Cell-Free DNA Sequencing Captures Tumor Dynamics, Microenvironment Interactions, and Treatment Pressures in Glioblastoma.

BACKGROUND: Glioblastoma (GBM) is a highly aggressive brain tumor for which cell-free DNA (cfDNA) has shown promise as a minimally invasive biomarker, yet the biological processes governing cfDNA release and composition in GBM remain incompletely understood. METHODS: We investigated cfDNA release dynamics, fragmentation patterns, variant allele frequencies (VAF), and copy number profiles under controlled experimental conditions using patient-derived GBM cultures. RESULTS: Longitudinal sampling of conditioned media from monocultures revealed progressive increases in cfDNA yield that correlated more strongly with viable cell numbers than with cell death, suggesting that cfDNA production in these models is not solely driven by apoptosis. In co-culture experiments combining GBM cells with normal human astrocytes (NHA), distinct tumor- and astrocyte-specific variants enabled deconvolution of mixed-cell populations, and cfDNA composition shifted over time, consistent with increasing astrocyte death under competitive co-culture conditions. Temozolomide (TMZ) treatment altered cfDNA release dynamics, shifting the dominant source from viable cells to cell-death-associated pathways, accompanied by increased cfDNA yield, nucleosomal fragmentation, and evidence of reduced variant diversity under therapeutic pressure. Tumor-derived cfDNA was also detected and deconvolved from plasma in GBM orthotopic xenograft models, with copy number profiles recapitulating those of the parental tumor cells. CONCLUSION: These findings suggest that cfDNA composition is shaped by tumor proliferation, microenvironmental context, and therapeutic stress, establishing a preclinical foundation for interpreting cfDNA-based liquid biopsy signals in GBM.