et al 14 and others. 15 Their behaviour seems to be the same as that of solitary fibrous tumour. Once recognized, these lesions are in fact usually quite easy to recognize - their patternlessness and variable cellularity are remarkably distinctive! Immunohistochemistry generally facilitates distinction from lesions showing nerve sheath or smooth muscle differentiation. especially when combined with cytological features. Solitary fibrous tumour is quite often mislabeled as hemangiopericytoma but the latter, if it exists as a specific tumour type, tends to be much more cellular, to be more diffusely hypervascular, to be composed of smaller more basophilic cells and to show a consistent pericellular reticulin pattern. Cellular examples may closely resemble monophasic synovial sarcoma (especially in small biopsies) but positivity for epithelial markers and/or cytogenetic analysis allows that distinction. References: 1. England DM et al. Loca.lized benign and malignant fibrous tumors of the pleura: a clinicopathologic review of 223 c.ases. Am J Surg Pathol 1989; 13:640-658. 2. Goodlad JR, Fletcher COM. Solitary fibrous tumor arising at unusual sites: analysis of a series. Histopathology 1991; 19:515-522. 3. Witkin GB, .Rosai J. Solitary tibrous tumor of the upper respiratory tract: a report of six cases. Am J Surg Patholl99l; 15:842-848. 4. Westra WHet a!. Solitary fibrous tumor. Consistent CD34 immunoreactivity and occurrence in the orbit. Am J Surg Pathol 1994; 18:992-998. 5. Suster S .er al. Solitary fibrous rumors of soft tissue: a clinicopathologic and immunohistochemical study of 12 cases. Am J Surg Parho/1995; 19:1257-1266. 6. Hanau C, Miettinen M. Solitary fibrous tumor: Histological and immunohistochemical spectrum of benign and · malignant variants presenting at different sites. Hum Pathol 1995; 26:440-449. 7. Hasegawa T et al. Solitary fibrous tumor of the soft tissue.. An immunohistochemical and ultrastructural study. Am J Clin Pathol 1996; 106:217- 224. 8. 'Brunnemann RB et al. Extrapleural solitary fibrous tumqr. a clinicopathologic study of24 cases. Mod Parho/1999; 12:1034-1042. 9. Carneiro SS et al. Solitary fibrous tumor of the meninges: a lesion distinct from fibrous meningioma. Am J ClirlPatholl996; 106:217-224. CITR l09'' Cancer Seminar Soft Tissue Tumors 25 10. Moran CA, Ishak KG, Goodman ZD. Solitary fibrous tumor of the liver: A clinicopathologic and immunohistochemical study of nine cases. Arm Diagn Pathol1998; 2:19-24. 11. Westra WH, Grenko RT, Epstein J. Solitary fibrous tumor of the lower urogenital tract: a report offive cases involving seminal vesicles, urinary bladder and prostate. Hum Pathol2000; 31 :63-68. 12. Vallat-DeCouvelaere A-V, Dry SM, Fletcher CDM. Atypical and malignant solitary fibrous tumors in extrathoracic locations. Am J Surg Pathol 1998; 22: 1"5Q 1-1511 . 13. de Saint Aubain Somerhausen N, Rubin BP, Fletcher CDM. Myxoid solitary fibrous tumor - a study of seven cases with ·emphasis on differential diagnosis. Mod Pathol 1999; 12:463-471. 14. Nielsen GP et al. Lipomatqus bemangiopericytoma. A histologic, ultrastructural and immunohistochemical study. Am J Surg Patlro/1995; 19:748-756. 15. Folpe AL, Devaney K, Weiss SW. Lipomatous hemangiopericytoma - a rare variant of hemangiopericytoma that may be confused with liposarcoma. Am J Surg Patho/1999; 23:1201-1207, CTTR 109'' Cancer Seminar Soft Tissue Tumors 26 Case#9 History -Ace. #28446 This 49-year-old Caucasian female was found to have a mediastinal mass. The 1830 gram encapsulated mass was 20 x 18 x 10 em. The cut surface was firm white-tan with small foci of necrosis and hemorrhage. (Contributed by Xuedong Wang, M.D., Ph.l).) Diagnosis: Monophasic Synovial Sarcoma 1 5 Synovial sarcoma ' occurs at any age but peak incidence is between the ages of I 0 and 35 years wi"th a slight male predominance. Anatomic location is widely distributed but more than 60% arise in the lower limb, especially the thigh. A small bu~ significant proportion of cases arise on the trunk, especially in the abdominal wall, in the head and 6 neck region and in the medistinum , as in this case. It is also now well recognized that these tumors may arise primarily in the lung or pleura1.8. Exceedingly rare isolated cases have been reported to arise within a vessel, a nerve or a joint but, although many cases arise close to major articular stnictures, it is no longer believed that these lesions originate from (or have any true. relationship with) synovium. Most lesions are deep-seated and size is very variable. It is distinctive. that many patients present with longstanding pain, often preceding the appearance of a palpable mass, and, even in painless cases, some patients describe a protracted preoperative history, even of I0 years or more. Radiologic evidence of lesional calcification is a frequent tinding. Overall 5-year survival probability is about 50% but falls to only 20-30% at 10 years. In general, small tumor size (<5 em), early clinical stage and early age at presentation (<10 years) are signs of better prognosis. Morphologically, synovial sarcoma falls imo two main groups: biphasic and monophasic spindle cell, of which the latter is more common, depending on sampling. Both variants share a spindle cell population arranged in fascicles with uniform, tapering nuclei and pale, poorly defined cytoplasm set in a variably colhigenous stroma which often has a somewhat wiry or hyaline appearance. These. spindle cells tend to be monomorphic, show a variable mitotic rate and are often associated with stromal mast cells and a branching, hemangiopericytoma-like vascular pattern. Biphasic lesions, in classkal fom1 , contain variably numerous glandular structures lined by well-differentiated cuboidal to columnar epithelium. Occasional cases show squamous metaplasia. Other (not strictly biphasic) cases are recognizable by the presence of clusters of plumper, more rounded cells which are packeted by reticulin within an otherwise spindle cell lesion. Approximately 5- 4 10% of cases have a poorly differentiated appearanee, ,9.IO most often characterized by undifferentiated round cell morphology resembling Ewing's sarcoma ilJld less often by a non-specific fibrosarcoma-like pattern. Rare cases, part.icularly in older patients, may show focally notable pleomorphism. Tumors with round cell morphology quite commonly CTTR 1.09~ Cancer Seminar Soft Tissue Tumors 27 show eosinophilic rhabdoid cytoplasmic inclusions. Al!hough !here is no convincing prognostic difference between monophasic and bipha~ic variants, lesions with poorly 4 10 differentiated morphology are associated with a notably worse outcome. • Very rare cases of synovial sarcoma have an almost monophasic epithelial pattern, but insufficient cases have been described to allow any meaningful clinical correlation. A common finding in all types of synovial sarcoma is small foci of stromal (or sometimes intraglandular) calcification or ossification. Such changes may be extensive and then are believed to 11 12 confer an improved prognosis. ' Immunohistochemically,6 in addition to positive staining in !he obviously epi!helial component, !he spindle cell element also shows (in almost all cases) at least focal positivity for EMA (epithelial membrane antigen) and keratin and this, in combination with !he morphologic clues, is generally the best way to distinguish monophasic lesions fiom a malignant peripheral nerve sheath tumor (MPNST) or fibrosarcoma. Generally speaking, EMA staining is more sensitive and stains more cells than keratin.13 Around 30% of cases of synovial sarcoma are S-100 protein-positive,13 which may be potentially misleading, although EMA staining in MPNST is infrequent. Similarly almost two thirds of synovial sarcomas stain positively for CD99,14 rendering such antibodies essentially useless in !he distinction from Ewing's sarcoma Recent data suggest !hat cytokeratin subtyping may be helpful in !he distinction between cases of synovial sarcoma and MPNST which show morphologic overlap, since cytokeratins 7 and 19 are more frequently positive in the former. 15 It should also be noted that most synovial sarcomas are bcl-2 positive,6 although (in this author's experience) this has limited discriminant value. Ultrastructurally, both biphasic and monophasic lesions usually show clear evidence of epithelial differentiation, marked in the latter by the presence of desmosome-like junctions and microvilli. Cytogenetically, it ba~ been clearly demonstrated that bo!h biphasic and monophasic forms (as well as poorly differentiated lesions) share a reproducible tumor-specific chromosome translocation, t(X;I8)(pll.2:qll.2),16 which results in !he production of two principal 17 fusion genes, SYT-SSX I and SYT-SSX2. ·'8 This has proved of interest, not only because biphasic lesions show only the SYT-SSXJ fusion gene (whereas monophasic cases show cither)19 but, more importantly cases showing SYT-SSXI fusion (irrespective of histotype) 9 20 have a poorer prognosis.' · This seems to be associated with increased proliferative activity in tumor cells,20 in accord with the fact that the proliferative (MlB-1) index is also prognostic in synovial sarcoma.21 References: I. Wright PH, Sim FH. Soule EH et al. Synovial sarcoma. J Bone Joint Surg 1982; 64A:II2-122. 2. Schmidt D, Thun P, Med C et al. Synovial sarcoma in children and adolescems. A repon from the Kiel Pediatric Tumor Registry. Cancer 1991 ; 67:1667-1672. 3. Bergh P, Meis-Kindblom JM. Gherlinzoni F et al. Synovial sarcoma: £dentification of low and high risk groups. Cancer 1999; 85:2596-2607. 4. Skytting B. Bauer HCF, Perfekt R et al. Clinical course in synovial sarcoma. A Scandinavian Sarcoma Group study of I 04 patients. Acta Orthop Scand 1999; 70:536-542. 5. Fisher C. Synovial sarcoma Ann Diagn Patho/1998; 2:401-421. crfR 109°' Cancer Seminnr Soft Tl.s.we Tumors 28 6. Witkin GB, Miettinen M, Rosai J. A biphasic tumor of the mediastLoum with featw·es of synovial sarcoma. A report of four cases. Am J Surg ?athol 1989; 13:490-499. 7. Zeren H, Moran CA, Suster S, Fishback NF, Koss MN. Primary pulmonary sarcomas with features of monophasic synovial sarcoma: A clinicopathological, immunohistochemical and ultrastructural study of 25 cases. Hum ?athol 1995; 26:474-480. 8. Gaertner E, Zeren EH, Fleming MV, Colby TV, Travis WD. Biphasic synovial sarcomas arising in the pleural cavity. A clinicopathologic study of five cases. Am J Surg Patho/1996; 20:36-45. 9. F.olpe AL, Schmidt RA, Chapman D, Gown AM. Poorly differentiated synovial sarcoma: irnmunohistocnemical distinction from primitive neuroectodermal tumors and high grade malignant peripheral nerve sheath tumors. Am J Surg Pathol I 998; 22:673-682. 10. Van de Rijn M, Barr FG, Xiong QB eta!. Poorly differentiated synovial sarcoma: An analysis of clinical, pathologic and molecular genetic features. Am J Surg Patho/ 1999 ; 23:106-112. 11. Varela Duran J, Enzinger FM. Calcifying synovial sarcoma. Cancer 1982; 50:345-352. 12. Milchgmb S, Ghandur-Mnaymneh L, Dorfman HD et al. Synovial sarcoma with extensive osteoid and bone formation. Am J Surg Pathol1993; 17:357-363. 13. Guillou L, Wadden C, Kraus MD, Dei Tos AP, Fletcher CDM. S-100 protein reactivity in synovial sarcomas - a potentially frequent diagnostic pitfall. Immunohistochemical analysis of 100 cases. Appllmmunohistochem 1996; 4:167- 175. 14. Dei Tos AP, Wadden C, Calonje E et a!. Immunohistochemical demonstration of glycoprotein p30/32M1c2 (CD99) in synovial sarcoma. A potential cause of diagnostic confusion. Appllmmunohistochem 1995; 3:168-173. 15. Smith TA, Machen SK, Fisher C, Goldblum JR. Usefulness of cytokeratin subsets for distinguishing monophasic synovial sarcoma from malignant peripheral nerve Sheath tumor. Am J Clin Parhol I 999; 112:641-648. 16. Dal Cin P, Rao U, Jani-Sait Setal. Chromosomes in the diagnosis of soft tissue tumors. l Synovial sarcoma. Mod Patholl992; 5:357-362. 17. Clark J, Rocques PJ, Crew AJ et a!. Identification of novel genes, SYT and SSX, involved in the t(X; l8)(p ll.2;q 11.2) translocation found in human synovial sarcoma Nm Genet 1994; 7:502-508. 18: Crew AJ , Clark J, Fisher C eta!. Fusion of SYT to two genes, SSXl and SSX2. encoding proteins with homology to the Kruppel-associared box in synovial sarcoma EMBO J 1995; 14:2333-2340. 19. Kawai A, Woodruff J, Healey JH eta!. SYT-SSX gene fusion as a determinant of morphology and prognosis in synovial sarcoma N Eng/ J Med 1998; 338:153-160. 20. Nilsson G, Skytting B, Xie Yet al. The SYT-SSXI variant of synovial sarcoma is associated with a high rate of tumor cell proli feration and poor clinical outcome. Cancer Res 1999; 59:3180-3184. CrT!~ 109'" Cancer Seminar Soft Tissue Tumors 29 21. Skytting B, Bauer HC, Perfekt R, Nilsson G, Larsson 0. Ki-67 is stwngly prognostic in synovial sarcoma: analysis based on 86 patients from the Scandinavian Sarcoma Group register. Br J Cancer 1999; 80:1809-1814. CTTR I 09'b Cancer Seminar Soft Tissue Tumors 30 Case #10 Histor v- Ace. #27731 Following complaints of fullness in the rectal area noted for several months, a 55- year-old Taiwanese woman was found to have a mass in the pelvis, which appeared to be fixed to the rectum, vagina and levator muscles. The reseeted specimen consisted of a 5.0 em diameter well-circumscribed, soft nodule pushing into the wall of the rectum and attached to the posterior wall of the vagina. (Contributed by Wilmier Talbert, M.D.) Diagnosis: Gastrointestinal Stromal Tumour Neoplasms arising from the stromal (or mural) components of the gut can be divided broadly in two categories. Some tumours are nor unique to the gastrointestinal wall and appear similar to their counterparts in other locations. This category includes schwannomas, usual leiomyomas (mostly located in the oesophagus and rectum) and leiomyosarcomas, showing characteristic morphological, immunohistochemical and ultrastructural features of smooth muscle differentiation, as well as some uncommon neoplasms such as lipomatous and vascular tumours. The other category is composed of spindle and epithelioid neoplasms histologically resembling smooth muscle tumours but either lacking (or presenting only limited) immunohistochemical and ultrastructural features of myogenic, neural or neuronal differentiation and for which prediction of behaviour has proved to be problematic. This latter group, which represent the substantial majority, are known as GI stromal tumours. Stromal tumours of the gastrointestinal tract (GIST) occur over a wide age range but affect predominantly middle aged and elderly individuals, with a slight female predominance. The biological behaviour of GISTs is difl1cuil to determine accurately from data available in the literature, given the different morphological and diagnostic criteria used, and the tendency for late metastases in some cases. with spread sometimes occurring after 20 to 30 years. The overall 5 and 10-year survivals of malignant GISTs, have been estimated at between 25 and 50%, although in one series, only 10% of patients remained free of disease after a median follow-up of 68 months.' Most such patients succumb to disseminated intraabdominal disease (with metastasis to the omenrum, mesentery, peritoneum and liver), although distant metastases may occur very rarely. GTSTs were originally thought to arise from mural smooth muscle of the GT tract, based on their histological resemblance to leiomyomas and leiomyosarcomas at other sites, 23 as well as their usually intimate association with the wall of the gm. • Epithelioid tumours, popularised by Stout,' were also considered to be smooth muscle neoplasms on the basis of the transition between spindly and epithelioid areas in some lesions. Subsequently, some authors noticed morphological differences between these CITR I09" Cancer Seminar Soft Tissue Ttwwrs 'leiomyomas' and 'leiomyosarcomas' of the gastrointestinal tract and their counterpartS in other sires: most gastrointestinal lesions appeared more cellular and the tumour cells had more elongated nuclei and less brightly eosinophilic, rather fibrillary cytoplasm. It also rapidly became clear that GISTs, unlike their apparent counterparts in other sites, could rnetastasise despite the absence of usual histological features of malignancy (in particular, in the absence of significant mitotic activity) and that their behaviour was much harder to predict.2.3.5 The L970s and the early 1980s saw the debate concerning these lesions focusing on the ultrastructural features