Background: Osteosarcoma (OS) is a highly metastasizing bone malignancy despite wide surgical resection of the primary lesion. A liquid biopsy approach to detect residual disease and identify therapeutic targets is still lacking. In this report, we aimed to track the metastasis of OS via extracellular vesicle (EV) RNA profiling in a non-invasive manner.
Methods: We applied RNA sequencing for 10 matched metastatic and primary OS EV samples, including two pairs of cell lines and three pairs of plasma, and compared the expressed mutation, gene expression, fusion transcript, and alternative splicing (AS) between metastatic and primary OS at the transcriptome-wide level. Additional paired tissue/EVs were sequenced and public datasets were used to validate the EV-based metastatic biopsy.
Results: EVs were characterized through size-profiling, immunolabeling, and morphological examination. A drastic increase of mutation burden was observed in metastatic OS versus the non-metastatic counterpart. Hierarchical clustering of the expression profiles differentiated the metastatic EVs from the non-metastatic, with a signature enriched in cell-adhesion signaling and tyrosine kinase pathways. Moreover, 30 cancer-related gene fusions were identified in EV RNA as AS events tend to be more frequently observed in metastatic EVs. Further investigation suggested that over 70% of expressed point mutations from EVs could be validated in paired cell line/EV and tissue/EV analyses, and the expression signature significantly predicted 5-year survivorship of 42 patients from a public dataset.
Conclusion: We have demonstrated a liquid biopsy-based approach for tracking cancer transcriptomic alterations, which is a promising source of prognostic and therapeutic biomarkers for metastatic OS.
Clinical trial registration: NCT03108677.