A Historical Perspective and Exposé on Serrated Polyps of the Colorectum

Serrated architecture in colonic mucosa means having sawtooth-like notches along the luminal edges of the crypts. The serrations are often the result of small clumps of epithelial cells, sometimes referred to as micropapillary processes, that protrude into the lumen and are separated by single cells along the sides of crypts and mucosal surface.¹  Less commonly, the serrations result from small cryptlike buds along the sides of the main tubules, which have been given the name ectopic crypt formations.²  Serrated polyps are now commonly identified in colorectal biopsies, but 60 years ago they were not recognized. How did this change in our concept of serrated polyps develop? To answer this question, let us look into the history and evolution of serrated polyps.

Two German pathologists, Feyrter in 1929³ and Westhues in 1934,⁴  are credited as being the first to use the term hyperplastic polyp in German-language literature.^1,5  In a 1959 histochemical study on oxidative enzymes in colonic lesions, Wattenberg⁶  used the term focal mucosal hyperplasia to refer to a benign proliferative lesion; the photomicrographs of this lesion are identical to what we call a hyperplastic polyp (Figure 1, A).

Basil Morson from London, United Kingdom, perhaps the best known and revered gastrointestinal pathologist of the mid to late twentieth century, used the term metaplastic polyp in 1962 for what we now call hyperplastic polyps in order to distinguish them from true adenomatous proliferations.⁵  Morson explained, “The term ‘metaplasia' is only meant to describe a change in the appearance of the mucous membrane when compared with the normal.” He questioned the accuracy of the term hyperplasia given there was “…little evidence of any increase in cellular activity in these polyps.” ⁵ The term metaplastic polyp remained the preferred designation in the United Kingdom until around 2006, when a group of gastrointestinal pathologists decided it was time to change it to the hyperplastic polyp designation.⁷ 

Lane and Lev⁸  in 1963 seem to be the first or among the first in the United States to call polyps “hyperplastic” to distinguish them from adenomas. Their study mainly involved examination of serial sections of minute polyps in patients with familial adenomatous polyposis, but they also digressed to discuss tiny polyps in general, and it was in that digression that they called polyps with sawtooth architecture hyperplastic polyps. This name, however, did not achieve widespread acceptance in the surgical pathology community until the late 1960s. What exactly were these small serrated polyps so named before we became familiar with the hyperplastic polyp term? It's unclear, but it is possible these were called adenomas. Why would we call them adenomas back in those old days? The answer is simple: because those were the days when men were men and women were women… and all polyps were adenomatous. Actually, the proliferative zone at the base of tubules in hyperplastic polyps sometimes resembles adenomatous epithelium, which may have been the reason for calling them adenomas.

Some studies looked at the cell kinetics of hyperplastic polyps. In a 1973 study by Kaye et al,⁹  the authors found that when compared to the normal progressive maturation of cells in colonic crypts, “the hyperplastic epithelium exhibits a similar progression, the primary difference being that most of the morphological features of maturing and mature cells are found either lower in the crypt or in exaggerated form at the same level of the crypt when compared with normal mucosa in the same colon.” A year later, Hayashi et al¹⁰  used electron microscopy to show that the cells on the surface of hyperplastic polyps looked hypermature and demonstrated a decreased rate of migration in cells from base to surface when using autoradiography. The authors suggested that there was a longer turnover time and delayed migration as compared to normal mucosa. In a 2004 publication, Higuchi and Jass¹¹  updated these concepts, stating: “Using more modern terminology, the hyperplastic polyp can be defined as a lesion caused by the inhibition of programmed cell death or apoptosis,” which they explain is probably a special form of apoptosis called anoikis that is associated with programmed exfoliation of surface epithelial cells. They go on to say, “Serration arises when cells continue to be generated… but are blocked from ascending to their rightful location within the epithelial surface,” ¹¹  presumably by surface cells that were not properly exfoliated. Putting all of these studies in context, hyperplastic polyps are defined by serrations and hypermature epithelium that result from early cellular maturation, a slower than normal migration of cells from base to surface, and delayed loss of surface cells due to inhibition of apoptosis.

