Citation for published version (APA): ten Heuvel, S. (2014). Diagnostic and prognostic aspects in soft tissue sarcomas [S.l.]: [S.n.

Transcription

1 University of Groningen Diagnostic and prognostic aspects in soft tissue sarcomas ten Heuvel, Suzan IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2014 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): ten Heuvel, S. (2014). Diagnostic and prognostic aspects in soft tissue sarcomas [S.l.]: [S.n.] Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Diagnostic and prognostic aspects in soft tissue sarcomas Suzan ten Heuvel

3 Colofon Layout and cover design FYN Werk, grafische vormgeving Fleur Bominaar Printed by Gildeprint Drukkerijen, Enschede Diagnostic and prognostic aspects in soft tissue sarcomas Proefschrift ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties. De openbare verdediging zal plaatsvinden op woensdag 19 november 2014 om uur door ISBN: Boek (Book) ISBN: Eboek : PDF zonder DRM (PDF without DRM) 2014 Suzanne ten Heuvel, The Netherlands All rights reserved. No part of this book may be reproduced, stored in a retreival system or transmitted in any form or by any means, without prior permission of the author. Suzanne Elisabeth ten Heuvel geboren op 24 september 1980 te Ruinen

4 Promotores Prof. dr. H.J. Hoekstra Prof. dr. A.J.H. Suurmeijer Beoordelingscommissie Prof. dr. J.H.W. de Wilt Prof. dr. H. Hollema Prof. dr. J.A. Gietema Inhoudsopgave General introduction 7 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas 19 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up 51 Epithelioid sarcoma: still an only surgically curable disease 67 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins 83 Summary 97 Samenvatting 103 General discussion and future perspectives 109 Dankwoord & curriculum vitae 127

5

6 General introduction 1

7 General Introduction 1 Soft tissue sarcomas configure a remarkably heterogeneous group of relatively uncommon mesenchymal tumors with distinctive histology and a wide spectrum of clinicopathological features and behavior. They comprise about 1% of adult and 15 % of pediatric malignancies. 1,2,3,4 Originally soft tissue sarcomas were classified on a histogenetic basis according to histologic differentiation towards muscle, fat, fibrous tissue, blood or lymphatic vessels and the peripheral nervous system. The current World Health Organization (WHO) classification (2013) of soft tissue sarcomas includes more than 50 different histological subtypes. 3 Most common types of soft tissue sarcoma are: pleomorphic undifferentiated sarcoma / malignant fibrous histiocytoma (MFH) (28%), liposarcoma (15%), leiomyosarcoma (12%), synovial sarcoma (10%), malignant peripheral nerve sheath tumor (MPNST) (6%) and rhabdomyosarcoma (5%). 2,3,4 The heterogeneity in their morphologic appearance and clinico-biological behavior make soft tissue sarcomas one of the most complex topics of oncology. In the past soft tissue sarcomas were often studied as a group, whereas nowadays it is well accepted that sarcomas need to be recognized as individual entities with their own specific clinico-pathological characteristics. 3,4 In this thesis various diagnostic, prognostic and treatment aspects are covered for four separate soft tissue sarcoma entities, in particular synovial sarcoma, myxoid liposarcoma, epithelioid sarcoma and dermatofibrosarcoma protuberans. Synovial sarcoma accounts for approximately 10% of all soft tissue sarcomas and most frequently presents as a painless deep-seated tumor in the extremity of a young adult patient. Histologically, synovial sarcoma is characterized by a biphasic or a monophasic spindle cell histology, defined respectively by the presence or absence of (glandular) epithelial differentiated areas. Despite improved management of local disease, still more than half of synovial sarcoma patients die of chemotherapy-resistant metastatic disease. The prevalent metastatic pattern is hematogenous and in some cases lymphatic. 3,5 Myxoid liposarcoma is the second most common type of liposarcoma, representing almost 10% of all adult soft tissue sarcomas, and frequently presents as a slowgrowing, deep-seated tumor in the lower extremity of a relatively young adult patient. Histologically, myxoid liposarcomas are characterized by the presence of primitive mesenchymal cells and scattered multivacuolar lipoblasts showing minimal nuclear 9

8 General introduction 1 pleomorphism. These cells are deposited in a myxoid stroma with a distinct plexiform capillary network. Metastases are often detected in other deep soft tissue locations, such as the retroperitoneum or the extremities. 6 Epithelioid sarcoma accounts for less than 1% of soft tissue tumors and usually presents as a slow growing painless swelling, with a predilection for the proximal extremity and especially the finger, the hand and forearm. It is most often found in young males between 10 and 35 years. Histologically, epithelioid sarcoma is composed of spindle-to polygonal epithelioid cells arranged in nodules that commonly exhibit central necrosis. As such, the tumor has a deceivingly benign histologic appearance and may not be recognized initially. In many cases epithelioid sarcoma presents in a locally advanced stadium with infiltration of contiguous structures. The tumor can originate superficially in the dermis or deeply under the fascia. Epithelioid sarcoma has a high propensity for locoregional recurrence and metastasis. 7,8 Dermatofibrosarcoma protuberans (DFSP) is a locally aggressive sarcoma arising from the skin. It accounts for approximately 4% of all soft tissue sarcomas. The tumor often presents as an indolent growing plaque, usually localized on the trunk, the proximal extremities, or the head-neck region of young or middle-aged adults. Histologically, DFSP is composed of monotonous spindle cells arranged in an irregularly whorled or storiform pattern. Typically, DFSP has a strong tendency to recur, due to its infiltrative asymmetric growth pattern, far beyond the macroscopic boundaries with asymmetric tentacle like projections. Metastases are rare in dermatofibrosarcoma protuberans and are usually preceded by multiple local recurrences and fibrosarcomatous change. 9 Diagnostics The four soft tissue sarcoma types discussed in this thesis illustrate the advances that have been made in the past decades in diagnosing these heterogenic soft tissue sarcomas. Clearly, advances in tumor diagnostics have translated into our ability to make an accurate sarcoma diagnosis, which is needed to determine appropriate individual therapeutic approach. 10,11 Immunohistochemistry (IHC) may assist in detecting tumor differentiation that is difficult to assess with routine histology. IHC is indispensable in rendering an accurate diagnosis of soft tissue sarcomas. Most soft tissue sarcomas harbor relatively specific morphologic and immunophenotypic features and usually the tumor can be classified by immunohistochemistry in concert with clinical presentation. However, some entities share overlapping immunophenotypes, complicating the histologic differential diagnosis. 3,4 In addition to IHC, molecular biology and in particular the discovery of specific chromosomal translocations in a number of sarcoma types has significantly advanced the quality of sarcoma typing. These chromosomal translocations are often unique and found consistently within the same sarcoma category, by which they have become very accurate diagnostic markers. 10,11,12,13 For instance, more than 95 % of synovial sarcomas harbour the translocation t(x;18) (p11.2;q11.2). The breakpoint of this translocation fuses the SYT gene from chromosome 18 to one of the two homologous genes, SSX1 or SSX2 on the X chromosome. 5 Myxoid liposarcoma is characterized by t(12;16)(q13;p11) or t(12;22)(q13;q12) which leads to the fusion of the FUS and CHOP genes or EWS and CHOP genes, respectively. 2,6 Dermatofibrosarcoma protuberans is characterized by a t(17;22)(22;q13) aberration with fusion of the COL1A1 gene on chromosome 17 with the platelet-derived growth factor B-chain (PDGFB) gene on chromosome In contrast, epithelioid sarcoma does not harbor a specific and reproducible chromosomal translocation, although other genetic events have been identified. 7,8 Increased knowledge of cytogenetics and molecular genetics of soft tissue sarcomas has led to the redefining of tumor categories and classifications. 3,10,11,12,13 Currently the World Health Organization (WHO) classification (2013) integrates morphology with genetics, whereas in the past soft tissue tumor classifications were principally based on morphology. 3 New entities have become established and other entities gradually disappeared. Myxoid liposarcoma and round cell liposarcoma (discussed in this thesis) are good examples of redefined categories: in the past these two tumors were regarded as different entities, but their shared chromosomal translocation has established that these tumors represent the ends of a morphological spectrum of one and the same tumor entity. Round cell liposarcoma is now regarded as a high grade type of myxoid liposarcoma. 3,14,15 In chapter 4 we have compared the prognostic impact and differences between these two subtypes of myxoid liposarcoma. In addition to paucicellular myxoid liposarcoma and round cell liposarcoma we investigated whether the histological transitional type of myxoid liposarcoma, which harbors an increased cellular pattern, is associated with adverse prognosis. The chromosomal translocations and their fusion products serve as useful and reliable diagnostic markers that can be detected with reverse transcriptase-polymerase chain reaction (RT-PCR) and fluorescence in situ hybridization (FISH). These techniques can be applied on relatively small amounts of tumor material, usually even in formalin 10 11

9 General introduction 1 fixed paraffin-embedded specimens and are therefore very well applicable in routine practice as well as research settings In chapter 2 of this thesis we have studied and compared the accuracy of RT-PCR and FISH in diagnosing synovial sarcoma. Both techniques turned out being highly reliable methods for diagnosing synovial sarcoma. Prognostification Improved diagnostics of soft tissue sarcoma have had a significant impact on prognostic studies of soft tissue sarcoma-types. The quality of prognostic studies has improved tremendously with the stratification of soft tissue sarcoma-types. 21,22,23 For example, before 1984 synovial sarcoma was mainly diagnosed as a soft tissue sarcoma showing biphasic histology. Lateron, the advent of immunohistochemistry of keratin intermediate filaments and molecular biologic techniques led to the recognition of monophasic synovial sarcoma and undifferentiated synovial sarcoma. The same holds for myxoid / round cell liposarcoma, as discussed above. In this thesis we have studied classical clinicopathological prognostic factors for synovial sarcoma, myxoid liposarcoma, epithelioid sarcoma and dermatofibrosarcoma protuberans specifically. The histopathological diagnosis of myxoid liposarcoma, epithelioid sarcoma and dermatofibrosarcoma protuberans is usually highly accurate, but in our series of 50 synovial sarcoma patients the tumor material of all patients was verified to bear the synovial sarcoma specific translocation by studying the diagnostic accuracy of RT-PCR and FISH in an earlier study (reported in chapter 2 of this thesis), which has significantly improved the reliability of our prognostic evaluation. The most important classical clinicopathological prognostic parameters are tumor grading and staging. Grading is applicable to most but not all soft tissue sarcomas. Grading of untreated primary soft tissue sarcomas is based on histologic parameters, such as tumor differentiation, mitotic activity and amount of necrosis. Grading is an important indicator of the malignant propensity of the tumor and a predictor of distant metastases and patient survival The two most widely used grading systems are the one of the National Cancer Institute (NCI) and the French Fédération Nationale des Centres de Lutte Contre le Cancer (FNCLCC). 28,29 Both are 3-grade systems and are mainly based on histologic subtype, tumor necrosis, and mitotic activity. There are remarkable discrepancies in these two commonly used grading systems. For example, the FNCLCC system recognizes synovial sarcoma grade 2 and grade 3, whereas the NCI system regards all synovial sarcomas as high grade tumors. 30,31 We have evaluated the value of grading in our series of synovial sarcomas, as discussed in chapter 3 of this thesis. Staging of soft tissue sarcoma is a tool for displaying the amount of tumor spread locally and systemically. Several different staging systems are used for soft tissue saroma, like the AJCC American Joint Committee on Cancer (AJCC/ International Union Against Cancer (UICC). Staging systems generally incorporate histologic grade, tumor size, depth of the tumor, lymph node metastases, and the presence or absence of distant metastases. 32 Using tumor grading and tumor staging parameters, researchers from the Memorial Sloan-Kettering Cancer Center in New York have developed a statistically based nomogram as a tool for predicting the probability that a patient will die of a soft tissue sarcoma. 33 The most common types of soft tissue sarcomas are integrated in their sarcoma nomogram that incorporates well-established predictors of survival. The predictive value of the nomogram will be discussed for myxoid liposarcoma in chapter 4 of this thesis. The staging and grading of soft tissue sarcoma types is far from fully explored and will remain under intensive investigation. Besides the ongoing evaluation of classical prognostic factors there is a persisting search for new factors. The advanced understanding of the molecular biology of soft tissue sarcoma leads researchers to the identification of new biomarkers that may have an impact on the survival of sarcomas and may even lead to the development of new targeted treatments. Some of the new prognostic factors are derived from the genetic characteristics of soft tissue sarcomas. For example it has been reported by others that the two different SYT-SSX translocation variants in synovial sarcoma are correlated with patient survival, but two relatively large studies have reported opposite conclusions about the predictive value of SYT-SSX fusion type. 34,35 We have analysed the prognosis of patients with the SYT-SSX1 versus SYT- SSX2 fusion-type in our series of 45 synovial sarcomas. Ezrin is an example of a biomarker that has been shown to be associated with early metastasis in different tumor types, including osteosarcoma and rhabdomyosarcoma. Ezrin is a cell membrane-cytoskeletal linker protein which plays a key role in the coordination of signals and cellular complexes required for the development of metastasis We hypothesized and investigated a possible prognostic impact of the immunohistochemical expression of ezrin in synovial sarcomas

10 General introduction Multimodality treatment of patients with local disease at presentation The therapeutic approach of soft tissue sarcomas depends on the disease stage at presentation. In this thesis we have mainly evaluated the treatment of patients with local disease at presentation, which generally consists of adequate local excision of the primary tumor and adjuvant radiotherapy in case of narrow resection margins. Soft tissue sarcomas often present as locally advanced large masses and / or at unfavourable locations, thereby restricting the assessment of optimal surgical margins. Local disease control of soft tissue sarcomas has improved tremendously during the last three decades due to advanced imaging techniques, improved surgical techniques, adjuvant radiotherapy and isolated limb perfusion. Although the four series of soft tissue sarcomas in this thesis are of limited size and collected retrospectively, they comprise collected cases with long term follow up. The impact of surgical margins and adjuvant radiotherapy on local disease control and patient survival is evaluated and discussed with respect to the existing literature. References Toro JR, Travis LB, Wu HJ, Zhu K, Fletcher CD, Devesa SS. Incidence patterns of soft tissue sarcomas, regardless of primary site, in the surveillance, epidemiology and end results program, : An analysis of 26,758 cases. Int J Cancer. 2006;19: Fletcher, CD, Bridge JA, Hogendoorn P et al. (eds). WHO Classification of Tumours of Soft Tissue and Bone. 4th ed. Lyon: IARC; Fletcher CD. The evolving classification of soft tissue tumours: an update based on the new WHO classification Histopathology 2006;48(1): de Silva MV, McMahon AD, Paterson L, et al. Identification of poorly differentiated synovial sarcoma: a comparison of clinicopathological and cytogenetic features with those of typical synovial sarcoma. Histopathology. 2003;43: Enzinger FM, Weiss SM. Liposarcoma. In: Enzinger FM, Weiss SM, eds, Soft Tissue Tumors. St. Louis; Mosby 2001: Enzinger FM. Epithelioid sarcoma. A sarcoma simulating a granuloma or a carcinoma. Cancer. 1970;26: Weiss SW, Goldblum JR. Epithelioid sarcoma. In: EnzingerFM, WeissSM, editors. Soft Tissue Tumors. 4th ed. St. Louis: Mosby; 2001: Kleihues P, Sobin L, Dermatofibrosarcoma protuberans. In: LeBoit PE, Burg G, Weedon D, Sarasin A. World Health Organization Classification of Tumours: pathology and Genetics of Skin Tumors. Lyon: IARC Press; 2007; Ladanyi M, Bridge JA. Contribution of molecular genetic data to the classification of sarcomas. Hum Pathol. 2000;31(5): Nielsen TO, West RB, Linn SC, Alter O, Knowling MA, O Connel JX, et al. Molecular characterisation of soft tissue tumours: a gene expression study. Lancet 2002;359(9314): Mertens M, Panagopoulos J, Mandahl N. Genomic characteristics of soft tissue sarcomas. Virchows Arch. 2010;456: Bovée JV, Hogendoorn PC. Molecular pathology of sarcomas: concepts and clinical implications. Virchows Arch. 2010;456: Panagopoulos I, Mandahl N, Ron D, et al. Characterization of the CHOP breakpoints and fusion transcripts in myxoid liposarcomas with the 12;16 translocation. Cancer Res 1994;54: Knight JC, Renwick PJ, Cin PD, Van den Berghe H, Fletcher CD. Translocation t(12;16)(q13;p11) in myxoid liposarcoma and round cell liposarcoma: molecular and cytogenetic analysis. Cancer Res 1995;55: Guillou L, Coindre J, Gallagher G, et al. Detection of the synovial sarcoma translocation t(x;18) (SYT;SSX) in paraffin-embedded tissues using reverse transcriptasepolymerase chain reaction: a reliable and powerful diagnostic tool for pathologists. A molecular analysis of 221 mesenchymal tumors fixed in different fixatives. Hum Pathol. 2001;32: Hostein I, Menard A, Bui BN, et al. Molecular detection of the synovial sarcoma translocation t(x;18) by real-time polymerase chain reaction in paraffin-embedded material. Diagn Mol Pathol. 2002;11: Tsuji S, Hisaoka M, Morimitsu Y, et al. Detection of SYT-SSX fusion transcripts in synovial sarcoma by reverse transcription-polymerase chain reaction using archival paraffin-embedded tissues. Am J Pathol. 1998;153: Terry J, Barry TS, Horsman DE, et al. Fluorescence in situ hybridization for the detection of t(x;18)(p11.2;q11.2) in a synovial sarcoma tissue microarray using a breakapart-style probe. Diagn Mol Pathol. 2005;14: Amary MF, Berisha F, Bernardi FC, et al. Detection of SS18-SSX fusion transcripts in formalin-fixed paraffin-embedded neoplasms: analysis of conventional RT- PCR, qrt-pcr and dual color FISH as diagnostic tools for synovial sarcoma. Mod Pathol. 2007;20:

11 General introduction Fritsch MK, Bridge JA, Schuster AE, et al. Performance characteristics of a reverse transcriptase-polymerase chain reaction assay for the detection of tumor-specific fusion transcripts from archival tissue. Pediatr Dev Pathol. 2003;6: Geurts van Kessel A, de Bruijn D, Hermsen L, et al. Masked t(x;18)(p11;q11) in a biphasic synovial sarcoma revealed by FISH and RT-PCR. Genes Chromosomes Cancer. 1998;23: Coindre JM. Grading of soft tissue sarcomas: review and update. Arch Pathol Lab Med. 2006; 130: Deyrup AT, Weiss SW. Grading of soft tissue sarcomas: the challenge of providing precise information in an imprecise world. Histopathology. 2006; 48: Borden EC, Baker LH, Bell RS, Bramwell V, et al. Soft tissue sarcomas of adults: state of the translational science. Clin Cancer Res. 2003;9: Oliveira AM, Nascimento AG Grading in soft tissue tumors: principles and problems. Skeletal Radiol. 2001;30(10): Coindre, J. M., P. Terrier, and L. Guillou. et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer 2001;91: Costa J, Wesley RA, Glatstein E, et al: The grading of soft tissue sarcomas. Results of a clinicohistopathologic correlation in a series of 163 cases. Cancer 1984;53: Trojani M, Contesso G, Coindre JM, et al: Soft-tissue sarcomas of adults: Study of pathological prognostic variables and definitions of a histopathological grading system. Int J Cancer 1984;33: Costa, J. The grading and staging of soft tissue sarcomas. In: Pathobiology of soft tissue tumors, Fletcher, CD, McKee, PH, (Eds), Churchill Livingstone, Edinburgh, p Coindre JM. Grading of soft tissue sarcomas: review and update. Arch Pathol Lab Med. 2006;130(10): AJCC (American Joint Committe on Cancer) Cancer Staging Manual. 7th ed., ed. S.B. Edge, Byrd, D.R., Compton, C.C, et al. 2010, Springer: New York. p Kattan MW, Leung DH, Brennan MF. Postoperative nomogram for 12-year sarcoma-specific death. J Clin Oncol 2002;20(3): Ladanyi M, Antonescu CR, Leung DH, et al. Impact of SYT-SSX fusion type on the clinical behavior of synovial sarcoma: a multi-institutional retrospective study of 243 patients. Cancer Res. 2002;62: Guillou L, Benhattar J, Bonichon F et al. Histologic grade, but not SYT-SSX fusion type, is an important prognostic factor in patients with synovial sarcoma: a multicenter, retrospective analysis. J Clin Oncol 2004;22: Hunter KW. Ezrin, a key component in tumor metastasis. Trends Mol Med. 2004;10: Khanna C, Wan X, Bose S, et al. The membrane-cytoskeleton linker ezrin is necessary for osteosarcoma metastasis. Nat Med. 2004;10: Yu Y, Khan J, Khanna C, Helman L, Meltzer PS, Merlino G. Expression profiling identifies the cytoskeletal organizer ezrin and the developmental homeoprotein Six-1 as key metastatic regulators. Nat Med. 2004;10: Weng WH, Ahlén J, Aström K, Lui WO, Larsson C. Prognostic impact of immunohistochemical expression of ezrin in highly malignant soft tissue sarcomas. Clin Cancer Res. 2005;11:

12

13 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas 2

14 Abstract Background Molecular detection of SYT-SSX fusion genes is the most reliable tool for diagnosing synovial sarcoma (SS). The objective of this study was to investigate the diagnostic accuracy of RT-PCR and a commercially available FISH technique for formalin-fixed and paraffin-embedded (FFPE) tumor tissue. Patients and Methods Fifty tumors with typical SS histology and 12 histologic mimics of SS were included. RT-PCR for SYT-SSX1/SSX2 gene fusions and FISH analysis for SYT gene breaks were performed on these 62 FFPE tumors. Results All 50 SS were positive by either RT-PCR or FISH. Forty-seven SS (94%) were positive by RT-PCR and 41 SS (82%) were positive by FISH. FISH and RT-PCR results were interpretable and concordant in 38 cases (76%). Two cases were not interpretable by RT-PCR and 6 cases were not interpretable by FISH. One SS was false-negative with RT-PCR and 3 SS were false-negative with FISH. RT-PCR and FISH had a sensitivity of 94% and 82%, a specificity and positive predictive value of 100% and 100% and a negative predictive value of 80% and 75%, respectively. Conclusion RT-PCR had a higher sensitivity than FISH. One of both methods was always positive, whereas both methods were concordant in 76% of cases. From an economic point of view, we advocate to use FISH as a method of first choice, because it allows microscopic control of a true positive result (unpaired fluorescent signals in a break apart assay). Using this approach, 80% of SS can be diagnosed by FISH only and 20% would need to be confirmed by RT-PCR. S.E. ten Heuvel H.J. Hoekstra A.J.H. Suurmeijer Applied Immunohistochemistry & Molecular Morphology May;16(3): Introduction Synovial sarcoma (SS) accounts for 5% to 10% of all soft tissue sarcomas. Although SS has a wide range of age at presentation and may occur at quite unusual localizations, the tumor most commonly presents as a deep-seated tumor in the extremity of a young adult patient. The histologic spectrum of SS encompasses biphasic SS, monophasic SS, and poorly differentiated SS. Biphasic SS is differentiated from the 2 other SS subtypes by the presence of (glandular) epithelial differentiation, a pattern it shares with carcinosarcoma and glandular malignant peripheral nerve sheath tumor (MPNST). Monophasic SS and poorly differentiated SS have to be differentiated from other types of monotonous spindle cell sarcomas and round cell sarcomas, in particular MPNST, fibrosarcoma, solitary fibrous tumor, and Ewing sarcoma/peripheral neuroectodermal tumor (PNET). 1,2 At least 95% of all SSs bear a unique chromosomal translocation, which results in a fusion of the SYT gene on chromosome 18 with either the SSX1 gene or the SSX2 gene or, more rarely, the SSX4 gene on the X chromosome. Because these gene fusions are highly specific for SS, their detection with molecular genetics allows the pathologist to render a correct diagnosis. 3,4 Molecular genetic methods will often be applied when SS has an unusual clinical presentation or when the diagnostic tumor sample does not allow easy differentiation from other spindle cell or round cell sarcomas, which is often the case with small biopsies. 5,6 For application with formalin-fixed and paraffin-embedded (FFPE) tissue, the 2 molecular biologic methods currently available are reverse transcription-polymerase chain reaction (RT-PCR) and fluorescence in situ hybridization (FISH). Compared with RT-PCR, FISH has many practical advantages, in particular for pathology laboratories without expensive molecular pathology facilities. FISH is relatively easy, quick, inexpensive, and suitable for automation. Moreover, because RT-PCR may give false-positive test results due to tissue contamination, many pathologists favor to use FISH, because this allows visual control of the presence of specific translocations. Using a break apart probe, for instance, SYT gene breaks in tumor cell nuclei can be visualized in histologic context. We herein describe our experience with the commercially available Vysis SYT break apart probe. In a retrospective study of 50 SS and 12 other soft tissue sarcomas potentially confused with SS, we compared the test sensitivity and specificity of this Vysis FISH method with that of a RT-PCR method proven to be very sensitive with FFPE tumor tissue samples. 2 21

