Association of chromosome complexity with age and metastasis in synovial sarcomas: Validation of expression and genomic prognostic signatures.

Abstract

9536 Background: Synovial sarcoma (SS) is one of the rare sarcomas that occurs in adolescents as well as in adults. Nevertheless, metastasis occurs with a lowest frequency in the former, 10 vs 50% respectively. In almost all cases, a characteristic translocation t(X;18)(p11.2;q11.2) exists and the importance of this translocation in SS oncogenesis is well established, but the genetic basis of SS metastasis is still poorly understood. We recently published expression (CINSARC) and genomic (GI) signatures related to chromosome integrity control that predict outcome in undifferentiated sarcomas and GIST and ask whether these signatures could also predict outcome in SS. Methods: To asses this issue we selected in the European sarcoma database CONTICABASE 92 primary untreated SS for expression and genomic profiling. Results: As demonstrated by metastasis-free survival, CINSARC and GI have strong and independent prognostic values (p = 1.6 x 10-5 and p = 4 x 10-6, respectively). Comparing expression profiles of tumors with or without metastasis in a training series of 58 SS, a 52-genes signature was identified and validated in an independent series of 34 SS. Fourteen of these genes are common with CINSARC and these 52 genes are involved the same pathways than CINSARC, mitosis checkpoints and chromosome integrity. Comparing genomic profiles of adult versus pediatric SS we show that in both situations metastasis is associated to genome complexity and that the adult genome is highly more frequently rearranged. In line with this, the pediatric good-prognosis patients, according to GI, do not develop metastasis. Conclusions: Results clearly indicate that SS metastasis development is strongly associated with chromosome complexity and that CINSARC and GI are powerful prognostic factors. Data also mean that the metastasis frequency difference between adult and children likely is associated to the genome instability which is highly more frequent in adults. Finally, among these tree signatures, GI is potentially the best overall and clearly the most clinically relevant considering that CGH from FFPE samples is already used in the daily practice of pathology.