Serrated Colorectal Polyps: An Update

Selective expression of gastric mucin MUC6 in colonic sessile serrated adenoma but not in hyperplastic polyp aids in morphological diagnosis of serrated polyps

Colonic sessile serrated adenoma, in contrast to hyperplastic polyp, is thought to be related to sporadic colorectal cancers with high microsatellite instability. However, the morphological distinction between these entities is difficult and subject to observer and sampling variation. Therefore, we elected to investigate the expression of gastric mucin MUC6 as a potential marker to separate the two in the hope of finding an objective and reproducible adjunct to morphological diagnosis. Endoscopic biopsies of colonic polyps with serrated architecture, but without cytological dysplasia were studied and categorized as sessile serrated adenoma or hyperplastic polyp, using previously published morphological criteria. Smaller groups of serrated polyps with cytological dysplasia (traditional serrated adenomas, filiform serrated adenomas and sessile serrated adenomas with cytological dysplasia) were also included. In total, 94 polyps were immunohistochemically stained with antibodies to MUC6 and to MLH-1. MUC6 was found to have 100% specificity in distinguishing sessile serrated adenoma (N=26; positive staining) from hyperplastic polyp (N=48; negative staining). Traditional serrated adenomas and filiform serrated adenomas were also negative for MUC6. Sessile serrated adenomas with cytological dysplasia were found to lose expression of MLH-1 in dysplastic areas, while retaining MUC6 expression. Neither anatomic location in the right or left colon nor polyp size appears to account for the differences in MUC6 expression.

Sessile serrated polyps of the colorectum are rare in patients with Lynch syndrome and in familial colorectal cancer families

Whereas the generally accepted carcinogenesis pathway of the microsatellite instabile high (MSI-H) colorectal carcinoma (CRC) involves the traditional adenoma in patients with Lynch syndrome, a serrate pathway involving serrate adenomas (SA) and sessile serrate polyps (SSP) characterize the sporadic MSI-H counterpart. Recent studies have, however, challenged such simple one-pathway models, inviting the consideration of alternative, unexpected pathways. Here, the issue as to the possible role of SSP, primarily in the context of Lynch syndrome, but also in subjects from familial CRC families (FCF) is addressed. Polyps coded as hyperplastic polyps (HP) from subjects with Lynch syndrome and FCF enrolled in the HNPCC-register at the Hvidovre University Hospital as well as adenomas from this population were retrieved and reviewed for features of SSP. Ninety-eight polyps coded as HP and 41 polyps coded as adenoma from 14 individuals with Lynch syndrome as well as 17 individuals from FCF constituted the study material. Seven of the 98 polyps coded as HP displayed histological features that, to varying extent, deviated from the traditional HP (THP), yet, merely two of these, both from the FCF, were considered examples of probable SSP. None of the 41 cases coded as adenoma possessed a morphology that qualified as SSP. The prevalence of SSP was not increased as compared to the background population and thus, this serrated lesion does not appear to play a tumorigenic role in Lynch syndrome, nor in FCF.

Colorectal serrated adenocarcinoma

Colorectal cancer (CRC) ranks among the three most common cancers in terms of both cancer incidence and cancer-related deaths in most Western countries. Serrated adenocarcinoma is a recently described, distinct variant of CRC, accounting for about 7.5% of all CRCs and up to 17.5% of most proximal CRCs. It has been postulated that about 10-15% of sporadic CRCs would have their origin in serrated polyps that harbour a significant malignant potential. These lesions include hyperplastic-type aberrant crypt foci, hyperplastic polyps, sessile serrated adenomas, admixed polyps and serrated adenomas, and constitute the so-called 'serrated pathway', which is distinct from both the conventional adenoma-carcinoma pathway and the mutator pathway of hereditary non-polyposis CRC and is characterized by early involvement of oncogenic BRAF mutations, excess CpG island methylation (CIM) and subsequent low- or high-level DNA microsatellite instability (MSI). Methylation of hMLH1 is likely to explain the increased frequency of high-level MSI (16%) and methylation of MGMT is postulated to explain the low-level MSI (29%) in serrated adenocarcinomas. Reproducible histopathological criteria for serrated adenocarcinoma have recently been established and they have been qualified by DNA expression analysis for 7928 genes, showing clustering of serrated adenocarcinomas into a molecular entity apart from conventional adenocarcinoma, and representing with distinct down-regulation of EPHB2, PTCH and up-regulation of HIF1alpha.

Morphology and microsatellite instability in sporadic serrated and non-serrated colorectal cancer

Colorectal serrated adenocarcinoma originates from serrated adenoma, but definite histological criteria have not yet been established. It presents with frequent DNA microsatellite instability (MSI), but the frequency of low-level (MSI-L) and high-level MSI (MSI-H) and the expression of mismatch-repair (MMR) enzymes in serrated adenocarcinoma are not known. To address these questions, morphological criteria for serrated cancers were established, their validity was tested, and MSI analysis was performed with NIH consensus markers and MMR enzyme immunohistochemistry for hMLH1, hMSH2, and hMSH6 in 35 serrated and 75 non-serrated colorectal carcinomas. Serrated carcinomas frequently showed a serrated, mucinous or trabecular growth pattern; abundant eosinophilic cytoplasm; chromatin condensation; preserved polarity; and the absence of necrosis. With these features, it was possible to distinguish them from non-serrated cancers, with the mean kappa score for five observers being 0.509. MSI analysis was successful in 31 serrated and 73 non-serrated carcinomas. 54.8% of serrated carcinomas were microsatellite-stable (MSS), 29.0% presented with MSI-L, and 16.1% presented with MSI-H, whereas 78.1% of non-serrated carcinomas were MSS, 13.7% were MSI-L, and 8.2% were MSI-H. MSI-L was more common in serrated cancers (p=0.035) and it was associated with patchy immunohistochemical staining (33.3%) of MLH1. MSI-H did not differ between serrated and non-serrated cancers (p=0.14). These results suggest that the biological background of serrated carcinomas differs from sporadic non-serrated colorectal cancer, but is not directly related to MSI.