Unravelling Therapeutic Vulnerabilities in Sarcoma Using Patient-Derived $\textit{Ex Vivo}$ Models

Sarcomas are a heterogeneous group of mesenchymal tumors comprising over 100 histological subtypes. Beyond histopathology, sarcomas are genomically classified into simple karyotype sarcomas, driven by recurrent genetic fusions, and complex karyotype sarcomas, which lack clear oncogenic drivers and are characterized by widespread genomic instability. Despite advances in molecular characterization, clinical management remains dependent mainly on conventional treatments such as surgery, chemotherapy, and radiation therapy, leaving progressive disease with a poor patient prognosis. While a small subset of simple karyotype sarcomas has benefited from the introduction of targeted therapies, these treatments apply to only a minority of patients, and resistance commonly develops within a year, further limiting their clinical durability. There is an urgent need for novel therapeutic strategies to treat patients with sarcoma. However, progress has been hindered by three persistent challenges: the scarcity of well-characterized, patient-derived cell models; the rapid emergence of resistance to the few available targeted therapies; and the absence of actionable vulnerabilities by sequencing. This thesis comprises three manuscripts, each exploring strategies to address one of the key challenges limiting therapeutic advancement in sarcoma. In the first publication, we established and characterized a panel of patient-derived sarcosphere models from multiple sarcoma subtypes and identified drug candidates by functional drug screens. These models preserve the molecular and phenotypic features of the original tumors and serve as a valuable platform for mechanistic and therapeutic investigations. In the second publication, we modelled TRK inhibitor resistance with an LMNA::NTRK1-rearranged soft-tissue sarcoma cell model. We identified an off-target NF2 mutation that conferred resistance by upregulating MAPK and mTOR/AKT pathways. Furthermore, we identified single drugs and drug combinations to overcome drug resistance. We provided a strategy to model and overcome acquired drug resistance. Finally, to identify therapeutic vulnerabilities in complex karyotype sarcomas, we performed whole-genome CRISPR-Cas9 dropout screens, complemented by medium-throughput drug screens. These studies revealed a dependence on protein homeostasis pathways as a key survival mechanism in most sarcomas with complex karyotypes. Our findings therefore suggest that proteostasis-targeted therapies may represent a promising strategy for treating these tumors. Taken together, this work provides novel biological insights and practical tools that help advance the therapeutic landscape for sarcomas.