Pathologic Features

Grossly, inflammatory pseudopolyps may be sessile or pedunculated. They are almost always smaller than 2 cm, but so-called giant inflammatory polyps may grow to large sizes and cause obstruction. Filiform polyposis refers to the presence of numerous dense, filamentous polyps that can project several centimeters above the surrounding mucosa ( Fig. 22.1 ). This form of polyposis is usually associated with IBD or, rarely, juvenile polyposis.

A, Endoscopic appearance of a giant “filiform” inflammatory polyp in a patient with chronic ulcerative colitis. B, The inflammatory polyp projects several centimeters above the mucosal surface.

Histologically, some polyps, particularly those adjacent to ulcers or anastomotic sites, are formed entirely, or almost entirely, by inflamed granulation tissue. However, most inflammatory pseudopolyps are composed of a mixture of inflamed lamina propria and distorted colonic epithelium; surface erosions may or may not be present ( Fig. 22.2 ). Colonic crypts are often dilated, branched, and hyperplastic, and neutrophilic cryptitis and crypt abscesses may be prominent. In the later stages of development, inflammatory polyps may consist of many finger-like projections of normal or near-normal mucosa surrounding a core of submucosal tissue.

Inflammatory pseudopolyps. A, Endosopic appearance of diffuse (severe) inflammatory polyposis in a patient with chronic ulcerative colitis. B, This colonic resection specimen shows broad, shallow ulcers surrounded by raised, erythematous tags of inflamed mucosa. C, Inflammatory pseudopolyps in patients with inflammatory bowel disease are frequently composed of a mixture of acute inflammation, dense mucosal lymphoplasmacytic infiltrate, and distorted dilated crypts with surface erosion, cryptitis, and crypt abscesses. D, Crypt abscesses with crypt rupture may be present and may lead to mucin granulomas or mucin extravasation.

Three potential pitfalls can occur in the histologic evaluation of inflammatory pseudopolyps. First, dysplasia may rarely develop in these lesions, particularly in those associated with IBD. Much more commonly, regenerating epithelium can be extreme, particularly in inflamed or eroded areas, simulating a neoplastic process. Careful attention to surface maturation, which is usually observed in regenerating epithelium and almost never in dysplasia, can help in making this distinction in difficult cases. Second, bizarrely shaped, enlarged, or multinucleated stromal cells that can mimic sarcoma (termed “pseudosarcoma”) may develop in inflammatory pseudopolyps. These cells may be spindled or epithelioid, and they often aggregate at the surface of the polyp, underneath areas of ulceration and granulation tissue ( Fig. 22.3 ).

“Pseudosarcoma” in an inflammatory polyp. Bizarrely-shaped, enlarged, atypical stromal cells may occasionally be seen in granulation tissue of an inflammatory pseudopolyp, particularly beneath an area of surface erosion. These cells can be confused with sarcoma, carcinoma, or viral infection (e.g., from cytomegalovirus).

Finally, in many instances, the histologic appearance of inflammatory pseudopolyps may be indistinguishable from that of juvenile polyps. Distinction between these two types of lesions is based largely on clinical information, such as the age of the patient (almost all such polyps in young children are juvenile polyps), the presence or absence of a clinical history of a hamartomatous polyposis syndrome (juvenile polyposis, Cowden disease, and Bannayan-Riley-Ruvalcaba syndrome [BRRS] can all have juvenile-type polyps), and the presence or absence of an underlying inflammatory disorder.

Natural History and Treatment

Inflammatory pseudopolyps tend to persist even after healing of the surrounding mucosa has taken place. They are frequently found in quiescent ulcerative colitis. These polyps have no increased tendency for neoplastic transformation. Treatment is usually directed at the underlying inflammatory condition. Rarely, surgical excision is indicated when large or numerous polyps cause symptoms due to bleeding or obstruction.

Inflammatory Cap Polyp (and Cap Polyposis)

Inflammatory cap polyps were first described in abstract form in 1985 by Williams and co-workers. These rare lesions usually occur in the setting of cap polyposis, in which dozens of these types of polyps develop. There is no sex predilection, and they occur over a wide age range. Clinically, patients with cap polyposis often are seen with diarrhea, mucoid stools, gastrointestinal (GI) bleeding, and/or tenesmus. Occasionally, these symptoms are accompanied by severe hypoproteinemia. A direct loss of protein from cap polyps was confirmed in one 54-year-old woman by scintigraphy with the aid of technetium 99m–labeled diethylenetriaminepentaacetic acid (DTPA) complexed to human serum albumin.

Endoscopically, most cap polyps are small, sessile or semipedunculated lesions that range in size from a few millimeters to 2 cm. The most common location is the rectum or rectosigmoid; less commonly, the descending colon is involved. Rarely, cap polyps may involve the entire colon and even the stomach. Multiple polyps are typically located at the crests of mucosal folds, separated by normal or edematous mucosa.

Histologically, cap polyps are non-neoplastic lesions composed of elongated, dilated, or tortuous hyperplastic colonic crypts with abundant inflammation in the lamina propria and a characteristic “cap” of inflamed and ulcerated granulation tissue ( Fig. 22.5 ). Goblet cells of cap polyps have been shown by immunohistochemistry to express nonsulfated mucins. The intervening lamina propria typically contains increased acute and chronic inflammatory cells. The endoscopic “cap” is composed of an inflammatory exudate. Some polyps contain splayed smooth muscle fibers or fibrosis suggestive of a mucosal prolapse etiology.

Inflammatory cap polyp. This type of inflammatory pseudopolyp shows an overlying “cap” of necroinflammatory exudate. These polyps may occur in the setting of cap polyposis, at anastomotic sites, in association with inflammatory bowel disease (as in this case), or in many other conditions that induce mucosal ulceration.

Many patients with multiple polyps require surgical resection for resolution of their symptoms, although some patients improve spontaneously or with therapy to eradicate Helicobacter pylori infection. Infliximab may also be effective therapy for cap polypsis. Isolated polyps are treated by simple polypectomy.

Colitis Cystica Profunda/Polyposa

Pathologic Features

Grossly, colitis cystica profunda can be a focal, segmental, or diffuse lesion. Focal and segmental lesions can mimic invasive adenocarcinoma. Transrectal ultrasonography and other imaging modalities may aid in making the distinction by demonstrating multiple cysts limited to the submucosa and lack of lymph node involvement in the colitis cystica profunda lesion.

Histologically, the condition is characterized by the presence of multiple cystically dilated, mucin-filled crypts in the submucosa and occasionally in the muscularis propria or serosa. The stroma surrounding misplaced crypts usually consists of lamina propria; this is helpful in distinguishing it from adenocarcinoma, which typically has a desmoplastic stroma. Furthermore, misplaced glands in colitis cystica profunda often grow in a lobular configuration without jagged borders or the unusual, irregularly shaped glandular profiles characteristic of adenocarcinoma. Misplaced crypts typically show either normal or reactive-appearing colonic epithelium. A mild degree of mucin depletion, pseudostratification, and increased mitotic activity may be observed as well. However, loss of nuclear polarity, an increased nucleus-to-cytoplasm (N : C) ratio, and atypical mitoses should alert one to the possibility of adenocarcinoma. Lamina propria surrounding misplaced glands and lack of dysplastic-appearing epithelium are key features that help distinguish colitis cystica profunda from invasive adenocarcinoma ( Table 22.1 ; Fig. 22.6 ).

Table 22.1

Characteristic Features of Epithelial Misplacement (Colitis Cystica Profunda in Mucosal Prolapse and Enteritis Cystica Profunda in Peutz-Jeghers Polyps) Versus Invasive Adenocarcinoma

Feature	Misplaced Epithelium	Invasive Adenocarcinoma
Architecture	Lobular arrangement of cystically dilated, mucin-filled crypts	Irregular crypts with back-to-back configuration, may be cystically dilated; variability in size and shape of crypts
Lamina propria around crypts	Present (may show fibrosis or fibromuscular obliteration)	Absent
Hemorrhage/hemosiderin	Usually present	Usually absent
Desmoplasia	Absent	Usually present
Dysplastic cytologic features	Absent	Present
Atypical mitoses	Absent	May be present

 

Colitis cystica profunda. A, A polypoid lesion (colitis cystica polyposa) was resected endoscopically; at low power, multiple dilated, mucin-filled spaces are shown expanding the submucosa. B, A case of cystica profunda occurring in association with mucosal prolapse. The thickened muscularis mucosae can be seen at the top of the figure, where it extends around the crypts at the base of the lamina propria. Within the submucosa, at the bottom, is a cystically dilated, mucin-filled crypt. C, Keys to recognizing the benign nature of cystica profunda include the lack of cytologic atypia in the epithelium of the misplaced crypts, which may appear regenerative but maintains the usual low nucleus-to-cytoplasm ratio and basal nuclear polarity. D, Hemosiderin deposition in the stroma and the presence of lamina propria surrounding misplaced crypts are additional features that separate cystica profunda from infiltrating adenocarcinoma. E, One of the most common associations of colitis cystica profunda is with solitary rectal ulcer syndrome (occult rectosigmoid prolapse), in which case the surrounding nonpolypoid mucosa shows characteristic fibromuscular proliferation in the lamina propria.

