OUP user menu

Serrated Lesions of the Appendix
A Morphologic and Immunohistochemical Appraisal

Andrew M. Bellizzi MD, Jonathan Rock MD, William L. Marsh MD, Wendy L. Frankel MD
DOI: http://dx.doi.org/10.1309/AJCP1UJPX6UURLCH 623-632 First published online: 1 April 2010


We performed a histologic and immunohistochemical assessment of 53 noninvasive appendiceal epithelial proliferations, appropriating terminology and using markers shown useful in differentiating serrated colorectal polyps. These were classified as hyperplastic polyp (HP), sessile serrated adenoma (SSA), mixed serrated and adenomatous lesion (MSAL), mucinous cystadenoma (MCA), or conventional adenoma (CAD). Immunohistochemical analysis for cytokeratin (CK) 20, Ki-67, MUC6, and β-catenin was performed. Diagnoses were as follows: HP, 6; SSA, 12; HP vs SSA, 3; MSAL, 16; MCA, 14; and CAD, 2. All HPs showed expanded (beyond surface) CK20 and expanded or normal (base) Ki-67; 1 was MUC6+. Most SSAs and MSALs were CK20-expanded or expanded with random expression in deep crypts (Ex/I) and Ki-67-expanded, Ex/I (expanded with asymmetry), or normal. All SSAs and 8 of 16 MSALs were MUC6+. CADs were CK20-Ex/I, Ki-67-Ex, and MUC6–; 1 showed nuclear β-catenin expression. Serrated appendiceal lesions can be categorized using colorectal terminology. MUC6 is associated with SSA morphologic features. Similar immunohistochemical patterns in SSA and MSAL suggest a link between these lesions.

Key Words:
  • Appendix
  • Sessile serrated adenoma
  • Hyperplastic polyp
  • Mucosal hyperplasia
  • Mucinous cystadenoma
  • Immunohistochemistry
  • Cytokeratin 20
  • Ki-67
  • MUC6
  • β-Catenin

Two decades of morphologic and molecular research has culminated in the delineation of a serrated pathway of colorectal neoplasia, distinct from the canonical adenoma-carcinoma sequence.112 A number of serrated polyps are now recognized, including hyperplastic polyp (with micro-vesicular and goblet cell variants), sessile serrated adenoma (also known as sessile serrated polyp and serrated polyp with abnormal proliferation), and traditional serrated adenoma (TSA). In the best-characterized serrated sequence, a sessile serrated adenoma (with a BRAF mutation and extensive promoter methylation) develops overt cytologic dysplasia (ie, becomes a “mixed polyp”) and ultimately invades as a micro-satellite unstable (MSI-H) adenocarcinoma.

Much less is known about morphologically similar serrated lesions of the appendix, described under various overlapping terms, including mucosal metaplasia, mucosal hyperplasia (focal or diffuse), hyperplastic (metaplastic) polyp, mixed adenomatous/hyperplastic lesion, and, more recently, serrated adenoma and sessile serrated adenoma.1318 Given the pace of research on serrated polyps in the colorectum and the connotation of the diagnosis “sessile serrated adenoma,” when confronted with a serrated appendiceal lesion at sign-out, we wrestle with diagnostic terminology.

The aim of this study was to characterize a series of serrated appendiceal lesions. For serrated lesions without cytologic dysplasia, we appropriated diagnostic terminology from the colorectum. Several groups have recently reported the usefulness of immunohistochemical analysis (cytokeratin [CK] 20, Ki-67, MUC6, and β-catenin) in distinguishing classes of serrated polyps.1921 We examined the relationship between light microscopic diagnoses and the results of immunohistochemical analysis. We conclude that there is considerable morphologic and immunophenotypic overlap between sessile serrated adenoma in the colorectum and appendix.

