A genomic approach to identify therapeutic targets in histologic subtypes of soft tissue sarcoma.

10577 Background: Sarcomas are a rare, heterogeneous group of tumors with varying degrees of aggressiveness and responsiveness to therapy. In the advanced setting, chemotherapy results in unsatisfactory long-term outcomes. Novel therapeutic strategies that rationally target specific sarcoma subtypes are desired to improve outcomes and limit chemotherapy resistance. Our research uses a genomic approach to identify unique biological mechanisms and signaling pathways associated with specific sarcoma subtypes so as to guide future rational combinations of standard cytotoxic agents with novel "targeted" therapies. METHODS: Publicly-available gene expression datasets ( GSE6481 , GSE2719 ) were obtained from Gene Expression Omnibus (GEO) representing 141 human sarcoma samples across 12 histologic subtypes. Datasets were combined, adjusted and standardized using ComBat to limit batch effect and assessed using hierarchical clustering (Gene Pattern) with filtering. Gene set enrichment analysis (GSEA) was performed using a two-class discriminator (i.e., synovial sarcoma vs. other sarcomas) to identify biologic pathways unique to specific histologic subtypes. Gene sets with a false discovery rate (FDR) <0.25 were deemed statistically significant as originally described. RESULTS: Hierarchical clustering identified well-defined clusters for synovial sarcoma, GIST, and myxoid/round cell liposarcoma with less organized clustering for the remaining subtypes. GSEA identified 11 enriched gene sets, representing biological pathways of activity, at FDR <0.25 for synovial sarcoma with 3 of 11 gene sets related to Wnt signaling (WNTPATHWAY, PITX2PATHWAY, PS1PATHWAY). Twenty-nine gene sets were significant at FDR <0.25 for myxoid/round cell liposarcoma with 2 gene sets related to mTOR signaling (MTORPATHWAY, IGF1MTORPATHWAY). Additional pathways and subtypes are being analyzed and will be reported. CONCLUSIONS: These preliminary data support the use of a genomic approach to identify pathways associated with specific histologic sarcoma subtypes. This strategy may be used to identify novel agents that can be used alone or in combination with conventional cytotoxic chemotherapy. No significant financial relationships to disclose.

Duke Scholars

Author Richard Francis Riedel Medicine, Medical Oncology