In many patients with rectal prolapse, treatment of the defecation disorder itself (e.g., education to avoid straining at defecation, a high-fiber diet, bulk laxatives) leads to remission of colitis cystica profunda. Some patients require surgical resection because of obstruction, confusion with carcinoma, severe symptoms, or coexist­ent IBD.

Diverticular Disease–Associated Polyp

Polyps that occur in association with diverticulosis are of two types: inverted diverticula and polypoid prolapsing mucosal folds. Polyps of the former type have been described in patients with inverted Meckel diverticula of the ileum (wherein they may cause intussusception), isolated inverted colonic diverticula, or inverted sigmoid diverticula in the setting of sigmoid diverticulosis. Inverted colonic diverticula usually range from 0.2 to 2 cm. They may be sessile or pedunculated but characteristically have the same color as the surrounding mucosa. These polyps tend to vanish with gentle pressure from the biopsy forceps or from air insufflation at endoscopy. Bowel perforation may occur as a result of endoscopic polypectomy in cases in which the inverted diverticulum mimics an adenoma.

Polypoid prolapsing mucosal folds are the more common form of diverticular disease–associated polyps. Grossly, these appear as bright red, polypoid, or slightly elevated patches of mucosa in patients with sigmoid diverticulosis. Swollen mucosal folds seen between diverticular ostia may reveal an apical brown discoloration. More advanced lesions characteristically show small, brownish, polypoid protrusions of the mucosal folds. Polyps normally range from 0.5 to 3 cm.

Histologically, early lesions show vascular congestion, hemorrhage, and hemosiderin deposition. More advanced lesions reveal edema, capillary thrombi, lamina propria fibrosis, and crypt architectural changes such as dilatation and branching. The most advanced, “leaflike” polyps show changes typically seen in mucosal prolapse, such as smooth muscle ingrowth into the lamina propria, crypt hyperplasia, mucin depletion, and serrated hyperplastic changes of the epithelium. Epithelial hyperplasia may be marked and should not be mistaken for an adenoma. Rarely, one may see pseudosarcomatous changes of the stroma, characterized by enlarged and hyperchromatic myofibroblasts that should not be mistaken for sarcoma.

Redundant mucosal folds occur in a large majority (90% to 100%) of patients with advanced sigmoid diverticulosis, although only a minority develop grossly polypoid lesions or frank prolapse-like histologic alterations of the mucosa ( Fig. 22.7 ). Thickening of the taenia coli, which leads to shortening of the sigmoid, is believed to be the initiating pathogenetic event in the development of these lesions.

Diverticular disease–associated polyp. A, Redundant mucosal folds in patients with sigmoid diverticular disease are produced by thickening and shortening of the muscularis propria, as seen in this section. A more prominent redundant fold at center has the low-power appearance of a pedunculated polyp. B, Biopsy specimen of a polyp in a different patient with sigmoid diverticular disease shows features indistinguishable histologically from those of inflammatory pseudopolyps in chronic inflammatory bowel disease, including dense lamina propria chronic inflammation, surface erosion, and distortion of crypts.

Patients with diverticular disease–associated polyps may develop chronic low-grade blood loss. Treatment is similar to that for diverticulosis in general and in many instances has been shown to result in regression of polyps.

Clinical Features

Juvenile polyps occur in four distinct settings: sporadic or isolated juvenile polyps of the colon, infantile JPS, juvenile polyposis coli, and generalized JPS (involving the stomach, small bowel, and colon) ( Table 22.2 ). Sporadic or isolated juvenile polyps are the most common type of colon polyp among patients in the first decade of life, occurring in 2% of the pediatric population, but they are not uncommonly detected in adults as well. One study reported a mean age at occurrence of 5.9 years. Children usually are seen with either painless rectal bleeding or prolapse of the polyp through the rectum. Adults, if they are symptomatic at all, often are seen with rectal bleeding.

First described in 1964 by McColl and co-workers, JPS is the most common GI hamartomatous polyposis syndrome. Separation of isolated juvenile polyps in childhood from JPS is important because of their divergent clinical behaviors and neoplastic risks (discussed later). Criteria for JPS are controversial and are rapidly evolving with the advent of newer molecular diagnostic techniques. At this point, the diagnosis should be suggested in any patient who has (1) any number of polyps in probands with a family history of JPS, (2) polyposis involving the entire GI tract, or (3) three or more colonic juvenile polyps. (This last figure has ranged from 3 to 10 polyps according to various diagnostic algorithms. ) Hamartomas do not occur outside the GI tract, but congenital birth defects are reported in 15% of JPS patients; these include malrotation of the gut and cardiac and genitourinary defects.

In 1998, linkage analysis in a large Iowa kindred demonstrated that a gene for familial JPS maps to chromosome 18q21.1, a region containing both the DCC and the SMAD4 ( DPC4, MADH4 ) tumor suppressor genes. Germline truncating mutations in SMAD4 were shown to be responsible for a subset of JPS patients. In 2001, germline mutations in the gene encoding bone morphogenic protein receptor 1A, BMPR1A , were also reported in JPS families lacking SMAD4 mutations. SMAD4 mutations account for 15% to 20% of patients with JPS, and BMPR1A mutations are found in 20% to 25% of JPS patients, suggesting that there is additional (as yet undefined) genetic heterogeneity in JPS. ENG mutations have now been reported in a small number of patients with JPS. Patients harboring SMAD4 mutations have a higher risk of generalized JPS, including massive gastric polyposis. Although JPS is transmitted in an autosomal dominant fashion, 25% of newly diagnosed patients have no family history and likely represent de novo mutations.

JPS of infancy is a rare disorder with a poor prognosis that typically manifests in the first 2 years of life as generalized polyposis complicated by GI bleeding, diarrhea, rectal prolapse, and/or protein-losing enteropathy. Unlike conventional JPS, no family history is found in JPS of infancy. In 2006, Delnatte and colleagues reported that four patients with JPS of infancy were heterozygous for a de novo germline deletion of chromosome 10q, including the PTEN and BMPR1A genes, providing the first evidence of a genetic basis for this disorder.

Pathologic Features

Isolated juvenile polyps are most common in the rectosigmoid colon (54%). However, 37% of patients have polyps proximal to the splenic flexure. In juvenile polyposis coli (polyps limited to the colon), polyps are most common in the rectosigmoid region and typically number as high as 200 ( Fig. 22.9 ). The classic form of juvenile polyp ( Fig. 22.10 ), found most frequently in isolated cases, is unilobulated and has a smooth, round surface contour. Histologically, these polyps are characterized by numerous cystic and dilated, often tortuous, crypts, some filled with neutrophils and inspissated mucin (which reflects the older term, mucous retention polyp ). The intervening lamina propria is edematous and expanded by lymphocytes and plasma cells and occasional neutrophils and eosinophils. A few strands of muscle fibers may be present as well. In ulcerated cases, the epithelium may be markedly regenerative in appearance, simulating dysplasia. In patients with JPS, both classic (typical) and nonclassic (atypical) polyps are usually found. Nonclassic polyps ( Fig. 22.11 ) are often larger, multilobulated, and villiform, frequently giving the gross appearance of several polyps attached to a single stalk. Histologically, compared with typical cases, they contain less abundant lamina propria and greater epithelial overgrowth, with many elongated, tortuous, and irregularly shaped crypts. The clonal origin of both the epithelium and the fibroblasts in juvenile polyps from SMAD4 mutation carriers has been shown by fluorescence in situ hybridization.

Juvenile polyposis coli. The resected colon harbors numerous sessile and pedunculated, reddish polyps. These polyps are less numerous than those seen in familial adenomatous polyposis and are characteristically grouped in the region of the rectosigmoid.

(Courtesy of Emma E. Furth, MD, University of Pennsylvania, Philadelphia, PA.)

Classic juvenile polyps. A, These polyps grossly appear rounded, smooth, and unilobular with an erythematous cap of eroded tissue. ( A, Courtesy of Thomas C. Smyrk, MD, Mayo Clinic, Rochester, MN.) B, Their cut surface reveals multiple dilated, mucin-filled crypts, leading to the term mucus retention polyp . C, At higher power, the crypts are dilated and branched. Some, like the one at center, contain crypt abscesses, collections of neutrophils, or eosinophils. The surrounding stroma is also expanded and contains numerous mixed inflammatory cells.

Atypical juvenile polyp in juvenile polyposis syndrome. A, These polyps have a villiform architecture and frequently give the appearance of several polyps attached to the same stalk. B, Atypical polyps exhibit more epithelial overgrowth and less abundant stroma in comparison with classic juvenile polyps. In this case, there is prominent crypt branching and distortion and cytologic atypia approaching that of dysplasia.

