ORIGINAL ARTICLE
BCOR-CCNB3 (Ewing-like) Sarcoma A Clinicopathologic Analysis of 10 Cases, In Comparison With Conventional Ewing Sarcoma Florian Puls, MD, FRCPath,* Angela Niblett, FIBMS,* Gillian Marland, BSc,* Czar Louie L. Gaston, MD,w Hassan Douis, FRCR,z D. Chas Mangham, FRCPath,*y Vaiyapuri P. Sumathi, FRCPath,* and Lars-Gunnar Kindblom, MD, PhD*
Abstract: BCOR-CCNB3 fusion transcripts resulting from an X-chromosomal paracentric inversion were recently identified in a series of unclassifiable soft tissue and bone sarcomas with Ewing sarcoma–like morphology. The morphologic and clinical features of these sarcomas are, as yet, not well characterized. Here we describe the clinicopathologic features of 10 cases of BCOR-CCNB3 sarcoma and compare their clinical course with typical Ewing sarcoma. Nine of 10 patients were male, and all were 11 to 18 years of age. Seven tumors were located in the bone and 3 in the deep soft tissues. The histomorphologic spectrum was quite wide, with 7 tumors predominately showing small primitive cell morphology with angulated nuclei simulating so-called atypical Ewing sarcoma and 3 predominately showing spindle cell morphology. Recurrent and metastatic lesions showed increased cellularity and marked pleomorphism. Immunohistochemistry showed expression of CCNB3 (100%), bcl2 (90%), CD99 (60%), and CD117 (60%). Reverse transcription polymerase chain reaction for BCOR-CCNB3 fusion transcripts was positive in all 9 cases, which yielded sufficient extracted RNA. Five- and 10-year survival rates were 75% and 56%, respectively. BCOR-CCNB3 sarcomas located in axial skeleton and soft tissues showed a significantly shorter survival. The Ewing sarcoma overall survival was not statistically different, although there was a trend for longer survival of patients with BCOR-CCNB3 sarcomas in the extremities. In conclusion, this study provides a detailed description of the histologic From the Departments of *Musculoskeletal Pathology; wOncology; zMusculoskeletal Radiology, Royal Orthopaedic Hospital NHS Foundation Trust, Birmingham; and yDepartment of Histopathology, Robert Jones & Ages Hunt Orthopaedic Hospital, Oswestry, UK. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Florian Puls, MD, FRCPath, Department of Musculoskeletal Pathology, Royal Orthopaedic Hospital NHS Foundation Trust, Robert Aitken Institute of Clinical Research, University of Birmingham, Birmingham B15 2 TT, UK (e-mail: fl[email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.ajsp.com. Copyright r 2014 by Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
spectrum, immunohistochemical features, and clinical characteristic of BCOR-CCNB3 sarcoma justifying distinction from Ewing sarcoma with its typical EWS/FUS-ETS translocations. Ideally immunohistochemistry is used in combination with reverse transcription polymerase chain reaction for definitive diagnosis. Key Words: BCOR-CCNB3 sarcoma, Ewing sarcoma, translocation-associated sarcoma, small round cell tumor, RT-PCR, immunohistochemistry (Am J Surg Pathol 2014;38:1307–1318)
A
subset of soft tissue and bone tumors with primitive “small round blue cell” morphology defies exact classification even after extensive diagnostic workup with immunohistochemistry and molecular techniques, thus posing a recurrent diagnostic and therapeutic challenge to pathologists and clinicians. Recently, Pierron et al1 identified recurrent gene fusions of BCOR (encoding Bcl6 interacting corepressor) and CCNB3 (cyclin B3) in primitive and, so far, unclassifiable small cell sarcomas in adolescents and young adults. BCOR-CCNB3 mRNA originates from a paracentric inversion on the X-chromosome and splicing of the end of the BCOR coding sequence to the CCNB3 exon 5 splice acceptor site.1 The resultant fusion protein is composed of full-length BCOR, a transcriptional repressor, and the C-terminus of CCNB3, a cyclin normally expressed in leptotene and zygotene of meiosis.2 In vitro studies suggest that the fusion BCOR-CCNB3 protein is oncogenic and drives proliferation in this sarcoma.1 BCOR-CCNB3 sarcoma showed a distinct gene expression profile when compared with classic Ewing sarcoma.1 Although currently listed under Ewing sarcoma in the recent World Health Organization classification, it could be argued that, on the basis of the distinct gene expression profiles, BCOR-CCNB3 sarcoma should be viewed as a separate entity.3,4 The original report of BCOR-CCNB3 fusion sarcoma focusing on its genetic and pathophysiological aspects is the only series reported to date.1 In this study we aimed to further characterize the morphologic, clinical, and www.ajsp.com |
1307
Puls et al
Am J Surg Pathol
prognostic features as well as the relative incidence of this sarcoma in comparison with its purported closest mimic, classical Ewing sarcoma.
MATERIALS AND METHODS Patients Archival cases from 1998 to 2013 of soft tissue or bone tumors with monotonous or primitive small cell morphology without a specific line of differentiation or specific identifiable transcripts, as well as cases of Ewing sarcoma predating routine molecular confirmation as part of the diagnostic workup, were screened for expression of CCNB3 by immunohistochemistry. These included cases diagnosed as Ewing-like sarcoma, small cell malignant peripheral nerve sheath tumor, poorly differentiated synovial sarcoma, or extraskeletal myxoid chondrosarcoma without molecular confirmation. In total, 134 cases were screened for CCNB3 expression by immunohistochemistry. The typical Ewing sarcoma control cases with confirmed translocations had their initial diagnosis made or their tumor resected at the Royal Orthopaedic Hospital in Birmingham between 2005 and August 2013. Second opinion pathology referral cases or patients not receiving treatment and follow-up at the Royal Orthopaedic Hospital were excluded. Sample collection and data analysis were conducted according to local ethical guidelines.
Histology and Immunohistochemistry For immunohistochemistry, paraffin wax sections were cut at 2 mm, dried, and antigen retrieved in an epitope retrieval solution of pH8 (RE7116; Novocastra, Newcastle, UK) at 681C for 17 hours in a stirred water bath. For detection of nuclear CCNB3, the primary rabbit polyclonal anti-CCNB3 antibody (HPA000496; Sigma, Gillingham, UK) was used in 1:200 dilution. Normal testis was used as positive control on the same slide. Other antibodies, clones, dilutions, and sources were: CD99 (12E7, 1:25; Dako, Ely, UK), desmin (D33, 1:100; Dako), S100 (polyclonal, 1:1000; Novocastra), cytokeratins (AE1/AE3, 1:100; Dako), Ki-67 (MIB-1, 1:200; Dako), CD117 (polyclonal, 1:100; Dako), bcl2 (124, 1:50; Dako), b-catenin (17C2; Novocastra, 1:50), and bcl-6 (LN22; Novocastra, 1:100). The DakoCytomation EnVision Detection System peroxidase/DAB visualization kit (K5007; Dako) was used throughout except for bcl-6 immunohistochemistry, wherein the Ultraview Universal DAB Detection and Amplification Kits (Ventana Medical Systems, Tucson, AZ) were used for visualization. All kits were used according to the manufacturer’s instructions, and same slide positive controls were used throughout. Mitoses were counted in ten 40 objective fields (field area of 0.238 mm2).
Reverse Transcription Polymerase Chain Reaction and Sequence Analysis Reverse transcription polymerase chain reaction (RT-PCR) was performed on formalin-fixed, paraffinembedded tissue except for 3 cases in which fresh frozen
1308 | www.ajsp.com
Volume 38, Number 10, October 2014
tumor tissue was available. Total RNA was extracted from fresh frozen tumor tissue using the RNeasy MiniKit (Qiagen, Hilden, Germany) and from formalin-fixed, paraffin-embedded sections using acid phenol chloroform separation and isopropanol precipitation as previously described.5 Amplification of a 98 bp amplicon of b-actin mRNA (forward 50 -ATCACCATTGGCAATGAGCG-30 , reverse 50 -TTGAAGGTAGTTTCGTGGAT-30 ) was used to confirm the presence of intact and amplifiable RNA. For BCOR-CCNB3 transcript detection a forward primer located in BCOR exon 15, 50 -GGCTCCACCCCAGTGAT CT-30 (BCOR-P1), and a reverse primer in CCNB3 exon 5, 50 -GGGTGTTTTGGAGGTGGTGGAT-30 (CCNB3-P2), were chosen to give an expected 140 bp amplicon spanning the fusion site within the BCOR-CCNB3 transcript (GenBank: JN813375.1, accession number JN813375).1 RNA from a confirmed BCOR-CCNB3 case (kindly provided by G. Pierron, Institut Curie, Unite´ de Ge´ne´tique Somatique– Centre Hospitalier, Paris, France) was used as a positive control, and 21 cases of small round blue cell tumors with known fusion transcripts (see Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/ PAS/A199) were used as negative controls. All primers were purchased from Eurofins MWG Operon-Biotech (Ebersberg, Germany), and RT-PCR reactions were performed with the OneStep RT-PCR kit (Qiagen) for 40 cycles with annealing temperatures of 601C and 611C for bactin and BCOR-CCNB3 primer pairs, respectively. Reaction products were visualized using GelRed staining (Cambridge Bioscience Ltd, Cambridge, UK) after separation by 10% polyacrylamide gel electrophoresis. PCR products of all BCOR-CCNB3 cases were Sanger sequenced at the Functional Genomics Laboratory at The University of Birmingham using an ABI 3730 Genetic Sequencer (Applied Biosystems, Warrington, UK).
Statistical Analysis The statistical analysis of baseline demographics and clinical outcome is based on all the data available up to January 15, 2014. Overall survival time was calculated as the number of completed months between the date of diagnostic biopsy and whichever occurred first: the date of death or January 15, 2014 (BCOR-CCNB3 sarcoma cases)/September 1, 2013 (classic Ewing sarcoma cases). Survival rates were calculated using the Kaplan-Meier method. For survival curve comparison the log rank (Mantel-Cox) test was performed using GraphPad Prism v5.00 (GraphPad, San Diego).
RESULTS In total 10 cases of BCOR-CCNB3 sarcomas were identified by immunohistochemical screening for CCNB3 expression, of which 9 cases underwent successful RTPCR analysis.
Clinical Features Nine of 10 BCOR-CCNB3 sarcoma patients identified were male, and all were in their second decade of life (age range 11 to 18 y, median 15.5 y). Seven tumors had r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
arisen in the bone, of which 6 had soft tissue extension. Radiologically, the largest tumor diameter ranged from 50 to 160 mm (median 107.5 mm). Computed tomography performed at the time of diagnosis showed metastatic disease in 4 cases (lungs 3, bone 1). Five patients received neoadjuvant chemotherapy before curative resection (cases 1, 4, 6, 8, 10; Table 1), 3 patients with axially located sarcomas received chemotherapy in combination with radiotherapy as first-line treatment (cases 2, 3, 5; Table 1). Curative surgical resection of the primary sarcoma in the extremity was performed in 7 patients, none of whom experienced local recurrence. One locally recurrent pelvic sarcoma was resected 25 months after the patient had received chemotherapy and radiotherapy (case 5). One patient was treated with local surgical decompression for spinal canal involvement on disease progression (case 2, Table 1). Surgical resection of metastatic disease was performed in 2 patients. Postoperative radiotherapy to the primary site was given to 4 patients, and 2 patients received whole-lung radiotherapy. Clinical follow-up information was available in all 10 patients (5 to 189 mo, median 69.5 mo). Six patients were without clinical evidence of disease at 5, 13, 78, 132, 145, and 189 months of follow-up, and 1 patient was alive with lung metastases at 61 months. Three patients died of disease after 27, 28, and 110 months (Table 1, Figs. 5C, E, F).
