Synovial Sarcoma of the Nerve—Clinical and Pathological Features: Case Series and Systematic Review

Abstract

BACKGROUND

Synovial sarcoma of the nerve is a rare entity with several cases and case series reported in the literature. Despite an improved understanding of the biology, the clinical course is difficult to predict.

OBJECTIVE

To compile a series of patients with synovial sarcoma of the peripheral nerve (SSPN) and assess clinical and pathological factors and their contribution to survival and recurrence.

METHODS

Cases from 2 institutions collected in patients undergoing surgical intervention for SSPN. Systematic review including PubMed and Scopus databases were searched for related articles published from 1970 to December 2018. Eligibility criteria: (1) case reports or case series reporting on SSPN, (2) clinical course and/or pathological features of the tumor reported, and (3) articles published in English.

RESULTS

From patients treated at our institutions (13) the average follow-up period was 3.2 yr. Tumor recurrence was seen in 4 cases and death in 3. Systematic review of the literature yielded 44 additional cases with an average follow-up period of 3.6 yr. From pooled data, there were 10 recurrences and 7 deaths (20% and 14%, respectively). Adjuvant treatment used in 62.5% of cases. Immunohistochemical markers used in diagnosis varied widely; the most common are the following: Epithelial membrane antigen (EMA), cytokeratin, vimentin, cluster of differentiation (CD34), and transducin-like enhancer of split 1 (TLE1). Statistical analysis illustrated tumor size and use of chemotherapy to be negative predictors of survival. No other factors, clinically or from pathologist review, were correlated with recurrence or survival.

CONCLUSION

By combining cases from our institution with historical data and performing statistical analysis we show correlation between tumor size and death.

ABBREVIATIONS

CD34

cluster of differentiation

CI

confidence interval

EMA

epithelial membrane antigen

GTR

gross total resection

HPF

high-powered field

MPNST

malignant peripheral nerve sheath tumor

STS

soft tissue sarcomas

SSPN

synovial sarcoma of the peripheral nerve

TLE1

transducin-like enhancer of split 1

The first published presentation of synovial sarcoma dates back to 1910 when Lejars and Rubens-Duval¹ described their findings as a “primitive sarcoma of synovial joints,” which illustrated glandular and spindle cell components. Knox² in 1936 suggested the name synovial sarcoma because of the characteristic pattern of differentiation resembling that of the synovial membrane of joints and bursae. This early description is now known as a biphasic synovial sarcoma, one of 2 main types of synovial sarcomas with the other being monophasic. It is made of 2 cellular components: a spindle cell layer and an epithelial component.³ Synovial sarcomas began to be recognized as a separate entity from normal synovia because of the lack of consistent synovial features.⁴ Miettinen and Virtanen⁵ showed that synovial sarcomas and normal synovia differed in patterns of intermediate filament, epithelial membrane antigen (EMA), and lectin binding site expression.

Imaging, clinical features, and histology, however, have many similarities with other malignant primary peripheral nerve tumors, particularly malignant peripheral nerve sheath tumors (MPNST).⁶ The higher prevalence of these primary nerve sheath tumors may have contributed to a relative neglect of the diagnosis of synovial sarcoma. Further work illustrated a variety of immunohistochemical features all with limitations given their overlap with other soft tissue tumors, such as MPNST, fibrosarcoma, inflammatory myofibroblastic tumor, low-grade myofibrosarcoma, leiomyosarcoma, spindle cell rhabdomyosarcoma, and endothelial neoplasms.⁶ In 1986 and 1987, Turc-Carel et al^7,8 were the first to show the translocation association of the SS18 (synovial sarcoma translocation) gene on chromosome 18 to homologous genes on the X chromosome in synovial sarcomas excluding any connection with normal synovia. This is now the gold standard for differentiating and diagnosing synovial sarcoma.

Most of the literature surrounding synovial sarcoma of the peripheral nerve (SSPN) focuses on the pathological characteristics of this tumor and differentiating these tumors from MPNST can be quite difficult.^6,9-16 Once the diagnosis is made treatment can vary widely and no published consensus exists. Aggressive surgery is typically undertaken, with or without margins. Amputation is utilized in some cases but the benefit is unknown. Similarly, adjuvant therapy is usually employed without any particular standard. We aim to review published cases in the literature and combine them with our own case series to better understand the clinical behavior of SSPN with respect to different treatment modalities. Specifically, we asked the question: in patients with SSPN do any clinical or pathology-based features contribute to patient survival or tumor recurrence, with inclusion of research articles, case reports, and review articles.

METHODS

Case Selection

The case series was approved by the institutional review boards at both institutions, with the consent process waived because of the retrospective nature of the study. Case archives of the 2 senior authors (R.J.S. and A.D.L.) were reviewed and all patients with the diagnosis of “synovial sarcoma” arising from peripheral nerve tissue were identified and included from 2006 to 2018. When possible tumors were confirmed via the presence of the characteristic gene fusion; t(X; 18). Histologic and immunohistochemical characteristics of each were reviewed that will be detailed below.

Histology and Immunohistochemistry

Comment on grade was made in some cases, and in other cases, the number of mitoses per high-powered field (HPF) was expressed. Tumor markers included EMA, SMA, S-100 protein, cluster of differentiation (CD34), Bcl-2, cytokeratin (CK), TLE-1, neurofilament protein, and others.