Patients with JPS may have a combination of other types of polyps as well. Ganglioneuromatous proliferation of mature ganglion cells and nerve bundles may occur in the mucosa and the submucosa of polyps. Many patients also have separate adenomas or polyps with combined features of a juvenile polyp and an adenoma/dysplasia. In fact, true dysplasia ( Fig. 22.12 ) has been detected in as much as 30% of polyps of JPS patients, but it is almost never observed in patients with isolated juvenile polyps. Gupta and associates reported dysplasia in only 1 of 331 juvenile polyps from 184 nonpolyposis patients.

Dysplasia arising in a juvenile polyp. A, A classic juvenile polyp contains an area of hyperchromatic crypts at top. B, At higher power, the involved crypts and surface epithelium, which contain enlarged, hyperchromatic, and crowded nuclei, are cytologically indistinguishable from an adenoma. The crypt at center shows a mixture of dysplastic and nondysplastic epithelium.

Natural History and Treatment

Patients with sporadic juvenile polyps have no increased risk for malignancy. Furthermore, they are not predisposed to the development of new juvenile polyps, and they do not require any particular type of follow-up. In contrast, estimates of the risk of GI carcinoma in JPS have varied widely. In one large kindred, the risk of upper and/or lower GI cancer in affected patients was 55%. In 1995, Desai and colleagues, in a reevaluation of data from the St. Mark’s Polyposis Registry, projected that the incidence of colorectal carcinoma through age 60 years was 68% (see Table 22.2 ). Surveillance recommendations should take into account the risks of both lower and upper tract cancer, including gastric cancer. Upper and lower endoscopy (including visualization of the entire colon) is usually recommended by age 15 years and should be repeated annually. Endoscopic polypectomies and histologic examination of all removed polyps should be performed until the patient is free of polyps, at which point the surveillance interval may be lengthened to 3 years. Consideration should be given to prophylactic gastrectomy or colectomy if diffuse polyposis cannot be controlled by endoscopic polypectomy or if there is a family history of GI carcinoma or dysplasia.

Clinical Features

PJS is an autosomal dominant syndrome that is characterized by mucocutaneous pigmentation; distinctive hamartomatous polyps of the small intestine, colon, and stomach; and an increased risk for both intestinal and extraintestinal malignancies (see Chapters 20 and 21 for details). Estimates of the incidence of PJS vary widely, from 1 in 50,000 to 1 in 200,000 births. Significant phenotypic variability can occur within PJS kindreds, and patients may not have a family history of PJS. Most patients seek treatment during the second or third decade of life. In one study, the average age at presentation was 18 years, with a range from 2 to 62 years. Peutz-Jeghers polyps have also been described in neonates. The main clinical manifestations of PJS include mucocutaneous pigmentation (most characteristically on the vermilion border of the lips), GI polyposis, and an increased risk for GI carcinomas, as well as for extraintestinal benign and malignant tumors.

In 1997, the PJS locus was mapped to chromosome 19p13.3; this was achieved through the use of comparative genomic hybridization of PJS polyps in one patient, followed by targeted linkage analysis in 12 affected families. In 1998, two separate reports identified the STK11 (formerly LKB1 ) gene on 19p13.3 as the main causative gene in PJS. Germline mutations in the STK11 tumor suppressor gene, which encodes a novel serine threonine kinase, disrupt the function of the protein’s kinase domain. Overall, however, STK11 mutations have been found in only approximately 60% of familial and 50% of sporadic cases, suggesting the possibility of genetic heterogeneity in PJS.

Pathologic Features

Hamartomatous polyps of PJS may occur throughout the GI tract but are most common in the small bowel (65% to 95%), followed by the colon (60%) and stomach (20% to 50%). Involved segments of bowel may harbor from 1 to 20 polyps, and these may vary in size from 0.5 to 3 cm. Well-developed polyps located in the small intestine and colon tend to be pedunculated.

Microscopically ( Fig. 22.13 ), the most prominent component is a distinctive, arborizing pattern of smooth muscle derived from the underlying muscularis mucosae. The lamina propria is usually normal in composition and cellularity and does not typically demonstrate increased inflammation. The overlying epithelium is composed of the normal epithelial cells typical of the involved segment of the GI tract. In most colonic Peutz-Jeghers polyps, the epithelium shows epithelial overgrowth and areas of hyperplasia, but dysplasia is unusual. Nevertheless, nondysplastic epithelium has been shown to be clonal.

Peutz-Jeghers polyps. A, Colonic Peutz-Jeghers polyp shows the characteristic overgrowth of epithelium and an arborizing smooth muscle core. B, The branching smooth muscle is derived from the muscularis mucosae. C, The overlying mucosa is typically nondysplastic and maintains its normal architecture. In contrast to juvenile polyps, the mucosa has a normal ratio of lamina propria to epithelium and the crypts are not cystically dilated. Cell types normal for that region of the gastrointestinal tract (in this case, goblet cells and absorptive cells) are found within the epithelium.

Studies of histologic and molecular alterations have shown that a subset of polyps progress through a sequence of hamartoma, adenoma (dysplasia), and then carcinoma. Somatic loss of the 19p13.3 allele occurs in a substantial portion of nondysplastic polyps; this (in conjunction with germline mutation at this locus) accounts for hamartoma formation. Dysplasia and carcinoma arise in hamartomas through the acquisition of additional genetic alterations at loci such as TP53 and β-catenin. Hizawa and co-workers documented “adenomatous” (dysplastic) foci in 9 (12.7%) of 71 Peutz-Jeghers hamartomas in the GI tract. In a study of 52 colonic Peutz-Jeghers polyps, Narita and associates found dysplasia in 3 (5.8%). Intramucosal and invasive carcinomas have been documented in Peutz-Jeghers polyps of the stomach, small bowel, and colon. Carcinomas are more likely to arise in patients with familial PJS. Only a single case report has described the occurrence of carcinoma in a nonsyndromic setting.

Some Peutz-Jeghers polyps show foci of epithelial misplacement, pseudoinvasion, or herniation of benign epithelium into the intestinal wall (all synonymous for the phenomenon of enteritis cystica profunda) ( Fig. 22.14 ). Shepherd and colleagues studied 491 Peutz-Jeghers polyps and reported epithelial misplacement in 10% of small intestinal polyps; however, this phenomenon was not observed in gastric or colonic polyps (see Chapters 20 and 21 ). Polyps with epithelial misplacement are characterized by intramural mucinous cysts that can involve all layers of the bowel wall, including the serosa. Grossly and histologically, these polyps can simulate invasive adenocarcinoma. Helpful histologic discriminators between pseudoinvasion and true carcinoma include the following: (1) lack of cytologic atypia in the deep glands, with a normal composition of epithelial cell types; (2) hemosiderin deposition; and (3) the presence of deep mucinous cysts in pseudoinvasion. Another potential pitfall is the overdiagnosis of misplaced dysplastic epithelium as invasion. Helpful discriminating features in these cases include the presence of both nondysplastic and dysplastic epithelium within the dilated misplaced crypts, the presence of lamina propria around the displaced crypts, and the lack of a desmoplastic response in noninvasive cases.

Epithelial misplacement in Peutz-Jeghers polyps. A, At low power, epithelium in the stalk of a Peutz-Jeghers polyp mimics infiltrating carcinoma. Dilated crypts are found within and deep to the muscularis propria ( center ). B, A dilated, mucin-filled crypt is seen in the muscularis propria. An epithelial lining with lamina propria is seen at right. C, In addition to mucin pools and lamina propria, histologic clues to the benign nature of the misplaced crypts include the benign cytology of the involved crypts, which contain a normal mix of goblet cells and absorptive cells.

Natural History and Treatment

Non-neoplastic GI complications of PJS include GI bleeding and abdominal pain related to intussusception or luminal obstruction by large polyps. Intussusception is frequent (occurring in 47% of 222 Japanese PJS patients in a series by Utsunomiya and co-workers ) and is most common in the small intestine, caused in part by the pedunculated nature of the small bowel and colonic polyps. Prolapse of pedunculated PJS polyps through the rectum may also occur.

Neoplastic complications of PJS include an increased frequency of both GI and extraintestinal neoplasms. Numerous early case reports attest to the increased risk of gastric, small intestinal, and colonic malignancy in PJS, although the absolute risk was debated and was initially thought to be relatively low. Subsequent studies suggested a high risk for cancer development. In a retrospective study from the Mayo Clinic spanning 50 years, noncutaneous cancers developed in 18 (53%) of 34 PJS patients, with a mean age of 39.4 years at the time of cancer diagnosis. In particular, the relative risk of GI carcinoma (predominantly colon cancer) in this study was 50.5, and in women with PJS, the relative risk of breast and gynecologic cancers was 20.3. In a meta-analysis of cancer risk in PJS, Giardiello and associates calculated a cumulative risk for malignancy of 93% in patients 15 to 64 years of age. In that study, the cumulative risks were as follows: breast cancer, 54%; colon cancer, 39%; pancreatic cancer, 36%; gastric cancer, 29%; ovarian cancer, 21%; and small intestinal cancer, 13% ( Table 22.3 ). Furthermore, significantly increased relative risks were found for cancers of the following organs: esophagus (57), stomach (213), small intestine (520), colon (84), pancreas (132), lung (17), breast (15.2), uterus (16), and ovary (27). Although a statistically significant increased risk for either testicular cancer or cervical cancer could not be demonstrated in Giardiello’s study, distinctive tumors in these sites have been reported in PJS. A study by Lim and colleagues of 240 PJS patients harboring germline mutations in LKB1 also reported high cancer risks for these patients. By the age of 60 years, the risk for pancreatic carcinoma in PJS patients was 8%, the risk for breast cancer was 32%, and the risk for colorectal cancer was 30%.