of GI mesenchymal tumours. Most studies failed to identify myofilaments with focal densities in the majority of cases; in fact, smooth muscle differentiation in gastrointestinal stromal tumours was most often supported only by relatively non-specific features (su_ch as the presence of pinocytic vesicles or focal basal lamina). In addition, several authors started to identify Schwannian or neuroaxonal characteristics in some cases microscopically indistinguishable form other GISTs, and in 1984, a distinctive subset of tumours, showing features of autonomic neural differentiation was first described by Herrera et al.6 These were initially known as plexosarcoma, later as gastrointestinal autonomic nerve tumour (GANT)6a and are now being subsumed once more into the general GIST category (see below). Numerous series of GIST have been analysed imrnunohistochemically with 7 10 inconsistent resuhs. ' Although a large proportion (between 30% to 80%)of GISTs have been shown to express muscle markers, the most specific of those, desmin, has usually stained only a minority of tumours. Variable proportions of tumours with a neural phenotype have been identified (from none to approximately 40%) and, interestingly, divergent differentiation (coexpression of muscular and neural markers) was identified in up to 20% of cases by Newman et al.9 Up to 41% of tumours have been characterised by a ·null' or 'uncommitted' phenotype, being stained by vimentin only. In the early 1990s, CD34, initially identified as a myeloid cell progenitor antigen., but also expressed in endothelial cells, in some mesenchymal cells as well as in a variety of soft tissue neoplasms, was shown to stain up to 80% of GISTs, with or without markers of other specific differentiation. Several authors tried to correlate the immunophenotype of GISTs with tumour location. the histological appearance and, more importantly, with prognosis. No significant differences bave been demonstrated in terms of immunophenotype between epithelioid and spindle cell lesions. Regarding the relationship with tumour site, in contrast to most gastric and intestinal tumours, oesophageal and rectal tumours frequently fail to stain for CD34 and tend to express desrnin, indicating that most of these tumours likely represent 'true' smooth muscle tumours that should be differentiated from GISTs, which more rarely occur in these locations. Some differences in immunophenotype between gastric and intestinal tumours, the latter tending to more commonly express a neural phenotype. have also been suggested but have not been further investi g11ted. ClfR 109"' Cancer Seminar Soft Tis.rut Tumors 32 HiStologic prediction of behaviour The difficulty in dassifying GISTs into benign and malignant categories has been recognised since the description of 'smooth muscle tumours' of the gut, by Golden and Stout in 1941, who noted that tumours showing usual histological criteria of malignancy did not consistently behave a~gressively, while occasional well differentiated low-grade lesions gave rise to metastases. In a series of 87 GISTs, Kempson and Ranchbd identified the mitotic count as the most useful indicator of malignancy. However, while the presence of 5 or more mito.ses per 10 HPF was closely correlated with aggressive behaviour, 40% of 'leiomyosarcomas' had fewer mitotic figures.3 ln order to refine the separation of benign and malignant GISTs, a profusion of studies subsequently analyzed the correlation between malignancy and various clinical and pathological parameters, often with variable 11 14 results. ' • The mitotic count has been most widely accepted as the best prognostic indicator'· 15 and it has been shown that, among clinically malignant tumours, a high mitotic 15 CQimt was associated with a shorter disease free interval and shortened overall survivaL Various cutoff levels, separating GISTs into benign and malignant categories, sometimes including a 'borderline category' , or into low and high-grade subsets have been proposed. However, because of the overlap in terms of mitotic activity between clinically benign and malignant GISTs, and in view of the rare occurrence of metastasis in histologically bland, mitotically inactive tumours, none of these has proven entirely reliable in th.e management of individual patients. Tumour size has also been shown to be strongly correlated with the occurrence of metastases. In Appelman's series of 127 cases, only one tumour smaller than 6 em 5 metastasised . These results have been confirmed by most studies and, again, various cutoff levels have been proposed, usually being set at around 5 and 6 em. Unfortunately, as for the mitotic count, some exceptions have been encountered, and, tumours as small as 2 em have been reported to metastasise.15 ln relation to tumour size, it is interesting to note that tumours found incidentally during operation performed for another unrelated conditron usually carry an ex.cellent prognosis; in Cooper et al's study, none of 19 incidentally discovered tumours resulted in the patient's death.11 In order to refine this discrimination between malignant and benign lesions, numerous other clinical, macroscopic and histological parameters have been assessed, most of which have been shown, at least in some univariate studies, to have some correlation with survival or the development of metastases. Cellularity has been considered useful by several authors. but this is extremely subjective and its interpretation is complicated by the variability between areas of the same tumour. Although the presence of unequiv.ocal tumour cell necrosis is usually regarded as higbiy suspicious for malignancy, this has been reported in rare clinically benign cases.16 Ulceration of the overlying mucosa has also 16 been considered as a worrisome feature by some authors. -' 8 Aside from oesophageal and colorectal tumours, which usually display fully developed features of smooth muscle differentiation (and therefore can be excluded, at ClTR I09" ' Cancer Seminar Soft Tisst.e Tumors 33 least conceptually, from the GIST spectrum), significant differences in terms of outcome according to tumour site have only been observed in one study, in which the ten year survival reached 74% for gastric lesions, while it was·only 17% for small bowel tumours.19 Interestingly, some studies have shown a better prognosis in rare pediatric cases and in young adults, with long survival despite metastatic disease. 20 In fact, most of these patients appear to be affected by Carney's triad. ln this syndrome, of which the genetic basis is still unclear, patients tend to develop gastric epithelioid 'leiomyosarcomas', functioning extraadrenal paragangfioma and pulmonary chondromatous hamartomas (which are sometimes clinically and radiologically misinterpreted as metastases from the gastrointestinal tumours). These tumours usually appear at a relatively young age and prolonged survival (more than 20 years) is commonly observed in the presence of metastases, even without surgical trean;nent. Because of this imperfect separation between benign and malignant GISTs using conventional pathological criteria, as well as the subjectivity and/or interobserver variability in the evaluation of some of these parameters, ancillary techniques such as the evaluation of ploidy (by DNA flow cytometry or computerised image analysis) and proliferation markers were introduced with enthusiasm in the early 1990s. Most studies have suggested that DNA ploidy, determined by flow cytometry, was significantly correlated (by univariate analysis) with histological grading and that aneuploidy was 11 14 associated with decreased Sl,lrvival. ' However, most of these authors compared tumour ploidy with malignancy defined either clinically but with a limited follow-up or defined only by histological criteria. Moreover, most of these studies included aneuploid cases that did not show clinical evidence of malignancy and, more importantly, a few patients with diploid tumours (even of small size), developed disseminated disease. Because of this overlap, ploidy does not appear to be more discriminant than morphology when applied in individual cases. Similar problems have occurred with studies utilising proliferation markers such as PCN A, Mffi-1 and AgNORs. Recent developments C-Kit is a proto-oncogene encoding a transmembrane tyrosine-kinase receptor. KIT or CD! I 7. The interaction of this receptor with its ligand, the stem cell factor (SCF) has been shown to play an important role in the development of me)anocytes, germ cells, mast cells and the interstitial cells of Cajal (ICCs). The latter, which are located between the muscular layers of the ga~tric. and intestinal wall in. association with the myenteric plexus, are known to regulate the autonomous contraction of the gastrointestinal tract and, ultrastructurally, show features of both incomplete myogenic and neural differentiatiou. These cells are characterised immunohistochemically by dual immunopositivity for CD34 and CDl 17. Recently, the hypothesis that GISTs might differentiate towards an ICC phenot¥Re· has been supported by immunohistochemical, ultrastructural and molecular studies· '23 and it has been demonstrated that immunoreactivity for CD 117 is seen in 85 to I 00% of GISTs whi le most other sarcomas or otber neoplasms are negative. In addition, activating c-kit mutations were identified in 5 of 6 cases from the original series21 and our own personal experience with larger case numbers is similar.24 C1TR 109"' Cancer Seminar Soft Tiss11e T11mors 34 These studies included spindle cell and epithelioid neopl a~ms, confirming that they repres·ent morphological variations of the same entity. CD117 positive tumours were also positive for CD34 in 72% of cases, while the (poorly specific) neural marker PGP9.5 was positive in about 70% of cases. and smooth muscle-actin was positive in 15 to 30% of cases. Interestingly, all tumours showing immunohistochemical evidence of smooth muscle differentiation i.e. desmin positivity and I or diffuse-actin positivity failed to stain for CDLI7, further justifying the separation of GISTs and conventional smooth muscle tumours in this anatomical location. These findings have led to a unifying concept, regarding GISTs as a morphologically (and immunophenotypically) heterogeneous group of tumours characterised by c-kit immunoreactivity differentiating towards an ICC phenotype.25 GANTs in our experience, are usually immunoreactive for CD117 but often negative for CD34 and, in truth, fall within the spectrum of GISTs. The recognition of GISTs as a cohesive but phenotypically heterogeneous group of tumours, defined by the expression of c-Kit, and their more objective separation from other mesenchymal neoplasms (mostly true smooth muscle tumours) will certainly allow more reproducibility between studies and could help to refine our criteria for malignancy and/or prognostic factors. 1'he expression of c-Kit also appears as a useful adjunct in the diagnosis of those GISTs arising in less usual sites, such as the mesentery, omentum or retroperitoneum, as well as in their separation from other intraabdominal neoplasms, such as desmoid fibromatosis for spindle cell lesions, and melanoma or metastatic carcinomas, which may be confused with epithelioid GISTs. The possibility .that c-Kit plays a central role in tumour developmept or progression has recently received additional support, with the identification of a germli11e mutation in a family with multiple GISTs, in the san1e domain where mutations had been found in sporadic cases. Future investigations in such patien(s might therefore provide valuable information concerning the biology of GISTs. Remarkably, novel therapies aimed at blocking the receptor tyrosine kina~e function of C· kit are already proving to be remarkably effective in GISTs which, until now have been regarded as a chemo-insensitive group of tumours. References: L Ng ER, Pollock RE, Munsell MF, Atkinson EN, Romsdalll MM. Prognostic factors influencing survival in gastrointestinal leiomyosarcomas. Ann Surg 1992; 215:68-77. :2. Golden T, Stout AP. Smooth muscle tumors of the gastrointestinal tract and retroperitoneal tissues. Surg Gynecol Obstet!941; 73:784-810. 3. Ranched M. Kempson RL. Smooth muscle tumors of the gastrointestinal tract and retroperitoneum. Cancer 1977; 39:255-62. 4. Stout AP. Bizarre smootll muscle tumors of the stomach. Cancer 1962; 15:400· 409. 5. Appelman HD. Helwig EB. CellullU" leiomyomas of the stomach in 49 patients. Arch Pathol Lab Med 1977; 101:373-377. 6. Herrera GA. Cerezo L, Jones JE et al. Malignant small bowel neoplasm of enteric plexus derivation (plexosarcoma). Light and electron microscopic study confirming the origin of the neoplasm. Dig Dis Sci 1984; 29:275-84. CITR 109'" Cancer Seminar Soft Tissue Tumors 35 6a. Lauwers GY, Erlandson RA, Casper ES, Brennan MF, Woodruff JM. Gastrointestinal autonomic nerve tumors: A clinicopathologic, immunohistochemical, and ultrastructural study of 12 cases. Am J Surg Pathol 1993; 17:887-97. 7. Mazur MT. Clark HB. Gastric stromal tumors: Reappraisal of histogenesis. Am J Surg Patholl983; 7:507-519. 8. l'ike AM, Lloyd RV, Appelman HD. Cell markers in gastrointestinal stromal tumors. Hum Pathol1988; 19:830-4. Sa. Hurlimann J, Gardiol D. Gastrointestinal stromal rumours: An immunohistochemical study of 165 cases. Histopathology 1991; 19:311-20. 9. Newman PL, Wadden C, Fletcher CDM. Gastrointestinal stromal tumours: Correlation of immunophenotype with clinicopathologic features. J Pathol 1991; 164:107-17. 10. Miettinen M, Virolaineo M, Saarlomo-Rikala M. Gastrointestinal stromal tumors: Value of CD34 antigen in their identification and separation from true leiomyomas and schwannomas. Am J Surg Pathol1995; 19:207-16. 11. Cooper PN, Quirke P. Hardy GJ, Dixon MF. A flow cytometric, clinical, and histological study of stromal neoplasms of the gastrointestinal tract. Am J Surg Pathol 1992; 16:163 -70. 12. Cunningham RE, Federspiel BH, McCarthy WF, Sobin LH, O'Leary TJ. Predicting prognosis of gastrointe.stinal smooth muscle tumors: Role of clinical and histologic evaluation; flow cytometry, and image cytometry. Am J Surg Pathol 1993; 17:588-594. n . Franquemom DW. Differentiation and risk assessment of gastrointestinal sttomal tumors. AmJ Clin PathoL 1995; 103:41-7. 14. Rudolph P et al. Immunophenotype, proliferation, DNA ploidy, and biological behavior of gastrointestinal stromal tumors: A multivariate clinicopathologic study. Hum Patholl998; 29:791-800. 15. Evans HL. Smooth muscle tumors of the gastrointestinal tract: A study of 56 cases followed for a minimum of lOyears. Cancer 1985; 56:2242-50. 16. Brainard JA, Goldblum JL. Stromal tumors of the jejunum and ileum: a clinicopathologic study of 39 cases. Am J Surg Patho/1997; 21:407-416. 17. Tworek JA, Appelman HD, Singleton HP, Greenson JK. S~omal tumors of the jejunum and ileum. Mod PathoL 1997; 10:200-9. 18. Goldblum JR, Appelman HD . . Stromal tumors of the duodenum. A histologic and immunohistochemical study of 20 cases. Am J Surg Pathol 1995; 19:71-80. 19. Ueyama T et aL A clinicopathologic and immunohistochemical study of gastrointestinal strorrial tumors. Cancer 1992; 69:947-9_55. 20. Persson S, Kindblom LG. Angervall L, Tisell LE. Metastasizing gastric epithelioid leiomyosarcomas (leiomyoblastomas) in young individuals with long-term survival. Cancer 1992; 70:721-32. 21. Hirota S, lsozaki K, Moriyama Y et a!. Gain-of-function mutations of c-Kit in human gastrointestinal stromal tumors. Science 1998; 279:577~80. 22. Kindblom LG, Remotti HE, Aldenborg F, Meis-Kindblom JM. Gastrointestinal pacemaker cell tumor (GlPACT) : Gastrointestinal stromal tumors sbow CTTR I 09'" Cancer Seminar Soft Tissue Tumors 36 phenotypic characteristics of the interstitial cells of Cajal. Am J Pathol 1998: 152: 1259-69. 23. Sarlomo-Rikala M, Kovatich AJ, Barusevicius A, Mieninenen M. CDI77: A sensitive marker for gastroinrestinal stromal rumors that is more specific than CD34. Mod Patlto/1998; 11:728-34. 24. Lux ML et al. KIT extracellular and kinase domain murations in gastrointestinal stromal tumor;s. Am J Pathol2000; 156:791-795. 