An investigation into the impact of screening on tumour and host determinants of outcome in colorectal cancer

Mansouri, David (2016) An investigation into the impact of screening on tumour and host determinants of outcome in colorectal cancer. PhD thesis, University of Glasgow.

Abstract

Colorectal cancer is the third most common cancer and the second most common cause of cancer death in the UK. Outcome is directly related to stage at diagnosis with over 90% of patients with Stage I disease surviving their disease to 5 years compared to less than 10% of those with Stage IV disease. Symptoms for colorectal cancer can be non-specific, particularly when the disease is at its early stage, and hence screening has been introduced. Population screening in the UK, using faecal occult blood testing (FOBt) has been introduced over the past 10 years following several major randomized control trials and a Cochrane review that has shown improved cancer specific mortality in the region of 15% in those individuals invited. This has been attributed to the detection of early stage disease with around 50% of all tumours detected through screening being Stage I.

However, it has previously been shown that there are additional tumour and host prognostic factors outside of stage that can determine outcome. For example, the presence of venous invasion and the presence of an elevated host systemic inflammatory response have been associated with poorer cancer specific survival. These additional factors have not previously been studied within the context of a population screening programme or indeed within early stage disease. Moreover, the FOBt screening programme itself is not without its pitfalls. Uptake of the test is below that of other established cancer screening programmes and it is recognised that repeated screening rounds are required to achieve an acceptable sensitivity of the test.

This thesis sought to examine the first round of the Scottish Bowel Screening Programme within the West of Scotland and assess its effect on tumour and host determinants of outcome. In Chapter 1 an overview of colorectal cancer and current determinants of outcome is provided. In addition, colorectal cancer screening is explored in detail including the evidence behind the current screening programme. Chapter 2 presents original data, utilising population databases, examining the changes in mode, site and stage of presentation across the West of Scotland that have accompanied the introduction of the national screening programme. It identifies that within non-metastatic disease there has been a shift towards a higher proportion of Stage I disease being present following screening introduction. Chapter 3 presents a detailed examination of the first round of screening in NHS Greater Glasgow & Clyde (NHS GG&C) emphasising the importance of the impact of deprivation throughout the screening programme. For example, deprived patients were less likely to take part, more likely to test positive, less likely to proceed to colonscopy following a positive test and less likely to have cancer detected at colonoscopy following a positive test.

Chapters 4, 5 and 6 utilise an original dataset of over 4000 patients who underwent colonoscopy following a positive test in the first round of screening in NHS GG&C generated through work from Chapter 3. Firstly, in Chapter 4, a theoretical model proposing a flexible sigmoidoscopy as a first line test, rather than a colonoscopy, is examined. It found a missed cancer rate of 17% and that around a third would require a completion colonoscopy, concluding that this would not be a desirable change to the current screening algorithm. Chapter 5 then examines the importance of potentially chemopreventative medications such as statins and aspirin, on the risk of neoplasia at colonoscopy, determining that patients on such medications did indeed have lower rates of neoplasia, significant neoplasia and cancer than those not on them. Chapter 6 then looks at symptoms in this population, identifying that around 40% had at least one bowel symptom however that these correlated poorly with the risk of significant neoplasia at colonoscopy. Chapter 7 explores outcomes in those who were invited but did not have a screen-detected cancer in order to examine the incidence of interval cancers (colorectal cancer within 2 years of a negative FOBt) and cancers in non-responders. Overall it identified a 30% interval cancer rate. The chapter then explores differences in tumour and host factors between screen-detected and non screen-detected disease reporting that stage for stage, patients with non screen-detected disease had higher rates of systemic inflammation. Furthermore it characterises the similarity between interval and non-responder tumours suggesting that rather than representing biologically more aggressive tumours, interval cancers arise due to limitations of the test itself. Chapter 8 presents long-term outcomes in patients who have undergone a resection for Stage I disease prior to the introduction of screening. The results report an excellent 5-year cancer specific survival of 95% however an overall survival of 76%. It identifies the presence of an elevated pre-operative host inflammatory response as being associated with a worse overall outcome.

Tissue work exploring the local immune-cell microenvironment of both early stage and pre-malignant disease is the focus for Chapters 9 and 10. This characterisation of immune cell infiltrate identifies similar rates of peritumoural inflammation between T1 and T2 disease and validates a previously published automated scoring system. When exploring local inflammation within premalignant polyps there appears to be a change from low-grade to high-grade dysplasia signifying a specific response to early disease progression suggesting host immunosurveillance. Chapter 11 summarises the main findings of the thesis and presents future directions.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Colorectal cancer, screening, prognostic factors, early-stage disease, host inflammatory response.
• Subjects: R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine; R Medicine > RD Surgery
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
• Supervisor's Name: McMillan, Professor Donald and Paul, Professor Horgan
• Date of Award: 2016
• Depositing User: Mr David Mansouri
• Unique ID: glathesis:2016-7195
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 12 May 2016 08:53
• Last Modified: 23 May 2016 10:52
• URI: https://theses.gla.ac.uk/id/eprint/7195