Surveillance in PJS is indicated for removal of GI polyps and for the early detection of both GI and extraintestinal neoplasms. Surveillance recommendations for the GI tract include (1) colonoscopy every 2 to 3 years beginning at age 18 years and (2) upper endoscopy with either push enteroscopy or double-contrast radiology for imaging of the entire small bowel every 2 to 3 years beginning at age 18 years. In women, monthly breast self-examinations should begin at age 18 years, and semiannual clinical breast examinations and annual mammography should begin at age 25 years. Minimum gynecologic surveillance recommendations include annual pelvic examinations and Papanicolaou smears in women and annual testicular examinations in men. Because of the clearly documented increased risk for pancreatic carcinoma, some authors also advocate pancreatic ultrasound examinations every 1 to 2 years.

Clinical Features

Cowden syndrome is an autosomal dominant hamartoma/neoplasia syndrome that bears the family name of the original patient described by Lloyd and Dennis in 1963. In 1996, the susceptibility locus for Cowden syndrome was mapped to chromosome 10q22-23, and in 1997, germline mutations in the PTEN gene (phosphatase and tensin homolog, deleted on chromosome 10) on 10q23 were first reported in families with this syndrome. Since that time, mutations of the PTEN tumor suppressor gene or its promoter have been identified in 80% to 85% of patients who meet the strict operational diagnostic criteria set forth by the International Cowden Consortium.

Hamartomas in Cowden syndrome affect all three germ cell layers but most commonly arise from ectodermal and endodermal elements. Almost all patients (90% to 100%) have mucocutaneous lesions that include trichilemmomas, acral keratoses, and oral papillomas. Breast lesions affect the majority of female patients and include fibroadenomas, fibrocystic disease, and adenocarcinomas (25% to 50%). Breast carcinoma is occasionally reported in men as well. Thyroid abnormalities, such as multinodular goiter and follicular adenoma, are found in one half to two thirds of patients. Thyroid carcinoma occurs in 3% to 10% of patients. Macrocephaly, cerebellar gangliocytoma, and genitourinary malformations are also frequent components of Cowden syndrome. Increased risks for endometrial carcinoma and renal cell carcinoma have been added to the operational criteria for Cowden syndrome.

Pathologic Features

GI hamartomas are found in only 35% to 40% of patients with Cowden syndrome. Hamartomatous polyps may affect the stomach, small bowel, and colon ( Fig. 22.15 ). The most common esophageal manifestation is glycogenic acanthosis. Most polyps resemble juvenile polyps and consist of cystically dilated, mucin-filled crypts set within an abundant lamina propria. Colonic lipomas, fibrolipomas, fibromas, ganglioneuromas, and adenomas have also been reported in Cowden syndrome. Among these polyps, polypoid ganglioneuromas are most specific for Cowden syndrome. Loss of expression of the wild-type PTEN allele, coupled with the germline PTEN mutation, was demonstrated in both GI hamartomas and colonic adenomas of several patients with Cowden syndrome.

Colonic polyp sampled from a patient with Cowden syndrome. There is mild architectural distortion of the crypts, which are arranged in a slightly fibrotic stroma. The epithelium is nondysplastic and contains the usual composition of goblet cells and absorptive cells found in the colon.

Natural History and Treatment

Recent studies have demonstrated that the lifetime risk of colorectal adenocarcinoma in patients with Cowden syndrome with PTEN mutations is 10% to 15%. For this reason, endoscopic surveillance of the GI tract has been recommended. Perhaps more importantly, screening and surveillance are directed at breast and thyroid malignancies. Recommendations for women include monthly breast self-examinations, annual clinical breast examinations beginning in late adolescence, and mammography beginning at age 25 years. Annual clinical examination of the thyroid in both sexes should begin in late adolescence with or without parallel use of thyroid ultrasonography every 1 to 2 years.

Clinical Features

Cronkhite-Canada syndrome is a nonhereditary polyposis syndrome of unknown etiology (see also Chapter 17 ). Since the first description of this syndrome by Cronkhite and Canada in 1955, hundreds of cases have been reported. Although it is worldwide in distribution, most cases originate from Japan, Europe, or the United States. Unlike most polyposis syndromes, most patients are seen in middle adulthood (mean age, 59 years). Infantile cases of Cronkhite-Canada syndrome are only rarely reported.

Colonic polyps associated with the Cronkhite-Canada syndrome occur in association with distinct ectodermal abnormalities, such as alopecia of both the scalp and body hair, dystrophy of the nails, and skin hyperpigmentation. Skin lesions are usually manifested by the presence of light to dark brown macules on the extremities, face, neck, palms, and soles. Affected patients often are seen with a variable combination of diarrhea, weight loss, nausea and vomiting, anorexia, and GI bleeding. In some cases, diarrhea may result in severe electrolyte abnormalities, with the development of seizures or tetany.

Pathologic Features

Polyps in Cronkhite-Canada syndrome occur throughout the GI tract but spare the esophagus. Histologically, in the colon, they closely resemble juvenile polyps, with cystically dilated, tortuous crypts containing inspissated mucin ( Fig. 22.16 ). The intervening lamina propria is edematous and contains increased numbers of mononuclear cells and eosinophils. Burke and Sobin compared the histology of Cronkhite-Canada polyps with that of juvenile polyps. The histologic features of each type are microscopically indistinguishable, although colonic juvenile polyps are more frequently pedunculated. However, in patients with Cronkhite-Canada syndrome, the intervening nonpolypoid intestinal mucosa shows pathologic changes such as edema, cystically dilated glands, and increased inflammatory cells (see Fig. 22.16 ), whereas the intervening mucosa in JPS is normal.

Cronkhite-Canada syndrome. A, Biopsy specimen of one of numerous colonic polyps in a patient with Cronkhite-Canada syndrome shows dilated and branched crypts within an expanded lamina propria. B, At higher power, cystically dilated crypts are filled with mucin; the surrounding lamina propria is edematous and contains increased numbers of mononuclear cells. C, Although individual Cronkhite-Canada polyps are histologically indistinguishable from juvenile polyps, biopsy of the intervening nonpolypoid colonic mucosa in a patient with Cronkhite-Canada syndrome shows lamina propria edema and scattered cystically dilated crypts.

Natural History and Treatment

The malignant potential of colonic Cronkhite-Canada polyps is controversial. There have been multiple reports of affected patients who have also had adenoma, colorectal carcinoma, or gastric cancer. Although some authors indicate that dysplastic changes may occur within the polyps, it is not clear whether these might represent separate incidental adenomas. The mortality rate in Cronkhite-Canada syndrome is high (50% to 60%). Death usually results from malnutrition, GI bleeding, or infection. Treatment includes aggressive nutritional support, antibiotics, corticosteroids, and surgical resection of symptomatic segments of the GI tract.

Nondysplastic Serrated Polyps

The terminology and classification of serrated colonic polyps is controversial. The scheme used here, and outlined in Box 22.1 , is based on the most recent WHO classification. It is clinically relevant and is focused on separating dysplastic from nondysplastic lesions. The growth pattern (architecture) and proliferative characteristics are also important parameters to evaluate in distinguishing these polyps. An excellent review on serrated lesions of the colorectum was published by Rex and colleagues in 2012.

Hyperplastic Polyp

Clinical Features.

“Traditional” hyperplastic polyps are serrated polyps with normal architecture and proliferative characteristics. They are small, innocuous lesions that may be found throughout the colon of adults but are especially common in the rectum. As much as 90% occur in the left colon. The prevalence rate increases with age. As much as 35% of asymptomatic individuals older than 50 years of age demonstrate hyperplastic polyps. Specific lifestyle and dietary factors commonly associated with adenomas are also associated with hyperplastic polyps. These include cigarette smoking, alcohol consumption, low folate intake, and obesity. These lesions are typically asymptomatic and are often identified as an incidental finding during endoscopic examination of the colon. Endoscopically, hyperplastic polyps are small, sessile, smooth bumps or nodules that appear pale and flatten on insufflation of air. In some instances, these lesions cannot be distinguished with certainty from adenomas endoscopically; therefore, patients typically undergo biopsy and fulguration.

Classification and Pathogenesis.

In a study by Torlakovic and co-workers, small left-sided “hyperplastic” polyps were classified into three general types based on their morphologic growth pattern and lack of proliferative or maturation abnormalities—the microvesicular, goblet cell, and mucin-poor types ( Tables 22.4 and 22.5 ). These polyps are normally smaller than 0.5 cm and show crypt epithelial serration but do not possess the architectural and proliferative abnormalities of large atypical hyperplastic polyps (SSA/Ps). The microvesicular type is the most common and correlates with lesions considered by most pathologists to represent a typical hyperplastic polyp. It is believed that these polyps develop as a result of increased crypt epithelial cell turnover combined with delayed maturation from the crypt bases to the surface of the mucosa. Because of decreased apoptosis and a slower rate of surface cell exfoliation, these lesions acquire a hypermature appearance and their characteristic histologic serration. Crypt cell kinetic studies have shown that hyperplastic polyps do not, in fact, exhibit epithelial cell hyperplasia. These polyps, at least the microvesicular type, contain various genetic and cell cycle regulatory defects, and this has led to the belief that these lesions have malignant potential. Some authorities assert that some SSA/Ps arise from microvesicular hyperplastic polyps, but this belief is controversial. Hyperplastic polyps have also been shown to demonstrate dysregulation of cell proliferation; altered BCL2 and BAX gene expression; loss of heterozygosity (LOH) of chromosome 1p, the APC gene, chromosome 3p, and CDKN2A ; TP53 overexpression; a high frequency of KRAS mutations; and altered CDKN1B expression. A variety of biochemical and oncofetal blood antigen expression patterns are altered in hyperplastic polyps as well.

Table 22.4

Pathologic Features of Serrated Colonic Polyps

Feature	Nondysplastic Serrated Polyps				Dysplastic Serrated Polyps
	Microvesicular	Goblet Cell	Mucin-Poor	Sessile Serrated Adenoma/Polyp	Sessile Serrated Adenoma/Polyp with Dysplasia	Traditional Serrated Adenoma
Size	<0.5 cm	<0.5 cm	<0.5 cm	>0.5 cm	>0.5 cm	>0.5 cm
Location	↑ left colon	↑ left colon	↑ left colon	↑ right colon	↑ right colon	↑ left colon
Sessile	+	+	+	++	±	±
Pedunculated	−	−	−	±	±	++
Eosinophilic cells	−	−	−	± (focal)	+	++
Abnormal crypt architecture	−	−	−	++	++	+
Basal crypt serration	−	−	−	++	++	±
Surface maturation	+	+	−	±	−	−
Dystrophic goblet cells	±	−	±	++	++	−
Goblet cells	+	++	−	+	+	+
Mucin rich	++	−	±	+	+	−
Neuroendocrine cells	++	++	++	±	±	±

 

Table 22.5

Molecular Features of Serrated Colonic Polyps

Feature	Nondysplastic Serrated Polyps				Dysplastic Serrated Polyps
	Microvesicular	Goblet Cell	Mucin-Poor	Sessile Serrated Adenoma/Polyp	Sessile Serrated Adenoma/Polyp with Dysplasia	Traditional Serrated Adenoma
Upper crypt proliferation	Absent	Absent	Low	Present	Present	Variable
BRAF mutation	High	Low	High	High	High	Low
KRAS mutation	Rare	High	Rare	Low	Moderate	Low
Hypermethylation	Moderate	Low	Moderate	High	High	Moderate
Microsatellite instability	Absent	Absent	Absent	Rare	Low	Rare

 

Little is known about the molecular characteristics of the mucin-poor type of hyperplastic polyp. However, goblet cell hyperplastic polyps typically show low levels of DNA methylation (CIMP) and absence or low levels of BRAF or APC mutations. In contrast, KRAS mutations have been identified in as much as 54% of cases, whereas TP53 is uniformly negative. In contrast, microvesicular hyperplastic polyps show DNA methylation (CIMP) in 68% of cases, and almost 80% show BRAF mutations. A small proportion show MGMT loss (14%). However, in contrast to goblet cell hyperplastic polyps, only a small proportion (4%) shows KRAS mutations. All three types of hyperplastic polyps are typically microsatellite stable.

There is abundant epidemiologic and morphologic evidence to suggest that hyperplastic polyps, at least the microvesicular type, may progress to carcinoma. Affected patients share risk factors with patients in whom colon cancer ultimately develops. Patients with hyperplastic polyps are also more likely to have adenomas than patients without these lesions. Foci with the appearance of microvesicular, or even goblet cell, hyperplastic polyps may be seen adjacent to, or within, portions of SSA/Ps, which are lesions considered to have higher malignant potential. On rare occasions, dysplasia, of either the conventional or the serrated adenomatous type, may be seen in association with microvesicular or even goblet cell hyperplastic polyps. Some traditional serrated adenomas (TSAs) show goblet cell hyperplastic polyp features in adjacent mucosa. Certainly, an increased risk of colon cancer has been detected in patients with SPS (see later discussion).

Pathologic Features.

Grossly, most hyperplastic polyps measure less than 0.5 cm. All types are characterized histologically by a “mature” appearance and display an irregular sawtooth or luminal serrated surface contour. This feature is most pronounced in the upper portions of the crypts but in some cases may also involve the middle and basal portions of the crypts. These polyps show simple tubular architecture without branching, budding, or crypt dilatation; they consist of either mature goblet cells or a combination of goblet cells and mucous cells. These lesions characteristically show proliferation (mitoses or Ki67 staining) limited to the lower half of the crypts. Most hyperplastic polyps also show increased numbers of neuroendocrine cells at the bases of the crypts.

The microvesicular type of hyperplastic polyp is characterized by the presence of abundant (microvesicular) mucin and a paucity of goblet cells, compared with normal mucosa ( Fig. 22.17 ). These lesions are more common in the left colon and are characterized by marked luminal serration and an expanded proliferative compartment that occupies most of the basal half of the crypts. Nuclear atypia is minimal, but some polyps show a mild degree of nuclear stratification at the base of the polyp. Dystrophic goblet cells and cells with round vesicular nuclei and prominent nucleoli are rare; if present, they are usually confined to the lower portions of the crypts. Most notably, these lesions show distinct evidence of surface maturation and absence of eosinophilic cytoplasm characteristic of TSAs (see later discussion). The superficial portion may show hypermucinous cells that may appear crowded focally. However, the cells maintain a mature appearance with small nuclei and absence of hyperchromasia or atypia. The deep regenerative portion of a microvesicular hyperplastic polyp may be confused with an adenoma because the cells may appear hyperchromatic compared with surrounding nonlesional colonic crypts. Nevertheless, the hyperchromatic regenerative zone maintains an orderly process of maturation to the surface, and this can be easily exemplified with the use of Ki67 or proliferating cell nuclear antigen (PCNA) staining. Some cases also show thickening and extension of the muscularis mucosae into the lamina propria.

A, Hyperplastic polyp, microvesicular type, shows marked luminal serration and a mixture of goblet cells and columnar cells with microvesicular mucin. B, These polyps show surface maturation with basally located small nuclei and abundant microvesicular mucinous cytoplasm. Notice the absence of mitoses in the upper crypt and surface epithelium.

Rarely, larger microvesicular hyperplastic polyps (>0.5 cm) may be slightly distorted or show mild crypt dilatation. In this instance, the features approach those of an SSA/P, and this has led some investigators to hypothesize that the microvesicular type of hyperplastic polyp is an early form of SSA/P. Both of these types of polyps show frequent BRAF mutations and DNA methylation abnormalities. Microvesicular hyperplastic polyps may, uncommonly, show misplacement of epithelium into the submucosa, but this is more common in SSA/Ps. This phenomenon has been termed inverted hyperplastic polyp, and it may be mistaken for an adenoma with epithelial misplacement or even an invasive adenocarcinoma. Some authorities have postulated that this phenomenon occurs as a result of trauma and subsequent protrusion of epithelium into the submucosa, similar to the pathogenesis of epithelial misplacement in adenomas.

Goblet cell hyperplastic polyps are the second most common type of hyperplastic polyp. They are typically small (<0.5 cm), sessile lesions and are more commonly found in the left colon. These polyps show elongated crypts that are rich in goblet cells but without microvesicular mucin ( Fig. 22.18 ). They are normally sessile but demonstrate a far lesser degree of luminal serration, which is often limited only to the surface and uppermost portions of the crypts. The crypts may be mildly elongated as well. Nuclear atypia, stratification, and mitoses are not seen in this type of hyperplastic polyp. KRAS mutations are common, but they only rarely show evidence of DNA methylation.

Goblet cell–rich hyperplastic polyps show less luminal serration compared with the microvesicular type. The epithelium contains abundant goblet cells and lacks microvesicular mucin.

Mucin-poor hyperplastic polyps are the least common type. Some authorities believe that they represent microvesicular hyperplastic polyps that have undergone irritation, mucin depletion, and inflammation as a result of injury or prolapse. These lesions are also sessile, but they exhibit small cells with less cytoplasm, micropapillary architecture, mucin depletion, absence of goblet cells, and a striking regenerative appearance ( Fig. 22.19 ). The lack of mucin and the presence of hyperchromatic nuclei are the main features of these polyps. Inflammation in the lamina propria is characteristic. The molecular features of this subtype of hyperplastic polyp are poorly understood.

