University of Edinburgh
Exploring the utilisation of Faecal Immunochemical Testing (FIT) within colorectal cancer diagnostic pathways
Abstract
dc:description.abstractColorectal cancer (CRC) is the third most common cancer globally with nearly 2 million cases diagnosed annually. Despite advances in local and systemic therapies, CRC remains the second leading cause of cancer death worldwide. A key factor in long term CRC survival is stage of disease at diagnosis. To help achieve early diagnosis, there are programmes and pathways in place to detect early asymptomatic cases (screening), expedite urgent investigation for those with concerning symptoms and offer regular surveillance for pre-determined high-risk patients. Bowel Cancer Screening is limited by endoscopy capacity, as such a two-step approach is adopted whereby the eligible population receive a faecal immunochemical test (FIT) and if this is positive, at a level set by each programme, a colonoscopy is performed. For those with symptoms suggestive of CRC, guidelines advise urgent specialist review and/or investigation. The symptoms that trigger these investigations are often vague and are a poor discriminator for significant bowel pathology as they are more commonly found in benign conditions such as irritable bowel syndrome. Consequently, fewer than 3% of all urgent symptomatic referrals are found to have CRC but nearly all undergo invasive investigations. Patients with inherited genetic conditions which predispose to CRC are required to undergo intensive endoscopic surveillance. For example, patients with Lynch Syndrome are at increased risk of cancers including colorectal and undergo biennial colonoscopy from the age of 25 to 75. They therefore could have over 20 colonoscopies in their lifetime. This thesis project aimed to explore how FIT can be best utilised in the diagnostic pathways for high-risk symptomatic patients, in asymptomatic screening and surveillance for patients with Lynch Syndrome who harbour a significant lifetime risk of CRC. First the use of a single FIT in the risk stratification for patients with symptoms suggestive of CRC was investigated in 2,260 patients. FIT performed better than symptoms at identifying those with a CRC or other significant bowel pathology, however the sensitivity for CRC was 84% which whilst in keeping with the literature, would mean that 16% of the cases of CRC would be missed. Thereafter, it was explored if missed cases of CRC by FIT could be reduced by performing a second test a few days later. This double-FIT strategy, completed in 2,637 patients, found that half of the cases of CRC and 30% of all significant bowel pathology that would be missed with one test was detected by the second. Furthermore, having two positive tests increased risk stratification for CRC. Discordance in test results was seen in 16.8% of all double test returners and this was the case both in the presence and absence of significant bowel pathology. Two negative tests yielded a CRC prevalence of 0.17%, meaning this group of patients did not require urgent investigations and could be managed in a watch and wait manor. Based on this study a double-FIT referral pathway was proposed which would see investigations reduce to as much as 31% of referrals. Discordance between FITs was observed in the double-FIT cohort. To study where this variation in results was being generated, a further study was conducted where aside from testing two different stools, the same stool was sampled in multiple areas and one sample kit was tested repeatedly through the analyser. This study suggested that the variance originated from both time between different stools and heterogeneity of sampling within each stool, supporting the hypothesis that multiple testing in FIT is beneficial. Further risk stratification of symptomatic patients was undertaken to evaluate if FIT can be improved by addition of other patient factors, whilst keeping the resulting workload manageable. Univariant and subsequent multivariant analysis of single FIT was used to create a scoring system that could provide a 99.5% sensitivity for CRC but at a positivity rate of 44.6%. This was not superior to double testing alone where the positivity rate was only 31.0%. No further factors where significantly associated with CRC than using two FITs. With the double-FIT study showing promise of a new way of processing symptomatic referrals, and concomitantly guidelines advocating FIT directed symptomatic pathways, assessment was made of those who do not return a FIT. In our study this accounted for 11.9% of all referrals. These patients were more likely to be male, younger in age and from lower socioeconomic backgrounds. These are all features of already disadvantaged groups regarding early colorectal cancer diagnosis. Added to that, the prevalence of CRC was not significantly different between the returners and non-returners, meaning a healthcare system that only considers FIT returners would create inequality to investigative services and disadvantage further a cohort of patients