Virchows Archiv

, Volume 459, Issue 6, pp 615–621

TLE1 is expressed in the majority of primary pleuropulmonary synovial sarcomas

Authors

    • Department of Anatomic Pathology, Instituto Nacional de Cancerología
  • Juan Pablo Flores-Gutiérrez
    • Department of Anatomic Pathology, Hospital UniversitarioUniversidad Autónoma de Nuevo Leon
  • Natalia Vilches-Cisneros
    • Department of Anatomic Pathology, Hospital UniversitarioUniversidad Autónoma de Nuevo Leon
  • Hugo Ricardo Domínguez-Malagón
    • Department of Anatomic Pathology, Instituto Nacional de Cancerología
Original Article

DOI: 10.1007/s00428-011-1160-4

Cite this article as:
Lino-Silva, L.S., Flores-Gutiérrez, J.P., Vilches-Cisneros, N. et al. Virchows Arch (2011) 459: 615. doi:10.1007/s00428-011-1160-4

Abstract

Pleuropulmonary synovial sarcoma (PPSS) is a rare entity, similar to synovial sarcoma of soft tissue (STSS). There are 120 published cases of PPSS, but no studies have explored the expression of TLE1. In soft tissues, it has been proven a useful marker, but in tumors of other sites, its expression has not been explored. The main objective was to study the expression and diagnostic sensitivity and specificity of TLE1 in a group of PPSS, of which the diagnosis was corroborated by fluorescence in situ hybridization confirming t(X;18) in a tissue microarray. Immunohistochemistry including TLE1, vimentin, CD99, CD56, bcl-2, AE1-AE3, EMA, CD34, CK7, CK19, calponin, and S-100 was performed on all PPSS and on 25 control cases (five carcinomas, ten mesotheliomas, and ten thoracic sarcomas). TLE1 was positive in 11 cases (73.3%); bcl-2 and vimentin in 100%; calponin and CD56 in 26.6%; CD99, CK AE1-AE3, CK19, CK7, and EMA in 80%; and S100 negative in all. The only biphasic PPSS was positive for epithelial markers only in the epithelial component. TLE1 was negative in all control cases. TLE1 is expressed in 73% of PPSS, a value inferior to that reported in STSS, but is highly specific for PPSS. TLE1 may therefore be of value in the differential diagnosis of PPSS, but should be used in a panel of antibodies.

Keywords

Synovial sarcomaPleuropulmonary sarcomasTLE1

Introduction

Synovial sarcoma (SS) is a high-grade soft tissue tumor that occurs mainly in para-articular locations of the extremities, affecting young adults. It has well-defined histological features with monophasic and biphasic types. The SS in the pleuropulmonary region (PPSS) has histological features similar to soft tissue SS and was first described by Zeren et al. [14], but it was Essary et al. [23] who considered it a different clinicopathological entity due to the specific translocation t(X; 18)(p11,q11) that causes the fusion of SYT and SSX genes which can be detected by gene fusion analysis or evidence of break-apart test fluorescence in situ hybridization (FISH) [1]. Initially, PPSS included tumors arising not only in the pleura and lung but also those arising in the chest wall, heart, and mediastinum. When only tumors arising in the pleura and lung are considered as genuine PPSS, there are less than 100 cases published to date. Clinical symptoms include cough, chest pain, and dyspnea. On chest radiographs, PPSS typically appears as a well-delimited mass with uniform opacity, situated in the pleura or in the lung and often accompanied by an ipsilateral pleural effusion [2]. CT scan shows a well-circumscribed, heterogeneously enhanced non-calcified lesion without bone involvement. Magnetic resonance imaging provides a better demonstration of the solid and multilocular fluid-filled internal components of the lesion, with peripheral rim enhancement after the intravenous administration of a gadolinium-based contrast material [3].

The main objective of this study was to report 15 new cases of genuine PPSS seen in our institution, all with a rearrangement at the SS18 gene (SYT), to describe its clinical characteristics and to determine its immunohistochemical profile, including the recently developed antibody TLE1. The other purpose was to explore the feasibility of the FISH test on a tissue microarray (TMA).

Material and methods

The cases analyzed were collected from the archives of the surgical pathology department of our institution. This study was performed in accordance with the protocols and rules of the ethics and research committee of our institution. A search over a period of 10 years (2000–2010) revealed a total of 37 cases of SS located in the lung and pleura. Of these, 17 represented metastatic SS and were excluded. Five additional cases were eliminated for lack of tissue for molecular analysis. In total, 15 cases were eligible for analysis (Table 1).
Table 1

Clinical characteristics of pleuropulmonary synovial sarcoma

Case no.

