Synovial sarcoma is a soft-tissue malignancy that can arise at any age and anatomic area but tends to favor the distal extremities in young adults.1–4 It can be confused with other mesenchymal tumors.5 Small (<5 cm) tumors have a good prognosis, whereas larger ones are at greater risk for metastases and local recurrences.6,7 Recent developments in basic and translational research have provided fresh insights into the pathogenesis and treatment of this disease.

Pathology and Biology

SS18-SSX Fusion Genes

The majority of synovial sarcomas carry the pathognomonic t(X;18) translocation, resulting in fusion of the SS18 (formerly SYT) gene on chromosome 18 with an SSX gene on chromosome X (SSX1, SSX2, or rarely SSX4) (Fig. 1).5,8 This genetic trigger is the only consistent cytogenetic abnormality in the disease, which thus far has been found to have a relatively stable genome and few additional mutations.10–12 Although SS18-SSX exhibits both transcriptional activating and repressing properties, it is not a transcription factor and does not bind DNA directly.13–16 SS18 tends to facilitate transcriptional activation, whereas SSX tends to affect transcriptional repression.17,18

Fig. 1
figure 1

Genetic translocations in synovial sarcoma. On the left, normal chromosomes X and 18 are shown, as well as wild-type SSX (SSX1, SSX2, and SSX4) and SS18 genes, respectively. Characteristic synovial sarcoma fusions (SS18-SSX1, SS18-SSX2, and SS18-SSX4) are pictured on the right

Full size image

SS18-SSX1 is the most frequent translocation followed next by SS18-SSX2 and finally SS18-SSX4.16 SS18-SSX1 and SS18-SSX2 behave somewhat differently. It has been suggested that SS18-SSX1 acts more as a “proliferation oncogene,” conferring proliferative, migratory, and invasive advantages to cells, whereas SS18-SSX2 functions more as a “position oncogene,” with effects on architectural, adhesive, and cytoskeletal properties, but this concept is perhaps best regarded as a preliminary one at present.19

Cellular Origin and Differentiation

Synovial sarcoma is a misnomer, because it does not arise from synovium, and the cells do not express synovial markers.7 Instead, it exhibits mesenchymal and epithelial differentiation.4 No cellular origin has yet been proven, but current research suggests that it develops from primitive mesenchymal cells or myoblasts.10,20–22 Naka et al. found that silencing of SS18-SSX activated mesenchymal lineage genes in synovial sarcoma cells.23 Garcia et al. found that expression of SS18-SSX2 in myoblasts inhibited further myogenic differentiation.24 Early-stage MYF5-positive myoblasts expressing SS18-SSX2 develop sarcomas with 100% penetrance in mice.22 This study also showed that SS18-SSX2 expression in more mature muscle cells caused only myopathies.22

Epigenetic Modifications

SS18-SSX alters chromatin remodeling via epigenetic alterations through SWI/SNF- and histone deacetylase (HDAC)-associated mechanisms, thereby activating or abrogating DNA interactions with transcription factors.12,25–28 It competes with wild-type SS18 for assembly within SWI/SNF complexes, ejecting SNF5, a tumor suppressor.29

Oncogenic Pathways

Several cellular pathways seem to be important in synovial sarcoma.12 The expression of genes pertaining to the Notch and Hedgehog pathways is notably altered in some tumors.30–33 Genetic anomalies in the Wnt network also have been described.34–39 TLE1, a mediator of the Wnt pathway, is upregulated in certain cases.1,12,40,41 TLE1 may also affect HDAC activity and mediate gene silencing in this disease.42,43

Some tumors exhibit PTEN and/or PIK3CA mutations, which can lead to upregulation of the Akt-mTOR pathway.12,36,44–48 Akt stimulation can also occur via the activity of receptor tyrosine kinases (RTKs), including EGFR, IGF-1R, VEGFR, and PDGFR.49–55 SS18-SSX stimulates IGF-2 expression in tumor cell lines and thereby activates IGF-1R, Akt, and SRC.55,56 SS18-SSX also can induce IGF-2 in fibroblast cells.57 VEGF serum levels are increased in some patients.58 Several studies have shown overexpression of FGF and FGFR in tumor cells.59,60 Garcia et al. showed that SS18-SSX upregulates FGFR2, which was critical for proliferation.24 Ephrin RTKs may promote metastasis in synovial sarcoma. SS18-SSX2-positive tumors activate this pathway through upregulation of EphB2 receptor and ephrin B1 ligand, leading to cytoskeletal modifications and loss of cellular adhesion.19

Diagnosis

The pathologic diagnosis of synovial sarcoma remains a challenge, because there is histological overlap with other tumor types. Ideally, recognition of the disease should be based upon a combination of findings, including traditional morphology, identification of the chromosomal t(X;18) translocation, and a panel of immunohistochemical markers.

