Spindle cell sarcomas of soft tissue; beyond herringbones and - - PDF document

1

Spindle cell sarcomas of soft tissue; beyond herringbones and fibrosarcoma goals UCSF Current Issues in Anatomic Pathology 2016 Andrew Horvai MD PhD Clinical Professor, Pathology UCSF Medical Center Mission Bay 1825 4th Street, Room M2354 San Francisco, CA 94158

2

INTRODUCTION AND HISTORICAL BACKGROUND Historically, a soft tissue tumor composed of a highly cellular proliferation of spindle cells, typically with a “herringbone” growth pattern was classified as

• fibrosarcoma. However, this diagnosis extended well beyond this narrow definition so

that fibrosarcoma was one of the most common subtypes of soft tissue malignancy. One of the earliest reviews of the subject by Bick at Mount Sinai defined fibrosarcoma as a “malignant tumor of connective tissue in which the cells are non-osseous mesoblastic mesoblastic origin.” 1 This definition included spindled, round and giant cells or a mixture, arising from fascia, blood vessels, nerve sheath or periosteum and almost certainly represented entities that would not be classified differently. At the Mayo clinic, the diagnosis of fibrosarcoma has decreased in incidence from a high of 66% in 1936 to <1% of adult soft tissue sarcomas by 2010. 2, 3 This approximately mirrors the use of the diagnosis in the pathology literature over the same time period. The decrease in fibrosarcoma incidence stems from improved diagnostic accuracy and classification of soft tissue sarcomas and a more strict definition of

• fibrosarcoma. The 2013 WHO defines fibrosarcoma as a “malignant tumor, composed
• f fibroblasts with variable collagen and, in classical cases, a herringbone architecture.

It is a diagnosis of exclusion”. 4 Note that the emphasized clause at the end of the definition was added in the most recent edition of the WHO. Despite this important qualifier, the definition is still problematic since there is no morphologic, immunophenotypic or genetic “gold standard” for a neoplastic fibroblast. From a practical perspective, two question arise from these observations:

• How have tumors formerly known as fibrosarcoma been

reclassified?

• Are there any true fibrosarcomas left?

To answer these questions, the current presentation will (1) review the soft tissue tumors that look like the classical definition of fibrosarcoma: namely spindle cells in a tightly packed fascicles, occasionally herringbone, pattern and (2) discuss the few remaining entities that are still called fibrosarcoma (irrespective of whether that designation is accurate in a biological sense). MONOPHASIC SYNOVIAL SARCOMA Apart from undifferentiated spindle cell sarcoma (discussed later), monophasic synovial sarcoma represents the most common mimic of fibrosarcoma.3 This is not surprising since monophasic synovial sarcoma is characterized by a highly uniform population of fusiform spindle cells growing in tightly packed fascicles. Although the herringbone

3

growth pattern is not always present, it can be seen in a subset of cases. Unlike biphasic synovial sarcoma, which demonstrates more overt epithelial differentiation in the form of pseudoglands, monophasic lacks this feature. Monophasic synovial sarcoma may demonstrate a branching, staghorn-like, vasculature and calcifications, which is not typical of fibrosarcoma. Keratin and EMA expression may be helpful to support the diagnosis of synovial sarcoma, but the monophasic type usually shows only very focal staining with these antigens or they can be entirely absent.5 TLE1 has shown promise as a helpful ancillary marker6 although, in our experience, specificity is not very

• high. Molecular or genetic methods to detect the t(X;18) translocation or resultant SS18

rearrangement has proven extremely helpful for definitive diagnosis.7 To the best of

• ur knowledge, this translocation remains highly specific for synovial sarcoma.

Clinically, synovial sarcoma typically affects young adults, and although it can occur anywhere in the body, the lower extremity in the vicinity of the knee is most common. Incidentally, it is this anatomic predilection that is responsible for the name synovial

• sarcoma. They are high grade tumors that are treated with surgery and adjuvant
• chemotherapy. Five year survival is between 36 and 76% depending on the series.

