ANATOMIC PATHOLOGY Original Article

ANATOMIC PATHOLOGY Original Article Histology of the Mucosa in Sigmoid Colon Specimens With Diverticular Disease Observations for the Interpretation of Sigmoid Colonoscopic Biopsy Specimens NEAL S. GOLDSTEIN, MD, AND EJAZ AHMAD, MD We examined retrospectively 100 sigmoid colon resection specimens removed for diverticulitis (DD [diverticular diseasel-diverticulitis), 53 adenocarcinoma specimens that also had diverticulosis (DD-adenocarcinoma), and 50 adenocarcinoma specimens that did not have DD (adenocarcinoma only) to study the mucosal changes that occur in DD. Documenting these histologic features could be helpful in deciphering changes seen in colonoscopic biopsy specimens from the sigmoid colon in older patients. Prominent mucosal folds were present in approximately 90% of all DD specimens. Increased mucosal lymphoplasmacytic inflammation at the bases of the prominent folds was present in 15% and 9% of DD-diverticulitis and DD-adenocarcinoma specimens, respectively. Eleven percent of the DD-diverticulitis and 4% of the DD-adenocarcinoma specimens had prolapselike mucosal abnormalities of the mucosa on the surface of the prominent mucosal folds. Mildly increased lymphoplasmacytic inflammation surrounded the diverticulosis ostia in approximately 25% of all DD specimens. All the diverticulitis ostia had neutrophilic and lymphoplasmacytic inflammation in the surrounding mucosa. No specimens had crypt distortion. Diverticular disease-related inflammation may be one cause of mild patchy inflammation that is occasionally observed in sigmoid colon biopsy specimens. Diverticular disease also should be considered as a cause of mucosal prolapse changes in sigmoid colon biopsy specimens. Other diseases should be considered when markedly increased mucosal inflammation, crypt distortion, or granulomas are present. Distinction between a DD-related incidental finding and a significant pathologic abnormality frequently can be made with the procurement of multiple biopsy specimens. (Key words: Colon; Diverticulosis; Diverticulitis; Prolapse; Inflammation) Am J Clin Pathol 1997,107:438-444. Diverticular disease (DD) is common in the sigmoid colon. In Western countries, it affects more than 30% of people older than 60 years. 1-3 Diverticulosis describes the condition of having diverticula, usually asymptomatically, while diverticulitis refers to inflammation of the outpouchings. Colonoscopy has provided physicians frequent contact with sigmoid DD. Usually the colonoscopy is performed for reasons other than DD. The frequent contact prompted us to document the histologic changes that can be observed in the mucosa between and around diverticular ostia in the sigmoid colon. Documenting these features could be helpful in assessing the changes seen in colonoscopic biopsy specimens from the sigmoid colon in older patients. Knowledge of the histologic features that could be attributed to DD and those that might better be ascribed to other causes would be beneficial in distinguishing innocuous histologic findings from significant abnormalities. We retrospectively examined sigmoid colon resection specimens removed for diverticulitis, adenocarcinoma that also had diverticulosis, and adenocarcinoma that did not have DD, to study the mucosal changes that occur in DD, emphasizing the histologic features that could be seen on mucosal biopsy specimens. From the Department of Anatomic Pathology, William Beaumont Hospital, Royal Oak, Michigan. Manuscript received June 20,1996; revision accepted September 12,1996. Address reprint requests to Dr Goldstein: Department of Anatomic Pathology, William Beaumont Hospital, 3601 West 13 Mile Rd, Royal Oak, MI 48324. MATERIALS AND METHODS The study consisted of three specimen groups: sigmoid colons with DD resected for diverticulitis (DDdiverticulitis), sigmoid colons with diverticulosis resected for adenocarcinoma (DD-adenocarcinoma), and sigmoid colons with adenocarcinoma and no DD 438