The epithelium of a mucin-poor hyperplastic polyp has a regenerative appearance, with hyperchromatic nuclei, mucin depletion, micropapillary architecture, and a distinct lack of goblet cells.

Differential Diagnosis.

Hyperplastic polyps should be differentiated from SSA/Ps, TSAs, prolapse-type inflammatory polyps, and conventional adenomas. In contrast to hyperplastic polyps, SSA/Ps show dilated crypts, prominent full crypt serration, basal crypt dilatation with horizontal and branched crypts, increased numbers of dystrophic goblet cells, focal nuclear stratification, and upper crypt mitoses. However, some microvesicular polyps may show one or more of these features focally; in these instances, correlation with the size, endoscopic appearance, and location of the lesion may be helpful. If the lesion is larger than 0.5 cm, sessile, and present in the right or transverse colon, it most likely represents an early SSA/P.

TSAs are usually easy to differentiate from hyperplastic polyps because the former are larger, are often villiform, and show marked cytoplasmic eosinophilia at all levels of the crypt. Nuclear stratification is typical.

Prolapse-type inflammatory polyps are frequently inflamed or even ulcerated. They may have superficial ischemic-type features, contain markedly regenerative epithelium, and show characteristic fibromuscular hyperplasia of the lamina propria.

Conventional adenomas have hyperchromatic, pseudostratified, cigar-shaped (pencillate), atypical nuclei. In adenomas with misplacement, the crypts also show atypical features, with frequent mitoses and even single-cell necrosis.

Natural History and Treatment.

Until recently, small hyperplastic polyps were not believed to require definitive treatment, although they are typically removed in the process of endoscopy with biopsy. However, the biologic characteristics and natural history of hyperplastic polyps, particularly the microvesicular type, are controversial. Certainly, large lesions that contain some features of SSA/Ps, particularly if they are located in the right colon, should be removed in total.

Further studies are needed to determine the natural history and malignant potential of hyperplastic polyps and each of the different subtypes. There is a strong correlation between the occurrence of hyperplastic polyps and that of adenomas in similar populations in the world. Although some studies suggest that distally located hyperplastic polyps predict an increased risk of proximal adenomas, or even advanced adenomas, this idea is controversial, and at present, the finding of a distal hyperplastic polyp at sigmoidoscopy is not considered to be an absolute indication for full colonoscopy. Furthermore, surveillance for detection of metachronous hyperplastic polyps or adenomas after removal of hyperplastic polyps is not recommended.

Sessile Serrated Adenoma/Polyp

Clinical Features and Synonyms.

In contrast to left-sided “hyperplastic” polyps, right-sided “serrated” polyps are often larger (>0.5 cm) and sessile and contain significant architectural, proliferative, and maturation abnormalities. Crypt distortion and abnormalities in proliferative characteristics, such as asymmetry in proliferation and lack of a readily identifiable proliferative zone, are hallmarks of these lesions. Endoscopically, these lesions show a smooth surface contour, often covered with mucus. Some polyps show a characteristic pattern of orderly arranged circular dots similar to that of the surrounding normal mucosa; they can be identified and reliably distinguished from adenomas with a relatively high degree of accuracy by using high-resolution chromoendoscopy.

There is ongoing debate regarding the terminology, diagnostic features, natural history, and risk of malignancy of this family of serrated polyps. Other commonly used terms to describe this lesion include atypical hyperplastic polyp, large hyperplastic polyp, hyperplastic polyp with architectural abnormalities, hyperplastic polyp with dysmaturation , and hyperplastic polyp with abnormal proliferation . However, based on the most recent edition of the WHO classification of digestive tumors, the recommended terms are sessile serrated adenoma or sessile serrated polyp (i.e., SSA/P). Both of these terms are considered acceptable in clinical practice. We prefer the term sessile serrated polyp (SSP) for several reasons, most importantly because most clinicians (and pathologists) associate the term “adenoma” with dysplastic changes that SSA/Ps do not possess until they become dysplastic.

As mentioned earlier, SSA/Ps show a high rate of DNA methylation (CIMP), approaching 92% in some studies, as well as BRAF mutations (80%), but a low prevalence of APC or KRAS mutations (<5%) and TP53 abnormalities. Microsatellite instability is not usually present until morphologic dysplasia, or carcinoma, has developed. Large lesions, or those with dysplasia, may also show loss of MLH1, MSH2 , or MGMT , particularly in the left colon. Some studies have shown increased MUC2, MUC5AC, and MUC6 staining as well.

Pathologic Features.

Morphologically, SSA/Ps show distinctive features characterized by crypt dilatation, crypt irregularity (horizontally shaped crypts), prominent and often exaggerated lower (and upper) crypt serration, mitoses in the upper levels of the crypts, vesicular nuclei in the upper crypts, reduced amounts of lamina propria between crypts, hypermucinous epithelium, and, occasionally, an inverted (epithelial misplacement) growth pattern ( Fig. 22.20 ; see Tables 22.4 and 22.5 ). The basal portion of the crypts are usually branched and appear flask or boot-shaped (similar to an inverted T), indicative of horizontal rather than vertical growth. Although their conclusion was controversial, a recent expert panel of pathologists suggested that the presence of one or more distorted crypts of this kind, at the base of the lesion, is enough evidence for diagnosis of an SSA/P rather than a hyperplastic polyp. These are sessile lesions, rich in mucin, and they contain round vesicular nuclei with prominent nucleoli. Mature goblet cells, or mucinous cells, show an irregular distribution by being uncharacteristically located at the bases of the crypts. Some polyps also show thickening of the basement membrane and perineurial-like proliferations in the lamina propria.

A, Sessile serrated adenoma/polyp showing prominent serration at all levels of the crypts; crypt dilatation; horizontally shaped crypts, particularly evident at the base of the mucosa; and abundant goblet cells, some dystrophic. B, The basal portions of the crypts are branched and appear flask shaped. The epithelium contains an irregular distribution of mature and dystrophic goblet cells and mucinous cells. Focally ( center ) the crypt nuclei show atypia with cytoplasmic eosinophilia, nuclear stratification, and loss of cell polarity, similar to a serrated adenoma. C, More advanced sessile serrated adenoma/polyp shows markedly hyperserrated and micropapillary architecture at all levels of the crypts. D, High-power view of the polyp in C shows a greater degree of nuclear stratification and atypia than in A, but the cells lack the hypereosinophilic cytoplasm characteristic of traditional serrated adenomas. These polyps have also been referred to as “mixed” polyps.

SSA/Ps may show a variable degree of nuclear atypia. At the low end of the spectrum, lesions show little or no stratification, low mitotic rate, and evidence of surface maturation. At the high end, SSA/Ps may show an increased degree of nuclear stratification, consisting of cells with open vesicular chromatin and prominent nucleoli, increased eosinophilia of the cytoplasm (simulating the cellular appearance of a TSA), marked hyperserration, and lack of surface maturation. These features approach those of a serrated adenoma but without long filiform, or villous, projections. Often these lesions grow horizontally, and they may measure more than 1 to 2 cm. Some authors have referred to this lesion as advanced SSA/P, but this term actually refers to SSA/P with dysplasia (see later discussion) and is not recommended in clinical practice.

Dysplastic Serrated Polyps

Sessile Serrated Adenoma/Polyp with Dysplasia

With histologic and molecular progression (often characterized by loss of MLH1 immunostaining and microsatellite instability), SSA/Ps may acquire morphologic evidence of dysplasia that leads to a variable histologic appearance (see Tables 22.4 and 22.5 ). Histologically, these lesions have been referred to as “mixed hyperplastic/adenomatous polyps”; however, it is now well recognized that these polyps do not occur as a result of a coincidental growth of hyperplastic and adenomatous elements but as a result of dysplastic change within an SSA/P. Therefore, the term “mixed polyp” is not recommended. Morphologically, these lesions may show discrete areas of hyperplastic change typical of a microvesicular polyp, but more often they show features of an SSA/P combined with dysplasia ( Fig. 22.21 ). Dysplastic changes often resemble a conventional adenoma, composed of elongated, cigar-shaped hyperchromatic nuclei with stratification and architectural serration. This is referred to as “intestinal-type” dysplasia. In some SSA/Ps, dysplastic changes show features of a TSA (“serrated dysplasia”), being composed of open, oval, or slightly elongated nuclei with prominent nucleoli, nuclear stratification, eosinophilic “pink” cytoplasm, and lack of surface maturation ( Fig. 22.22 ). Some polyps may show a combination of conventional (intestinal) and serrated dysplastic features. Rarely, a combination of hyperplastic or SSA/P–appearing epithelium, dysplastic epithelium, and adenocarcinoma may be present in a single polyp.

Sessile serrated adenoma/polyp with dysplasia. The left side of the field shows features of a sessile serrated polyp with prominent serration. The right side of the field shows conventional adenoma–type low-grade dysplasia. These lesions have also been referred to as mixed hyperplastic/adenomatous polyps.

