Longitudinal genomic and epigenomic changes in glioblastoma brain tumours

Abstract

Glioblastoma is a highly aggressive brain tumour with a poor prognosis and inevitable recurrence following standard treatment. Understanding the molecular basis of treatment resistance and tumour progression is critical to improving therapeutic outcomes. This PhD thesis aimed to explore the genetic and epigenetic evolution of GBM through three phases: optimisation of sequencing pipelines, identification of altered biological pathways under therapeutic pressure, and DNA methylation profiling of recurrent disease. In the first phase, whole-exome and whole-genome sequencing pipelines were optimised for use with challenging clinical material, including FFPE-derived samples. Custom adjustments, including the correction of overlapping read pairs and mitigation of FFPE artefacts, significantly improved variant calling accuracy and tumour mutational burden estimation. The second phase focused on uncovering treatment-associated pathway alterations using paired primary and recurrent GBM samples from 2 cohorts. By tracking changes in variant allele frequency pre- and post-treatment, I identified variants either selected for or against by therapy. Pathway analysis using PathScore revealed several significant biological pathways under selection pressure, notably involving the ERBB signalling family. Disruption of ERBB4 signalling was associated with treatment sensitivity, suggesting that its inhibition may enhance therapeutic efficacy in a subset of patients. The final phase applied genome-wide DNA methylation profiling using Illumina Infinium arrays. Although recurrence-associated changes were subtle at the cohort level, stratification by JARID2-related transcriptional response revealed subtype-specific epigenetic dynamics. A quadrant-based analysis highlighted greater methylation shiPs in Down responders, potentially reflecting adaptive responses to treatment. Altogether, this work provides insight into GBM evolution under therapy, demonstrating how both genetic and epigenetic shiPs contribute to recurrence. The identification of ERBB4 signalling as potentially associated with treatment sensitivity highlights a candidate pathway that warrants further functional validation. Future work, including targeted experimental studies of ERBB4 function, alongside single-cell and spatial profiling, may reveal actionable therapeutic insights and refine strategies to overcome treatment resistance.