Histopathologic Features of BCOR-CCNB3 Sarcomas Biopsies of primary sarcomas were available from all 10 patients. Resection specimens without previous chemotherapy for unaltered morphologic assessment were available from 2 patients, and 5 patients received chemotherapy before tumor resection. In all but 1 case, areas with high cellularity predominated. Seven tumors were predominately composed of small round to plump tumor cells with a small amount
BCOR-CCNB3 Sarcoma
of ill-defined pale to eosinophilic cytoplasm (Figs. 1C–E). Nuclei were angulated and hyperchromatic with finely dispersed to glassy chromatin, and nucleoli were not readily discernible in the majority of cases. There were numerous apoptotic bodies, and the mitotic rate ranged from 5 to 37/10 high-power fields. Small areas of tumor cell necrosis were seen in 9 of 10 biopsies. In 3 cases tumor cells predominately showed a short spindle cell morphology (Figs. 1F, 2A). In all tumors the cells were discohesive without formation of a distinct architectural pattern and lay within an edematous to myxoid matrix with small myxoid pools in 1 case (Fig. 2A). The matrix was richly vascularized with thin-walled, sometimes gaping, vessels. In 5 cases intravascular thrombi were seen (Fig. 1E). Resections of recurrent or metastatic tumors were available from 2 patients. When compared with the primary tumor, 1 recurrent sarcoma showed a higher degree of pleomorphism with large, highly atypical spindle cells within a fibrotic matrix simulating undifferentiated pleomorphic sarcoma (Fig. 2F, case 5). Subsequent resection of a pulmonary metastasis in this patient showed a similar pleomorphic morphology (Fig. 2G). Several resections of pulmonary metastases from another patient showed increased cellularity, elongated spindle cells focally simulating synovial sarcoma, and scattered foci of osteoblastic differentiation (Figs. 2C, D, case 9). Despite the markedly different morphology compared with the primary tumor, these metastases were confirmed as originating from the known primary by nuclear expression of CCNB3 immunohistochemically and the detection of a BCORCCNB3 fusion transcript by RT-PCR. Resection shortly after completion of first-line chemotherapy was performed in 6 patients (5 resections of primary tumors, 1 metastasectomy) allowing histologic assessment of tumor response. In 4 cases there was complete response showing total replacement of tumor
TABLE 1. Clinical Features of BCOR-CCNB3 Sarcomas
Case
Age*/ Sex
Site
Size (mm)
Treatment
1 2 3 4 5
13/M 18/M 14/F 15/M 13/M
Femur (B) Psoas (ST) Pubic ramus (B) Tibia (B) Pelvis (B)
130130 60 9030 30 110 100 100 5030 30 160 160 120
6 7 8 9 10
16/M 16/M 14/M 17/M 11/M
Fibula (B) Quadriceps (ST) Calcaneus (B) Gastrocnemius (ST) Fibula (B)
13070 65 10575 50 10070 70 7065 55 150 largest dimension
I, D, MT; rx 11; rtx V, I, D, E; surgical decompression of spine; rtx; V, A, I V, I, D, E; rtx 11 and met; V, A, I V, I, D, E; rx 11; V, A I V, I, D, E; rtx to 11; whole-lung rtx; V, A, I, D for rec 11; rx rec 11 and lung met; palliative Rtx met; I, C, E V, I, D, E; rx of 11; V, A, I Rx of 11; V, I, D, E E, V, I, A, D; rx 11; whole-lung rtx Rx 11; rtx; V, I, D, E; rx lung met V, I, D, E; rx 11; V, A, I
Mets at Diagnosis
Outcome (mo of FU)
None None Bones Lung Lung
NED (78) DOD (28) DOD (27) NED (13) DOD (110)
None None Lung None None
NED (5) NED (132) NED (189) AWD (61) NED (145)
*Age at diagnosis. 11 indicates primary; A, actinomycin D; AWD, alive with disease; B, bone; C, carboplatin; D, doxorubicin; DOD, died of disease; E, etoposide; F, female; FU, followup; I, ifosfamide; M, male; Met, metastatic disease; MT, methotrexate; NED, alive, no evidence of disease; rec, recurrent; rtx, radiotherapy; rx, resection; ST, soft tissue; V, vincristine.
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1309
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
FIGURE 1. Morphology of BCOR-CCNB3 sarcoma. A, Sagittal PD fat-saturated MRI demonstrates a large soft tissue mass of predominately increased signal intensity within the lateral head of the gastrocnemius muscle compatible with a soft tissue sarcoma. The resection specimen shows a fleshy partially necrotic tumor (case 9). B, MRI-scan of the fibula demonstrates an aggressive, destructive lesion within the proximal fibula with a very extensive, circumferential soft tissue mass on the coronal STIR images. The imaging findings were interpreted as suspicious of Ewing sarcoma. The postoperative resection specimen shows regression of both extraosseous and intraosseous components (case 6). C, Intraosseous tumor with Ewing-like morphology and CD99 positivity (inset). CCNB3 shows strong nuclear positivity (case 4). D, Soft tissue sarcoma with small and round cell morphology and weak CD99 positivity (case 7). E, Primitive sarcoma with plump, short spindle cell morphology (inset) around gaping thin-walled vessels (case 1). F, Other example of BCOR-CCNB3 sarcoma with predominant short spindle cell morphology (case 8). CD99 is negative, and CCNB3 is only moderately positive. MRI indicates magnetic resonance imaging.
1310 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
FIGURE 2. Morphologic change of BCOR-CCNB3 sarcomas during disease progression. A, Primary moderately cellular sarcoma with spindle cell morphology, abundant myxoid matrix, and thin-walled vessels (prechemotherapy resection specimen, case 9). B–D, Several resections of pulmonary metastases showed increasing cellularity with focal osteoblastic differentiation; CCNB3 expression is retained (case 9). E, Primary pelvic BCOR-CCNB3 sarcoma with monomorphic oval cell morphology (case 5, pretreatment biopsy). F and G, Local recurrence after radiotherapy and chemotherapy showed similar morphology juxtaposed to undifferentiated pleomorphic sarcoma-like morphology (F), also seen in the pulmonary metastases (G). CCNB3 expression is retained throughout (inset).
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1311
Puls et al
with hypocellular loose fibrous tissue or in 1 case complete coagulative tumor necrosis (Figs. 3A, B). In 2 cases there were scattered residual neoplastic cells with altered morphology showing elongated bland-appearing nuclei without mitotic activity (Figs. 3C, D). In comparison with the prechemotherapy specimen, Ki-67 proliferative index was markedly reduced in these 2 cases, and CCNB3 immunostaining was weak to negative, although RT-PCR for BCOR-CCNB3 fusion transcript was positive (Figs. 3C, D).
Immunohistochemistry By immunohistochemical screening for nuclear CCNB3 accumulation, this series of BCOR-CCNB3 sarcomas was identified from 134 cases of undifferentiated, monomorphic sarcomas without a known fusion transcript or morphologically typical Ewing sarcomas without molecular confirmation (cases predating routine molecular confirmation of Ewing sarcoma in our archive). In
Am J Surg Pathol
Volume 38, Number 10, October 2014
normal tissues CCNB3 is expressed in a short time span in meiosis, and testis was used as a positive control (Fig. 4A).2 In confirmed BCOR-CCNB3 sarcomas, nuclear positivity was strong and uniform in 8 biopsies of primary tumors (Figs. 1C–E, insets). Two cases showed patchy but nevertheless distinct nuclear CCNB3 immunoreactivity (Fig. 1F; cases 8, 9). On some resection specimens tested, CCNB3 immunoreactivity was also only patchy with some background cytoplasmic staining (Figs. 2G, 4B). One postchemotherapy pulmonary metastasis did not show any CCNB3 immunoreactivity despite positive RT-PCR (Fig. 3D). Nonspecific weak cytoplasmic CCNB3 staining was found in some undifferentiated sarcomas, 4 of which were tested resulting in no BCORCCNB3 fusion transcripts being detected by RT-PCR (Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/PAS/A200, cases 1 to 4). Weak to moderate cytoplasmic staining was also identified in some
FIGURE 3. Chemotherapy response of BCOR-CCNB3 sarcomas. A, Complete response of 1 intraosseously confined tumor (inset, case 4). B, Complete response with focal coagulative tumor cell necrosis as well as granulation tissue replacement (inset, case 8). C, Partial response of pulmonary metastasis to chemotherapy showing a hypocellular lesion with bland spindle cell morphology and patchy, but distinct, CCNB3 expression (inset, case 9). D, Partial response to preoperative chemotherapy (case 1) with markedly reduced cellularity and relatively bland tumor cells, which were negative on CCNB3 immunohistochemistry (inset); RTPCR for BCOR-CCNB3 was positive.
1312 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
FIGURE 4. CCNB3 immunohistochemistry. A, Testes used for positive control showing a layer of positive spermatocytes correlating with strictly controlled expression during spermatogenesis. B, Patchy and moderate CCNB3 expression (resection pulmonary metastasis, case 9). C, Ewing sarcoma cases with variant translocations did not show nuclear expression for CCNB3. D, Alveolar rhabdomyosarcoma showed strong cytoplasmic positivity, although tumor cell nuclei were negative. E, In areas of necrotic Ewing sarcoma, nuclear “pseudopositivity” was observed. This case was EWS-FLI1 positive and BCOR-CCNB3 negative on RT-PCR. Viable areas of this case were negative for CCNB3. F–I, Additional immunohistochemical markers in BCOR-CCNB3 sarcomas. There was cytoplasmic expression of bcl2 (9/10), membranous expression for b-catenin (7/8), membranous expression for CD117 (6/10), and weak expression of bcl-6 (3/5).
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1313
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
FIGURE 5. RT-PCR for detection of BCOR-CCNB3 fusion transcripts. A, 140 bp amplicons spanning the fusion of BCOR-CCNB3 for several cases and for metastasis. B, Confirmation by sequencing. C, Survival by primary site for BCOR-CCNB3 sarcoma patients. D, Survival by primary site for patients with classical Ewing sarcoma. E, Survival according to detectable metastases at diagnosis for BCOR-CCNB3 and classical Ewing sarcoma patients. F, Overall survival of BCOR-CCNB3 and classical Ewing sarcoma patients (M, 25 bp DNA ladder; POS, positive control1; EWS, Ewing sarcoma; BC3, BCOR-CCNB3 sarcoma).
cases of classic Ewing sarcoma, alveolar rhabdomyosarcoma, and very focally in some synovial sarcomas, all of which did not show any BCOR-CCNB3 fusion transcripts
1314 | www.ajsp.com
(Figs. 4C–E, Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/PAS/A199). In the confirmed cases of BCOR-CCNB3 sarcomas, staining for r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
additional immunohistochemical markers was performed, and positive fractions were as follows: CD99 (7/10), CD117 (6/10), Ki-67 (30% to 50% in 9 tested), bcl2 (9/10), bcl-6 (weakly positive 3/5), desmin (0/10), CD45 (0/10), AE1/ AE3 (0/10), and membranous/cytoplasmic b-catenin (7/8). Immunophenotypes are shown in Table 2, and representative sections of immunohistochemistry are depicted in Figures 4F–I.