Molecular Genetics

The typical translocation between chromosomes 18 and X was sought using either reverse transcriptase polymerase chain reaction (RT-PCR) or fluorescence in situ hybridization (FISH). These methods were able to identify the gene fusion product SS18-SX1 or SS18-SX2.

Literature Review: Search Strategy and Information Sources

This systematic review was conducted in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines (Figure 1).¹⁷ Database searches of MEDLINE (PubMed) and Scopus from 1970 to December 2018 were performed. Key word search terms were as follows: (“synovial sarcoma”) AND (“nerve” OR “peripheral nerve” OR “plexus” OR “neuronal”). Studies were limited to those published in the English language. Studies were included if a case of primary synovial sarcoma of the nerve was presented with pathological and clinical details. Both case series and case reports were included.

FIGURE 1.

PRISMA flow diagram. Two databases were thoroughly reviewed (MEDLINE and Scopus) using MeSH terms, as described in Methods, to best capture articles discussing synovial sarcomas arising from peripheral nerves. Case reports and case series were captured, as this is a rare tumor, and no large-scale studies have been carried out.

Statistical Analysis

Data analysis was performed using custom software within MATLAB 2018a (MathWorks, Natick, Massachusetts). Univariate logistic regression was performed to identify clinical and histological factors associated with disease progression or death with a P value < .05 indicating statistical significance. Factors associated with outcome variables on univariate regression with P ≤ .05 were included as candidate variables in multivariate logistic regression models. Regression results are described by the corresponding odds ratio (OR), 95% CI of the OR, and the corresponding P value. Kaplan-Meier (KM) analysis was also performed on time to disease recurrence and time to death. KM results are summarized with a corresponding hazard ratio (HR), 95% confidence for the HR, and P value from the log-rank test to assess for differences between groups with a P value < .5 indicating statistical significance.

RESULTS

Clinical Features in Local Series

A total of 13 patients were included in this case series; clinical data and interventions preformed are represented in Table 1. The average patient age at presentation was 47.5 yr with a 7:6 F: M ratio. Patients presented with 3 primary symptoms: neuropathic and local pain (6 and 2), motor deficit (5), and a palpable mass (3). Note that 8 patients, 61.5%, were neurologically intact at the time of presentation. A deep vein thrombosis was the presenting feature in one case. Nerves involved included the following: brachial plexus (extensive: 2, lower trunk: 1, proximal divisions: 1, medial cord: 1), tibial (2), distal radial (1), median (2), deep peroneal (1), sciatic (1), femoral (1). Such that 8 tumors involved the upper and 5 involved the lower limbs. Tumor size ranged from 1.3 to 28 cm in maximal dimension; 2 cases with extensive involvement of the brachial plexus could not be specifically measured. These features are presented in Table 1.

TABLE 1.

Synovial Sarcoma of the Nerves (Institutional Data)

Case	Age; sex	Nerve involved	Clinical features (presentation; neurological deficit)	Size (max dimension; cm)	Resection	Follow-up time (months)	Recurrence (months from surgery)	Adjuvant therapy (preop chemo-; radiation; postop chemo-; radiation) with initial surgery	Clinical course
1	53; M	Brachial plexus	Neuropathic pain and weakness; motor deficit	Diffuse	GTR with amputation	17	Yes (8)	Y; N; N; N	Death
2	38; M	Tibial	Neuropathic pain; intact	6.2	STR	N/A	N/A	N; N; N; N	N/A
3	32; M	Radial	Palpable mass; intact	2	GTR with margins	17	No	N; Y; N; N	NED
4	58; F	Median	Neuropathic pain; intact	1.3	STR	82	No	N; N; N; N	NED
5	54; M	Peroneal (deep)	Palpable mass; motor deficit	7.3	Biopsy	N/A	N/A	N/A	N/A
6	58; M	Brachial plexus (lower trunk)	Neuropathic pain; motor deficit	Diffuse	GTR with amputation	29	Yes (17)	N; N; N; N	Death; chemo and radiation given postop after recurrence
7	45; M	Sciatic	Local pain/cramping; intact	28	GTR	14	Yes (4)	Y; Y; N; N	Death
8	15; F	Median	Palpable mass; intact	1.4	GTR	129	No	Y; Y; N; N	NED
9	73; F	Tibial	N/A	2	GTR with amputation	63	No	N; N; N; N	NED
10	52; F	Brachial plexus (proximal divisions)	Neuropathic pain, palpable mass; intact	Diffuse	STR	43	Yes (84)	N; N; N; Y	NED; postop chemo given at recurrence
11	30; F	Femoral	Local pain; motor deficit	4.4	GTR	12	No	N; N; N; Y	NED
12	42; F	Brachial plexus	Deep vein thrombosis; intact	6.8	STR	12	No	Y; N; Y; Y	NED
13	68; F	Brachial plexus (medial cord)	Weakness; motor deficit	6.7	GTR with margins	6	No	Y; N; Y; N	NED

Clinical features of the 13 cases captured between 2 home institutions are displayed.