Histological Classification

Three distinct subtypes are recognized: (1) monophasic, which contains predominantly spindle cells; (2) biphasic, which contains spindle and epithelial-like cells, with areas recapitulating gland formation; and (3) poorly differentiated, which can be characterized by necrosis, bizarre mitoses, high cellularity, and nuclear atypia, but perhaps more often is seen as a proliferation of small, round cells (Fig. 2).13,61 Separation into subtypes is based upon somewhat subjective criteria, and there is a certain degree of overlap.

Fig. 2
figure 2

Pathology and immunohistochemistry of synovial sarcoma. High-power H&E views of: a monophasic; b biphasic, and c small cell, poorly differentiated synovial sarcoma. Immunohistochemistry characteristically reveals: d patchy cytoplasmic cytokeratin expression, and e strong nuclear reactivity for TLE1

Full size image

Early studies suggested a relationship between tumor histology and fusion type.62 SS18-SSX1 tumors tended to be biphasic and had a higher proliferative Ki67 index, whereas SS18-SSX2 tumors were more likely to be monophasic with a lower Ki67 index.2,10,11,15,63–65 In an attempt to explain the presence of both epithelial and mesenchymal components, Saito et al. hypothesized that the cell of origin may have innate potential for undergoing epithelial differentiation, but loses the trait in certain cellular contexts and acquires mesenchymal features.62

Cytogenetic Diagnosis

Monophasic and poorly differentiated subtypes can sometimes be difficult to distinguish from other tumors. The translocation t(X;18) has been identified in synovial sarcoma only, and its sensitivity and specificity have been both established.5,8 In certain difficult cases, the detection of this fusion event (by RT-PCR or cytogenetic studies) has been instrumental to the correct diagnosis.1,42

Synovial sarcoma cannot be entirely excluded from the differential diagnosis if tumors have the morphological and clinical features of synovial sarcoma but do not bear an SS18-SSX fusion gene. In rare instances (estimated to be <5% of all cases), synovial sarcomas do not carry the characteristic SS18-SSX transcripts. These tumors may arise from alternative gene fusions (such as SS18L1/SSX1) or cryptic rearrangements.9,10,66 Apart from these exceptional cases, t(X;18) analysis remains an important tool for diagnosing synovial sarcoma.67

Immunohistochemical Markers

The diagnostic value of various markers has been limited by their lack of sensitivity and/or specificity. Immunomarkers with some utility include SMARCB1/INI1, cytokeratins, epithelial membrane antigen (EMA), carcinoembryonic antigen (CEA), vimentin, calponin, TLE1, Bcl2, CD34, CD99, and S100 protein (Fig. 2).1,42,68–75 Cytokeratins and EMA show a characteristic patchy pattern in the spindle cell component and a more uniform staining in the epithelial component. Strong, diffuse nuclear TLE1 reactivity may be a helpful finding in certain cases. Whereas Bcl2 and CD99 are usually reactive in synovial sarcoma, they are seen in many other tumors as well, thus limiting their specificity. Lai et al. found that NY-ESO-1, a cancer testis antigen, in synovial sarcoma, was highly expressed in 76% of tumors, which was greater than other spindle cell tumors.76 In contrast, however, Endo et al. showed that NY-ESO-1 is expressed in 49% of synovial sarcoma and that its distribution is not unique to this disease, because it is seen in myxoid liposarcomas, myxofibrosarcoma, and chondrosarcoma as well.77

Prognostic Factors

Many clinical factors have been studied to determine their potential prognostic value. Age has been reported to predict survival; young patients generally fare better than older ones.2,6,7,13,21,61,63,78–85 However, some authors found no effect of age.86–89 Most studies report no effect of gender on outcome with the exception of one study that found worse survival for males.13,81–83,87,88

Several tumor-related variables may be important. Tumor size has consistently predicted local recurrence and survival in many studies.2,6,13,63,84,90 Tumor location also seems to be significant with central locations having worse prognosis than extremities.13,82,86,90,91 Undifferentiated tumors, high histologic grade, high mitotic rate, and necrosis have been associated with worse outcome.2,6,7,13,84,90 Similarly, bone and neurovascular invasion are poor prognosticators.82,83,89 Local recurrence has been associated with greater risk of metastasis and shorter survival.13,79,81,87,90