(P = 0.0857; Figs. 5C, D, Table 3). In classical Ewing sarcoma, the presence of metastatic disease at the time of diagnosis was significantly associated with shorter overall survival (P < 0.0001), whereas in BCOR-CCNB3 no association between survival and metastatic disease at diagnosis was apparent (P = 0.3118; Fig. 5E, Table 3).
Reverse Transcription Polymerase Chain Reaction
The recent identification of sarcomas with BCORCCNB3 fusion within the problematic and heterogeneous group of non-Ewing small round cell sarcoma is yet another example of the power of high-throughput technology, which is rapidly expanding the number of recognized translocation-associated sarcomas.6–9 The BCOR-CCNB3 sarcomas have, to date, only been reported in the original series by Pierron et al1 focusing on the genetic and pathogenetic aspects of BCOR-CCNB3 fusion sarcoma. The present study of BCOR-CCNB3 sarcomas diagnosed and treated at a single institution was carried out to estimate relative incidence to Ewing sarcoma, to further characterize the histopathologic features and clinical behavior of this newly identified sarcoma, and to aid further decisions regarding classification. The group of fusion-defined Ewing sarcoma treated by the same multidisciplinary team over the same period served as an ideal control group. The morphologic spectrum of BCOR-CCNB3 sarcomas in this series was quite varied. Most cases were highly cellular and dominated by fairly small cells with angulated nuclei, different from the more uniform round cell morphology of Ewing sarcoma and somewhat resembling so-called “atypical Ewing sarcoma.”4,10,11 A few tumors had a spindle cell morphology with, in areas, a fibrous or myxoid stroma. These findings are in line with those of Pierron et al,1 some of whose cases were initially labeled as “fusiform sarcoma.” As in their series, malignant peripheral nerve sheath tumor and poorly differentiated synovial sarcoma were also among the diagnoses initially considered. One of their cases was thought to be a small cell osteosarcoma, most likely on the basis of focal osteoid formation, which was also observed in 1 of our cases.1 The solid, nonmyxoid, small cell variant of extraskeletal myxoid chondrosarcoma may also mimic BCORCCNB3 sarcoma.12 Spindle cell, sclerosing, and large cell variants of Ewing sarcoma are also recognized.11,13 Another feature in our series was the pronounced pleomorphism seen in local recurrences and metastases, occurring after chemotherapy and/or radiotherapy and simulating undifferentiated pleomorphic sarcoma, a feature not typically encountered in Ewing sarcoma. Awareness of the broad morphologic spectrum of these sarcomas and of the resultant long list of differential diagnoses justifies CCNB3 immunohistochemistry and molecular analysis for BCOR-CCNB3 fusion in all poorly differentiated, difficult-to-classify sarcomas. In our experience, CCNB3 immunohistochemistry for screening and RT-PCR for confirmation appears to be the preferred approach. RT-PCR can be utilized efficiently because the fusion transcript appears to be constant requiring only a
RT-PCR generated 140 bp amplicons, and sequencing confirmed identical fusion points of exon 15 of BCOR to exon 5 of CCNB3 (Fig. 5B) in 9/10 cases with strong nuclear CCNB3 staining by immunohistochemistry. Case 10 showed strong and diffuse nuclear immunohistochemical positivity for CCNB3, but attempted RNA extraction did not yield sufficient mRNA, presumably because of previous decalcification of biopsy material by nitric acid. RT-PCR was negative in 20 cases of other translocation sarcomas (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/PAS/ A199) and in 9 cases of randomly chosen CCNB3-negative, not otherwise specified, small round cell sarcomas (Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/PAS/A200, cases 5 to 13).
Comparison With Classic Ewing Sarcoma With Known Translocations When compared with 242 fusion-positive classical Ewing sarcomas diagnosed at our institution, a higher proportion of BCOR-CCNB3 sarcomas occurred in soft tissue (30% vs. 5.8%, P < 0.05). The median age at the time of diagnosis was identical (15.5 y), and the mean ages at diagnosis for Ewing sarcoma and BCOR-CCNB3 sarcoma were 18 and 15 years, respectively (Table 3). At the time of diagnosis, metastatic disease was present in 40% of BCOR-CCNB3 sarcomas and in 39% of Ewing sarcomas. Since 2005, when the screening for classic Ewing sarcoma translocations EWS-FLI1, EWS-ERG, FUSERG, and later EWS-ETV1 and FUS-FEV had been routinely performed at our institution, the ratio of identified BCOR-CCNB3 sarcomas to classical Ewing sarcomas was 1:27. Follow-up data were available for 234 patients with classical Ewing sarcoma (median follow-up 30.5 mo, range 1 to 175 mo) and all 10 patients with BCOR-CCNB3 sarcoma (median follow-up 69.5 mo, range 5 to 189 mo). Five-year survival of all patients with classical Ewing sarcoma and BCOR-CCNB3 sarcoma was 54% and 75%, respectively, but this difference was not statistically significant (P = 0.3541; Fig. 5F, Table 3). For both entities, longer overall survival was significantly linked to location of the primary tumor in the extremities versus axial skeleton and soft tissues (P < 0.01). The survival difference between BCOR-CCNB3 and Ewing sarcoma within these subgroups was not significant, although there was a trend toward longer survival in patients with BCOR-CCNB3 sarcomas of the extremities compared with Ewing sarcoma in the same location r
2014 Lippincott Williams & Wilkins
DISCUSSION
www.ajsp.com |
1315
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
TABLE 2. Morphologic and Immunohistochemical Features of BCOR-CCNB3 Sarcomas Case
Tumor Examined
1
Bone primary (bx, rx)
2
Soft tissue primary (bx) Bone primary (bx) Bone primary (bx, rx)
3 4 5
6 7 8 9
Original Diagnosis
Primitive STS, Angulated short spindle cells possibly MPNST
37
Primitive sarcoma, Short spindle cells with overlapping nuclei Ewing-like Ewing-like Small densely packed plasmacytoid cells, sarcoma PAS+ cytoplasmic inclusions Ewing-like Small round blue cell tumor, Ewing-like sarcoma
Bone primary (bx) Recurrent tumor (bx, rx) Metastatic tumor, lung (rx) Bone primary (bx) Soft tissue primary (bx, rx) Bone primary (bx, rx) Soft tissue primary (bx, rx) Metastatic tumor, lung (rx post Ct) Metastatic tumor, lung (rx)
10
Mitoses/ CCNB3 10 hpf IHC (N)
Morphology
Bone primary
Ewing-like sarcoma MPNST Metastatic highgrade STS BCOR-CCNB3 sarcoma Ewing sarcoma
Plump, short, cells, edematous matrix thin-walled vessels Spindle cells, scattered marked pleomorphic cells, fibrotic matrix Undifferentiated pleomorphic sarcoma-like, fibrotic matrix Small round blue cell tumor, Ewing-like
8 10 22 32 20 34 32
CD99 , CD117+, bcl2+, b-cat+ (m), ki-67 50%, bcl-6 (w) Strong, CD99 ,CD117+, ki-67 30% diffuse Strong, CD99+, CD117+, ki-67 diffuse 50% Strong, CD99+, ki-67 40%, CD117 , bcl2+, diffuse b-cat+ (m) Strong, CD99+, CD117 , ki-67 30%, bcl2+, b-cat+ (m) diffuse Strong, CD99+, bcl2+, b-cat+ (m), bcl-6 (w) diffuse Strong, CD99+, ki-67 30%, bcl2+, b-cat+ (m), diffuse Strong, CD99+, bcl2+, b-cat+ (m), diffuse ki-67 40%, CD117+, bcl-6 (w) Strong, CD99+, ki-67 30%, bcl2+, diffuse b-cat-, CD117 Moderate, CD99 , ki-67 30%, CD117 , bcl2+, patchy b-cat+ (m) Strong, CD99+, ki-67 40%, patchy CD117+, bcl2+, b-cat+ (m) Strong, CD99+, ki-67 5%, CD117+ patchy Strong, diffuse
31
High-grade spindle cell sarcoma, NOS Primitive STS, NOS
Round cells with pale cytoplasm, fibrotic matrix Small spindle cells with marked pleomorphism, vacuolated cytoplasm, edematous matrix Small plump elongated cells in edematous, focally myxoid matrix
Metastatic STS
Relatively bland elongated spindle cells
39
Metastatic STS
Variable morphology, synovial-sarcomalike, osteoblastic foci
24
Strong, patchy
Ewing sarcoma
Small round blue cell tumor, Ewing-like
5
Strong diffuse
18 30
Other IHC
CD99+, ki-67 40%, bcl2+, b-cat+ (m), CD117+, bcl-6 (w) CD99+, ki-67 10%, bcl2 , b-cat+ (m), CD117+
RTPCR Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Fail
bx indicates biopsy; C, cytoplasmic positivity; Ct, chemotherapy; m, membranous positivity; MPNST, malignant peripheral nerve sheath tumor; n, nuclear positivity; n/p, not performed; NOS, not otherwise specified, rx, resection; STS, soft tissue sarcoma; w, weak.
