Current biological models envisage homeostasis of normal renewal of epithelial cells in the Lieberkhun’s crypt as the result of the plastic and modulable balancing of the interactions between cell proliferation and programmed cell death. The balancing keeps uniform both the architectural (height and number of cells per crypt) and functional (cell cycle duration, speed of cell transit ) features of the crypt in the entire large bowel.
Colonic epithelial serration, namely the saw-toothed outline derived from infolded epithelial tufts inthe crypt and in the luminal surface, occurs in consequence of the undue accumulation of cells, following the loss of the homeostatic trophism and can be found in several types of colorectal polyps.
Transient serration associated with hyperproliferation can befound in reparative and regenerating colonic epithelium, mainly in certain subsets of inflammatory polyps (Myoglandualar inflammatory polyps or in cloacogenic inflammatory polyps included in the mucosal prolapse syndrome).
Steady serration in the absence of hyperproliferation is displayed in the common hyperplastic polyps of the large bowel, in consequence of the inhibition of programmed cell death: dysregulation of apoptosis is therefore the basic event triggering and sustaining serrated tumourigenesis (“bottom-up morphogenesis”).
Epithelial serration together with minimal architectural changes, and without cytological atypia are the diagnostic features of hyperplastic polyps of the large bowel. They are very common (30% of resected colorectal polyps), asymptomatic, small-sized (0.2–0.5 cm) mucosal bumps of the sigmoid colon and rectum, consisting of straight, parallel, slightly elongated crypts lined by serrated epithelium in the intermediate and upper third, and by undifferentiated cells in the lower third.
Three cytotypes are recognized (microvesicular, goblet cell, mucin-poor),microvesicularhyperplastic polyps being the most frequent ones.
They are regarded as non-neoplastic lesions since no firm evidence has yet been presented of a greater likelihoodofcancerisation of these polyps in comparison with the normal mucosa. In effect, the absence of dysplasia, normal cell proliferation and cell hypermaturation are all characteristics supporting the non-neoplastic nature of hyperplastic polyps. On the other hand, hyperplastic polyps are more frequent in colonic segments with carcinoma and share some phenotypic features with colorectal adenocarcinoma. Hyperplastic polyps are likely to be markers of the action of an environmental factor interacting in the initiation stage of colorectal tumourigenesis, but not influencing promotion: they are paraneoplastic rather than neoplastic lesions, even though the specific mechanisms that orientate the mucosa towards hyperplastic differentiation remain to be elucidated.
They do not require additional surveillance.
Dysplasia can affect the serrated epithelium as seen in hyperplastic polyps, featuring mainly right-sided colonic neoplastic polyps, the serrated adenomas, and providing the rationale for a new path of colonic dysplasia: the serrated neoplasia. It represents a unique precancerous morphogenetic pathway in that architectural abnormalities precede and are mostly uncoupled from nuclear and cytological dysplastic changes.
Sessile serrated adenomas/polyps (SSA/Ps) are mainly right-sited (75%), large-sized (36%, <5 mm; 47%, 6–10 mm; 17%,>11 mm)sessile, serrated lesions displaying patchy or diffuse distorsions o ftissue organization: branching of crypts, serration of foveolar cell phenotype at the base of the crypt, dilatation at the base of the crypt, horizontal crypt growth.
Architectural features at the bases of the crypts, however, are often not visible in routine clinical practice because the slides are not histologically well-oriented, reducing histologic interpretation accuracy. The correct handling of the endoscopically resected specimens increases the number of longitudinally oriented crypts and allows the microscopic visualization of the deepest parts of the crypts, improving the accuracy of pathological diagnosis.
Intermingling and/or intermediate features of SSA/Ps and hyperplastic polyps can often be seen in the same histological section, above all in small polyps, impairing diagnostic reproducibility, so that the frequency of SSA/Ps diagnoses is widely ranging in the reported literature (1.1% - 14.7%). It has been estimated, however, that only 8.3% of the polyps previously diagnosed as hyperplastic polyps would now be reclassified as SSA/Ps, and such a level of diagnostic accuracy is felt acceptable to support clinical decision making.
The natural history of SSA/Ps, harbouring architectural changes only, is not well defined yet. They have been associated with an increased risk of both synchronous…..
….. and metachronous advanced colorectal neoplasia.
Shortened (3-5 years) surveillance intervals after endoscopic resection are therefore recommended.
A serrated dysplasia–carcinoma sequence exists, being histologically proven the transition to malignancy of serrated adenomas, that parallels the classical adenoma–carcinoma sequence of the large bowel and leads from serrated adenomas towards colorectal carcinoma through the transition from architectural to nuclear and cytologic serrated dysplasia or the conversion to conventional adenomatous dysplasia, the latter being even more common than the evolution to advanced serrated dysplasia.