25. Rubin BP, Fletcher JA, fletcher CDM. Molecular insights into the histogenesis and pathogenesis of gastrointestinal stromal tumors. lnt J Surg Pathol 2000; 8:5- LO. 26. Nishida T, Hirota S, Tanigushi M. Familial gastrointestinal stromal tumours with gerrnline mutation of the KIT gene. Nat Genet 1998; 19:323-4. CTTR I09'" Olnccr Seminar Soft Tissue Tumors 3? Case #11 History - Ace. #26932 During work-up for gastroenteritis and indigestion, CT scan revealed a large retroperitoneal mass in this 77-year-old Caucasian female. The II x 9 x 7.5 em specimen had a lobular, focally glistening, gray-tan, solid cut surface. (Contributed by K. Greg Peterson, M.D.) Diagnosis: Spindle Cell and Epithelioid Neoplasm most consistent with Malignant Peripheral Nerve Sheath Tumor The appearances in this case <)re rather unusual, in particular the non-distinctive cytomorphology and relatively abrupt transition between epithelioid and spindle cell areas - hence the somewhat descriptive diagnosis above. Malignant peripheral nerve sheath tumor (MPNST) is the preferred term for sarcomas of nerve sheath type because of the heterogeneity of their constituent cells and inconsistent or poorly reproducible relationship to conventional benign schwannoma ·or neurofibroma. While it was believed formerly that to make a diagnosis of MPNST one must demonstrate either (l) origin from a nerve or pre-existing benign nerve sheath tumor, (2) ultrastructural evidence of Schwann cell differentiation, or (3) development of a spindle cell sarcoma in a patient with neurofibromatosis (NF- I ), it has become apparent over time that many cases of MPNST share reproducible morphologic features (see below) which can be sufficient to enable their recognition, especially if supported immunohistocbemically, in the absence of the criteria listed above.' MPNST arises in two principal forms, either sporadically or, in 30-50% of patients, 2 in association with stigmata or a family history of NF- 1. .5 The lifetime risk of a patient with NF-1 developing an MPNST is estimated to be 2%.6 The age range in MPNST is broad bur the majority of cases arise in adulthood; in sporadic- cases the sex incidence is equal and peak onset is in the fifth decade of life, while in NP-1 patients the peak is 10-15 years younger and there is a male predominance. Up to 10% occur in childhood. Anatomic distribution is wide, with the limbs being involved more commonly than the trunk (including retroperitoneum) and origin in the head imd neck region being least frequent. Tumors associated with NF-1 more often show convincing evidence of origin from a nerve or preceding neurofibroma. Rare cases arise in a benign scbwannoma and others may arise in a ganglioneuroma, either de novo or following irradiation of a neuroblastic tumor. Up ro I 0% cases of MPNST appear to be radiation-induccd/·8 with a post-treatment latency period generally exceeding 10 years. Overall 5-year survival probability in MPNST is CTIR J091h Cancer Seminar Sofr Tissue T11mors 38 around 50% in sporadic cases, 20-25% in NF-I patients and no more than 10-15% in postirradiation cases. Most patients succumb.to pulmonary metastases. Unless associated with significant neurologic symptoms (which is rare), most MPNSTs are greater than I 0 em in maximum ·diameter by the time of presentation, refleeting their generally deep location. Those arising from a major nerve commonly sbow fusiform swelling of the adjacent nerve and those arising in .a pre-existing benign tumor may show a zoned appearance macroscopically, although often this is not evident. Histologically the majority of cases have a spindle celled fascicular appearance; distinctive features that may suggest neural differentiation are the abrupt alternation between cellular and more myxoid areas and the apparent perivascular accentuation or whorling of tumor cells, which sometimes extend directly into vessel walls and lead to thrombosis. Other cases have a uniformly cellular and fascicular pattern throughout and may be almost indistinguishable from monopha~ic synovial sarcoma or fibrosarcoma in H&E stained sections. Up to 10% of cases have a copious myxoid stroma (myxoid MPNST). Helpful diagnostic clues at the cellular level are the presence of pale, poorly defined cytoplasm and nuclei with a narrow, tapering outline, often with a wavy or buckled eonfiguration in some areas; the latter feature, however, is not as common as in neurofibromas. Nucle.i tend to be hype·rchromaiic and at least focally pleomorphic with inconspicuous nucleoli Nuclear palisading is infrequent (and. in fact, is probably no more common than in leiomyosarcoma or synovial sarcoma) and Meissnerian differentiation is rare. Mitoses are generally easy to find except in those infrequent examples of cellular or mildly atypical neurofibroma, usually in a patient with NF-1, in which the distinction be.tween benignity and malignancy rests entirely on the finding of any mitoses, however scarce. Up to 5% of MPNSTs are believed to be perineurial in type.9 Rare cases of MPNST are indistinguishable from so-called pleomorphic MFH on H&E staining and require .ancillary techniques (usually electron microscopy) for their accurate diagnosis. Some cases of MPNST in childhood are remarkably plexiform. A distinctive feature in approximately 10-15% of MPNSTs, especially in those 0 11 arising in NF-I patients, is the presence of heterologous differentiation.' ' This most commonly takes the form of a rhabdomyosarcomatous component (the so-called malignant Triton tumor) which is associated with an especially poor prognosis. Also frequently found are areas of osteosarcomatous or chondrosarcomatotis differentiation, sho,wing cytologic features of malignancy to a varying degree. By tradition these tumors are not termed malignant mesenchymoma. Very rare cases show epithelial, most often glandular, differentiation which, on occasion, may require distinction from ~ynovial sarcoma. lmmunohistochemically, only around 50% of case-s of MPNST are S- 100 protein-positive and this seems to reflect the fact that, at the ultrastructural level, no more than about 50% of cases show extensive, well-developed Schwannian ·differentiation, while the other cases show either fibroblastic, perineurial or mixed features to a varying 12 13 degree. ' Even in positive cases, the proportion of S-1 DO-reactive cells is often less than 20-30%. The rare cases showing perineurial differentiation exhibit EMA positivity and characteristic features at the ultrastructural level. Other putative neural markers, such as C'ITR 109'" Cancec Seminar Soft Tissqe Tumors 39 Leu-7 and myelin basic protein, have proved to be insufficiently reliable or sensitive for general use in the diagnosis of MPNST, although personal experience suggests that GFAP can sometimes be a valuable 'second Hne' marker in MPNST. 14 16 Epithelioid MPNST - accounts for approximately 5% of MPNSTs and is clinically non-distinctive but for the fact that up to 50% of cases arise in the deep dermis or subcutis, at which superficial location the prognosis seems to be greatly improved.16 Histologically, epithelioid MPNST is most often. multinodular and is composed, to a variable extent, of nests and cords of round, eosinophilic or amphophilic epithelioid cells with ve.~icular nuclei and a prominent nucleolus, somewhat resembling amelanotic malignant melanoma. The presence of spindle cell areas typical of usual MPNST and the immunohistochemical negativity for melanoma antigens (such as HMB-45) combined with electron microscopy, when necessary, allows distinction from metastatic melanoma. In contrast to ordinary MPNST, the proportion of S-1 DO-positive cells is often much higher. Differentiation from a carcinoma (primary or metastatic) is readily achieved by immunohistochemistry. References: I. Fletcher CDM. Malignant peripheral nerve sheath tumours. In: Harms D. Schmidt D (eds) Current topics in pathology: soft tissue tumours. Vol 89. Springer-Verlag, Heidelberg, p. 333-354, 1994. 2. Ghosh BC, Ghosh L, Huvos AG et al. Malignant schwannoma. A clinicopathologic study. Cancer 1973; 31:184-190. 3. Gucdon JG, Enzinger FM. Malignant schwannoma associated with von Recklinghausen's neurofibromatosis. Virchows Arch [A] 1979; 383:43-57. 4. Ducatman BS, Scheithauer BW, Piepgras DG et al. Malignant peripheral nerve sheath tumors. A clinicopathologic study of 120 cases. Cancer 1986; 57:2006- 2021. 5. Hruban RH, Shiu MH, Senie RT et a!. Malignant peripheral nerve sheath tumors of the buttock and lower extremiiy. Cancer 1990; 66:1253-1265. 6. Sorensen SA, Mulvihill JJ, Nielsen A. Long-term follow-up of von Recklinghauseo neurofibromatosis. Survival and malignant neoplasms. N Engl J Med 1986; 314:1010-1015. 7. Foley KM, Woodruff JM, Ellis FT et al. Radiation-induced malignant and atypical peripheral nerve sheath tumors. Aim Neuro/1980; 7:311-318. 8. Ducatman BS, Scheithauer BW. Postirradiation neurofibrosarcoma. Cancer 1983; 51: 1028-1033. 9. Hirose T, Scheithauer BW, Saio T. Perineurial malignant peripheral nerve sheath tumor (MPNST). A clinicopathologic, immunohistochemical and ultrastructural study of seven ca~es. Am J Surg Pathol 1998; 22:1368-1378. 10. Ducatman BS, Scheitbauer BW. Malignant peripheral nerve sheath tumors showing divergent differentiation. Cancer 1984; 54: 1049-1057. 11. Woodruff JM, Perino G. Non"germ cell or teratomatous malignant tumors showing additional rhabdomyoblastic differentiation, with emphasis on the malignant Triton tumor. Semin Diagn Pathol1994; 11:69-81. ClTR I09'" Cancer Seminar Soft Tissue Tumors 40 12. Fisher C, Carter RL, Ramachandra S et al. Peripheral nerve sheath differentiation in malignant soft tissue tumours: an ultrastructural and immunohistochemical study. Histopathology 1992; 20:11S-12S. 13. Hirose T, Hasegawa T, Kudo E et al. Malignllnl peripheral nerve sheath tumors: an immunohistochemical study in relation to ultrastructural features. Hum Pathol 1992; 23:86S-870. 14. Lodding P, Kindblom L-G, Angervall L. Epithelioid malignant schwannoma. A study of 14 cases. VirchowsArch {A] 1986; 409:433-4S I. IS. DiCarlo EF, Woodruff JM, Bansal Metal. The purely epithelioid malignant peripheral nerve sheath tumor. Am J Surg Pathol 1986; I 0:478-490. 1.6. Laskin WB, Weiss SW, Bratthauer GL. Epithelioid variant of malignant peripheral nerve sheath tumor (malignant epithelioid schwannoma). Am .I Surg Palhol 1991 ; IS: 1136-1145. CITR 109" Cancer Seminar Soft Tissue Tumo rs 41 Case#12 History- Ace. #28747 During surgery for removal of total knee hardware, this 79-year-old male underwent resection of a large mass associated with the synovium of the knee. The 780 gram. 22.0 x 9.0 x 2.0 em specimen was composed of yellow-tan, mucoid material separated by pink-tan, fibrous septa. (Contributed by Mia Perez, M.D. ) Diagnosis: Diffuse-Type Giant Cell Tumour (So-Called Pigmented Villonodular Tenosynovitis) Diffuse-type giant cell tumor, •·l sometimes also known as pigmented villonodular tenosynovitis, is much less common than the localised type and usually arises either closely adjacent to, or within, a large joint, especially the knee. These lesions present mainly in young adu ll~ (with a slight female predominance) as an often painful, slowly growing swelling associated with joint effusions. Radiologically there may be erosion and invasion of adjacent bone. Around 10-15% of cases develop at a completely extra a.tticular. extra-ynovial location, most usually in the limbs, notably the thigh/knee region, feet and hands. Un less wide excision is achieved then local recurrence is common (developing in at least 30% of cases) and may be repeated. In the literature, until receotly, there were only a very few convincing cases which had been reported to metastasise but, histolo~ically, this was not predictable in a reliable fashion . However a study from Mayo Clinic, while acknowledging the problem of prognostication, described a small series of convincingly malignant cases in which the hallmarks were cytologic monomorphism. worrisome nuclear morphology and necrosis and we have had similar experience with extra-articular lesions: Diffuse-type giant cell tumor is typically a large, soft, lobulated brownish mass with poorly defined, unencapsulated margins. Its histologic components are the same as in the more common digital lesions - but their relative proportions are different. Generally, in the diffuse lesions. cellularity is higher and multinucleate cells and xanthoma cells are less conspicuous and often onl y localised: the predominant cell type usually is the eosinophilic histiocyte-like cell, resembling Type B synoviocytes and t!lese cells often have nuclear grooves. A subset of these histiocytoid cells also often have eosinophilic cytoplasmic inclusions (and may almost mimic rhabdomyoblasts). There are generally admixed chronic innammatory cells and hyalinization is often limited. Cleft-like spaces are common and have sometimes been mistaken for either the glandular spaces of synovial sarcoma or else for large vessels: however these cleftS lack any epithelial or endothelial lining. CTTR 109''' Cancer Seminar Soft Tis.<11e T11mors 42 While CD-68 positivity is common in the hi~tiocytoid and osteoclastic tumor cells, immune in fact plays little role in this diagnosis other than through negative results which may help to exclude other diagnostic opt~ons (see below). Interestingly. however, 40-50% 4 6 of cases contain d~smin-positi ve dendritic cells ' which can cause diagnostic confusion. In the past there was some suggestion that these were reactive lesions, but their locally aggressive and occasionally metastatic behaviour, combined with the demonstration of clonal cytogenetic aberrations, most notably trisomy 5 and 7, as well as rearrangements of lpl l-13 in many cases,7 argues strongly in favour of a neoplastic process. Differential Diagnosis The biggest problem in these cellular infiltrative lesions is that, if their true nature is not recognised, then they are often mislabeled as some type of unspecified sarcoma. Distinetion from synovial sarcoma, in truth, should be simple since there is neither morpholo~c nor immune. evidence of epithelial differentiation in these lesions and the rather basophilic spindle cell sarcomatous element of synovial sarcoma is Jacking. The cytoplasmic eosinophilia might perhaps suggest a diagnosis of (epithelioid) leiomyosarcoma or some form of rhabdomyosarcoma but the cytologic monomorphism and negativity for myogenic markers (other than desmin in strikingly dendritic cells in some cases) exclude these possibilities. Some cases may bear a resemblance to angio.matoid 'MFH' but the latter tends to have usually a more obviously noc:lular growth pattern, shows greater cytoplasmic eosinophilia and is positive for desmin or HHF-35 in 40-50% of cases. References: 1. Myers BW, Masi T, Feigenbaum SL. Pigmented villonodular synovitis and tenosynovitis: a clinical epidemiologic study of 166 cases. Medicine 1980; 59:223- 238. 2. Ushijima Met al. Giant cell tumor of the tendon sheath. A study of 207 cases to compare the large joint group with the common digit group. Cancer 1986; 57:875· 884. 3. Schwartz HS, Unni KK, Pritchard DJ. Pigmented villonodular synovitis: A retrospective review of large affected joints. Clin Orthop 1989; 247:243-255. 4. de Saint Aubain Somerhausen N, Fletcher CDM. Diffuse-type giant cell tumor. Clinicopathologic and immunohistochemical analysis of 50 cases with extra articular disease. Am J Surg Patho/2000; 24:479-492. 5. Benoni F eta!. Malignant giant cell tumor of the tendon sheaths and joints (malignant pigmented villonodular synovitis). Am J Surg Pathol !997; 21:153- 163. 6. Folpe AL, Weiss SW, Fletcher COM, Gown AL. Tenosynovial giant cell tumors: evidence for a desmin-positive dendritic cell population. Mod Patholl998; 11:939-944. 