It also became apparent that hyperplastic polyps have an abundance of cells with finely vacuolated apical cytoplasm, which by electron microscopy were identified as absorptive cells^12,13  (Figure 1, B). The proportion of these absorptive cells to goblet cells varies in hyperplastic polyps. In a reappraisal of serrated colorectal polyps by Torlakovic et al,¹⁴  hyperplastic polyps were subdivided to include the “microvesicular cell type,” which are the more frequently encountered type with a predominance of absorptive cells, and “goblet cell type,” which are rich in goblet cells with fewer absorptive cells¹⁴  (Figure 1, B and C). In our practice, we call both hyperplastic polyps and do not separate these by cell type.

Hyperplastic polyps also have some other peculiarities. For instance, those in the rectum are likely to have large bundles of smooth muscle extending into the base from the muscularis mucosae, a change often considered to be a manifestation of prolapse, yet it is very difficult to understand why a tiny polyp should undergo prolapse changes. Some hyperplastic polyps have atypical epithelium and mitoses in the basal proliferative zone that mimic dysplasia or viropathic change (Figure 1, D). These have been studied by different groups, most recently in 2007 by Kambham et al,¹⁵  who examined a series of serrated polyps with atypical multinucleated giant cells that were characterized by enlarged, smudgy, overlapping nuclei, and large inclusion-like nucleoli in some cases. The authors tried to identify viral proteins in the polyps primarily by using immunohistochemistry, but they did not find evidence of their presence; an etiology was not determined and they suggested this was a degenerative phenomenon in response to inflammation or injury.¹⁵  These changes, which seemed to be fairly common a couple of decades ago, are rarely seen any more. If these hyperplastic polyps resulted from some kind of viral epidemic in the past, it appears that the epidemic has passed.

Following the earlier detailed studies about cell kinetics and cell types in hyperplastic polyps, there was a long period of complacency. We recognized only 1 type of colorectal serrated polyp, the hyperplastic polyp, which was not considered precancerous. This changed with the publication of a 1996 article by Torlakovic and Snover¹⁶  in which they examined 6 patients with numerous serrated polyps who also had colonic adenocarcinoma or numerous adenomatous foci. The polyps tended to be large (average size 1.2 cm) and were distributed throughout the colon in 3 cases, predominantly left sided in 2 and right sided in 1.¹⁶  The authors gave the name serrated adenomatous polyposis to the condition observed in this group of patients, and this became the first documentation of carcinoma-associated serrated polyps.

In their morphologic reappraisal of serrated polyps, Torlakovic et al¹⁴  recommended that “…for practical purposes, large lesions with abnormal proliferation, dilated distorted crypts, dilatation of the base of the crypts, and/or focal nuclear atypia with pseudostratification… be given a different designation, ‘sessile serrated adenoma.'” There were, and still are, pathologists who prefer the name sessile serrated polyp rather than sessile serrated adenoma, since these lack typical adenoma-like dysplasia. The ultimate compromise in the 2010 WHO Classification of Tumours of the Digestive System was to name these entities sessile serrated adenomas/polyps, abbreviated as SSA/Ps.¹⁷  Regardless of the name, the important thing is for pathologists and clinicians to recognize that these are carcinoma-associated serrated polyps. We prefer the term sessile serrated adenoma in our practice, which we feel more clearly conveys its cancer association.