Age (years)

Sex

Tumor site

Size (cm)

Subtype

Follow-up/outcome

1

38

M

Pleura

10

Monophasic

12 m/AWD

2

37

F

Lung

8

Monophasic

31 m/AWD

3

59

F

Lung

31

Monophasic

18 m/DOD

4

28

F

Lung

10

Monophasic

6 m/DOD

5

73

M

Pleura

1.9

Monophasic

40 m/AWD

6

32

M

Pleura

3.2

Monophasic

37 m/DOD

7

41

M

Lung

3.5

Monophasic

13 m/AWD

8

40

F

Pleura

20

Monophasic

17 m/DOD

9

22

F

Lung

12

Biphasic

12 m/AWD

10

35

M

Pleura

5

Monophasic

46 m/AWD

11

27

M

Lung

11.4

Monophasic

12 m/AWD

12

37

F

Lung

18

Monophasic

22 m/DOD

13

20

F

Lung

4.5

Monophasic

39 m/DOD

14

35

F

Pleura

4

Monophasic

13 m/DOD

15

30

M

Lung

5.2

Monophasic

12 m/AWD

Survival expressed in months (m)

DOD dead of disease, AWD alive with disease, M male, F female

Chest radiographs, computed tomography, magnetic resonance imaging, or positron emission tomography/computed tomography studies were available for all 15 cases. Clinical information and follow-up data were obtained from patient records. All cases were biopsy specimens obtained by core needle or open incisional biopsy. Hematoxylin- and eosin-stained sections and paraffin-embedded tissue were available in all cases. Tumors were classified as monophasic or biphasic according to the World Health Organization criteria [3]. The monophasic type was composed of densely packed spindle cells arranged in short fascicles often with a hemangiopericytoid vascular pattern (Fig. 1a). Biphasic tumors contained spindle cell as well as epithelial components with malignant epithelial cells forming gland-like spaces and tubulo-papillary structures (Fig. 1b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00428-011-1160-4/MediaObjects/428_2011_1160_Fig1a_HTML.gifhttps://static-content.springer.com/image/art%3A10.1007%2Fs00428-011-1160-4/MediaObjects/428_2011_1160_Fig1b_HTML.gif
Fig. 1

a Monophasic PPSS showing a spindle cell neoplasm with hemangiopericytoma-like pattern and little nuclear atypia. b Biphasic PPSS with glandular epithelial structures adjacent to a malignant spindle cell component. c TLE1, diffuse and strong nuclear expression in pleuropulmonary synovial sarcoma. d Bcl-2 cytoplasmic expression in pleuropulmonary synovial sarcoma. e Tissue microarray and FISH test show at least tree fluorescent signals of chromosome X and 18 representing the X:18 translocation

Histological preparations stained with hematoxylin and eosin were evaluated separately by two pathologists (HRDM and LSLS), and from each case, a representative area of 5 × 5 mm from the tumor was selected to construct the TMA, allowing 15 cases in a single paraffin block (Fig. 1e). As the synovial SS is a homogeneous tumor with little variability between histological fields, we took only one fragment of tissue. For biphasic tumors, we verified that both components were included, through an initial evaluation with hematoxylin and eosin prior to immunohistochemistry and molecular studies (Fig. 1e). Additionally, we used 25 control cases for immunohistochemistry for TLE1, which were evaluated in full sections. To this end, we chose pleuropulmonary or mediastinal tumors that resemble SS, including five spindle cell carcinomas, ten spindle cell mesotheliomas, six malignant fibrous histiocytomas, three malignant peripheral nerve sheath tumors, and one leiomyosarcoma.

Immunohistochemistry was performed on sections obtained from the TMA and on whole sections using commercially available antibodies according to the manufacturer’s instructions, with recommended dilutions (Table 2). After deparaffinization, antigen retrieval was performed using an automated immunostainer (Ventana, Tucson, AZ) with mild Tris-based buffer (CC1, Ventana). Tissue sections were incubated with isotypic antibodies (immunoglobulins); this was followed by washing and detection with the i-View DAB kit (Ventana). For TLE1, nuclear staining was considered positive; cytoplasmic staining for all the other antibodies. Based on intensity, a three-point scale was chosen: mild, moderate, or strong. In addition, we considered the staining diffuse when >30% of neoplastic cells were positive; focal if a smaller percentage of cells stained. For each of the antibodies, we performed positive and negative controls on full histological sections using tissues recommended by the manufacturer.
Table 2