single pair of primers. Alternatively, a fluorescence in situ hybridization approach for detection of the X-chromosomal inversion can be used.14 Apart from CCNB3 positivity, the immunohistochemical profile showed broad overlap with that seen in classic Ewing sarcoma, although CD99 staining was generally weaker in most BCOR-CCNB3 sarcomas or, occasionally, completely absent. Despite evidence of activation of the Wnt pathway in expression profiling of BCOR-CCNB3 sarcoma, there was no nuclear staining for b-catenin in our cases.1 BCOR is a transcriptional repressor interacting with bcl-6.15 Interestingly, we observed nuclear bcl-6 staining in 3 of 5 cases tested. Staining was weaker than in lymphoid follicular centers, which were used as positive controls, and does not appear suitable for diagnostic exploitation. In gene expression profiling, there was no evidence of increased transcription of wild-type BCOR mRNA, and therefore nuclear positivity for bcl-6 may indicate that BCOR-CCNB3 fusion protein interacts with bcl-6 in BCOR-CCNB3 sarcomas.1
1316 | www.ajsp.com
BCOR-CCNB3 sarcoma appears to be rare. By screening 594 undifferentiated sarcomas Pierron et al1 identified 24 cases. To estimate incidence and minimize selection bias we compared the ratio of classical Ewing sarcoma with BCOR-CCNB3 sarcoma primarily diagnosed at the Royal Orthopaedic Hospital Birmingham. Although these findings need to be interpreted with caution because of low numbers, the finding of 27 cases of classical Ewing sarcoma for every case of BCOR-CCNB3 sarcoma is also in keeping with a very rare sarcoma. As in the series of Pierron and colleagues there was a predominance of sarcomas originating in the bone and a majority of patients presenting during the second decade. Male predominance was more pronounced in our series (1:9) than in the series of Pierron and colleagues (1:2). This differs markedly from classic Ewing sarcoma both in our control population and other series (1:1.4).4 However, the exact mechanism for this obviously increased male susceptibility for development of this paracentric inversion of the X-chromosome remains unclear. r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
TABLE 3. Comparison With Classic Ewing Sarcoma
Sex F:M Median age (range) [y] Location (n [%]) Bone Soft tissue Axial (spine, pelvis) Trunk, chest wall Extremity Metastatic disease at diagnosis (%) Ratio (translocationpositive Ewing sarcoma:BCORCCNB3 sarcoma; all cases January 2005 to August 2013) Outcomes Duration of follow-up
Ewing Sarcoma
BCOR-CCNB3 Sarcoma
100:142, 1:1.42 (242 informative cases) 15.5 (3-62)
1:9 (10 informative cases) 15.5 (11-18)
228 (94.2) 14 (5.8) 63 (26.0) 28 (11.6) 151 (62.4) 87 (39) (223 informative cases) n = 190, 1:27.1
7 (70) 3 (30) 3 (30) 0 7 (70) 4 (40) (10 informative cases) n = 7, 1:27.1
1-175 mo, mean 5-189 mo, mean 42 mo, median 79 mo, median 30.5 mo (234 69.5 mo (10 informative cases) informative cases) 5-year survival, 54%, median survival 75%, median survival 72 mo (234 undefined (10 median survival, informative cases) informative cases) all cases 5-year survival, 29,8%, median 33.3%, median median survival, survival 31 mo survival 28 mo axial soft tissues and (64 cases) (3 cases) skeleton (without trunk) 5-year survival, 60.3%, median 100%, median median survival, survival undefined survival undefined location in extremity (142 cases) (7 cases) 5-year survival, 29.3%, median 66.7%, median median survival, survival 26 mo (85 survival, 110 mo mets at time of cases) (4 cases) diagnosis 5-year survival, 69%, median survival 80%, median survival median survival, no undefined (131 cases) undefined (6 cases) mets at time of diagnosis
In our series, overall survival of BCOR-CCNB3 sarcoma (75% at 5 y) was not statistically different from our large series of classic Ewing sarcoma (54% at 5 y). In view of the low number of cases, possible variation in treatment over time, and the retrospective analysis, the findings have to be interpreted with caution. Similar to Ewing sarcoma, patients with axial/pelvic BCOR-CCNB3 sarcomas had a shorter survival than patients with tumors in the extremities or trunk.16,17 Comparison of extra-axially located BCOR-CCNB3 sarcomas with classical Ewing sarcoma showed a trend toward longer survival in BCOR-CCNB3 sarcoma. In Ewing sarcoma, the presence of metastatic disease at diagnosis is an adverse prognostic factor.17 In our series of Ewing and BCOR-CCNB3 sarcomas, a similar proportion of patients presented with metastases. Four patients with BCOR-CCNB3 sarcoma presenting with metastasis, however, are too few for any definite conclusion on prognostic significance. The presence of 1 long-term survivor (>15 y) within this subgroup r
2014 Lippincott Williams & Wilkins
BCOR-CCNB3 Sarcoma
indicates that aggressive treatment, including whole-lung radiation, may be justified. Because new therapeutic approaches are likely to selectively target specific fusion transcripts and downstream pathways, correct identification of oncogenic fusion transcript in Ewing and Ewing-like sarcoma will be vitally important for patient care.18,19 Distinction from other “Ewing-like” small blue cell sarcomas, such as those associated with CIC-DUX4 fusion, appears important. Within the last few years sarcomas with CIC-DUX4 fusions have been increasingly identified in soft tissue.20–25 Molecular and morphologic similarities of CIC-DUX4 and classical Ewing sarcoma have led to speculation that classical Ewing sarcoma and CIC-DUX4 sarcoma may be closely related.22,26 Other authors have pointed out morphologic and clinical differences of CIC-DUX4 sarcoma compared with Ewing sarcoma.20 Compared with BCOR-CCNB3 sarcoma in the present and the previous series, reported cases of CIC-DUX4 sarcoma show a wider age range, exclusive occurrence in soft tissues, and absence of clear male preponderance.1,20,21,24 Its clinical behavior is aggressive, with 7 of 19 and 4 of 4 patients dying of disease within 16 months of diagnosis.20,21 Despite overlapping morphologic and immunohistologic features, distinction of BCOR-CCNB3 sarcoma, CICDUX4 sarcoma, and classical Ewing sarcoma appears to be of prognostic relevance. In conclusion, this study provides a first detailed description of the histologic spectrum, immunohistochemical features, and clinical characteristic of BCORCCNB3 sarcoma. Bearing in mind the morphologic spectrum, we believe that combined CCNB3 immunohistochemistry and RT-PCR analysis should be performed on primitive high-grade bone and soft tissue sarcoma not fitting into any defined entity in terms of morphology, immunoprofile, and genetic aberration, particularly if these occur in childhood and adolescence. ACKNOWLEDGMENTS The authors are indebted to Dr G. Pierron, Institut Curie, Unite´ de Ge´ne´tique Somatique–Centre Hospitalier, Paris, France, for providing control material from BCORCCNB3 cases and Dr Y. Hock, Department of Histopathology, Walsall Healthcare NHS Trust, Walsall, UK, for performing bcl-6 immunohistochemistry. The authors thank Elaine Haywood, Karen Joynes, and Jade Clarke for excellent technical assistance. REFERENCES 1. Pierron G, Tirode F, Lucchesi C, et al. A new subtype of bone sarcoma defined by BCOR-CCNB3 gene fusion. Nat Genet. 2012;44: 461–466. 2. Nguyen TB. Characterization and expression of mammalian cyclin B3, a prepachytene meiotic cyclin. J Biol Chem. 2002;277:41960–41969. 3. Fletcher CDM, Chibon F, Mertens F. Undifferentiated/unclassified sarcomas. In: Fletcher CDM, Bridge JA, Hogendoorn PCW, et al, eds. WHO Classification of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2013:236–238. 4. de Alava E, Lessnick SL, Sorensen PH. Ewing sarcoma. In: Fletcher CDM, Bridge JA, Hogendoorn PCW, et al, eds. WHO Classification
www.ajsp.com |
1317
Puls et al
5.
6. 7.
8. 9. 10. 11.
12.
13.
14. 15.
of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2013:306–309. Williams A, Bartle G, Sumathi VP, et al. Detection of ASPL/TFE3 fusion transcripts and the TFE3 antigen in formalin-fixed, paraffinembedded tissue in a series of 18 cases of alveolar soft part sarcoma: Useful diagnostic tools in cases with unusual histological features. Virchows Arch. 2011;458:291–300. van de Rijn M, Guo X, Sweeney RT, et al. Molecular pathological analysis of sarcomas using paraffin-embedded tissue: current limitations and future possibilities. Histopathology. 2014;64:163–170. Wang L, Motoi T, Khanin R, et al. Identification of a novel, recurrent HEY1-NCOA2 fusion in mesenchymal chondrosarcoma based on a genome-wide screen of exon-level expression data. Genes Chromosom Cancer. 2012;51:127–139. Chmielecki J, Crago AM, Rosenberg M, et al. Whole-exome sequencing identifies a recurrent NAB2-STAT6 fusion in solitary fibrous tumors. Nat Genet. 2013;45:131–132. Robinson DR, Wu Y-M, Kalyana-Sundaram S, et al. Identification of recurrent NAB2-STAT6 gene fusions in solitary fibrous tumor by integrative sequencing. Nat Genet. 2013;45:180–185. Machado I, Noguera R, Mateos EA, et al. The many faces of atypical Ewing’s sarcoma. A true entity mimicking sarcomas, carcinomas and lymphomas. Virchows Arch. 2010;458:281–290. Folpe AL, Goldblum JR, Rubin BP, et al. Morphologic and immunophenotypic diversity in Ewing family tumors: a study of 66 genetically confirmed cases. Am J Surg Pathol. 2005;29: 1025–1033. Meis-Kindblom JM, Bergh P, Gunterberg B, et al. Extraskeletal myxoid chondrosarcoma: a reappraisal of its morphologic spectrum and prognostic factors based on 117 cases. Am J Surg Pathol. 1999; 23:636–650. Llombart-Bosch A, Machado I, Navarro S, et al. Histological heterogeneity of Ewing’s sarcoma/PNET: an immunohistochemical analysis of 415 genetically confirmed cases with clinical support. Virchows Arch. 2009;455:397–411. Antonescu C. Round cell sarcomas beyond Ewing: emerging entities. Histopathology. 2014;64:26–37. Huynh KD, Fischle W, Verdin E, et al. BCoR, a novel corepressor involved in BCL-6 repression. Genes Dev. 2000;14:1810–1823.
1318 | www.ajsp.com
Am J Surg Pathol
Volume 38, Number 10, October 2014
16. Bacci G, Ferrari S, Mercuri M, et al. Multimodal therapy for the treatment of nonmetastatic Ewing sarcoma of pelvis. J Pediatr Hematol Oncol. 2003;25:118–124. 17. Cotterill SJ, Ahrens S, Paulussen M, et al. Prognostic factors in Ewing’s tumor of bone: analysis of 975 patients from the European Intergroup Cooperative Ewing’s Sarcoma Study Group. J Clin Oncol. 2000;18:3108–3114. 18. Stegmaier K, Wong JS, Ross KN, et al. Signature-based small molecule screening identifies cytosine arabinoside as an EWS/FLI modulator in Ewing sarcoma. PLoS Med. 2007;4:e122. 19. Lessnick SL, Ladanyi M. Molecular pathogenesis of Ewing sarcoma: new therapeutic and transcriptional targets. Ann Rev Pathol. 2012;7:145–159. 20. Choi EY, Thomas DG, McHugh JB, et al. Undifferentiated small round cell sarcoma with t(4;19)(q35;q13.1) CIC-DUX4 fusion: a novel highly aggressive soft tissue tumor with distinctive histopathology. Am J Surg Pathol. 2013;37:1379–1386. 21. Italiano A, Sung YS, Zhang L, et al. High prevalence of CIC fusion with double-homeobox (DUX4) transcription factors in EWSR1negative undifferentiated small blue round cell sarcomas. Genes Chromosom Cancer. 2012;51:207–218. 22. Kawamura-Saito M. Fusion between CIC and DUX4 up-regulates PEA3 family genes in Ewing-like sarcomas with t(4;19)(q35;q13) translocation. Hum Mol Genet. 2006;15:2125–2137. 23. Machado I, Cruz J, Lavernia J, et al. Superficial EWSR1-negative undifferentiated small round cell sarcoma with CIC/DUX4 gene fusion: a new variant of Ewing-like tumors with locoregional lymph node metastasis. Virchows Arch. 2013;463:837–842. 24. Graham C, Chilton-MacNeill S, Zielenska M, et al. The CIC-DUX4 fusion transcript is present in a subgroup of pediatric primitive round cell sarcomas. Hum Pathol. 2012;43:180–189. 25. Kajtar B, Tornoczky T, Kalman E, et al. CD99-positive undifferentiated round cell sarcoma diagnosed on fine needle aspiration cytology, later found to harbour a CIC-DUX4 translocation: a recently described entity. Cytopathology. 2014;25:129–132. 26. World Health Organization. Fletcher CDM, Bridge JA, Hogendoorn PCW, et al. WHO Classification of Tumours of Soft Tissue and Bone. Lyon: International Agency for Research on Cancer; 2013.