15 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas Materials and Methods Patients and tumors Cases diagnosed as SS between 1984 and 2006 were retrieved from the archives of the Department of Pathology of the University Medical Center Groningen. Hematoxylin and eosin-stained and epithelial membrane antigen (EMA) immunostained slides were reviewed. Typical histopathologic features were used as the diagnostic gold standard for SS. Fifty tumors with a typical SS histology were included. Histologic subtyping was performed according to the definitions described by Guillou et al.7 Twenty-nine SS had monophasic, 6 biphasic, and 15 poorly differentiated histology. In 28 cases, the diagnosis of SS had been previously confirmed with either cytogenetics (11 cases) or with RT-PCR on fresh frozen tissue material with a consensus primer as described by Panagopoulos et al (17 cases).8 Twelve non-ss cases were included as negative controls, including 10 cellular spindle cell tumors (MPNST, solitary fibrous tumor, and fibrosarcoma) and 2 round cell sarcomas (1 Ewing sarcoma/pnet and 1 round cell sarcoma not otherwise specified). From each case 1 representative paraffin tumor tissue block containing the largest area of viable tumor tissue was selected for FISH and RT-PCR analysis. FISH Four-micrometer paraffin sections of selected tissue blocks were transferred to aminopropyltriethoxysilane-coated glass slides. Slides were dried overnight at 60 C, dewaxed in xylene, and dehydrated in descending concentrations of ethanol and MilliQ water. To make target DNA accessible to the DNA probes, the slides were heated in a pressure cooker in TRIS/ethylenediaminetetraacetic acid buffer, ph 9.0 at 120 C for 7 minutes and cooled twice in saline sodium citrate (SSC) at room temperature. The slides were incubated in 0.1 µg/l RNAse solution in SSC at 37 C for 1 hour, followed by three 5-minutes washes in SSC, after which they were placed in a 1 g/l pepsine MilliQ water solution for 10 minutes at ph 2 and 37 C, followed by three 5-minutes washes in phosphate-buffered saline. After RNAse pretreatment and pepsin pretreatment, slides were dehydrated in ascending concentrations of ethanol and air-dried. Hybridization was performed in a humidified chamber overnight at 37 C using the Vysis dual color break apart rearrangement probes suspended in Vysis LSI hybridization buffer according to the manufacturer s instructions (Vysis, Downers Grove, IL). Probes were a telomeric 650-kb probe labeled with Spectrum Orange and a centromeric 1040-kb probe labeled with Spectrum Green. These probes are separated by a gap of 56 kb within the SYT gene. After posthybridization washes, slides were dehydrated in ascending concentrations of ethanol, air-dried, covered with 4,6-diamidino-2-phenylindole in antifade mounting solution and coverslipped. Slides were visually evaluated using a Leica DMR fluorescence microscope, equipped with appropriate filters for 4,6-diamidino-2-phenylindole, Spectrum Orange, and Spectrum Green. Tumors (SS) scored positive for SYT translocations contained several areas where at least 15% of tumor cell nuclei had clearly split orange and green signals (separated by at least 3 signal diameters from an oppositely colored signal). Typical examples of a SS with split signals in many nuclei and a negative control with paired signals are shown in Figure 1. Figure 1. Typical examples of unpaired split signals in many nuclei of a SS (A) and paired signals in nuclei of a histologic mimic of SS (B)

16 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas RT-PCR for SYT-SSX1 and SYT-SSX2 Five-micron thick sections of the selected FFPE tumor tissue blocks were cut and placed in an RNA-free cup. Sections were dewaxed using 2 changes of xylene for 10 minutes. After centrifugation, the sections were rehydrated by 100% ethanol washes. RNA extraction was performed by adding lysis buffer containing 10 mmol/l Tris ph 8, 0.1 mmol/l ethylenediaminetetraacetic acid, 2% sodium dodecyl sulphate, and 500 µg/ml proteinase K, as described by Specht et al.9 Sections were incubated in lysis buffer at 60 C for 15 hours, after which proteinase K was heat-inactivated at 95 C for 5 minutes. Total RNA was precipitated by adding 1/10 volume of 2 M RNA-free sodium acetate ph 4 to 4.5, 1/5 volume chloroform, and 1 volume water saturated acidic phenol. This mixture was vortexed for 1 minute, placed on ice for 15 minutes, and centrifuged for 20 minutes at 13,000 rpm at 4 C. The upper aqueous phase was taken and precipitated with 1 µl of 10 mg/ml carrier glycogen and an equal volume of isopropanol. This was incubated at -20 C for 2 hours and centrifuged for 30 minutes. The pellet was washed with 70% alcohol and dissolved in H2O. Before reverse transcription, extracted RNA was treated with a Turbo DNA-free kit (Ambion). RNA was reverse transcribed using Superscript II reverse transcriptase (Invitrogen, Paisley, UK). PCR amplification was performed in a 30 µl final reaction volume containing 3 µl of 2 mm of each dntp, 3 µl with 10 µm SYT forward primer, 10 µm SSX reverse primer, 12.8 µl RNAse-free water, and 1 U of Taq polymerase (Applied Biosystems, Foster City, CA). Amplification conditions consisted of denaturation at 94 C for 5 minutes, extension by 35 cycles of 30 seconds at 94 C, annealing at 62 C for 30 seconds, and final extension at 72 C for 45 seconds. After the last cycle an extended period of 10 minutes at 72 C was followed by cooling to 4 C. The amplified products were analyzed by electrophoresis on a 1% agarose gel. The specific primers used were forward primer SYT (5 AGACCAACACAGCCTGGAC- CAC 3 ) and reverse primers SSX1-rev2 (5 ACACTCCCTTCGAATCATTTTCG 3 ) and SSX2-rev2 (5 GCACTTCCTCCGAATCATTTC 3 ). Appropriate positive and negative controls were included. To verify the presence of intact RNA and amplifiable cdna, each reversed transcriptase reaction product was also assayed with the following primers of phosphoglycerate kinase (PGK) PGK F (5 CAGTTTGGAGCTCCTGGAAG 3 ) and PGK R (5 TGCAAATC- CAGGGTGCAGTG 3 ). Results The results of RT-PCR for SYT-SSX1/2 gene fusions and FISH for SYT breaks in the 50 tumors with typical SS histology are summarized in Table 1. All 50 SS were positive by RT-PCR and/or FISH, confirming that these tumors were indeed SS. Forty-seven SS (94%) were positive by RT-PCR and 41 SS (82%) were positive by FISH. RT-PCR and FISH showed concordant results in 38 cases (72%). Two cases with monophasic histology were not interpretable by RT-PCR analysis due to poor RNA quality and 6 tumors showed no fluorescent signals and were not interpretable by FISH (5 monophasic and 1 poorly differentiated SS). One poorly differentiated case was false-negative with RT-PCR (but positive with FISH) and 3 monophasic cases were false-negative with FISH (but positive with RT- PCR). In all discordant and not interpretable cases both FISH and RT-PCR analysis were repeated, which rendered the same results. All 12 negative control samples were successfully analyzed by RT-PCR and all these lacked SYT-SSX1 or SYT-SSX2 fusion genes. FISH results were interpretable in 9 negative controls, in which no split signals were seen. Thus, RT-PCR and FISH had a sensitivity of 94% and 82%, a specificity and positive predictive value of 100% and 100%, and a negative predictive value of 80% and 75%, respectively. Table 1. Results of RT-PCR and FISH in 50 SS cases with typical histology SYT-SSX1/2 positive SYT-SSX1/2 negative SYT-SSX1/2 TF FISH positive FISH negative FISH TF Total TF indicates technical failure. Total

17 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas Discussion SS is a distinctive soft tissue sarcoma with a variable degree of epithelial differentiation that most commonly presents as a deep-seated sarcoma in the lower or upper extremity of a young adult patient. The biphasic type, which comprises about 20% to 30% of all SS, shows histologic evidence of (glandular) epithelial differentiation, whereby it is diagnosed relatively easy. On the other hand, to make a conclusive diagnosis of monophasic SS (50% to 60%) or poorly differentiated SS (15% to 25%), the surgical pathologist will often have to apply immunohistochemistry and molecular methods. Concerning the value of immunohistochemistry, it has been shown that, within a panel of antibodies including cytokeratins, EMA is the most sensitive and specific marker for SS. Nearly all biphasic and monophasic SS and at least 90% of poorly differentiated SS are immunoreactive for EMA. 10,11 However, EMA expression is often focal, which may be problematic when only small biopsies are available. Moreover, some sarcomas histologically mimicking SS, for example, MPNST, solitary fibrous tumor, and Ewing sarcoma/pnet, may show focal positive immunostaining for EMA. This is in accordance with our experience. Ninety percent of the 50 SS cases and 75% of the 12 diagnostically difficult non-ss cases in this study were EMA positive (data not shown). Clearly, EMA immunostaining is not absolutely specific and of limited value in these diagnostically challenging cases. To establish an accurate diagnosis of SS in these cases, molecular methods will have to be applied, taking advantage of the fact that at least 95% of SS bear a unique t(x;18) translocation, by which the SYT gene on chromosome 18 fuses with either the SSX1 or SSX2 or, very rarely, the SSX4 gene on the X chromosome. The t(x;18) translocation was first detected with conventional cytogenetics in As a diagnostic technique, conventional cytogenetics is only rarely used, because the method relies on tissue culture and chromosome banding of metaphases, by which it is labor intensive and failure prone. Nowadays, most pathology laboratories use molecular genetic methods, in particular RT-PCR and/or FISH, for the demonstration of the specific SYT-SSX gene fusions. At present, RT-PCR is the most commonly used technique. Although some RT-PCR protocols require frozen tissue for optimal results, sensitive RNA primers have also been designed for successful application with FFPE tumor samples. 5,13 16 A major disadvantage of RT-PCR is that tissue contamination in the molecular pathology laboratory may lead to false-positive results. The O Sullivan paper, in which falsepositive SYT-SSX fusions were found with RT-PCR in 75% of MPNST cases, among others, is often cited to stress the importance of quality control measures. 17,18 To avoid false-positive interpretation, demonstration of SYT-SSX gene fusions by combining RT-PCR with FISH is a good alternative. Compared with RT-PCR, FISH has the advantage of being a microscopic in situ technique. For instance, FISH break apart assays for the SYT gene, which are now commercially available and easy to interpret, allow the pathologist to see with his or her own eyes whether split fluorescent signals are present in tumor cell nuclei. In other words, break apart assays allow highly reliable visual control of SYT gene breaks and these gene breaks can be used as specific diagnostic markers for all known SS translocations. Technically, FISH is easy to perform, which makes it very appealing, in particular for pathology laboratories that do not have expensive molecular pathology facilities. In this study, we compared the sensitivity and specificity of FISH and RT-PCR as diagnostic molecular methods in routinely processed FFPE tissue samples of 50 cases of SS and 12 histologic mimics of SS. For FISH we used the commercially available Vysis break apart assay, which applies dual color fluorescent DNA probes flanking the SYT gene. We found that FISH (82%) is less sensitive than RT-PCR (94%). To date, only 2 studies have addressed the diagnostic accuracy of the commercially available FISH SYT break apart assay for diagnosing SS with FISH. Terry et al found a sensitivity of 81% in a tissue microarray study of 45 FFPE cases and Amary et al found a sensitivity of 86% in a tissue microarray study of 101 FFPE cases. 19,20 In contrast to Terry et al and Amary et al, we included the number of technical failure (TF) in calculating the accuracy of FISH. If we would not have included the TF, the sensitivity of FISH would be 93%. A meta-analysis of the diagnostic accuracy of our study and these 2 other studies is summarized in Table 2, using this approach (inclusion of TF). The sensitivity calculated from the meta-analysis amounts 80%. For RT-PCR we used primers designed by researchers from the Memorial Sloan Kettering Cancer Center, which allows sensitive detection of a 110-bp product of specific SYT-SSX1 and SYT-SSX2 fusion genes in FFPE SS samples. 4 Using this particular primer Amary et al found that RT-PCR had a sensitivity of 92%. 20 Moreover, similar sensitivities have been reported for other primers used to detect SYT-SSX or SSX1/SSX2 in FFPE material. The 94% sensitivity of RT-PCR in our 50 FFPE SS cases compares to that reported in these other studies (range: 92% to 96%). 14,16,20,21 All 50 tumors with typical SS histology were positive with either FISH and/or RT-PCR, indicating that these tumors are indeed SS. Although the sensitivity of both RT-PCR and FISH were relatively high, concordant FISH and RT-PCR results were found in only 76% of cases. This was mainly due to the rather high number of TF. FISH had a considerable higher rate of TF than RT-PCR (15% vs. 3%, respectively). Amary et al and Terry et al reported 7% and 12% TF with FISH, respectively. Amary et al reported a 2% TF for RT-PCR

18 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas Table 2. Accuracy of the Vysis FISH SYT break part assay for molecular confirmation of SS in comparison with the results of two other studies Study Ten Heuvel et al Terry et al 19 Amary et al 20 Meta- Analysis No. SS cases* No. non-ss cases TP TN FP FN including TF TF (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy TF are included in calculating the sensitivity, specificity, PPV, NPV, and accuracy. * Only SS with an RT-PCR confirmed translocation were evaluated. FN indicates false negative; FP, false positive; NPV, negative predictive value; NR, not relevant; PPV, positive predictive value; TF, technical failure; TN, true negative; TP, true positive. The 3 false-negative FISH results, meaning that paired in stead of split dual color fluorescence signals were observed in the presence of a SYT-SSX1 or SYT-SSX2 fusion gene with RT-PCR, may be explained by incidental cryptic or masked chromosomal rearrangements as described in the literature. 22,23 At least in 2 of these 3 false-negative FISH cases, a false-positive RT-PCR result could be ruled out, because these 2 SS were also positive with RT-PCR using another consensus SYT-SSX primer with frozen samples. No fresh frozen tissue material of the third discordant case was available. We faced 1 false-negative RT-PCR result, which may have been due to the presence of a translocation other than SYT-SSX1 or SYT-SSX2, for example, SYT-SSX4, which could be detected with FISH but not with the PCR primers used in this study. Amary et al also reported 8% (11/131) false-negative RT-PCR results. Six out of 11 false-negatives in their series proved to be true SS after application of a quantitative RT-PCR method and/or FISH. Five other SYT-SSX fusion gene negative cases did not show SYT-SSX with alternative primers (including primers for rare variants). After review of the clinical and histopathologic characteristics the investigators eventually concluded that other diagnosis than SS were still probable in these 5 cases. In our meta-analysis in Table 2 we regarded these 5 false-negative cases as true negative cases. In accordance with other studies, false-positive results were not found. Both methods had a specificity and positive predictive value of 100%. 19,20 In conclusion, when used with FFPE material, RT-PCR has a higher sensitivity than FISH. In this series of 50 SS, one of both methods was always positive, whereas both methods showed concordant results in 76% of cases. From an economic point of view, we advocate to use FISH as an initial technique, because it allows microscopic control of a true positive result (unpaired fluorescent signals in a break apart assay). Using this approach, 80% of SS can be diagnosed by FISH only and 20% would need additional confirmation by RT-PCR. References 1. Fisher C, Ladanyi M. Synovial sarcoma. In: Fletcher CDM, Unni KK, Mertens F, eds. World Health Organization: Classification of Tumours. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon: IARC Press; 2002: de Silva MV, McMahon AD, Paterson L, et al. Identification of poorly differentiated synovial sarcoma: a comparison of clinicopathological and cytogenetic features with those of typical synovial sarcoma. Histopathology. 2003;43: Ladanyi M. Fusions of the SYT and SSX genes in synovial sarcoma. Oncogene. 2001;20: Antonescu CR, Kawai A, Leung DH, et al. Strong association of SYT-SSX fusion type and morphologic epithelial differentiation in synovial sarcoma. Diagn Mol Pathol. 2000;9: Argani P, Faria PA, Epstein JI, et al. Primary renal synovial sarcoma: molecular and morphologic delineation of an entity previously included among embryonal sarcomas of the kidney. Am J Surg Pathol. 2000;24: Coindre JM, Pelmus M, Hostein I, et al. Should molecular testing be required for diagnosing synovial sarcoma? A prospective study of 204 cases. Cancer. 2003;98:

19 Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas 7. Guillou L, Benhattar J, Bonichon F, et al. Histologic grade, but not SYT-SSX fusion type, is an important prognostic factor in patients with synovial sarcoma: a multicenter, retrospective analysis. J Clin Oncol. 2004;22: Panagopoulos I, Mertens F, Isaksson M, et al. Clinical impact of molecular and cytogenetic findings in synovial sarcoma. Genes Chromosomes Cancer. 2001;31: Specht K, Richter T, Muller U, et al. Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffin-embedded tumor tissue. Am J Pathol. 2001;158: Pelmus M, Guillou L, Hostein I, et al. Monophasic fibrous and poorly differentiated synovial sarcoma: immunohistochemical reassessment of 60 t(x;18)(syt- SSX)-positive cases. Am J Surg Pathol. 2002;26: Olsen SH, Thomas DG, Lucas DR. Cluster analysis of immunohistochemical profiles in synovial sarcoma, malignant peripheral nerve sheath tumor, and Ewing sarcoma. Mod Pathol. 2006;19: Smith S, Reeves BR, Wong L, et al. A consistent chromosome translocation in synovial sarcoma. Cancer Genet Cytogenet. 1987;26: Antonescu CR. The role of genetic testing in soft tissue sarcoma. Histopathology. 2006;48: Guillou L, Coindre J, Gallagher G, et al. Detection of the synovial sarcoma translocation t(x;18) (SYT;SSX) in paraffin-embedded tissues using reverse transcriptasepolymerase chain reaction: a reliable and powerful diagnostic tool for pathologists. A molecular analysis of 221 mesenchymal tumors fixed in different fixatives. Hum Pathol. 2001;32: Hostein I, Menard A, Bui BN, et al. Molecular detection of the synovial sarcoma translocation t(x;18) by real-time polymerase chain reaction in paraffin-embedded material. Diagn Mol Pathol. 2002;11: Tsuji S, Hisaoka M, Morimitsu Y, et al. Detection of SYT-SSX fusion transcripts in synovial sarcoma by reverse transcription-polymerase chain reaction using archival paraffin-embedded tissues. Am J Pathol. 1998;153: O Sullivan MJ, Kyriakos M, Zhu X, et al. Malignant peripheral nerve sheath tumors with t(x;18). A pathologic and molecular genetic study. Mod Pathol. 2000;13: Ladanyi M, Woodruff JM, Scheithauer BW, et al. Re: O Sullivan MJ, Kyriakos M, Zhu X, Wick MR, Swanson PE, Dehner LP, Humphrey PA, Pfeifer JD: malignant peripheral nerve sheath tumors with t(x;18). A pathologic and molecular genetic study. Mod Pathol. 2000;13: Mod Pathol. 2001;14: Terry J, Barry TS, Horsman DE, et al. Fluorescence in situ hybridization for the detection of t(x;18)(p11.2;q11.2) in a synovial sarcoma tissue microarray using a breakapart-style probe. Diagn Mol Pathol. 2005;14: Amary MF, Berisha F, Bernardi FC, et al. Detection of SS18-SSX fusion transcripts in formalin-fixed paraffin-embedded neoplasms: analysis of conventional RT- PCR, qrt-pcr and dual color FISH as diagnostic tools for synovial sarcoma. Mod Pathol. 2007;20: Fritsch MK, Bridge JA, Schuster AE, et al. Performance characteristics of a reverse transcriptase-polymerase chain reaction assay for the detection of tumor-specific fusion transcripts from archival tissue. Pediatr Dev Pathol. 2003;6: Geurts van Kessel A, de Bruijn D, Hermsen L, et al. Masked t(x;18)(p11;q11) in a biphasic synovial sarcoma revealed by FISH and RT-PCR. Genes Chromosomes Cancer. 1998;23: Mathew S, Dalton J, Riedley S, et al. Complex t(x;18)(p11.2;q11.2) with a pericentric inversion of the X chromosome in an adolescent boy with synovial sarcoma. Cancer Genet Cytogenet. 2002;132:

20

21 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression 3

22 Abstract Background The aim of this study was to investigate prognostic factors influencing the survival of synovial sarcoma, including the debated role of SYT-SSX fusion type and the newly suggested immunohistochemical marker ezrin. Patients and Methods From 1984 to 2005, 45 patients 25 men (56%) and 20 women (44%) with a median age of 31 (range: 2 to 81) years were diagnosed with a synovial sarcoma. Age at diagnosis, tumor site, tumor size, tumor histology (biphasic vs. monophasic), mitotic count, necrosis, histologic grade, SYT-SSX fusion type, and ezrin immunostaining were analyzed for influence on survival by univariate and multivariate methods. Results The median follow-up for all patients was 55 (range: 2 to 238) months. Five patients had metastatic disease at the time of presentation. Five-year disease-specific survivals (DSS) were 67% overall and 72% for the 40 patients presenting with localized disease at diagnosis. Nineteen patients (48%) developed metastases during follow-up. Five-year metastasis-free survival (MFS) for the 40 patients with localized disease at diagnosis was 60% and the 10-year MFS was 52%. Disease stage at presentation, tumor size > 5 cm, and histologic grade 3 were univariate significant factors associated with a worse DSS. Age 30 years, tumor size > 5 cm, necrosis, and histologic grade were univariate significant factors associated with a worse MFS. In multivariate analysis, tumor size and tumor grade remained significant prognostic factors for DSS and MFS. A role of SYT-SSX fusion type could not be confirmed in our patient group. Ezrin showed high expression in glandular and nonglandular epithelioid components in biphasic synovial sarcoma. Variable expression was found in the mesenchymal component of monophasic and biphasic synovial sarcoma. Low versus high ezrin expression levels in monophasic and/or biphasic synovial sarcoma did not correlate with patient outcome. Conclusion Disease stage at presentation, tumor size, and tumor grade were significant predictors of survival in synovial sarcoma. SYT-SSX fusion type was not correlated with survival in our series. Ezrin expression levels were not discriminative in predicting outcome. S.E. ten Heuvel H.J. Hoekstra E. Bastiaannet A.J.H. Suurmeijer Applied Immunohistochemistry & Molecular Morphology May;17(3): Introduction Synovial Sarcoma (SS) is a high-grade malignant soft tissue tumor accounting for 5 to 10% of all soft tissue sarcomas. Despite improved management of local disease, the high metastatic potential of SS remains the predominant cause of poor outcome, as more than half of SS patients die of chemotherapy-resistant metastatic disease. SS can arise almost anywhere in the body and at any age, but presents most frequently as a deep-seated tumor in the extremity of a young adult patient. Histologically, SS is either a monophasic spindle cell lesion or a biphasic lesion containing spindle cell and epithelioid/ glandular areas. Over 95% of SS are marked by a unique and specific chromosomal translocation that leads to a fusion between the SYT gene on chromosome 18 with either the SSX1 gene or the SSX2 gene on chromosome X. 1 The discovery of this specific chromosomal translocation led to a major advance in the accuracy of diagnosing SS. With FISH or RT-PCR, SS can nowadays be identified highly reliably. 2,3,4,5 In contrast to the improvements achieved in diagnosing SS, there still is a need for better prognostification and more effective treatment modalities. 6,7,8 New immunohistochemical and molecular markers are being investigated intensively, with the hope that they will be of prognostic value and might even lead to the development of a new targeted treatment. 7 Despite intensive efforts, convincing new markers have not been found for SS so far. Furthermore, no true consensus has been reached about some of the classic prognostic factors. Some studies report tumor grade to be the most important indicator of prognosis, others regard all SS as high grade tumors and do not differentiate between grade 2 or grade 3 tumors. 9,10 The prognostic impact of SYT-SSX fusion type is still a matter of debate Two large multi-institutional series show conflicting results with regard to the predictive role of the SYT-SSX fusion type. 9,10 In this study we evaluated classic prognostic factors including tumor grade and SYT- SSX fusion type in 45 patients treated for SS at the University Medical Center Groningen between 1984 and In addition to previously investigated prognostic factors, we also analysed the impact of ezrin expression by immunostaining. Ezrin is a cell membrane cytoskeletal linker protein. Upon phosphorylation, ezrin is activated and tethers F-actin to the cell membrane. Ezrin plays a key role in the coordination of signals and cellular complexes that are required for the development of metastasis in several tumor types. 14 It has been shown that high ezrin expression levels promote metastatic behaviour of sarcomas, in particular osteosarcoma and rhabdomyosarcoma. 15,16 To date, ezrin has not been studied systematically in SS. 3 35

23 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Patients and Methods Patient selection This study includes 45 patients that were treated for a SS at the University Medical Center Groningen between 1984 and In all cases the histological diagnosis of SS was confirmed molecularly by FISH and or RT-PCR. Pre-treatment tumor tissue of all patients was analyzed for SYT-SSX fusion type as described previously. 2 Clinical and pathologic features were retrieved from the Groningen Sarcoma Database, patient files and the pathology database. The 45 patients include 25 men (56%) and 20 women (44%) with a median age of 31 (range, 2-81) years. The median tumor size was 9 (range ) cm. Patient and tumor characteristics are listed in Table 1 and Table 2. Treatment Wide local excision with histologic confirmed free margins was the preferred surgical approach. In case of marginal or positive resection margins with no opportunity for further surgical management due to anatomic constraints, postoperative radiation (60-70 Gy) was administered. Ten patients with extremity based tumors received hyperthermic isolated limb perfusion (HILP) with tumor necrosis factor α and melphalan before resection of the tumor. In one case the tumor was not resected after the HILP procedure, because the patient was diagnosed with an incurable bowel malignancy during follow-up. Two patients received chemotherapy for the primary tumor as part of their curative treatment. The first patient was a 34 year old male with a 13 cm large tumor on the shoulder wherefore adjuvant DIME courses (consisting of doxorubicin, ifosfamid, mesna, and etoposide) were administered. The second patient, an 8-year-old boy with a relatively small tumor on the foot, received adjuvant ifosfamide, etoposide and vincristine. Five patients presented with metastasis at the time of diagnosis, of whom four patients had a primary tumor in the lower extremity and one patient had an intra-thoracic tumor. Histology Hematoxylin and eosin-stained sections of tumor biopsies and/or tumor resections were reviewed by an experienced sarcoma pathologist (A.J.H.S.) for histologic subtyping, mitotic count, tumor necrosis and tumor grade. Tumor grade was determined according to the definitions described by the French Federation of Cancer Centers Sarcoma Group grading system, as described by Guillou et al. 17 Histologic subtyping was performed according to the definitions in the World Health Organization classification of soft tissue tumors. 1 Ezrin immunostaining Paraffin sections were deparaffinized in xylene and rehydrated through graded alcohols to distilled water. Antigen retrieval was performed by microwaving slides for 15 minutes in 0.01 M citrate buffer (ph 6.0) and M ethylene diaminetetra acetic acid buffer (ph 8.0), respectively. Endogenous peroxidase was blocked with 3% hydrogen peroxide in phosphate-buffered saline (PBS). The sections were incubated with primary antibody diluted in 1% bovine serum albumin/pbs for 1 hour at room temperature. The primary antibody for the demonstration of ezrin was purchased from Sigma-Aldrich (St Louis, MO) and diluted at 1:2000. After washing with PBS, slides were incubated with the appropriate biotinylated secondary antibody (1:300; ABC complex Dako Cytomation, Glostrup, Denmark) for 30 minutes, followed by incubation with ABC complex DakoCytomation for 30 minutes. Peroxidase activity was visualized by incubating the slides with 3,3-diaminobenzidine substrate (DAB; Sigma-Aldrich, St Louis, MO) solution (0.05% DAB, 0.1% imadizol, 0.03% hydrogen peroxide in PBS). Sections were counterstained with hematoxylin. Tissue of a first trimester placenta served as a positive control. Analysis Immunostaining of ezrin in monophasic and biphasic SS was scored by intensity as either negative (0), weak positive (1+), moderately positive (2+), or strong positive (3+), as ezrin immunostaining was diffuse in all tumors. Staining of epithelial and mesenchymal components in biphasic SS were scored separately. For statistical analysis, we separated SS in 2 groups, tumors with low (0 and 1+) expression levels and tumors with high (2+ and 3+) expression levels of ezrin in the spindle cell component of monophasic and biphasic SS. Age at diagnosis, tumor location (extremity vs. nonextremity), tumor size, presence of metastases at presentation, histologic subtype (biphasic vs. monophasic), mitotic count, tumor necrosis, tumor grade, SYT-SSX fusion type, and ezrin expression levels were analyzed for their prognostic value. DSS curves were calculated with the Kaplan-Meier method and Fisher exact tests were used to assess associations between factors. Prognostic factors were analyzed by the log-rank test and univariate significant factors were subsequently analyzed in a multivariate fashion using Cox proportional hazard regression model. 18,

24 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Table 1. Clinicopathological features of the 45 synovial sarcomas treated at the UMCG Metastasis at months FU DSS (mo) Metastasis at Diagnosis Margins Grade Histotype Treatment primary tumor Site (size (cm)) Age (yr)/ sex Patient No. 1 33/M Knee (5) resection, RT neg 2 Monophasic SSX2 No NED (165) /M Thigh (10.30) resection, RT neg 2 Biphasic SSX2 No DOD (37) /F Groin (5) resection, RT neg 2 Biphasic SSX2 No NED (126) 4 41/F Knee (9) HILP, resection, RT pos 3 Monophasic SSX2 No NED (55) 5 3/M Thigh (3) resection neg 2 Monophasic SSX1 No NED (123) 6 25/M Knee (3) resection, RT neg 2 Biphasic SSX1 No NED (170) 7 29/M Knee (20) HILP, resection, RT neg 2 Monophasic SSX1 No NED (100) 8 65/F Foot (NA) HILP, resection, RT neg 2 Biphasic SSX1 No NED (17) 9 43/F Abdomen (NA) resection NA 2 Monophasic SSX1 No DOC (266) /F Head and neck (4) resection, RT NA 2 Monophasic SSX2 No AWD (60) /F Foot (4) resection neg 2 Biphasic SSX1 No NED (60) 12 30/M Head and neck (6) resection, RT pos 2 Monophasic SSX1 No DOD (29) /M Foot (9) resection neg 3 Monophasic SSX1 No AWD (8) /M Abdomen (2.80) resection neg 2 Monophasic SSX1 No NED (37) 15 24/M Thorax (6) resection, RT pos 2 Monophasic SSX2 No DOD (72) /F Upper arm (4) resection, RT neg 3 Monophasic SSX1 No DOD (112) /M Schoulder (13) resection, RT, chemo neg 3 Monophasic SSX1 No DOD (37) /F Knee (4) resection, RT neg 2 Monophasic SSX1 No DOD (34) /M Thorax (5.40) resection neg 3 Monophasic SSX1 No DOD (18) /M Upper arm (4) resection pos 2 Monophasic SSX2 No DOD (41) /M Thigh (6) HILP, resection, RT neg 3 Monophasic SSX2 No DOD (44) /M Shoulder (2.50) resection, RT NA 2 Monophasic SSX1 No DOD (26) /M Distal leg (3) HILP, resection, RT neg 2 Monophasic SSX1 No NED (131) Metastasis at months FU DSS (mo) Metastasis at Diagnosis Fusiontye Fusiontye Margins Grade Histotype Treatment primary tumor Site (size (cm)) Age (yr)/ sex Patient No /F Foot ( 5) resection, RT pos 2 Monophasic SSX1 No NED (20) 25 25/F Knee (5) HILP, resection, RT neg 2 Monophasic SSX1 No NED (146) 26 40/F Foot (1.60) resection, RT pos 2 Monophasic SSX2 No NED (22) 27 65/F Buttocks (12.50) resection, RT neg 2 Monophasic SSX2 No DWD (41) /F Foot (3) resection neg 2 Monophasic SSX2 No NED (131) 29 43/M Thigh (11) HILP, resection, RT neg 3 Monophasic SSX2 No NED (174) /F Head and neck (2.50) resection, RT neg 2 Monophasic SSX2 No NED (72) 31 52/F Distol Leg (NA) resection neg 2 Monophasic SSX1 No NED (97) 32 35/F Distal leg (6) resection, RT neg 2 Monophasic SSX2 No NED (102) 33 8/M Foot (4) resection, chemo neg 2 Monophasic SSX1 No NED (146) 34 31/F Distal leg (NA) resection neg 2 Biphasic SSX1 No DOD (283) /M Thorax (18) resection neg 3 Monophasic SSX1 No DOD (14) /F Elbow (8) HILP - 2 Monophasic SSX1 No DWD (48) 37 39/F Elbow (8) HILP, resection neg 3 Monophasic SSX2 No DOD (27) /F Thigh (8) resection neg 2 Biphasic SSX1 No NED (135) 39 28/M Thigh (3) resection, RT neg 3 Monophasic SSX1 No NED (122) 40 28/M Knee (3) HILP, resection neg 2 Biphasic SSX1 No NED (61) 41 38/M Thorax (NA) resection, chemo pos 3 Monophasic SSX1 Yes DOD (15) /M Thigh (15) RT, chemo - 3 Monophasic SSX2 Yes DOD (3) /M Thigh (14) resection neg 3 Biphasic SSX2 Yes AWD (2) /M Thigh (10) chemo NA 3 Monophasic SSX2 Yes DOD (3) /M Distal leg (14) Resection, chemo NA 2 Biphasic SSX1 Yes AWD (4) 0 DOC indicates death of other causes; DOD, death of disease; F, female; M, male; mo, months; NED, no evidence of disease; NA, not available; RT, radiotherapy; yr, year

25 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Table 2. Patient and tumor characteristics in the 45 synovial sarcomas studied Characteristics No. of patients % Male Female Age at diagnosis < 30 years Age at diagnosis 30 years Anatomic site Extremity Leg 28 Arm 4 Nonextremity Tumor Size 5 cm Tumor Size > 5 cm Tumor size unknown 5 Histology Monophasic Biphasic Mitotic count 0-9 mitoses/10 hpf (score 1) mitoses/10 hpf (score 2) > 19 mitoses/10 hpf (score 3) Necrosis absent Necrosis present 9 20 Grade Grade Fusion type SSX Fusion type SSX Distal Failure at presentation 5 11 hpf indicates high-power fields. Results Ezrin immunostaining Typical examples of positive staining for ezrin in monophasic and biphasic glandular and non-glandular SS are shown in Figures 1, 2 and 3 respectively. Variable ezrin immunostaining was seen in the mesenchymal spindle cell component of monophasic and biphasic SS. In the 35 monophasic SS, 12 samples were scored negative or weak positive (0-1+), 19 samples were scored moderately-strong positive (2-3+). For the other 4 monophasic SS, interpretable results for ezrin-staining were not available. The spindle cell component of the 10 biphasic SS was scored weak positive in 5 cases (1+) and moderately positive in 4 cases (2+). In one biphasic SS ezrin immunostaining results were not available. Ezrin immunoreactivity was uniformly and strongly (3+) present in epithelioid glandular and glandular areas of biphasic tumors. In fact, in all tumors strong ezrin staining allowed easy discrimination of monophasic and biphasic SS, even when the epithelioid component consisted of solid and trabecular, nonglandular elements. Associations between factors Most of the tumors (80%) that had developed metastasis at the time of first diagnosis were grade 3 (P = ) and tended to be larger (P = 0.053) when compared to tumors that had no evidence of metastatic spread. For all SS with either localized disease or metastasis at presentation, grade 3 tumors were significantly larger than grade 2 tumors (P = 0.003). Patients with SYT-SSX1 were more often male (male to female ratio 1.45:1) (n.s.). For SYT-SSX2 tumors the male to female ratio was 1:1. Twenty SS with SSX1 had monophasic histology and 7 had biphasic histology. Fifteen SYT-SSX2 tumors had monophasic histology and 3 had biphasic histology. Seventy percent of the biphasic tumors had a SYT-SSX1 fusion (P = 0.363). SYT-SSX1 tumors were more often extremity based and tended to be smaller (59% were 5cm) than SYT-SSX2 tumors (39% were 5cm) (n.s.). Twenty-six percent of the SYT-SSX1 tumors were grade 2, whereas 39% of the SYT-SSX2 tumors were grade 3 (P = 0.512). Ezrin expression levels in monophasic and/or biphasic SS were not associated with disease stage at presentation, tumor size and tumor grade

26 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Figure 1. Monophasic synovial sarcoma showing high positive expression of ezrin. Figure 2. Biphasic glandular synovial sarcoma showing intense immunoreactivity for ezrin in the glandular epithelial component. The Spindle cell component shows weak positive immunoreactivity. Figure 3. Nonglandular biphasic synovial sarcoma showing high positive immunoreactivity for ezrin in the glandular epithelial component. The Spindle cell component shows weak positive immunoreactivity. Survival analysis The median follow-up for all patients was 55 (range 2-238) months. Five year disease-specific survivals (DSS) were 67% including patients with metastasis at presentation and 72% for the group of 40 patients with localized disease at diagnosis. Because of the prognostic influence of distal failure at the time of diagnosis, univariate and multivariate analysis was performed for a subset of 40 patients with localized disease at presentation. No difference in survival was found for extremity based tumors when compared to non-extremity tumors. Therefore, a selective subset analysis of extremity based tumors was not performed. Nineteen patients (48 %) developed metastases during follow-up. Five-year metastasis-free survival (MFS) for the 40 patients with localized disease at diagnosis was 60% and the 10-year MFS was 52%. Univariate and multivariate results are presented in Table 3 and Table 4. Tumor size > 5 cm and histologic grade 3 were univariate significant factors associated with a worse DSS. Age 30 years, tumor size > 5 cm, necrosis and histologic grade were univariate significant factors associated with a worse MFS. Fusion-type and ezrin expression levels in monophasic and/or biphasic SS had no significant influence on patient survival. Univariate significant factors were included in multivariate analysis, except for necrosis, because this variable was included in the tumor grade. In multivariate analysis, tumor size and histologic grade remained significant prognostic factors for DSS and MFS. Discussion We will discuss prognostic factors in SS by comparing our single center results with those of the two largest multicenter studies reported by Ladanyi et al and Guillou et al. 10,11 Studying factors influencing outcome in 45 molecularly proven SS, we found that stage at presentation was the most significant variable influencing prognosis. Eleven percent of the patients treated at our institution had distant metastasis at the time of presentation. These patients had a considerably worse outcome than patients that only had localised disease at the time of presentation (P = 0.000). Tumors with metastasis at the time of diagnosis were mostly (80%) grade 3 (P = ) and tended to be larger (P = than tumors without metastasis. Apparently, the aggressive nature of these early metastatic tumors is mainly determined by histologic grade 3. In the series of Ladanyi et al and Guillou et al 12% and 14% of SS had metastasis at the time of diagnosis, respectively. 10,11 In accordance with our results, both studies found a strong

27 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression Table 3. Univariate analysis of classic prognostic factors in 40 SS with localized disease at presentation Characteristics DSS MFS Age < 30 or 30 yr Site extremity vs nonextremity Tumor Size 5 or > cm Histology monophasic vs. biphasic Mitotic count Necrosis absent/ present Grade 2 vs. grade Fusion type SSX1 vs. SSX DSS indicates disease specific survival; MFS, metastasis free survival. The bold P values present P values of 0.05 or less and were considered statistically significant for prognostic factors. Table 4. Multivariate analysis of univariate prognostic factors (with Cox proportional hazard regression) Characteristics DSS MFS Age < 30 or 30 yr Tumor size 5 or > cm Grade 2 vs grade DSS indicates disease specific survival; MFS, metastasis free survival. The bold P values present P values of 0.05 or less and were considered statistically significant for prognostic factors. P P correlation between disease stage at presentation and survival. 10,11 Only Guillou et al differentiated between grade 2 and grade 3 SS and observed that SS presenting with metastasis at diagnosis were more often grade Once metastatic disease has evolved, tumor stage overrules the influence of other prognostic factors. Patients with metastasis at presentation were therefore excluded from further statistical analysis. In the analysis of patients with localized disease, factors significantly associated with poor DSS in univariate analysis were FNCLCC tumor grade (grade 3 vs. grade 2; P = 0.009) and tumor size (> 5 vs. 5 cm; P = 0.037). Moreover, tumor size (P = 0.004), tumor grade (P = 0.006), tumor necrosis (present vs. absent; P = 0.009) and patient age ( 30 yr vs. < 30 yr; P = 0.028) were correlated with a poor MFS. By multivariate analysis, only tumor size (> 5cm) was independently associated with poor DSS (P = ). Tumor size (P = 0.024) and tumor grade (P = 0.047) remained significantly associated with MFS. In our series of patients with localized SS, tumor size was the strongest predictor of survival. We used cut-off values of 5 or > 5cm, because these values had the best predictive value in our series. Moreover, these values are also used for AJCC staging of soft tissue sarcomas 20 and allow optimal comparison with other studies. Ladanyi et al also used 5 cm cut-off values and found an association with survival for all patients (P = 0.04 in multivariate analysis), but not for patients with localized tumors at diagnosis. 10 Guillou et al used cut-off values of 5 cm and 7 cm and found a significant correlation with DSS by univariate analysis for patients with localized tumor at diagnosis (P = 0.10 for 5 cm and P = for 7 cm), but not by multivariate analysis. 11 After tumor stage and tumor size, tumor grade was the third most significant predictor in our series. We applied the three-tierced FNLCC grading system, which applies histologic type, mitotic rate and necrosis. 17 For SS, mitotic rate and necrosis determine whether tumors are either grade 2 or grade 3. Analysing the prognostic impact of mitotic rate and necrosis separately by univariate analysis, we found that only necrosis was significantly correlated with MFS by univariate analysis. Not unexpectedly, Guillou et al found that mitotic count was the most important predictor of reduced DSS and MFS in multivariate analysis. 11 Ladanyi et al applied the two-tierced UICC grading system, 21 which considers all SS as high grade tumors. 10 Our results substantiate those of Guillou et al by showing that grading SS according to the three-tierced FNLCC system has prognostic value. 11 Concerning patient age, a cut-off value of 30 years was most differentiating for outcome in our series and therefore used in our further analysis. Although age 30 yr was correlated with shorter MFS in the univariate analysis (P = 0.028), it was not significant in multivariate analysis. Guillou et al showed that the DSS was af

28 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression fected by age greater than 35 years in univariate analysis (P = ), but not in multivariate analysis. 11 In the series of Ladanyi et al age had no prognostic impact on survival. 10 Although patients with extremity tumors had a slightly better outcome, localization was not of significant influence in our series. In 15 patients (33%), SS was not extremity-based. One expects that non-extremity soft tissue sarcomas are associated with worse outcome since anatomic constraints limit curative options for surgery with wide margins. The individual cases in Table 1 do illustrate that patients with axial tumors have a low chance of curation. However, it is clear that, in our series, the high tendency of extremity-based SS to metastasize had a more significant impact on survival. In our series, SYT-SSX fusion type was not correlated with survival. SSX1 tumors had a slightly better outcome compared to SSX2 tumors in this study, which is in sharp contrast to the findings of Ladanyi et al, who reported that SYT-SSX fusion type was the only independent significant factor for overall survival in 133 patients with localized disease. 10 In their series, SSX1 tumors presented more often with metastasis and had a shorter survival. In contrast, Guillou et al found a shorter survival for SSX2 tumors. 11 Analysis of the association between fusion type and established clinicopathological prognostic factors (tumor stage, tumor size and tumor grade) does not provide a ready explanation for these conflicting results. Forty percent of SS in our series were SSX2 tumors, equal to Ladanyi et al. 10 In the series of Guillou et al 30% of SS were SSX2 tumors. 11 Associations of fusion type and tumor stage differed among these studies. We and Guillou et al found that SSX2 tumors had a higher tendency to present with metastasis at diagnosis (n.s. vs P = 0.17), in contrast to Ladanyi et al who found that SSX1 tumors were associated with metastasis at presentation (P = 0.05). 10,11 SSX2 tumors were more often grade 3 than SSX1 tumors, in both our series (P = n.s.) and those reported by Guillou et al (P = 0.13). 11 In all three studies discussed, fusion type was not associated with tumor size when using a cut-of value of 5 cm. However, Guillou et al observed that the majority of SSX2 tumors was larger than 7 cm (62%, P = 0.038). 10,11 In addition to classic clinicopathological prognostic factors and SYT-SSX fusion type we examined the possible prognostic impact of immunohistochemical expression of ezrin. Given the strong association between SS survival and tumor stage, we hypothesized that ezrin expression in either monophasic and/or biphasic SS was related to prognosis. In animal models and human tumors it has been shown that ezrin expression was clearly associated with early metastasis in different tumor types, including sarcomas. For instance, Khanna et al reported that high ezrin expression was related to poor outcome in pediatric osteosarcoma patients. 15 Yu et al showed that ezrin was associated with metastasis in rhabdomyosarcoma (RMS) cell lines. It was experimentally demonstrated in a mouse model that ezrin can induce a highly metastatic state in poorly metastatic RMS cell lines. 16 In high grade sarcomas of different subtypes, Weng et al observed that high ezrin expression was associated with development of metastasis during follow up. 22 To the best of our knowledge, we were the first to explore the possible significance of ezrin in a series of SS. In contrast to its prognostic value in other high grade sarcoma types, we found that ezrin expression levels had no impact on prognosis in monophasic and/or biphasic SS. In conclusion, this study confirms that disease stage at presentation, tumor size and histologic grade are major clinicopathological factors predicting survival in SS. Our results emphasize that it is worthwhile to differentiate between grade 2 and grade 3, as grade 3 tumors do worse. SYT-SSX fusion type and ezrin expression levels are not discriminative in predicting outcome of SS. References 1. Fisher C., de Bruijn D.R.H., Geurts van Kessel A.. Synovial sarcoma. In: Fletcher CDM, Unni KK, Mertens F eds. World Health Organization: Classification of Tumours. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon: IARC Press; 2002: Ten Heuvel SE, Hoekstra HJ, Suurmeijer AJH. Diagnostic accuracy of FISH and RT-PCR in 50 routinely processed synovial sarcomas. Appl Immunohistochem Mol Morphol 2008; 16: Amary MF, Berisha F, Bernardi FC et al. Detection of SS18-SSX fusion transcripts in formalin-fixed paraffin-embedded neoplasms: analysis of conventional RT-PCR, qrt-pcr and dual color FISH as diagnostic tools for synovial sarcoma. Mod Pathol 2007; 20: Terry J, Barry TS, Horsman DE, Hsu FD, Gown AM, Huntsman DG, Nielsen TO. Fluorescence in situ hybridization for the detection of t(x;18)(p11.2;q11.2) in a synovial sarcoma tissue microarray using a breakapart-style probe. Diagn Mol Pathol 2005;14: Guillou L, Coindre J, Gallagher G et al. Detection of the synovial sarcoma translocation t(x;18) (SYT;SSX) in paraffin-embedded tissues using reverse transcriptasepolymerase chain reaction: a reliable and powerful diagnostic tool for pathologists. A molecular analysis of 221 mesenchymal tumors fixed in different fixatives. Hum Pathol 2001;32:

29 The classic prognostic factors tumor stage, tumor size, and tumor grade are the strongest predictors of outcome in synovial sarcoma: no role for SSX fusion type or ezrin expression 6. Coindre JM, Pelmus M, Hostein I, et al. Should molecular testing be required for diagnosing synovial sarcoma? A prospective study of 204 cases. Cancer 2003;98: Bergh P, Meis-Kindblom JM, Gherlinzoni F, et al. Synovial sarcoma: identification of low and high risk groups. Cancer. 1999;85: Albritton KH, Randall RL. Prospects for targeted therapy of synovial sarcoma. J Pediatr Hematol Oncol. 2005;27: Ferrari A, Gronchi A, Casanova M, et al. Synovial sarcoma: a retrospective analysis of 271 patients of all ages treated at a single institution. Cancer. 2004;101: Ladanyi M, Antonescu CR, Leung DH, et al. Impact of SYT-SSX fusion type on the clinical behavior of synovial sarcoma: a multi-institutional retrospective study of 243 patients. Cancer Res. 2002;62: Guillou L, Benhattar J, Bonichon F et al. Histologic grade, but not SYT-SSX fusion type, is an important prognostic factor in patients with synovial sarcoma: a multicenter, retrospective analysis. J Clin Oncol 2004;22: Kawai A, Woodruff J, Healey JH, Brennan MF, Antonescu CR, Ladanyi M. SYT- SSX gene fusion as a determinant of morphology and prognosis in synovial sarcoma. N Engl J Med. 1998;338: Nilsson G, Skytting B, Xie Y, et al. The SYT-SSX1 variant of synovial sarcoma is associated with a high rate of tumor cell proliferation and poor clinical outcome. Cancer Res. 1999;59: Hunter KW. Ezrin, a key component in tumor metastasis. Trends Mol Med. 2004;10: Khanna C, Wan X, Bose S, et al. The membrane-cytoskeleton linker ezrin is necessary for osteosarcoma metastasis. Nat Med. 2004;10: Yu Y, Khan J, Khanna C, Helman L, Meltzer PS, Merlino G. Expression profiling identifies the cytoskeletal organizer ezrin and the developmental homeoprotein Six-1 as key metastatic regulators. Nat Med. 2004;10: Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading systems in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol. 1997;15: Kaplan El, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53: Cox DR. Regression models and life tables (with discussion). J R Stat Soc B 1972;34: Greene FL, Page DL, Fleming ID, et al, eds. American Joint Committee on Cancer Staging Manual. 6th ed. Philadelphia: Springer, Sobin LH, Wittekind CH. International Union Against Cancer: TNM Classification of malignant tumours (ed 6). New York, NY, Wiley-Liss, Weng WH, Ahlén J, Aström K, Lui WO, Larsson C. Prognostic impact of immunohistochemical expression of ezrin in highly malignant soft tissue sarcomas. Clin Cancer Res. 2005;11:

30

31 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up 4

32 Abstract Background The main goal of this retrospective study was to investigate prognostic factors influencing the survival of myxoid liposarcoma (MLS) with emphasis on the role of transitional areas (TLS) and round cell morphology (RCLS). Patients and Methods From 1977 to 2004, 49 patients-28 men (57%) and 21 women (43%) with a median age of 44 years (range, 7-83 years)-were diagnosed with an MLS. In 42 patients, the histology could be reviewed, and tumors were classified as MLS, TLS, or RCLS. Clinicopathologic factors were analyzed for influence on survival by univariate and multivariate methods. Results The median follow-up of 49 patients was 101 months (range, months). Of the 42 patients for whom histology was reviewed, 16 tumors were classified as MLS (38%), 19 as TLS (45%), and 7 as RCLS (17%). Sixteen patients (33%) developed a local recurrence after a median follow-up of 21 months (range, months). Thirteen patients (27%) developed metastases. The median interval between diagnosis and metastasis was 41 months (range, months). Median survival after metastasis was 18 months (range, months). The 5- and 10-year disease-specific survival rates were 85% and 72%, whereas the 5- and 10-year overall survival rates were 83% and 68%, respectively. Age at presentation (P = 0.02), tumor grade (P = 0.01), and tumor size (P = 0.005) were significant prognostic factors associated with survival. Tumor grade was the only independent prognostic variable that remained significant with multivariate analysis. A TLS presentation had no negative influence on patient survival. Conclusions Age at presentation, tumor grade, and tumor size had a negative influence on survival by univariate analysis, whereas tumor grade was the only independent prognostic factor by multivariate analysis. TLS was not associated with poor outcome. S.E. ten Heuvel H.J. Hoekstra R.J. van Ginkel E. Bastiaannet A.J.H. Suurmeijer Annals of Surgical Oncology jan;14(1): Introduction Myxoid liposarcoma (MLS) belongs to the group of soft tissue sarcomas with lipomatous differentiation; this group also comprises well-differentiated liposarcoma/atypical lipomatous tumor, pleomorphic liposarcoma, and dedifferentiated liposarcoma. MLS is the second most common type of liposarcoma, representing approximately one third of all liposarcomas and 10% of all adult soft tissue sarcomas. 1,2 MLS has a typical clinical presentation and metastatic behavior, quite different from the other liposarcoma histotypes. MLS frequently presents as a slow-growing, deep-seated tumor in the lower extremity of a relatively young adult patient. Metastases are often detected in other deep soft tissue locations, such as the retroperitoneum or extremities. 1 6 MLS has a distinct morphology, which is rarely confused with other monomorphic soft tissue tumors with myxoid stroma and lipomatous differentiation. 1,2 Moreover, specific chromosomal translocations have been discovered in MLS, which consists of the fusion of the FUS and CHOP genes [(t12;16)(q13;p11)] in 90% of tumors and fusion of the EWS and CHOP genes [(t12;22)(q13;q12)] in > 5% of tumors. In difficult cases, detection of these translocations with polymerase chain reaction methods allows the pathologist to make a specific diagnosis of MLS In the past, paucicellular MLS and hypercellular round cell liposarcoma (RCLS) were considered separate entities, even though transitions between MLS and RCLS were often recognized. However, with the identification of the specific translocations in both MLS and RCLS, it became clear that these lesions represent the ends of a morphological and biological spectrum of the same tumor entity. 1,2 MLS with > 5% RCLS has a higher tendency to metastasize, and this is associated with poorer survival As a consequence, pure MLS is considered a low-grade sarcoma, and MLS with > 5% round cell morphology is considered a high grade sarcoma. 14 In addition, transitional morphological phenotypes intermediate between paucicellular MLS and hypercellular RCLS have been recognized. The biological significance of this transitional type of MLS with increased cellularity (TLS) is still uncertain. 1,11,12 In this retrospective study of MLS, the prognostic effect of myxoid, transitional, and round cell morphology, in addition to other more established clinicopathologic prognostic factors, including age at presentation, tumor localization, tumor size, and tumor grade, was studied. Clinical end points evaluated were times to local recurrence, metastasis, and death of disease. The results of this study are discussed with respect to the limited existing literature. 4 53

33 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up Patients and Methods Forty-nine patients were treated between 1977 and 2004 for an MLS at the University Medical Center Groningen. Clinical and pathologic features were retrieved from the Groningen Sarcoma Database patient files and the pathology database. All patients underwent a surgical resection of the tumor with curative intent. Seven patients received hyperthermic isolated limb perfusion with tumor necrosis factor α and melphalan before surgery. 15 Two patients with a tumor in the thigh received preoperative radiotherapy and neoadjuvant chemotherapy according to the so-called Eilber protocol. 16 These two patients also received postoperative radiotherapy. Twenty-four patients (49%) received 60 to 70 Gy of postoperative radiotherapy. Six patients received chemotherapy: five patients for metastatic relapse and one patient for an irresectable local recurrence. Tumor size was recorded at 5- and 10-cm cutoff values, in accordance with the tumor-node-metastasis staging system for soft tissue sarcomas. 17 Hematoxylin and eosin stained sections from tumor resections were available for review in 42 cases. All specimens were reviewed by an experienced sarcoma pathologist (A.J.H.S.). All sections were scored by dividing MLS into paucicellular MLS, TLS, and RCLS according to the criteria described by Smith et al. 12 MLS was characterized by the presence of lipoblasts in varying numbers, primitive mesenchymal cells with minimal nuclear pleomorphism deposited in a myxoid stroma, and a distinct plexiform vascular pattern composed of thin-walled capillaries. TLS areas were hypercellular compared with the low cellularity of MLS, but the cells remained spindled, the plexiform vascular pattern remained easily discernible, and the cells were separated by at least some myxoid stroma. RCLS areas were characterized by a marked increase in cellularity in which the cells were round and separated by little or no stroma. In these areas, the mitotic index was generally increased, and a plexiform vascular pattern was difficult to recognize secondary to the overgrowth of primitive round cells. 12 In this study, tumors with > 5% TLS areas or 5% RCLS areas were categorized as TLS and RCLS, respectively. Tumors with > 5% TLS and 5% RCLS were categorized as RCLS. Typical examples of TLS and RCLS are shown in Fig 1. All tumors could be classified on histological analysis. Polymerase chain reaction for specific chromosomal translocations was not performed. Tumor grade was determined according to the French sarcoma grading system. 14 MLS and TLS were considered low-grade tumors (grade 1), whereas RCLS was considered a high-grade tumor (grade 2). Tumor recurrences and tumor metastases were diagnosed with imaging studies and confirmed with histological analysis. Time to recurrence and metastasis, disease-free survival (DFS), and overall survival (OS) were calculated by using the date of the histological diagnosis as the starting point. Clinical follow-up Figure 1 Typical histology of transitional areas (A), and round cell areas (B) in myxoid liposarcoma (hematoxylin and eosin, at high power magnification) data were obtained from all patients by using information from our tumor registry and medical records. OS and disease-specific survival (DSS) curves were generated with the Kaplan-Meier method. Fisher s exact tests were used to assess associations between factors. Clinical and pathologic prognostic factors in Kaplan-Meier curves were analyzed by the log-rank test, and all univariate significant factors were subsequently analyzed in a multivariate fashion with Cox proportional hazard regression

34 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up Results 1,0 Patient and Tumor Characteristics There were 28 men (57%) and 21 women (43%) with a median age at presentation of 44 years (range, 7 83 years). All patients in this series presented with localized disease. Forty-one tumors (84%) were located in the lower extremity, with the following distribution: 21 in the thigh, 7 in the lower leg, 5 in the gluteal region, 5 around the knee, and 3 in the inguinal region. Of the 8 remaining tumors, 3 were intrathoracic, 2 were retroperitoneal, 1 was intra-abdominal, 1 was in the lower arm, and 1 was in the head and neck area. Histological review of the 42 MLS resection specimens resulted in the following subtyping: 16 MLS (38%), 19 TLS (45%), and 7 RCLS (17%). All 7 RCLS cases had areas with > 5% round cell morphology, whereas none of the cases showed > 25% round cell areas. Tumor size could be retrieved from the pathology reports in 39 cases. The median tumor size was 12 cm (range, 1-40 cm); 7 tumors (18%) were 5 cm, 11 tumors (28%) were > 5 cm but 10 cm, and 21 tumors (54%) were > 10 cm. Treatment In all patients, surgical management of the tumor was with curative intent. Reviewing the specimens, a macroscopic radical (R0) resection with narrow margins of the tumor was documented in 37 cases (76%), and a microscopic involved resection margin (R1 resection) was involved in 12 cases. The 24 patients (49%) who received postoperative radiotherapy consisted of 6 patients with an R1 resection and 18 of 37 patients with a R0 resection (including all 6 patients with preoperative hyperthermic isolated limb perfusion). Proportion surviving Proportion surviving 0,8 DSS 0,6 OS 0,4 0,2 DFS 0, Years NED Local recurrence Metastasis 1,0 0,8 0,6 0,4 0,2 0, Years Figure 2 Kaplan-Meier curve showing overall survival (OS), disease-specific survival (DSS), and disease-free survival (DFS) of 49 patients with myxoid liposarcoma. Figure 3 Kaplan-Meier curves showing survival of 11 patients with local recurrence (and no metastasis) and 12 patients with metastatic relapse as opposed to patients with no evidence of disease (NED) during follow-up. Note the long survival of patients with metastatic relapse. 4 Follow-Up Results The OS, DSS, and DFS with a median follow-up of 101 months (range, months) are presented in Figure 2. Differences in OS between patients who developed local recurrence and patients who developed metastasis (with or without local recurrence), as opposed to patients without local recurrence or metastasis, are shown in Figure 3. Sixteen patients (33%) developed a local recurrence after a median follow-up of 21 months (range, months). Eight patients treated for local recurrence were alive without evidence of disease, one patient was alive with disease, and seven patients had died of disease, of whom five patients had developed metastatic disease, whereas the other two patients had uncontrollable local recurrence in the retroperitoneum and mediastinum. Proportion surviving 1,0 MLS 0,8 0,6 RCLS 0,4 0,2 0, Years Figure 4 Kaplan-Meier curves showing survival of 43 patients by histological type and grade (P = ). Myxoid liposarcoma (MLS) is grade 1, and round cell liposarcoma (RCLS) is grade

35 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up The risk of local recurrence was associated with tumor location; 6 (86%) of the 7 patients with a nonextremity tumor location developed a local recurrence, compared with 10 (24%) of 42 patients with a tumor in the lower or upper extremity (P = 0.003). The development of local recurrence was not related to the histological subtype of the primary tumor. The risk of local recurrence for a primary tumor classified as MLS, TLS, or RCLS was 38%, 32%, and 29%, respectively, which was not significant (P = 0.897). After the development of a local recurrence, the 5-year survival was 81% (Figure 3). According to current Dutch soft tissue sarcoma guidelines, 18 all 42 patients with R1 or marginal R0 resection of a tumor that was located in an extremity were eligible for postoperative radiotherapy. However, only 24 of these 42 patients actually received postoperative radiotherapy. Comparison of these patient groups revealed 8 local recurrences (44%) in the nonirradiated group of 18 patients, compared with 2 local recurrences (8%) in the irradiated group of 24 patients (P = 0.01). Of the 34 patients with an R0 tumor that was located in an extremity, 7 (21%) developed a local recurrence. Of the eight patients with an R1 tumor that was located in an extremity, three (38%) developed a local recurrence (P = 0.369). Thirteen patients (27%) had a metastatic relapse after a median follow-up of 41 months (range, months). Clinical and pathologic features, including treatment and follow-up events of the 13 patients found to have metastatic disease, are listed in Table 1. Ten patients (77%) had extrapulmonary metastases, of whom three developed lung metastases during follow-up. The three remaining patients developed lung metastases only. At last contact, 1 patient with metastatic disease was alive without evidence of disease, 2 patients were alive with disease, and 10 patients had died of disease. The 5-, 7-, and 10-year survival rates of patients after development of metastatic disease were 67%, 58%, and 34%, respectively (Figure 3). The OS, DFS, and DSS according to Kaplan-Meier analysis are presented in Figure 2 and Table 2. All seven patients with a tumor 5 cm were alive without evidence of disease with a median follow-up of 178 months (range, months). In univariate analysis, a worse DSS was significantly associated with tumor size by using > 5 cm as a cutoff value (P < 0.05), age at presentation > 45 years as a cutoff value (P < 0.02), and presence of > 5% round cell morphology, defined as grade 2 RCLS (P = ). Tumor location (extre mity vs. nonextremity), tumor size > 10 cm, and tumors with transitional morphology defined as TLS were not significantly associated with survival. Moreover, preoperative hyperthermic isolated limb perfusion and postoperative radiotherapy were not significantly associated with survival. In multivariate analysis of age at presentation, tumor size, and tumor histology, grade 2 RCLS was the only variable that remained significantly associated with DSS (P = 0.033). A Kaplan-Meier curve of patient survival related to tumor grade is presented in Figure 4. Table 1. Clinical and pathologic features of the 13 patients with metastatic disease Survival (mo) Treatment of metastasis Location of metastasis (mo) Treatment local recurrence (mo) Treatment Location of primary tumor (size; cm) Histotype Age(y)/ sex Patient No. Mesenterial (27) Chemotherapy DOD (33) Hemipelvectomy, RT 1 57/M NA R. buttock (9) Resection, RT DOD (233) Retroperitoneal, R. buttock (199) Resection, RT(108) 2 19/M TLS L. thigh (NA) Resection 3 49/M MLS R. calf (8) Resection, RT Amputation (12) CS (104) RT DOD (110) 4 68/M RCLS retroperitoneal (35) Resection (29) Liver (29) DOD (39) 5 37/M MLS L. thigh (14) Resection, RT R. thigh (222) Resection, RT NED (238) DOD (61) Chemotherapy, resection, RT R. retroperitonealparavertebral (7) Perfusion, Resection, RT R. popliteal fossa (15,5) 6 53/M RCLS 7 62/M TLS L. calf (11) Resection, RT Intra-abdominal (27) Chemotherapy AWD (28) Resection DOD (9) L. breast, lung (L. neck, TS) (0) Perfusion, Resection 8 50/V RCLS L. thigh (17,5) Chemotherapy DOD (181) Sacral, lung, intraabdominal (158) Resection, RT, chemotherapie 9 43/M TLS L. thigh (22) Resection AWD (135) R. upper arm, retroperitoneal, intra-abdominal, lung (115) HILP, amputation, RT (2) 10 38/V TLS R. thigh (NA) Resection Lung (28) DOD (35) Perfusion, Resection, RT 11 59/M MLS R. thigh (20) Lung (60) DOD (96) Resection, RT, chemotherapy (35) 12 48/M NA L. parotid gland (7) Resection DOD (106) Chemotherapy, resection, RT 13 43/V TLS R. thigh (23) Resection Lung (41) NA indicates not available R., right; RT, radiotherapy; DOD, death of disease; TLS, transitional liposarcoma; MLS, myxoid liposarcoma; L., left; CS, cervical spine; RCLS, round cell liposarcoma; NED, no evidence of disease; HILP, hyperthermic isolated limb perfusion; AWD, alive with disease; TS, thoracic spine

36 Table 2. Literature review of clincicopathological features and their relation to prognosis in MLS Mayo Study UMCG MSKCC 11 Cleveland Clinic Clinicopathologic features Year of diagnosis Median Age Site: extremity vs. nonextremity (%) Tumor size 5 cm (%) 14 4 NA 17 Tumor size > 10 cm (%) > 5% RC component Survival features No. of cases with follow-up Median follow-up (months) % local recurrence a 31 % LR tumor in extremity % metastatic relapse year survival (%) DSS 85 / OS 83 DSS 80 bos 82 OS 70 7-year survival (%) DSS 82 / OS 80 OS year survival (%) DSS 72 / OS 68 OS 67 OS 47 Univariate analysis of factors associated with survival Age (< or > 45 yr) p = 0.02 NA p = NS Tumor size (cut off in cm) p = ( 5) NS p = 0.04 (10) NS TLS NS NS NA NS RCLS > 5% p = p = 0.01 NS p < 0.05 RCLS > 25% NA NS p = NA Multivariate analysis of univariate prognostic factors (with Cox proportional hazard regression) Age ( or > 45 yr) NS NA p = Tumor size (cut off in cm) NS ( 5) NS (>10) RCLS > 5% (loc. disease) p = 0.03 p = 0.02 NA RCLS > 25% NA NS p = 0.04 NA MLS indicates myxoid liposarcoma; UMCG, University Medical Center Groningen; MSKCC, Memorial Sloan- Kettering Cancer Center; NA, no data available; RC, round cell; LR, local recurrence; DSS, disease-specific survival; OS, overall survival; NS, not significant; TLS, transitional liposarcoma; RCLS, round cell liposarcoma. a Only a subgroup of 55 patients who underwent primary surgical treatment at the Mayo Clinic was evaluated statistically regarding the development of local recurrence. b Only a subgroup of 70 patients presenting with localized disease at diagnosis was evaluated for calculating survival. Discussion Researchers at the Memorial Sloan-Kettering Cancer Center (MSKCC) have created a nomogram after statistical analysis of DSS in 2163 adult patients with the most common types of soft tissue sarcoma. This nomogram uses significant and well-established predictors of survival, including histological tumor type, tumor site (e.g., lower extremity), superficial or deep localization, tumor size, tumor grade, and patient age. Although not perfectly accurate, the nomogram can be used to predict the probability that a patient will die of soft tissue sarcoma; it has been calculated that nomogram predictions are within approximately 8% of the actual probability on average. 18 Using the calculator on the MSKCC Web site ( cfm), we entered the common clinicopathologic features of MLS (a low-grade liposarcoma > 10 cm deeply situated in the lower extremity of a 45-year-old patient), and this resulted in an estimated 4-, 8-, and 12-year DSS ranging from 85% to 100%, 78% to 94%, and 74% to 90%, respectively. As might be expected, the actual 5- and 10-year DSS rates of 85% and 72% in our clinicopathologic analysis of 49 patients with MLS are comparable to those obtained with the nomogram. The MSKCC nomogram can also be used to further gain insight into the relative importance of each separate prognostic variable. For instance, when using the 8-year DSS of 86% ± 8% of the average MLS patient as baseline, the difference in estimated DSS is 25% for a patient with a high-grade tumor (instead of a low-grade tumor), 6% for a patient age of 65 years (instead of 45 years), and +9% for tumor size 5 cm (instead of > 10 cm). From these calculations, one might conclude that tumor grade is prognostically much more important than patient age and tumor size. However, it should be pointed out that the sarcoma survival odds in the MSKCC nomogram are based on all histological types of liposarcomas and not just on MLS. Therefore, we have compared the results in our MLS study with those of previous retrospective prognostic studies performed by three large cancer centers in the United States, including the MSKCC, as summarized in Table As pointed out by others, direct comparison of results derived from retrospective prognostic studies is hampered by inevitable problems due to differences in patient and tumor characteristics, treatment, and statistical analysis. Clearly, statistical power of significance of prognostic factors depends on the number of cases and follow-up time. In this respect, it is notable that the median follow-up of 101 months in our series is the longest follow-up reported to date. The two largest studies (from MSKCC and Mayo Clinic) have also used Kaplan-Meier methods and multivariate analysis with Cox proportional hazards analysis. 11,13 In the comparatively small study performed by the Cleveland Clinic Foundation, only event-free survival could be measured. 12 We have 61 4