Summary of antibodies performed

Antibody

Laboratory

Clone

Dilution

Bcl-2

Dako, Carpintera CA

124

1:20

CD34

Dako, Carpintera CA

QBEnd 10

1:40

CKAE/1AE3

Dako, Carpintera CA

AE1/AE3

1:200

CK7

Dako, Carpintera CA

OV-TL 12/30

1:160

CK19

Dako, Carpintera CA

RCK108

1:50

CD99

Dako, Carpintera CA

12E7

1:80

Calponin

Dako, Carpintera CA

CALP

1:600

EMA

Dako, Carpintera CA

E29

1:100

S-100

Dako, Carpintera CA

Policolnal

1:800

TLE-1

Santa Cruz Biochemicals, Santa Cruz, CA

sc-9121

1:20

Vimentin

Dako, Carpintera CA

Vim3B4

1:400

All cases were studied by break-apart FISH on 3- to 4-μm unstained sections of the TMA using the LSI SYT dual-color break-apart probe, flanking the SS18 [SYT]18q11.2 region, according to previously described standard laboratory methods [5]. The signal was developed by 10-μL Locus Specific Identifier (Vysis, Inc., Downers Grove, IL). The FISH slides were read separately by two pathologists (JPFG and NVC) at ×1,000 magnification on an Olympus BX51 fluorescence microscope equipped with appropriate filters (Olympus, Tokyo, Japan). Fluorescent signals from a minimum of 100 intact non-overlapped nuclei were counted. FISH was interpreted positive when it showed at least one red and one green split nuclear signal.

Results

Clinical data were available for all patients. There was no gender predilection: seven cases (47%) were in males and eight (53%) in females. The median age was 37 years, (range, 20–73 years). The median tumor size was 10 cm (range, 1.9–31 cm). The typical clinical presentation was a pleural or pulmonary mass associated with thoracic pain and cough.

The median follow-up was 6–46 months. Five patients died of disease (33%) and the remaining ten are alive with progressive disease not amenable for surgical treatment. Nine of the 15 cases were localized in the lung (60%) and six in the pleura (40%). One case was a biphasic tumor and the other 14 monophasic spindle cell synovial sarcomas.

Immunohistochemical and FISH results are summarized in Table 3 and Fig. 1e. The FISH test was positive in all 15 cases (100%), showing splitting of the fluorescent signal consistent with a rearrangement involving SYT (Fig. 1e).
Table 3

FISH and immunohistochemical findings

Case No.

FISH (X:18)

TLE-1

Bcl2

VIM

EMA

AE1/AE3

CK7

CK19

Calponin

CD34

CD99

S-100

CD56

1

+

++

++

+

+

2

+

++

++

+

+

3

+

++

+

+

4

+

+

F

+

5

+

++

++

+

+

6

+

+

+

+

7

+

+

+

+

+

8

+

+

F

+

9

+

+

++

++

++

++

++

+

10

+

++

++

+

F

+

11

+

++

+

+

12

+

+

++

++

+

+

13

+

+

+

+

+

F

F

14

+

+

+

F

+

+

+

+

15

+

+

+

+

+

+

− Negative; + weak positive, diffuse; ++ strong positive, diffuse; f focal positive

TLE1 was positive in 11 cases (73%), with strong and diffuse staining in 50% and diffuse–weak in the remaining cases (Fig. 1c) on the TMA and the full sections. All control cases were negative. This result shows 73% sensitivity and 100% specificity of TLE1 staining for PPSS.

Staining for bcl-2 protein and vimentin was positive in 100% of cases (Fig. 1d), CD56 was positive in ten cases (67%), calponin in four cases (27%), and EMA, CK AE1/AE3, CK19, CK7, and CD99 in two cases each (12%) in TMA sections, but positive in whole sections of 12 cases (80%). CD34 was focally positive in one case (7%) and S-100 negative in all cases.

Discussion

The term synovial sarcoma was coined in 1936 because the morphological features of the tumor are similar to those of the early developmental stages of joints, namely, cleft-like spaces surrounded by subsynovial spindle cells [6]. Currently, SS is regarded as a mesenchymal tumor characterized by a spindle cell component and an inconstant and variable epithelial component. In the 2002 World Health Organization Classification of Tumors of Soft Tissues and Bones, SS is classified as a malignant soft tissue tumor of uncertain differentiation [7]. Unusual locations of SS have been reported, including in the head and neck region, thorax, and internal organs such as the lung, kidney, liver, or prostate [8].