7. Sciot Ret al. Analysis of 35 cases of localized and diffuse tenosynovial giant cell tumor: A report from the CHAMP Study Group. Mod Patlro/1999; 12:576-579. CTfR· I09'" Cancer Seminar Soft Tissue Tumors 43 Case#13 History -Ace. #28703 A mass had been present in the right buttock area of this 47-year-old Caucasian male for many years. The 5900 gram, 32 x 28 x 20 em gluteal tumor had a gelatinous cut surface showing regions of hemorrhage, necrosis, and calcification separated by tan fibrous bands. (Contributed by Craig Zuppan, M.D.) Diagnosis: Extraskeletal Myxoi4 Chondrosarcoma Extraskeletal rnyxoid chondrosarcoma, l-J also sometimes known as chordoid sarcoma, arises principally in the 4th to 6th decades, with a slight male predominance, as a painless, slowly growing subfascial or deep subcutaneous mass, most often situated in the lower limb. Occasional cases at other sites such as the trunk or retroperitoneum do occur, and, although often denied, it is clear that rare examples may originate in bone." Occasional cases occur in young children. s Most tumors are lobulated, well circumscribed and measure between 5 and 15 ern in maximum dimension. Traditionally these were regarded as low grade tumors prone to local recurrence but with a metastatic rate of around 3 only 10-15%. However, more recent data ·6 suggests that many cases eventually metastasize and cause the patient's death, often LO or more yeari after diagnosis, yet again proving the importance of long-term follow-up in patients with soft tissue sarcoma. Myxoid chondrosarcoma typically has a multinodular growth pattern and consists of cords or strands of small uniform cells, with little eosinophilic cytoplasm and rounded nuclei, set in a copious rnyxoid matrix composed principally of sulphated mucopol ysaccharides.7 The periphery of the lobules often appears more cellular while, more centrally, there may be cytoplasmic vacuolation and formation of chondroblastic looking lacunae. However, frank cartilaginous differentiation is evident in less than 5% of cases. Approximately 15-20% of cases contain cytoplasmic hyaline (rhabdoid) inclusions. In some cases tumor cells are more spindle shaped and may be mildly pleomorphic but hypercellularity is usually only focal. Mitoses are generally sparse and there appear to be no histologic parameters that re liably predict aggressive behavior, with the possible exception of those very rare cases that contain an obviously high grade epithelioid or round cell component.3.8 More reliable prognostic indicators are tumor size, proximal location and patient age.3 Special stains reveal copious intracytoplasmic glycofen in most cases. S-100 protein positivity is a feature of no more than 20% of cases and, in the past, has been greatly overstated. However, a potentially important (although as yet unexplained) recent finding has been the immunohistochemical demonstration of apparent neuroendocrine 10 11 differentiation in many cases. • Ultrastructural examination confirms chondroblastic C1TR t09"' Cancer Seminar Soft Ti>sue Tumors 44 differentiation and, in 30% of cases, reYeals prominent microtubular aggregates within the rougli endoplasmic reticulum. A notable finding is the consistent presence of a reciprocal chromosome translocation t(9;22) in this tumor type, 12 the gene fusion product of which has been cloned.13 Variant translocations are also recognised in a minority of cases. The principal differential diagnosis is mixed tumor of soft tissues, 14 which is generally more superficial, may show focally obvious ductal differentiation and typically is also keratin and 'actin positive, in addition to staining for S-100 protein. Additional diagnostic p0ssibilities are chordoma, which is definitionally axial in location, contains physaliferous cells and is kerati.n and EMA as well as S-100 protein positive or, less likely. myxoid liposarcoma, which, in addition to having lipoblasts, has a more ·prominent vascular network and lacks the lace-Like or corded growth pattern. Epithelioid hemangioendothelioma may have areas reminiscent of myxoid chondrosarcoma but can be distinguished by the presence of cytoplasmic vacuoles (lumina) and immunopositivity for endothelial markers. References: I. Enzinger FM, Shiraki M. Extraskeletal myxoid chondrosarcoma. An analysis of 34 cases. Hum Patho/1972; 3:421-435. 2. Dardick. I, Lagace R, Carlier MR, Jung RC. Chordoid sarcoma (extraskeletal myxoid chondrosarcoma). A light and electron microscopic study. Virch Arch A Pathol Anat 1983; 399:61-78. 3. Meis-Kindblom JM, Bergh P, Gunterberg B, Kindblom LG. Extraskeletal myxoid chondrosarcoma: A reappraisal of its morphologic spectrum and prognostic factors based on 117 cases. Am J Surg Patholl999; 23: 636-650. 4. Kilpatrick SE, Inwards CY, Fletcher CDM, Smith MA, Gitelis S. Myxoid chondrosarcoma (chordoid sarcoma) of bone: a report of two cases and review of the literature. Cancer 1997; 79:1903-1910. 5. Hachitanda Y, Tsuneyoshi M, Daimaru Y et al. Extraskeletal myxoid chondros11rcoma in young children. Cancer 1988; 61:2521 ~2526. 6. Saleh G, Evans HL, Ro JY, Ayala AG. Extraskeletal myxoid chondrosarcoma. A clinicopathologic study of ten patients with long-Jerm follow-up. Cancer 1992; 70:2827-2830. 7. Fletcher CDM, Powell G, McKee PH. Extrask.eletal rnyxoid chondrosarcoma: A histochemical and immunohistochemical study. Histopathology 1986; 10:489-499. 8. Lucas DR, Fletcher CDM, Adsay NV, Zalupski MM. High grade extrask.eletal myxoid chondrosarcoma: A high grade epithelioid malignancy. Histopathology 1999; 35:201-208. . 9. Dei Tos AP, Wadden C, Fletcher CDM. Extraskeletal myxoid chondrosarcoma: An immunohistochemical reappraisal of 39 cases. Appl lmmunohistochem l 997; 5:73-77. 10. Chhieng DC et a!. Neuroendocrine differentiation in aduh soft tissue sarcomas with features o[ extraskeletal myxoid chondrosarcoma: Report of seven cases. Mod Pathol 1998; I I :SA (Abstract). II. Oliveira AM et aL Extraskeletal myxoid chondrosarcoma: a clinicopathologic, immunohistochemical and ploidy analysis of 23 cases. Mod Pathol 2000; 13:900- 908. CTIR 109"' Cancer Seminar Soft Tissue Tumors 45 12. Sciot R, Dal Cin P, Fletcher CDM, et al. t(9;22)(q22-31;qll- 12) is a consisteDl marker of extraskeletal myxoid chondrosarcoma: Evaluation of three cases. Mod Patlwll995; 8:765-768. 13. Labelle YK et a!. Oncogenic conversion of a novel orphan nuclear receptor by chromosome translocation. Hum Malec Genet 1995; 4:2219-2226. 14. Kilpatrick SE et a!. Mixed tumors and myoepitheliomas of soft tissue: A clinicopathologic study of 19 cases with a unifying concept. Am J Surg Pathol 1997; 2 1:13-22. CITR 109" Cancer Seminar Soft Tissue Tumors 46 Case.#14 History- Ace. #28474 This 57-year-old male discovered a cyst on his right ankle. The 6.3 x 3.5 x 3.5 em resected specimen consisted of multiple fragments of gray-white, myxoid, mucinous tissue. (Contributed by Philip Robinson, M.D.) Diagnosis: High Grade Myxofibrosarcoma 1 3 Myxofibrosarcoma • is a cohesive, distinct and definable entity which shows a broad spectrum of histologic grade. The high grade, poorly differentiated end of this spectll!m resembles so-called pleomorphic "MFH" in terms of cellularity, atypia and pleomorpbism and this is bow the term "myxoid MFH" originated. In practical terms, however, this is a fibroblastic neoplasm, more easily defined in its low grade form, hence the designation of myxofibrosarcoma seems more appropriate. Myxofibrosarcoma affects mainly adults in the 6th to 8th decades, although the overall age Iange is wide. There is a slight predominance in males. The great majority of these tumors arise in the limbs; retroperitoneal origin is extremely infrequent and, in fact, this morphologic pattern in that location almost always represents part of a: dedifferentiated liposarcoma. Of particular note, about two thirds of cases develop in subcutaneous rather than deeper tissue. Clinically there is a tendency for multinodular, diffusely infiltrative growth and the true anatomical extent of these tumors, especially the superficial ones, often is underestimated. Survival correlates with histologic grade but overall 5 year survival is 60-70%. The lowest grade lesions have no capacity to metastasize but may bewme higher grade and thus acquire metastatic potential in a recurrence. The high grade lesions have !In unusual but distinct tendency to demonstrate lymph node metastaSis, in additiorlto anticipated pulmonary and osseous spread in around 30-35% of cases. The principal features shared by tumors of all grades are hypocellular areas, comaining thin-walled curvilinear vessels, in which hyperchromatic small spindle and ste1late cells with poorly defined, sometimes vacuolated cytoplasm are found. These small, bubbly vacuoles contain acid mucin rather than lipid ('pseudolipobla~.ts ' ). Most tumors have a discernibly multinodular growth pattern. Tlie cellularity and degree of pleomorphism in these lesions parallel the histologic grade, but it is the characteristic low grade areas which defme this entity. Many cases are notable for a prominent inflammatory component (chronic, acute or mixed) especially at the periphery, and some areas of the tumor may closely simulate granulation tissue. In addition some of the higher grade Jesions have rather eosinophilic fa~cicular foci (likely myo(ibrobla~tic in nature) and these tend to account for occasional actin positivity (see below). A further distinctive feature of ern~ I09., Cancer Seminar Soft Tissue Tumors 47 these lesions is to show subtle but extensive infiltration along adjacent fibrous septa, often growing several centimeters beyond .the main mass. Immunostains reveal consistent vimentin positivity and, often, focal actin positivity in higher grade lesions. The spindle and stellate cells have ultrastructural features of fibroblasts or, occasionally, myofibroblasts. In high grade lesions, areas of the rumor are indistinguishable from the non-specific pattern of pleomorphic 'MFH' and, ultrastructurally, these bizarre cells may appear undifferentiated or 'histiocyte-like'. Principal differential diagnoses are myxoid liposarcoma, intramuscular myxoma and superficial angiomyxoma. Myxoid liposarcoma does not show such a degree of nuclear atypia and pleomorphism, its vessels have a branching (trow's feet) pattern and it contains lipoblasts. Intramuscular myxoma contains very few vessels and shows no nuclear atypia. Superficial angiomyxoma does not show nuclear atypia or pleomorphism and has an epithelial component in around 20% of cases. Myxoid variants of nerve sheath or smooth muscle tumors generally are less pleomorphic and most often show immunohistochemical evidence of their line of differentiation. Low grade fibromyxoid sarcoma is more collagenous, has a swirling growth pattern and shows less nuclear atypia. References; l. Merck C, Angervall L, Kindblom L-G, Oden A Myxofibrosarcoma. A malignant soft tissue tumour of fibroblastic-histiocytic origin. A clinicopathologic and prognostic study of II 0 cases using multivariate analysis. A era Patho.l Microbial Immunol Scand A 1983; 91 (Suppl. 282): l-40. 2. Weiss SW, Enzinger FM. Myxoid variant of malignant fibrous histiocytoma. Cancer 1977; 39:1672-1685. 3. Mentzel T, Calonje E, Wadden C et a!. Myxofibrosarcoma: clinicopathologic analysis of 75 cases with emphasis on the low grade variant Am J Surg Pathol 1996: 20:391-405. CITR I09° ' Cancer Seminar So/r Tissue Tumors 48 Case#15 History - Ace. #28513 This 71-year-old female found a nodule on her back. Past history included a maste<;tomy for brea.!;l carcinoma 25 years earlier aild four previous resections of a soft tissue tumor. The 50 gram, 9.0 x 5.5 x 2.0 em resected mass was irregular, soft and rubbery. The cut surface wa~ soft, bright yellow, gelatinous with a slight whorled appearance. (eontributed by Thomas Heinz, M.D.) Diagnosis: Atypicql Lipomatous Tumour (Well Differentiated Liposarcoma) The traditional histologic classification of liposarcoma was introduced by Enzinger and Winslow in 19621 and included the category of well differentiated liposarcoma (WDLS). This entity has been the subject of considerable debate, provoked by the proposal that it shows significant site, but not morphology, related differences in clinical and biologie behavior. In particular, several studies suggested that WDLS occurring in the extremities recurred less often than its re.troperitoneal counterpart, did not metastasize and showed no propensity to develop a high grade sarcomatous component (so-called 2 5 dedifferentiated liposarcoma). · To emphasize their indolent behavior, to prevent unnecessarily radical surgery and to imply uncert~inty about their malignant designation, Evans et a1 in 19792 assigned all WDLS arising in the extremities to the category of 'atypical lipoma' - a term tentatively introduced 4 years earlier by Kindblom et al6 to denote the most highly differentiated forms of WDLS. Subsequently Evans' concept was endorsed by others. One of the basic precepts upon which the introduction of atypjcal lipoma was based was the absence of metastatic potential among these lesions, unless they dedifferentiated. 7 8 However, several more recent studies have documented both de novo dedifferentiation • (non-lipogenic sarcomatous areas and WDLS present in the primary tumor) and dedifferentiation in recurrent WDLS9 in deep-seated lin1b lesions. This has led some to recommend that the term atypical lipoma be· restricted to WDLS of the subcutaneous 9 10 fat · 1f one adopts this rationale. there would be a case. for abandoning the term altogether, given the subsequent reports of de novo dedifferentiated liposarcomas arising in subtUiltiiOOUS fat! In practice I regard these ierms as synonymous and usually sign out such cases (arising at surgically amenable locations) as 'atypical lipomatous tumour' (with a note that this is the term currenrly preferred for lesions formerly known as 'well differentiated liposarooma') since. I know of no convincing clinical-, morphologic or molecular genetic 11 differences between them. I retain the term WDLS for such lesions when they occur at CITR I09 '' Cancer Seminar Soft Tissue Tumors 49 sites in which wide excision is often not achievable (or which may necessitate very radical surgery), since tumors arising in these locations (e.g. retroperitoneum, mediastinum) may be associ Although some studies have included spindle cell and pleomorphic lipomas within the concept of atypical lipoma, these invariably benign neoplasms show sufficiently 12 3 distinctive clinicopathologic features to merit separate classification. .1 This is funher supported by their separate and distinctive karyotypes.14 Spindle cell/pleomorphic lipomas show rearrangements of 13q and 16q, whereas atypical lipomatous tumours (irrespective of depth or location) have rings or giant marker chromosomes representing amplified material from the region of 12ql3-15. The fact that this is the same region involved in the commonest rearrangements (without amplification) in benign lipoma and that the same region is amplified in dedifferentiated liposarcomas suggests the possibility of a molecular genetic continuum.' s Atypical lipoma/WDLS shows a slight predilection for males and the peak incidence is between 40 and 70 years of age. The major sites of origin are the thigh and the retroperitoneum. A significant minority arise in the inguinal region, usually in relation to the spermatic cord. Other sites, including the trunk, limb girdles, mediastinum and head/neck are affected more rarely. The commonest presenting features are those of a slowly growing, painless mass or abdominal enl argement. Macroscopically most cases consist of a poorly lobulated fatty mass which contains a variable, usually limited proportion of fmn grey-white fibrous tissue. A majority of subcutaneous example.~ are weU circumscribed whereas inter/intramuscular and retroperitoneal tumors may sometimes show macroscopic evidence of infiltration. Most (except those in subcutis) are greater than 10 em in diameter and retroperitoneal rumors may attain a very considerable size. Classically WDLS has been subdivided into 3 histologic subtypes - lipoma-l ike, sclerosing and inJ:lanunatory. Especially in the retroperitoneum, these patterns common.l y are mixed. Lipoma-like WDLS is composed of lobules of fat cells closely resembling normal mature adipocytes and, hence, simulating usual lipoma. However, WDLS characteristically shows considerable variation in cell size and shape although this may be focaL In fact, this feature may be the only clue to indicate that a lipomatous tumor should be sampled more extensively, in order to demonstrate focL containing either fat cells showing nuclear hyperchromasia, bizarre (often multinucleate) stromal cells or scattered typical multivacuolated lipoblasts. The adipocytic lobules often are traversed by variably sized fibrous trabecula in which the atypical, hyperchromatic stromal cells are most commonly found. Such bizarre cells may also be found in the wall of lesional vessels. CTTR I 09<> Cancer Seminar Soft Tissue Twm>rs 50 This lipoma-like histologic variant is the commonest type of WDLS found in subcutaneous tissue, albeit such superficial location is relatively infrequent. The sclerosing variant, which is commonest in retroperitoneum or spermatic cord, is characterized by hypocellular, often rather fibrillary collagenous fibrous tissue containing bizarre, hyperchromatic stromal cells and usually occasional lipoblasts, although the$.e may be hard to find. Admixed lipoma-like areas are quite common and again may show considerable cell size and shape variation. Occasionally, there is a superimposed, florid chronic inflammatory cell infiltrate often with germinal centers (intlammatory type WDLS), producing a close resemblance to inflammatory pseudotumor or even Castleman's disease, and this may almost 16 17 completely obscure the fatty compopent. • The principal clue. to accurate diagnosis, a5ide from identification of neoplastic fat, is the presence of atypical, usually multinucleate stromal cells, similar to those seen in sclerosing liposarcoma More recently, Dei Tos et al18 described a rare, but histologically distinctive, fourth subtype, known as spindle cell liposarcoma, which is characterized by a limited lipoma like component, admixed with hypocellular myxoid or hyaline areas and a predominant spindle cell proliferation. The latter consists of fairly uniform spindle cells showing only mild nuclear at:t.pia arranged in short fascicles or whorls. Typical lipoblasts are usually found. Reeurrence may revert to a lipoma-like panem or, more rarely, dedifferentiate. Although the original series suggested a marked predilection for subcutaneous tissue and truncal location, subsequent experience with larger case numbers has revealed a topographic distribution similar to WDLS as a whole, albeit with a somewhat increased incidence in subcutis. Differential Diagnosis Atypical lipoma!WDLS, whether deep or superficial, can easily be misinterpreted a.