Colorectal cancer in inflammatory bowel disease : aspects on post-colonoscopy colorectal cancer

Abstract

Colorectal cancer (CRC) is the third most common cancer in the world. It is diagnosed by and can be prevented with colonoscopy combined with polyp removal. A post-colonoscopy CRC (PCCRC) is a cancer detected within 3 years after a colonoscopy in which no cancer was found and is a central quality measure for colonoscopy since it mainly arises from missed or incompletely resected lesions. Sub-group analyses have indicated that IBD, also at increased risk for CRC, is a risk factor for PCCRC; however, studies exploring the risk factors within the IBD group or reporting nationwide PCCRC rates for IBD are limited. Hence, this thesis aims to increase the knowledge of PCCRC among the IBD population and gain a deeper understanding of its causes and identify the addressable risk factors.

In Study I, a registerbased cohort study, 348 000 colonoscopies performed on 271 000 individuals in Sweden from 2001–2010 were identified. The PCCRC rate was 28% for Crohn’s disease (CD) and 41% for ulcerative colitis (UC). The relative risks compared to the non-IBD group were 3.8 (95% confidence interval [CI] 2.9–5.0) for CD and 5.9 (95% CI 5.1–6.8) for UC. The risk factors for PCCRC differed between the IBD and non-IBD groups, with young age, male sex and rectal CRC having a higher risk. The highest PCCRC rate was observed among patients with CD and rectal cancer at 68%.

Study II, a descriptive population-based case review study, included Swedish patients with CD and rectal cancer who underwent colonoscopy within 36 months of cancer diagnosis from 2001–2015. The medical records were reviewed, and of 24 patients with rectal PCCRC, 79% were men. The median disease duration was 21 (interquartile range 19–30) years, and the median age was 50 (45–59) years. Distal rectal location was most common, and when retroversion was added to the algorithm for determining the adequateness of the prior examination, the most plausible explanation for 77% of the PCCRCs was ‘possible missed lesion, index colonoscopies negative but adequate’. Noninterval PCCRC type C (54%) and B (37%) were the most common PCCRC sub-categories, while none were interval cancers.

Study III, a register-based cohort study, examined the notion of ‘post-endoscopy CRC’ (PECRC) in IBD patients who have undergone colectomy, with either an ileorectal anastomosis (IRA) or a rectal remnant. The patients underwent an endoscopy between 2001 and 2012. We identified 12 patients with an IRA and 21 patients with residual rectal stumps diagnosed with CRC within 36 months of endoscopy. Of these, 67% were PECRCs, and the PECRC rate was 51%. Younger age and shorter time from the colectomy to endoscopy were identified as risk factors for PECRC while there was no difference in risk between the IBD subtypes, or between IRA or rectal remnant.

In conclusion, PCCRC risk is significantly increased in IBD patients, and the risk factors differed from those previously known in non-IBD patients. Sub-categorisation and root-cause analysis of a subgroup with the highest PCCRC rate showed several areas where PCCRC can be prevented and theoretically lower its rate in IBD patients. Adjustment in the root-cause algorithm is suggested in the context of rectal cancer, and PECRC is recommended for the evaluation of endoscopy quality in colectomised patients with a residual rectum.