A, Less frequently, dysplasia (left side of field) arising in a sessile serrated adenoma/polyp (right side of field) may have a traditional serrated adenoma appearance. B, High-power view shows an abrupt transition from “hyperplastic” epithelium on the right to dysplastic epithelium on the left with stratified nuclei and hypereosinophilic cytoplasm reminiscent of a traditional serrated adenoma.

Natural History and Treatment.

The natural history and risk of progression to malignancy of SSA/P are poorly understood but are under intense investigation. It is commonly believed that loss of DNA repair ability and subsequent microsatellite instability are mechanisms that drive rapid neoplastic progression. Although SSA/Ps may contain dysplasia or even carcinoma, in most cases, the acquisition of dysplasia or carcinoma is related to the size of the lesion and take years to progress. Therefore, it is likely that the risk of progression to malignancy varies significantly with the size and location of the lesion. In one study of 55 patients with SSA/P who were monitored for a mean of 7 years, colorectal cancer or adenomas with high-grade dysplasia developed in 15% of patients. In other studies, patients with SSA/P were found to be at increased risk for synchronous advanced neoplasia and interval neoplasia during surveillance. Some recent studies have suggested that the potential for malignant transformation is similar to that of conventional tubular adenomas. Nevertheless, a rational proposal for the management of serrated polyps is based on the fact that most traditional hyperplastic polyps (goblet cell or microvesicular type) are unlikely to progress to carcinoma and that SSA/Ps without cytologic dysplasia are probably slow, but progressive, lesions.

Chapter 2 provides a summary of recommendations of surveillance intervals after endoscopic resection of serrated lesions. For SSA/Ps without dysplasia, complete endoscopic removal is warranted. However, if this is not possible, then repeat endoscopy with biopsies should be performed within 1 year to evaluate for signs of cytologic dysplasia, and this should be followed by continued surveillance at shorter intervals until the lesion is completely removed. Surgical excision for large SSA/Ps without dysplasia, particularly those that are recurrent, is also a reasonable alternative. Many investigators recommend surveillance every 5 years for SSA/Ps that are smaller than 10 mm in size; every 3 years for lesions that are larger than 10 mm as well as those that are smaller than 10 mm but multiple (three or more polyps); and every 1 to 3 years for lesions that are larger than 10 mm and multiple (two or more polyps) or for any polyp with dysplasia. In contrast, complete excision is considered mandatory for SSA/Ps with dysplasia, based on the likelihood that these lesions have undergone hypermethylation and microsatellite instability and therefore are prone to carcinoma progression. Complete excision should be accomplished by endoscopy or surgical resection. Patients usually undergo repeat endoscopy within 1 year to ensure that the lesion has been removed in its entirety and that there has been no progression.

Traditional Serrated Adenoma

TSAs are broadly defined as polyps that show a prominent serrated architectural growth pattern, a villiform (or filiform) growth pattern, and hypereosinophilic epithelium with phenotypic evidence of dysplasia, either low or high grade (see Tables 22.4 and 22.5 ). Although many lesions, in particular SSA/Ps (with or without dysplasia), have been referred to as “serrated adenomas,” TSAs are characterized by epithelium with confluent pink eosinophilic cytoplasm and a papillary, or villiform, growth pattern ( Fig. 22.23 ). Because of confusion in the literature with regard to inclusion of different types of lesions within the TSA category, the clinical, epidemiologic, and molecular features of these uncommon lesions are poorly understood. Nevertheless, TSAs are uncommon, representing less than 1% to 2% of all colonic polyps in most studies. In one Japanese study, TSAs accounted for 1.8% of more than 10,000 colonic polyps. They are more common in females and occur much more commonly in the left colon, particularly the sigmoid and rectum. Right-sided lesions often represent SSA/P with exuberant and villiform serrated dysplasia. The mean age at diagnosis is 60 to 65 years, which is typically older than for patients with hyperplastic polyps or SSA/Ps.

A, The traditional serrated adenoma differs from a sessile serrated polyp by the presence of hypereosinophilic cytoplasm and confluent nuclear stratification. B, High-power view shows nuclear stratification. C, Filiform serrated adenomas show finger-like villous projections and dilated lymphatics in the lamina propria. These lesions are typically found in the rectum.

Endoscopically, TSAs manifest variable growth morphology and are more often pedunculated rather than sessile. In one study, 63% of TSAs were pedunculated, 29% were sessile, and 8% were described as flat or carpet-like. However, in SPS, TSAs are more common and may be seen in combination with hyperplastic polyps, SSA/Ps, SSA/Ps with dysplasia, and conventional adenomas.

The pathogenesis of TSAs is uncertain. Some studies suggest that they develop from precursor hyperplastic polyps (particularly the goblet cell type) or from SSA/Ps. However, other studies have failed to demonstrate this association and suggest that these lesions may develop via an alternate pathway. Methylation levels are intermediate between those of hyperplastic polyps and those of SSA/Ps, being present in approximately 65% of TSAs in one study, compared with 18% in conventional adenomas. In contrast, BRAF mutations are present in one third of cases. KRAS mutations have been identified in as much as 27% of TSAs, APC mutations in 5% to 28%, MGMT loss in 17% to 26%, and TP53 mutations in 10% to 20%. In one study, the prevalence of LOH of TP53, APC , 3p, and CDKN2A was low, similar to that for conventional adenomas. In a more recent study of 112 TSAs from Korea, one third of cases showed a hyperplastic or SSA/P precursor lesion, particularly in TSAs from the right colon. In that study, the presence of KRAS mutations and MGMT methylation was associated with an aggressive phenotype (high-grade dysplasia or carcinoma). Therefore, there is heterogeneity in the molecular pathogenesis of serrated adenomas.

Pathologic Features.

TSAs are characterized by prominent crypt serration with confluent epithelial dysplasia. These polyps are normally pedunculated and may occasionally be rather large and filiform. Filiform serrated adenomas show elongated, finger-like villous projections and often show inflammation, ulceration, and dilated lymphatics within the lamina propria (see Fig. 22.23 ). In one study, filiform serrated adenomas exhibited exclusive involvement of the rectum. The epithelium of serrated adenomas shows “dysplastic” characteristics, marked by nuclear atypia and nuclear stratification at all levels of the polyp, including the surface epithelium. Micropapillation of the surface epithelium and the presence of cells with eosinophilic cytoplasm are characteristic features.

The cells show a low N : C ratio with vesicular nuclei, an open chromatin pattern, and occasional inconspicuous nucleoli. These polyps are normally mucin depleted. The number of goblet cells may be variable. Dystrophic goblet cells are uncommon. On Ki67 or PCNA staining, the basal crypt compartment is often positive, but the luminal compartment is typically negative. Some TSAs, particularly those that are filiform, show abnormal horizontal crypt budding on the villi, which may be indicative of an abnormality in anchoring of the crypts to the basement membrane. This is referred to as “ectopic crypt formation.” Ectopic crypt formation is characteristic of TSA but is not specific, because it can also be seen in villous adenomas and in SSA/Ps with dysplasia.

On occasion, TSAs, particularly those that are large, show areas of conventional, adenoma-like, dysplastic foci within the lesion. It is likely that these lesions represent TSAs that have progressed, both morphologically and molecularly, along the APC/chromosomal instability carcinogenetic pathway, representing molecular heterogeneity within the polyp. TSAs may also show high-grade dysplasia (serrated or intestinal type, or both), characterized by increased cytologic atypia, marked stratification of nuclei, and a back-to-back epithelial growth pattern. Intraluminal “gland-in-gland” formation and cribriforming are common features of high-grade dysplasia.

Natural History and Treatment.

The natural history of TSA and the risk of progression to malignancy are poorly understood. Progression to high-grade dysplasia was reported in 37% of TSAs in one study. In that study, 11% contained intramucosal adenocarcinoma. In another study of filiform TSAs by Yantiss and colleagues, 22% showed high-grade dysplasia and another 6% showed invasive adenocarcinoma. In that study, 33% of polyps showed an adjacent region of hyperplastic or SSA/P morphology, indicating those areas as possible precursor lesions. It is possible that, as in other serrated polyps, loss of DNA repair mechanisms may drive rapid neoplastic progression. Some studies have suggested that the rate of malignant transformation is similar to that of conventional adenomas. However, it is likely that the risk of progression is related to the size and location of the lesion. Large TSAs in the proximal colon may progress at a more rapid rate than those in the left colon. The treatment of TSAs is similar to that of conventional adenomas. Complete endoscopic removal is warranted, and future surveillance intervals similar to those for patients with conventional adenomas (i.e., every 3 to 5 years) are recommended.

