Use this URL to cite or link to this record in EThOS:
Title: An investigation of volatile organic compounds as biomarkers for gastrointestinal neoplasia
Author: Bond, A.
Awarding Body: University of Liverpool
Current Institution: University of Liverpool
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Early identification of malignancy has the ability to improve long term morbidity and mortality. This is certainly the case for colorectal cancer and hepatocellular carcinoma (HCC), two of the commonest gastrointestinal malignancies. Both diseases are subject to screening programmes. The UK national Bowel Cancer Screening Programme (BCSP) has been shown to significantly improve 5 year cancer survival by identifying colorectal cancer at an earlier stage, but also by identifying advanced, pre-malignant adenomatous disease and removing it. Screening of cirrhotic patients at risk of HCC is a bit more controversial but is recommended by national and international bodies, with some evidence to suggest an improved survival from cancer associated death. Biomarkers are currently employed in the diagnosis and patients selection process for these screening programmes, in particular alpha fetaprotein (AFP) for HCC. The sensitivity and specificity, and thus the valid application of these biomarkers has been brought in question, in the case of alpha feta protein, leading to its removal from screening protocols. Volatile organic compounds have been proposed as biomarkers for various disease processes, including gastrointestinal malignancies. They may therefore have an application in disease screening and/or monitoring. The work presented here explores volatile organic compounds (VOCs) emitted from stool and urine, in order to detect disease specific differences that may be utilised as biomarkers for colorectal cancer and hepatocellular carcinoma. It also explores the driver-passenger model of colorectal cancer and biological plausibility via the detection of volatile organic compounds emitted from cultures of Fusobacterium nucleatum and Campylobacter showae. Finally, it assesses the utility of the stool based tM2-PK assay as a marker of colorectal neoplasia in a novel secondary care cohort. Solid phase micro-extraction of headspace gas followed by gas chromatography mass spectrometry was used to isolate and identify candidate volatile organic compounds. Statistical analysis, including logicistic regression modelling and 10 fold cross validation, were applied to assess biomarker utility. Analysis of VOCs emitted from stool was able to differentiate those with higher risk neoplastic disease with the greatest confidence, including established colorectal cancer. When comparing those with cancer to no neoplasia isopropyl alcohol was significantly more abundant in the colorectal cancer samples (p= < 0.0001, q=0.004), producing an AUROC curve of 0.76. When isopropyl alcohol is combined with butanoic acid, 3-methyl, the AUROC was 0.82, sensitivity 87.9% (95% CI 0.87-0.99) and specificity 84.6% (95% CI 0.65-1.0). Further logisitic regression analysis of VOC presence identified a three VOC panel (isopropyl alcohol, 2-Hexanone and butanoic acid,3-methyl-,ethyl ester) with an AUROC of 0.86: a person being 6 times more likely to have cancer if all 3 VOCs were present in their stool(p= < 0.0001). A number of the VOCs identified as important in those with colorectal neoplasia were also identified in the assessment of Fusobacterium nucleatum and Campylobacter showae, namely butanoic acid based compounds and isopropyl alcohol. VOCs emitted from urine failed to demonstrate any candidate biomarkers for colorectal neoplasia. With regards VOCs emitted from urine as biomarkers for HCC, AUROC comparing all those with and without HCC was 0.76 (Sensitivity 65% [95% CI 0.61-0.69] Specificity 74% [95% CI 0.69-0.78]). When assessing treatment naive HCC patients, 3 compounds were found to have significantly different abundance (p=<0.01), when combined and modelled these VOCs demonstrated a superior AUROC of 0.81 (Sensitivity 77% [95% CI 0.71-0.83], Specificity 75% [95% CI 0.71-0.79]). Of this treatment naive group patients defined by Barcelona Clinic Liver Cancer (BCLC) staging as having early disease and therefore potentially curative, demonstrated an AUROC of 0.82. VOC emitted from stool and urine show a clear ability to act as biomarkers for the diagnosis of colorectal cancer and hepatocellular carcinoma respectively.
Supervisor: Probert, C. ; Pritchard, Mark Sponsor: Not available
Qualification Name: Thesis (M.D.) Qualification Level: Doctoral