(P 25) IDENTIFICATION AND VALIDATION OF NON-GENETICALLY ENCODED VULNERABILITIES IN EWING SARCOMA

Objective: Prognosis for relapsed or metastatic Ewing sarcoma (EWS) remains poor with few actionable targets. We previously described a transcriptome-based systems biology approach, OncoTarget for identifying targetable Master Regulator (MR) proteins representing critical non-genetically encoded vulnerabilities from bulk RNA-seq across pediatric sarcomas. To better understand regulatory interactions and therapeutic vulnerabilities in EWS, we applied regulatory network-based analysis of EWS single-cell RNA-seq, hypothesizing that identifying and targeting cell state-specific MRs will prioritize effective and durable drug combinations.

Methods: A EWS-specific gene regulatory network was constructed from snRNA-seq of 6 EWS tumors, regulatory protein activity computed, and protein activity-based tumor cell states identified. OncoTarget analysis identified the cell state-specific activity of directly targetable MRs. Functional validation included ex vivo drug screens and in vivo testing in 2 EWS patient-derived xenograft (PDX) models (TP53 and STAG2 mutant). Anti-tumor activity and progression-free survival (PFS) were assessed.

Results: OncoTarget analysis identified histone deacetylases (HDACs), notably HDAC6, 9 and 3 (in rank order) as top significantly activated targets (p < 10-5) across tumor cell states. Ex vivo screens showed synergistic activity of citarinostat (selective HDAC6 inhibitor) combined with vorinostat (pan-HDAC inhibitor). In vivo assessment of citarinostat with vorinostat in 2 EWS PDXs showed statistically significant differences in tumor volume (p = 0.005, EWS-83033; p = 0.02, EWS-83311) and PFS (p = 0.02, EWS-83033; p = 0.04, EWS-83311) compared to vehicle control. Combination therapy also showed improved efficacy in terms of best response to treatment compared to vehicle control in both models. (p = 0.003). No significant effect on tumor volume or PFS with HDAC inhibitor monotherapy was observed.

Conclusion: Single-cell regulatory network analysis identified HDACs as targetable vulnerabilities in EWS. These findings highlight this approach’s ability to identify non-genetically encoded therapeutic targets in genomically “silent” tumors and provides rationale for combinatorial strategies targeting HDACs in EWS.