Roel Bogie

Chapter 12

On the other hand, leaving pre-malignant colorectal neoplasms in situ or applying inappropriate surveillance intervals may increase PCCRC risk and should be avoided. Patients’ acceptance of relying on the endoscopist’s optical diagnosis may also be limited according to an Australian questionary study. 54 In Chapter 8 the performance of endoscopists who qualified for and were certified to participate in the Dutch national colorectal cancer screening program, was tested. A local database was used, containing the first 2470 national CRC screening colonoscopies in the South Limburg region after implementation in 2014. Suspected histology had to be documented for each polyp as a requirement from the national screening organization. The diagnostic accuracy of diminutive lesions (≤5 mm) was low with 76% (95% CI: 74 – 77) throughout the colon. Selecting only diminutive lesions of the rectum, the diagnostic accuracy was even worse with 71% (95% CI: 69 – 74). The negative predictive value, however, was 84% (95% CI: 80 – 87). Chapter 8 concludes that with these results, leaving diminutive lesions in the distal colon is not yet a safe option. Resection without histopathological assessment is another strategy. When applying this so-called ‘resect and discard’ strategy, surveillance interval advices are identical to the advice after histopathological assessment in more than 90% of the cases. The 9% discrepancies led in about 6% of the cases to shorter intervals and in about 3% to longer surveillance intervals. Improvements in tools for optical polyp assessment, such as the BASIC classification, could help to improve the diagnostic accuracy, making these strategies more favorable. 55 PCCRCs in IBD patients So far, patients with inflammatory bowel disease (IBD) were excluded in all studies because of a much higher incidence of CRC. 56 Chronic inflammation is thought to stimulate carcinogenesis through specific molecular pathways. 57 The adenoma-carcinoma dwell time may be shorter because of this. Furthermore, neoplasia in IBD patients are predominantly flat, leading to more difficult detection, especially in case of active inflammation. Taken together, these data suggest that IBD patients are at higher risk for developing PCCRCs. In Chapter 11 , the frequency and etiology of PCCRCs within all CRCs detected in IBD patients was studied. The IBD South Limburg cohort was used, containing 1644 patients with ulcerative colitis and 1157 patients with Crohn’s disease at the time of analysis. This database comprised a total of 25,931 person years at risk for CRC diagnosis, in which 20 CRCs were detected. Nine of these CRCs complied with the definition of PCCRC. The adapted Pabby algorithm was applied, resulting in five probably missed lesions, one inappropriate surveillance interval, one inadequate bowel examination, one incomplete resection, and one newly developed cancer. A similar subdivision of etiology, with the highest proportion of missed lesions, was also seen in the PCCRCs within the general population based cohort of the same region. 58 The conclusion was that the default interval between IBD diagnosis and start of surveillance colonoscopies of eight years, may be too long. The active inflammation at IBD diagnosis impairs neoplasm detection and could reduce adenoma dwell time. In IBD patients, adequate visualization of the mucosa is a key factor for successful surveillance. A repeat colonoscopy after successful treatment of inflammation was suggested to assure a neoplasm free colon as start of surveillance. A large population-based study from Sweden confirmed the higher risk of developing PCCRC among IBD patients. Compared to non-IBD patients, the relative risk of having a CRC diagnosis within 3 years after last colonoscopy was 3.82 (95% CI: 2.94 – 4.96) for Crohn’s disease and 5.89 (95% CI: 5.10 – 6.80) for ulcerative colitis. 59

228