Materials and Methods

The surgical pathology files of the Ohio State University Medical Center, Columbus (1989 to 2008), were searched for appendiceal lesions originally diagnosed as any of the following: mucosal hyperplasia, hyperplastic polyp, sessile serrated adenoma, serrated adenoma, villous adenoma, tubular adenoma, tubulovillous adenoma, cystadenoma, or mucocele. Cases with an associated mural invasive appendiceal adenocarcinoma were excluded from further analysis.

Corresponding glass slides were retrieved. Cases were reviewed at a multiheaded light microscope by 4 pathologists with an interest in gastrointestinal pathology and reassigned by consensus to the following diagnostic categories: hyperplastic polyp (HP), sessile serrated adenoma (SSA), mixed serrated and adenomatous lesion (MSAL), mucinous cystadenoma (MCA), or conventional adenoma (CAD). HPs and SSAs were classified as such based on the presence of epithelial serration in the absence of cytologic dysplasia. The distinction of SSA from HP was based on a constellation of cytoarchitectural features, as described previously.5,6,8,19 MSALs combined epithelial serration with overt cytologic dysplasia. MCAs were composed of dysplastic mucinous epithelium without obvious serration; CADs resembled their colorectal counterparts.

Immunohistochemical analysis for CK20, Ki-67, MUC6, and β-catenin was performed on 4-μm sections derived from a representative tissue block from each of the study cases using a DakoCytomation Autostainer (DakoCytomation, Carpinteria, CA) after heat-induced epitope retrieval in Target Retrieval Solution (DAKO, Carpinteria, CA) at pH 6.1. Antibody characteristics including clone, vendor, and dilution are given in Table 1. Reactions were developed using the Labeled Streptavidin Biotin Plus Detection Kit (DAKO), after blockade of endogenous biotin.

View this table:
Table 1

CK20 was graded as normal if staining was confined to the surface epithelium; expanded if staining extended beyond the upper quarter of the crypt height; expanded/irregular if, in addition to expanded staining, random strong expression was noted in the deep crypts; diffuse if all lesional cells stained; and no pattern (NP) if there was haphazard staining. Ki-67 was graded in analogous manner, with normal staining confined to the crypt bases, expanded staining extending beyond the lower quarter of the crypt height, expanded/irregular staining combining expanded staining with asymmetrical staining of crypts and the staining of patches of surface epithelium, and NP characterized by haphazard staining; diffuse Ki-67 expression was not encountered. This CK20/Ki-67 grading scheme was modified from Torlakovic et al.19 Any definitive cytoplasmic reactivity for MUC6 was considered positive. β-Catenin was scored as normal given a membranous staining pattern or abnormal in the presence of significant nuclear accumulation.

Immunohistochemical stains were assessed without reference to the light microscopic diagnosis. When present, morphologically normal mucosa was also evaluated.

The mean age, sex distribution, frequency of synchronous or metachronous serrated colorectal polyp (ie, HP or SSA), and frequency of synchronous or metachronous colorectal adenocarcinoma for the MCA category vs the combined serrated category (HP, SSA, HP vs SSA, and MSAL) were compared using a 2-tailed Mann-Whitney U test or Fisher exact test, as appropriate. A P value of less than .05 was considered statistically significant.


In the cases studied, 6 HPs occurred in 1 woman and 5 men at a mean age of 75 years (range, 58–90 years); 12 SSAs were noted in 6 women and 6 men at a mean age of 64 years (range, 29–83 years); 16 MSALs were diagnosed in 7 women and 9 men at a mean age of 63 years (range, 25–97 years); 14 MCAs occurred in 11 women and 3 men at a mean age of 48 years (range, 24–68 years); and 2 CADs were noted in 1 woman and 1 man at a mean age of 62 years (range, 60–64 years). In 3 serrated polyps without cytologic dysplasia, a consensus diagnosis (HP vs SSA) could not be reached (3 women; mean age, 53 years; range, 31–81 years). Patients with a diagnosis of MCA (mean age, 48 years) were significantly younger than were patients in the combined serrated polyp group (mean age, 64 years; P = .004); the female/male ratio for these 2 groups (MCA, 3.7:1; serrated polyp, 0.85:1) approached statistical significance (P = .06). Of 14 MCA cases, 2 (14%) had a synchronous/metachronous serrated colorectal polyp compared with 10 (27%) of 37 in the combined group (P = .47), and 2 (14%) of 14 MCA cases had a synchronous/metachronous colorectal adenocarcinoma, compared with 15 (41%) of 37 in the combined group (P = .10). These data (and the immunohistochemical results) are summarized in Table 2 and Table 3.

