The use of salivary biomarkers in the detection of oral squamous cell carcinoma

Matthews, April The use of salivary biomarkers in the detection of oral squamous cell carcinoma. Doctor of Medicine thesis, University of Liverpool.

Text MatthewsApr_May2015_2025699.pdf - Unspecified Available under License Creative Commons Attribution. Download (4MB)

Abstract

Background Oral squamous cell carcinoma (OSCC) is the 15th most common cancer worldwide but has poor five year survival (50%). Late stage presentation and limitations of early diagnostic techniques are persistent clinical problems. Sixty percent of patients present with advanced stage disease and with the attendant increase in mortality, morbidity and risk of recurrent disease it is particularly burdensome for both patients and health economies. Early diagnosis and treatment of OSCC improves prognosis. There is an opportunity to diagnose OSCC early in patients with oral epithelial dysplasia however currently there is no way of accurately predicting which lesions will undergo malignant transformation. Aberrant methylation of tumour suppressor genes plays a significant role in the biology of early cancer and is detectable in both tumour and saliva. Saliva is a non-invasive method of longitudinal sampling and has potential as a tumour surrogate in disease surveillance programmes. This study aims to compare rates of methylation of a panel of genes in OSCC patients and a normal cohort to establish a threshold by which we could determine future disease testing in a dysplastic population. Methods Saliva samples were collected from 219 individuals from three diagnostic groups: Normal (defined as no oral malignant or premalignant disease) n=97, OSCC n=62 and dysplasia n=60. For statistical analysis the dysplasia cohort was sub-divided into lesions of low and high risk of malignant transformation based on the histological diagnosis of the index lesion. DNA was extracted and bisulphite treated from 258 saliva samples before duplex quantitative methylation specific PCR (qMSP) assays were performed on all samples to detect the frequency of methylation in saliva of a panel of genes. The five target genes (ADAMTS9, CCNA1, CYGB, P16, TMEFF2) were selected using a candidate approach on the basis of tumour specificity from studies on tumour/normal matched tissue pairs. Clinicopathological data was correlated with the qMSP data and analysed using SPSS v.21 statistical software to look for associations with tumour and survival characteristics. Results Only 3/97 individuals from the control normal cohort had saliva samples with detectable methylation above the analytical sensitivity of the P16 assay. Methylation of the remaining target genes (ADAMTS9, CCNA1, CYGB, TMEFF2) was not detected in normal saliva at levels above the analytical sensitivity of the qMSP assays. The most significant finding in this study was that methylation of four of the target genes (CCNA1, CYGB, P16, TMEFF2) in saliva, individually and when considered as a panel, was significantly associated with OSCC and as such could aid discrimination between malignant disease and normal saliva samples. Methylation of at least one gene in the panel was discovered in 29/67 of the binned OSCC saliva samples but only 3/97 of normal samples (Fisher’s exact p=0.001). Furthermore methylation of the gene panel is associated with high risk lesions when detected in saliva of patients with premalignant lesions (Fisher’s exact p=0.03). Conclusions This exploratory data supports the utility of duplex qMSP as a detection method for methylation markers in saliva. The detection of methylation of this gene panel in saliva is significantly more associated with oral malignancy and high risk premalignant lesions than normal and low risk disease. This implies saliva may have merit as a surrogate tissue in an adjunctive role to clinical assessment and biopsy. The assays are specific but have limited sensitivity. However with further work, inclusive of additional genes, this methodology may identify predictive biomarkers that can be introduced into a trial surveillance of premalignant lesions.

Item Type:	Thesis (Doctor of Medicine)
Additional Information:	Date: 2015-05-01 (completed)
Subjects:	?? RD ?? ?? RK ??
Depositing User:	Symplectic Admin
Date Deposited:	14 Jan 2016 16:13
Last Modified:	17 Dec 2022 01:06
DOI:	10.17638/02025699
URI:	https://livrepository.liverpool.ac.uk/id/eprint/2025699