r
2014 Lippincott Williams & Wilkins
BCOR-CCNB3 (Ewing-like) Sarcoma A Clinicopathologic Analysis of 10 Cases, In Comparison With Conventional Ewing Sarcoma Florian Puls, MD, FRCPath,* Angela Niblett, FIBMS,* Gillian Marland, BSc,* Czar Louie L. Gaston, MD,w Hassan Douis, FRCR,z D. Chas Mangham, FRCPath,*y Vaiyapuri P. Sumathi, FRCPath,* and Lars-Gunnar Kindblom, MD, PhD*
Abstract: BCOR-CCNB3 fusion transcripts resulting from an X-chromosomal paracentric inversion were recently identified in a series of unclassifiable soft tissue and bone sarcomas with Ewing sarcoma–like morphology. The morphologic and clinical features of these sarcomas are, as yet, not well characterized. Here we describe the clinicopathologic features of 10 cases of BCOR-CCNB3 sarcoma and compare their clinical course with typical Ewing sarcoma. Nine of 10 patients were male, and all were 11 to 18 years of age. Seven tumors were located in the bone and 3 in the deep soft tissues. The histomorphologic spectrum was quite wide, with 7 tumors predominately showing small primitive cell morphology with angulated nuclei simulating so-called atypical Ewing sarcoma and 3 predominately showing spindle cell morphology. Recurrent and metastatic lesions showed increased cellularity and marked pleomorphism. Immunohistochemistry showed expression of CCNB3 (100%), bcl2 (90%), CD99 (60%), and CD117 (60%). Reverse transcription polymerase chain reaction for BCOR-CCNB3 fusion transcripts was positive in all 9 cases, which yielded sufficient extracted RNA. Five- and 10-year survival rates were 75% and 56%, respectively. BCOR-CCNB3 sarcomas located in axial skeleton and soft tissues showed a significantly shorter survival. The Ewing sarcoma overall survival was not statistically different, although there was a trend for longer survival of patients with BCOR-CCNB3 sarcomas in the extremities. In conclusion, this study provides a detailed description of the histologic From the Departments of *Musculoskeletal Pathology; wOncology; zMusculoskeletal Radiology, Royal Orthopaedic Hospital NHS Foundation Trust, Birmingham; and yDepartment of Histopathology, Robert Jones & Ages Hunt Orthopaedic Hospital, Oswestry, UK. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Florian Puls, MD, FRCPath, Department of Musculoskeletal Pathology, Royal Orthopaedic Hospital NHS Foundation Trust, Robert Aitken Institute of Clinical Research, University of Birmingham, Birmingham B15 2 TT, UK (e-mail: fl[email protected]). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.ajsp.com. Copyright r 2014 by Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
spectrum, immunohistochemical features, and clinical characteristic of BCOR-CCNB3 sarcoma justifying distinction from Ewing sarcoma with its typical EWS/FUS-ETS translocations. Ideally immunohistochemistry is used in combination with reverse transcription polymerase chain reaction for definitive diagnosis. Key Words: BCOR-CCNB3 sarcoma, Ewing sarcoma, translocation-associated sarcoma, small round cell tumor, RT-PCR, immunohistochemistry (Am J Surg Pathol 2014;38:1307–1318)
A
subset of soft tissue and bone tumors with primitive “small round blue cell” morphology defies exact classification even after extensive diagnostic workup with immunohistochemistry and molecular techniques, thus posing a recurrent diagnostic and therapeutic challenge to pathologists and clinicians. Recently, Pierron et al1 identified recurrent gene fusions of BCOR (encoding Bcl6 interacting corepressor) and CCNB3 (cyclin B3) in primitive and, so far, unclassifiable small cell sarcomas in adolescents and young adults. BCOR-CCNB3 mRNA originates from a paracentric inversion on the X-chromosome and splicing of the end of the BCOR coding sequence to the CCNB3 exon 5 splice acceptor site.1 The resultant fusion protein is composed of full-length BCOR, a transcriptional repressor, and the C-terminus of CCNB3, a cyclin normally expressed in leptotene and zygotene of meiosis.2 In vitro studies suggest that the fusion BCOR-CCNB3 protein is oncogenic and drives proliferation in this sarcoma.1 BCOR-CCNB3 sarcoma showed a distinct gene expression profile when compared with classic Ewing sarcoma.1 Although currently listed under Ewing sarcoma in the recent World Health Organization classification, it could be argued that, on the basis of the distinct gene expression profiles, BCOR-CCNB3 sarcoma should be viewed as a separate entity.3,4 The original report of BCOR-CCNB3 fusion sarcoma focusing on its genetic and pathophysiological aspects is the only series reported to date.1 In this study we aimed to further characterize the morphologic, clinical, and www.ajsp.com |
1307
Puls et al
Am J Surg Pathol
prognostic features as well as the relative incidence of this sarcoma in comparison with its purported closest mimic, classical Ewing sarcoma.
MATERIALS AND METHODS Patients Archival cases from 1998 to 2013 of soft tissue or bone tumors with monotonous or primitive small cell morphology without a specific line of differentiation or specific identifiable transcripts, as well as cases of Ewing sarcoma predating routine molecular confirmation as part of the diagnostic workup, were screened for expression of CCNB3 by immunohistochemistry. These included cases diagnosed as Ewing-like sarcoma, small cell malignant peripheral nerve sheath tumor, poorly differentiated synovial sarcoma, or extraskeletal myxoid chondrosarcoma without molecular confirmation. In total, 134 cases were screened for CCNB3 expression by immunohistochemistry. The typical Ewing sarcoma control cases with confirmed translocations had their initial diagnosis made or their tumor resected at the Royal Orthopaedic Hospital in Birmingham between 2005 and August 2013. Second opinion pathology referral cases or patients not receiving treatment and follow-up at the Royal Orthopaedic Hospital were excluded. Sample collection and data analysis were conducted according to local ethical guidelines.
Histology and Immunohistochemistry For immunohistochemistry, paraffin wax sections were cut at 2 mm, dried, and antigen retrieved in an epitope retrieval solution of pH8 (RE7116; Novocastra, Newcastle, UK) at 681C for 17 hours in a stirred water bath. For detection of nuclear CCNB3, the primary rabbit polyclonal anti-CCNB3 antibody (HPA000496; Sigma, Gillingham, UK) was used in 1:200 dilution. Normal testis was used as positive control on the same slide. Other antibodies, clones, dilutions, and sources were: CD99 (12E7, 1:25; Dako, Ely, UK), desmin (D33, 1:100; Dako), S100 (polyclonal, 1:1000; Novocastra), cytokeratins (AE1/AE3, 1:100; Dako), Ki-67 (MIB-1, 1:200; Dako), CD117 (polyclonal, 1:100; Dako), bcl2 (124, 1:50; Dako), b-catenin (17C2; Novocastra, 1:50), and bcl-6 (LN22; Novocastra, 1:100). The DakoCytomation EnVision Detection System peroxidase/DAB visualization kit (K5007; Dako) was used throughout except for bcl-6 immunohistochemistry, wherein the Ultraview Universal DAB Detection and Amplification Kits (Ventana Medical Systems, Tucson, AZ) were used for visualization. All kits were used according to the manufacturer’s instructions, and same slide positive controls were used throughout. Mitoses were counted in ten 40 objective fields (field area of 0.238 mm2).
Reverse Transcription Polymerase Chain Reaction and Sequence Analysis Reverse transcription polymerase chain reaction (RT-PCR) was performed on formalin-fixed, paraffinembedded tissue except for 3 cases in which fresh frozen
1308 | www.ajsp.com
Volume 38, Number 10, October 2014
tumor tissue was available. Total RNA was extracted from fresh frozen tumor tissue using the RNeasy MiniKit (Qiagen, Hilden, Germany) and from formalin-fixed, paraffin-embedded sections using acid phenol chloroform separation and isopropanol precipitation as previously described.5 Amplification of a 98 bp amplicon of b-actin mRNA (forward 50 -ATCACCATTGGCAATGAGCG-30 , reverse 50 -TTGAAGGTAGTTTCGTGGAT-30 ) was used to confirm the presence of intact and amplifiable RNA. For BCOR-CCNB3 transcript detection a forward primer located in BCOR exon 15, 50 -GGCTCCACCCCAGTGAT CT-30 (BCOR-P1), and a reverse primer in CCNB3 exon 5, 50 -GGGTGTTTTGGAGGTGGTGGAT-30 (CCNB3-P2), were chosen to give an expected 140 bp amplicon spanning the fusion site within the BCOR-CCNB3 transcript (GenBank: JN813375.1, accession number JN813375).1 RNA from a confirmed BCOR-CCNB3 case (kindly provided by G. Pierron, Institut Curie, Unite´ de Ge´ne´tique Somatique– Centre Hospitalier, Paris, France) was used as a positive control, and 21 cases of small round blue cell tumors with known fusion transcripts (see Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/ PAS/A199) were used as negative controls. All primers were purchased from Eurofins MWG Operon-Biotech (Ebersberg, Germany), and RT-PCR reactions were performed with the OneStep RT-PCR kit (Qiagen) for 40 cycles with annealing temperatures of 601C and 611C for bactin and BCOR-CCNB3 primer pairs, respectively. Reaction products were visualized using GelRed staining (Cambridge Bioscience Ltd, Cambridge, UK) after separation by 10% polyacrylamide gel electrophoresis. PCR products of all BCOR-CCNB3 cases were Sanger sequenced at the Functional Genomics Laboratory at The University of Birmingham using an ABI 3730 Genetic Sequencer (Applied Biosystems, Warrington, UK).
Statistical Analysis The statistical analysis of baseline demographics and clinical outcome is based on all the data available up to January 15, 2014. Overall survival time was calculated as the number of completed months between the date of diagnostic biopsy and whichever occurred first: the date of death or January 15, 2014 (BCOR-CCNB3 sarcoma cases)/September 1, 2013 (classic Ewing sarcoma cases). Survival rates were calculated using the Kaplan-Meier method. For survival curve comparison the log rank (Mantel-Cox) test was performed using GraphPad Prism v5.00 (GraphPad, San Diego).
RESULTS In total 10 cases of BCOR-CCNB3 sarcomas were identified by immunohistochemical screening for CCNB3 expression, of which 9 cases underwent successful RTPCR analysis.
Clinical Features Nine of 10 BCOR-CCNB3 sarcoma patients identified were male, and all were in their second decade of life (age range 11 to 18 y, median 15.5 y). Seven tumors had r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
arisen in the bone, of which 6 had soft tissue extension. Radiologically, the largest tumor diameter ranged from 50 to 160 mm (median 107.5 mm). Computed tomography performed at the time of diagnosis showed metastatic disease in 4 cases (lungs 3, bone 1). Five patients received neoadjuvant chemotherapy before curative resection (cases 1, 4, 6, 8, 10; Table 1), 3 patients with axially located sarcomas received chemotherapy in combination with radiotherapy as first-line treatment (cases 2, 3, 5; Table 1). Curative surgical resection of the primary sarcoma in the extremity was performed in 7 patients, none of whom experienced local recurrence. One locally recurrent pelvic sarcoma was resected 25 months after the patient had received chemotherapy and radiotherapy (case 5). One patient was treated with local surgical decompression for spinal canal involvement on disease progression (case 2, Table 1). Surgical resection of metastatic disease was performed in 2 patients. Postoperative radiotherapy to the primary site was given to 4 patients, and 2 patients received whole-lung radiotherapy. Clinical follow-up information was available in all 10 patients (5 to 189 mo, median 69.5 mo). Six patients were without clinical evidence of disease at 5, 13, 78, 132, 145, and 189 months of follow-up, and 1 patient was alive with lung metastases at 61 months. Three patients died of disease after 27, 28, and 110 months (Table 1, Figs. 5C, E, F).
Histopathologic Features of BCOR-CCNB3 Sarcomas Biopsies of primary sarcomas were available from all 10 patients. Resection specimens without previous chemotherapy for unaltered morphologic assessment were available from 2 patients, and 5 patients received chemotherapy before tumor resection. In all but 1 case, areas with high cellularity predominated. Seven tumors were predominately composed of small round to plump tumor cells with a small amount
BCOR-CCNB3 Sarcoma
of ill-defined pale to eosinophilic cytoplasm (Figs. 1C–E). Nuclei were angulated and hyperchromatic with finely dispersed to glassy chromatin, and nucleoli were not readily discernible in the majority of cases. There were numerous apoptotic bodies, and the mitotic rate ranged from 5 to 37/10 high-power fields. Small areas of tumor cell necrosis were seen in 9 of 10 biopsies. In 3 cases tumor cells predominately showed a short spindle cell morphology (Figs. 1F, 2A). In all tumors the cells were discohesive without formation of a distinct architectural pattern and lay within an edematous to myxoid matrix with small myxoid pools in 1 case (Fig. 2A). The matrix was richly vascularized with thin-walled, sometimes gaping, vessels. In 5 cases intravascular thrombi were seen (Fig. 1E). Resections of recurrent or metastatic tumors were available from 2 patients. When compared with the primary tumor, 1 recurrent sarcoma showed a higher degree of pleomorphism with large, highly atypical spindle cells within a fibrotic matrix simulating undifferentiated pleomorphic sarcoma (Fig. 2F, case 5). Subsequent resection of a pulmonary metastasis in this patient showed a similar pleomorphic morphology (Fig. 2G). Several resections of pulmonary metastases from another patient showed increased cellularity, elongated spindle cells focally simulating synovial sarcoma, and scattered foci of osteoblastic differentiation (Figs. 2C, D, case 9). Despite the markedly different morphology compared with the primary tumor, these metastases were confirmed as originating from the known primary by nuclear expression of CCNB3 immunohistochemically and the detection of a BCORCCNB3 fusion transcript by RT-PCR. Resection shortly after completion of first-line chemotherapy was performed in 6 patients (5 resections of primary tumors, 1 metastasectomy) allowing histologic assessment of tumor response. In 4 cases there was complete response showing total replacement of tumor
TABLE 1. Clinical Features of BCOR-CCNB3 Sarcomas
Case
Age*/ Sex
Site
Size (mm)
Treatment
1 2 3 4 5
13/M 18/M 14/F 15/M 13/M
Femur (B) Psoas (ST) Pubic ramus (B) Tibia (B) Pelvis (B)
130130 60 9030 30 110 100 100 5030 30 160 160 120
6 7 8 9 10
16/M 16/M 14/M 17/M 11/M
Fibula (B) Quadriceps (ST) Calcaneus (B) Gastrocnemius (ST) Fibula (B)
13070 65 10575 50 10070 70 7065 55 150 largest dimension
I, D, MT; rx 11; rtx V, I, D, E; surgical decompression of spine; rtx; V, A, I V, I, D, E; rtx 11 and met; V, A, I V, I, D, E; rx 11; V, A I V, I, D, E; rtx to 11; whole-lung rtx; V, A, I, D for rec 11; rx rec 11 and lung met; palliative Rtx met; I, C, E V, I, D, E; rx of 11; V, A, I Rx of 11; V, I, D, E E, V, I, A, D; rx 11; whole-lung rtx Rx 11; rtx; V, I, D, E; rx lung met V, I, D, E; rx 11; V, A, I
Mets at Diagnosis
Outcome (mo of FU)
None None Bones Lung Lung
NED (78) DOD (28) DOD (27) NED (13) DOD (110)
None None Lung None None
NED (5) NED (132) NED (189) AWD (61) NED (145)
*Age at diagnosis. 11 indicates primary; A, actinomycin D; AWD, alive with disease; B, bone; C, carboplatin; D, doxorubicin; DOD, died of disease; E, etoposide; F, female; FU, followup; I, ifosfamide; M, male; Met, metastatic disease; MT, methotrexate; NED, alive, no evidence of disease; rec, recurrent; rtx, radiotherapy; rx, resection; ST, soft tissue; V, vincristine.
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1309
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
FIGURE 1. Morphology of BCOR-CCNB3 sarcoma. A, Sagittal PD fat-saturated MRI demonstrates a large soft tissue mass of predominately increased signal intensity within the lateral head of the gastrocnemius muscle compatible with a soft tissue sarcoma. The resection specimen shows a fleshy partially necrotic tumor (case 9). B, MRI-scan of the fibula demonstrates an aggressive, destructive lesion within the proximal fibula with a very extensive, circumferential soft tissue mass on the coronal STIR images. The imaging findings were interpreted as suspicious of Ewing sarcoma. The postoperative resection specimen shows regression of both extraosseous and intraosseous components (case 6). C, Intraosseous tumor with Ewing-like morphology and CD99 positivity (inset). CCNB3 shows strong nuclear positivity (case 4). D, Soft tissue sarcoma with small and round cell morphology and weak CD99 positivity (case 7). E, Primitive sarcoma with plump, short spindle cell morphology (inset) around gaping thin-walled vessels (case 1). F, Other example of BCOR-CCNB3 sarcoma with predominant short spindle cell morphology (case 8). CD99 is negative, and CCNB3 is only moderately positive. MRI indicates magnetic resonance imaging.
1310 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
FIGURE 2. Morphologic change of BCOR-CCNB3 sarcomas during disease progression. A, Primary moderately cellular sarcoma with spindle cell morphology, abundant myxoid matrix, and thin-walled vessels (prechemotherapy resection specimen, case 9). B–D, Several resections of pulmonary metastases showed increasing cellularity with focal osteoblastic differentiation; CCNB3 expression is retained (case 9). E, Primary pelvic BCOR-CCNB3 sarcoma with monomorphic oval cell morphology (case 5, pretreatment biopsy). F and G, Local recurrence after radiotherapy and chemotherapy showed similar morphology juxtaposed to undifferentiated pleomorphic sarcoma-like morphology (F), also seen in the pulmonary metastases (G). CCNB3 expression is retained throughout (inset).
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1311
Puls et al
with hypocellular loose fibrous tissue or in 1 case complete coagulative tumor necrosis (Figs. 3A, B). In 2 cases there were scattered residual neoplastic cells with altered morphology showing elongated bland-appearing nuclei without mitotic activity (Figs. 3C, D). In comparison with the prechemotherapy specimen, Ki-67 proliferative index was markedly reduced in these 2 cases, and CCNB3 immunostaining was weak to negative, although RT-PCR for BCOR-CCNB3 fusion transcript was positive (Figs. 3C, D).
Immunohistochemistry By immunohistochemical screening for nuclear CCNB3 accumulation, this series of BCOR-CCNB3 sarcomas was identified from 134 cases of undifferentiated, monomorphic sarcomas without a known fusion transcript or morphologically typical Ewing sarcomas without molecular confirmation (cases predating routine molecular confirmation of Ewing sarcoma in our archive). In
Am J Surg Pathol
Volume 38, Number 10, October 2014
normal tissues CCNB3 is expressed in a short time span in meiosis, and testis was used as a positive control (Fig. 4A).2 In confirmed BCOR-CCNB3 sarcomas, nuclear positivity was strong and uniform in 8 biopsies of primary tumors (Figs. 1C–E, insets). Two cases showed patchy but nevertheless distinct nuclear CCNB3 immunoreactivity (Fig. 1F; cases 8, 9). On some resection specimens tested, CCNB3 immunoreactivity was also only patchy with some background cytoplasmic staining (Figs. 2G, 4B). One postchemotherapy pulmonary metastasis did not show any CCNB3 immunoreactivity despite positive RT-PCR (Fig. 3D). Nonspecific weak cytoplasmic CCNB3 staining was found in some undifferentiated sarcomas, 4 of which were tested resulting in no BCORCCNB3 fusion transcripts being detected by RT-PCR (Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/PAS/A200, cases 1 to 4). Weak to moderate cytoplasmic staining was also identified in some
FIGURE 3. Chemotherapy response of BCOR-CCNB3 sarcomas. A, Complete response of 1 intraosseously confined tumor (inset, case 4). B, Complete response with focal coagulative tumor cell necrosis as well as granulation tissue replacement (inset, case 8). C, Partial response of pulmonary metastasis to chemotherapy showing a hypocellular lesion with bland spindle cell morphology and patchy, but distinct, CCNB3 expression (inset, case 9). D, Partial response to preoperative chemotherapy (case 1) with markedly reduced cellularity and relatively bland tumor cells, which were negative on CCNB3 immunohistochemistry (inset); RTPCR for BCOR-CCNB3 was positive.
1312 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
FIGURE 4. CCNB3 immunohistochemistry. A, Testes used for positive control showing a layer of positive spermatocytes correlating with strictly controlled expression during spermatogenesis. B, Patchy and moderate CCNB3 expression (resection pulmonary metastasis, case 9). C, Ewing sarcoma cases with variant translocations did not show nuclear expression for CCNB3. D, Alveolar rhabdomyosarcoma showed strong cytoplasmic positivity, although tumor cell nuclei were negative. E, In areas of necrotic Ewing sarcoma, nuclear “pseudopositivity” was observed. This case was EWS-FLI1 positive and BCOR-CCNB3 negative on RT-PCR. Viable areas of this case were negative for CCNB3. F–I, Additional immunohistochemical markers in BCOR-CCNB3 sarcomas. There was cytoplasmic expression of bcl2 (9/10), membranous expression for b-catenin (7/8), membranous expression for CD117 (6/10), and weak expression of bcl-6 (3/5).
r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
1313
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
FIGURE 5. RT-PCR for detection of BCOR-CCNB3 fusion transcripts. A, 140 bp amplicons spanning the fusion of BCOR-CCNB3 for several cases and for metastasis. B, Confirmation by sequencing. C, Survival by primary site for BCOR-CCNB3 sarcoma patients. D, Survival by primary site for patients with classical Ewing sarcoma. E, Survival according to detectable metastases at diagnosis for BCOR-CCNB3 and classical Ewing sarcoma patients. F, Overall survival of BCOR-CCNB3 and classical Ewing sarcoma patients (M, 25 bp DNA ladder; POS, positive control1; EWS, Ewing sarcoma; BC3, BCOR-CCNB3 sarcoma).
cases of classic Ewing sarcoma, alveolar rhabdomyosarcoma, and very focally in some synovial sarcomas, all of which did not show any BCOR-CCNB3 fusion transcripts
1314 | www.ajsp.com
(Figs. 4C–E, Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/PAS/A199). In the confirmed cases of BCOR-CCNB3 sarcomas, staining for r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
BCOR-CCNB3 Sarcoma
additional immunohistochemical markers was performed, and positive fractions were as follows: CD99 (7/10), CD117 (6/10), Ki-67 (30% to 50% in 9 tested), bcl2 (9/10), bcl-6 (weakly positive 3/5), desmin (0/10), CD45 (0/10), AE1/ AE3 (0/10), and membranous/cytoplasmic b-catenin (7/8). Immunophenotypes are shown in Table 2, and representative sections of immunohistochemistry are depicted in Figures 4F–I.