GOLDSTEIN AND AHMAD 439 Histology of the Mucosa in Sigmoid C Specimens With Diverticular Disease (adenocarcinoma only). The first specimen group (DD-diverticulitis) consisted of 100 consecutive diverticulitis sigmoid resection specimens from 1994 and 1995. All patients with diverticulitis underwent surgery for abdominal pain associated with diverticulitis-induced perforation or pericolonic abscess. No patients had obstruction or acute bleeding as the primary reason for surgical intervention. There were no cases with coexistent primary inflammatory bowel disease or DD-related active colitis. 4 ' 5 All the patients had diverticulosis and diverticulitis. The DD-diverticulitis specimens were evaluated in three different sites: grossly normal mucosa between diverticular ostia, mucosa around diverticulosis ostia, and mucosa around diverticulitis ostia. Diverticulosis ostia were defined as grossly normal ostia that were at least 2 cm from the acutely inflamed diverticulitis ostium. The second specimen group (DD-adenocarcinoma) served as a comparison group for diverticulosis ostia. It consisted of 53 consecutive sigmoid colon specimens resected during 1970 through 1995 for adenocarcinoma that had coexistent diverticulosis. Specimens from this time period with coexistent diverticulitis and adenocarcinoma also were sought, but none were found. The specimens were evaluated in two sites: grossly normal mucosa between the diverticular ostia and mucosa around the diverticulosis ostia. The third specimen group (adenocarcinoma only) functioned as a normal mucosa control group and consisted of 50 consecutive sigmoid colon specimens resected for adenocarcinoma from 1993. None of these cases had coexistent DD or inflammatory bowel disease. Examination was limited to the slides of routinely sectioned, grossly normal colonic bowel procured distant from the adenocarcinoma. A mean of 6 slides per DD-diverticulitis resection specimen was examined (range, 4-13). Slides consisted of cross sections of at least two diverticulosis ostia and two diverticulitis ostia. Two slides from each DD-adenocarcinoma specimen were examined; one slide was of diverticulosis ostia and the other was of the routinely sampled, grossly normal bowel wall distant from adenocarcinoma or DD ostia. One slide of grossly normal bowel from each adenocarcinomaonly specimen was examined. The histologic features in the grossly normal mucosa of the three specimen groups that were evaluated included prominent mucosal folds, lamina propria lymphoplasmacytic inflammation, neutrophils, length and number of foci of inflamed mucosa, intraepithelial lymphocytes, lamina propria basal plasmacytosis, granulomas, lamina propria fibrosis, ulceration, crypt branching, shortening, and dilatation. A prominent mucosal fold was arbitrarily defined as extending greater than 5 mm above the inner aspect of the muscularis propria. This distance was measured with a ruler placed on the stage using a x2 stage and xlo eyepiece objective. The diameter of the viewing field was 12.5 mm at this magnification. A mucosal fold was considered prominent if by visual estimation it extended greater than one third of the diameter across the field at this magnification. The number and length of the inflamed mucosa foci were recorded in the regions between DD ostia and in the normal mucosa control slides. The intensity and distance of mucosal abnormalities around the diverticular ostia were measured in the DD specimen groups. The normal constituent mucosal inflammatory cells was a sprinkling of lymphocytes and plasma cells between occasional lymphoid follicles, and no neutrophils. The white color of the areolar lamina propria connective tissue was the predominant color compared with the blue inflammatory cell nuclei (Fig 1). Inflammation was graded on a two-grade scale. Lymphocytes and plasma cells were mildly increased when the white spaces of the areolar lamina propria connective tissue was obscured by the inflammatory cells but there was no crypt architectural shortening or increased intercrypt distance. Markedly increased lymphoplasmacytic inflammation was present when there was crypt shortening or increased intercrypt distance due to the cells expanding the lamina propria. Mildly increased neutrophilic inflammation was considered FIG 1. Prominent mucosal fold formed by redundant submucosa. The mucosa on the surface is normal (see Fig 2). The mucosa of the right-fold base has a mildly increased number of lymphocytes and plasma cells (x22.5). Figure 3 is a higher magnification of the area of the right fold base. Vol.: No. 4