Recent data suggest that fusion type does not have prognostic value, despite early studies suggesting that SS18-SSX1 produces more aggressive disease than SS18-SSX2.2,13,63,64,86 Similarly, biphasic histology is of questionable predictive value. In one study, monophasic tumors were to be indolent than biphasic tumors.82 However, in a different study, when FNCLCC grade was taken into account, biphasic histology was not an independent factor for outcome.13

Gene Expression Profiles

The various genes and pathways that exhibit perturbations in synovial sarcoma include Wnt (LEF1, TCF7, ZIC2, WNT5A, and FZD10), Hedgehog (PTCH1), NY-ESO-1 (CTAG1A), and Notch (JAG1, JAG2, and HES1), as well as RTKs (FGF2, FGF3, EGFR, PDGFR, and IGFBP3). Because these pathways are not consistently affected in all cases, efforts have been made to identify a genetic signature that predicts survival or tumor progression.6,12,14,54,60,92–95 The Complexity Index in Sarcoma (CINSARC) and the Genomic Index signatures are 67- and 97-gene panels, respectively, which have been found to have predictive value for metastasis in synovial sarcoma.95–97

Therapeutic Options

Therapeutic approaches vary according to stage and prognostic factors. The principles of surgical management are similar to those that apply to soft-tissue sarcomas in general. Patients with nonmetastatic, T1 (<5 cm), superficial tumors in favorable extremity sites may be treated with wide surgical excision alone. In one recent study of T1 tumors, it was found that microscopic disease was present in 43% of reexcised tumor beds, and reexcision of unplanned resections was strongly recommended.98 Larger tumors in deeper, more unfavorable locales may require radiation and surgery. For more advanced disease, multimodal treatment that entails surgery, radiotherapy, and systemic chemotherapy may be indicated.32,35 In this disease, however, the efficacy of current chemotherapy is less than optimal, and newer systemic therapies need to be developed (Fig. 3).

Fig. 3
figure 3

Novel therapeutic strategies in synovial sarcoma. Blunt arrows (⊥) indicate inhibition, sharp arrows (→) indicate stimulation, and straight lines (–) indicate interference. A Naïve T cells are cultured with NY-ESO-1 to expand T-cell populations and enhance their activity. Genetically engineered anti-NY-ESO-1 T cells elicit an immune response when reintroduced into the patient. B HDAC inhibitors downregulate HDAC, being responsible for the shift of DNA from transcriptionally active to inactive status. SS18-SSX interferes with SWI/SNF complex, leading to SNF5 ejection and degradation, mediating tumorigenesis. C Trabectedin inhibits synovial sarcoma by blocking transcription factors as well as depleting tumor macrophages. D Pazopanib represses many RTK pathways simultaneously or acts through a yet uncharacterized network in synovial sarcoma. E Dacarbazine alkylates DNA strands and inhibits cellular division. F Gemcitabine blocks DNA synthesis and docetaxel interferes with tubulin to repress mitosis. G By intercalation, doxorubicin inhibits DNA transcription and synthesis. Ifosfamide is an alkylating agent that blocks DNA transcription. H Injected SS18-SSX-loaded dendritic cells present SS18-SSX peptides to T cells to induce T-cell-mediated immunity

Full size image

Doxorubicin (60–75 mg/m2) and ifosfamide (7.5–9 g/m2) comprise front-line therapy for synovial sarcoma.99–103 Together, the agents produce better outcomes in advanced disease than other chemotherapy regimens.7,82,103–107 Ferrari et al. reported 5-year, metastasis-free survival rates of 60 and 40% for patients treated with and without chemotherapy, respectively.82 Edmonson et al. showed partial tumor regression in 5 of 12 patients with residual, recurrent, or metastatic tumors, with a median overall survival of 11 months.103 Ifosfamide-based chemotherapy increased the 4-year, disease-specific survival rate from 67 to 88% in 101 patients with primary high-risk disease.106

High-dose ifosfamide (14 g/m2) alone has been used, but relatively few studies have been published to quantify its efficacy. Lee et al. reported that 2 of 7 patients experienced a complete response, and 4 of 11 patients had a partial response or disease stabilization.108 Median progression-free survival was 2.9 months, and median overall survival was 8.7 months.