~ an .ordinary lipoma, often because the tumor has been poorly sampled and the most obvious clue of variat.lon in adipocyte size and shape has been overlooked. Recognition of this feature should stimulate a se!lfch for atypical stromal cells and lipoblasts but it should be borne in mind that atypia can be remarkably focal. Intramuscular lipoma is composed of unif01m adipocytes showing no nuclear hyperchromasia or pleomorphism and never contains lipoblasts. Mosr often it has a poorly defined infiltrative margin with entrapment of skeletal muscle cells. Retroperitoneal lipomas are exceedingly rare, therefore, regardless of any clinical impression of benignity, retroperitoneal adipocytic tumors should be extensively sampled, and their cytologic detail scrutinized -·since most (if not all!) will prove to be malignant. Histologically, sclerosing type WDLS, in which large bizarre stromal cells are common, may be misinterpreted as pleomorphic lipoma. However this type of WDLS is unusual in the subcutis whereas pleomorphic lipoma is invariably a subcutaneous tumor (with a predilection of the upper back/shoulder region of older males) and generally laeks a prominent collagenous stroma. Although occasional lipoglasts are found in this lipoma C1TR 109"' Cancer Seminar Safr Tisst~e Tumors 51 variant, the ratio between typical floret type giant cells to lipoblasts is high, in contrast to the low ratio in liposarcoma. Nevertheless foci of WDLS may very closely resemble prototypical pleomorphic lipoma, empha~izing the requirement in such rumors for extensive sampling and close clinicopathologic correlation. The recently recognized spindle cell variant of WDLS has the potential for misdiagnosis as spindle cell lipoma. However, although nuclear hyperchromasia and pleomorphism are commonly mild they are invariably present unlike the bland spindle cell morphology of spindle cell lipoma. Furthermore, given adequate satnpling lipoma-like WDLS foci and lipoblasts will be present Additional points of distinction are the hyaline ropey collagen bundles and stromal mast cells of spindle cell lipoma. The prognosis of atypical lipoma/WDLS is determined principally by its site of origin. Those arising in subcutaneous tissue recur locally Jess often than deeper lesions and dedifferentiation in this superficial setting is exceedingly rare. Wide excision with clear margins is the treatmem of choice for atypical lipomatous tumours in deep soft tissue of the extremities, which otherwise have at least a 30% local recurrence rate and significant potential to dedifferentiate. Retroperitoneal WDLS may be life threatening, even without dedifferentiation and metastatic sp~ad, because of its local effects and tendency to repeated, oft.en unconttolled local recurrence, emphasizing the importance of accurate diagnosis - although even then it may be difficult to excise completely. Interestingly, while dedifferentiation undoubtedly worsens the prognosis, survival is not as poor as in other histologically compamble high grade sarcomas. McCormick er al reported a recurrence rate of SO% in a series of 32 de novo dedifferentiated WDLS but a metastatic rate of only 13%8 in the first five years of follow-up, results which have been substantiated by a larger more recent series.19 References: I. Enzinger FM, Winslow DJ. Liposarcoma. A study of 103 cases. Virch Arch Parhol AtUJI 1962; 335:367-388. 2. Evans HL, Soule EH, Winkelmann RK. Atypical lipoma, atypical intramuscular lipoma and well differentiated retroperitoneal liposarcoma. A reappraisal of 30 cases formerly classified as well differentiated liposarcoma. Cancer 1979; 43:574-584. 3. Evans H.L. Liposarcoma. A study of 55 cases with a reassessment of its classification. Am J Surg Pathol1979; 3:507-523. 4. Kindblom L-G, Angervall L, Fassina AS. Atypical lipoma Acta Parlwl Microbiol Scand (A) 1982; 90:27-36. 5. Azumi N, Curtis J. Kempson RL et al. Atypical and malignant neoplasms showing lipomatous differentiation. A.m J Surg Pmhol 1987; II: 161-183. 6. Kindblom L-G, Angervall L, Svendsen P. Liposarcoma. A cl in icopathological, radiographic and prognostic study. Acta Pathol Microbiol Scand Suppl 1975:253:1- ?l. 7. Brooks JJ, Connor AM. Atypical lipoma of the exttemities and peripheral soft tissues with dedifferentiation: implications for maoagemem. Surg Pmlwl 1990: 3:169-178. CTTR 109'• Cancer Seminar Soft Ti.t."" Tumors 52 8. McCormick D, Mentzel T, Beham A, Aetcher COM. Dedifferentiated liposarcoma: clinicopathologic analysis of 32 cases suggesting a better prognostic subgroup among pleomorphic sarcomas. Am J Surg Patho/1994; 18: 1213-1223. 9. Weiss SW, Rao VK. Well differentiated liposarcoma (atypical lipoma) of deep soft tissue of the extremities, retroperitoneum and miscellaneous sites. A follow-up study of 92 cases with analysis of the incidence of dedifferentiation. Am J Surg Pathol 1992; 16: 1051-1058. 10. Weiss SW, Sobin LH. W.H.O. histological typing of sofl tissue tumors. Berlin: Springer Verlag, 1994. II. Rosai J, Akerman M, Oal Cin Petal. Morphologic-karyotypic study of 59 atypical lipomatous tumors: evaluation of their relationship and differential diagnosis with other adipose tissue tumors. Am J Surg Pathol 1996; 20: 1182-1189. 12. Enzinger FM, Harvey OA. Spindle cell lipoma. Cancer 1975; 36:1852-1859. 13. Shmookler BM, Enzinger FM. Pleomorphic lipoma: a benign tumor simulating liposarcoma. Cancer 1981; 47: 126-133. 14. Fletcher COM, Akerman M. Dal Cin Petal. Correlation between clinicopathologic features and karyotype in lipomatous tumors. A report of 178 cases from the Chromosomes and Morphology (CHAMP) collaborative study group. Am J Parhol 1996; 148:623-630. 15. Dei Tos AP et al. Coordinated expression and amplification of the MDM2, COK4 and HMGI-C genes in atypical lipomatous tumours. J Patho/2000; 190:531-536. 16. Kn1us MD, Guillou L, Aetcher COM. Well differentiated inflanumllory liposarcoma: an uncommon and easily overlooked variant of a common sarcoma. Am J Surg Patho/1997; 21:518-527. 17. Argani P et al. Lymphocyte-rich well differentiated liposarcoma: report of nine cases. Am J Surg Patho/1997; 21:884-895. 18. Dei Tos A, Mentzel T, Newman PL et al. Spindle cell liposarcoma, a hitherto unrecognized variant of liposarcoma. Analysis of six cases. Am J Surg Pathol 1994; 18:913-921. 19. Henricks WH, Chu Y-C, Goldblum JR, Weiss SW. Oedifferemiated liposarcoma: a clinicopathologic analysis of 155 cases. Am J Surg Parhol 1997; 2:271-281 . CITR 109'" Cancer Seminar Soft Tiss11e T11mors Case#l6 History - Ace. #28503 On routine physical examination, lhis 44-year-old, gravida 2, para I, Caucasian female wa.~ found lO have a 12 em mass in the right cul-de-sac. A sonogram showed a uterus with small fibroids and a cystic and solid right pelvic mass. CT scan showed a complex cystic solid pelvic mass on the right with no ascites and no evidence of metastasis. Past history included a partial thyroidectomy for papillary carcinoma. The 600 gram resected specimen consisted of a 13.0 x 10.0 x 12.0 em irregular soft gelatinous mass. with nodular areas of tan-white ~nn tissue. (Contributed by Wafa Michael, M.D.) Diagnosis: Dedifferentiated Liposarcoma The classification of liposarcoma, in line with many tumor classifications, is 1 evolving. .2 This is due largely to the application of cytogenetic analysis to solid tumors over the past J0-15 years, which, among fatty tumors, has revealed a significant number of tumor-specific karyotypic aberrations.3 The most. logical classification of liposarcoma, in biologic terms, is into three main groups: well·differentiated, characterized by ring or marker chromosomes derived from the long arm of chromosome 12, myxoid (including myxoid and round cell), characterized in most cases by a reciprocal translocation t(l2;16)(ql3;pll) and pleomorphic, which has a complex karyotype, usually with a large number of extra chromosomes. All the other morphologic subtypes can be categorized under these main subheadings. Well-differentiated liposarcoma is nowadays widely regarded as the conunonest form of this disease, accounting for more than 30% of cases. In my view, the term well differentiated liposarcoma is best regarded as synonymous with atypical lipoma (or atypical lipomatous rumor}, since both designations connote a tumor with significant risk of local recurrence (approx 30% in marginally excised lesions) but no potential for metastasis. Local recurrence is quite often delayed for some years but carries an attendant risk (perhaps 5%) of being associated with dedifferentiation and hence the acquisition of metastatic potential. Because of the significant mortality of retroperitoneal lesions due to local effects (even in the absence of dedifferentiation or metastasis"), many people have preferred to retain the term well-differentiated liposarcoma in this specific anatomic location, although it has to be admitted that using different terms for exactly the same tumor type at different locations is somewhat illogical! As discussed for Case 15, key diagnostic criteria for recognizing these well -differentiated lesions at all sites are vari ation in udipocyte size, ad ipocytic nuclcur atypia and the presence of atypical hyperchromatic stromal cells which may either be spindled or quite often multinucleate. Lipoblasts, although often present CTrR 109"' Cancer Seminar Soft Tissue Tumors 54 (even if only focally), are no longer regarded as a diagnostic prerequisite, nor do they usually jmpact on the diagnostic nomenclature used. Remember, however, that many pleomorphic lipomas (in their classical clinical setting of non-recurring shoulder/neck/head lesions in older patients, especially males) also often contain small numbers of Hpoblasts, but, importantly, their karyotype is usually entirely different from that of atypical lipom~tous tumours, thus justifying the separate classification of spindle cell/pleomorphic lipomas? Atypical lipoma (or well-differentiated liposarcoma) can be subclassified into adipoc.yiic (lipoma-like), sclerosing, inflammatory, spindle cell and dedifferentiated subtypes ($ee Discussion of Case 15). Lipoma-like lesions are by far the most common, followed by sclerosing lesions which arise most often in the retroperitoneum or paratesticulaF region. Tumors shoYJing mixed features are frequent, especially in the retroperitoneum. Spindle cell and inflammatory lesions are both fairly rare. Dedifferentiated liposarcoma is definitional! y a variant of well-differentiated liposru-coma (or atypical lipomatous tumour) which is characterised by transition to high grade non-lipogeruc tumour. 5 Such transition much more often occurs de novo (i.e. ·in a primary tumour) than it does in the recurrence of a previously excised well-differentiated tumour. When first described -by Harry Evans,5 the transition was generally regarded as occurring in an abrupt fashion but it is increasingly recognised that, io some small groportion of cases, the transition may be gradual and, in very exceptional cases, the well differentiated and dedifferentiated areas may be co-mingled. Distinction of the latter lesions from pleomorphic liposarcoma may be problematic and depends entirely on the recogrution of a prior or adjacent purely well-differentiated component. It was, however, almest universa:Uy agreed (until very recently - see below) that to fulfill Evan's originally pro(J9Sed criteria for dedifferentiation then the dedifferentiated areas should be morphologically high grade. In terms of anatomical location, dedifferentiated liposarcoma is commonest in the retroperitoneum and, in contrast to what was formerly believed, it may also occur in 6 9 somatic deep soft tissue and rare! y even in subcutaneous tissue. ' The variable incidence at these different sites mainly seems to reflect the greater size and generally longer pre operative duration of more deep-seated lesions, thus presumably enhancing the possibility of dedifferentiation taking place. Among retroperitoneal spindle cell sarcomas, dedifferentiated liposarcoma is by far the most common, accounting for > 80% of cases so careful sampling of adjacent fatty tissue (to identify the well differentiated areas) is always advisable. With regard to the development of dedifferentiation in the recurrence of a prior well-differentiated neoplasm, the overall risk appears to be around 5%10 and, again, the relative risk is increased in deeper or (especially) retroperiton~al tumours because of their higher rate(s) of local recurrence due to the difficulty in obtaining wide excision margins. Morphologically the dedifferentiated component may· show a wide variety of appearances but most commonly resembles so-called pleomorphic 'MFH', storiform fibrob,lastic 'MFH' or my?Cofibrosarcoma (myxoid 'MFH'). A smaller proportion of cases resemble fibrosarcoma or inflammatory 'MFH' and, in truth, almost any pattern may be CTTR I09'" Cancer Seminar Soft Tissue Tumors 55 encountered on occasion - for e1tample, I have seen ca~es in which the dedifferentiated component had a small round cell appearance or even a very epithelioid, trabecular carcinoma~llke appearance. A small but distinctive subset of cases is characterised by hypercellular whorled micronoduJes which may resemble a menillgotheliaJ, follicular dendritic cell or vascular neoplasm: such cases often also contain areas of metaplastic ossification. Depending on the elttent to which one makes use of immunohistochemistry, it has become clear that around 10% of dedifferentiated Liposarcomas also show alternative heterologous differentiation, most often myogenic (rbabdo or leio) and less often osteo- or 6 8 11 12 choodrosarcomatous. • • ' Such heterologous components appear to have no independent effect on outcome. More controversially, it has been proposed recently by Henricks et al7 to expand the definition of dedifferentiation to. include a variety of low grade morphologic patterns and, in my opinion, this is a reasonable plan given that there undoubtedly exist cases which are more cellular and worrisome than conventional variants of well differentiated liposarcoma but which do not fulfill traditional criteria for dedifferentiation. In Henricks' series such cases accounted for 9% of the total and, interestingly, their clinical behaviour was no different from high grade lesions. My only concern with regard to such cases is that reproducible diagnostic criteria need to be defined since it would be inappropriate for some cases of sclerosing or spindle cell liposarcoma to end up in the dedifferentiated category. Clinically it is interesting to note that the extent of dedifferentiation does not influence the outcome and that overall, the metastatic rate of such tumours (which is 25- 30%) is much less than that of other comparable pleomorphic sarcomas. Naturally many patiems ultimately succumb to the local effects of inoperable retroperitoneal disease but the general 5 year survival probability in dedifferentiated liposarcoma is at least 70%; the LO year surviva.l is undoubtedly less and 20 year survival does not exceed 10-15%, at least among retroperitoneal cases. With regard to differential diagnosis, this is not usually a problem so long as the well differentiated component is identified. This requires thorough sampling and (especially in the retroperitoneurn) macroscopic awareness of the more coarsely lobulated appearance of neoplastic fat. Aside from typical leiomyosarcomas, it is a useful rule of thumb to regard all high grade retroperitoneal sarcomas as' potential dedifferentiated liposarcomas until proved otherwise. Refere nces: I. Mentzel T, Fletcher CDM. Lipomatous tumours of soft tissue: an update. Virchows Arch 1995; 427:353-363. 