Patients with Inflammatory Bowel Disease (IBD) are at an increased risk of bowel cancer. However, the level of risk is currently uncertain. Recent studies report an initial increased risk with reduction over-time, whilst some report no increased risk. The reduction may be partly explained by improvements in IBD treatments which control inflammation within the bowel.
Another common disease of the digestive system causing inflammation is gallstones disease. The number of patients with gallstones has increased dramatically over the last decade and is expected to continue to rise. We will assess the risk of bowel cancer diagnosis and death in patients with IBD or gallstones compared to the general population. We are particularly interested in assessing the risk of bowel cancer by subsite; proximal (first section of the large bowel) and distal (final section of the bowel). We will examine trends over-time and risk in subgroups by age, sex, IBD type, disease duration and cholecystectomy (operation to remove gallbladder).
We will also study signs and symptoms of bowel cancer by subsite for patients with IBD, gallstones and the average risk population. We will compare the rates of these clinical features prior to diagnosis in patients with bowel cancer with the rates in the general population. We will also examine whether the time taken to be diagnosed after a GP visit with clinical features of bowel cancer is associated with how advanced the cancer is. The results of this study could be used to inform cancer surveillance, referral and investigative examination guidelines.
Using data from Clinical Practice Research Datalink (CPRD) with linkage to Hospital Episode Statistics (HES), the cancer registry (NCRAS) and Office of National Statistics (ONS), we will investigate the risk of colorectal cancer (CRC) incidence and mortality by subsite in patients with IBD or gallstones (objective one). Additionally, we will examine the clinical features of CRC by subsite and investigate the association between diagnostic interval and stage at diagnosis (objective two).
Methods
Objective one: Population-based cohort study design. Incident cases of IBD or gallstones diagnosed within a 25-year period will form two exposure cohorts, followed-up for CRC incidence and mortality, and compared with reference cohorts matched for age, sex and GP practice. We will use Cox regression to estimate hazard ratios (HR) adjusted for confounders for CRC incidence and mortality by subsite. HRs will be calculated for each cohort (gallstones/IBD) and stratified by age, gender, IBD type and record of cholecystectomy for the gallstone cohort. We will also calculate HRs by calendar period, stratified by time since IBD/gallstone diagnosis, to examine temporal trends in CRC incidence and mortality by subsite.
Objective two: Nested case-control study design. We will use conditional logistic regression to select and quantify key clinical features or combinations of features indicative of high risk of CRC overall and by subsite for three separate patient groups; patients diagnosed with IBD, patients diagnosed with gallstones and an average risk population. We will then go on to examine the relationship between diagnostic interval and stage at diagnosis (early stage versus advanced disease), controlling for important patient characteristics such as age, sex, co-morbidity and route of presentation (emergency) using logistic regression with restricted cubic splines. The initial model will assess total diagnostic interval for all CRC subsites. Subsequent models will assess primary and secondary care intervals separately and CRC by subsite.
Primary outcome measures:
Incidence of colorectal cancer (overall and by subsite);
Signs and symptoms associated with colorectal cancer (overall and by subsite);
Colorectal cancer diagnostic interval;
Colorectal cancer stage at diagnosis.
Secondary outcome measures:
All-cause mortality;
Colorectal cancer mortality (overall and by subsite).
Amanda Cross - Chief Investigator - Imperial College London
Ruth Disney - Corresponding Applicant - Imperial College London
Amanda Cross - Collaborator - Imperial College London
Andrew Hart - Collaborator - University of East Anglia
David Muller - Collaborator - Imperial College London
Katherine Wooldrage - Collaborator - Imperial College London
Konstantinos Tsilidis - Collaborator - Imperial College London
HES Admitted Patient Care;HES Outpatient;NCRAS Cancer Registration Data;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation