Discussion
Ewing sarcoma is a rare childhood tumor, accounting for 1% of all pediatric malignancies, that typically presents in males ages 5 to 13 years.
3
, 10
, 11 EWS is the second most common type of primary bone cancers, arising chiefly from long bones or the axial skeleton.
1
, 2 Extraosseous EWS is less common, and primary intracranial EWS—whether dural or osseous in origin—is exceptionally rare.
2
Although EWS is a type of “small round blue cell tumor,” it is unrelated to medulloblastoma, CNS embryonal tumors, or embryonal tumors with multilayered rosettes, C19MC altered.
1 Because therapy for EWS differs considerably from embryonal CNS tumors, correct diagnosis is critical and typically involves documentation of the characteristic gene rearrangement.
Genetically, EWS is characterized by a translocation of the
EWS gene, located on chromosome 22q12.
12 While translocations involving a number of genes can be identified in EWS, the most common involves the
FLI1 gene located on 11q24, which creates a t(11;22)(q24;q12) translocation. This
FLI1 translocation, present in 90–95% of EWS cases
13, creates the
EWSR1 gene, which is an
EWS-
FLI1 fusion; the resulting protein is a known aberrant transcription factor that interacts with
TP53 and
p21.
12
, 13 Downstream effects of the
EWS-FLI1 fusion gene include dysregulation of cell proliferation, differentiation, apoptosis, angiogenesis, invasion, and metastasis.
14 By utilizing either fluorescence in situ hybridization (FISH) or reverse transcriptase polymerase chain reaction (rtPCR), it is possible to detect translocations diagnostic for EWS, specifically the
EWSR1 rearrangement, with 91–100% sensitivity and specificity.
15 Within our series of four cases, all four patients exhibited the
EWSR1 rearrangement.
Overexpression of CD99, a cell surface glycoprotein and a product of the
MIC2 gene, is seen in essentially all cases of EWS.
16 Positive CD99 immunohistochemical staining on cell membranes is a highly sensitive marker of EWS, although it is not highly specific, being seen in other CNS tumors including lymphoblastic lymphomas, ependymomas, and rhabdomyosarcomas.
17 In contrast with EWS, however, CNS embryonal tumors do not express the
MIC2 gene, and thus will not positively stain for CD99.
18 Of the four cases in our series, tumors from three patients exhibited strongly positive staining for CD99, which further supports the diagnosis of primary intracranial Ewing sarcoma.
Treatment of EWS includes maximal surgical resection with aggressive chemotherapy and focal radiation.
19 In systemic EWS, neoadjuvant chemotherapy has also been shown to improve cytoreduction and achieve local control of tumors prior to surgical resection; it also has the benefit to assess tumor response to chemotherapy.
20 Current first-line chemotherapy for EWS includes vincristine, doxorubicin, and cyclophosphamide, alternating with ifosfamide and etoposide.
21 The completed COG AEWS0031 protocol improved clinical outcomes using interval-compressed chemotherapy; this has now become the standard of care therapy for EWS in addition to local control with maximal surgical resection and focal radiation.
9 EWS usually requires only focal radiation (unless dissemination is present at diagnosis), unlike CNS embryonal tumors, such as medulloblastoma, that receive full craniospinal irradiation plus a boost to the tumor bed.
6
Three of the four patients in our series received initial maximal surgical resection, followed by the either standard of care interval-compressed chemotherapy (Case #2, #3, and #4) or high-dose chemotherapy with autologous stem cell transplant (Case #1) and focal radiation. Two of these patients have had one and 10.5-year disease-free survival, respectively. The third patient pursued alternative therapies including homeopathy, diet, and IV vitamin C in addition to interval-compressed chemotherapy and succumbed to her illness after relapsing with widespread extracranial bony metastasis. The final patient has completed therapy and end of therapy scans demonstrated no evidence of disease. This patient did not undergo gross total resection as it was felt that total resection at the petrous bone was unlikely and the surgical risk was not justified.
Previous case studies of primary EWS involving the central nervous system have been reported.
2
, 5
, 6
, 17 Srivastava et al. described extraosseous primary intracranial EWS occurring in the cavernous sinus.
17 Primary EWS of the petroclival bone was reported by Balasubramanian et al.
2 Navarro et al. detailed a 3-year-old boy with primary intracranial EWS of the tentorium who presented with intracranial hemorrhage.
6 Upon literature review, Navarro et al. found that 41% of 17 patients with intracranial EWS presented with tumoral hemorrhage, and all but one of those tumors were found to have meningeal/dural origin.
6
Salunke et al. examined a series of 10 cases of primary osseous intracranial EWS, six of which occurred in pediatric patients.
5 Four involved the temporal bone, two the frontal bone, two the sphenoid bone, one the occipital bone, and one the ethmoid bone. In only one of 10 cases was primary intracranial EWS found to have metastasized at the time of presentation. The prognosis of primary intracranial EWS was found to be better than that of peripheral EWS, with a 5-year survival rate ranging from 39 to 65% for primary intracranial EWS.
5
Within our case series, none of the four patients presenting with primary intracranial EWS had metastases at time of diagnosis. However, one patient did develop metastases during her third relapse following interval compression chemotherapy, re-irradiation, and high-dose vitamin C therapy. Two patients have shown disease-free survival (one at 1 year, the other at 10.5 years) and one patient is too early in her post-therapy monitoring to assess outcome, although end of therapy scans demonstrate no signs of disease.
Our case series is small with some patients still relatively close to completion of therapy making it difficult to make recommendations for therapy based on this experience, but important considerations for therapy can be gleaned from larger studies of extracranial EWS. As discussed above, large collaborative studies have shown that interval-compressed chemotherapy results in improved survival outcomes and is now standard of care.
9 Less aggressive surgical approaches may be required in some cases of intracranial EWS, as demonstrated in Case #4, where the risk associated with total resection of the petrous bone was not justified. This could result in decreased survival for these patients. A review of outcomes for EWS in the National Cancer Database found a decreased 5-year survival in patients with radiation alone (52.5%) compared to patients treated with surgery alone (77.2%).
22 A further analysis of patients treated in a Brazilian collaborative study also found that patients treated with radiation alone for local control had considerably worse outcomes with only a 17.8% survival at 5 years when treated with radiation alone.
23 These results should be considered when evaluating the potential for complete resection of intracranial lesions and consideration for more aggressive surgery including the risks and benefits of cranioplasty.
With an increased understanding of the molecular diagnosis of CNS tumors, the importance of molecular studies for the definitive diagnosis of tumors is becoming more apparent. This has resulted in new classifications as outlined by the WHO Classification of Tumors of the Central Nervous System.
24 For EWS specifically, an integrated genomic analysis of 323 tumors previously diagnosed as cPNET found 2% of those patients were subsequently molecularly identified as EWS. While this would have a consideration for surgical approaches as noted above, it would also significantly affect chemotherapy choice and potential relapse therapies.
25
In conclusion, Ewing sarcoma typically presents as a PNET most commonly arising from the long bones or axial skeleton. Primary intracranial Ewing sarcoma is extremely rare and can easily be mistaken for CNS embryonal tumors. The cytogenetics, treatment regimen, and disease-free survival vary considerably between primary intracranial EWS and other more commonly seen CNS embryonal tumors. All supratentorial small round blue cell tumors should be tested for the EWSR1 gene rearrangement for accurate diagnosis, to ensure all patients are started on the most appropriate treatment to optimize outcomes.