Promiscuous genes involved in recurrent chromosomal translocations in soft tissue tumours.

Soft tissue tumours represent a heterogeneous group of mesenchymal lesions and their classification continues to evolve as a result of incorporating advances in cytogenetic and molecular techniques. In the last decade, traditional diagnostic approaches were supplemented with a significant number of reliable molecular diagnostic tools, detecting tumour type specific genetic alterations. Additionally, the successful application of some of these techniques to formalin fixed, paraffin embedded tissue enabled a broader range of clinical material to be subjected to molecular analysis. However, despite all these remarkable advances, the realisation that some of the genetic abnormalities are not fully histotype specific and that certain gene aberrations can be shared among different sarcoma types, otherwise completely unrelated clinically or immunophenotypically, has introduced some drawbacks in surgical pathology practice. One such common example is the presence of EWSR1 gene rearrangements by fluorescence in situ hybridisation (FISH), a test now preferred over the elaborate RT-PCR testing, in a variety of benign and highly malignant soft tissue tumours, in addition to a subset of carcinomas. Furthermore, the presence of identical gene fusions in completely different sarcoma types (i.e., EWSR1-ATF1, EWSR1-CREB1) or in non-mesenchymal malignancies (epithelial or haematological) has raised skepticism as to their diagnostic utility, and their lack of specificity has been compared to the limitations of other ancillary techniques, in particular immunohistochemistry. This review catalogues the main groups of genes that behave in a promiscuous manner within recurrent fusion events in soft tissue tumours. Although we acknowledge that the present molecular classification of soft tissue tumours is much more complex than two decades ago, when EWSR1 gene rearrangements had been described as the hallmark of Ewing sarcoma, we make the strong argument that with very few exceptions, the prevalence of fusion transcripts in most sarcomas is such that they come to define these entities and can be used as highly specific molecular diagnostic markers in the right clinical and pathological context.