HPs tended to involve a portion of the appendiceal circumference (on cross-section). They were characterized by superficial epithelial serration with the lesion's configuration tapering down to a preserved proliferative zone. By definition, they lacked cytologic dysplasia. SSAs tended to involve the entire appendiceal circumference. They exhibited an array of characteristic architectural abnormalities, including crypt dilatation and branching, transverse-lying crypts, extension of serration toward the crypt base, and the presence of differentiated mucous cells in the crypt base. In general, these features involved at least 20% of crypts. MSALs ranged from those largely resembling SSA involved by small foci of cytologic dysplasia (morphologically analogous to SSA with superimposed cytologic dysplasia or mixed hyperplastic/adenomatous polyp) to those entirely composed of cytologically dysplastic epithelium (morphologically analogous to serrated adenoma). MCAs were composed of dysplastic mucinous epithelium without obvious serration; the lining epithelium was generally gently undulating or flat. CADs resembled their colorectal counterparts, composed of elongated, stratified, hyperchromatic nuclei without a prominent mucinous component. The histologic features of these lesions are summarized in Table 4, and representative photomicrographs from each diagnostic category are given in Image 1.

View this table:
Table 2
View this table:
Table 3
View this table:
Table 4
Image 1

Appendiceal epithelial proliferations. A and B, Hyperplastic polyp. A, Hyperplastic polyps generally involve a portion of the appendiceal circumference and are characterized by superficial serration (arrow) (H&E, ×40). B, On higher power, a mixture of goblet and microvesicular mucous cells are apparent (H&E, ×200). C and D, Sessile serrated adenoma. Sessile serrated adenomas tend to involve the entire appendiceal circumference. C, Crypt dilation, including transverse-lying crypts, are apparent at low power (H&E, ×40). D, On higher power, near-full-thickness serration and mucous cells in the deep crypts are noted (H&E, ×200). E, F, and G, Mixed serrated and adenomatous lesion. Mixed serrated and adenomatous lesions combine serrated architecture with overt cytologic dysplasia. E, The diffuse nature of this lesion and crypt dilation are noted at low power (H&E, ×20). On higher power, serrated foci indistinguishable from hyperplastic polyp (F) alternate with foci composed of villi lined by clear-cut, dysplastic, mucinous epithelium (G) (H&E, F, ×200; G, ×200). H and I, Mucinous cystadenoma. Mucinous cystadenomas tend to involve the appendix circumferentially. H, At low power, loss of the normal lamina propria complement and fibrosis of the appendiceal wall are noted (H&E, ×20). I, Higher-power examination reveals the undulating, dysplastic, mucinous epithelium to advantage (H&E, ×200). J and K, Conventional adenoma. Conventional adenomas resemble their colorectal counterparts. J, Low-power examination reveals a polypoid tubulovillous proliferation (H&E, ×20). K, On higher power, typical intestinal-type dysplasia with scattered goblet cells is noted (H&E, ×400).