Ten patients had preoperative biopsy. Surgery included subtotal resection (4), gross total resection (5), gross total resection with negative microscopic margins (2), and gross total resection with amputation (3). One patient received biopsy only. Recurrence occurred in 4 patients, with 2 undergoing reresection. The average time from initial treatment to recurrence was 28.3 mo. Death from tumor progression occurred in 3 patients at an average of 20 mo from the time of surgery. Two patients did not have adequate follow-up, and the remaining 8 patients are still alive at most recent follow-up; the average follow-up among these patients is 38.5 mo (3.2 yr). One illustrative case is presented in Figure 2.

FIGURE 2.

Illustrative case. A and B, Preoperative MR images prior to A and after chemotherapy B. C and D, Intraoperative photographs. Note arrows in each image highlighting the lesion. This corresponds to case number 13 from institutional series; a 68-yr-old female presented with severe hand weakness and was noted to have a 6.7 cm tumor associated with the medial cord of the brachial plexus. A gross total resection was achieved with negative margins. The subclavian artery was resected along with the tumor and reconstructed with a saphenous vein graft. This SS stained positive for TLE1, p53, and D2-40. Negative stains included pancytokeratin and CD34. It housed the typical genetic translocation; t(x;18). The patient underwent both preoperative and postoperative chemotherapy. Interestingly, this tumor occurred near a prior radiation field for the treatment of breast cancer 20 yr prior. The patient had a rather dramatic response to preoperative chemotherapy (doxorubicin and ifosfamide), see panel A vs B, with approximately an 80% reduction in size of the tumor. The patient remains tumor-free at latest follow-up (6 mo).

Histology and Immunohistochemistry

Of the 13 patients, a histologic grade was given in 10 cases. A total of 7 cases noted to be high grade (3/3) and 3 noted to be intermediate grade (2/3). All 3 deaths occurred in the high-grade cases. Monophasic morphology was noted in 7 cases with biphasic morphology noted in 2 cases, others were not discussed. Immunohistochemical analysis yielded widely variable results. Typical assessment included EMA, CK, keratin markers, transducin-like enhancer of split 1 (TLE1), S100, CD34, and p53, among others (Table 2).

TABLE 2.

Pathological Characteristics of Synovial Sarcomas (Institutional Data)

Case	Appearance	Grade; mitoses per HPF	t(X;18)	SYT-SSX1/2	EMA	S100	CK	CD31/CD34/CD57/CD99	Desmin/MSA	Bcl-2	TLE1	KRT	VIM	Cam5.2	Collagen IV	AE1/AE3	Other
1		Grade 3	+
2		Grade 3			−	−	+	−CD34									Pos: p53, p16
3	Mono	Grade 2				+						+
4	Mono				+	−		−CD34				+
5		Grade 3	+		+	−			−/−	+		−
6	Mono	Grade 3	+			+	+	−CD34			+
7	Mono	Grade 3			+	−	+PanCK	−CD34	−/				+				Neg: HER2/Neu expression
8		Grade 3			+		+PanCK	−CD31−CD34
9	Mono	Grade 3	+	2
10	Mono		+		+	+		−CD34+ CD99	−/	+		+		+	+	+	Neg: GFAP, caldesmon
11	Bi	Grade 2; 8 m/10 HPF
12	Mono	Grade 2; 13 m/10 HPF	+		+	+		−CD34	+/		+	+					Pos: SMA Neg: myogenin, myoD1
13	Bi		+	2		−	−PanCK	−CD34			+						Pos: p53, D2-40 Neg: SMA, myogenin, GFAP, SOX-10

Detailed and comprehensive histology, immunohistochemistry, and genetic analysis are presented. HPF: high powered field; t(x;18): translocation between chromosomes X and 18; SYT-SSX1/2: gene fusion product produced from characteristic translocation in SS; EMA: epithelial membrane antigen; CK: cytokeratin; MSA: muscle specific actin; KRT: keratin; VIM: vimentin; NF: neurofilament; ER: estrogen receptor; PR: progesterone receptor; GFAP: glial fibrillary acidic protein.

Molecular Pathology

RT-PCR or FISH was performed in 7 cases to confirm typical gene translocation. SS18-SSX2 was detected in 2 cases, whereas the others expressed SS18-SSX1.

Adjuvant Treatment

Combinations of chemotherapy and radiation were used as follows: preoperative adjuvant treatment alone in 4 cases (radiation 1, chemotherapy 1, and radiation/chemotherapy 2). Postoperative adjuvant treatment alone used in 2 cases (radiation 2). Combination of preoperative and postoperative adjuvant treatment used in 2 cases (preoperative chemotherapy with postoperative radiation/chemotherapy 1 and preoperative chemotherapy with postoperative chemotherapy 1).

Review of the Literature

Literature review yielded 27 published articles in which 44 cases were identified.^9-16,18-35 These are detailed in Tables 3 and 4. Of these, 24 are case reports and 3 case series (with 2, 5, and 15 patients, respectively). Combined with the current 13 cases these total 57 cases of SSPN.

TABLE 3.