Sessile serrated adenomas/polyps have clinical and histologic features that differ from typical hyperplastic polyps. These SSA/Ps tend to be large and located in the cecum and ascending colon. Endoscopically, they often have a mucous covering and a sessile appearance. Some, however, are difficult to detect because they form “flat” rather than sessile endoscopic lesions. Histologically, both SSA/Ps and hyperplastic polyps have a mix of absorptive and goblet cells, but as it was pointed out by Torlakovic and Snover,¹⁶  SSA/Ps have greater architectural complexity (Figure 2, A and B). This includes crypt dilation with mucus retention, lateral growth at the base of crypts, and exaggerated serrations along the middle of the crypts that sometimes extend to the base. Goblet cell maturation can be seen at the base of SSA/Ps, and in addition, goblet cells on the surface or superficial portion of crypts may be stratified, which we refer to as “floating” goblet cells. Peculiarly, some of these big serrated polyps overlie large amounts of submucosal adipose tissue resembling a lipoma (Figure 2, A). Some SSA/Ps contain small foci of a bland spindle cell proliferation in the lamina propria that resembles a perineurioma. It is also not unusual for these abnormal dilated basal crypts to herniate into the superficial submucosa through breaks in the muscularis mucosae (Figure 2, C). Quite likely, this herniation or “inversion” occurs into lymphoid follicles, which form natural breaks in the muscularis mucosae.

Not only do SSA/Ps have these peculiar architectural changes, but they also have unusual proliferative dynamics. Sessile serrated adenomas/polyps are characterized by movement of the proliferative zone away from its usual location at the base, resulting in abnormal downward growth and leading to architectural distortion and basal goblet cell maturation.^2,18  This can be seen with a Ki-67 (MIB1) immunohistochemical stain. In normal crypts, Ki-67 labeling is symmetric and limited to the base of crypts.²  A typical SSA/P, in contrast, has variably irregular and asymmetric staining²  that can include loss of staining at the base and extension much higher in the crypts, sometimes almost approaching the surface, a reflection of the proliferative zone's movement away from the base (Figure 2, D). Sessile serrated adenomas/polyps also have cytologic abnormalities seen at the light-microscope level such as large vesicular nuclei with nucleoli in cells along the sides of the crypts and mitoses in the middle and upper crypts (Figure 2, E), further evidence to support aberrant proliferation away from its normal location at the crypt base.

What are the minimal criteria for the light microscopic diagnosis of an SSA/P? The World Health Organization (WHO) 2010 classification of tumors states, “…if more than two or three contiguous crypts demonstrate features of SSA/P, the lesion should be classified as SSA/P.” ¹⁷ However, this is not helpful as this definition includes 2 numbers, namely, 2 and 3. Both cannot be the minimal criteria, it must be one or the other. In 2012, a group of experts on serrated polyps from around the world put out a set of recommendations for diagnosing and managing all serrated colon polyps.¹⁹  This group suggested that “…the presence of at least one unequivocal architecturally distorted, dilated, and/or horizontally branched crypt, particularly if it is associated with inverted maturation, is sufficient for a diagnosis of SSA/P” ¹⁹  (Figure 2, F). Inverted maturation means the presence of goblet cells at the very base of the crypt. Most of the SSA/P features are architectural, so it is usually easier to find the diagnostic changes in large polyps, since small polyps provide less material for observation. The distinction, however, cannot be achieved in all cases and there are polyps with borderline features between SSA/P and hyperplastic polyp.

Some SSA/Ps are complicated by typical adenoma-like dysplasia (Figure 3, A). We used to refer to these as mixed hyperplastic/adenomatous polyps. Now they are referred to as sessile serrated adenoma/polyp with cytologic dysplasia. In keeping with the overall architecture, the dysplastic crypts are also often serrated. It appears that when carcinomas develop in these polyps, they develop after the initial development of cytologic dysplasia (Figure 3, B and C). This progression of SSA/P to SSA/P with cytologic dysplasia to invasive carcinoma led to the concept of the serrated pathway for colorectal carcinoma.^20,21  Serrated pathway carcinomas have diverse histologic and molecular profiles, but they have in common a RAS-RAF-MAPK signaling pathway activation, usually BRAF or KRAS mutation, and a CpG island methylator phenotype that can result in either microsatellite-instable or microsatellite-stable colorectal adenocarcinomas.²²  In the best characterized serrated pathway of SSA/P progression to sporadic microsatellite-instable-high carcinomas, a BRAF activating mutation is an early event that results in the formation of an SSA/P from normal mucosa or possibly a microvesicular hyperplastic polyp precursor, and subsequent CpG island methylation of the hMLH1 promoter region results in epigenetic silencing of this DNA mismatch repair gene; these lesions tend to accumulate additional mutations at a rapid rate, which coincides with the development of dysplasia and progression to invasive carcinoma.^17,18,22  Because of this cancer association, recommendations for follow-up or surveillance for patients with serrated polyps are available that take into consideration the type, number, size, and location of serrated polyps.¹⁹ 