37 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up chosen cutoff values for grade (5% round cell morphology), age (45 years), and tumor size (5 and 10 cm) to allow optimal comparison with the two larger studies. Moreover, the cutoff values applied for tumor size and tumor grade are well established and already used in the 2002 World Health Organization classification for histological typing and the 2002 American Joint Committee on Cancer classification for staging of soft tissue sarcoma. 1,17 In addition, age 45 is the median age at presentation of patients with MLS, as found in the other studies, and is thus likely to provide the strongest statistical power. 13 With respect to clinicopathologic features potentially influencing prognosis, important differences between the studies discussed herein are differences in the relative number of tumors presenting in the (lower) extremity, of tumors 5 cm or > 10 cm, and of grade 2 tumors with > 5% round cell morphology (Table 2). The prognostic factors significantly associated with worse DSS by univariate analysis in our study were tumor grade (grade 2 with > 5% RCLS vs. grade 1; P = ), patient age (> 45 vs. 45 years; P = 0.02), and tumor size (> 5 vs. 5 cm; P = ). By multivariate analysis, only 5% RCLS (tumor grade 2) was significantly associated with worse DSS (P = 0.03), in accordance with the MSKCC study. 11 In the study from the Cleveland Clinic, 5% RCLS (tumor grade 2) was significantly associated with event-free survival. 12 The higher cutoff value of 25% RCLS was significantly associated with OS by multivariate analysis in the Mayo Clinic series, but not in the MSKCC series. 11,13 We were only able to consider 5% round cell morphology as a cutoff value for grade 2 MLS. None of our cases contained areas with > 25% round cell morphology. The percentage of RCLS (17%) was lower compared with the other studies (27% 43%), and this may explain why the 5-year OS of 83% and the 5-year DSS of 85% were slightly better than the 5-year OS of 82% found in the Mayo Clinic series and the DSS of 80% observed in the MSKCC subgroup of 70 patients with localized disease We and others observed that there is no association between TLS and patient survival. Moreover, in a meta-analysis only considering cases of TLS of the lower extremity in the Cleveland Clinic and University Medical Center Groningen series, TLS was not associated with poorer outcome. 12 The metastatic rate in our patient group was 27%, which is lower than that found by other groups (range, 30% 35%) This may also be due to the lower percentage of RCLS in our series. Although the median survival of patients with metastatic disease is quite long (Figure 3), their outcome was poor. Of the 13 patients with metastatic relapse in our series, 10 had died of disease (after 5 to 65 months), 2 patients were alive with disease (after 28 and 135 months), and 1 patient had no evidence of disease after 16 months. Most metastases were observed in unusual soft tissue locations, which is a typical feature of MLS, as found in other studies. 3 6 The local recurrence rate in this study was 33%, which is higher than that found in the MSKCC (27% in localized disease) and Mayo Clinic (14%) series. 11,12 The relative number of tumors located in the extremities in our series was comparable to that in the MSKCC series, whereas most tumors in the Mayo Clinic series were in the extremities. 11,12 In our series, 24% of tumors in the extremities recurred, as opposed to 86% of tumors with a nonextremity location. Surgical margins and postoperative radiotherapy are the main factors that are associated with the development of local recurrence. We were not able to study the significance of surgical margin status in this study, but not surprisingly, in the larger MSKCC and Mayo Clinic series, local recurrence free survival was significantly related to both negative surgical margins and extremity versus nonextremity location. 11,12 In our study, the chance of local recurrence was not related to histological subtype (MLS, TLS, or RCLS), whereas in the MSKCC study, development of local recurrence was related to > 5% increased cellularity, defined as TLS in our study (P < 0.01 by univariate analysis). 11 In our series, 42 tumors that were located in an extremity were resected with an R1 or a narrow R0 margin, which is an indication for radiotherapy according to current soft tissue sarcoma guidelines in The Netherlands. 15 Patients who did not receive radiotherapy had more recurrences than patients who did not. In the Mayo Clinic study, no significant difference was seen in local recurrence rate between patients treated with marginal resection with radiotherapy versus wide surgical resection alone. 13 In summary, this study in 49 patients with MLS showed that tumor size, age at presentation, and tumor grade (grade 2 being defined by 5% RCLS) were associated with worse DSS by univariate analysis, whereas tumor grade was the only independent prognostic factor that remained significant by multivariate analysis. In a meta-analysis of our study and one previous study, TLS of the lower extremity was not associated with a worse outcome. References 1. Antonescu CR, Ladanyi M. Myxoid Liposarcoma. In: Fletcher CDM, Unni KK, Mertens F, eds. World Health Organization Classification of Tumours: Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2002: Enzinger FM, Weiss SM. Liposarcoma. In: Enzinger FM, Weiss SM, eds, Soft Tissue Tumors. St. Louis; Mosby 2001:

38 Clinicopathologic prognostic factors in myxoid liposarcoma: a retrospective study of 49 patients with long-term follow-up 3. Cheng EY, Springfield DS, Mankin HJ. Frequent incidence of extrapulmonary sites of initial metastasis in patients with liposarcoma. Cancer 1995;75: Pearlstone DB, Pisters PW, Bold RJ, et al. Patterns of recurrence in extremity liposarcoma: implications for staging and follow-up. Cancer 1999;85: Spillane AJ, Fisher C, Thomas JM. Myxoid liposarcoma the frequency and the natural history of nonpulmonary soft tissue metastases. Ann Surg Oncol 1999;6: Estourgie SH, Nielsen GP, Ott MJ. Metastatic patterns of extremity myxoid liposarcoma and their outcome. J Surg Oncol 2002;80: Panagopoulos I, Mandahl N, Ron D, et al. Characterization of the CHOP breakpoints and fusion transcripts in myxoid liposarcomas with the 12;16 translocation. Cancer Res 1994;54: Knight JC, Renwick PJ, Cin PD, Van den Berghe H, Fletcher CD. Translocation t(12;16)(q13;p11) in myxoid liposarcoma and round cell liposarcoma: molecular and cytogenetic analysis. Cancer Res 1995;55: Dal Cin P, Sciot R, Panagopoulos I, et al. Additional evidence of a variant translocation t(12;22) with EWS/CHOP fusion in myxoid liposarcoma: clinicopathological features. J Pathol 1997;182: Panagopoulos I, Hoglund M, Mertens F, Mandahl N, Mitelman F, Aman P. Fusion of the EWS and CHOP genes in myxoid liposarcoma. Oncogene 1996;12: Antonescu CR, Tschernyavsky SJ, Decuseara R, et al. Prognostic impact of P53 status, LS-CHOP fusion transcript structure, and histological grade in myxoid liposarcoma: a molecular and clinicopathologic study of 82 cases. Clin Cancer Res 2001;7: Smith TA, Easley KA, Goldblum JR. Myxoid/round cell liposarcoma of the extremities. A clinicopathologic study of 29 cases with particular attention to extent of round cell liposarcoma. Am J Surg Pathol 1996;20: Kilpatrick SE, Doyon J, Choong PF, Sim FH, Nascimento AG. The clinicopathologic spectrum of myxoid and round cell liposarcoma. A study of 95 cases. Cancer 1996;77: Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading system in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol 1997;15: Nijhuis PH, Pras E, Sleijfer DT, Molenaar WM, Koops HS, Hoekstra HJ. Long-term results of preoperative intra-arterial doxorubicin combined with neoadjuvant radiotherapy, followed by extensive surgical resection for locally advanced soft tissue sarcomas of the extremities. Radiother Oncol 1999;51: Greene FL, Page DL, Fleming ID, et al. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag, Richtlijn diagnostiek bij weke delen tumoren en behandeling van weke delen sarcomen (herziening). Utrecht: Nederlandse Werkgroep Weke Delen Tumoren, Available at: Accessed on January 1, Kattan MW, Leung DHY, Brennan MF. A postoperative nomogram for 12-year sarcoma-specific death. J Clin Oncol 2002;20:

39

40 Epithelioid sarcoma: still an only surgically curable disease 5

41 Abstract Background Epithelioid sarcoma is a rare soft tissue sarcoma with a known high propensity for locoregional recurrence and distant metastases. The clinical behavior and prognostic factors that influence the survival of patients with epithelioid sarcoma were studied. Patients and Methods Twenty-three patients, including 16 men (70%) and 7 women (30%), who were treated for epithelioid sarcoma between at the University Medical Center Groningen and Radboud University Nijmegen Medical Center, were reviewed retrospectively. The median age at diagnosis was 22 years (range, 1 54 years). At the time of diagnosis, 11 patients (48%) had metastases. Six patients with distant metastasis and 1 patient with an unresectable tumor received palliative treatment (30%). The remaining 16 patients underwent surgical treatment of local disease (11 patients) or locoregional disease (5 patients). Five patients in that group received isolated limb perfusion with tumor necrosis factor and melphalan. Results The 5-year and 10-year disease-free survival rates for all patients were 34% and 17%, respectively; for the 16 patients who received curative treatment, both rates were 56%. In the latter group, 8 patients developed local recurrence (50%) after a median follow-up of 4 months (range, 1 14 months). Nine patients were disease free after a median follow-up of 50 months (range, months). Tumor size > 5 cm (P < ) at diagnosis and local recurrence (P < ) were significant predictors of survival. Conclusions The prognosis for patients with epithelioid sarcoma is poor, because a substantial number of patients present with extensive disease, lymph node metastases, and/or distant metastases. Treatment consists of radical surgical excision of the tumor and, if indicated, therapeutic lymph node dissection. In patients who have large tumors, isolated limb perfusion may be useful. S.A.H.J. de Visscher R.J. van Ginkel T. Wobbes R.P.H. Veth S.E. ten Heuvel A.J.H. Suurmeijer H.J. Hoekstra Cancer Aug;107(3): Introduction Epithelioid sarcomas are rare soft tissue tumors that originate from mesenchymal tissue and have unclear histogenesis. 1 3 Clinically, the tumor usually is characterized as a slow-growing, painless swelling. It shows preference for the distal part of the extremities, particularly the hand Other localizations, although very rare, are the perineum, penis, and vulva. 1,2,4,6 18 The tumor can originate superficially in the dermis or deeply under the fascia. The more deeply lying tumors often are larger and are fixed to tendons, tendon sheaths, or fascia, and they are poorly differentiated. 1,2,5,12 Treatment depends mainly on the presence or absence of regional and/or distant metastases. When surgery can be applied with curative intent, there is still a high risk of local recurrence, regional metastases, and/or distant metastases. 1,4 8,11,17,19,20 To gain greater insight into the current treatment modalities for epithelioid sarcomas and treatment outcomes, a retrospective study was performed at the University Medical Center Groningen and the Radboud University Nijmegen Medical Center, the Netherlands. In this report, new treatment options for epithelioid sarcomas are discussed, and the literature is reviewed. Materials and methods From 1979 to 2003, epithelioid sarcomas were diagnosed in 23 patients with a median age of 22 years (range, 1 54 years), including 16 men (70%) and 7 women (30%). For the purpose of staging, all patients underwent computer tomography (CT) imaging of the lungs and CT or magnetic resonance imaging (MRI) of the tumor. At the time of diagnosis, 5 patients had lymph node metastases (22%), 5 patients had lymph node and distant metastases (22%), and 1 patient had distant metastases only (4%). According to the American Joint Committee on Cancer/International Union Against Cancer staging system, there were 8 patients with Stage IIA/B tumors (35%), 4 patients with Stage III tumors (17%), and 11 patients with Stage IV tumors (48%). 21,22 Table 1 shows the patient and tumor data. Sixteen patients (70%) underwent curative treatment, which consisted of a local resection (n = 5 patients), a local resection and lymph node dissection (n = 1 patient), a lymph node dissection plus isolated limb perfusion (ILP) with tumor necrosis factor (TNF) and melphalan and a local excision (n = 3 patients), an amputation (n = 5 patients), an amputation and lymph node dissection combined with ILP (n = 1 patient), and an amputation and ILP (n = 1 patient). Lymph nodes were examined for metas- 5 69

42 Epithelioid sarcoma: still an only surgically curable disease Table 1. Tumor and treatment characteristics of 23 patients with epithelioid sarcoma Characteristic Gender No. of patients Male Female 7 30 Median age, y (range) 22 (1 54) Treatment delay, mo (range) 3.5 (1 36) Site of primary tumor Distal upper limb 4 17 Hand 4 17 Proximal upper limb 2 9 Distal lower limb 5 22 Proximal lower limb 2 9 Perineal region 1 4 Trunk 3 13 Head and neck 2 9 Tumor size, cm < to < to < to < > Stage* IIA/B 8 35 III 4 17 IV Palliative treatment 7 30 Curative treatment * Stage was determined according to the International Union Against Cancer/ American Joint Committee on Cancer 2002 staging classification. 21,22 % tases by palpation and ultrasound scanning or fine-needle aspiration. Indications for regional lymph node dissection included regional lymph node metastases without distant metastases. Elective regional lymph node dissections were not performed. Indications for ILP consisted of patients with in-transit metastases and large or multifocal tumors without distant metastases. When a curative resection (R0 resection) was achieved, no postoperative radiation therapy (RT) was given. Indications for the use of postoperative RT were a marginal primary resection (R1), local regional recurrence (without distant metastases), or palliative treatment. One patient who underwent an R1 resection and 6 patients who were treated with palliative intent received RT. The remaining 7 patients (6 patients with distant metastases and 1 patient with an inoperable tumor) received palliative treatment (30%) with chemotherapy (doxorubicin, cyclophosphamide, iphosphamide, and/or vincristine) and RT. None of these patients were eligible for resection of the primary tumor and/or metastases. Overall survival and disease-free survival were calculated with the Kaplan Meier method. Prognostic factors for survival were analyzed with a univariate log-rank test. 23 Results The median interval between observing the first symptoms, making the diagnosis, and starting treatment (delay) was 3.5 months (range, 1 36 months). The median patient age at the time of diagnosis was 22 years (range, 1 54 years). The male: female ratio was 2:1. After a median follow-up of 17 months (range, months), 14 of 23 patients died of disease (61%), and 9 patients were alive with no evidence of disease (39%). In the group of 16 patients who received curative treatment, the median follow-up was 47 months (range, months); 7 patients were tumor free (44%) after a median follow-up of 50 months (range, months); and, 9 patients developed tumor recurrence (56%), including 8 patients with local recurrence (median follow-up, 4 months; range, 1 14 months) who underwent local resection and 1 patient with locoregional recurrence (after 8 months) who underwent inguinal lymph node dissection and received adjuvant RT (40 grays). In the group of 8 patients who underwent surgical treatment with curative intent for their recurrence, 1 patient developed locoregional recurrence and underwent axillary lymph node dissection; 7 patients developed distant metastases, and in 3 of them, this was combined with locoregional recurrence. Four of those 7 patients received palliative RT. Figure 1 shows the follow-up of the 16 patients after their first treatment with curative intent. The survival of 16 patients who were treated with curative intent and 7 patients

43 Epithelioid sarcoma: still an only surgically curable disease who were treated with palliative intent is shown in Figure 2. The 5-year and 10-year survival rate of the curative group was 56% (P < 0.001). In the curative group, tumor size (> 5 cm; P = ) and the development of local recurrence (P = ) influenced survival significantly (Table 2). Furthermore, in the total group of 23 patients, female gender (P = ) and the presence of distant or lymph node metastases (P = ) were unfavorable prognostic factors. Patients who had distant metastases (Stage IV) had a significantly poorer prognosis compared with patients who had regional lymph node metastases (P = ). Nine of 13 patients (69%) who presented with or developed distant metastases had lung metastases. Five patients underwent lymph node dissection plus ILP, including 3 patients who died from regional and distant metastases and 2 patients who were disease free after 12 months and 218 months follow-up, respectively. Discussion Epithelioid sarcomas are very rare, high-grade, soft tissue sarcomas that constitute < 1% of all soft tissue sarcomas. 11,17 The current results study confirmed earlier reports that the tumor is diagnosed mainly between ages 10 years and 35 years in ma les. 1,2,4 6,8 11,13,18 20,24,25 The tumor can consist of 1 or more lymph nodes that vary in size from a few millimeters to several centimeters. 1 6,18 The surface of the tumor often is yellowy-brown because of focal necrosis and bleeding. 1,2,4,5,9 In view of the innocent appearance of the tumor, sometimes, there is great delay, even up to a number of years. It should be realized that clinical manifestation can vary widely. The current study showed that the prognosis depended on the initial tumor stage at presentation. The high percentage of patients with initial metastases and local recurrence, lymph node metastases, and distant metastases after surgical treatment with curative intent is characteristic of epithelioid sarcomas. 1,2,4 8,10 12,16,17,20,26 The size of the tumor (> 5 cm) is important for the prognosis, although it remains difficult to determine tumor size in patients who have multifocal disease. 1,2,4,6,8,10,11,16,18,27,28 Epithelioid sarcomas are localized mainly on the upper extremity. In the current study, no significant difference was observed between proximal or distal tumor localization and survival (Table 2). This is in contrast to the literature, in which it was reported that patients with distal localization had a significantly better prognosis. 1,4,8,16,29 It is possible that the biologic behavior of proximally localized epithelioid sarcoma is more aggressive, which can be explained by possible rhabdoid characteristics and/or classic prognostic characteristics, such as size, depth, vascular invasion, and resectability. 8, LNM 8 LR Surgery + RT Surgery NED NED 1 LNM Surgery NED 4 DM 3 LNM + DM DOD Figure 1. Follow-up results are illustrated for 16 patients who had epithelioid sarcoma after they were treated with curative intent. LNM indicates lymph node metastases; LR, local recurrence; DM, distant metastases; NED, no evidence of disease; DOD, dead of disease; RT, radiotherapy. % survival Months Curative intent Palliative intent Figure 2. This Kaplan Meier curve compares 16 patients who received treatment with curative intent versus 7 patients who received treatment with palliative intent

44 Epithelioid sarcoma: still an only surgically curable disease Table 2. Statistical analysis of treatment outcomes for 16 patients with epithelioid sarcoma treated with curative intent Table 3. Sites of primary epithelioid sarcoma as published in various studies in the literature Characteristic Tumor size, cm No. of patients Mean distant metastases-free survival (95% CI), months < ( ) > ( ) Localization Proximal ( ) Distal ( ) Gender Male ( ) Female ( ) Local recurrence Yes ( ) No 8 Mean follow-up, 98.3 Lymph node metastases Yes ( ) No ( ) P Study Year No. of patients Upper extremity (%) Lower extremity (%) Perineum and trunk (%) Head and neck (%) Enzinger Prat et al Chase and Enzinger Bos et al Whitworth et al Ross et al Halling et al Spillane et al Callister et al Matsushita et al Livi et al De Visscher et al. (current study) Total AJCC stage Stage II ( ) Stage III/IV ( ) AJCC indicates American Joint Committee on Cancer Table 3 shows an overview of the anatomic localizations of the epithelioid sarcomas described in this series and in the literature. In the majority of studies, the prognosis for patients with epithelioid sarcoma was better among women. 2,3,6,16,17,20,29 Ross et al. reported a significant correlation between male gender and shorter survival in a metaanalysis of 269 patients with epithelioid sarcomas. 17 We could not confirm this correlation in our analyses on the total population or in the group of patients who were treated with curative intent. In the total population, the prognosis was significantly poorer among women (P = ) because of their presentation with widespread and metastatic disease

45 Epithelioid sarcoma: still an only surgically curable disease All 16 patients in our study who were treated with curative intent underwent curative resection (R0 resection). The degree of radicality of the excision determined the risk of local recurrence. In 2 studies in the literature, local recurrence rates after an R0 (radical resection) were 10% and 12%, respectively. After an R1 resection (microscopic, nonradical resection), these rates were 36% and 60%, respectively. 8,11 A central issue in local treatment, thus, is to achieve an R0 resection. Amputation does not lead to better survival than a successful local R0 resection. The benefit of an R0 resection also was emphasized in a series of patients with epithelioid sarcomas who were diagnosed incorrectly. Local recurrence rates and deaths from the disease were increased as a result of an inadequately performed surgical resection. 1,6 When multiple local recurrences were present, local resection generally no longer was possible, and amputation of the affected extremity was the only remaining option. 6,11,16,20 Our experience with ILP with TNF and melphalan during surgical treatment for epithelioid sarcoma is limited. Patients who present with multifocal and/or unresectable primary tumors probably will benefit from perfusion. ILP enables the administration of 10 times higher dosages of cytostatics without causing systemic toxicity. In the current study, the size of the tumor decreased after ILP, which presented the opportunity to perform an R0 resection. The role of adjuvant RT in the treatment of marginally resected, high-grade sarcomas already has been established. Although epithelioid sarcomas belong under the classification of sarcomas, they behave more like malignant skin tumors. Several studies reported a lower risk of local recurrence of epithelioid sarcomas after preoperative or postoperative RT in combination with radical surgery. 8,12,17,18,20 The studies by Bos et al. and Livi et al., however, did not confirm this lower risk. 6,16 In our study, of 16 patients who received treatment with curative intent, 1 patient received postoperative RT after an R1 resection, and 5 patients received RT after recurrence. Only 1 of those 5 patients presented with lymph node metastases only. The other 4 patients presented with distant metastases and/or lymph node metastases, and they received RT with palliative intent. In contrast to many other studies, we did not observe that the presence of lymph node metastases was a significantly unfavorable factor. 2,4-6,8,10,11,16,26,29,32 Callister et al. suggested that lymph node metastases may be a (first) symptom of widely disseminated disease rather than a purely locoregional process. 8 If this is true, then lymph node metastases should be ascribed the same value as distant metastases. Fifteen patients (65%) in our series developed lymph node metastases after a median of 7 months (range, months), and 10 of those patients subsequently manifested distant metastases (Table 4). Therefore, it is worthwhile to examine the regional lymph nodes after the primary diagnosis has been made. 5 Currently, this can be accomplished noninvasively by means of ultrasound scanning of the regional lymph node basin in combination with fine-needle aspiration or sentinel lymph node biopsy. If lymph node metastases are present, then therapeutic lymph node dissection is indicated. 5,11,26 It would be difficult to prove the extra value of ultrasound scanning or sentinel lymph node biopsy in relation to clinical palpation in the treatment of epithelioid sarcomas; however, on a theoretical basis, we strongly advise this approach in the current treatment for epithelioid sarcomas. In the group of patients who were treated with curative intent, 7 patients developed distant metastases (43%); whereas, in the total series, 13 patients (57%) had distant metastases initially or during follow-up (Table 4). The most common site for dissemination was the lung. 1,4,5,11,19 Other less frequent sites were the digestive tract, liver, kidneys, and skeletal system. 1 In our series, 69% of the distant metastases were localized in the lungs. Distant metastases can differ from the primary tumor or local recurrence in terms of poorer differentiation and increased necrosis. 1,6 In the current study, distant metastases occurred after a median follow-up of 10 months (range, 2 29 months) in patients who had received treatment with curative intent. The median survival after a diagnosis of distant metastases in this study was only 5 months (range, 1 44 months). Others have reported a survival duration between 8 months and 28 months. 11,16,19 Unfortunately, our study and other studies showed that chemotherapy is of little benefit in the palliative treatment of patients with disseminated epithelioid sarcomas. 4,6,16 In the literature, there is wide variation in the 5-year and 10-year survival rates of between 25% to 78% and 25% to 74%, respectively. In the current study, the 5-year and 10-year survival rate was only 32% for 23 patients (Table 4). This finding raises thee question of whether these differences are based on the stage distribution. In agreement with Spillane et al. and Ramanath et al., we did not observe any differences in survival between the stages. 11,24 However, we did observe a significant difference in survival between patients with Stage IV disease who had lymph node metastases and patients with Stage IV disease who had distant metastases. It may be concluded that surgery remains the only treatment option for patients with primary and disseminated epithelioid sarcoma. The role of postoperative RT after an R0 resection and the value of ILP with TNF and melphalan in patients with epithelioid sarcoma need further research. However, we believe that modern, minimally invasive diagnostics of the regional lymph node basins with ultrasound scanning and sentinel lymph node biopsy should become part of the preoperative staging procedure for patients with epithelioid sarcoma