Clinical Features of Published Cases

Prior case	Age; sex	Nerve involved	Clinical features (presentation; neurological deficit)	Size (max dimension: cm)	Resection	Follow-up time (months)	Recurrence	Adjuvant therapy (preop chemo-; radiation; postop chemo-; radiation) with initial surgery	Clinical course
1; Cugola and Pisa18	16; F	Radial	Muscle swelling, extensor weakness; motor weakness	2	GTR	48	None	N; N; N; N	NED
2; Rinehart et al19	23; F	Median	Radiating pain; intact	2.5	GTR with margins	8	None	N; N; N; N	NED
3; O’Connell et al9	16; F	Common digital	Palpable mass; intact	2.5	GTR	N/A	N/A	N/A	N/A
4; Tacconi et al10	44; F	Brachial plexus	Palpable mass with radiating pain; diminished sensation in C5-6 dermatomes with motor weakness	2	GTR	12	None	N; N; Y; Y	NED
5; Spielmann et al11	43; M	Tibial	Palpable mass, radiating pain; intact	3.5	GTR with amputation	8	None	N/A	NED
6; Chesser et al20	16; M	Median	Palpable mass; intact	2	STR	12	None	N; N; N; Y	NED
7; Vang et al21	16, F	Medial plantar nerve	Palpable mass; intact	Initial size N/A; on repeat resection size: 15 cm	STR	4	Local recurrence at 4 mo and metastatic lung nodules	N; N; N; N	BKA performed at 4 mo after STR
8; Zenmyo et al22	58, F	S1 nerve root	Radiating pain; intact	5.5	N/A	6	None	N; N; N; N	New lung adenoca; death at 6 mo
9; Lestou et al23	54; M	Peroneal	Weakness; motor deficit	N/A	N/A	N/A	N/A	N/A	N/A
10; Chu et al12	46; F	Facial	Palpable mass;	0.8	GTR	60	None	N; N; N; Y	Initial surgery with STR followed shortly by GTRwith margins; NED
11; Chu et al/de Rubaupierre et al12,24	11; F	C7 nerve root	Pain and weakness; motor deficit	0.4	GTR	48	Local recurrence at 36 mo	N; N; N; Y	Local recurrence at 3 yr
12; Weinreb et al13	48; F	Tibial	Radiating pain; intact	8.7	STR	14	Local recurrence at 3 mo	N; N; N; N	NED after GTR and amputation
13; Uehara et al14	33; F	Median	Pain, palpable mass; motor deficit	1.5	GTR	14	Local recurrence at 14 mo	N; N; N; Y	Local recurrence at 14 mo - unknown after this time
14; Ghiya et al35	10; M	C8 nerve root	Palpable mass; neurological deficit	5.5	STR	6	None	N; N; N; Y	NED
15; Naphade et al34	14; M	C7 nerve root	Radiating pain; neurological deficit	3.1	GTR	6	None	N; N; N; N	NED
16; Tosi et al33	59; F	Median	Palpable mass; intact	3.7	GTR	N/A	N/A	N; N; N; N	N/A
17; Lipira et al15	66; F	Median	Radiating pain and hand weakness; motor deficit	2.7	GTR with margins	6	None	N; N; N; Y	Initial surgery with STR followed shortly by GTRwith margins; NED
18; Pirouzmand et al29	53; F	Brachial plexus (upper trunk)	Radiating pain; intact	4.5	STR	72	None	N; N; N; N	Stable residual
19; Peia et al16	7; F	L4 nerve root	Radiating pain; motor deficit	2.5	GTR with margins	60	None	Y; N; Y; Y	Initial surgery with STR followed shortly by GTR with margins; NED
20; Kim et al30	26; M	Digital nerve	Palpable mass; intact	1.5	GTR with margins	18	None	N; N; N; Y	NED
21; Bhat et al25	39; F	Median	N/A	N/A	N/A	N/A	N/A	N/A	N/A
22; Chrisinger et al26: 1	46, M	Median	Neuropathic symptoms; N/A	4	STR	32	None	N; N; N; Y	NED
23; Chrisinger et al26: 2	50; F	Peroneal	Neuropathic symptoms; N/A	7.4	GTR	31	None	N; N; Y; Y	Death at 31 mo; developed pulmonary mets at 7 mo
24; Chrisinger et al26: 3	38; M	L4	Neuropathic symptoms; N/A	6.4	STR	12	None	N; N; Y; Y	Death at 12 mo due to pulmonary mets
25; Chrisinger et al26: 4	44; F	Sciatic	No neuropathic symptoms; N/A	3.5	GTR	38	None	N; Y; N; N	NED
26; Chrisinger et al26: 5	37; F	Ulnar	Neuropathic symptoms; N/A	3	GTR	131	None	N; N; N; Y	NED
27; Chrisinger et al26: 6	58; F	Ulnar	Neuropathic symptoms; N/A	9.5	STR	9	None	N; N; Y; Y	Local tumor present and being observed
28; Chrisinger et al26: 7	62; F	Ulnar	Neuropathic symptoms; N/A	3.5	STR	129	Recurrence at 94 mo & 127 mo	N; N; N; Y	NED
29; Chrisinger et al26: 8	25; M	Peroneal	Neuropathic symptoms; N/A	2	STR	94	Recurrence at 49 mo	N; N; N; N	NED
30; Chrisinger et al26: 9	58; F	Median	Neuropathic symptoms; N/A	13	GTR with amputation	14	None	Y; N; N; N	Death at 14 mo; developed pulmonary mets at 10 mo
31; Chrisinger et al26: 10	19; F	Ulnar	Neuropathic symptoms; N/A	2.5	STR	10	None	N; N; N; N	NED
32; Chrisinger et al26: 11	33; F	Median	Pain; N/A	N/A	N/A	N/A	N/A	N/A	N/A
33; Chrisinger et al26: 12	19; F	Peroneal	Neuropathic symptoms; N/A	4.5	GTR	139	None	N; N; Y; Y	NED; STR followed by GTR
34; Chrisinger et al26: 13	39; F	Ulnar	Neuropathic symptoms; N/A	2.5	GTR	173	None	N; N; N; N	NED
35; Chrisinger et al26: 14	34; M	Sciatic	N/A; N/A	7	GTR	16	None	N; N; N; N	NED
36; Chrisinger et al26: 15	44; F	T1 nerve root	Neuropathic symptoms; N/A	3.6	STR	2	None	N; N; Y; N	Local tumor present and being observed
37; Larque et al27: 1	31; F	Saphenous	Palpable mass; intact	4.5	GTR	237	None	N; N; N; Y	NED
38; Larque et al27: 2	31; F	Tibial	Pain; intact	2.4	GTR with margins	32	None	N; N; N; Y	NED
39; Larque et al27: 3	50; M	Tibial	Palpable mass, swelling; intact	8.1	GTR with amputation	44	None	N; N; Y; N	NED
40; Larque et al27: 4	39; F	Median	Palpable mass; intact	N/A	GTR with margins	58	None	N; N; N; Y	NED
41; Larque et al27: 5	28; M	Median	Palpable mass; motor weakness	6.6	GTR	33	None	N; N; N; Y	NED
42; Hashimoto et al28	11; F	Tibial	Pain and palpable mass; intact	4.7	GTR with margins	40	None	Y; N; Y; N	NED
43; Mohaidat et al32	48; F	Ulnar	Palpable mass; intact	4.2	GTR with margins	9	None	N; N; N; N	NED
44; Raveendran et al31	24; M	C8 nerve root	Palpable mass; intact	13.6	GTR	6	None	N; N; N; N	NED