Some data suggest that small, superficial tumors and those with calcifications have better prognosis. 8, 9 MALIGNANT PERIPHERAL NERVE SHEATH TUMOR Malignant peripheral nerve sheath tumor (MPNST) is defined as a malignant neoplasm that either derives from, or differentiates toward, a peripheral nerve sheath cell type. 10,

11 Approximately half of cases arise in patients with Neurofibromatosis 1 (NF1). Familial

NF1 cases more frequently arise from pre-existing benign nerve sheath tumors (usually neurofibroma) whereas sporadic cases arise de novo. In the absence of NF1 history or a pre-existing benign neurofibroma, the diagnosis is very difficult to make with certainty. This is because no “gold standard” morphologic, immunophenotypic or genetic changes have been identified to prove MPNST with certainty. In fact, one could argue that the clinical setting of NF1 is hardly sufficient evidence to diagnose every spindle cell sarcoma in these patients as MPNST. Low-grade MPNST have a high recurrence risk (~50%) whereas high-grade forms frequently metastasize. Though MPNST can have a number of growth patterns, the most common tightly packed fascicular growth mimics “fibrosarcoma.” Unfortunately, there are no pathognomonic microscopic features unless a precursor benign nerve sheath tumor (usually neurofibroma) is present. The nuclei are hyperchromatic and fusiform but more tapered at their tips than synovial sarcoma. MPNST typically does not show calcifications or staghorn vessels as seen in synovial sarcoma. A peculiar herniation of

4

tumor cells into the media or lumen of vessels has been described as a useful feature of

• MPNST. Mitotic activity and pleomorphism is generally higher than synovial sarcoma.

Immunohistochemistry can be helpful. Approximately half of MPNST express at least focal S100 or SOX10.12, 13. Somewhat paradoxically, diffuse and strong S100 or SOX10 staining in a spindle cell neoplasm argues against MPNST and should raise the possibility of a spindle cell melanoma. Focal S100 can be seen in synovial sarcoma, but SOX10 is much less common.13 Unfortunately, about half of MPNST do not express either S100 or SOX10. Recently, loss of H3K27Me3 expression has been described as highly sensitive for MPNST.14 Currently, there are no reproducible genetic changes to confirm the diagnosis of MPNST. 10 Mutations in the NF1 gene are of course found in patients with Neurofibromatosis 1, but the finding is not specific. MPNST do not harbor rearrangements of the SS18 gene. DERMATOFIBROSARCOMA PROTUBERANS and FIBROSARCOMATOUS PROGRESSION Dermatofibrosarcoma protuberans (DFSP) is a locally aggressive, dermal and subcutaneous neoplasm of fibroblasts.15 Although DFSP is composed of monomorphic spindle cells, the storiform growth and fat infiltration are so characteristic that, despite the “fibrosarcoma” name, its microscopic appearance is distinct from classic “fibrosarcoma” (i.e. fascicular, herringbone). However, DFSP can occasionally progress to a higher grade sarcoma, known as fibrosarcomatous DFSP. The latter tumors, like DFSP, arise on the trunk and proximal extremities of middle aged adults with either a synchronous or metachronous DFSP at the same site. There is some controversy whether a fibrosarcomatous component invariably carries a worse prognosis. Though a higher recurrence and metastatic risk was reported in earlier studies, a more recent study in which patients received wide local excision (as they would for conventional DFSP) did not show a worse prognosis if fibrosarcoma was present.16 Both DFSP and fibrosarcomatous DFSP contain a monotonous population of spindle cells with scant cytoplasm, and hyperchromatic nuclei with tapered ends. Microscopically, compared to the storiform pattern of DFSP, the fibrosarcomatous component is fascicular with the classic herringbone pattern, more diffusely hypercellular and larger (>5 cm). Necrosis and abundant mitoses (> 7 / 10 hpf) are

• common. The residual DFSP can often be identified at the periphery of the lesion.

The tumors express CD34, albeit less intensely and diffusely than the precursor DFSP in a manner analogous to loss of S100 staining in MPNST compared to benign

• schwannoma. Fibrosarcomatous DFSP notably lacks keratin, S100 or desmin.

Genetically, DFSP and fibrosarcomatous DFSP share the same t(17;22)(q21;q13)

5

translocation and resulting COL1A1:PDGFβ fusion. Such genetic confirmation is seldom necessary for classic DFSP, but for a dermal-subcutaneous spindle cell sarcoma lacking a DFSP precursor, it may be useful. INFANTILE FIBROSARCOMA Infantile fibrosarcoma is a very rare mesenchymal malignancy in neonates. It is almost never seen after 1 year of age.17 Although retaining the name “fibrosarcoma” for historical reasons, it is unclear whether the tumor is truly fibroblastic and, as with many translocation-driven sarcomas, it represents a clinicopathologic and genetic entity rather than a histologic subtype. Most tumors involved the distal lower extremity though the anatomic distribution is wide. Despite the high grade morphology, the tumors have a relatively good prognosis compared to other pediatric sarcomas such as rhabdomyosarcoma or adult fibrosarcoma. Therapy consists of either neoadjuvant chemotherapy to reduce tumor size and conservative excision or wide excision with or without adjuvant chemotherapy.18 The prognosis does not correlate with histologic grading. Microscopically, the tumors are highly cellular. They consist of monotonous small spindled to ovoid cells usually with little intervening stroma. Although a herringbone pattern may be observed in some cases, sheet like or haphazard growth is more common in our experience. Pleomorphism is absent. Hemangiopericytomatous vasculature can be seen. Mitotic activity can vary within a given tumor sometimes exceeding 20 per 10 high power fields. However, as mentioned above, this is not