ANATOMIC PATHOLOGY 440 Original Article present if there were clusters of neutrophils in the lamina propria between crypts and no infiltration of the c r y p t or surface e p i t h e l i u m. N e u t r o p h i l s w e r e markedly increased when they formed crypt abscesses or extensively infiltrated the surface epithelium. RESULTS Histologic changes observed in the bowel between DD ostia are listed in Table 1. Prominent mucosal folds were present in 9 1 % of DD-diverticulitis and 87% of DD-adenocarcinoma specimens. The prominent mucosal folds were due to redundant mucosa and submucosa (Figs 1 and 2). Fifteen percent of DDdiverticulitis and 9% of the DD-adenocarcinoma specimens had mildly increased mucosal lymphoplasmacytic inflammation at the bases of the prominent folds (Fig 3). All of the inflamed mucosal segments along the fold bases were less than 5 mm long (mean, 3.8 mm). The mean number of inflamed mucosal foci was 3.5 (range, 1-6) per DD-diverticulitis and DD-adenocarcinoma specimen. There was no lamina propria basal plasmacytosis or intraepithelial lymphocytic infiltration within the inflamed regions. The inflamed mucosal segments did not have neutrophils, granulomas, ulcerations, crypt branching, shortening, dilatation, or lamina propria fibrosis. In contrast, 6% of the adenocarcinoma-only control group specimens had p r o m i n e n t mucosal folds, and n o n e of these h a d mucosal inflammation along the fold bases. Eleven percent of the DD-diverticulitis and 4% of the DD-adenocarcinoma specimens h a d mucosal a b n o r m a l i t i e s on t h e surface of t h e p r o m i n e n t mucosal folds. The mucosal changes in those regions included lamina propria fibrosis and slips of hyperplastic muscularis mucosae that extended up toward the surface between elongated crypts (Fig 4) (see Table 1). A few specimens also had mild lymphoplasmacytic inflammation, p a t c h y mild n e u t r o p h i l i c inflammation, and dilated granulation tissuelike capillaries. These mucosal changes were limited to small regions on the tips of the folds in most specimens. In four specimens (DD-diverticulitis, three; DD-adenocarcinoma, one), they produced a polypoid lesion. The histologic features of the mucosa surrounding DD ostia are listed in Table 2. There was increased lymphoplasmacytic inflammation around the diverticulosis ostia in 27% and 23% of DD-diverticulitis and DD-adenocarcinoma specimens, respectively. This inflammation was limited to within 0.2 cm of the edges of the diverticulosis ostia in 14 (52%) of the 27 DD-diverticulosis specimens and 10 (83%) of the 12 FIG 2. Normal colonic mucosa covering the surface of a prominent mucosal fold pictured in Figure 1. There is no crypt architectural distortion and the number of lymphocytes and plasma cells is normal (xl50). FIG 3. Higher magnification of the mucosa from the right-fold base of the prominent mucosal fold pictured in Figure 1. There is mildly increased lymphocytes and plasma cells but no crypt architectural distortion. The presence of a Paneth cell suggests that there has been chronic injury. Compare this inflamed region with the normal mucosa from the surface of the prominent mucosal fold (see Fig 2). A colonoscopic biopsy specimen from this region would be abnormal. Procurement of additional biopsy specimens and providing the exact location (base of a prominent fold) would assist in the interpretation that this change is DD-related incidental finding (xl98). AJCP April 1997

GOLDSTEIN AND AHMAD 441 Histology of the Mucosa in Sigmoid Colon Specimens With Diverticular Disease TABLE 1. PROMINENT MUCOSAL FOLDS Specimen Groups DD-Diverticulitis DD-Adenocarcinoma Adenocarcinoma-Only (n = 100) (n = 53) (n = 50) Number of specimens with prominent mucosal folds (%)* 91 (91%) 46 (87%) 3 (6%) Inflammation at bases of prominent mucosal folds (%)* 15 (15%) 5 (9%) 0 Lamina propria fibrosis, crypt elongation, and muscularis 11 (11%) 2 (4%) 0 mucosa hypertrophy(%)* + 'Percent of all specimens. + Mucosa on the surface of the prominent mucosal folds. TABLE 2. HISTOLOGIC FEATURES OF THE MUCOSA SURROUNDING DIVERTICULOSIS OSTIA Distance From Ostia, cm <0.2 0.2-0.5 0.5-1.0 1.0-2.0 Total DD-Diverticulitis specimens (n = 100) Lymphoplasmacytic inflammation 14 8 5 0 27(27%) Neutrophilic inflammation 3 0 0 0 3(3%) DD-Adenocarcinoma specimens (n = 53) Lymphoplasmacytic inflammation 10 2 0 0 12(23%) Neutrophilic inflammation 1 0 0 0 1 (2%) DD = diverticular disease. DD-adenocarcinoma specimens. The lymphoplasmacytic inflammation was mildly increased in all specimens. No basal lamina propria plasmacytosis was present. Three percent of DD-diverticulitis and 2% of DD-adenocarcinoma specimens also had mildly increased neutrophils around the diverticulosis ostia. The mucosal crypts around diverticulosis ostia in DDdiverticulosis and DD-adenocarcinoma specimens were normal. They were evenly spaced and were devoid of branching, shortening, and dilatations. No ulcerations or granulomas were found. All diverticulitis ostia had lymphoplasmacytic inflammation in the