already at risk for delayed CRC diagnosis. Next, it was investigated if FIT could be of benefit to patients with Lynch syndrome. By acting as a non-invasive adjunct, the proposed benefit would be to reduce the number of colonoscopies required during surveillance. Tested in 78 patients who returned at least one FIT, the sensitivity for advanced and non-advanced adenomas (the target of surveillance) was only 66.7% and 22.2% respectively. The miss rate, and accelerated CRC formation potential of the polyps means that there was not the evidence to propose using FIT in Lynch syndrome surveillance. It has been reported that cases of CRC in younger patients (under the age of 50 years) is increasing. The incidence of this early onset CRC (EOCRC) was investigated regionally in NHS Lothian and then compared with Scotland as a whole, England and Wales. Between 2002 and 2019 there has been a constant significant year on year increase in the incidence of EOCRC in the population of NHS Lothian. This was being driven by males and those aged 30-39. Similar patterns of increase were observed in Scotland, England, and Wales. When diagnosed EOCRC was at a worse stage with unfavourable histopathological features. FIT shows promise as an objective way of identifying the minority of patients under the age of 50 with significant bowel disease from the vast majority who have self-resolving symptoms. In the asymptomatic screening population of NHS Lothian, since 2017 FIT is used at a higher threshold with the aim of diagnosing CRC early before it becomes symptomatic. The results of 316,583 screening FITs performed in the region were reviewed along with CRC diagnosis within 2 years of screening. 289/546 (52.9%) of CRC diagnosed in that timeframe following bowel screening had a 'positive' bowels screening result (³80µg Hb/g). Of the 257 CRCs with a 'negative' bowel screening, 121 (47.1%) had detectable blood in their test result. CRCs detected at screening had earlier stage disease and better two-year survival. The workload implications of lowering the threshold for screening colonoscopy against the extra CRCs that could be detected was calculated. To lower the threshold to 30µg Hb/g, there would need to be an extra 48 colonoscopies performed each week for a relative reduction in missed CRC of 28.0%. It was shown that the CRC prevalence in the screening group 30-79µg Hb/g was greater than in the symptomatic double-FIT group with two tests led than 10µg Hb/g. Therefore, a better utilisation of resource would be to lower the bowel screening threshold rather than investigate the double negative symptomatic referrals. This would be resource neutral within NHS Lothian. The majority of patients with a raised FIT do not have CRC. Whilst FIT is marketed as being specific for lower gastrointestinal bleeding, a tumour within the upper gastrointestinal (UGI) tract can also result in bleeding into the digestive tract which would ultimately end up in the stool. The incidence of UGI cancers in FIT positive patients was therefore compared to see if a positive FIT with no CRC was a marker for possible UGI malignancy. The double-FIT pathway has been implemented into clinical practice in NHS Lothian since November 2021. The performance of double-FIT in the first year of real-world practice was reviewed. The test performed as expected from the previous study. It was able to risk stratify well for CRC and reduce the overall number of patients requiring invasive investigation whilst maintain an expected number of CRC diagnoses. In summary, the utilisation of FIT in the diagnostic pathway of patients with symptoms suspected of CRC has shown benefits of double testing over single. The double-FIT pathway was created based on these findings and has performed as anticipated throughout the first year of real-world clinical practice. FIT directed pathways are superior to symptoms alone, but strategies must be put in place for non-returners or else inequality to CRC diagnosis will be created. Variation in repeated FIT results can come from sampling areas of the same stool and between stool motions. The evidence generated from studying FIT in Lynch syndrome would not support the use of FIT in surveillance as a substitute or temporising measure to colonoscopy. CRC incidence is increasing in those under 50 across Scotland, England and Wales, it is a challenge in primary care to differentiate the small numbers of people under 50 who have CRC from the majority with transient symptoms, FIT can help objectively with risk assessment. Bowel screening programmes are limited by their endoscopic capacity; the detection of more CRCs earlier can be made by lowering the screening FIT thresholds. The extra workload this creates can be offset by savings in symptomatic double negative FIT patients.
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Adam Daniel, Gerrard
- Advisors dc:contributor.advisor
-
- Din, Farhat
- Dunlop, Malcolm
- Theodoratou, Evropi
Subjects
dc:subject × 6Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- https://doi.org/10.7488/era/7968