Image 2

Immunohistochemical analysis using cytokeratin (CK) 20, Ki-67, MUC6, and β-catenin. A and B, CK20. A, Expanded CK20 expression is noted in a hyperplastic polyp (arrow); the surrounding mucosa demonstrates normal, surface-confined expression. B, The expanded/irregular pattern in this sessile serrated adenoma combines expanded expression with strong staining in deep crypts (asterisks) (immunoperoxidase [IP], A, ×40; B, ×100). C and D, Ki-67. C, Expanded Ki-67 expression in this hyperplastic polyp extends nearly half-way up lesional crypts. D, The expanded/irregular pattern combines expanded expression with asymmetrical crypt staining or stretches of surface staining (arrow) (IP, C, ×100; D, ×100). E, MUC6 expression is noted in the deep crypts of all appendiceal sessile serrated adenomas, many mixed serrated and adenomatous lesions, and a rare hyperplastic polyp (IP, ×200). F, With β-catenin, abnormal nuclear localization is noted in a single conventional adenoma (IP, ×400). Inset, This serrated polyp exhibits normal membranous expression (IP, ×400).

Normal mucosa largely demonstrated normal CK20 and Ki-67 expression; occasional cases displayed expanded expression. MUC6 was negative. β-Catenin was normally expressed in normal mucosa and in all lesions, except for 1 CAD. HPs were uniformly characterized by expanded CK20 expression, and 3 (50%) of 6 also demonstrated expanded Ki-67 staining; MUC6 staining was noted in 1 (17%) of 6. SSAs were generally characterized by expanded/irregular or expanded expression of CK20 and Ki-67. MUC6 immunoreactivity was noted in all 12 (100%). The majority of MSALs demonstrated an expanded or expanded/irregular CK20 pattern, while Ki-67 staining was frequently haphazard; MUC6 immunoreactivity was noted in 8 (50%) of 16. MCAs frequently demonstrated diffuse CK20 and haphazard Ki-67 labeling; they were negative for MUC6. The 2 CADs showed expanded/irregular CK20 and expanded Ki-67 staining; MUC6 was negative. As noted, 1 CAD displayed abnormal β-catenin localization. Overall, MUC6 staining in the serrated appendiceal lesions involved 3 to 115 crypts (median, 19 crypts). The staining intensity ranged from modest to strong and was restricted to crypt bases, with most intense staining in the basalmost 10 to 20 cells. In cases with multiple crypts staining, staining tended to involve groups of contiguous crypts. Regarding the Ki-67 expanded/irregular pattern, groups of surface cells (always at least 10 cells and in most cases >25 cells) were never the sole manifestation of this pattern. These results are summarized in Table 2. Individual results for the 3 serrated lesions for which a light microscopic consensus (HP vs SSA) could not be reached are given in Table 3. Photomicrographs depicting representative immunohistochemical patterns are given in Image 2.


By appropriating diagnostic criteria from the colorectum, we were able to classify the vast majority (18/21 [86%]) of serrated appendiceal lesions lacking cytologic dysplasia. As such, we consider the term hyperplastic polyp to be roughly equivalent to focal mucosal hyperplasia (metaplasia) and sessile serrated adenoma to be roughly equivalent to diffuse mucosal hyperplasia (metaplasia). The inability to achieve a consensus diagnosis for 3 lesions underscores the difficulty in distinguishing SSA from HP anywhere in the gastrointestinal tract.22,23 In these 3 cases, immunohistochemical analysis proved useful in adjudicating the differential diagnosis. We encountered lesions combining architectural serration with varying amounts of cytologic dysplasia nearly as frequently as the “nondysplastic” serrated lesions (n = 16 vs n = 18). For the purposes of this study we classified these cases as MSALs. This category is probably heterogeneous, including lesions morphologically analogous to “SSA with superimposed cytologic dysplasia” and “serrated adenoma.” We also noted several cases of MCA. This seems to be a uniquely appendiceal lesion, and given a distinct mean age and a sex distribution approaching statistical significance, it may not be related to the serrated appendiceal lesions. Finally, as has been noted by others, rare adenomatous polyps exhibit morphologic features indistinguishable from conventional colorectal adenomas.24,25