SS is characterized cytogenetically by a t(X;18)(p11.2;q11.2) translocation that results in the fusion of the SYT gene on chromosome 18 with the SSX1 or SSX2 gene on chromosome X. Epithelial and spindle cells of the biphasic type both carry the translocation [9]. The SYT–SSX1 fusion is present in the majority of biphasic tumors, while SYT–SSX1 and SYT–SSX2 fusions are observed with equal frequency in tumors of the monophasic subtype. This distinctive chromosomal translocation is present in >90% of synovial sarcomas regardless of subtype, and it appears to be specific. However, there are reports of other tumors like fibrosarcoma and malignant fibrous histiocytoma that have shown the same translocation [10, 11]. It has been shown that this translocation correlates with TLE1 expression by immunohistochemistry [26]. It has also been demonstrated in mesothelioma [22], but to our knowledge, no studies have been performed by FISH for the t(X;18) in mesothelioma. Two rare fusion genes variants, SYT–SSX4 = t(X;18)(p11;q13) and SYTL1–SSX1 = t(X;20)(p11;q13) have been shown recently in a small proportion of cases [2731]. Regarding the molecular analysis of t(X;18) in PPSS, in two previous studies, t(X;18) was reported in 36 of 39 patients (92%) and in 40 of 40 cases [2, 21]. We found the SYT translocation in 15 of 15 tested cases (100%); five cases were excluded because they were core needle biopsies with scant material insufficient for analysis. In our study, the FISH test was performed in a tissue microarray and showed high specificity. It is therefore a practical and reliable method to analyze several cases together, resulting in a highly cost-effective method.

Most SS display at least focal immunoreactivity for epithelial markers. Reportedly, 90% of all synovial sarcomas are cytokeratin-positive, more pronounced in the epithelial component than in the spindle component [1214]. The more frequently expressed are CK7 (79%), CK19 (60%), and CK8/18 (45%); no other spindle cell sarcomas exhibit these cytokeratins consistently [15, 16]. Up to 30% show focal immunoreactivity for S-100 protein. CD99, the product of the MIC2 gene, shows membranous or cytoplasmic staining in 60–70% of SS. Bcl-2 protein is diffusely positive in virtually all SS, especially in the spindle cells, but is of limited diagnostic value since many other tumors express this antigen [17]. Calponin has also been frequently found in SS; it may be useful in recognizing poorly differentiated variants since other round cell tumors are negative for this antigen [4]. CD34 is almost always negative in synovial sarcoma. The immunohistochemical profile in our series of PPSS is similar to that reported in soft tissue SS with some differences: 100% of cases were positive for vimentin and Bcl-2. CD56 was positive in 67% and calponin in 27%. However, epithelial markers and CD99 were expressed in a minority of cases studied by TMA (12%), however in whole sections were positive in 80% of cases and CD34 focally positive in one case.

A new immunohistochemical marker is Transducin Like Enhancer of Split 1 (TLE1), a protein that interacts with TLE2, glycoprotein 130, UTY, HES6, RUNX3, SIX3, and RUNX1 proteins involved in cellular proliferation. It belongs to the Groucho/TLE/Grg family of co-repressors that affect many signaling pathways. Human TLE1 operates with some critical transcription factors that regulate the survival, hematopoiesis, and differentiation of normal cells [5]. It is overexpressed when t(X;18) occurs. It was shown that TLE1 in SS causes the overexpression of Bcl-2. Experimental knockdown of TLE1 in fibroblasts and synovial cells did not affect cell proliferation or enhance the cytotoxicity of chemotherapy, which suggests that TLE1 can be a good target for selectively treating synovial sarcoma without causing damage to the host tissue. However, further studies will be needed to determine a potential therapeutic role for TLE1 [32]. In the largest series of 259 cases of molecularly confirmed synovial sarcomas using a TMA, TLE1 was reported positive in 96%, but negative in others [1820, 24, 25]. However, in another large series using whole tissue sections, it was found positive in other sarcomas, especially malignant peripheral nerve tumors [21]. In a recent series of spindle cell mesotheliomas, Matsuyama et al. [22] described focal and weak positivity in 28 malignant mesotheliomas regardless of histologic subtype. We included in our control series ten spindle cell mesotheliomas; all were negative. We did not, however, include epithelioid or mixed subtypes.

In conclusion, the molecular, immunohistochemical, and histological features of PPSS are very similar to SS in soft tissues. They are diffusely positive for vimentin and bcl-2, focally for EMA and keratin, and negative for CD34 and S100. TLE1 expression is highly specific, but less frequent than reported in soft tissue SS. Detection of SYT rearrangement by FISH is a valuable diagnostic adjunct that can be effectively performed on TMA.

Conflict of interest statement

We declare that we have no conflict of interest.

Copyright information

© Springer-Verlag 2011