In a large population-based study, 13.2% of SSA/Ps showed dysplasia, accounting for 0.17% of all colorectal polyps. In these lesions (Sessile Serrated Adenoma with Dysplasia, SSAD previously reported in the literature as “mixed hyperplastic/adenomatous polyps) an abrupt transition from otherwise typical SSA/P to variably extensive,cytologically dysplastic tissue microsectorsis seen. Nuclear and cellular dysplastic changes can be similar to those found in conventional colorectal adenomas (i.e.: nuclear pleomorphism, stratification and loss of polarity, atypical mitoses, basophilic cytoplasm) or maintain serrated features (eosinophilic cytoplasm, vesicular andbasally located nuclei).
The presence within a single lesion of a serrated and a conventional dysplastic component is in accordance with the proposed “fusion model”, which includes the two major routes of colorectal tumourigenesis, the adenoma–carcinoma sequence and the serrated pathway. Such polyps are thought to be relatively aggressive, exploiting both hyperproliferation of adenomas and apoptosis inhibition of serrated neoplasia…..
…….and occur in large number in serrated polyposis and in MUTYHgene-associated polyposis. It is now established that nearly half of MAP patients harbor serrated polyps causally associated to MUTYH gene deficiency, in addition to conventional adenomas, overlapping the phenotype of serrated polyposis. In this regard, three points should be considered for the differential diagnosis:
- Conventional adenomas are found in Serrated Polyposis (median number/patient: 2; range: 0-22)
- MUTYH mutation analysis should be considered in patients with more than 25 adenomas
- Bi-allelic MUTYH gene mutations with serrated polyps in the absence of adenomas have not been reported.
Stochastic modelling is conceivable for serrated neoplasia, although the evolutive rates of the lesions are unknown at the present time. On the whole, 30% of colorectal cancers are expected to originate from serrated polyps and approximately one-third of these cancers maintain serrated morphology. The overall association between serrated dysplasia and cancer (5.8%) peaks 55% in cancers showing microsatellite instability and 54–70% in serrated polyposis, the number of hyperplastic polyps and serrated adenomas being associated with CRC incidence.
The evolutive speed is not completely known yet: roughly, the time elapsed from the diagnosis of SSA/P and the onset of advanced carcinoma is 15 years, significantly longer than canventional adenoma – carcinoma transition time.
A subset of these lesions, however, appears to rapidly give rise to carcinoma often less than a few millimetres in size ( 4 out of 11 cases in Sheridan’s study).
Moreover, single cases have been reported with the malignat conversion of SSA occurring within few months.
This subgroup of serrated lesions is thought to be responsible of the vast majority of interval colorectal cancers (i.e.: tumors diagnosed in the interval between serial colonoscopy). The clinical (proximal location), pathological (mucinous histology), and molecular features (CIMP machinery) of such cancers are consistent with this interpretation.
It has however to be noted that the detection rates of serrated lesions vary dramatically between endoscopists, indicating that these lesions are significantly underdiagnosed in clinical practice. Among the the pitfalls impairing the detection rate, the presence of a thick mucus cap is the most frequent and important one. So, it is unclear whether the biological aggressiveness, the rate of missed lesions or both link serrated neoplasia to interval colorectal cancers.
Accounting for 1% of colorectal polyps, Traditional Serrated Adenoma (TSA) (serrated adenoma, polypoid type) is the least frequent serrated polyp. Mainly polypoid (67%) and – differently from SSA/Ps - located in the distal colon and rectum, TSAs display nuclear (elongated, hyperchromatic, stratified nuclei) and cytological dysplastic features within the serrated epithelium. They typically show tubule villous architecture, eosinophilic tall columnar epithelium with prominent serration, and ectopic crypt foci.
Both conventional and serrated dysplasia can be observed in TSA. With neoplastic progression, it is believed that TSAs acquire increasing degrees of cytological atypia before the development of carcinoma: a total of 25% of TSA show high-grade dysplasia, and 8% show intramucosal adenocarcinoma
This slide outlines the molecular genetics underlying serrated tumorigenesis. Methylation of CpG islands in the promoter regions of several genes is the genetic machinery sustaining the evolution of serrated neoplasia (CpG islands methylator phenotype, CIMP) by silencing pro-apoptotic (e.g., RASSF1, RASSF2) and cell-cycle inhibitory (p16, p14, p19, Rb) genes, whilst epigenetic suppression of DNA repair genes hMLH1andMGMT, inducing high-level MSI, would represent the rate-limiting step for tumour progression.
BRAF-proto-oncogene mutations are likely to represent the initiating event, unlinked with the action of the CIMP machinery.
Fusion pathways involving the sequential alterations of genes KRAS, APC, TP53 and MGMT could lead to SSA/P with conventional dysplasia and TSA.
To sum up, epithelial serration can occur in several types of colorectal polyps. Transient serration in inflammatory polyps, steady serration in hyperplastic polyps, in pre-neoplastic serrated polyps (SSA), and in neoplastic serrated polyps (SSSAD and TSA): the accurate histotyping drives the management of patients.