2. Dei Tos AP. Liposarcoma: new entities and evolving concepts. Ann Diagn Pathol 2000; 4:252-266. 3. Fletcher CDM, Akerman M, Dal Cin P et al. Correlation between clinicopathologic features and karyotype in lipomatous tumors. A report of 178 cases from the Chromosomes and Morphology (CHAMP) collaborative study group. Ami Patholl996; 148:623-630. CTTR I 09'" Cancer Seminar Soft Tissue Tumors 56 4. Linehan DC et al. Influence of biologic factors and anatomic site in incompletely resected liposarcoma. J Clin Oncol2000; 18:1637-1643. 5. Evans HL. Liposarcoma. A study of 55 cases with a reassessment of its classification. Am J Surg Patholl979; 3:507-523. 6. McCormick D, Mentzel T, Beham L, Fletcher CDM. Dedifferentiated liposarcoma: clinicopathologic ·analysis of 32 cases suggesting a better prognostic subgroup among pleomorphic sarcomas. Am J Surg Pathol 1994; 18:1213-1233. 7. Henricks WH, Chu YC, Goldblum JR, Weiss SW. Dedifferentiated liposarcoma. A clinicopathologic analysis of l 55 cases with a proposal for an expanded definition of dedifferentiation. Am J Surg Patho/1997;. 21:271-281. 8. Hasegawa T eta!. Dedifferentiated liposarcoma of retroperitoneum and mesentery: vatied growth patterns and histologic grades - a clinicopathologic study of 32 cases. Hum Pathol2000; 31)17-727. 9. 'Yoshikawa H et a!. Dedifferentiated liposarcoma of the subcutis. Am J Surg Palho/1996; 20:1525-1530. 10. Weis_s SW, Rao VK. Well differentiated liposarcoma (atypical lipoma) of deep soft tissue of the extremities, retroperitoneum and miscellaneous sites: a follow-up study of 92 cases with analysis of the incidence of dedifferentiation. Am J Surg Pathol1992; 16: 105 1-1058. 11. Tallini G et al. Divergent myosarcomatous differentiation in retroperitoneal liposarcoma. Am J Surg Pathol1993; 17:546-556. 12. Evans HL et al. Heterologous elements in the dedifferentiated component of dedifferentiated liposarcoma. Am] Surg Patho/1994; 18:1150-1157. CTfR 109" Cancer Seminar Soft Tissue Tumors 57 Case #17 Historv- Ace. #26726 After presenting with sudden onset of severe right upper quadrant pain, thls 46- year-o1d male was found to have a palpable hard mass on the right side. A cr scan showed a large retroperitoneal mass abutting and displacing the right kidney and inferior vena cava. The 4300 gram specimen included kidney. adrenal and a 23 x 16 x 10 ern gray tan, irregularly-shaped, rubbery to focally firm mass. Sectioning revealed whorling bundles of interlacing rubbery tissue with areas of necrosis, hemorrhage and white calcification. (Contributed by Douglas Kahn, M.D.) Diagnosis: High Grade Spindle Cell Sarcoma with Osteogenic Differentiation ? Dedifferentiated Liposarcoma ? 'MFH' with Bone ? Soft Tissue Osteosarcoma This is a morphologically high grade pleomorphlc spindle cell sarcoma associated with bone formation; in large areas the osseous component appears bland ('metaplastic') but, at least focally, there is clear-cut deposition of osteoid by cytologically malignant ·cells. In tractitional terms this tumour could then be classified as soft tissue osteosarcoma or so-called 'MFH' with bone, since many authorities have accepted neoplastic bone formation in MFH in the pa~t. 1 "2 In reality, statistical probability suggests that this is a dedifferentiated liposarcoma with heterologous osteosarcomatous differentiation - but we don't have the well differentiated fatty component required to prove this - thus allowing discussion of the other two diagnostic options! Malignant fibrous histiocytoma (MFH) has been a fashionable diagnosis for more than 25 years and has been regarded by many as the commonest type of soft tissue sarcoina. Although five variants - pleomorphic, myxoid, giant cell, inflammatory and angiomatoid - have been described, it has been increasingly appreciated that the 'founder member' of the group, the pleomorphic type, is not a discrete entity and that the links between all the variants are tenuous to say the least. Certainly it is no longer believed that any of these lesions show true hlstiocytic differentiation. Although fonnerly held to be the single commonest sarcoma, pleomorphic 'MFH ' 3 5 is no longer defensible as a cohesive entity. ' Rather it represents the shared, largely anaplastic morphology of a wide range of other definable neoplasms in which a specific line of differentiation can be demonstrated in most cases. Perhaps the most important clue 10 this reassessment of pleomorphic 'MFH' came from the realization that there were no positive, reproducible criteria for iEs recognition and that it was essentially a diagnosis of exclusion. Most of the tumors which have been classified traditionally into this CITR I09' ' Cancer Seminar Soft Tissue Trmiors 58 ·wast.ebasket' category arise in deep soft tissue of adults, with an ever-increasing incidence beyond the age of 40 years. If one troubles to define the differentiation which they show, tbe most common histologic types to have been subsumed in this category are the pleomorphic variants of leiomyosarcoma, liposarcoma and rhabdomyosarcoma, as well as many cases of high grade myxofi brosarcoma, which has a relatively better prognosis (see Discussion of Case 14). In reality, on occasion, almost any sarcoma may share this morphology. It is important to note that carcinomas, melanoma and lymphoma may also show the same morphologic pattern and clearly this has major clinical and therapeutic implications. The extent to which subclassification of the pleomorphic sarcomas will be of direct clinical relevance remains to be proven, but available data already su~est, for example, that pleomorphic rhabdomyosarcomas are especially aggressive, while dedifferentiated liposarcomas are relatively indolent when compared with other 8 9 pleomorphic sarcomas. · In fact, l\ recent retrospective reclassification of 100 so-called 'MFH's has revealed that pleomorphic sarcomas showing any evidence of myogenic differentiation are twice as likely to metastasize when compared to non-myogenic lesions and they also metasta~ize earlier. 10 Only if pathologists undertake such categorization on a routine basis will prognostic and therapeutic differences emerge. From the morphologic point of view, all lesions designated as pleomorphic 'MFH' bave tended to be large masses often with areas of necrosis. Irrespective of their specific differentiation, they almost all share considerable cytologic pleomorphism, the presence of bizarre multinucleate cells, the frequent presence of a storiform architecture and the frequent presence of a collagenous stroma containing variably prominent chronic inflammatory cells, sometimes with foamy macrophages. Only by thorough sampling, combined with judicious use of immunohistochemistry and/or electron microscopy, is precise categorization achieved. Inevitably there remain a group of high-grade pleomorphic soft tissue neoplasms which cannot be classified, at least at present. Some of these are almost certainly fibroblastic or myofibroblastic in nature but a small, though significant, proportion remain enigmatic. Such cases merit continued vigorous study but the ' box' in which they wait. at the present time, to be sorted out does not warrant a specific name! Extraskeletal osteosarcoma 11 14 Soft tissue osteosarcoma ' accounts for about 5% of all osteosarcomas and affects mainly adults in the sixth to eighth decades with a slight male predominance. The majority of cases arise in the limbs, especially the thigb. but the anatomic distribution is wide. Approximately 15% of cases arise at the site of previous tissue injury or irradiation. Tbe majority of cases pursue an aggressive, metastasizing clinical course with an overall monality of 60-70% within 5 years. Whether or not these lesions respond reliably to the same chemotherapeutic regimen employed for osseous lesions in young patients has not been determined properly but, in general, such intensive treatment protocols are not so well tolerated in older adults, thus reducing the chance of an effective response. The morphologic features of soft tissue osteosarcoma are essentially the same as those in primary bone tumours. Among soft tissue lesions, the commonest pattern is 'MFH' -like and, less often, osteobla~tic. Chondroblastic osteosarcoma in soft tissues is CITR 109"' Cancer Seminar Soft Tissue Tumors 59 12 12 15 relatively infrequent, as is the telangiectatic variant. • If recognition of convincing osteoid matrix is problematic, some authors have found either alkaline phosphatase staining14 (as in conventional bone tumors) or the immunohistochemical demonstration of osteocalcin 16 to be helpful. The subgroup of giant cell-rich lesions, formerly known as giant cell 'MFH', which might mo(e logically be classified as osteosarcoma, in light of the frequent presence of malignant bone or osteoid, have simflar clinicopathologic features to the other types of osteosarcoma. References: 1. Bhagavan BS, Dorfman HD. The signifi<;ance of bone and cartilage formation in malignant fibrous histiocytomas of soft tissue. Cancer 1982; 49:480-488. 2. Enzinger FM. Malignant fibrous histiocytoma 20 years after Stout. Am J Surg Pathol 19.86; lO(Suppl 1):4 _3-53. 3. Fletcher CDM. Pleomorphic malignant fibrous histiocytoma: fact or fiction? A critical reappraisal based on 159 tumors diagnosed as pleomorphic sarcoma. Am 1 Surg Pathol 1992; 16:213-228. · 4. Hollowood K, Fletcher CDM. Malignant fibrous histiocytoma: Morphologic pattern or pathologic entity? Semin Diagn Patho/1995; 12:210-220. 5. Akerman M. Malignant fibrous histiocytoma - the commonest soft tissue-sarcoma or a nonexistent entity? Acta Orthop Scand 1997; 68(Suppl 273):41-46. 6. Gaffney E F, Dervan P A, Fletcher CDM. Pleomorphic rhabdomyosarcoma in adulthood: Analysis of 11 cases with definition of diagnostic criteria. Am J Surg Parho/1993; 17:601-609. 7. Schurch W, Begin LR, Seemayer TA et a!. Pleomorphic soft tissue myogenic sarcomas of adulthood. A reappraisal in the mid-1990s. Am J Surg Pathol1996; 20:131-147. 8. McCormick D et al. Dedifferentiated liposarcoma: Clinicopathologic analysis of 32 cases suggesting a better prognostic subgroup among the pleomorphic sarcomas. Am J Surg Patho/1994.; 18:1213-1223. 9. Henricks WH et al. Dedifferentiated liposarcoma: A clinicopathologic analysis of 155 cases with a proposal for an expanded definition of dedifferentiation. Am 1 Surg Patho/1997; 21:271-281. . 10. Fletcher CDM, Gustafson P, Rydholm A, Willen H. Akerman M. Clinicopathologic re-evaluation of 100 malignant fibrous histiocytomas: the prognostic relevance of subclassification. 2000 (Submitted for publication.) 11. Sordillo PP, Hajdu SI, Magill GB et p.L Extraosseous osteogenic sarcoma. A review of48 patients. Car~cer 1983;51:727-734. 12. Chung EB, Enzinger FM. Extraskeletal osteosarcoma. Car~cer 1987; 60:1132- 1142. . 13. Lee JS, Fetsch JF, Wasdahl DA, et al. A review of 40 patients with extraskeletal osteosarcoma. Cancer 1995; 76:2253-3259. 14. Lidang Jensen M, Schumacher B. Myhre Jensen 0 et a!. Extraskeletal osteosarcomas: a clinicopathologic study of 25 cases. Am J Surg Pathol 1998; 22:588-594. 15. Mirra JM, eta!. Extraskeletal telangiectatic osteosarcoma. Cancer 1993; 71:3014- 3019. CITR I 09"' Cancer Seminar Soft Tissue Tumors 60 J6. Fanburg-Smith JC et al. Osteocalcin and osteonectin immunoreactivity in extraskeletaJ osteosarcoma: A study of 28 cases. Hum. Pathol 1999; 30:32-38. CTTR 109'' Cancer Scminnr Soft Tissue Tumors 61 Case#18 History - Ace. #18433 For one year, this 16-year-old African American female had noticed a mass in her right hand, not associated with trauma. Physical eJtamination revealed a 5.0 Jt 6.0 em, soft, non-tender mass of the right hand interposed between the first and second metacarpals. Radiographs showed no evidence of bony involvement. The 46.5 gram specimen consisted of a roughly oval, 6.0 x 5.5 Jt 3.8 em mass. The cut surface was monied yellow tan with patcby central necrosis. (Contributed by J. H. Cremin, M.D.) Diagnosis: Alveolar Rhabdomyosarcoma Rhabdomyosarcoma falls into three main groups - embryonal. alveolar and pleomorphic - and, in general terms, the first two groups occur mainly in children, while pure pleomorphic lesions occur almost exclusively in adults. Rhabdomyosarcoma in children14 occurs principally before the age of 10 years, with a peak before the age of 4, and shows a moderate male predominance. Commonest sites are the head and neck region (including orbit and meninges), followed by the genitourinary tract, then followed by the limbs and then the trunk. As a general rule, embryonal lesions affect somewhat younger patients than the alveolar type and limb tumours are much more commonly a.lveolar than embryonal. Two other infrequent, but distinctive, clinical traits in alveolar rhabdomyosarcoma are the tendency to develop breast metastases in female patients and to present rarely in a manner resembling leukaemia. due to extensive marrow replacement, without a clearly evident primary lesion. This Iauer pattern of disease may well represent malignant transformation of bone marrow mesenchymal stem cells ("rhabdomyoblastaemia"). By vinue of the anatomically awkward si tes of predilection, most childhood examples of rhabdomyosarcoma are not amenable to complete excision hence the drive towards effective radiotherapy and chemotherapy over the past 15 years. Taking into account the IRS system of clinical grouping,24 which allows for resectability as well as extent of tumour spread, overall 5 year survival figures now approach 60-70% and almost 50% of all childhood rhabdomyosarcomas are now said to be cured, which is a remarkable advance from the position just 25 years ago. Botryoid and spindle cell variantS of embryonal rhabdomyosarcoma (see below) have an es~ially excellent prognosis. while the presence of an even focal alveolar pattern confers a diminished survival probabil ity ~see below). Relapse after treatment in paediatric rhabdomyosarcoma also represents a poor prognostic sign. By contrast, rhabdomyosarcoma in adults5 predominantly arises in the limbs irrespective of histological type, appears to be somewhat more chemoinsensitive and has a 5 year survival probability of only around 20%. In this age group the pleomorphic subtype CTTR 109"' Cancer Seminar Soft Tissue Tumors 62 significantly outnumbers embryonal and alveolar lesions, although the latter do occur. As a proportion of the whole, adult cases account for no more than 10%. Embryonal rhabdomyosarcoma, which accounts for around 60% of childhood caSes, in its classical form consists of small round or spindle-shaped, undifferentiated (basophilic) cells, admixed with variable numbers of round. strap-or-tadpole-shaped eosinophilic rhabdomyoblasts, in a myxoid stroma. Cytoplasmic cross-striations are present in no more than 20-30% of cases but, even in tumours with almost no discernible rhabdomyoblasts, desmin or muscle actin immunostains are positive in more than 95% of 6 7 cases and this is therefore the most reliable means of making the diagnosis. • Additionally, well characterised antibodies to the protein products of the nuclear transcription factors MyoD l and myogenin (myf4) are now proving to be very sensitive and Specific markers. 