University of Glasgow
Prediction of local relapse of high-risk oropharyngeal squamous cell carcinoma following radical (chemo)-irradiation using a multi-omics approach
Abstract
dc:description.abstractBackground Oropharyngeal cancer (OPSCC) incidence has risen dramatically in Western countries over the last 3 decades. The mainstay of treatment for locally advanced disease is (chemo-) radiotherapy (ChT-RT), but despite advances in treatment planning and delivery, approximately two thirds of patients are left with at least moderate long-term toxicity. Relapse rates remain suboptimal even with maximal non-surgical management, with survival closely correlated with radiatioten response in the treated tumour volumes. Tobaccosmoking associated disease and lack of association with the human papillomavirus (HPV) are long-established risk factors for radiation resistance, however mechanisms of resistance in these high-risk groups are poorly understood. Methods Clinical, radiomic, bulk RNAseq and multiplex immunofluorescence histochemistry datasets were used to explore potential predictors of relapse after radical ChT-RT. Data was drawn from a cohort of 86 patients treated at the Beatson West of Scotland Cancer Centre. Results Tumour shrinkage during RT appeared to be predictive of long-term control, although there were limited radiomic textural differences between relapse and response tumours. Comparably, the transcriptomes of relapse and response tumours were similar, with more variation noted between HPV-positive (HPV+ve) and HPV-negative (HPV-ve) tumours. In HPV+ve subgroup analysis relapsing tumours displayed a phenotype more in keeping with HPV-ve tumours: partial epithelial-mesenchymal transition enrichment, higher percentage of fibroblasts and lower lymphocytes. High-stroma, low lymphocyte tumours were associated with poorer overall survival, regardless of HPV status. Discussion This study demonstrated early markers of radiation resistance in a homogeneous cohort of high-risk OPSCC. Tumour volume changes and tumour microenvironment analysis could be easily incorporated into the current clinical treatment pathway.
Degree
thesis:*- Level dc:type.qualificationlevel
- PhD
- Grantor dc:publisher.institution
- University of Glasgow
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Mactier, Karen Elizabeth
Subjects
dc:subject × 2Identifiers
dc:identifier.*- DOI dc:identifier.doi
- https://doi.org/10.5525/gla.thesis.85956