Ewing’s sarcoma family of tumors

Epidemiology

Second only to osteosarcoma among bone tumors that occur in the pediatric population, ES affects approximately 250–500 adolescents and young adults in the United States each year [ ]. The annual incidence is low (0.6 per million) before age 5 but peaks during puberty, with a rate of 5 per million and slight predominance in males (1.3:1). ES is much more common in Caucasians than Asians or Hispanics and is nearly absent in African-Americans or Africans [ ]. Recent genome-wide association studies (GWASs) suggest the higher incidence of ES in people of European descent stems from single-nucleotide polymorphisms (SNPs) in susceptibility loci that increase the length of GGAA motifs, thereby enhancing the tumor-promoting activity of the EWS-FLI1 fusion protein [ , ].

Though possible links between ES and environmental toxins and familial cancer syndromes have been reported, no statistically significant data exists to suggest environmental or drug exposures, Mendelian inheritance, diseases, or events (history to trauma, for example) induce ES. Anecdotal evidence suggests ES can occur at previously radiated sites. However, a large case series reports that only 1.3% of pediatric patients exposed to radiation as part of their original cancer diagnosis will develop a secondary bone sarcoma of any type (osteosarcoma, ES, chondrosarcoma, etc.); of those, the risk of radiation-induced osteosarcoma is 35-fold higher than ES [ ]. There are no known modifiable risk factors and no screening tests available for early detection.

Morphology and pathogenesis

Microscopically, ES consists of densely packed homogeneous small round to oval-shaped cells and most express CD99, a 32-kDa cell surface glycoprotein encoded by the MIC2 gene ( Fig. 42.1A and B ). Alternative diagnoses such as lymphoma/leukemia, rhabdomyosarcoma, medulloblastoma, and neuroblastoma must be considered in the differential diagnosis in patients under age 30, as should small cell lung cancer in older patients. An immunostain panel for muscle (MyoD1: myogenic differentiation 1 and/or myogenin), lymphoid (leukocyte common antigen [CD45] and/or TdT), or neural tissues (neuron-specific enolase) can aid in obtaining the correct diagnosis, and molecular studies using reverse transcription polymerase chain reaction (RT-PCR) or fluorescence in situ hybridization (FISH) can readily exclude other SRCSs (e.g., synovial sarcoma and DSRCT) from the diagnosis.

Though experimental evidence suggests ES is derived from a primitive pluripotent mesenchymal stem cell with neuroendocrine features, the precise cell of origin remains elusive and an active topic of debate. It is further complicated by the lack of spontaneous ES tumors in xenograft models [ ]. Nevertheless, the molecular etiology of ES is well studied and paradigmatic of a class of sarcoma subtypes caused by balanced chromosomal translocations. EWSR1-FLI1, the prototypical translocation that occurs in 85% of ES, results from the apposition of the N-terminal portion of the EWS gene (located at 22q12) with the C-terminal FLI1 gene of the ETS transcription factor family ( Fig. 42.2 ) [ ].

Both genes are, of course, present in normal tissues where they exert physiological roles. The full-length EWS protein encoded by EWSR1 has a C-terminal RNA-binding domain, and gene knock-down of two recently identified zebrafish ES orthologues (ewsr1a and ewsr1b) suggests they act to maintain mitotic integrity and proneural cell survival of the CNS in early embryonic development [ ]. FLI1 is a transcription factor that exerts physiological effects on embryological development, hematopoiesis, and cell growth and differentiation through its sequence-specific DNA-binding domain. Interestingly, neither EWS nor FLI1 overexpression induces ES. Instead, they act in concert to promote malignant transformation [ ].

Less common chimeric pairings include EWSR1-ERG (5%–8% incidence) [ ] and EWSR1-ETV1 [ ], EWSR1-EIAF [ ], and EWSR1-FEV [ ], which each occur in less than 1% of reported cases [ , ]. Rarely, FUS (one of three TET genes, also known as TLS) substitutes for EWSR1 to produce an FUS-ERG-positive ES [ ], and non-ETS pairings have also been reported. Table 42.1 summarizes the classic TET-ETS fusion protein structure linked to ES and highlights variations that exist in the TET, ETS, or both translocation partners. As discussed previously, since EWSR1 and FUS may serve as N-terminal TET translocation partners that join with non-ETS genes in other sarcoma subtypes (DSRCT, myxoid liposarcoma, etc.)—which are distinct from ES, both morphologically and phenotypically—one should avoid designating the growing number of non-ETS containing SRCS as ES even if these tumor types express CD99. Nonetheless, because these non-ETS containing ES-like tumors are too rare to study prospectively as a group (≤25 patients per year in the United States), most will understandably be treated as ES-like variants.

In addition to variation in the ETS and non-ETS C-terminal translocation partners, different in-frame EWS-FLI1 transcripts are possible; all include at least the first seven exons of EWS and exon 9 of FLI1 (containing the DNA-binding domain). The two most common fusions, joining EWS exon 7 to FLI1 exon 6 (type 1) or FLI1 exon 5 (type 2), make up 85% of all EWS-FLI1 fusion proteins. Of note, though the type 1 EWS-FLI1 fusion was historically considered to confer a better prognosis [ , ], as it is a less potent transactivator [ ], this effect on prognosis is no longer observed in the modern chemotherapy era. Though beyond the scope of this chapter, as many as 20% of ESs exhibit complex chromosomal arrangements, including a gain of 1q, loss of 16q, changes in copy number (e.g., trisomy 8 and/or 12), or mutations or deletions of the p53 tumor suppressor gene and the ink4 A gene, respectively. The latter change to p53 and ink4 A, unlike the type 1 EWS-FLI1 fusion, clearly portends a worse prognosis [ ].