Bailey, James Alexander
(2024)
Optimisation of Faecal Immunochemical Test (FIT) for symptomatic colorectal cancer diagnosis.
PhD thesis, University of Nottingham.
Abstract
Colorectal cancer (CRC) is one of the most common cancers in the UK and a leading cause of cancer death. Despite concerted efforts to detect more cases at an earlier stage in asymptomatic patients, most diagnoses are still made in patients with symptoms of the disease. The clinical features which typically prompt investigation for suspected CRC include a persisting change in bowel habit, iron deficiency anaemia, weight loss, presence of an abdominal mass or pain, and rectal bleeding or a rectal mass. The Two-Week-Wait (2WW) pathway was introduced as part of the National Health Service (NHS) Cancer Plan in 2000 to streamline the investigation and management of those at highest risk of cancer, and therefore improve cancer-related mortality. Referral on an urgent 2WW pathway is determined by age and symptom-based criteria set out by the National Institute for Health and Care Excellence (NICE) national guidelines. Unfortunately, the clinical features described are often vague and non-specific, and are more likely to be associated with benign disease than CRC.
The Faecal Immunochemical Test (FIT) is an investigation that identifies microscopic amounts of human haemoglobin in a stool sample which can indicate the presence of CRC. FIT has become the test of choice for the Bowel Cancer Screening Programme (BCSP) and has been endorsed by NICE guidelines to guide urgent secondary care referral in patients with “low-risk symptoms”. Recently, FIT has been shown to accurately stratify risk of CRC across all symptom groups and can therefore be used to facilitate prioritisation of patients for invasive investigations.
This thesis seeks to explore how FIT can be optimally utilised to detect CRC in symptomatic patients within a 2WW pathway. Chapter 1 introduces the topic and gives context to the clinical pathways in Nottingham, where FIT has been incorporated since 2017.
The first study presented in this thesis in chapter 2 is an observational study reporting a service evaluation of the Rapid Colorectal Cancer Diagnosis (RCCD) pathway in Nottingham. Two years after the inception of the RCCD, FIT was shown to accurately stratify risk of CRC with excellent “rule out” performance at the lowest threshold. CRC detection in patients with a FIT less than 4 µg Hb / g faeces was 0.1%, and less than 0.3% for those with a FIT below 20 µg Hb / g faeces. Furthermore, patients with a faecal haemoglobin (f-Hb) over 100 µg Hb / g faeces had a 20.7% risk of CRC, corroborating the value of FIT to identify patients who require urgent investigation.
The association between iron-deficiency anaemia and CRC is well established, but other blood tests can also identify those at risk of cancer. In chapter 3, a retrospective review of 2WW referrals between August 2014 and August 2017 examines the presence of thrombocytosis as an independent risk factor for CRC. A multivariate logistic regression analysis showed that patients with thrombocytosis were significantly more likely to have CRC than those with normal platelet values (OR 2.62, 95% CI 1.6-4.3). CRC diagnosis was significantly higher in males with thrombocytosis (16.1% vs 7.9%, χ2 4.62, p=0.032) and females (10.3% vs 2.9%, χ2 19.41, p<0.001), confirming the stratification value of thrombocytosis in a 2WW population.
In chapter 4, an observational study of a 2WW population in Nottingham is presented, evaluating the value of symptomatology in the assessment of symptomatic patients. 1784 patients were included in this analysis, with 181 CRC detected. CRC was diagnosed in 3.5% (24/684) with CIBH compared to 8.1% (6/74) with both CIBH and IDA. No individual or combination of referring clinical features were associated with an increased diagnosis of CRC (χ2 8.03, p=0.155). 3 patients with negative FIT results (<4 μg Hb / g faeces) were diagnosed with CRC (3/1027, 0.3%). The highest proportion of cancers detected was in the ≥100 μg Hb / g faeces group (55/181, 30.4%). In the multivariate model presented, FIT outperformed age, gender and all symptoms prompting referral. FIT has greater stratification value than any referral symptoms and demonstrated value in patients with IDA.
Chapter 5 evaluates whether there are sociodemographic variations in the uptake of FIT when used in a primary care symptomatic pathway for CRC. A retrospective study of 38920 patients referred from primary care over a 4-year period was completed, with multivariate regression analysis performed to identify disparities in the age, sex, ethnicity and socioeconomic status of those referred for the investigation of CRC. Males accounted for 44% of the study population and had a significantly lower FIT return on multivariate analysis (OR 1.11, 95% CI 1.03-1.19). FIT return was significantly higher in patients ≥65 years compared to those aged 18-64 years (adjusted OR 0.78, 95% CI 0.72-0.83). The multivariate model showed the most socially deprived patients had more than double the rate of unreturned FIT compared to the least deprived (OR 2.20, 95% CI 1.99-2.43). Patients from Asian (OR 1.82, 95% CI 1.58-2.10), Black (OR 1.21, 95% CI 0.98-1.49) and Mixed/Other ethnic groups (OR 1.29, 95% CI 1.05-1.59) were also more likely to not return their FIT kits compared to the White ethnic group after adjustment. This confirmed that FIT return varies by gender, age, ethnicity, and socioeconomic deprivation.
Chapter 6 is an examination of whether repeat FIT samples confer added diagnostic value compared to a single sample. A prospective clinical study evaluated 44 patients recently diagnosed with CRC by serially collecting FIT samples over a 4-week period. 4/44 (9.1%) of the first samples returned were below the local 20 µg Hb / g faeces threshold. Of the 4 who returned a falsely low FIT at this threshold, none of them returned a second sample which was below this positivity threshold.
Chapter 7 constitutes a summary discussion of the work undertaken for each chapter of this body of work and reports strengths and limitations of the thesis. Finally, a discussion of future research work required in the field of FIT for symptomatic patients is offered in Chapter 8.
This thesis provides evidence contributing towards the optimal utilisation of FIT for the investigation of CRC in symptomatic patients. Within this work the stratification value of FIT in both “high-risk” and “low-risk” symptoms is confirmed. Thrombocytosis is confirmed as an independent risk-factor for CRC which can be used synergistically with a haemoglobin level, ferritin, and FIT result to more accurately stratify risk of CRC. This thesis has presented evidence of sociodemographic variation in the uptake of FIT across a symptomatic population in Nottingham for the first time, identifying a need for strategies to mitigate differential impact as the use of FIT in primary care expands. Finally, the validity and added diagnostic performance potentiated by a further FIT sample is described, to further minimise the risk of a false-negative leading to a missed cancer.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