The distinction of SSA from HP largely rests on the identification of architectural features collectively referred to in various investigations as “abnormal proliferation,” “dysmaturation,” or “crypt compartmentalization aberration.”5,6,19 Torlakovic and colleagues19 recently explored the ability of CK20 and Ki-67 immunohistochemical analysis to delineate the compartmentalization of maturation and proliferation. SSAs exhibited a characteristic CK20 and Ki-67 pattern, reflecting the presence of crypt compartmentalization aberration, which the authors termed the “SSA pattern.”19 The SSA pattern distinguished SSA from HP, TSA, and CAD. Many of the appendiceal SSAs in our study exhibited an analogous pattern, which we referred to as expanded/irregular. The presence of an expanded/irregular pattern for CK20 and/or Ki-67 was 67% sensitive and 100% specific for the diagnosis of SSA vs HP. The presence of an expanded/irregular pattern in a significant minority of MSALs suggests a link between SSA and at least a subset of MSAL. The presence of expanded/irregular Ki-67 and/or CK20 expression (along with MUC6 reactivity) in 1 of 3 “no consensus” polyps supports the diagnosis of SSA; expanded Ki-67 and/or CK20 in the other 2 polyps (along with a lack of MUC6 reactivity), supports the diagnosis of HP.

Serrated polyps have long been known to exhibit a mixed intestinal and gastric phenotype; this finding has been highlighted with apomucin immunohistochemical analysis.2628 More recent attention has focused on the ability of apomucin immunohistochemical analysis to discriminate the various classes of serrated polyps. Owens et al20 found immunoreactivity for the gastric pyloric mucin MUC6 in 33 of 33 SSAs (including 7 cases with superimposed dysplasia), 0 of 48 HPs, and 0 of 13 TSAs. Other studies have been less dramatic. For example, Mochizuka et al28 found MUC6 expression in 76% of 29 SSAs, 27% of 15 HPs, and 6.3% of 16 CADs. We found MUC6 expression in all 12 appendiceal SSAs (100%), highlighting the phenotypic similarity between appendiceal and colorectal lesions. MUC6 was also noted in 1 morphologically unequivocal HP. The presence of MUC6 immunoreactivity in 8 of 16 MSALs again suggests a link between SSA and a portion of MSALs. MUC6 was absent from all 14 typical MCAs, further supporting their distinction from the serrated appendiceal lesions; 2 CADs were also MUC6–.

β-Catenin functions in the cell membrane (as a component of adherens junctions) and in the nucleus (where it mediates the transcription of target genes).29 β-Catenin compartmentalization can be visualized with immunohistochemical analysis. In the colorectum, nuclear immunoreactivity is generally restricted to the crypt bases, reflecting physiologic Wnt signaling. Widespread nuclear accumulation generally reflects loss of APC (adenomatous polyposis coli) function or “activating” β-catenin mutation, features of the canonical adenoma-carcinoma sequence. Wu and colleagues21 were somewhat surprised, then, to find aberrant nuclear labeling in 9 (41%) of 22 SSAs vs 0 (0%) of 19 HPs. They argued that this may reflect Wnt signaling rather than APC/β-catenin mutation.21 In our series of appendiceal lesions, we failed to demonstrate abnormal nuclear accumulation of β-catenin in 37 serrated lesions and 14 MCAs. Abnormal nuclear accumulation in 1 CAD is not surprising, likely reflecting loss of APC function. Similarly, Yantiss et al30 failed to detect significant nuclear localization in a series of 56 serrated appendiceal lesions and 17 cystadenomas.