(P = 0.0857; Figs. 5C, D, Table 3). In classical Ewing sarcoma, the presence of metastatic disease at the time of diagnosis was significantly associated with shorter overall survival (P < 0.0001), whereas in BCOR-CCNB3 no association between survival and metastatic disease at diagnosis was apparent (P = 0.3118; Fig. 5E, Table 3).
Reverse Transcription Polymerase Chain Reaction
The recent identification of sarcomas with BCORCCNB3 fusion within the problematic and heterogeneous group of non-Ewing small round cell sarcoma is yet another example of the power of high-throughput technology, which is rapidly expanding the number of recognized translocation-associated sarcomas.6–9 The BCOR-CCNB3 sarcomas have, to date, only been reported in the original series by Pierron et al1 focusing on the genetic and pathogenetic aspects of BCOR-CCNB3 fusion sarcoma. The present study of BCOR-CCNB3 sarcomas diagnosed and treated at a single institution was carried out to estimate relative incidence to Ewing sarcoma, to further characterize the histopathologic features and clinical behavior of this newly identified sarcoma, and to aid further decisions regarding classification. The group of fusion-defined Ewing sarcoma treated by the same multidisciplinary team over the same period served as an ideal control group. The morphologic spectrum of BCOR-CCNB3 sarcomas in this series was quite varied. Most cases were highly cellular and dominated by fairly small cells with angulated nuclei, different from the more uniform round cell morphology of Ewing sarcoma and somewhat resembling so-called “atypical Ewing sarcoma.”4,10,11 A few tumors had a spindle cell morphology with, in areas, a fibrous or myxoid stroma. These findings are in line with those of Pierron et al,1 some of whose cases were initially labeled as “fusiform sarcoma.” As in their series, malignant peripheral nerve sheath tumor and poorly differentiated synovial sarcoma were also among the diagnoses initially considered. One of their cases was thought to be a small cell osteosarcoma, most likely on the basis of focal osteoid formation, which was also observed in 1 of our cases.1 The solid, nonmyxoid, small cell variant of extraskeletal myxoid chondrosarcoma may also mimic BCORCCNB3 sarcoma.12 Spindle cell, sclerosing, and large cell variants of Ewing sarcoma are also recognized.11,13 Another feature in our series was the pronounced pleomorphism seen in local recurrences and metastases, occurring after chemotherapy and/or radiotherapy and simulating undifferentiated pleomorphic sarcoma, a feature not typically encountered in Ewing sarcoma. Awareness of the broad morphologic spectrum of these sarcomas and of the resultant long list of differential diagnoses justifies CCNB3 immunohistochemistry and molecular analysis for BCOR-CCNB3 fusion in all poorly differentiated, difficult-to-classify sarcomas. In our experience, CCNB3 immunohistochemistry for screening and RT-PCR for confirmation appears to be the preferred approach. RT-PCR can be utilized efficiently because the fusion transcript appears to be constant requiring only a
RT-PCR generated 140 bp amplicons, and sequencing confirmed identical fusion points of exon 15 of BCOR to exon 5 of CCNB3 (Fig. 5B) in 9/10 cases with strong nuclear CCNB3 staining by immunohistochemistry. Case 10 showed strong and diffuse nuclear immunohistochemical positivity for CCNB3, but attempted RNA extraction did not yield sufficient mRNA, presumably because of previous decalcification of biopsy material by nitric acid. RT-PCR was negative in 20 cases of other translocation sarcomas (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/PAS/ A199) and in 9 cases of randomly chosen CCNB3-negative, not otherwise specified, small round cell sarcomas (Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/PAS/A200, cases 5 to 13).
Comparison With Classic Ewing Sarcoma With Known Translocations When compared with 242 fusion-positive classical Ewing sarcomas diagnosed at our institution, a higher proportion of BCOR-CCNB3 sarcomas occurred in soft tissue (30% vs. 5.8%, P < 0.05). The median age at the time of diagnosis was identical (15.5 y), and the mean ages at diagnosis for Ewing sarcoma and BCOR-CCNB3 sarcoma were 18 and 15 years, respectively (Table 3). At the time of diagnosis, metastatic disease was present in 40% of BCOR-CCNB3 sarcomas and in 39% of Ewing sarcomas. Since 2005, when the screening for classic Ewing sarcoma translocations EWS-FLI1, EWS-ERG, FUSERG, and later EWS-ETV1 and FUS-FEV had been routinely performed at our institution, the ratio of identified BCOR-CCNB3 sarcomas to classical Ewing sarcomas was 1:27. Follow-up data were available for 234 patients with classical Ewing sarcoma (median follow-up 30.5 mo, range 1 to 175 mo) and all 10 patients with BCOR-CCNB3 sarcoma (median follow-up 69.5 mo, range 5 to 189 mo). Five-year survival of all patients with classical Ewing sarcoma and BCOR-CCNB3 sarcoma was 54% and 75%, respectively, but this difference was not statistically significant (P = 0.3541; Fig. 5F, Table 3). For both entities, longer overall survival was significantly linked to location of the primary tumor in the extremities versus axial skeleton and soft tissues (P < 0.01). The survival difference between BCOR-CCNB3 and Ewing sarcoma within these subgroups was not significant, although there was a trend toward longer survival in patients with BCOR-CCNB3 sarcomas of the extremities compared with Ewing sarcoma in the same location r
2014 Lippincott Williams & Wilkins
DISCUSSION
www.ajsp.com |
1315
Puls et al
Am J Surg Pathol
Volume 38, Number 10, October 2014
TABLE 2. Morphologic and Immunohistochemical Features of BCOR-CCNB3 Sarcomas Case
Tumor Examined
1
Bone primary (bx, rx)
2
Soft tissue primary (bx) Bone primary (bx) Bone primary (bx, rx)
3 4 5
6 7 8 9
Original Diagnosis
Primitive STS, Angulated short spindle cells possibly MPNST
37
Primitive sarcoma, Short spindle cells with overlapping nuclei Ewing-like Ewing-like Small densely packed plasmacytoid cells, sarcoma PAS+ cytoplasmic inclusions Ewing-like Small round blue cell tumor, Ewing-like sarcoma
Bone primary (bx) Recurrent tumor (bx, rx) Metastatic tumor, lung (rx) Bone primary (bx) Soft tissue primary (bx, rx) Bone primary (bx, rx) Soft tissue primary (bx, rx) Metastatic tumor, lung (rx post Ct) Metastatic tumor, lung (rx)
10
Mitoses/ CCNB3 10 hpf IHC (N)
Morphology
Bone primary
Ewing-like sarcoma MPNST Metastatic highgrade STS BCOR-CCNB3 sarcoma Ewing sarcoma
Plump, short, cells, edematous matrix thin-walled vessels Spindle cells, scattered marked pleomorphic cells, fibrotic matrix Undifferentiated pleomorphic sarcoma-like, fibrotic matrix Small round blue cell tumor, Ewing-like
8 10 22 32 20 34 32
CD99 , CD117+, bcl2+, b-cat+ (m), ki-67 50%, bcl-6 (w) Strong, CD99 ,CD117+, ki-67 30% diffuse Strong, CD99+, CD117+, ki-67 diffuse 50% Strong, CD99+, ki-67 40%, CD117 , bcl2+, diffuse b-cat+ (m) Strong, CD99+, CD117 , ki-67 30%, bcl2+, b-cat+ (m) diffuse Strong, CD99+, bcl2+, b-cat+ (m), bcl-6 (w) diffuse Strong, CD99+, ki-67 30%, bcl2+, b-cat+ (m), diffuse Strong, CD99+, bcl2+, b-cat+ (m), diffuse ki-67 40%, CD117+, bcl-6 (w) Strong, CD99+, ki-67 30%, bcl2+, diffuse b-cat-, CD117 Moderate, CD99 , ki-67 30%, CD117 , bcl2+, patchy b-cat+ (m) Strong, CD99+, ki-67 40%, patchy CD117+, bcl2+, b-cat+ (m) Strong, CD99+, ki-67 5%, CD117+ patchy Strong, diffuse
31
High-grade spindle cell sarcoma, NOS Primitive STS, NOS
Round cells with pale cytoplasm, fibrotic matrix Small spindle cells with marked pleomorphism, vacuolated cytoplasm, edematous matrix Small plump elongated cells in edematous, focally myxoid matrix
Metastatic STS
Relatively bland elongated spindle cells
39
Metastatic STS
Variable morphology, synovial-sarcomalike, osteoblastic foci
24
Strong, patchy
Ewing sarcoma
Small round blue cell tumor, Ewing-like
5
Strong diffuse
18 30
Other IHC
CD99+, ki-67 40%, bcl2+, b-cat+ (m), CD117+, bcl-6 (w) CD99+, ki-67 10%, bcl2 , b-cat+ (m), CD117+
RTPCR Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Pos Fail
bx indicates biopsy; C, cytoplasmic positivity; Ct, chemotherapy; m, membranous positivity; MPNST, malignant peripheral nerve sheath tumor; n, nuclear positivity; n/p, not performed; NOS, not otherwise specified, rx, resection; STS, soft tissue sarcoma; w, weak.
