Development of personalized medicine in osteosarcoma by exploiting novel, representative, genetically characterized experimental models

Osteosarcoma (OS) is the most common primary malignant tumor affecting bone. Its frequency is low (3.4 cases per year per million people), but because it primarily affects children and adolescents, its social impact is significant. The survival rate is around 70% but decreases dramatically to 20-30% for patients who are metastatic at diagnosis or do not respond to treatment. Managing OS remains challenging, necessitating a multifaceted approach. Current treatment protocols involve aggressive multidrug therapy and surgical resection, yet survival rates have remained unaltered for the past four decades. Molecularly, OS exhibits elevated levels of chromosomal structural variations, including whole genome duplication, chromotripsis, and kataegis. Somatic copy number alterations and structural variants, with few recurrent point mutations in protein-coding genes (except for TP53 and RB1), contribute to its complex yet low mutational burden. Intra- and inter-tumor heterogeneity in genomic alterations underscore the necessity for personalized, genome-driven therapeutic strategies. Recent research efforts have focused on identifying targeted genetic alterations for new therapeutic interventions. In this study, a sixty-seven-cancer gene panel was utilized through High-Throughput Sequencing to analyze twenty-one osteosarcoma samples, 21 Patient Derived Xenografts (PDXs), and 6 PDX-derived cell lines. Genomic correlations between OS samples, OS-PDX, and OS-PDX-derived cell lines were established and validated at the protein level. A highly aggressive OS cell line was employed to assess the efficacy of 2880 FDA-approved drugs via High Throughput Screening (HTS). Analysis revealed a low mutational burden with notable copy number level structural variations, identifying C-MYC, DDR2, CCNE1, and CDK4 as the most amplified genes. PDXs and PDX-derived cell lines mirrored the original tumors' genomic characteristics. Several compounds exhibited significant cell growth inhibition, prompting further investigation. Seventeen promising compounds are scheduled for D/R curve testing, indicating the potential of PDX models and associated cell lines as reliable preclinical platforms for personalized OS therapy.