Yantiss and colleagues30 recently reported a largely molecular-based survey of serrated appendiceal lesions. “Cystadenomas” served as controls. They detected reduced MLH-1 and MGMT immunoreactivity across the spectrum of serrated lesions (while cystadenomas failed to reveal loss of staining). Loss of MLH-1 immunoreactivity was not accompanied by MSI-H. The authors also detected BRAF and KRAS mutations, as seen in serrated polyps of the colorectum, albeit at lower overall rates and, in particular, with less frequent BRAF mutations.7,9,10 They also failed to demonstrate MSI-H in 4 appendiceal adenocarcinomas arising adjacent to serrated polyps; BRAF mutation was detected in only 1. These results led the authors to question whether serrated lesions in the appendix and colorectum are directly analogous. Instead of embracing identical diagnostic terminology, they preferred more generic language (ie, “nondysplastic or dysplastic serrated polyps” of the appendix).30

The lack of demonstration of MSI-H or loss of mismatch repair (MMR) function in appendiceal adenocarcinomas has previously led others to question aspects of the serrated pathway of neoplasia vis-à-vis the appendix. Kabbani et al31 failed to detect MSI in 30 appendiceal adenocarcinomas (including 22 low-grade mucinous and 8 high-grade tumors). Misdraji et al32 detected loss of MSH-2 and MSH-6 immunoreactivity in a single mural-invasive tumor (likely representing a germline mutation); 8 low-grade appendiceal mucinous neoplasms with extra-appendiceal spread (essentially equivalent to low-grade mucinous tumors) and 7 other mural-invasive tumors demonstrated intact expression of MLH-1, MSH-2, and MSH-6; similarly, Pham et al33 previously reported intact MMR protein expression in 15 low-grade and 14 high-grade appendiceal tumors.

Before discarding the concept of an “appendiceal SSA equivalent,” it should be noted that the majority of tumors studied for MSI/MMR status to date represent low-grade mucinous tumors; it is generally accepted that the MCA represents the precursor lesion in many of these cases.25,3437 And as our demographic and immunophenotypic data and the molecular data of Yantiss and colleagues30 suggest, most cases of MCA may occur independent of serrated appendiceal lesions. The precursor lesion of high-grade/mural-invasive appendiceal adenocarcinomas remains ill-characterized. The resulting tumors exhibit frequent loss of heterozygosity, suggesting chromosomal instability; as such, lack of MSI-H is not surprising.38,39 Higuchi and Jass40 estimated the risk of MSI-H cancer in patients with right-sided serrated polyps at 1 in 25 (4%). Given this and the likelihood of a large polyp coming to clinical attention as acute appendicitis, the difficulty in identifying MSI-H appendiceal adenocarcinomas is not entirely unexpected.18


Serrated lesions morphologically analogous to those seen in the colorectum are found in the appendix. Our data indicate immunophenotypic overlap as well. Appendiceal HPs and SSAs demonstrate CK20, Ki-67, and MUC6 staining patterns similar to their colorectal counterparts. The results of coordinate CK20/Ki-67/MUC6 immunohistochemical analysis may allow for the adjudication of serrated polyps indeterminate for a diagnosis of HP vs SSA. A proportion of MSALs demonstrated a CK20/Ki-67/MUC6 pattern similar to SSA, suggesting a link between these lesions. Given its demographic and immunophenotypic features, MCA seems to represent a distinct lesion. Appendiceal CADs are rare; β-catenin nuclear localization most likely reflects loss of APC function.

Despite significant morphologic and immunophenotypic overlap between serrated appendiceal and colorectal lesions, the presence of differential rates of BRAF mutation and the lack of demonstrable resultant MSI-H appendiceal adenocarcinomas calls into some question the presence of a directly analogous serrated pathway of appendiceal neoplasia. Whatever diagnostic terminology one uses, we recommend a note addressing the unresolved nature of these lesions. Further study (eg, methylation patterns of serrated appendiceal lesions and molecular features of colorectal cancers occurring in patients) is warranted.


Upon completion of this activity you will be able to:

  • describe the features of the serrated pathway of colorectal neoplasia, compared to the canonical adenoma-carcinoma sequence.

  • list the various noninvasive epithelial proliferations of the appendix and describe their histologic features.

  • apply a panel of immunohistochemical stains to assist in the differential diagnosis of a serrated appendiceal lesion.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit™ per article. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 670. Exam is located at www.ascp.org/ajcpcme.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
View Abstract