8 Botryoid rjlabdomyosarcoma represents simply an anatomical subset of embryonal tumours arising beneath a mucosal surface and growing in an exophytic, grape-like manner. Such tumours are characterised in 50% or more of cases by the presence of a subepithelial cambium layer consisting of condensed, relatively undifferentiated tumour cells. Other distinctive variants in the embryonal category are: spindle cell rhabdomyosarcoma,9 which arises at a variety .of sites but principally in the paratesticular region and less often in the head and neck, is characterised predominantly by amphophilic or eosinaphilic spindle-shaped cells in a tascicular or storiform pattern and carries a 5 year survival of more than 90%; anaplastic rhabdomyosarcoma, 10 which is now the preferred tenn for pleomorphic childhood cases and describes the · presence, at least in areas, of sheets of bizarre cells with giant nuclei and atypical mitoses set in a collagenous stromai such cases, which almost always have identifiable embryonal (or less often alveolar) foci, have a relatively poor survival and, finally, rhabdomyosarcoma witb cytoplasmic hyaline ('rhabdoid') inclusions, 11 which needs to be distinguished from so-called rhabdoid tumour, since its prognosis is comparable to usual rhabdomyosarcoma and is far better than that for so-called 'rhabdoid' tumour. 7 12 13 Alveolar rhabdomyosarcoma ' • is characterised generally by rather larger, more rounded undifferentiated cells, with larger, more vesicular nuclei than those in the embryonal vatiant. The,~e are admixed with variable numbers of eosinophilic. rhabdornyoblasts and multinucleate giant cells with peripheral (wreath-like) nuclei are often present. The cells usually are arranged in an alveolar pattern, such that they line or cover collagenous septa and tend to be ' shed' in a discohesive manner into the centre of these alveolar structures. Cross-striations are evident in no more than 10% of ca~es . With regard to the present case, it is worth noting that 5-l 0% of alveolar rhabdomyosarcomas are CD99 .Positive (see differential diagnosis below). Occasional cases show a remarkable deg~;ee of cytoplasmic differentiation, becoming much more obviously rhabdomyoblastic, after chemotherapy. It had long been realised that the tumours with a very obvious alveolar architecture behaved in an aggressive "manner but, importantly, it has also been 14 16 confimted more recently ' that tumours which appear to have a more solid embryonal type growth pattern, but in which .either the cells (and their nuclei) are larger than the embryonal type or in which a nascent pattern of alveolar nesting is evident either by H&E ClTR I09 ." Cancer Seminar Soft Tissue TumQrs 6'3 or reticulin staining, have a significantly diminished prognosis. Such tumours are known as the solid variant of alveolar rhabdomyosarcoma. It is increasingly realised that such lesions can be diagnosed on cytological grounds alone, based largely on nuclear size. It is worth commenting that the solid variant of alveolar rhabdomyosarcoma took more than 10 years to be accepted generally by the paediatric pathology community, due largely to resistance in various American centres. Once this 'taboo' had been broken it seems reasonably clear that the survival curves of the embryonal and alveolar subsets have spread further apart, largely because a significant subset of the apparently more aggressive embryonal lesions are now reclassified as alveolar. Alveolar cases may also, albeit less frequently, show anaplastic features or rhabdoid inclusions, as described in embryonal tumours above. In addition occasional cases may show a misleading clear cell appearance. 17 18 19 Pleomorphic rhabdomyosarcoma, " as presently defined, is a high grade pleomorphic sarcoma which is composed principally of a patternless array of large, bizarre, polygonal or spindle shaped rhabdomyoblasts, set in a variably collagenous stroma. Some tumours are very storiform and approach closely the appearance of so-called pleomorphic 'MFH'. Cross-striations are virtually never found. Childhood examples almost always show at least small foci of embryonal or (less often) alveolar tumour elsewhere and the term 'anaplastic' is preferred for such tumours, as mentioned above. There have been considerable recent advances in molecular genetic means of confirming and refining the diagnosis of rhabdomyosarcoma and, at the same time, facilitating distinction from other round cell tumours, especially in children in whom this is 16 2 2 now becoming the standard of care. • 1). J In brief summary here, from rhe cytogenetic aspect, alveolar rhabdomyosarcoma carries a reproducible tumour-specific chromosome translocation, t(2;13)(q35;q l4), which is karyotypically quite different from the deletions typically seen on the long arm of chromosome II in embryonal rhabdomyosarcoma. A small proponion of cases show a variant translocation r( I; 13}(p36;q 14). These differences may be utilised diagnostically either in classical karyotyping, RT-PCR or by fluorescence in-siiu hybridisation. Furthermore, there is evidence that the specific fusion genes generated by these trans locations may be of prognostic relevance? 4 Differential diagnosis of alveolar rhabdomyosarcoma is generally not too broad, especially in classical cases with a well-formed alveolar architecture. Principal considerations are extraskeletal Ewing's sarcoma!MPNET and sometimes poorly differentiated synovial sarcoma. Ewing's/MPNET is fairly consistently CD99 positive and desmin negative, but shows (usually dot-like) keratin staining in 10-15% of cases. It generally lacks any hint of eosinophilic cytoplasm and often s·hows a filigree pattern of necrosis or nuclear pyknosis. Poorly differentiated (round cell) synovial sarcoma is best distinguished immunophenotypically (EMA andlor keratin positive, CD99 often positive, S-100 occasionally positive, desmin consistently negative). Occasional cases raise other differential possibilities e.g. rarely angiosarcoma (if alveolar structures are set in a sclerotic stroma) or perhaps desmoplastic small cell tumour. A more frequent problem may be the reliable distinction from embryonal rhabdomyosarcoma, in which case the presence of large, rounded nuclei and often focal nesting on a reticulin stain are generally most useful. crfR 109"' Cancer Seminar Soft Tiss11e T11mors 64 References: I. Bale PM, Parsons RE, Stevens MM. Diagnosis and behaviour of juvenile rhabdomyosarcoma. Hum Patilol 1983; 14:596-611. 2. Maurer HM, Beltangady M, Gehan EA et al. The Intergroup Rhabdomyosarcoma Study - I. A final report. Cancer 1988; 61:209-220. 3. Crist WM, Garnsey L, Beltagady MS et al. Prognosis in children with rhabdomyosarcoma: a report of the Intergroup Rhabdomyosarcoma Studies 1 and 11. J Clin Oncol 1990; 8:443-452. 4. Maurer HM. Geban EA. Beltangady Met al. The Intergroup Rhabdomyosarcoma Study- II. Cancer 1993; 71: 1904-1922. 5. Hollowood K, Fletcher COM. Rhabdomyosarcoma in adults. Semin Diagn Parhol 1994; II :47-57. 6. Parham DM, Webber B, HGlt H, Williams WK, Maurer H. Immunohistochemical study of childhood rhabdomyosarcomas and related neoplasms. Cancer 1991; 67:3072-3080. 7. Tsokos M. The diagnosis and classification of childhood rhabdomyosarcoma. Semin Diagn Patilol 1994; 11:26-38. 8. Wang NP et al. Expre$sion of myogenic regulatory proteins (myogenin and MyoD I) in small blue round cell tumors of childhood. Am J Pathol 1995; 147:1799-1810. 9. Leuschner I, Newton W A, Schmidt D et al. Spindle cell variants of embryonal rhabdomyosarcoma in the paratesticular region. A report of the Intergroup Rhabdomyosarcoma Study. Am J Surg Patilo/ 1993; 17:221-230. 10. Kodet R, Newton W A, Hamoudi AB et al. Childhood rhabdomyosarcoma with anaplastic (pleomorphic) features. A report of the fntergroup Rhabdomyosarcoma Study. Am J Surg Pathol 1993; 17:443-453. II. Kodet R, Newton W A, Hamoudi AB, Asmar L. Rhabdomyosarcomas with intennediate filament inclusions and features of rhabdoid tumors. Ught microscopic and immunohistochemical study. Am J Surg Patlzol 1991; 15:257- 267. 12. Enterline HT, Hom RC. Alveolar rhabdomyosarcoma. A distinctive tumor type. Am J Clin Patholl958; 29:356-366. 13. Enzinger FM, Shiraki M. Alveolar rhabdomyosarcoma. An analysis of 110 cases. Cancer 1969; 24:18-31. 14. Tsokos M, Webber BL. Parham DM et al. Rhabdomyosarcoma. A new classification scheme related to prognosis. Arch Patltoli.Ab Med 1992; 116:847- 855. 15. Parham DM et al. Solid alveolar rhabdomyosarcomas with t(2; 13). Am J Surg Pathol 1994; 18:474-478. 16. Qualman SJ et al. Intergroup Rhabdomyosarcoma Study: update for pathologists. Pediatr Dev Patltol1998; 1:550·56 1. 17. Chan JKC, Ng H-K, Wan KY et al. Clear cell rhabdomyosarcoma of the nasal cavity and paranasal sinuses. Histopathology 1989; 14:39 1-399. CITR I 09"' Cancer Seminar Soft Ti.uue Tumors 65 18. Gaffney EF, Dervan P, Aetcher COM. Pleomorphic rhabdomyosarcoma in adulthood: analysis of II cases with definition of diagnostic criteria. Am J Surg Patho/1993; 17:601-609. !9. Schurch W et al. Pleomorphic soft tissue myogenic sarcomas of adulthood. A reappraisal in the mid-1990s. Am l Surg Pathol 1996; 20:131-147. 20. Scrable H, Wine D. Shimada H et al. Molecular differential pathology of rhabdomyosarcoma Genes Chromes Cancer 1989; 1:23-35. 21. Parham OM. The molecular biology of childhood rhabdomyosarcoma. Semin Diagn Pathoi 1994; II :39-46. 22. Kushner BH et al. Clinically critical impact of molecular genetic studies in pediatric solid tumors. Med Pediatr Oncol 1999; 33:530-535. 23. Anderson J et al. Genes. chromosomes and rhabdomyosarcoma. Genes Chromos Cancer 1999; 26:275-285. 24. Kelly KM et al. Common and variant gene fusions predict distinct clinical phenotypes in rhabdomyosarcoma. J C/in Onco/1991; 15:1831-1836. C1TR 10911> Cancer Seminar Soft Tissue Tumnrs 66 Case #19 History - Ace. #28362 During work-up for an inguinal hernia, this 19-year-old male was found to have a mass in the, left inguinal area. A 5.0 x 2.2 x 1.0 em portion of firm, pink-gray tissue was resected. (Contributed by Kenneth Frankel, MD.) Diagnosis: Intra-Abdominal Desmoplo.stic Small Cell Tumour This distinctive member of the small round cell group of sarcomas has been recognised in small case numbers since 1989.1 The first substantive series wa~· published in 1991 2 and more than 200 cases have now been reported. Most cases of intra-abdominal desmoplastic small cell tumour present in chiloren and young adults, usually with symptoms of abdominal pain and distension.H There is a striking male predominance. An abdominal mass is often palpable and, characteristically, laparotomy reveals multiple peritoneal masses within the abdominal cavity, frequently involving the pelvis. Extension into the retroperitoneum may also be seen. Despite aggressive chemotherapy, the majority of patient succumb within three years, usually to the effects of extensive intra-abdominal tumour, and often they do not live long enough to develop systemic metastaSes. Small 7 8 series of this entity have been described in the paratesticular Tegion an~ ovary , and occasional cases located outside the abdominal/peritoneal cavity are now recognised and 9 1 have been confJIDJed molecularly (see below). " The histology is both characteristic and reproducible in most ca~es and consists of sheets and islands of small basophilic cells set in a prominent desmoplastic stroma. The .relative proportions of cellular islands and stroma are variable, but the latter is a distinctive and almost invariable component of the tumour and is also present in distant metastases. The individual tumour cells have small hyperchromatic nuclei showing frequent mitoses. Cytoplasm is usually scanty and eosinophilic but occasional cells with clear cytoplasm are seen, and cells with a 'rhabdoid' appe.arance, characteri.sed by hyaline cytoplasmic inclusions, are common. Foci of glandular differentiation and rosette-like structures have been described in a minority of cases. The .small cell islands often show central necrosis and may be bounded peripherally by a thin layer of eosinophilic hyaline collagen. With tlte passage of time and additional exp,erience, it has also been appreciated that this tumor has a broader morphologic spectrum ·6 not only in terms of relative proponions of small cell and desmoplastic components, but also in variability of tumour cell Size and in cytOmorphology, which includes spindle cell, epithelioid, clear cell, trabecular and even relatively pleomorphic variants. Perhaps the most characteristic feature of these tumours is their extraordinary immunophenotype. Nearly all cases show positive staining with antibodies t() keratin, CITR I09" Cancer Seminar Softlisme Tumors 67 epithelial membrane antigen, desmin and neuron-specific enolase. Moreover, immunohistochemical staining of serial sections clearly indicates that this divergent pattern of differentiation takes place within the same tumour cells. In addition, up to 50% of these lesions show CD99 (0-13) immunopositivity but, most usefully in differential diagnosis, these tumours can be separated from Ewing's sarcoma/PNET and from rhabdomyosarcoma by their almost invarable positivity for WT- 1 (using antibodies to the C-terminal of the 12 13 protein). • Ultrastructurally rumour cells are largely undifferentiated with paranuclear whorls of intermediate filaments. Up to 10% of cases have been said to show dense core granules. The precise nosologic status of intra-abdominal desmoplastic small cell tumour remains uncertain. The histological appearances, together with its unique immunophenotype, have been used to argue that this tumour represents a blastomatous neoplasm arising from primitive pfuripotential cells and differentiating either towards mesothelium ("mesothelioblastoma") or neural crest. The demonstration of a distinctive 14 15 t( ll;22)(pl3;q12) reciprocal translocation, in these lesions, ' quite different from that in Ewing's sarcoma/PNET, has validated the discrete nature of these lesions and, since the gene fus ion was rapidly cloned,16 this has provided a valuable means of confirming the 13 17 18 diagnosis,6. • particularly in clinicopathologically a.typical cases. It is remarkable to note the speed with which the gene fusion product (EWS- WTJ) resulting from the tumour specific translocation above was cloned - thus defining the molecular genetic signature of this neoplasm Jess than 5 years after its first description! While clearly this reflects the substantial and rapid advances in positional cloning techniques, it is noteworthy that, more than 5 years after the breakpoint was cloned, effective therapy for maintaining prolonged disease-free remission remains unavailable. References: I. Gerald WL, Rosai J. Desmoplastic small cell rumor with divergent differentiation. Pediatr Pathol 1989; 9:177-183. 2. Gerald WL et al. Intra-abdominal desmopla~t ic small round cell tumor. Report of 19 cases of a distinctive type of high-grade polyphenOlype malignancy affecting young individuals. Am J Surg Pathol 1991; 15:499-513. 3. Ordonez NG et al. Intra-abdominal desmoplastic small cell tumor: a tight microscopic, immunocytochemical, ultrastructural and flow cytometric study. Hum Patlwl 1993; 24:850-865. 4. Wills EJ. Peritoneal desmoplastic small round cell tumors with divergent differentiation: a review. Ultrastruct Parho/1993; 17:295-306. 5. Ordonez NG. Desmopla~tic small round cell tumor. A histopathologic study of 39 cases with empha~is on unusual histologic partems. · Am J Surg Pathol 1998: 22:1303-1313. 6. Gerald WL el a!. Clinical, pathologic and molecular spectrum of tumors associated with t( ll;22)(p l3;q 12): Desmoplastic small round cell tumor and its variants . .! Clin Onco/ 1998; 16:3028-3036. 7. Cummings OW et uJ. Desmoplastic small round cell tumors of the paratesticular region. A report of six cases. Am J Surg Patlrol 1997; 21:219-225. C'ITR 109"' Cancer Seminar Soft Ti. CTrR J09'" Cancer Seminar Soft Tiss11e T11mors 69 Case#20 History- Ace. #25337 This 74-year-old Caucasian male had a nine year history of elevated alkaline phosphatase but was otherwise clinically healthy. A liver scan showed multiple defects. At laparotomy, multiple umbilicated nodules ranging from 1 to 4 em in diameter were observed. A 5.0 x 3.5 x 2.4 em ponion of liver was resected. (Contributed by Joseph Carberry, M.D.) Diagnosis: Epithelioid Haema.hgioendothelwma Epithelioid hemangioendothelioma, described as a distinctive entity in soft tissues in 1982, is a lowfrade malignant vascular neoplasm in the spectrum of epithelioid endothelial tumors.2 I regard this tumor, however. as fully malignant in view of its high morbidity and monality (see below).4 Previously, similar cases were classified with other epithelioid endothelial lesions under the term histiocytoid hemangiorna.s Identical tumors occur in other organs including the lung (where thef were known formerly as in1.ravascular 6 89 10 11 bronchioalveolar tumor), liver .