Clinical evaluation of a functional combinatorial precision medicine platform to predict treatment outcomes and enhance combination therapy design in soft tissue sarcomas

Background Soft tissue sarcomas (STS) are a highly heterogeneous collection of tumors that arise from primitive mesenchymal cells. Survival outcomes for advanced STS patients remain poor, with effective agents limited primarily to cytotoxic chemotherapies, achieving responses below 15% as the standard of care. Although efforts to understand sarcomagenesis have identified potential biomarkers, the development of effective targeted therapies has been hindered by the diversity of STS subtypes and interpatient tumour heterogeneity. Given the uniquely individualized nature of STS, we hypothesized that the application of an ex vivo drug sensitivity platform, Quadratic Phenotypic Optimization Platform (QPOP), in primary STS patient samples can improve identification of effective drug regimens on a per-patient basis. In this study, we evaluated clinical concordance of QPOP to predict treatment outcomes in an STS cohort, and explored QPOP’s ranking function for drug combination discovery. Methods Freshly dissociated tumor samples were treated with a predesigned array of 155 test combinations to rank and compare all possible therapeutic combinations from a 12-drug set comprising STS standard of care, FDA approved drugs and promising investigational drugs. Concordance analysis was performed by comparing QPOP-defined outcomes with clinical outcomes from either QPOP-guided therapies or prior lines of treatment. Efficacy of the most frequently top-ranked drug combination was evaluated in both cell line and patient-derived models, and potential mechanisms were investigated using transcriptomic, ChIP-qPCR and in vivo studies. Results Across a total of 45 patient samples, QPOP demonstrated a total predictive value of 72.4% and an AUCROC of 74%, highlighting its discriminative ability in predicting treatment response in patients. We identified BETi and pazopanib as most frequently top-ranked in 67.8% of all patient samples, outperforming standard of care ifosfamide and doxorubicin. Validation in a panel of established patient lines and in vivo models supported its synergistic interaction, which shows repressed oncogenic MYC and related pathways. Conclusions Work here provides preliminary clinical evidence for QPOP to predict STS treatment outcomes and scientific rationale for an effective combinatorial therapeutic strategy validated in vivo that may offer new therapeutic options for sarcoma patients.