Roel Bogie

General discussion

Molecular features of PCCRCs As stated above, in IBD patients reduced adenoma dwell time could explain the high incidence of PCCRCs (with respect to all CRCs). In healthy individuals (at risk for sporadic CRC), the precursor lesions of PCCRCs have also been hypothesized to have reduced dwell time. If this is the case, we expect them to have unique molecular features. Otherwise, other features, such as shape and location, may contribute to their occurrence. In Chapter 10 we examined the molecular profile of all PCCRCs diagnosed between 2001 and 2010 in the South Limburg region. We also studied the molecular profile of randomly selected prevalent CRCs from the same population. Based on an often flat appearance and proximal colonic location of LSTs ( Chapter 3 ), on the fact that non-polypoid neoplasms and serrated lesions have a subtle and more difficult to detect appearance 60, 61 and on a higher recurrence risk of non-polypoid neoplasms, 36 we hypothesized that non-polypoid and serrated lesions may frequently contribute to PCCRCs. Results showed that MSI, CIMP and BRAF gene mutations are associated with PCCRCs. Gain of the 13q chromosome and loss of the 18q chromosome were less frequent in PCCRCs. A PCCRC specific molecular profile was not found. Instead, our findings support the hypothesis that non-polypoid neoplasms and serrated lesions are important precursor lesions. After correction for colonic location, the differences in MSI, CIMP and BRAF gene mutation were no longer present. We already know that non-polypoid and serrated lesions are predominantly located in the proximal colon, so the result that these molecular findings are predominantly found proximally further supports the hypothesis. A study focusing on molecular profiles, used histology specimens of CRC developed within a serrated neoplasm and confirmed the proposed serrated pathway. 62 In that study, 5 of the 6 studied CRCs contained the same mutations as the surrounding sessile serrated neoplastic tissue; CIMP, MSI and BRAF gene mutations. The remaining CRC had features of the CIN pathway and lacked the BRAF mutation of the surrounding tissue. The authors state that the sessile serrated precursor lesions not always had signs of cytological dysplasia, meaning that serrated lesions should always be considered as precursor lesion. Several other studies also pointed to an increased risk of developing CRC in patients with sessile serrated lesions. 63, 64 Using national databases for a case-control study, it was demonstrated that patients with SSA/Ps had a OR of 3.07 (95% CI: 2.30 – 4.10) for developing CRC when compared to controls. 63 Sessile serrated lesions with cytology markers or a proximal location additionally increased the risk. In that study, hyperplastic polyps were reviewed by pathologists to apply the new definition for serrated lesions. The conclusion of the authors was that patients with sessile serrated adenoma/polyps or traditional serrated adenoma, had a similar to or even higher risk of developing CRC than patients with conventional adenomas. 63 Another study that used data from a randomized controlled study to CRC screening, largely agrees with this conclusion. 64 However, in that study several serrated lesions remained in situ after taking biopsies. Years later some of the patients from that study developed CRC, but not at the location of the previously diagnosed serrated lesion. This led to the conclusion that patients with sessile serrated lesions may have a higher risk profile for CRC, but not directly from malignant growth of the serrated lesions. 64 Furthermore, previous studies showed that high risk serrated lesions coincide with advanced adenomas. 61 Since only 23 serrated lesions were left in situ, data is scarce and definite conclusions cannot be drawn.

12

229