46 Table 4. Literature overview 10-year survival (%) 5-year survival (%) Distant metastases (%) Lymph node metastases (%) Local recurrence (%) Mean follow-up, years Mean age at first symptoms, years Mean symptom duration before diagnosis, months Female (%) Male (%) Study No. Enzinger, 1970 (AFIP) 1 * Median, 9 Median, 23 7 (median, 5) NA NA Prat et al., 1978 (MSKCC) NA Chase and Enzinger, NA (AFIP) 4* Bos et al., 1988 (MAYO) 16 * Median, 10.5 Median, Whithworth et al., NA NA 58 (MDACC) 20 Steinberg et al., 1992 (MGH) 7* NA NA Ross et al., 1996 (MSKCC) NA Halling et al., 1996 (MAYO) 18* Spillane et al., 2000 (RMH) NA 29 Median, Callister et al., NA 35 Median, (MDACC) 8* Matsushita et al., NA (CIH) 19* Current study, (UMCG, RUNMC) Variation AFIP indicates Armed Forces Institute of Pathology; NA, not available; MSKCC, Memorial Sloan-Kettering Cancer Center; MAYO, Mayo Clinic; MDACC, The University of Texas M. D. Anderson Cancer Center; MGH, Massachusetts General Hospital; RMH, Royal Marsden Hospital; CIH, Cancer Institute Hospital (Tokyo); UMCG, University Medical Center Groningen ; RUNMC, Radboud University Nijmegen Medical Center * No patients were included who had lymph node metastasis or distant metastasis at first diagnosis. Patients were included who had distant metastasis at first diagnosis. Patients were included who had lymph node metastasis, but not distant metastasis, at first diagnosis. References 1. Enzinger FM. Epithelioid sarcoma. A sarcoma simulating a granuloma or a carcinoma. Cancer. 1970; 26: Weiss SW, Goldblum JR. Epithelioid sarcoma. In: EnzingerFM, WeissSM, editors. Soft Tissue Tumors. 4th ed. St. Louis: Mosby; 2001: Molenaar WM, De Jong B, Dam-Meiring A, Postma A, De Vries J, Hoekstra HJ. Epithelioid sarcoma or malignant rhabdoid tumor of soft tissue? Epithelioid immunophenotype and rhabdoid karyotype. Hum Pathol. 1989; 20: Chase DR, Enzinger FM. Epithelioid sarcoma. Diagnosis, prognostic indicators, and treatment. Am J Surg Pathol. 1985; 9: Prat J, Woodruff JM, Marcove RC. Epithelioid sarcoma: an analysis of 22 cases indicating the prognostic significance of vascular invasion and regional lymph node metastasis. Cancer. 1978; 41: Bos GD, Pritchard DJ, Reiman HM, Dobyns JH, Ilstrup DM, Landon GC. Epithelioid sarcoma. An analysis of fifty-one cases. J Bone Joint Surg Am. 1988; 70: Steinberg BD, Gelberman RH, Mankin HJ, Rosenberg AE. Epithelioid sarcoma in the upper extremity. J Bone Joint Surg Am. 1992; 74: Callister MD, Ballo MT, Pisters PW, et al. Epithelioid sarcoma: results of conservative surgery and radiotherapy. Int J Radiat Oncol Biol Phys. 2001; 51: Sarica K, Bakir K, Yagci F, Ozalpat O, Korkmaz C. Epithelioid sarcoma of the penis. Urol Int. 1999; 62: Evans HL, Baer SC. Epithelioid sarcoma: a clinicopathologic and prognostic study of 26 cases. Semin Diagn Pathol. 1993; 10: Spillane AJ, Thomas JM, Fisher C. Epithelioid sarcoma: the clinicopathological complexities of this rare soft tissue sarcoma. Ann Surg Oncol. 2000; 7: Shimm DS, Suit HD. Radiation therapy of epithelioid sarcoma. Cancer. 1983; 52: Huang DJ, Stanisic TH, Hansen KK. Epithelioid sarcoma of the penis. J Urol. 1992; 147: Konefka T, Senkus E, Emerich J, Dudziak M. Epithelioid sarcoma of the Bartholin s gland primarily diagnosed as vulvar carcinoma. Gynecol Oncol. 1994; 54: Guillou L, Wadden C, Coindre JM, Krausz T, Fletcher CD. Proximal-type epithelioid sarcoma, a distinctive aggressive neoplasm showing rhabdoid features. Clinicopathologic, immunohistochemical, and ultrastructural study of a series. Am J Surg Pathol. 1997; 21:

47 Epithelioid sarcoma: still an only surgically curable disease 16. Livi L, Shah N, Paiar F, et al. Treatment of epithelioid sarcoma at the Royal Marsden Hospital. Sarcoma. 2003; 7: Ross HM, Lewis JJ, Woodruff JM, Brennan MF. Epithelioid sarcoma: clinical behavior and prognostic factors of survival. Ann Surg Oncol. 1997; 4: Halling AC, Wollan PC, Pritchard DJ, Vlasak R, Nascimento AG. Epithelioid sarcoma: a clinicopathologic review of 55 cases. Mayo Clin Proc. 1996; 71: Matsushita Y, Ahmed AR, Kawaguchi N, Matsumoto S, Manabe J. Epithelioid sarcoma of the extremities: a dismal long-term outcome. J Orthop Sci. 2002; 7: Whitworth PW, Pollock RE, Mansfield PF, Couture J, Romsdahl MM. Extremity epithelioid sarcoma. Amputation vs local resection. Arch Surg. 1991; 126: GreeneFL, PageDL, FlemingID, et al., editors. AJCC Cancer Staging Manual. 6th ed. New York: Springer-Verlag; SobinLH, WittekindC, editors. TNM Classification of Malignant Tumors. 6th ed. New York: John Wiley & Sons; Kaplan EL, Meier P. Nonparametric estimates from incomplete observations. J Am Stat Assoc. 1958; 53: Ramanath RC, A Hern R, Fisher C, Thomas JM. Modified staging system for extremity soft tissue sarcomas. Ann Surg Oncol. 1999; 6: Posma-Ilic D, Van Voorst Vader PC, Kardaun SH, et al. Epithelioid sarcoom als oorzaak van ulcus cruris. Ned Tijdschr Geneeskd. 1995; 139: Fong Y, Coit DG, Woodruff JM, Brennan MF. Lymph node metastasis from soft tissue sarcoma in adults. Analysis of data from a prospective database of 1772 sarcoma patients. Ann Surg. 1993; 217: Harrison LB, Franzese F, Gaynor JJ, Brennan MF. Long-term results of a prospective randomized trial of adjuvant brachytherapy in the management of completely resected soft tissue sarcomas of the extremity and superficial trunk. Int J Radiat Oncol Biol Phys. 1993; 27: Eggermont AM, Schraffordt Koops H, Lienard D, et al. Isolated limb perfusion with high-dose tumor necrosis factor-alpha in combination with interferon-gamma and melphalan for nonresectable extremity soft tissue sarcomas: a multicenter trial. J Clin Oncol. 1996; 10: Komdeur R, Hoekstra HJ, van den Berg E, et al. Metastasis in soft tissue sarcomas: prognostic criteria and treatment perspectives. Cancer Metastasis Rev. 2002; 21:

48

49 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins 6

50 Abstract Background The aim of the study was to investigate the results of surgical treatment in primary and recurrent dermatofibrosarcoma protuberans (DFSP), with respect to local tumor control. Patients and Methods Thirty-eight patients were treated between at the University Medical Center Groningen (UMCG). Thirty patients presented with primary disease (79%) and 8 patients with locally recurrent disease (21%). The treatment consisted of surgical resection and in case of marginal or positive resection margins (R1 resection) adjuvant radiotherapy. Results Adequate surgical margins as a single modality was associated with 100% local control in all primary DFSPs. Two patients whose resection specimens had microscopically positive resection margins had withdrawn from adjuvant radiotherapy and developed local recurrence (LF rate 7%). Two of the 8 patients referred with a local recurrence developed a second recurrence (LF rate 25%); one of these patients developed distant disease and ultimately died of systemic disease. Non of the five patients with DFSP-FS developed LF after treatment at the UMCG. After a median follow-up of 89 (12-271) months, the 10-year disease-free survival was 85% and the 10-year disease specific survival was 100%. Conclusion After wide surgical resection of a DFSP or DFSP-FS, or an R1 resection combined with adjuvant radiotherapy the risk of local recurrence is extremely low. S.E. ten Heuvel A.J.H. Suurmeijer E. Pras R.J. van Ginkel H.J. Hoekstra European Journal of Surgical Oncology Jan;36(1): Introduction Dermatofibrosarcoma protuberans (DFSP) is a locally aggressive sarcoma arising from the skin. DFSP is relatively rare and accounts for approximately 4% of all soft tissue sarcomas (STS). The tumor presents as an indolent growing plaque, usually localized on the trunk, the proximal extremities, or the head-neck region (Figure 1). DFSP is most often diagnosed in young or middle-aged adults, but it has been described in infants and the elderly as well. Most tumors are less than 5 cm, but untreated DFSP can attain massive dimensions. 1-3 Histologically, DFSP is composed of spindle cells arranged in an irregularly whorled or storiform pattern. The histological diagnosis DFSP can be confirmed reliably with immunohistochemical staining for CD34, which is especially helpful in the differential diagnosis from dermatofibroma, also called benign fibrous histiocytoma (BFH). 4,5 Approximately % of tumors contain a high-grade fibro-sarcomatous component (DFSP-FS), which is associated with a higher risk of recurrence. 6,7 Most DFSP bear a specific chromosomal translocation, t(17;22)(q22;q13), that leads to a fusion between the collagen type 1α1 (COL1A1) on chromosome 17 and the platelet-derived growth factor B-chain gene (PDGFB) on chromosome 22. The COL1A1-PDGFB fusion product is detectable by RT-PCR. 8 DFSP is especially feared for the high tendency to recur, due to its infiltrative asymmetric growth pattern. The histologic tumor margins occasionally extend far beyond the macroscopic boundaries with asymmetric tentacle like projections. Metastases are rare in DFSP and are usually preceded by multiple local recurrences. 1,9,10 DFSP is not particularly chemosensitive, but imantinib mesylate seems to emerge as an effective new targeted therapy for locally advanced or uncontrollable DFSP. 11 There are still many issues to resolve regarding the optimal local treatment of DFSP. In this study we evaluated the outcome of surgical treatment of primary and recurrent DFSP treated at the University Medical Center Groningen over a three decade period. 6 Figure 1. Dermatofibrosarcoma protuberans manifesting as an irregular red plaque on the right shoulder. 85

51 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins Patients and Methods Patients This study includes 38 patients treated for a primary or recurrent DFSP at the University Medical Center Groningen (UMCG) between 1971 and The group consisted of 25 males and 13 females with a median age of 38 (range 8-77) years. Clinical and pathologic features were retrieved from the Groningen Sarcoma Database, patient files and the pathology database. Thirty patients presented with a primary tumor, and 8 patients were referred with a local recurrence after previous surgery. Patient and tumor characteristics are listed in Table 1 and 2. Treatment Wide local skin excision that includes the underlying subcutaneous tissue and the fascia, with histologic confirmed free margins (R0 resection) was the preferred surgical approach. If anatomical structures allowed it, we tried to resect the tumor with a margin of at least 2-3 cm normal tissue to improve local control. Of the 20 patients who were primarily treated at the UMCG 13 underwent a single primary excision and 7 patients had to have a re-excision because of inadequate margins. Of the 10 patients that were referred for re-excision at the UMCG 3 patients had to be operated on twice at our center. Postoperative radiotherapy (50-70 Gy) was administered in case of microscopic involved resection margins (R1 resection), or a narrow excision with no opportunity for further surgical management due to anatomic constraints. In all patients, surgical management was with curative intent. One patient who had a 15 cm large, primarily irresectable DFSP on the proximal arm, was treated with hyperthermic isolated limb perfusion (HILP) with tumor necrosis factor α (TNF α) and Melphalan. 12 After six weeks the tumor had successfully reduced in volume and a radical resection was performed. The wound was closed with a split skin graft. The resection wound was closed primarily in 21 patients, a split skin graft was applied in 13 patients and a plastic surgical reconstruction with a rotation flap was performed in 3 patients. Histopathology and resection margins Histopathology of the resected DFSPs was performed in a standard version, including CD34 immunohistochemistry. Grossly specimens were bread-loafed after inking of the resection margins. The resection margin status was examined in multiple tissue sections taken perpendicular to the nearest margin with at least 1 section per cm. Overall Table 1. Clinical and pathological features of the 30 patients treated for primary dermatofibrosarcoma protuberans at the UMCG No. Sex Male 19 Female 11 Median age (range) 38 (8-77) Median tumor size (cm) (range) 3 (0.5-22) Histo-type DFSP 26 DFSP-FS 4 Localization Trunk 9 Shoulder 7 Upper arm 5 Head-neck 5 Buttock 2 Groin 2 Reason Referral Diagnosis and treatment primary tumor 11 Treatment primary tumor 9 Re-excision primary tumor 10 Treatment Resection 30 Times operated on at the UMCG Once 20 Twice 10 Resection margins Negative 24 Negative, but marginal 3 Positive 3 Adjuvant radiotherapy 5 Survival Median follow-up (range) 80 (12-261) LF 2 6 LF indicates local failure

52 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins Table 2. Clinical and pathological features of the 8 patients that were referred with recurrent dermatofibrosarcoma protuberans Survival (mo) Metastasis (mo) Recurrence after treatment UMCG (mo) Treatment LF (times operated on at the UMCG) (resection margin) (mo) Treatment primary tumor (margins) Localization primary tumor (size (cm)) Histo-type primary tumor Age (yr) at diagnosis primary tumor/sex Reason Referral Patient No. 1 LF 38/M DFSP trunk ( 5) resection (neg) resection (1) (neg) (50) NED (70) 2 LF 29/M DFSP abdomen ( 5) resection (pos) resection (1) (neg) (6) NED (112) 3 LF 35/M DFSP-FS upper arm (5) resection (pos) resection (1) (pos), RT (6) NED (118) 4 LF 38/M DFSP shoulder (NA) resection (pos), resection (1) (neg) (53) NED (156) 5 LF 23/F DFSP trunk (8) resection (NA) resection (1) (neg) (144) LF (81) NED (261) DOD (182) Pulmonary (26) LF (3) resection (1) (neg, marg), RT (53) 6 LF 38/F DFSP shoulder (NA) resection (pos) 7 LF 8/M DFSP trunk (1) resection (NA) resection (1) (NA) (163) NED (269) 8 LF 55/M DFSP head (9) resection (NA), resection (1) (pos), RT (26) NED (251) DOD indicates death of disease; F, female; LF, local failure; M, male; marg, resection margins marginal; mo, months; NED, no evidence of disease; NA, not available; neg, resection margins negative; pos, resection margins positive; Rx, resection margin unknown; RT, radiotherapy; yr, year. a radical excision of the primary tumor was documented in 27 patients (R0). Marginal or microscopically involved resection margins (R1) were documented in 7 patients. In 4 cases of which 3 patients were primarily treated elsewhere, the resection margins of the primary tumor were not documented in the pathology report. The tumors of the three patients that were treated elsewhere were initially not recognized to be malignant and unfortunately not all the tumor tissue material was preserved to allow review of the resection margins retrospectively. Adjuvant radiotherapy Adjuvant radiation (50-70 Gy). was applied in 8 (5 primary and 3 recurrent) patients. One patient with a primary tumor received radiation because of microscopically involved margins with no opportunity for further resection. The two other patients with microscopically involved margins refused adjuvant radiotherapy. Three patients with negative but marginal resection margins also received radiation. The patient who underwent a radical resection after a HILP treatment was treated with adjuvant radiotherapy as well, according to the institutional perfusion protocol. [12] Patient 6 in Tabel 2 was referred after repeated recurrences that were treated surgically elsewhere. We were able to resect the recurrence marginally and adjuvant radiotherapy was administered. Results Patient treated for a primary tumor Two of the 30 patients who were treated for a primary tumor at our institute developed a local recurrence. These were the two before mentioned patients who had an R1 resection and choose to withdraw from adjuvant radiotherapy. The first patient, who had an R1 resection of a DFSP on the trunk, presented again with a local recurrence 81 months after resection of the primary tumor. The local recurrence was resected with microscopically free margins and at last follow-up the patient was disease free (NED 271 months). The second patient, who had had an R1 resection of a primary tumor in the gluteal region and refused adjuvant radiotherapy, presented with a large local recurrence near the anal sphincter 11 months after resection of the primary tumor. The local recurrence was primarily irresectable and the patient was treated with imatinib for two months with a dosage of 400 mg twice a day orally. 11 This resulted in a significant decrease of the tumor volume and after two months a complete resection was performed with preservation of the anal sphincter. Pathologic examination of the

53 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins Mohs micrographic surgery versus WLE Some professionals advocate Mohs micrographic surgery as an appropriate alternative surgical method in particular for DFSP in anatomically challenging areas. 28 In Mohs surgery the tissue layers are sequentially removed and examined microscopically during the operation to determine the extend of tumor invasion. Sequential layers are excised until the malignancy has been removed completely. 28 Mohs surgery is a timeresection specimen revealed a small residual tumor of 0.6 cm that was resected with free margins. At last follow-up the patient was disease free (NED 35 months). Patients referred with a local recurrence Two of the 8 patients that were referred with a local recurrence developed a second recurrence. The recurrences of these two patients showed histologic progression to DFSP-FS. As mentioned above one of these two patients had been treated for several local recurrences before referral to the UMCG. Despite adjuvant radiotherapy to the site of the local recurrence the patient developed a new recurrence and also a lung metastasis after 26 months follow-up, which was removed by thoracotomy. Pathologic review revealed a 3.5 cm large metastasis that had progressed into DFSP-FS. The tumor was removed completely and adjuvant radiotherapy was administered. Although the lungmetastasis was removed completely, the patient lateron developed new and extensive pulmonary and soft tissue metastases which were from then on treated palliatively with radiotherapy. This patient eventually died of disseminated disease 15 years after the initial treatment of the primary tumor. None of the other 37 patients developed metastasis. Local control rate None of the patients with negative surgical margins, as a single modality, developed a local recurrence, resulting in a 100% local control rate after adequate resection. Of the 8 patients who received adjuvant radiation, only the patient that had been referred after treatment of repeated local recurrence developed a local recurrence resulting in a 82.5% local control rate. Survival At last follow-up 9 of 10 patients that were treated for a local recurrence were disease free. The median follow-up of the entire group was 89 (12-271) months. Ten-years disease free survival (DFS) was 85% and 10-years disease specific survival (DSS) was 100%. Discussion Local recurrence DFSP is known for its high tendency to recur locally after surgical resection. Recurrence rates reported in literature range from 0-60% The chance however of developing a local recurrence was 7% in the UMCG series, which is comparable to low numbers reported by DuBay et al and Fiore et al. 17,18 Although local recurrences usually occur within three years after surgical resection of the primary tumor, late relapses may occur. 20,21 Wide local excision (WLE) with histologically confirmed free margins (R0 resection) is the mainstay of surgical treatment of DFSP. 1,14,15,18,22 In this series adequate surgical margins as a single modality was associated with 100% local control in all primary DFSPs. Resection margins The precise adequate resection margins to achieve local control remain unclear, since the level of evidence is low. Prospective randomized trials assessing resection margins of healthy tissue cannot be performed, due to the rarity of the disease. Kimmel et al investigated resection margins in DFSP and concluded that the clinically apparent size of a DFSP was a poor indicator of the true histologic tumor extent. 23 In a recently published review by Lemm et al it is reported that studies in which DFSP was excised with undefined or narrow margins, show local failure rates ranged from 26-60%. When a wide local excision (> 2cm) was performed, the local failure rate varied from 0-41%, with a total local failure rate of 8.8%. 19 In our experience some patients presented with long-existing lesions that were resected several times before. In some cases the patient was initially treated without recognition of the correct diagnosis. 14 Patients that presented in our hospital with recurrent lesions had a higher chance of developing another local recurrence and one of these patients eventually developed metastasis. Sometimes the clinical assessment of optimal surgical margins is restricted by anatomical structures. DFSP may arise in the head and neck area, for example, where it is not always possible to perform a wide local excision with 3-cm free margins. In case of a marginal or positive resection and no opportunity for further surgical treatment due to anatomical limitation, postoperative radiotherapy is indicated All patients that had positive or marginal resection margins of a primary tumor and received adjuvant radiotherapy remained disease free during follow-up. Strikingly the two patients that had chosen to withdraw from the indicated adjuvant radiotherapy both developed local relapse

54 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins consuming technique though, and for DFSP it is sometimes difficult to differentiate scattered malignant spindle cells from normal fibroblasts in cryostat frozen tissue sections. 19,29 Although CD34 immunostaining is positive in 90% of DFSP it can be negative in nodular areas of the tumor and it is therefore not a 100% reliable tool to detect residual tumor. 29,30 Our results show that DFSP and recurrent DFSP can usually be well managed with WLE as a single modality or if indicated combined with radiotherapy. Surgical reconstruction In case of extended resection, split skin grafts or plastic surgical reconstruction can be applied to cover the surgical defect. If a reconstruction is required, it is important that there be no doubt about the resection margin, since the (rotation) flap may prevent early detection of a local recurrence. In our series a plastic surgical reconstruction with a rotation flap was employed in three patients and all three were performed in one tempi. In case of initially doubtful resection margins however, extensive reconstructions should be delayed until appropriate surgical margins are confirmed by pathologic examination. 31,32 Metastasis Only one patient in our series developed pulmonary metastasis, which was from a locally recurred tumor on the shoulder that had histologically progressed to DFSP- FS. Metastasis is rare in DFSP and is usually preceded by multiple local recurrences. 3 DFSP-FS progressed tumors are associated with a more aggressive clinical course, and a significantly higher chance of metastasis than classic DFSP. 6,7 The four patients that were initially diagnosed with DFSP-FS in our series remained disease free though, with relatively long-term follow-up. Imatinib Imatinib mesylate has recently become available as a targeted therapy for primarily irresectable or metastatic DPFS. Unfortunately the patient that had developed pulmonary metastasis in our series had died before imatinib had been developed and recognized for DFSP. Lateron another patient in our series who had developed a large irresectable local recurrence in the gluteal region was successfully treated with imatinib prior to resection. The imatinib treatment had effectively reduced the tumor mass and an R0-resection could be performed with preservation of the anal sphincter. Few similar case studies have been reported and preliminary trial results seem promising. There is no role for imatinib in the adjuvant treatment of DFSP though, as the use of imatinib is still experimentally and no prospective data are currently available to prove a benefit in terms of progression free and overall survival and theoretically adjuvant use of imatinib may even facilitate the growth of resistant clones. The precise role of imatinib in a limited subgroup of locally advanced or uncontrollable DFSP is currently under intensive investigation in clinical trials. 11,33-35 Conclusion Clinicians often have to deal with unfavourable presentations of DFSP, due to patient delay, initial misdiagnosis and unfavourable anatomic localisation. The mainstay in treatment of DFSP is resection with a wide surgical margin (> 2 cm). If adequate surgical margins cannot be obtained without severe functional or cosmetic impairments postoperative radiotherapy should be applied. References 1. Kleihues P, Sobin L, Dermatofibrosarcoma protuberans. In: LeBoit PE, Burg G, Weedon D, Sarasin A. World Health Organization Classification of Tumours: pathology and Genetics of Skin Tumors. Lyon: IARC Press; 2007; Toro JR, Travis LB, Wu HJ, Zhu K, Fletcher CD, Devesa SS. Incidence patterns of soft tissue sarcomas, regardless of primary site, in the surveillance, epidemiology and end results program, : An analysis of 26,758 cases. Int J Cancer. 2006;19: Mendenhall WM, Zlotecki RA, Scarborough MT. Dermatofibrosarcoma protuberans. Cancer. 2004;101: Kutzner H. Expression of the human progenitor cell antigen CD34 (HPCA-1) distinguishes dermatofibrosarcoma protuberans from fibrous histiocytoma in formalin-fixed, paraffin-embedded tissue. J Am Acad Dermatol. 1993;28: Abenoza P, Lillemoe T. CD34 and factor XIIIa in the differential diagnosis of dermatofibroma and dermatofibrosarcoma protuberans. Am J Dermatopathol. 1993;15: Abbott JJ, Oliveira AM, Nascimento AG. The prognostic significance of fibrosarcomatous transformation in dermatofibrosarcoma protuberans. Am J Surg Pathol. 2006;30: Mentzel T, Beham A, Katenkamp D, Dei Tos AP, Fletcher CD. 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: Sandberg AA, Bridge JA. Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors. Dermatofibrosarcoma protuberans and giant cell fibroblastoma. Cancer Genet Cytogenet. 2003;140:

55 Dermatofibrosarcoma protuberans: recurrence is related to the adequacy of surgical margins 9. Rutgers EJ, Kroon BB, Albus-Lutter CE, Gortzak E. Dermatofibrosarcoma protuberans: treatment and prognosis. Eur J Surg Oncol. 1992;18: Häfner HM, Moehrle M, Eder S, Trilling B, Röcken M, Breuninger H. 3D-Histological evaluation of surgery in dermatofibrosarcoma protuberans and malignant fibrous histiocytoma: differences in growth patterns and outcome. Eur J Surg Oncol. 2008;34: McArthur GA, Demetri GD, van Oosterom A, Heinrich MC, Debiec-Rychter M, Corless CL, et al. Molecular and clinical analysis of locally advanced dermatofibrosarcoma protuberans treated with imatinib: Imatinib Target Exploration Consortium Study B225. J Clin Oncol 2005; 23: Van Ginkel RJ, Thijssens KM, Pras E, Van Der Graaf WT, Suurmeijer AJ, Hoekstra HJ. Isolated limb perfusion with tumor necrosis factor alpha and melphalan for locally advanced soft tissue sarcoma: three time periods at risk for amputation. Ann Surg Oncol. 2007;14: Monnier D, Vidal C, Martin L, Danzon A, Pelletier F, Puzenat E, et al. Dermatofibrosarcoma protuberans: a population-based cancer registry descriptive study of 66 consecutive cases diagnosed between 1982 and J Eur Acad Dermatol Venereol. 2006;20: Lindner NJ, Scarborough MT, Powell GJ, Spanier S, Enneking WF. Revision surgery in dermatofibrosarcoma protuberans of the trunk and extremities. Eur J Surg Oncol. 1999;25: Bowne WB, Antonescu CR, Leung DH, Katz SC, Hawkins WG, Woodruff JM, et al. Dermatofibrosarcoma protuberans: A clinicopathologic analysis of patients treated and followed at a single institution. Cancer. 2000;88: Chang CK, Jacobs IA, Salti GI. Outcomes of surgery for dermatofibrosarcoma protuberans. Eur J Surg Oncol. 2004;30: DuBay D, Cimmino V, Lowe L, Johnson TM, Sondak VK. Low recurrence rate after surgery for dermatofibrosarcoma protuberans: a multidisciplinary approach from a single institution. Cancer. 2004;100: Fiore M, Miceli R, Mussi C, Lo VS, Mariani L, Lozza L, et al. Dermatofibrosarcoma protuberans treated at a single institution: a surgical disease with a high cure rate. J Clin Oncol. 2005;23: Lemm D, Mügge LO, Mentzel T, Höffken K. Current treatment options in dermatofibrosarcoma protuberans. J Cancer Res Clin Oncol. 2009;135: Swan MC, Banwell PE, Hollowood K, Goodacre TE. Late recurrence of dermatofibrosarcoma protuberans in the female breast: a case report. Br J Plast Surg. 2005;58: Orlando R, Lumachi F, Lirussi F. Recurrent dermatofibrosarcoma protuberans sixteen years after radical excition. A case report. Anticancer Res. 2003;23: Khatri VP, Galante JM, Bold RJ, Schneider PD, Ramsamooj R, Goodnight JE, Jr. Dermatofibrosarcoma protuberans: reappraisal of wide local excision and impact of inadequate initial treatment. Ann Surg Oncol 2003;10: Kimmel Z, Ratner D, Kim JY, Wayne JD, Rademaker AW, Alam M. Peripheral excision margins for dermatofibrosarcoma protuberans: a meta-analysis of spatial data. Ann Surg Oncol 2007;14: Ballo MT, Zagars GK, Pisters P, Pollack A. The role of radiation therapy in the management of dermatofibrosarcoma protuberans. Int J Radiat Oncol Biol Phys. 1998;40: Suit H, Spiro I, Mankin HJ, Efird J, Rosenberg AE. Radiation in management of patients with dermatofibrosarcoma protuberans. J Clin Oncol. 1996;14: Dagan R, Morris CG, Zlotecki RA, Scarborough MT, Mendenhall WM. Radiotherapy in the treatment of dermatofibrosarcoma protuberans. Am J Clin Oncol. 2005;28: Haas RL, Keus RB, Loftus BM, Rutgers EJ, Coevorden F van, Bartelink H. The role of radiotherapy in the local management of dermatofibrosarcoma protuberans. Soft Tissue Tumours Working Group. Eur J Cancer. 1997;33: Snow SN; Gordon EM; Larson PO; Bagheri MM; Bentz ML; Sable DB Dermatofibrosarcoma protuberans: a report on 29 patients treated by Mohs micrographic surgery with long-term follow-up and review of the literature. Cancer 2004 Jul 1;101(1): Massey RA, Tok J, Strippoli BA, Szabolcs MJ, Silvers DN, Eliezri YD. A comparison of frozen and paraffin sections in dermatofibrosarcoma protuberans. Dermatol Surg 1998;24: Garcia C, Viehman G, Hitchcock M, Clark RE. Dermatofibrosarcoma protuberans treated with Mohs surgery. A case with CD34 immunostaining variability. Dermatol Surg 1996;22: Brabant B, Revol M, Vergote T, Servant JM, Banzet P. Dermatofibrosarcoma protuberans of the chest and the shoulder: Wide and deep excision with immediate reconstruction. Plast. Reconstr. Surg. 1993;92: Arnaud EJ, Perrault M, Revol M, Servant JM, Banzet P. Surgical treatment of dermatofibrosarcoma protuberans. Plast Reconstr Surg. 1997;100: Mehrany K, Swanson NA, Heinrich MC, Weenig RH, Lee KK, White CR, Jr., et al. Dermatofibrosarcoma protuberans: a partial response to imatinib therapy. Dermatol Surg. 2006;32: Sirvent N, Maire G, Pedeutour F. Genetics of dermatofibrosarcoma protuberans family of tumors: from ring chromosomes to tyrosine kinase inhibitor treatment. Genes Chromosomes Cancer. 2003;37: Labropoulos SV, Fletcher JA, Oliveira AM, Papadopoulos S, Razis ED. Sustained complete remission of metastatic dermatofibrosarcoma protuberans with imatinib mesylate. Anticancer Drugs. 2005;16:

56

57 Summary 7

58 Summary Chapter 1 comprises a basic introduction of soft tissue sarcomas in general and an introduction of the specific diagnostic and prognostic aspects that were presented in this thesis. Chapter 2 describes and compares the diagnostic accuracy of RT-PCR and FISH techniques applied on material of 50 routinely processed synovial sarcomas and 12 histologic mimics of synovial sarcoma. Both methods were applied on formalin-fixed and paraffin-embedded tumor tissue. RT-PCR and FISH had a sensitivity of 94% and 82%, a specificity and positive predictive value of 100% and 100% and a negative predictive value of 80% and 75%, respectively. We concluded that RT-PCR and FISH are highly reliable techniques for diagnosing synovial sarcomas. RT-PCR had a slightly higher sensitivity than FISH. FISH however allows microscopic control of a true positive result within the histological context. Using FISH as a method of first choice, 80% of synovial sarcomas can be diagnosed by FISH only and 20% would need to be confirmed by RT-PCR. In chapter 3 we aimed to assess factors that are supposed to influence the survival of synovial sarcoma, including the role of SYT-SSX fusion type and the immunohistochemical marker ezrin. These factors were investigated in 45 patients with diagnosis of synovial sarcoma confirmed in the study reported in chapter 2. Classical prognostic factors included age at diagnosis, tumor site, tumor size, tumor histology (biphasic versus monophasic) mitotic count, necrosis and histologic grade (according to the definitions described by the FNCLCC grading system). SYT-SSX fusion type was determined by RT-PCR. Ezrin immunostaining was scored by intensity and analyzed for prognostic value. Disease stage at presentation, tumor size and tumor grade were significant predictors of survival in synovial sarcoma. SYT-SSX fusion type was not correlated with survival in our series. Ezrin expression levels were not discriminative in predicting outcome. 7 Chapter 4 encompasses a retrospective clinicopathological prognostic study of myxoid liposarcoma patients. Forty-nine patients were included with relatively long term follow-up. The prognostic effect of myxoid, transitional, and round cell morphology in addition to other more established clinicopathological prognostic factors was studied and discussed with respect to the existing literature. Tumor size, age at presentation, and tumor grade were associated with worse DSS by univariate analysis, whereas tumor grade (grade 2 being defined by 5% RCLS morphology) was the only independ- 99

59 Summary ent prognostic factor that remained significant by multivariate analysis. A transitional morphology with increased cellularity was not associated with a worse outcome. In chapter 5 we address pathological and clinical prognostic indicators for epithelioid sarcoma based on a retrospective study of 23 patients treated at the University Medical Center Groningen (UMCG) and the Radboud University Medical Center, Nijmegen (Radboudumc). The 5-year and 10-year survival rate was 34% and 17% respectively. Disease stage at presentation and tumor size > 5 cm at diagnosis were significant predictors of poor survival. Survival was relatively poor because patients often present with extensive disease, lymph node metastases, and/ or distant metastases. Even after surgical treatment with curative intent a high percentage of patients develops local recurrence, lymph node metastases, and distant metastases, which is characteristic of epithelioid sarcomas. The risk of local recurrence was determined by the radicality of the excision. In chapter 6 the results of surgical treatment in primary and recurrent dermatofibrosarcoma protuberans are described for 38 patients that were treated for a primary or recurrent dermatofibrosarcoma protuberans at the UMCG. The chance of developing a local recurrence was 7% for patients that were initially treated at the UMCG. Resection with adequate surgical margins as a single modality was associated with a 100% local control in all primary tumors. Some patients presented with long-existing lesions that had been resected several times before. In some cases the patient was initially treated without recognition of the correct diagnosis. Patients that presented in our hospital with recurrent lesions had a higher chance of developing another local recurrence and one of these patients eventually developed metastasis. All patients that had positive or marginal resection margins of a primary tumor and received adjuvant radiotherapy remained disease free during follow-up. Two patients that had withdrawn from adjuvant radiotherapy both developed local relapse. 100

60

61 Samenvatting 8

62 Samenvatting Hoofdstuk 1 betreft een inleiding in de weke delen sarcomen in het algemeen en de specifieke diagnostische en prognostische aspecten die in dit proefschrift worden beschreven. Hoofdstuk 2 beschrijft de betrouwbaarheid van RT-PCR en FISH als diagnostische methoden. Beide methoden werden getest op het materiaal van 50 synoviaal sarcomen en 12 tumoren die histologisch op het synoviaal sarcoom lijken. De methoden werden toegepast op formaline-gefixeerd en in paraffine ingebed tumor weefsel. RT- PCR en FISH hadden een sensitiviteit van respectievelijk 94% en 82%, een specificiteit en positief voorspellende waarde van respectievelijk 100% en 100% en een negatief voorspellende waarde van 80% en 75%. We concluderen dat RT-PCR en FISH zeer betrouwbare technieken zijn om het synoviaal sarcoom te diagnosticeren. RT-PCR had een iets hogere sensitiviteit dan FISH, echter met FISH kan men het positieve resultaat microscopisch in de histologische context controleren. Bij het gebruik van FISH als methode van eerste keuze kan 80% van de synoviaal sarcomen gediagnosticeerd worden door alleen gebruik te maken van FISH, de overige 20% zal bevestigd moeten worden door RT-PCR. Hoofdstuk 3 betreft een studie naar de rol van factoren die verondersteld worden de prognose van het synoviaal sarcoom te beïnvloeden, inclusief de controversiële rol van het SYT-SSX fusie type en de hypothetische rol van de immunohistochemische marker ezrin. Deze factoren werden onderzocht in een groep van 45 patiënten waarvan de diagnose synoviaal sarcoom moleculair werd bevestigd in de eerdere studie beschreven in hoofdstuk 2. Klassieke prognostische factoren die bestudeerd werden betroffen: leeftijd bij diagnose, de lokalisatie van de tumor, de grootte van de tumor, de histologie van de tumor (bifasisch versus monofasisch), de mitotische activiteit, necrose en de histologisch gradering (volgens de definities van het FNCLCC graderingsysteem). Het SYT-SSX fusie type werd vastgesteld door middel van RT-PCR. De ezrinimmunokleuring werd gescoord op intensiteit en vervolgens geanalyseerd op prognostische waarde. Het stadium van de ziekte bij presentatie, tumor grootte en tumor graad waren significante voorspellers van de overleving van synoviaal sarcoom patiënten. In onze serie was het SYT-SSX fusie type niet gecorreleerd met de overleving. De ezrin expressie had geen onderscheidende rol bij het voorspellen van de prognose. 8 Hoofdstuk 4 betreft een retrospectieve klinisch-pathologische studie naar de prognose van myxoid liposarcoom patiënten. Negenenveertig patiënten werden geïnclu- 105

63 Samenvatting deerd met een relatief lange follow-up. Naast de bekende klinisch-pathologische fac toren werd de prognostische waarde van myxoid, transitionele en rondcellige morfologie onderzocht en vergeleken met de bestaande literatuur. Tumor grootte, leeftijd bij presentatie en tumor graad werden gecorreleerd met een slechtere DSS door middel van univariate analyse, terwijl tumor graad (graad 2 werd gedefinieerd door 5% RCLS morfologie) de enige onafhankelijke prognostische factor was die significant bleef in de multivariate analyse. Transitionele morfologie met verhoogde cellulariteit konden niet gecorreleerd worden met een slechtere prognose. In hoofdstuk 5 beschrijven we pathologische en klinische prognostische factoren bij het epithelioid sarcoom, retrospectief in kaart gebracht met de gegevens van 23 patiënten die behandeld zijn in het Universitair Medische Centrum Groningen (UMCG) en het Radboud Universitair Medisch Centrum, Nijmegen (Radboudumc). De 5-jaars en 10-jaars ziektevrije overleving betroffen respectievelijk 34 % en 17 %. Ziektestadium bij presentatie en tumor grootte > 5 cm bij de diagnose waren significante voorspellers van een slechte overleving. De overleving was relatief slecht omdat patiënt vaak kwamen met uitgebreide ziekte, lymfklier metastasering en/ of afstandsmetastasering. Zelfs na chirurgische behandeling met een curatieve opzet ontwikkelde een groot percentage van de patiënten lokale recidivering, lymfklier metastasering, en afstandsmetastasering, wat karakteristiek is voor het epithelioid sarcoom. Het risico op lokale recidivering werd bepaald door de radicaliteit van de excisie. In hoofdstuk 6 worden resultaten van de chirurgische behandeling van primaire en gerecidiveerde dermatofibrosarcoma protuberans beschreven. De kans om een lokaal recidief te ontwikkelen was 7% voor de patiënten die initieel in het UMCG werden behandeld. Resectie met adequate marges als enige behandelvorm leidde tot 100% lokale tumor controle van alle primaire tumoren. Enkele patiënten presenteerden zich met al lang bestaande laesies die al enkele malen eerder waren geopereerd. In sommige gevallen was de patiënt behandeld zonder dat de diagnose dermatofibrosarcoma protuberans gesteld was. Patiënten die bij ons kwamen met recidief laesies hadden een hogere kans om opnieuw een recidief te ontwikkelen en een van deze patiënten ontwikke l- de uiteindelijk een metastase. Alle patiënten die een positief snijvlak hadden van de primaire tumor en aanvullende radiotherapie kregen bleven ziekte vrij gedurende de follow-up. Twee patiënten die geen aanvullende radiotherapie kregen ontwikkelden een lokaal recidief. 106

64

65 General discussion and future perspectives 9

66 General discussion and future perspectives Determining the optimal treatment regimen for the individual soft sarcoma patient is a complex cascade of individually tailored empirical decisions by a collaboration of dedicated multidisciplinary sarcoma-specialists. Despite the recognition and better understanding of the distinctive clinicopathological behavior of the different soft tissue sarcoma (STS-) types only few entities have a subtype-specific evidence based treatment protocol. While there is a lack of effective subtype-specific treatment modalities current soft tissue sarcoma treatment guidelines still mainly provide general treatment principles for almost all STS entities. 1-5 For localized disease quite general surgical principles can usually be applied, but in particular for more advanced STS new targeted therapies are desperately needed, as most STS are not particularly sensitive to conventional chemotherapy. Most probably real breakthroughs in the treatment of progressed STS will have to emanate from a better understanding of the basic molecular biology of the specific STS. Imatinib mesylate forms the paradigm for the translation of tumor-biology research into an effective targeted therapy for gastrointestinal stromal tumors (GIST) that also turned out to be effective for progressed and uncontrollable dermatofibrosarcoma protuberans (DFSP). 6,7 Until more new targeted agents are developed it remains essential to validate and expand our understanding of the prognosis of STS and factors that are associated with prognostic outcome. Thorough understanding of these parameters is needed to enable better selection of patients for adjuvant therapy and stratification for clinical trials that delineate the best treatment regimens of existing chemotherapeutic agents, radiotherapy as well as newly developed targeted therapies. For long various STS subtypes were lumped together when studying prognostic factors and this may have masked specific behavioral and prognostic features of the individual subtypes. 8,9 The series that were studied in this thesis reflect the behavioral heterogeneity of STS and the necessity of accurate subtyping: the four specific STS types that were analyzed show quite distinctive characteristics, as well in their morphological appearance, their clinical presentation, anatomic localization, as in their metastatic behavior, obviously requiring different clinical approaches. We therefore underscore that STS should be handled as separate entities and that the results of prognostic studies should be presented categorized by diagnostic subtype Diagnosis Indisputably a correct diagnosis and pathological analysis is crucial. Studies that have reviewed the initial diagnosis of soft tissue sarcomas found a large number of discrepancies between the initial diagnosis (based on pathological features) and when 9 111

67 General discussion and future perspectives reviewed by a specialty panel of soft tissue sarcoma expert pathologists. 15,16 Logically, the consequences of a misdiagnosis may be dramatic for the individual patient. Although we did not evaluate the concordance of diagnosis systematically in this thesis, the series of patients with DFSP presented in chapter 6 show that patients that had a delayed diagnosis, or were primarily misdiagnosed had a higher chance of developing another local recurrence. DFSP requires a very specific approach because of the local infiltrative growth and initial misdiagnosis may lead to multiple reresections. Delayed treatment did not shorten the survival of DFSP as even progressive disease can usually be treated curatively as was demonstrated in this thesis, but extensive reresections may impact severely upon functional and cosmetic results. In contrast to DFSP, it can be assumed that a misdiagnosed synovial sarcoma (SS), myxoid liposarcoma (MLS) or especially epithelioid sarcoma (ES) will significantly reduce survival. In addition to the consequences for the individual patient, inaccurate diagnosis will also considerably decrease the quality of prognostic research. Although the pathological diagnosis of the four STS types in our thesis can usually be assessed reliably by routine light microscopic examination of the tumor histology complemented with immunohistochemistry, some STS harbor overlapping immunohistochemical features, especially some monophasic SS and poorly differentiated SS. The specific genetic translocations that are found in SS and MLS have proven to be powerful diagnostic markers that can be used in case of inconclusive immunohistochemistry. For daily practice detection of the translocation will not supplant immunohistological assessment by the pathologist, but especially for research purposes it can be used as a highly objective and consistent standard. 17,18 In chapter 3 we tested the diagnostic accuracy of RT-PCR and a commercially available FISH technique in a series of 50 SS tumors and 12 histological mimics. We demonstrated that both methods are highly accurate for diagnosing SS and can be routinely used. A practical advantage of these techniques is that they can be applied to small amounts of tumor material, making them useful for small biopsy specimens, as well as residual tumor tissue or disseminated disease. These techniques will help enhance the quality of prognostic studies by enabling more consistent and accurate prognostification and allowing larger numbers to be included for research purposes as archival tumor material can be accurately reviewed retrospectively. Future challenges include improvement of these diagnostic techniques by making the procedure more time-efficient, for instance for FISH with the development of automated counting systems of FISH signals. Perhaps the molecular detection of the SYT-SSX translocation by FISH will also provide chances for high sensitivity detection of submicroscopic disease, such as micrometastasis in the near-by future. Whereas our study was purely focused on the accuracy of RT-PCR and FISH in diagnosing SS, Neuville et al recently demonstrated the practical diagnostic utility of molecular analysis in an epidemiological, population based series. They included all tumors that were suspicious of sarcoma types that harbor specific genetic translocations or mutations or that had such a diagnosis in the differential diagnosis. On basis of histology, immunohistochemical analysis and the clinical context the premolecular diagnosis was formulated as certain, probable or possible. Among other types of sarcomas they respectively collected 68, 34 and 95 patients with a tumor suspicious of SS, MLS and DFSP. RT-PCR and/ or FISH was applied to most cases. Molecular analysis was applicable and accurate for all three types, but in particularly useful to definitively confirm a probable diagnosis and in some cases to allow or out rule a possible diagnosis. Although it was demonstrated for all three sarcoma types that in few specific cases the premolecular diagnosis was adjusted after molecular analysis, we agree with the authors that molecular analysis is more critical in SS given the prognostic and therapeutic complications, and less necessary for diagnosing DFSP and MLS unless the diagnosis is uncertain or unspecific. In this matter it should also be kept in mind that a negative result of molecular analysis does not rule out the diagnosis as the specific translocations are not found in 100% of cases. The authors also underline that histology remains the basis for diagnosing STS and they emphasize the importance of complete and systematic interpretation of histology, immunohistochemistry, clinical context and in some cases complemented with molecular analysis. 19 Currently, in case of MLS neoadjuvant radiotherapy has become part of the standard treatment regimen, henceforth preclinical tumor tissue samples are usually limited to biopsy material In our retrospective series on MLS we were able to retrieve sufficiently large samples of preclinical tissue material to perform an accurate diagnosis based on histology complemented with immunohistochemistry, but as variable morphologic regions generally coexist in one tumor, molecular analysis is becoming more essential in the diagnosis of MLS/ RCLS as well. Meanwhile also novel diagnostically applicable immunohistochemical markers are being identified. For example, recently Valte et al showed that TLE1 is a highly specific diagnostic marker for SS. TLE1 may especially be useful when molecular tests are noninterpretable due to technical reasons or in rare cases of a cryptic SYT-SSX rearrangement. 25 The complexity of prognostic research in STS Prognostic research in STS is complicated due to the heterogeneity of their behavior in combination with the low incidence of STS(sub-)types. 10 This is an extremely imprac

68 General discussion and future perspectives ticable combination, continuously challenging researchers to choose the most optimal methodological approaches but at the same time forcing them to make compromises. This complexity is reflected in our work and involves some limitations. The cases that were included in our series were collected and studied retrospectively. The relatively older cases provided our studies with the advance of long-term follow-up, but disadvantageously not all patients were treated homogeneously and/ or fully according to current treatment protocols. 26 On the other hand we tried to collect the cases for our studies carefully for proper diagnosis, availability of pretreatment tumor tissue material, representative treatment and sufficient clinical data, which may have led to a concomitant selection bias. It is important to point out that sarcomas do not only show substantial heterogeneous behavior themselves, they also present within a wide variety of clinical situations further complicating the clinico-pathological analysis of prognostic factors and subsequently impeding the generalizability of the results and thus the translation back to the individual patient. 27 Literature on prognostic STS research reflects this heterogeneity, as for example different studies use a broad variety of in- and exclusion criteria, follow-up end-points and parameters with different cut-of thresholds, thereby hampering optimal systematic comparison between different studies. This complexity causes the difficulty but at the same time the rationale for improvement of prognostic research in STS. In this thesis we demonstrated that it is worthwhile to reevaluate classical prognostic factors in well-selected subtype specific series. First we will discuss the most important results of our prognostic evaluations, later on we will make suggestions for future methodological refinement. Local tumor control In general, local tumor control has improved tremendously due to advanced surgical techniques, imaging techniques and the use of radiotherapy. Although the main general principles of conservative surgical treatment apply quite well to most STS subtypes, there are various unique subtype-specific aspects that are of importance in the treatment of localized STS, like the preferred anatomical localization, the local growth pattern and the sensitivity to radiotherapy and to systemic therapy. 28,29 Generally, wide local excision (WLE) with histologically confirmed free margins is the mainstay of surgical treatment. Assessing the exact resection margins that are needed to achieve local control is often complicated as STS present at diverse and sometimes unfavorable anatomical localizations and the incidence is low. In the subtype specific series of MLS and DFSP presented in this thesis we retrospectively validated that adequate surgical margins as a single modality usually gave 100% local tumor control in patients with a primary tumor and it was confirmed that in case of microscopically positive resection margins adjuvant radiotherapy gave a better local control rate. 21,22,30-38 Whereas in SS, MLS and DFSP local recurrence did not directly influence survival, a locally recurred ES is strongly associated with very poor survival outcome, reflecting the aggressive behavior of ES. Radiotherapy was also administered in our series with ES, but the number of patients was too small to draw solid conclusions about the effect of radiotherapy on the local control of ES. 39 Metastasis Not surprisingly our results in SS and ES confirmed that metastasis at the time of presentation is generally the strongest variable affecting outcome. Because of the known similar impact of metastasis on the prognosis in MLS we choose to only select cases with localized disease in our series of MLS. In our series of SS 11% of the cases that were included had metastasis at the time of diagnosis. For ES even 48% of the patients had distal failure at the time of presentation. In contrast to patients with localized disease, the prognosis of patients with metastatic disease has changed very little over the past four decades. 40 For the entities that we studied in this thesis the efficacy of chemotherapeutic agents is limited and so far novel targeted therapies have not been developed for these types of STS, with exception of imatinib for DFSP. SS, MLS, ES and DFSP show very diverge metastatic behavior, as well in their metastatic potential, as the time to appearance of metastasis, as in their routes of metastatic spread. Staging was not studied systematically in our work but it may be obvious that the different subtypes require slightly different staging approaches and follow-up care. Especially the metastatic behavior of MLS is remarkable, as metastasis may develop after many years and in unusual soft tissue localizations, as was the case for some patients in our series as well. 23,41-46 Most probably these second tumors are not primary tumors but metastasis as it was shown by De Vreeze et al that the first and the second tumor share similar specific genetic profiles. 47 It is evident that metastatic MLS cells have a strong predilection for fatty tissues, but it remains a puzzle how these silent cells can be indolent for so long and what triggers them to become activated after so many years. Also according to newer insights, the two patients in our series of MLS who had a retroperitoneal tumor must actually have either had a well or dedifferentiated liposarcoma or a retroperitoneal metastasis of a primary MLS outside the retroperitoneum. 48 Particularly challenging to control are ES due to their high propensity for lymphatic and distant metastasis. Overall about three-quarter of the ES patients in our cohort presented with or developed metastasis after a relatively short term follow-up. Of

69 General discussion and future perspectives the patients that were treated curatively, about half developed local recurrence and in most cases synchronal or subsequently distant metastasis. All patients that developed distant metastasis eventually died of disease. These are mostly relatively young adults in their 2 nd -4 th decade of life. Of interest is the prognostic and therapeutic significance of pathological staging of the regional lymph nodes in ES. 49 Larger scale validation is needed but as ES are so extremely rare it is unlikely that adequately powered trials will be established. 50,51 In contrast tot ES, DFSP seldom metastasize. One patient in our series developed pulmonary metastasis after multiple local recurrences and unfortunately for this patient she died before imatinib was introduced, which could probably have prolonged her life significantly. The importance of tumor grade In case of localized disease, our results demonstrate that grade is sustainable as a prognostic factor in relatively small practice based series of SSs and MLSs. Although we confirmed the prognostic value of grade for SS and MLS, grade should always be interpreted most carefully, as it is a conceptual approach. The first grading systems for STS were bases on different sorts of STS lumped together and STS histotype was included in the grading formula as a parameter. The one-model-fits all approach is still used in STS research publications, despite the growing consensus that grading systems should be modified for each sarcoma type. 11,52 Too excessive subdivision will reduce the prognostic value, but we pose that for those tumors that can be diagnosed accurately grading should be applied stratified for histotype. It can be rationalized that for established STS entities like SS and MLS, studied in our thesis, histotype is a consistent factor that should not be taken out of its context, therefor we agree that grade should be used complementary to the histologic typing, as stated by Coindre et al. For SS we used the three-tierced FNCLCC grading system that includes histological type, mitotic rate and necrosis. 56 Some guidelines endorsed a two-tierced grading version and consider all SS as high grade tumors. 57,58 Our results confirm that it is worthwhile to differentiate between grade 2 and grade 3 for SSs, as patients with grade 3 tumors had a significantly poorer outcome compared to patients with grade 2 tumors. Reasoning adversely, we observed that the primary tumors of SS patients with metastasis at the time of diagnosis mostly were grade 3. The patients that presented with metastasis at the time of diagnosis also had significantly larger tumors, while grade and size were both independently correlated in multivariate analysis. In our study of clinicopathological prognostic factors in the series of patients with MLS we demonstrated that tumor grade was the only independent prognostic factor by multivariate analysis. MLSs hardly show mitoses and the FNCLL grading system differentiates grade 1 and grade 2 MLS based on the presence of (> 5%) round cell morphology. In addition we investigated whether the transitional type of MLS with increased cellular morphology was also correlated with outcome, but this hypothesis could not be confirmed in our analysis. It is noteworthy to mention that a similar histologic analysis of cellularity and tumor grade like we performed, cannot easily be replicated with patients that are treated nowadays as neoadjuvant radiotherapy has currently become standard treatment for MLS, causing that pretreatment tumor material is usually limited to biopsy material. 55 Novel markers The search for the ideal prognostic biological marker is ongoing: a marker that is highly specific, objective and reproducible. Various areas are being explored intensively for markers that might contribute to the prediction of outcome and treatment results in STS. Many prospect markers are being supposed, but pioneering in this largely unexplored territory implies chances but also means dealing adequately with negative results. Inspired by previous findings in osteosarcoma and rhabdomyosarcoma reported by others, we hypothesized that the level of ezrin expression in SS might correlate with patient outcome Our series of molecularly confirmed SS with long term follow up provided us with a reliable cohort to explore a possible correlation between ezrin expression levels and prognostic outcome in SS. Our analysis showed that ezrin expression levels were not discriminative in predicting outcome though. Besides immunohistochemistry, new prognostic and predictive markers may be derived from the specific genetic alterations in STS, as has been established in other oncology fields like leukemias, lymphomas and breast cancer. We analyzed whether the two SYT-SSX different fusion types were correlated with outcome in our series of SS. Earlier, two large multi-center studies reported contradictory results regarding an assumed correlation between fusion type and patient outcome, but clearly in our series of SS solely fusion type was not correlated with outcome. 57,62 Recent genomic and expression profiling studies show that combinations of specific genetic features may have more prognostic significance than a single genetic alteration: Lagarde et al showed that the CINSARC complexity index and the Genomic Index are strong independent predictors of metastasis in SS. The finding that chromosome instability was strongly associated with the development of metastasis also helps explain that pediatric patients hardly develop metastasis during follow-up: comparison of the tumors of adults versus SS tumors in children showed that the genome of SS in adults was more frequently dearranged and non of the pediatric patients or adult

70 General discussion and future perspectives patients with an even genomic profile developed metastasis during follow-up in the series of Lagarde et al. 63 Another emerging area that may provide potential valuable biomarkers comprises the cell proteins. This resource is becoming efficiently and accurately accessible by recent technical advances allowing multiplexed protein analysis at the single cell level by antibody barcoding with photocleavable DNA (ABCD) platform. With this comprehensive method a large amount of proteomic information can be generated at the cell level and subsequently compared with morphology, clinical outcome and therapeutic effects. One of the major advantages of this method includes that only a limited amount of cells is needed, making this technique particularly suitable to fine needle aspiration samples (FNA). This approach may help gain more insight into pharmacological effects and therapeutic resistance. In a recent paper by Ullal et al the investigators performed a proof of concept study in both cell culture lines and human clinical samples of primary lung adenocarcinoma and human fibrosarcoma cells and demonstrated that the method is relatively simply and reliably applicable in both settings. They successfully demonstrated that more insight can be gained into intra- and intertumor heterogeneity. Also they demonstrated that therapeutic response can be linked with specific protein markers. 64 Although these experiments need to be replicated on a larger scale basis, these results are eminent. This approach forms an attractive scope to gain more insight into STS. Especially when combined with genetic features significant progress may be achieved into the understanding of the biology and the heterogeneous clinical behavior of STS subtypes in broad sense. A whole new resource of potentially clinically meaningful proteomic and genetic biomarkers or signatures is becoming accessible. These may help predict more precisely which subsets of patients will benefit from certain existing therapies, but also reveal tangible targets for the design of innovative therapies that will be more effective and less toxic. Improving the quality of prognostic research Although it is exciting to speculate about the discovery of novel innovative biomarkers as these embody the desirable major breakthroughs, it is at the same time essential to continue improving the quality of clinicopathological data collection and analysis. Larger series will be collected by the integration of different resources, including national and international collaborations. Preferably publications should include the rough data to allow systematic and accurate comparison between studies. The accuracy of data collection can be further improved and uniformed by systematic application of data reporting guidelines like the REMARK (Reporting Recommendations for Tumor Marker Prognostic Studies) guideline. 65 In the UMCG a standard pathologic reporting checklist is used for STS conform the recommendations of the ADASP (Association of Directors of Anatomic and Surgical Pathology). 66 All patients are registered at the IKNL (Integral Kankercentrum Nederland) database. 67 Conform our results and earlier statements we stress that accurate stratification based on accurate diagnosis and subtyping is essential, as these conditions form the foundation of good prognostic research. Grouping different sorts of STS together can sometimes be unavoidable, especially in trials that investigate specific treatment modalities or multimodal treatment including surgical, radiotherapeutic and for instance Isolated Limp Perfusion (ILP), but the results of these studies should be carefully interpreted with acknowledgment that the different STS subtypes should be considered different types of diseases. 9,13,14 Furthermore substantial efforts are put into the development of new comprehensive analytical approaches and innovative multivariate statistical methods. These are increasingly getting better in tracing independent correlations that may be hidden in the complex heterogeneous accumulation of variables in the relatively small cohorts of STS patients. 68 Prediction models A great challenge will remain in the careful integration of validated prognostic parameters into predictive schemes or statistic tools that help estimate prognostic outcome, like staging systems, algorithms and calculating models like for example nomograms. 54,69-71 Existing models that are used for STS usually only provide very rough estimations, they are often too complex and lack correct inclusion of critical STS(sub-)type specific prognostic variables. Most importantly we underscore that at least histopathological subtype should have a prominent place in any algorithm. We validated that for SS tumor stage, tumor size and tumor grade (according to the 3-tierced FNCLCC grading system) should be included. For localized MLS we validated that tumor grade should be included. In case of ES and DFSP mainly disease stage is of prognostic significance. We do want to emphasize that in daily practice algorithms should only be used with great cautiousness. The careful assessment and interpretation of the individual patient-tumor profile and subsequent translation to an optimally validated treatment schedule is a matter of precision medicine and in our opinion still at its best carried out by a multidisciplinary team of well skilled sarcoma specialists. Summary and conclusions In this thesis we documented outcome and treatment results of 4 stratified series of STS types treated at the University Medical Center Groningen (UMCG) and in case of ES at the Radboud University Medical Center, Nijmegen (Radboudumc) and the

71 General discussion and future perspectives UMCG and we analyzed diagnostic and prognostic aspects. In this concluding chapter we appraised the necessity of subtyping and stratification, we discussed the value and the role of molecular diagnosis and we explicated the complexity of prognostic research in STS. Mainly traditional clinicopathological prognostic factors were evaluated in our series and we particularly demonstrated the value of proper grading in SS and MLS. Our findings were discussed in perspective to more recent insights. Future directions in diagnostic and prognostic STS research include that gradually more clinically relevant markers and signatures will be identified at the molecular level, allowing that diagnosis and grading will be performed more accurately on small amounts of tissue material obtained by core or fine needle biopsies. Proper stratification and the discovery of such markers will help to better delineate subtype specific regimens of existing and newly developed treatment modalities. References 1. Borden EC, Baker LH, Bell RS et al. Soft tissue sarcomas of adults: state of the translational science. Clin Cancer Res. 2003;9: Eriksson M. Histology-driven chemotherapy of soft-tissue sarcoma. Ann Oncol. 2010;21 Suppl 7:vii ESMO/ European Sarcoma Network Working Group. Soft tissue and visceral sarcomas: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2012;23 Suppl 7:vii Blay JY, Sleijfer S, Schöffski et al. International expert opinion on patient-tailored management of soft tissue sarcomas. Eur J Cancer. 2014;50: Nederlandse Werkgroep Weke Delen Tumoren (NWWDT). Richtlijn Wekedelentumoren. 2011; Available at: 6. McArthur GA, Demetri GD, van Oosterom A et al. Molecular and clinical analysis of locally advanced dermatofibrosarcoma protuberans treated with imatinib: imatinib target exploration consortium study B225. J Clin Oncol. 2005;23: Rutkowski P, Van Glabbeke M, Rankin CJ et al. Imatinib mesylate in advanced dermatofibrosarcoma protuberans: pooled analysis of two phase II clinical trials. J Clin Oncol. 2010;28: Verweij J. Soft tissue sarcoma trials: one size no longer fits all. J Clin Oncol. 2009;27: Sleijfer S, Ouali M, van Glabbeke M, et al. Prognostic and predictive factors for outcome to first-line ifosfamide-containing chemotherapy for adult patients with advanced soft tissue sarcomas: an exploratory, retrospective analysis on large se- ries from the European Organization for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group (EORTC-STBSG). Eur J Cancer. 2010;46: Fletcher, CD, Bridge JA, Hogendoorn P et al. (eds). WHO Classification of Tumours of Soft Tissue and Bone. 4 th ed. Lyon: IARC; Brown FM, Fletcher CD. Problems in grading soft tissue sarcomas. Am J Clin Pathol Nov;114 Suppl:S Coindre JM. Grading of soft tissue sarcomas: review and update. Arch Pathol Lab Med Oct;130(10): Verweij J, Baker LH. Future treatment of soft tissue sarcomas will be driven by histological subtype and molecular aberrations. Eur J Cancer. 2010;46(5): Gonzalez-Angulo AM, Hennessy BT, Mills GB. Future of personalized medicine in oncology: a systems biology approach. J Clin Oncol. 2010;28(16): Meis-Kindblom JM, Bjerkehage B, Bohling T et al. Morphologic review of 1000 soft tissue sarcomas from the Scandinavian Sarcoma Group (SSG) Register The peer-review committee experience. Acta Orthop Scand Suppl. 1999;285: Ray-Coquard I, Montesco MC, Coindre JM et al. Sarcoma: concordance between initial diagnosis and centralized expert review in a population-based study within three European regions. Ann Oncol. 2012;23(9): Bovée JV, Hogendoorn PC. Molecular pathology of sarcomas: concepts and clinical implications. Virchows Arch. 2010;456: Coindre JM, Pelmus M, Hostein I et al. Should molecular testing be required for diagnosing synovial sarcoma? A prospective study of 204 cases. Cancer. 2003;98: Neuville A, Ranchère-Vince D, Dei Tos AP et al. Impact of molecular analysis on the final sarcoma diagnosis: a study on 763 cases collected during a European epidemiological study. Am J Surg Pathol. 2013;37: Engstrom K, Berg P, Cederlund CG et al. Irradiation of myxoid/ round cell liposarcoma induces volume reduction and lipoma-like morphology. Acta Oncologica. 2007;46: Guadagnolo BA, Zagars GK, Ballo MT et al. Excellent local control rates and distinctive patterns of failure in myxoid liposarcoma treated with conservation surgery and radiotherapy. Int J Radiat Oncol Biol Phys. 2008;70: Chung PW, Deheshi BM, Ferguson PC et al. Radiosensitivity translates into excellent local control in extremity myxoid liposarcoma: a comparison with other soft tissue sarcomas. Cancer. 2009;115: Moreau LC, Turcotte R, Ferguson P et al. Myxoid/round cell liposarcoma (MR- CLS) revisited: an analysis of 418 primarily managed cases. Ann Surg Oncol. 2012;19:

72 General discussion and future perspectives 24. Pitson G, Robinson P, Wilke D et al. Radiation response: an additional unique signature of myxoid liposarcoma. Int J Radat Oncol Biol Phys. 2004;60: 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: Nijhuis PH, Schaapveld M, Otter R et al. Soft tissue sarcoma compliance with guidelines. Cancer 2001;91: Penel N, Glabbeke MV, Mathoulin-Pelissier S et al. Performance status is the most powerful risk factor for early death among patients with advanced soft tissue sarcoma. The European Organisation for Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group (STBSG) and French Sarcoma Group (FSG) study. Br J Cancer. 2011;104: Gronchi A, Casali PG. Adjuvant therapy for high-risk soft tissue sarcoma in the adult. Curr Treat Options Oncol. 2013;14: Somaiah N, von Mehren M. New drugs and combinations for the treatment of soft-tissue sarcoma: a review. Cancer Manag Res. 2012;4: Antonescu CR, Tschernyavsky SJ, Decuseara R et al. Prognostic impact of P53 status, LS-CHOP fusion transcript structure, and histological grade in myxoid lipsarcoma: a molecular and clinicopathologic study of 82 cases. Clin Cancer Res. 2001;7: Smith TA, Easley KA, Goldblum JR. Myxoid/round cell liposarcoma of the extremities. A clinicopathologic study of 29 cases with particular attention to extend of round cell liposarcoma. Am J Surg Pathol. 1996;20: Khatri VP, Galante JM, Bold RJ et al. Dermatofibrosarcoma protuberans: reappraisal of wide local excision and impact of inadequate initial treatment. Ann Surg Oncol. 2003;10: Kimmel Z, Ratner D, Kim JY et al. Peripheral excision margins for dermatofibrosarcoma protuberans: a meta-analysis of spatial data. Ann Surg Oncol. 2007;14: Ballo MT, Zagars GK, Pisters P et al. The role of radiation therapy in the management of dermatofibrosarcoma protuberans. Int J Radiat Oncol Biol Phys. 1998;40: Suit H, Spiro I, Mankin HJ et al. Radiation in management of patients with dermatofibrosarcoma protuberans. J Clin Oncol. 1996;14: Dagan R, Morris CG, Zlotecki RA et al. Radiotherapy in the treatment of dermatofibrosarcoma protuberans. Am J Clin Oncol. 2005;28: Haas RL, Keus RB, Loftus BM et al. The role of radiotherapy in the local management of dermatofibrosarcoma protuberans. Soft Tissue Tumours Working Group. Eur J Cancer. 1997;33: Jebsen NL, Engellau J, Engström K et al. Patterns of local recurrence and dose fractionation of adjuvant radiation therapy in 462 patients with soft tissue sarcoma of extremity and trunk wall. Int J Radiat Oncol Biol Phys. 2013;86: Callister MD, Ballo MT, Pisters PW et al. Epithelioid sarcoma: Results of conservative surgery and radiotherapy. Int J Radat Oncol Biol Phys. 2001;51: Weitz J, Antonescu CR, Brennan MF. Localized extremity soft tissue sarcoma: improved knowledge with unchanged survival over time. J Clin Oncol. 2003;21: Cheng EY, Springfield DS, Mankin HJ. Frequent incidence of extrapulmonary sites of initial metastasis in patients with liposarcoma. Cancer. 1995;75: Pearlstone DB, Pisters PW, Bold RJ, et al. Patterns of recurrence in extremity liposarcoma: implications for staging and follow-up. Cancer 1999; 85: Spillane AJ, Fisher C, Thomas JM. Myxoid liposarcoma the frequency and the natural history of nonpulmonary soft tissue metastases. Ann Surg Oncol. 1999;6: Estourgie SH, Nielsen GP, Ott MJ. Metastatic patterns of extremity myxoid liposarcoma and their outcome. J Surg Oncol 2002;80: Sheah K, Ouellette HA Torriani M et al. Metastatic myxoid liposarcomas: imaging and histopathologic findings. Skeletal Radiol. 2008;37: Pisters PW, Sondack VK. Metastatic patterns of extremity liposarcoma and their outcome. J Surg Oncol. 2002;80; De Vreeze RS, de Jong D, Nederlof P et al. Multifocal myxoid liposarcoma metastasis or second primary tumor?: a molecular biological analysis. J Mol Diagn. 2010;12: De Vreeze RS, De Jong D, Tielen IHG et al. Primary retroperitoneal myxoid/round cell liposarcoma is a nonexisting disease: an immunohistochemical and molecular biological analyis. Mod Pathol. 2009;22: Blazer DG 3rd, Sabel MS, Sondak VK. Is there a role for sentinel lymph node biopsy in the management of sarcoma? Surg Oncol. 2003;12: Maduekwe UN, Hornicek FJ, Springfield et al. Role of sentinel lymph node biopsy in the staging of synovial, epithelioid, and clear cell sarcomas. Ann Surg Onvol. 2009;16: Andreou D, Tunn PU. Sentinel node biopsy in soft tissue sarcoma. Recent Results Cancer Res. 2009;179: Daugard S, Mooi WJ, Harris. Is histological grading of value for the prognosis of soft tissue sarcomas? Eur J Cancer. 1997;33: Coindre JM, Terrier P, Guillou L et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of

73 General discussion and future perspectives 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer. 2001;91: Demicco EG, Lazar AJ. Clinicopathologic considerations: how can we fine tune our approach to sarcoma. Semin Oncol. 2011;38 Suppl 3:S Deyrup AT, Weiss SW. Grading of soft tissue sarcomas: the challenge of providing precise information in an imprecise world. Histopathology. 2006;48: Guillou L, Coindre JM, Bonichon F et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading systems in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol. 1997;15: Ladanyi M, Antonescu CR, Leung DH et al. Impact of SYT-SSX fusion type on the clinical behavior of synovial sarcoma: a multi-institutional retrospective study of 243 patients. Cancer Res. 2002;62: Sobin LH, Wittekind CH. International Union Against Cancer: TNM Classification of malignant tumours (ed 6). New York, NY, Wiley-Liss, Hunter KW. Ezrin, a key component in tumor metastasis. Trends Mol Med. 2004;10: Khanna C, Wan X, Bose S et al. The membrane-cytoskeleton linker ezrin is necessary for osteosarcoma metastasis. Nat Med. 2004;10: Yu Y, Khan J, Khanna C, Helman L et al. Expression profiling identifies the cytoskeletal organizer ezrin and the developmental homeoprotein Six-1 as key metastatic regulators. Nat Med. 2004;10: Guillou L, Benhattar J, Bonichon F et al. Histologic grade, but not SYT-SSX fusion type, is an important prognostic factor in patients with synovial sarcoma: a multicenter, retrospective analysis. J Clin Oncol 2004;22: Lagarde P, Przybyl J, Brulard C et al. Chromosome instability accounts for reverse metastatic outcomes of pediatric and adult synovial sarcomas. J Clin Oncol. 2013;31: Ullal AV, Peterson V, Agasti SS et al. Cancer Cell Profiling by Barcoding Allows Multiplexed Protein Analysis in Fine-Needle Aspirates. Sci Transl Med. 2014;6:219ra Altman DG, McShane LM, Sauerbrei W et al. Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK): explanation and elaboration. PLoS Med. 2012;9:e Recommendations for the reporting of soft tissue sarcoma. Association of directors of anatomic and surgical pathology. Virchows Arch. 1999;434: Forsberg JA, Healey JH, Brennan. A probabilistic analysis of completely excised high-grade soft tissue sarcomas of the extremity: an application of a Bayesian belief network. Ann Surg Oncol. 2012;19: Canter RJ, Qin LX, Maki RG et al. A synovial sarcoma-specific preoperative nomogram supports a survival benefit to ifosfamide-based chemotherapy and improves risk stratification for patients. Clin Cancer Res. 2008;14: Maki RG, Moraco N, Antonescu CR et al. Toward better soft tissue sarcoma staging: building on american joint committe on cancer staging systems versions 6 and 7. Ann Surg Oncol. 2013;20: Kattan MW, Leung DHY, Brennan MF. A postoperative nomogram for 12-year sarcoma-specific death. J Clin Oncol. 2002;20:

74

75 Dankwoord & curriculum vitae