Conventional Adenoma with Serrated Architecture

Rarely, conventional tubular, tubulovillous, or villous adenomas show areas of architectural serration, but these polyps should be distinguished from TSAs by the cytologic features of the nuclei ( Fig. 22.24 ). Conventional adenomas with architectural serration show elongated, cigar-shaped nuclei with clumped chromatin, nuclear stratification, and mucin depletion, resembling conventional adenomas that develop through the APC/chromosomal instability pathway. It is unclear why some conventional adenomas acquire a serrated growth pattern, but this may be related to the acquisition of KRAS mutations, which has been proposed to add a “serrated” molecular signature to traditional adenomas and provides evidence of a “fusion” molecular pathway of carcinogenesis in some cases. In fact, molecular alterations characteristic of both the serrated pathway and the chromosomal instability carcinogenetic pathway coexist in a minority of advanced colorectal polyps that display morphologic features of both lesions. In one study, these lesions accounted for only 2% of colorectal polyps but were postulated to have a higher degree of biologic potential.

Conventional adenoma with a serrated architecture. In contrast to traditional serrated adenomas, conventional adenomas with architectural serration harbor cigar-shaped elongated nuclei with clumped chromatin, nuclear stratification, and mucin depletion, similar to traditional tubular or tubulovillous adenomas.

Unclassifiable Serrated Polyps

Rarely, because of overlapping features or processing artefact, hyperplastic and other serrated polyps may not be easily distinguished from each other. For lesions such as these, it is important that the pathologist make a great effort to determine whether the lesion has cytologic evidence of dysplasia, in order to place it into either the “nondysplastic” or “dysplastic” serrated polyp category for the purpose of adequate treatment. Use of the terms unclassifiable serrated polyp without dysplasia or unclassifiable serrated polyp with dysplasia may be a reasonable alternative in these cases. One interobserver study suggests a high degree of reliability in distinguishing dysplastic from nondysplastic features in hyperplastic or serrated polyps. However, for the purpose of clinical management, most, if not all, serrated polyps in the right colon, particularly if they are larger than 1 cm in size and flat (sessile) in appearance, should be classified as SSA/Ps even if they do not contain classic histologic features. Several studies have shown that the molecular characteristics of right-sided serrated lesions are similar regardless of the histologic features of the lesion or the presence of classic SSA/P morphology.

Serrated Polyposis Syndrome

SPS is a recently described, poorly defined entity with rapidly evolving diagnostic and molecular criteria and an increased risk for colorectal carcinoma that develops through the serrated pathway. SPS is considered a rare disorder, but evidence suggests that it is probably grossly underdiagnosed. Furthermore, it is now believed that SPS probably represents several overlapping disorders with considerable clinical, pathologic, and molecular heterogeneity. Relatives of patients with SPS are at significantly increased risk of colon and pancreas cancer. Although familial clustering has been documented, no consistent pattern of inheritance has been reported. Nevertheless, there are many factors, such as familial clustering, multiplicity of polyps, early age at onset, polyps and cancer in first-degree relatives of subjects with SPS, differences in incidence in different populations, and the finding of DNA methylation in normal colorectal mucosa, that raise the strong possibility of a genetic predisposition to SPS. It occurs almost equally among males and females and has been reported in children and young adults as well.

There seem to be at least two different types of SPS in terms of its phenotypic expression. One type comprises small, predominantly microvesicular or goblet cell hyperplastic polyps, often numerous in number but with little associated risk of malignancy. Many of the early descriptions of this disorder belong in this category and were referred to as “hyperplastic polyposis.” The other, more recently described form of SPS is composed of multiple SSA/Ps, often in combination with hyperplastic polyps, conventional adenomas, and TSAs. This type also shows a strong association with dysplastic SSA/P. In one recent study, 39% of patients with this type of SPS had colorectal cancer, and these patients were more likely to have conventional adenomas along with serrated polyps.

The following diagnostic criteria have been proposed for SPS : (1) at least five histologically confirmed serrated polyps located proximal to the sigmoid colon, of which at least two are larger than 10 mm in diameter; (2) any number of serrated polyps located proximal to the sigmoid colon in a subject with a first-degree relative with SPS; or (3) more than 30 serrated polyps of any size distributed evenly throughout the colon. The molecular characteristics of SPS are essentially unknown, but recent studies suggest that BRAF mutations and loss of MLH1/PMS2 expression play a major role in cancer development, as suspected.

Clinical Features

Adenomas are common lesions that are almost always asymptomatic. Some patients are seen with overt or occult rectal bleeding. Large polyps may lead to iron deficiency anemia. The clinical importance of adenomas is almost entirely related to their well-established premalignant nature. In general, the prevalence of adenomas increases dramatically with age. By the fifth decade of life, approximately 12% of individuals have adenomas, of which approxi­mately 25% are considered high-risk lesions. After age 50 years, the prevalence of adenomas continues to rise, to approximately 50% of the population in high-risk Western countries such as the United States. The likelihood that adenomas will develop in an individual is strongly influenced by family history and by a variety of nutritional factors.

Adenomatous polyps of the colon may be broadly classified as conventional (tubular, tubulovillous, or villous), serrated, or flat (discussed in the next section). Some studies suggest that flat adenomas have a higher incidence of high-grade dysplasia and carcinoma and therefore may pose a higher risk of malignant degeneration, but this is controversial. Serrated adenomas were discussed earlier in this chapter.

Pathologic Features

Adenomas are morphologically defined as dysplastic clonal proliferations of colonic epithelium ( Fig. 22.25 ). Grossly, adenomas are broadly classified as sessile (without a stalk) or pedunculated (with a stalk). However, intermediate forms between sessile and pedunculated may develop as well. Rarely, adenomas may be multilobulated or filiform or both. However, these features are unusual. Occasionally, surface ulceration may be observed.

Tubular adenoma featuring low-grade dysplasia and predominantly tubular architecture. The adenomatous epithelial cells show mild hyperchromasia with enlarged, oval nuclei that occupy the basal half of the cell cytoplasm. Basal polarization of the nuclei is retained, and no significant pleomorphism occurs. Occasional nondysplastic crypts are evident in the deep portion for comparison.

Microscopically, adenomas are categorized architecturally as tubular, tubulovillous, or villous ( Fig. 22.26 ). Precise histologic criteria for these separate categories vary widely, but a reasonable rule is that villous lesions should contain at least 75% villi, whereas tubular lesions should contain less than 25% villi. Tubulovillous lesions are those that contain between 25% and 75% villous epithelium. The degree of villous differentiation has been shown to increase with increasing size of the adenoma. Although several previous studies reported a correlation between tubulovillous or villous histology and subsequent advanced adenoma or carcinoma, the histologic type of adenoma has not been found to predict subsequent advanced neoplasia in randomized trials thus far. Nevertheless, villous lesions are considered “advanced” for the purpose of clinical management (see later discussion).

Each of these three adenomas displays low-grade dysplasia, but the architecture varies from tubular ( A ) to tubulovillous ( B ) to villous ( C ). The loss of architectural rigidity is a common finding in adenomas.

Microscopically, by definition, all adenomas contain at least low-grade (mild) dysplasia. Dysplasia in adenomas is generally classified as low grade (mild to moderate) or high grade (severe, including carcinoma in situ) based on a combination of cytologic and architectural features. Adenomas that show invasion of the mucosa or the muscularis mucosae (but not beyond) are considered intramucosal adenocarcinomas. Invasive carcinoma occurs when submucosal stalk invasion is present beyond the muscularis mucosae. This classification system is favored because (1) a decrease in the degree of interobserver variability has been documented regarding interpretation of dysplasia by pathologists; (2) improved clinical pathologic relevance is seen with regard to surveillance and treatment options; and (3) the term carcinoma in situ is often misinterpreted by clinicians as indicative of malignant behavior, which may lead to an unnecessary colonic resection.

Low-grade dysplasia is defined by the presence of architecturally noncomplex crypts containing nuclei that are pseudostratified, or partially stratified, such that the cell nuclei reach only the lower half of the cell cytoplasm. Mitotic activity may be brisk, but atypical mitoses, significant loss of polarity, and pleomorphism are minimal, if present at all. The crypts are arranged in a parallel configuration without significant back-to-back configuration, cribriforming, or complex budding.

High-grade dysplasia is defined by marked pseudostratification, or stratification, of neoplastic nuclei that extend toward the luminal half of the cells and usually contain significant pleomorphism, increased mitotic activity, atypical mitoses, and marked loss of polarity. Architectural changes such as back-to-back gland configuration and cribriforming may also be observed. With progression of neoplasia, glands lose their orderly configuration and become more irregular and complex. In addition, neoplastic nuclei become more “open” in appearance and may contain prominent nucleoli ( Fig. 22.27 ). The N : C ratio of the cells increases, and loss of polarity becomes marked. In this classification, the designation “carcinoma in situ” is encompassed and is replaced by the term high-grade dysplasia .

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Systemic Illnesses Involving the Gastrointestinal Tract Polyps of the Esophagus Epithelial Neoplasms of the Stomach Epithelial Neoplasms of the Large Intestine Molecular Pathogenesis and Diagnostics of Hepatocellular Tumors Toxin- and Drug-Induced Disorders of the Liver

Stay updated, free articles. Join our Telegram channel

Join