(as in this case), · • bone, pleura and peritoneum, 4 12 skin, ' lymph node and even stomach, brain and meninges. In lung, liver and bone, multicentricity is common, as in the case presented. Because about 50% of the lesions 14 13 arise from a large or medium-sized blood vessel, especially a vein, • tllis tumor can arise potentially from any organ. Epithelioid hemangioendothelioma occurs over a wide age range, but is commonest in middle-aged adults; it is distinctly rare in children. Soft tissue lesions show no sex predilection, as opposed to those in the lung and liver where females predominate by 2: I. Tumors in soft tissue are usually solitary, in contrast to the multicentricity at other sites which can be mistaken for metastasis. Clinical presentation varies according to the organ involved but in soft tissue, as well as a mass. symptoms such as intractable pain can be related to the effects of vascular occlusion by rumor. Up to 30% of both soft tissue and liver cases are associated with metastasis. Reported monality rates vary from I 7% in soft tissue4 to 43% in liver9 and 65% in lung.2 Many of the patients with hepatic disease succumb to liver failure, hence the increasing role of hepatic transplantation in such cases. Macroscopically, these tumors typically have a pale, very firm and sometimes rather hyaline appearance, especially when arising in deep soft tissue. Those involving the liver (and lung) are cypicaiJy characterised by multiple nodules. Most tumors are lU defined and infiltrative and are composed of epithelioid, polygonal, or less commonly, snort spindle-shaped endothelial cells with variable amounts of pink cytoplasm and a ves icular nucleus with inconspicuous nucleolus. In hepatic lesions, spindle or stellate tumour cell morphology is often striking. Tumor cells are arranged in cords, small nests or shon trabeculae surrounded by a variably hyaline or myxoid stroma. which commonly has CTIR 109"' Cancer Semi~ar Soft Tissue Twnors 70 a rather chondroid appearance. Hepatic lesions tend to be diffusely infiltrative between pre-existing hepatocyte plates, and are generally associated with a more fibrous stroma which, towards the centre of lesional nodules, may be notably hyalinised. Intervening hepatocytes may show striking reactive or proliferative changes. often producing their own cord-like or tubuloglandular appearance. Prominent cytoplasmic vacuoles containing occasional erythrocytes, reminiscem of primitive vascular channels. are frequently seen, but well -formed blood vessels are, at best, only focally present. When the tumor arises from a blood vessel, the cells fill the lumen and extend centrifugally through the wall into the surrounding tissue. Complete occlusion or obliteration of the pre-existing vessel is common. A reticulin stain shows a tubular pattern, highlighting the vascular architecture. In some cases. dystrophic calcification and metaplastic ossification are prominent features. Stromal inflammation is generally not a prominent feature but some cases are associated 13 15 with a prominent osteoclast-like gi!Uit cell reaction. ' Although tumor cells are usually only mildly atypical, with a low mitotic count, there is a wide spectrum of morphology even sometimes within the same lesion; a small proponion of cases show large solid nests of cells with prominent cytologic arypia and high mitotic count and even areas indistinguishable from epithelioid angiosarcoma can be seen. Cases with such features are usually associated with poor prognosis and have been 2 labeled malignant. .4 Although, there is no clear correlation between histologic features and prognosis, there is a trend for striking nuclear atypia, mitotic count and angiosarcoma like foci to be associated with poor clinical outcome. lmmunohistochemically, most epithelioid hemangioendotheliomas show a typical vascular phenotype with expression of endothelial markers, most notably CD31 and Von Willebrand factor. Up to 45% of cases show positivity for alpha-smooth muscle actin and 4 6 17 26% show positivity for cytokeratin. ·' • As opposed to epithelial tumors, EMA (epithelial membrane antigen) is usually negative. The keratin positivity, which is most frequent in lesions arising in bone, most likely reflects the high filament content of the cell cytoplasm. The main differential diagnosis, especially in parenchymal organs and bone, is with metastatic or primary carcinoma. Helpful distinguishing features are the presence of erythrocytes and absence of mucin in the cytoplasmic vacuoles of epi thel ioid hemangioendothelioma, coupled with the immunopositivity for vascular markers. Also the degree of nuclear pleomorphism in carcinomas is usually more pronounced. In soft tissues, the differential diagnosis also includes epithelioid sarcoma. The latter shows a more sheet-like growth pattern, only occasional cytoplasmic vacuoles, and it is positive for both keratin and EMA, often CD34 positive but negative for more specific endothel ial markers such as CD31 or VWF. Cases with a very prominent myxoid stroma can be confused with myxoid liposarcoma or myxoid chondrosarcoma but the latter has a lobular architecture, lacks cytoplasmic vacuoles and is S-100-positive. Myxoid liposarcoma is best distinguished by identification of the typical branching vascular pattern and small muhivacuolated lipoblasts. CITR 109'' Cancer Seminar Sofr Ti.r.t11 e Tumors 71 References: I. Weiss SW, Enzinger FM. Epithelioid hemangioendothelioma. A vascular tumor often mistaken for a carcinoma. Cancer 1982; 50:970-981. 2. Weiss SW, Ishak KG, Dail DH et al. Epithelioid hemangioendothelioma and related lesions. Semin Diagn Patho/1986; 3:259-287. 3. Tsang WYW, Chan JKC. The family of epithelioid vascular tumors. Histol Histopatho/1993; 8:187-212. 4. Mentzel T, Beham A, Calonje E, Katenkamp D, Retcher COM. Epithelioid hemangioendothelioma of skin and soft tissues: clinicopathologic and immunohistochemical study of30 cases. Am J Surg Parholl997; 2 1:363-374. 5. Rosai J et al. The histiocytoid hemangiomas. A unifying concept embracing several previously described entities of skin, soft tissue, large vessels, bone and heart. Hum Pathol1979; 10:707-730. 6. Dail DH, Leibow AA, Gmelich IT et al. Intravascular, bronchiolar and alveolar tumor of the lung (IVBAn. Cancer 1983; 51 :452-464. 7. Ishak KG et al. Epithelioid hemangioendothelioma of the liver: A clinicopathologic and follow-up study of 32 cases. Hum Pathol 1984; 15:839-852. 8. Dietze 0 et al. Malignant epithelioid haemangioendothelioma of the liver: A clinicopathological and histochemical study of 12 cases. Histopathology 1989: 15:225-237. 9. Mak.hlouf HR. Ishak KG, Goodman ZD. Epithelioid hemangioendothelioma of the liver: A clinicopathologic study of 137 cases. Cancer 1999; 85: 562-582. 10. Tsuneyoshi M, Dorfman HD, Bauer TW. Epithelioid hemangioendothelioma of bone. A clinicopathologic, ultrastructural and immunohistOchem ical study. Am J Surg Patho/1986; 10:754-764. II. Lin BT, Colby T, Gown AM et al. Malignant vascular tumors of the serous membranes mimicking mesothelioma. A repon of 14 cases. Am J Surg Patlwl 1996; 20:1431-1439. 12. Quante M, Patel NK, HillS et al. Epithelioid hemangioendothelioma presenting in the skin. A clinicopathologic study of eight cases. Am J Dermatopathol 1998; 20:541-546. 13. Suster S, Moran CA, Koss MN. Epithelioid hemangioendothelioma of the anterior mediastinum. Clinicopathologic, immunohistochemical and ultr.lstructural analysis of 12 cases. Am J Surg Par hoi 1994; 18:871-881. 14. Lamovec J, Sobel H, Zidon A et al. Epithelioid hemangioendothelioma of the anteri or mediastinum with osteoclast-like giant cells. Am J Clin Parhol 1990; 93:813-8 17. 15. Williams SB, Bulter CB, Gilkey GW et al. Epithelioid hemangioendothelioma with osteoclast like giant cells. Arch Patholl.Ab Med 1993; 117:315-318. 16. Van Haelst UJ GM, Pruszczynski M, Ten Cate LN et al. Ultrastructural and immunohistochemical study of epithelioid hemangioendothelioma of bone: Coexpression of epithelial and endothelial markers. Ultrasrruct Pathol 1990: 14: 141 - 149. 17. Gray MF, Rosenberg AE, Dickersin GRetal. Cytokeratin express ion in epithelioid va~cu l arneop lasms. Hum P(llho/1990; 21:212-217. CTTR 109"' Cancer Seminar Soft Tissue Tumor.< 72 .li1J7 , . f , :?j l L.· ( ,. q J • ,I \ }' ' ( • • •• _,. I ·--- I a' • • ..• ' ..,.J # , ·'":. ' lj ...... , ,. \ . -. \ J ' ' ...:.·-. ... '• # .. . .. ~ ;· ·. -' ~ ·- .-- ..._. --- ... _. . .. . • I . -. .-. -. .- . •. •. ~ • .. ' .. ~- · , -~ , • .• •. • ' .. __ ... •""':;r ~· ·; • • , • ' • • ·(" ('; -, ..... f ••• . •. • _, • .. - ~ ·', • ' . . . • Figure Legends Figure 1 Desmoplastic fibroblastoma. This is a typically hypocellular collagenous lesion(a) with ar.eas that are somewhat more myxoid (b). Note that the lesional fibroblasts have characteristically angulated or stellate nuclei (b and c). Figure 2 Cellular angiofibroma. This cellular spindle cell neoplasm has prominent vessels (a) vrith variably thick walls (b). Tumour cells resemble those in spindle cell lipoma, having short bland nuclei and indistinct cytoplasm (c). Note also the stromal mast cells. Figure 3 Mammary-type myofibroblastoma. This example is somewhat more cellular than an 'average' case but shows the usual rather patternless admixture of uniform spindle cells and hyaline collagen bundles (a and b). Closer examination confirms the cytologic uniforrility and (again) something of a resemblance to spindle cell lipoma, (c). These lesions are consistently immunopositive for CD34 and desmin (d). Figure 4 Inflammatory myofibroblastic tumour. Formerly known as inflammatory pseudotumour or plasma cell granuloma, these are generally highly cellular lesions (a) with a predominantly fascicular growth pattern (b). The spindle cells in this case have plump but uniform nuclei and are admixed with numerous lymphocytes and plasma cells (c). Figure 5 Pseudosarcomatous reactive myofibroblastic proliferation. This bladder lesion shows a typically infiltrative pattern into muscle (a). The lesional spindle cells are admixed with predominantly acute inflammatory cells (b). Note, in areas, the mimicry of rhabdomyosarcoma with elongated strap-like cells, but in this context being set in an inflamed myxoid stroma (c). Figure 6 Benign fibrous histiocytoma with cellular and aneurysmal features. This is a highly cellular dermal lesion (a), which is notable for prominent stromal haemorrhage and (in areas) blood-filled pseudovascular spaces (b). Closer examination reveals the typically polymorphic cytology of a cutaneous fibrous histiocytoma (c). Mitoses are often ·numerous adjacent to areas ofhaemorrhage. Figure 7 FibrosarcomatO\IS (higher grade) variant of dermatofibrosarcoma protuberans. Low power examination shows an abrupt transition from relatively liypocellular (left) to more cellular (right) areas (a). The tess cellular areas show relatively bland cytology, a storiforrn growth pattern and collagenous stroma. (b). The cellular areas are more. fascicular and mitotically active with less stromal collagen (c). There is striking CD34 positivity in the less cellular component (d). CTIR 109"' Cancer Seminar Soft tl.twe [umot.t 82 Figure 8 Solitary fibrous tumour. This is a patternless spindle cell neoplasm with clusters of thin-walled vessels (a), which is infiltrating adjacent salivary tissue (b). Tumour cells have unifonn tapering nuclei and are set in a collagenous stroma (c). There is immunopositivity for CD34 (d), CD99 and bcl-2. Figure 9 Monophasic synovial sarcoma. This mediastinal neoplasm is a fascicular spindle cell sarcoma with, focally, branching pericytoma-like blood vessels (a) and stromal calcifications (b). The tumour cells have monomorphic, somewhat tapering nuclei and there is intervening wiry stromal collagen (c). Figure 10 Gastrointestinal stcomal tumour. This tumour is located in the bowel wall (a) and is composed of variably epithelioid (b) and more spindle shaped (c) cells with relatively uniform nuclei. Tumour cells show striking CD117 (c-k:it) immunopositivity (d). Figure II Spindle cell and epithelioid neoplasm most consistent with MPNST. This unusual tumour has both spindle cell and epithelioid components (a) with relatively abrupt transition between them (b). The spindle cell component is bland, while the epithelioid areas are quite cellular (c). Tumour cells show striking immunopositivity for S-100 protein (d). Figure 12 Diffuse-type giant cell tumour (so called "pigmented villonodular tenosynovitis"). This is a variably cellular lesion with a loose, focally cystic stroma (a). Tumour cells are variably mono- or multi-nucleated and are associated \vith stromal haemorrhage (b). Closer examination shows that many of the cells have histiocytoid morphology and some contain intracytoplasmic haemosiderin (c). Figure 13 Extraskeletal myxoid chondrosarcoma. This tumour has a lobulated myxoid appearance (a) and is composed mainly of eosinophilic spindle cells in a lacy (reticular) pattern (b). Elsewhere there is transition (c) to areas in which the cells are smaller and more rounded (d). Some of the latter cells have eosinophilic cytoplasmic inclusions. Figure !4 High grade myxofibrosarcoma (so-called myxoid "MFH"). This is a distinctively pleomorphic myxoid sarcoma with notably curvilinear vessels (a). Tumour cells tend to cluster around vessels and there are vacuolated mucin-containing pseudolipoblasts (b). The solid high grade areas have a non-distinctive pleomorphic ("MFH"-like) appearance (c), shared by many pleomorpbic malignant neoplasms. Figure 15 Atypical lipomatous tumour (well differentiated liposarcoma). Even at low power, this tumour shows variation in adipocyte size, atypical stromal cells and adipocytic nuclear hyperchromasia (a), sufficient to make the CTTR I09\h Cancer Seminar Soft 7/s.\?~e Tumors 83 diagnosis. Although lipoblasts are present in this case (b), they are not a diagnostic requirement. Areas of the lesion show some increase in cellularity (c), worrisome for 'incipient' dedifferentiation. Figure 16 Dedifferentiated liposarcoma. There is abrupt transition from well differentiated areas to a more hypercellular non-lipogenic appearance (a). Most of the dedifferentiated component has a non-distinctive high gmde spindle cell appearance (b). In this case the dedifferentiated element shows focal heterologous rhabdomyoblastic differentiation (c), which is confirmed immunohistochemically. Figure 17 High grade spindle cell sarcoma with osteogenic differentiation. This tumour has two distinct components - high grade spindle ceU areas with ' MFH' -like features (a} and an extensive osteogenic component (b), which clearly shows malignant features (c). While these appearances are consistent with soft tissue osteosarcoma, this combination of features in the retroperitoneum is much more likely to represent dedifferentiated liposarcoma with heterologous differentiation and further sampling is indicated. Figure 18 Alveolar rhabdomyosarcoma. This tumour from the hand is a highly cellular malignant round cell neoplasm with both solid areas (a) and areas with a looser alveolar architecture (b). The tumour cells have larger nuclei than an embryonal rhabdomyosarcoma and some of the cells have eosinophilic cytoplasm (c). Figure 19 Intra-a bdominal desmoplastic small cell tumour. At low power, the tumour consists of basophilic small cells set in a prominent desmoplastic fibrous stroma (a). The small cell component is armnged in nests (a) and has a variably round cell or more angulated (? apoptotic) appearance (c). lmmunostains for keratin, desmin, NSE and WT-1 are positive. Figure 20 Epithelioid haemangioeodotbelioma. The hepatic architecture is largely effaced by neoplastic tissue with a fibrous stroma (a). The tumour consists of cords and strands of epithelioid and polygonal cells (b), some of which have vacuolated cytoplasm (intracytoplasmic 'lumina') (c). The tumour cells show diffuse immunopositivity for CD31 (d). CTIR I09"' Cancer Seminar Soft 1issue Tumors 84 California Tumor Tissue Registry Department of Pathology 11021 Campus Avenue, AH 335 Loma Linda, California 92350 Voice (909) 558-4788 Fax (909) 558-0188 Email: [email protected] Home page: www.cttr.org Case of the month: www.llu.edu/llu/cttr/cotm CTTR Acknowlcd•cments; Manager, aod Seminar Coordinator: Anne Chism Chief His!Ot:cbologist: Linda Healy Office Secretary: Christine Rose Co-Director and Editorial Supenisor: Resa L. Chase, M.D. E=utivc Director arul Graphics Editor: DooaldR. Chase, M.D. CTIR 109" Cancer Serninar Soft Ti:R>>te Tumors 85