Data taken from the systematic review of the literature and all available clinical details are presented. A total of 37 cases are discussed.

TABLE 4.

Pathological Characteristics of Synovial Sarcomas (Published Data)

Case	Appearance	Grade; mitoses per HPF	t(X;18)	SYT-SSX1/2	EMA	S100	CK	CD31/CD34/CD57/CD99	Desmin/MSA	Bcl-2	TLE1	KRT	VIM	Cam5.2	Collagen IV	AE1/AE3	Other
1	Bi
2	Mono	Grade 3			+	+	+		+/			+	+
3	Bi				+		+
4	Mono				+	+	+						+
5	Mono		+		+	−	−	+CD99	−/−				+				Neg: NF
6	Bi	10 m/HPF											+	+
7	Mono	6 m/HPF	+		−	−		−CD34 +CD57	−/−	+		−HMB45	+		+
8	Mono		+		+					+		+	+
9	Mono	Grade 3	+	2	+	+	+CK7	+CD99					+
10	Bi		+		+	−		−CD99				+K7 +K19	+
11	Mono		+		−	−		+CD99				−K7−K19	+
12	Bi		+				+CK7 −CK20			−			+	+		+	Neg: TTF-1, ER, PR, ChG
13	Mono		+	1	+		+CK7 +PanCK										Ki-67 2.9%
14	Bi				+	+	+CK7 +CK19	+CD99
15	Mono	Grade 3	+		+	−		−CD34 + CD99		+							Ki-67 40%
16	Bi	2 m/10 HPF	+	1	+	−	+CK7										Neg: ER, TTF-1
17	Mono		+		+	−	+CK7 +CK18 +CK19	−CD31−CD34	−/−	+			+			+	Neg: GFAP, p53,
18	Mono	5 m/HPF	+		+	−	−PanCK	−CD34	−							−	Neg: GFAP, myosin
19	Bi	Grade 2	−		+		+						+
20			+		+	+				+
21	Bi
22	Mono	4 m/HPF	+		+		+				+
23	Mono	40 m/HPF	+
24	Mono	1 m/HPF	+		+		+				+
25	Mono	3 m/HPF	+		+		+				+
26	Mono	1 m/HPF	+				−
27	Mono	4 m/HPF			−		+
28	Mono	12 m/HPF	+		+		−				+
29	Mono	1 m/HPF	+		+		+				+
30	Mono	21 m/HPF	+		+						+
31	Mono	1 m/HPF	+
32	Mono	2 m/HPF			+		+				+
33	Mono	2 m/HPF	+		+		−
34	Mono	1 m/HPF			+		+
35	Bi	21 m/HPF	+		+		+
36	Mono	14 m/HPF	+		+		+				+
37	Mono	Grade 2; 3 m/HPF			+	−	−
38	Mono	Grade 2; 14 m/HPF	+		+	−	−
39	Bi	Grade 2; 8 m/HPF	+		+	−	+
40	Mono	Grade 2	+		+	−	+
41	Mono	Grade 2; 16 m/HPF	+		+	−	+
42	Mono		+		−	−		+CD99				−	+
43	Mono
44		Grade 2	+	1	+	+		+CD56 +CD57			+

Detailed and comprehensive histology, immunohistochemistry, and genetic analysis is presented. HPF: high powered field; t(x;18): translocation between chromosomes X and 18; SYT-SSX1/2: gene fusion product produced from characteristic translocation in SS; EMA: epithelial membrane antigen; CK: cytokeratin; MSA: muscle specific actin; KRT: keratin; VIM: vimentin; NF: neurofilament; ER: estrogen receptor; PR: progesterone receptor; GFAP: glial fibrillary acidic protein.