single pair of primers. Alternatively, a fluorescence in situ hybridization approach for detection of the X-chromosomal inversion can be used.14 Apart from CCNB3 positivity, the immunohistochemical profile showed broad overlap with that seen in classic Ewing sarcoma, although CD99 staining was generally weaker in most BCOR-CCNB3 sarcomas or, occasionally, completely absent. Despite evidence of activation of the Wnt pathway in expression profiling of BCOR-CCNB3 sarcoma, there was no nuclear staining for b-catenin in our cases.1 BCOR is a transcriptional repressor interacting with bcl-6.15 Interestingly, we observed nuclear bcl-6 staining in 3 of 5 cases tested. Staining was weaker than in lymphoid follicular centers, which were used as positive controls, and does not appear suitable for diagnostic exploitation. In gene expression profiling, there was no evidence of increased transcription of wild-type BCOR mRNA, and therefore nuclear positivity for bcl-6 may indicate that BCOR-CCNB3 fusion protein interacts with bcl-6 in BCOR-CCNB3 sarcomas.1
1316 | www.ajsp.com
BCOR-CCNB3 sarcoma appears to be rare. By screening 594 undifferentiated sarcomas Pierron et al1 identified 24 cases. To estimate incidence and minimize selection bias we compared the ratio of classical Ewing sarcoma with BCOR-CCNB3 sarcoma primarily diagnosed at the Royal Orthopaedic Hospital Birmingham. Although these findings need to be interpreted with caution because of low numbers, the finding of 27 cases of classical Ewing sarcoma for every case of BCOR-CCNB3 sarcoma is also in keeping with a very rare sarcoma. As in the series of Pierron and colleagues there was a predominance of sarcomas originating in the bone and a majority of patients presenting during the second decade. Male predominance was more pronounced in our series (1:9) than in the series of Pierron and colleagues (1:2). This differs markedly from classic Ewing sarcoma both in our control population and other series (1:1.4).4 However, the exact mechanism for this obviously increased male susceptibility for development of this paracentric inversion of the X-chromosome remains unclear. r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 10, October 2014
TABLE 3. Comparison With Classic Ewing Sarcoma
Sex F:M Median age (range) [y] Location (n [%]) Bone Soft tissue Axial (spine, pelvis) Trunk, chest wall Extremity Metastatic disease at diagnosis (%) Ratio (translocationpositive Ewing sarcoma:BCORCCNB3 sarcoma; all cases January 2005 to August 2013) Outcomes Duration of follow-up
Ewing Sarcoma
BCOR-CCNB3 Sarcoma
100:142, 1:1.42 (242 informative cases) 15.5 (3-62)
1:9 (10 informative cases) 15.5 (11-18)
228 (94.2) 14 (5.8) 63 (26.0) 28 (11.6) 151 (62.4) 87 (39) (223 informative cases) n = 190, 1:27.1
7 (70) 3 (30) 3 (30) 0 7 (70) 4 (40) (10 informative cases) n = 7, 1:27.1
1-175 mo, mean 5-189 mo, mean 42 mo, median 79 mo, median 30.5 mo (234 69.5 mo (10 informative cases) informative cases) 5-year survival, 54%, median survival 75%, median survival 72 mo (234 undefined (10 median survival, informative cases) informative cases) all cases 5-year survival, 29,8%, median 33.3%, median median survival, survival 31 mo survival 28 mo axial soft tissues and (64 cases) (3 cases) skeleton (without trunk) 5-year survival, 60.3%, median 100%, median median survival, survival undefined survival undefined location in extremity (142 cases) (7 cases) 5-year survival, 29.3%, median 66.7%, median median survival, survival 26 mo (85 survival, 110 mo mets at time of cases) (4 cases) diagnosis 5-year survival, 69%, median survival 80%, median survival median survival, no undefined (131 cases) undefined (6 cases) mets at time of diagnosis
In our series, overall survival of BCOR-CCNB3 sarcoma (75% at 5 y) was not statistically different from our large series of classic Ewing sarcoma (54% at 5 y). In view of the low number of cases, possible variation in treatment over time, and the retrospective analysis, the findings have to be interpreted with caution. Similar to Ewing sarcoma, patients with axial/pelvic BCOR-CCNB3 sarcomas had a shorter survival than patients with tumors in the extremities or trunk.16,17 Comparison of extra-axially located BCOR-CCNB3 sarcomas with classical Ewing sarcoma showed a trend toward longer survival in BCOR-CCNB3 sarcoma. In Ewing sarcoma, the presence of metastatic disease at diagnosis is an adverse prognostic factor.17 In our series of Ewing and BCOR-CCNB3 sarcomas, a similar proportion of patients presented with metastases. Four patients with BCOR-CCNB3 sarcoma presenting with metastasis, however, are too few for any definite conclusion on prognostic significance. The presence of 1 long-term survivor (>15 y) within this subgroup r
2014 Lippincott Williams & Wilkins
BCOR-CCNB3 Sarcoma
indicates that aggressive treatment, including whole-lung radiation, may be justified. Because new therapeutic approaches are likely to selectively target specific fusion transcripts and downstream pathways, correct identification of oncogenic fusion transcript in Ewing and Ewing-like sarcoma will be vitally important for patient care.18,19 Distinction from other “Ewing-like” small blue cell sarcomas, such as those associated with CIC-DUX4 fusion, appears important. Within the last few years sarcomas with CIC-DUX4 fusions have been increasingly identified in soft tissue.20–25 Molecular and morphologic similarities of CIC-DUX4 and classical Ewing sarcoma have led to speculation that classical Ewing sarcoma and CIC-DUX4 sarcoma may be closely related.22,26 Other authors have pointed out morphologic and clinical differences of CIC-DUX4 sarcoma compared with Ewing sarcoma.20 Compared with BCOR-CCNB3 sarcoma in the present and the previous series, reported cases of CIC-DUX4 sarcoma show a wider age range, exclusive occurrence in soft tissues, and absence of clear male preponderance.1,20,21,24 Its clinical behavior is aggressive, with 7 of 19 and 4 of 4 patients dying of disease within 16 months of diagnosis.20,21 Despite overlapping morphologic and immunohistologic features, distinction of BCOR-CCNB3 sarcoma, CICDUX4 sarcoma, and classical Ewing sarcoma appears to be of prognostic relevance. In conclusion, this study provides a first detailed description of the histologic spectrum, immunohistochemical features, and clinical characteristic of BCORCCNB3 sarcoma. Bearing in mind the morphologic spectrum, we believe that combined CCNB3 immunohistochemistry and RT-PCR analysis should be performed on primitive high-grade bone and soft tissue sarcoma not fitting into any defined entity in terms of morphology, immunoprofile, and genetic aberration, particularly if these occur in childhood and adolescence. ACKNOWLEDGMENTS The authors are indebted to Dr G. Pierron, Institut Curie, Unite´ de Ge´ne´tique Somatique–Centre Hospitalier, Paris, France, for providing control material from BCORCCNB3 cases and Dr Y. Hock, Department of Histopathology, Walsall Healthcare NHS Trust, Walsall, UK, for performing bcl-6 immunohistochemistry. The authors thank Elaine Haywood, Karen Joynes, and Jade Clarke for excellent technical assistance. REFERENCES 1. Pierron G, Tirode F, Lucchesi C, et al. A new subtype of bone sarcoma defined by BCOR-CCNB3 gene fusion. Nat Genet. 2012;44: 461–466. 2. Nguyen TB. Characterization and expression of mammalian cyclin B3, a prepachytene meiotic cyclin. J Biol Chem. 2002;277:41960–41969. 3. Fletcher CDM, Chibon F, Mertens F. Undifferentiated/unclassified sarcomas. In: Fletcher CDM, Bridge JA, Hogendoorn PCW, et al, eds. WHO Classification of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2013:236–238. 4. de Alava E, Lessnick SL, Sorensen PH. Ewing sarcoma. In: Fletcher CDM, Bridge JA, Hogendoorn PCW, et al, eds. WHO Classification
www.ajsp.com |
1317
Puls et al
5.
6. 7.
8. 9. 10. 11.
12.
13.
14. 15.
of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2013:306–309. Williams A, Bartle G, Sumathi VP, et al. Detection of ASPL/TFE3 fusion transcripts and the TFE3 antigen in formalin-fixed, paraffinembedded tissue in a series of 18 cases of alveolar soft part sarcoma: Useful diagnostic tools in cases with unusual histological features. Virchows Arch. 2011;458:291–300. van de Rijn M, Guo X, Sweeney RT, et al. Molecular pathological analysis of sarcomas using paraffin-embedded tissue: current limitations and future possibilities. Histopathology. 2014;64:163–170. Wang L, Motoi T, Khanin R, et al. Identification of a novel, recurrent HEY1-NCOA2 fusion in mesenchymal chondrosarcoma based on a genome-wide screen of exon-level expression data. Genes Chromosom Cancer. 2012;51:127–139. Chmielecki J, Crago AM, Rosenberg M, et al. Whole-exome sequencing identifies a recurrent NAB2-STAT6 fusion in solitary fibrous tumors. Nat Genet. 2013;45:131–132. Robinson DR, Wu Y-M, Kalyana-Sundaram S, et al. Identification of recurrent NAB2-STAT6 gene fusions in solitary fibrous tumor by integrative sequencing. Nat Genet. 2013;45:180–185. Machado I, Noguera R, Mateos EA, et al. The many faces of atypical Ewing’s sarcoma. A true entity mimicking sarcomas, carcinomas and lymphomas. Virchows Arch. 2010;458:281–290. Folpe AL, Goldblum JR, Rubin BP, et al. Morphologic and immunophenotypic diversity in Ewing family tumors: a study of 66 genetically confirmed cases. Am J Surg Pathol. 2005;29: 1025–1033. Meis-Kindblom JM, Bergh P, Gunterberg B, et al. Extraskeletal myxoid chondrosarcoma: a reappraisal of its morphologic spectrum and prognostic factors based on 117 cases. Am J Surg Pathol. 1999; 23:636–650. Llombart-Bosch A, Machado I, Navarro S, et al. Histological heterogeneity of Ewing’s sarcoma/PNET: an immunohistochemical analysis of 415 genetically confirmed cases with clinical support. Virchows Arch. 2009;455:397–411. Antonescu C. Round cell sarcomas beyond Ewing: emerging entities. Histopathology. 2014;64:26–37. Huynh KD, Fischle W, Verdin E, et al. BCoR, a novel corepressor involved in BCL-6 repression. Genes Dev. 2000;14:1810–1823.
1318 | www.ajsp.com
Am J Surg Pathol
Volume 38, Number 10, October 2014
16. Bacci G, Ferrari S, Mercuri M, et al. Multimodal therapy for the treatment of nonmetastatic Ewing sarcoma of pelvis. J Pediatr Hematol Oncol. 2003;25:118–124. 17. Cotterill SJ, Ahrens S, Paulussen M, et al. Prognostic factors in Ewing’s tumor of bone: analysis of 975 patients from the European Intergroup Cooperative Ewing’s Sarcoma Study Group. J Clin Oncol. 2000;18:3108–3114. 18. Stegmaier K, Wong JS, Ross KN, et al. Signature-based small molecule screening identifies cytosine arabinoside as an EWS/FLI modulator in Ewing sarcoma. PLoS Med. 2007;4:e122. 19. Lessnick SL, Ladanyi M. Molecular pathogenesis of Ewing sarcoma: new therapeutic and transcriptional targets. Ann Rev Pathol. 2012;7:145–159. 20. Choi EY, Thomas DG, McHugh JB, et al. Undifferentiated small round cell sarcoma with t(4;19)(q35;q13.1) CIC-DUX4 fusion: a novel highly aggressive soft tissue tumor with distinctive histopathology. Am J Surg Pathol. 2013;37:1379–1386. 21. Italiano A, Sung YS, Zhang L, et al. High prevalence of CIC fusion with double-homeobox (DUX4) transcription factors in EWSR1negative undifferentiated small blue round cell sarcomas. Genes Chromosom Cancer. 2012;51:207–218. 22. Kawamura-Saito M. Fusion between CIC and DUX4 up-regulates PEA3 family genes in Ewing-like sarcomas with t(4;19)(q35;q13) translocation. Hum Mol Genet. 2006;15:2125–2137. 23. Machado I, Cruz J, Lavernia J, et al. Superficial EWSR1-negative undifferentiated small round cell sarcoma with CIC/DUX4 gene fusion: a new variant of Ewing-like tumors with locoregional lymph node metastasis. Virchows Arch. 2013;463:837–842. 24. Graham C, Chilton-MacNeill S, Zielenska M, et al. The CIC-DUX4 fusion transcript is present in a subgroup of pediatric primitive round cell sarcomas. Hum Pathol. 2012;43:180–189. 25. Kajtar B, Tornoczky T, Kalman E, et al. CD99-positive undifferentiated round cell sarcoma diagnosed on fine needle aspiration cytology, later found to harbour a CIC-DUX4 translocation: a recently described entity. Cytopathology. 2014;25:129–132. 26. World Health Organization. Fletcher CDM, Bridge JA, Hogendoorn PCW, et al. WHO Classification of Tumours of Soft Tissue and Bone. Lyon: International Agency for Research on Cancer; 2013.
r
2014 Lippincott Williams & Wilkins