There is 1 more type of serrated polyp, which may not be related to either hyperplastic polyps or sessile serrated adenomas, that we currently classify as the traditional serrated adenoma (TSA). Possibly the first report of these polyps was by Longacre and Fenoglio-Preiser²³  in a 1990 article, which they called mixed hyperplastic adenomatous polyps/serrated adenomas. They noted that, compared to hyperplastic polyps, these proliferations had certain differences including goblet cell immaturity, upper zone mitoses, prominent nucleoli, and lack of a thick basement membrane. Traditional serrated adenomas tend to be exophytic, villiform, and are characterized by epithelium that has prominent pink cytoplasm with scattered goblet cells and remarkably uniform elongated hyperchromatic nuclei that stratify generally no more than half the thickness of the cells (Figure 4, A and B). Cells of this type are very uncommon in more typical adenomas that generally have much more nuclear irregularity. It is uncertain if these cells are dysplastic; they might actually be metaplastic or senescent.²  Architecturally, TSAs are characterized by small cryptlike buds extending from the sides of the crypts into the lamina propria, which have been referred to as ectopic crypt formation and may account for their protuberant exophytic growth. Some experts believe that these ectopic crypts are actually the best defining features of these polyps.^2,24  In some studies, the TSAs tended to have an adjacent lesion such as a hyperplastic polyp or an SSA/P, which may or may not be precursors. For instance in one study, 24% of 55 left-sided TSAs had associated lesions that were mostly hyperplastic polyps.²⁵  In another study, 38% of 200 pancolonic TSAs had associated lesions that were mostly SSA/Ps.²⁶  How to best classify such polyps continues to be debated. There are reports of adenocarcinomas developing within TSAs that have cytologic dysplasia,²⁶  comparable to the situation in the serrated pathway for SSA/Ps, but the cancer risk for TSAs is unknown and additional long-term follow-up studies are needed to evaluate this risk.

Serrated adenomatous polyposis, the term initially proposed by Torlakovic and Snover,¹⁶  is now referred to as the serrated polyposis syndrome. The genetic basis is currently unknown, and various definitions have been proposed. The 2010 WHO working criteria for serrated polyposis include 3 definitions: (1) at least 5 serrated polyps proximal to the sigmoid colon of which at least 2 are greater than 1 cm, (2) any number of serrated polyps proximal to the sigmoid colon in an individual with a first-degree relative with serrated polyposis, or (3) 20 or more serrated polyps of any size distributed throughout the colon.¹⁷  These serrated polyps can be hyperplastic polyps, SSA/Ps, or TSAs. Recent clinical studies looking at risk factors associated with developing colorectal carcinoma in patients diagnosed with serrated polyposis syndrome have led to discussions on how to best manage and treat these patients,²⁷  and the criteria for diagnosing serrated polyposis syndrome will need to be refined as we gain better understanding of this entity.

Other lesions can have serrated change and be mistaken for polyps. For instance, prominent serrations can be found in prolapsed mucosa, including rectal prolapse associated with the solitary rectal ulcer syndrome and polypoid prolapsed mucosa associated with diverticulosis. Another challenge is that some serrated polyps do not conform to the classic definitions of hyperplastic polyp, SSA/P, or TSA and may have overlapping features of these entities. New serrated polyps will likely be recognized among these currently unclassifiable polyps, and our understanding of serrated polyps will undoubtedly change as further insights are made into the molecular pathways of serrated lesions.