• prognostic. Rare cases have more fibrous stroma or a prominent lymphoid infiltrate

(another feature typically absent in adult fibrosarcoma). The tumors are negative for keratins, desmin, myogenin, S100 and CD34. Some weak SMA staining may be seen but this is a very nonspecific result among spindle cell

• sarcomas. The ETV6:NTRK3 gene fusion resulting from a t(12;15) translocation is
• diagnostic. Interestingly, the same translocation is seen in congenital pulmonary

fibrosarcoma and congenital mesoblastic nephroma.19, 20 These may be closely related lesions but the finding of the same translocation in some breast cancers and myeloid leukemias suggests the genetics are not entirely specific for congenital sarcomas.21, 22 ADULT FIBROSARCOMA / UNDIFFERENTIATED SPINDLE CELL SARCOMA As previously discussed, the diagnosis of adult fibrosarcoma has become a diagnosis of

• exclusion. Ostensibly, fibrosarcoma is defined as a malignant mesenchymal tumor with

6

fibroblastic differentiation that usually forms collagen matrix. This definition is problematic for several reasons. First, fibroblastic differentiation is very difficult to “prove” morphologically and immunophenotypically and ultrastructure is rarely evaluated in soft tissue tumors. Furthermore, the presence of patchy SMA expression in a subset

• f “fibrosarcomas” has led some to conclude that the category also includes

myofibroblastic sarcomas.3 Third, collagen production is hardly specific for

• fibrosarcoma. Thus, one could argue that the diagnosis of undifferentiated spindle

cell sarcoma (USS) is a more accurate term to describe these tumors. By analogy, the diagnosis of malignant fibrous histiocytoma (MFH) has largely been replaced by undifferentiated pleomorphic sarcoma (UPS) to denote the exact same degree of uncertainty regarding histotype. Whichever term (fibrosarcoma, USS or even UPS) is used, the definition of this category is similar. Tumors typically arise in middle aged or older adults, involving the deep soft tissues of the proximal extremity or trunk. Superficial examples most likely represent fibrosarcomatous DFSP if analyzed at the genetic level. The prognosis is considerably worse than infantile fibrosarcoma (<55% 5 year survival). Treatment consists of wide excision with or without adjuvant (or intraoperative) radiotherapy and chemotherapy. Adult fibrosarcoma consists of monomorphic spindle cells in short to intermediate length fascicles, in parallel arrays that produce the classic “herringbone” pattern. That some authors also include tumors with storiform or haphazard growth, fibromatosis like areas

• r extensive myxoid matrix underscores the lack of a strict definition for this tumor since

these features are commonly seen in other, better characterized, subtypes of sarcoma. That said, fibrosarcoma is immunophenotypically negative for keratins, desmin, S100, CD34 but can show some weak SMA – essentially a “null” immunophenotype. Immunostains are, of course, helpful to exclude other spindle cell sarcomas discussed

• above. The usually show a genomic landscape that is complex.23, 24

7

SUMMARY

• 1. Adult fibrosarcoma is a diagnosis of exclusion
• 2. Some diagnoses that retain the diagnosis of “fibrosarcoma” are not

uniformly herringbone

• 3. Most herringbone spindle cell neoplasms in adults can be classified as

synovial sarcoma, MPNST or fibrosarcomatous DFSP

• 4. Immunohistochemistry and, in some cases, genetics are needed in most

cases to classify these tumors Table 1. Ancillary immunohistochemistry and genetic testing useful in the evaluation of monomorphic spindle cell sarcomas. Synovial sarcoma MPNST FS in DFSP Infantile FS Adult FS Keratin Rare cells

• S100

Weak, patchy Weak, patchy -

• CD34
• Rare cells

Weak

• Other

SS18 FISH SOX10 PDGF FISH ETV6 FISH - REFERENCES

1. Bick EM. Fibrosarcoma of the Extremities. Ann Surg. 1935;101:759-765. 2. Meyerding H, Broders A, R. H. Clinical aspects of fibrosarcoma of the soft tissues of the

• extremities. Surg Gynecol Obstet. 1936;62:1010-1019.