surrounding mucosa (Table 3). It was limited to the within 0.2 cm of the diverticulitis ostia in 14% of specimens, 0.5 cm from the ostia in 61%, and 1.0 cm in 25% of specimens. The lymphoplasmacytic inflammation was mildly increased in 66% and markedly increased in the other 34% of the specimens. Neutrophilic inflammation also was present in the mucosa around all diverticulitis ostia. The neutrophils were limited to 0.5 cm around the ostia in 52% of the specimens. In 40% of the specimens, neutrophils extended 0.5 to 1.0 cm from the diverticulitis ostia, and 1 to 2 cm in the other 8%. The neutrophilic inflammation was mild in 89% and marked in the other 11%. There was no crypt architectural distortion, branching, or dilatation, even in the most active areas of inflammation. DISCUSSION Diverticular disease of the sigmoid colon is a common abnormality in people older than 60 years. 1-3 Because of its frequency, DD can coexist with other pathologic conditions. We examined sigmoid colon specimens with diverticular disease to document the baseline mucosal abnormalities associated with diverticulitis and diverticulosis. Our findings can be applied to the interpretation of colonoscopic-procured biopsy specimens. Our results highlight that a single, colonoscopic sigmoid colon biopsy specimen from a patient with DD could easily be misinterpreted. Incidental DDrelated histologic findings could be confused with minimal, but significant, histologic abnormalities of several pathologic conditions if the morphology of a Vol. 107 No. 4

442 ANATOMIC PATHOLOGY Article single sigmoid colon biopsy specimen is evaluated without knowing pertinent clinical information, or the specific location from which the biopsy specimen was obtained. Knowledge of the specific location is key for the weighing of pathologic abnormalities toward DDrelated and, therefore, incidental abnormalities or significant abnormalities that would warrant a diagnosis. Limited information about the location, such as that the biopsy specimen is from the sigmoid colon, is sometimes insufficient. Distinction between incidental and significant abnormalities may only be possible if multiple tissue fragments are obtained. We make these points before discussing specific histologic features to emphasize their importance and to caution the reader about overinterpreting minor histologic changes in a specimen consisting of only one biopsy tissue fragment or insufficient clinical and specific location information. The mucosa around diverticulosis ostia in most DD specimens was normal. There was mild inflammation, predominantly confined to a 0.5-cm diameter region around a minority of diverticulosis ostia. No crypt architectural distortion was associated with the periostial inflammation around diverticulosis ostia. Others have also noted a similar lack of significant mucosal injury. 6-13 The relevance of these findings is that DD is the unlikely cause of moderate or severe lymphoplasmacytic inflammation, any neutrophilic inflammation, crypt architectural distortion, surface ulceration, or granulomas in biopsy tissue fragments procured from an area more than 1 cm from a DD ostia. Rather, other colitides should be considered as causes of the histologic changes, especially Crohn's colitis, chronic infectious colitis, or drug-related colitis. 11 Approximately 90% of DD specimens had prominent mucosal folds. This figure is in contrast with the 6% incidence of prominent mucosal folds in the DDfree sigmoid colon specimens resected for adenocarcinoma. Although prominent mucosal folds are an endoscopic observation and would not be identified on a colonoscopic biopsy specimen, there were several findings related to the prominent mucosal folds that could be observed on a colonoscopic biopsy tissue fragment. Approximately 15% of DD specimens with prominent mucosal folds had less than 0.5-cm lengths of mucosa with mild lymphoplasmacytic inflammation localized to the bases of prominent mucosal folds. If a biopsy specimen was procured from this region, it would seem that there was focal mucosal lymphoplasmacytic inflammation, confronting the pathologist with making a decision about the significance of the inflammation. Should the mild lymphoplasma- cytic inflammation be interpreted as incidental, or does it warrant a pathologic diagnosis? There are two factors related to evaluating the significance of focal lymphoplasmacytic inflammation, the patient's symptoms and the appearance of the colon. The diseases or histologic patterns that are commonly associated with a symptomatic patient and an endoscopically normal colon are Crohn's colitis, lymphocytic colitis, and collagenous colitis. Crohn's colitis enters into this discussion because it typically involves the sigmoid colon and rectum when it occurs in older patients. Crohn's colitis can have no or minimal symptoms and a normal-appearing endoscopy with microscopic abnormalities. 