PROGNOSTIC MARKERS IN OROPHARYNGEAL CANCERS

Abstract

Introduction: Human papillomavirus (HPV) is changing the prevalence, survival and treatment paradigms in oropharyngeal squamous cell carcinoma (OPSCC). Improved survival of patients with HPV positive compared to HPV negative OPSCC has led to trials of treatment de-escalation. Current HPV detection methods are imprecise, therefore standardised assessment of transcriptionally active HPV in OPSCC is required. Furthermore, the differences in immune characteristics and/or the hypoxia response/effects could explain observed differences in prognosis between HPV positive and negative OPSCC. Rigorous HPV detection and subsequent biomarker evaluation should provide additional information required before introduction of treatment de-escalation in broad patient groupings.Methods: The study cohort was 218 patients with OPSCC who received radiotherapy with curative intent. HPV status was determined on pre-treatment, formalin-fixed paraffin-embedded blocks using: 1) polymerase chain reaction (PCR); 2) in-situ hybridisation (ISH) and 3) immuno-histochemistry (IHC). QuantiGene multiplex assay was designed to detect mRNA of reference sequences of the common high-risk HPV types (16, 18, 33, 35, 45, 52 and 58). HPV detection methods were compared with mRNA quantification. Multimarker IHC of immune cell markers using chromogenic and fluorescent staining was performed, analysed and compared with single marker IHC using automated multispectral image analysis. A validated multiplex IHC method was used for a) chromogenic (CD3, CD4, CD8, and FoxP3) and b) fluorescent (CD8, CD68 and PD1/PD-L1) evaluation in tumour and stroma compartments. Single marker IHC was used to investigate tumour hypoxia markers (HIF-1alpha and CA-IX) in HPV positive and negative OPSCC. Results: p16 IHC and ISH were the most sensitive and specific, respectively, for classifying HPV status. The combination of the three tests had the highest positive/negative predictive values compared with QuantiGene mRNA detection. Multiplex validation showed that, for serial sections up to 6 micro metre apart, there were highly significant correlations (P

Date of Award	1 Aug 2016
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Catharine West (Supervisor), Peter Stern (Supervisor) & Ian Hampson (Supervisor)