3. Bahrami A, Folpe AL. Adult-type fibrosarcoma: A reevaluation of 163 putative cases diagnosed at a single institution over a 48-year period. Am J Surg Pathol. 2010;34:1504-1513. 4. Folpe AL. Adult fibrosarcoma. In: Fletcher CD, Bridge JA, Hogendoorn PC, et al., eds. WHO Classification of Tumours of Soft Tissue and Bone. Lyon: IARC Press; 2013:91-92. 5. Corson JM, Weiss LM, Banks-Schlegel SP, et al. Keratin proteins and carcinoembryonic antigen in synovial sarcomas: an immunohistochemical study of 24 cases. Hum Pathol. 1984;15:615-621. 6. Jagdis A, Rubin BP, Tubbs RR, et al. Prospective evaluation of TLE1 as a diagnostic immunohistochemical marker in synovial sarcoma. Am J Surg Pathol. 2009;33:1743-1751. 7. Ladanyi M. Fusions of the SYT and SSX genes in synovial sarcoma. Oncogene. 2001;20:5755- 5762. 8. Michal M, Fanburg-Smith JC, Lasota J, et al. Minute synovial sarcomas of the hands and feet: a clinicopathologic study of 21 tumors less than 1 cm. Am J Surg Pathol. 2006;30:721-726. 9. Varela-Duran J, Enzinger FM. Calcifying synovial sarcoma. Cancer. 1982;50:345-352. 10. Nielsen GP, Antonescu CR, Lothe RA. Malignant peripheral nerve sheath tumor. In: Fletcher CD, Bridge JA, Hogendoorn PC, et al., eds. WHO Classification of Tumours of Soft Tissue and Bone. Lyons, France: IARC Press; 2013:187-188. 11. Scheithauer BW, Louis DN, Hunter S, et al. Malignant peripheral nerve sheath tumor. In: Louis DN, Ohgaki H, Wiestler OD, et al., eds. World Health Organization Classification of Tumours of the Central Nervous System. Lyon, France: IARC Press; 2007:160-162. 12. Karamchandani JR, Nielsen TO, van de Rijn M, et al. Sox10 and S100 in the Diagnosis of Soft- tissue Neoplasms. Appl Immunohistochem Mol Morphol. 2012;20:445-450.

8

13. Kang Y, Pekmezci M, Folpe AL, et al. Diagnostic utility of SOX10 to distinguish malignant peripheral nerve sheath tumor from synovial sarcoma, including intraneural synovial sarcoma. Mod

• Pathol. 2014;27:55-61.

14. Prieto-Granada CN, Wiesner T, Messina JL, et al. Loss of H3K27me3 Expression Is a Highly Sensitive Marker for Sporadic and Radiation-induced MPNST. Am J Surg Pathol. 2015. 15. Mentzel T, Pedeutour F, Lazar AJ, et al. Dermatofibrosarcoma protuberans. In: Fletcher CD, Bridge JA, Hogendoorn PC, et al., eds. WHO Classification of Tumours of Soft Tissue and Bone. Lyon: IARC Press; 2013:77-79. 16. Goldblum JR, Reith JD, Weiss SW. Sarcomas arising in dermatofibrosarcoma protuberans: a reappraisal of biologic behavior in eighteen cases treated by wide local excision with extended clinical follow up. Am J Surg Pathol. 2000;24:1125-1130. 17. Chung EB, Enzinger FM. Infantile fibrosarcoma. Cancer. 1976;38:729-739. 18. Russell H, Hicks MJ, Bertuch AA, et al. Infantile fibrosarcoma: clinical and histologic responses to cytotoxic chemotherapy. Pediatr Blood Cancer. 2009;53:23-27. 19. Knezevich SR, McFadden DE, Tao W, et al. A novel ETV6-NTRK3 gene fusion in congenital

• fibrosarcoma. Nat Genet. 1998;18:184-187.

20. Rubin BP, Chen CJ, Morgan TW, et al. Congenital mesoblastic nephroma t(12;15) is associated with ETV6-NTRK3 gene fusion: cytogenetic and molecular relationship to congenital (infantile)

• fibrosarcoma. Am J Pathol. 1998;153:1451-1458.

21. Euhus DM, Timmons CF, Tomlinson GE. ETV6-NTRK3--Trk-ing the primary event in human secretory breast cancer. Cancer Cell. 2002;2:347-348. 22. Kralik JM, Kranewitter W, Boesmueller H, et al. Characterization of a newly identified ETV6- NTRK3 fusion transcript in acute myeloid leukemia. Diagnostic pathology. 2011;6:19. 23. Limon J, Szadowska A, Iliszko M, et al. Recurrent chromosome changes in two adult

• fibrosarcomas. Genes Chromosomes Cancer. 1998;21:119-123.

24. Dal Cin P, Pauwels P, Sciot R, et al. Multiple chromosome rearrangements in a fibrosarcoma. Cancer Genet Cytogenet. 1996;87:176-178.