Combined Data: Clinical Presentation

Grouping all cases together the average age at presentation was 38.3 yr (7-73 yr) with a M: F ratio of 1:3. The 3 most common presentations (not mutually exclusive) included neuropathic-type (25) or focal pain (7), palpable mass (20), or motor deficit (noted in 14 cases of the 39 cases with description of neurological status). Size of tumor ranged from 0.4 to 28 cm, average size of 5.6 cm in maximum dimension. Three tumors had diffuse involvement of the brachial plexus and size was not measurable. Of these 36 were present in the upper extremity, 20 in the lower extremity, and one along the facial nerve. Diffuse brachial plexus invasion was present in 3 cases. Complete list of nerves involved can be found in Tables 1 and 3.

Combined Data: Pathology and Immunohistochemistry (IHC)

In combining all cases we are able to review the trends in pathology and IHC and the complete data are presented in Tables 2 and 4. Of the 51 cases in which morphology was described, there were 38 monophasic tumors and 13 biphasic tumors. Grading was explicitly listed in 20 cases, 9 of which were noted to be high grade (grade 3), vs 25 cases were mitosis per HPF was listed. Average mitosis/HPF was 7.7 (range 0.8-40). The 5 most commonly used IHC markers are as follows: EMA (±total studied—39/44 or 88.6%+), CK (any molecular weight; 26/34 or 76.5%+), S100 (10/28 or 35.7%+), vimentin (14/14 or 100%+), CD34 (0/13 or 0%), and TLE1 (11/11 or 100%+).

Combined Data: Treatment Strategy

Four general surgical strategies were seen including biopsy only (1), subtotal resection (15), gross total resection without specific consideration for tumor at margins (21), gross total resection with microscopically negative margins (10), gross total resection with amputation (6), and 4 patients without complete operative data. Adjuvant treatment was divided into 4 categories. Preoperative adjuvant treatment was used alone in 6 cases: 2 isolated chemotherapy, 2 isolated radiation, and 2 combined chemotherapy/radiation. Postoperative adjuvant therapy was used alone in 23 cases; 2 chemotherapy, 16 radiation, and 5 combined chemotherapy/radiation. Other combinations included 2 patients with both preoperative and postoperative chemotherapy and 2 patients with preoperative chemotherapy and postoperative chemotherapy/radiation. Intraoperative radiation therapy was utilized in one case. A total of 18 patients received no adjuvant treatment (35.3%), and 6 patients did not have these details listed. See Tables 1 and 3.

Combined Data: Survival and Outcome

There were 7 patients without follow-up data, and thus, a total of 50 patients were studied. Patients were followed for an average of 42.3 mo (range 2-237). Tumor recurrence was seen in 10 cases (20%) at an average of 31.3 mo (range 4-94) from the time of initial treatment. Death was observed in 7 cases (14%) at an average of 17.6 mo (range 6-31) from initial treatment. Metastatic disease was noted in 3 cases and all of these involved lung metastasis. One patient died 6 mo after initial treatment for SSPN from lung adenocarcinoma and other cases from progression of their tumors.

Factors Predicting Survival and Recurrence

Univariate logistic regression on disease progression and death was performed using all variables, both clinical and pathological. With regards to survival, tumor size (OR 1.15 per unit cm increase, 95% CI 1.02-1.31, P = .03), use of preoperative adjuvant treatment (OR 6.29, 95% CI 1.05-37.58, P = .04), use of preoperative chemotherapy (OR 9.20, 95% CI 1.45-58.36, P = .02), and use of any chemotherapy (OR 20.5, 95% CI 2.15-195.59, P = .01) were found to increase the odds of death. However, only tumor size (OR 1.17, 95% CI 1.001-1.38, P = .04) remained statistically significant on the multivariate regression. Patient age, gender, and number mitoses per high power field were correlated (P < .1) with mortality but failed to reach statistical significance.

No factors were significantly correlated with tumor recurrence in univariate regression, and thus, a multivariate regression was not performed. KM curves are presented in Figures 3 and 4. Given the average tumor size of 5.4 cm in the pooled data, KM analysis shows results dichotomized using a size cutoff of 5.5 cm. Age demonstrated a trend toward increased risk of death when dichotomized at 50 yr (HR 4.43, 95% CI 0.79-25, P = .06). Male gender conferred a statistically significant increased risk of death on KM analysis (HR 5.37, 95% CI 0.89-32.5, P = .03). If groups were dichotomized above and below 15 mitoses per HPF, there was a trend toward significance with patients who had >15 mitoses per HPF having increased risk of death (P = .05). Gross total vs subtotal resection correlated with an improvement in disease-free survival on KM analysis but this did not reach statistical significance (P = .07). In terms of overall survival, patients undergoing gross total resection (GTR) with amputation tended to have worse outcomes but this did not reach statistical significance (P = .06).

FIGURE 3.

Survival related to clinical features. Kaplan–Meier curves are displayed relating patient age (greater or less than 50 yr), tumor size (greater or less than 5.5 cm), male vs female sex, and type of resection. The P values and total number of patients are displayed along with the plotted data.