An alternative treatment is the combination of gemcitabine and docetaxel, which may be considered in patients who cannot tolerate or are resistant to standard chemotherapy. Gemcitabine is an S-phase-specific nucleoside analog that blocks DNA synthesis. Docetaxel is a tubulin stabilizer and mitotic inhibitor. Early studies suggested that gemcitabine, despite its effectiveness in soft-tissue sarcomas, might not have much activity in synovial sarcoma.109,110 Similarly, in an early randomized study, patients receiving docetaxel exhibited no discernible responses.111 However a subsequent phase 2 trial showed objective but incomplete responses in 4 patients with synovial sarcoma treated with docetaxel.112 A recent, randomized, phase 2 trial compared gemcitabine plus docetaxel to gemcitabine alone in 122 patients with advanced soft-tissue sarcoma.113 For the combination treatment, the median progression-free and overall survivals were 6.2 and 17.9 months, respectively, whereas single agent gemcitabine resulted in 3 and 11.5 months survival, respectively. Two of the four synovial sarcoma patients treated with the combination of gemcitabine plus docetaxel exhibited stable disease. Similarly, two of four treated with gemcitabine alone had stable disease.

Multiple RTK networks are active in synovial sarcoma. Their co-inhibition may lead to synergistic antitumor effects.114,115 Clinical trials have been performed to analyze the benefit of RTK inhibitors (Table 1). Recent phase 2 and 3 studies suggest that pazopanib has activity in metastatic and refractory synovial sarcoma.116,117 In one phase 2 trial, the 3-month progression-free survival rate was 49% (18/37 patients), partial responses were noted in five patients, and the median overall survival duration was 310 days.116

Table 1 Investigational trials exploring multi-targeted receptor tyrosine kinase agents in the treatment of recurrent or metastatic synovial sarcoma
Full size table

Other trials have been designed to inhibit specific targets in synovial sarcoma. NCT00356031 is an ongoing phase 2 trial of a VEGFA antibody plus radiation on large (>5 cm) primary or recurrent synovial sarcoma. In a previous phase 2 study (NCT00831844), the IGF-1R antibody cixutumumab was found to have no benefit in 11 patients with recurrent refractory synovial sarcoma.118 An ongoing, phase 1 study (NCT00720174) examines the combination of cixutumumab and doxorubicin in advanced disease. It may be worth noting that in other soft-tissue sarcomas in which the IGF/IGF-1R axis is active, a shift from IGF-1R toward insulin receptor (IR) can occur. Hence, future efforts may need to target both IGF-1R and IR simultaneously.

Trabectedin is a promising agent, and partial regression of bilateral lung metastases was seen in one patient with advanced synovial sarcoma.119 In a recent retrospective study of 61 patients with advanced synovial sarcoma treated with trabectedin, 9 (15%) experienced partial responses and 19 (31%) had complete responses.120 The median progression-free survival was 3 months for the whole group and 7 months in the responding cohort. The mechanism of action is still being elucidated; it may affect transcription factors and tumor microenvironment through neoplastic macrophage depletion.121,122

HDAC- and SWI/SNF-mediated epigenetic modulation are potential therapeutic targets in synovial sarcoma. Radiotherapy induces DNA double strand breaks, stimulating DNA repair mechanisms, particularly those involving HDAC.123,124 In preclinical studies, HDAC inhibitors induced differentiation, apoptosis, and growth arrest of synovial sarcoma cells while increasing tumor cell sensitivity to radiation and chemotherapy.43,125,126 A phase 2 trial (NCT00112463) to study the efficacy of an HDAC inhibitor in synovial sarcoma has recently closed to accrual, and results of the trial are pending.

Wnt signaling inhibition via TCF/β-catenin complex inhibitors induces apoptosis and inhibits synovial sarcoma cell proliferation both in vitro and in vivo.34,127 A phase 1 clinical trial (NCT01469975) is being undertaken to evaluate the monoclonal antibody FZD10, which interrupts the Wnt pathway at the receptor-ligand level.

There have been some progress toward immunotherapeutic strategies. Particular mention should be made of NY-ESO-1, which is expressed in a substantial percentage of synovial sarcoma. Treatment of patients with NY-ESO-1-positive tumors with genetically engineered lymphocytes seems promising.76,128 Another immunotherapeutic approach involves SS18-SSX vaccine development, which may be employed with or without chemotherapeutic agents.6,90,129,130 Clinical studies exploring these therapeutic avenues are summarized in Table 2.

Table 2 (A) Ongoing phase 1 clinical trials exploring the efficacy of genetically engineered T lymphocytes targeting NY-ESO-1-positive synovial sarcomas; (B) Phase 1 clinical trials investigating immunotherapeutic vaccines in synovial sarcoma
Full size table

Conclusions

Synovial sarcoma is a malignant disease that frequently manifests in the extremities of young adults but can occur anywhere and in any age group. It has a tendency toward late recurrences and metastases, particularly in large tumors. Current multimodal treatment includes radiation, surgical resection, and chemotherapy. Multiple signaling networks and pathways have been identified in the disease. Understanding the molecular mechanisms of these phenomena may potentially lead to the development of newer and more effective therapies for patients with advanced and relapsed disease.