FIGURE 4.

Recurrence related to clinical features. Kaplan–Meier curves relating the following patient factors to tumor recurrence: age (greater or less than 50 yr), tumor size (greater or less than 5.5 cm), male vs female sex, and type of resection.

When time to death was dichotomized according to tumor size, there were no deaths observed in the group with tumors ≤5.5 cm compared to 7 deaths in patients with tumors >5.5 cm (log-rank test, P < .001). Similarly, smaller tumor (<5.5 cm) had a significantly lower risk for tumor recurrence (HR 6.32, 95% CI 1.09-36.7, P = .001).

DISCUSSION

Clinical Behavior of Synovial Sarcoma

SSPN can occur in nearly any site where nerves are present. Patients tended to present with painful symptoms but presentation varied to include a palpable mass, usually painful, and isolated motor deficit, especially with cases of brachial plexus involvement. As discussed the diagnosis of synovial sarcoma via traditional histology and immunohistochemistry is not straightforward when involving the nerve and in the case of monophasic morphology can be very similar to MPNST.²⁶

Considering the similarities between SSPN and MPNST, it is helpful to examine the behavior of MPNST and current treatment paradigms. One important distinction in clinical features between MPNST and SSPN is the association the former has with NF1, which is not noted in SSPN.^36-38 Besides this point, the clinical presentation between MPNST and SSPN are similar,^39,40 although MPNST has a higher incidence of motor deficits on presentation reflecting its locally aggressive nature. Most authors advocate for complete, gross total resection with wide negative margins^37,41-43 in the surgical treatment of MPNST; with evidence pointing to the fact that remaining tumor correlates with prognosis.^36,44,45 Interestingly, this may not be true of low-grade MPNST as data suggest surgical margins do not play a role in prognosis.⁴⁶ The story of adjuvant radiation is not completely elucidated. A retrospective analysis of 33 patients was not able to determine any beneficial effect of radiation therapy for MPNST.⁴⁷ In a prospective, randomized trial including all types of soft tissue sarcomas (STS) adjuvant radiation therapy improved rates of local control but did not improve overall survival.⁴⁸ With regard to chemotherapy, we must again consider STS in general to see adequate numbers for statistical analysis. In a large pooled study of 12 clinical trials for chemotherapy in STS, authors determined that MPNST responded similarly to STS with varying chemotherapy regimens in cases of advanced disease.⁴⁹ In another retrospective of STS, authors were not able to associate chemotherapy with increased survival.⁵⁰ Overall, MPNST is a more aggressive disease with local recurrence rates between 40% and 65% and metastasis seen in 30%-60% of patients.^36,39,51-54 In total, 5-yr survival for MPNST is about 20% to 50% with an overall mortality rate estimated up to 75%.^37,55,56 When MPNST metastasizes it tends to affect the lungs but lymph node involvement is not seen; this is in contrast to SSPN, where lymph nodes can be involved.^26,37 We determined a local recurrence rate of 20% in the average follow-up period of 3.5 yr.

As alluded to both MPNST and SSPN fall in the category of STS and are, thus, grouped as such for many large clinical studies. The standard treatment of STS involves surgery with supramarginal resection when possible, and usually with radiation therapy to control local recurrence.⁵⁷ As evidenced in our series and review, the treatment regimen is highly variable based on institution. Chemotherapy may play a role in synovial sarcoma as suggested by Ferrari and colleagues,⁵⁸ whereas it is typically not used in other STS owing to lack of efficacy. This was especially true for those patients with larger (>5 cm) tumors, which could not be resected completely. In the presented series, preoperative chemotherapy correlated with a worse overall survival, but this is likely attributed to the fact that clinicians tend to reserve chemotherapy for patients with more advanced disease or histologically aggressive tumors. In all synovial sarcoma patients, with nonmetastatic, surgically resected disease the 5-yr overall survival is about 70%.⁵⁸ Patients with surgical resection and negative margins, in this study, had a local recurrence rate of about 30% at 5 yr.⁵⁸ In the cases in which negative margins were not sought, but macroscopic tumor was removed completely the local recurrence rate rose to ∼53% at 5 yr. We did analyze the relationship between tumor recurrence and survival as they related to extent of resection. As noted, those patients who received gross total resection tended to have improved progression-free survival. Those patients with supramarginal resection tended to behave similarly to those who underwent GTR (see Figures 3 and 4). Interestingly, patients who underwent amputation with GTR had worse progression-free and overall survival.

Microscopic Considerations of SSPN

In accordance with much of the literature we attempted to cover histologic and immunohistochemical features of these tumors. Despite the aforementioned advances in molecular biology, SS still has certain diagnostic features on histology that can aid pathologists. Biphasic synovial sarcoma appears very similar to normal synovial tissue because of its typical 2 component structure. The spindle cell layer is uniform in appearance consisting of relatively small, short spindle, or oval pale staining nuclei appearing to overlap because of the small amount of cytoplasm and indistinct cell membranes. The epithelial component consists of glandular or adenopapillary structures made of epithelial cells with moderate amounts of cytoplasm and distinct cell membranes.³ Monophasic synovial sarcomas contain primarily the spindle cells found within fascicles in association with collagen and branching vascular pattern.³ This is the most common form of synovial sarcoma. In the combined series presented we note 74.5% to have monophasic morphology. Authors have proposed a third classification that has been linked to poor outcome named poorly differentiated type synovial sarcoma. This subtype is characterized by small, poorly differentiated round cells resembling Ewing sarcoma, which expresses Fis1. These poorly differentiated areas are characterized by necrosis and increased mitotic activity as visualized by Ki67.⁵⁹ With regards to grade, we noted high-grade tumors in 45% of cases and an average mitotic index of 7.7 mitoses per HPF. Neither grade nor mitotic index demonstrated any correlation with survival or recurrence. The 3 histologic subtypes are illustrated in Figure 5.

FIGURE 5.

Histologic variants of synovial sarcoma. The 3 morphologies of synovial sarcoma are illustrated in panels A-C. The monophasic variant A, is most commonly seen and is composed of densely cellular intersecting fascicles of uniform spindle cells. Biphasic synovial sarcoma B, contains aggregates of glands lined by epithelial cells and surrounded by fascicles of spindle cells. Poorly differentiated synovial sarcoma C, can also been seen, composed of sheets of round cells. All images at 20× magnification with H&E staining.

In contrast to most peripheral nerve neoplasms, synovial sarcomas are of mesenchymal origin. One of the most widely accepted markers for synovial sarcoma is transducin-like enhancer of split 1 or TLE1, a corepressor that binds to other transcription factors to suppress genetic transcription. In synovial sarcoma, TLE1 is a nuclear protein and it is thought to play a role in apoptotic suppression by inhibiting the actions of Bcl-2, which is important in the apoptotic pathway.⁶⁰ Originally, thought to be specific for synovial sarcoma, it has now been shown to be positive in nerve sheath tumors, solitary fibrous tumors, and mesothelioma, with up to 30% positivity in MPNST.^6,60 In the combined series presented, 11 cases tested for the presence of TLE1 and it was positive in all of these.

Vimentin is an intermediate filament found in epithelial or mesenchymal cells. It has been used for positive identification of nonmuscle cell sarcoma.⁶¹ Vimentin was tested in 14 cases in our series and present in all of these. EMA is a glycoprotein that normally acts as barrier to apical surface of epithelial cells and is usually absent in other STS.⁶² It was present in 39 of 44 cases tested and was the most commonly tested IHC marker. CK proteins have also been found to be expressed in synovial sarcoma, which are important for intermediate filaments dealing with mechanical stress. Common CK proteins tested included CK7, CK18, CK19, and CK20; CK of any kind was found in 26 of the 34 cases in which it was tested (Tables 2 and 4).⁶² S-100 has been found in cells with neural crest lineage; however, it varies in positivity in synovial sarcoma with some studies citing expression in 50% of cases.⁶ Many of the detailed pathology reports note S100 positivity in the nerve tissue present within the specimen, and its presence was noted in 10 out of 28 cases (35.7%) in which it was tested in our series. In our statistical analysis, no IHC marker showed any correlation with survival or recurrence. They clearly have a role in the diagnosis of this lesion, but their heterogeneity adds confounding information. Molecular biology was used to aid in the diagnosis of 39 cases using either RT-PCR or FISH.

Combined Cases and Statistical Analysis

Owing to the rarity of this tumor, we must rely on pooled historical and institutional data. Fortunately authors in past reports have included significant detail in both clinical and pathological characteristics of their cases, and this allowed for inclusion in the present study. As mentioned, we noted both size and use of preoperative chemotherapy to be associated with poorer prognosis. We used a size cutoff of 5.5 cm and generated a KM curve (Figures 3 and 4), as this size cutoff has been used in the literature. Increased size and decreased survival follows logical progression. The use of chemotherapy and decreased survival, on the other hand, we attribute to inherent bias in retrospective analysis.

CONCLUSION

The presented series of SSPN is the largest to date as it combines institutional data with published cases. We attempt to offer correlation between clinical and microscopic features with recurrence and survival. In accordance with literature on SS, we observed a significant negative correlation between size and survival. Similarly, we observed a trend toward reduced recurrence conveyed by GTR. In light of this pooled data, we recommend the following: aggressive adjuvant treatment for large tumors and gross total resection of all tumors often necessitating removal of parent nerve. This report illustrates the importance of the following: (1) establishing a prospective database of patients suffering this pathology, and (2) determining treatment guidelines to enhance uniformity and potentially target specific molecular markers.

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Dr Levi receives teaching honorarium from the American Association of Neurological Surgeons and grant support from the Department of Defense and the National Institutes of Health (NIH-NINDS).

Neurosurgery Speaks! Audio abstracts available for this article at www.neurosurgery-online.com.

COMMENT

This article represents an excellent review of the experience of two institutions supplemented by an extensive and rigorous review of the literature on synovial sarcomas involving peripheral nerve. This is a very challenging clinical disease to treat surgically as shown by their results. They found a strongest correlation between tumor size and both recurrence and survival. Patient age, gender, and mitoses per high field showed a weak correlation that was not statistically significant. An interesting observation was the finding of positive staining for TLE1, thought to play a role in the suppression of apoptosis, in all 11 cases tested. This finding raises the interesting possibility of modulating this molecule and/or pathway to improve clinical outcome.

Michel Kliot

Chicago, Illinois