14-18 Patients with Crohn's disease or DD can have similar signs and symptoms, making clinical distinction between the two diseases difficult. 12,19-21 One histologic pattern of Crohn's colitis has patchy inflammation with minimal or no crypt architectural distortion. A distinction between DDrelated inflammation and Crohn's disease may not be possible if only one biopsy tissue fragment is obtained. However, if numerous tissue fragments are procured, Crohn's colitis usually will have more than one abnormal biopsy fragment, while DD-related inflammation will probably be limited to a single biopsy fragment. Recently, DD-associated inflammatory bowel diseaselike colitis was described that mimics active inflammatory bowel disease. 4,5 The mucosal changes we describe in this study are usually in the context of a patient who has minimal or no symptoms. Biopsy specimens from a patient with lymphocytic or collagenous colitis have diffuse lymphoplasmacytic FIG 4. Prolapse-type mucosal changes on the surface of a prominent mucosal fold. The muscularis mucosa is hypertrophic and hyperplastic. Slips of smooth muscle extend up between elongated crypts toward the surface. The lamina propria is fibrotic (xl60). AJCP' April 1997

GOLDSTEIN AND AHMAD 443 Histology of the Mucosa in Sigmoid Colon Specimens With Diverticular Disease TABLE 3. HISTOLOGIC FEATURES OF THE MUCOSA SURROUNDING DIVERTICULITIS OSTIA FROM DD-DIVERTICULITIS SPECIMENS (N = 100) Distance From Ostia, cm <0.2 0.2-0.5 0.5-1.0 1.0-2.0 Total Lymphoplasmacytic inflammation, % Neutrophilic inflammation, % 14 17 61 35 25 40 0 8 100(100%) 100 (100%) DD = diverticular disease. mucosal inflammation without crypt architectural distortion. 22 ' 23 There is a concomitant increase in intrasurface epithelial lymphocytes and subsurface epithelial collagen in the latter colitis, features that were not observed in DD-related inflammation. These colitides could mimic DD-related inflammation when the intraepithelial lymphocytes or the amount of subsurface collagen are minimally increased. The inflammation in these two colitides is usually diffuse, involving all the biopsy fragments, while the inflammation in DD is patchy, limited to the bases of the prominent mucosal folds. Similar to Crohn's colitis, distinction between DD-associated inflammation and lymphocytic or collagenous colitis may not be possible if only one biopsy fragment is obtained. Procurement of more than 6 biopsy fragments, in our experience, provides sufficient tissue to detect the focal or diffuse nature of the inflammation, increased intraepithelial lymphocytes, and, in collagenous colitis, increased subsurface collagen. Finally, the clinical scenario of collagenous colitis or lymphocytic colitis is one of nonbloody diarrhea, which is unusual for DD. Focal lymphoplasmacytic inflammation in a biopsy specimen from a normal-appearing colon in an asymptomatic patient probably has no clinical significance. Our results suggest that DD is one cause of focal lymphoplasmacytic inflammation found in biopsy specimens from endoscopically normal colons. These changes are similar to the clinically insignificant focal active colitis that can also be observed in biopsy specimens procured during colonoscopy performed for adenoma surveillance. 24 Other changes related to the prominent folds of DD that could be observed on a colon biopsy specimen were areas of lamina propria fibrosis, muscularis mucosae hypertrophy and hyperplasia, and crypt elongation located on the tips of the prominent mucosal folds. Eleven percent of the DD-diverticulitis and 5% of DD-adenocarcinoma specimens had them. They were microscopically observed and did not form a grossly seen polypoid lesion in most cases. The lower incidence (5%) of these changes in DDadenocarcinoma may reflect the lower number of sections procured from the normal mucosa compared with the number procured from the DD-diverticulitis specimens. These histologic changes are similar to those of mucosal prolapse syndrome occurring in the rectum and anus that has been termed solitary rectal ulcer syndrome and inflammatory cloacogenic polyp, respectively. 25-28 We believe that these changes are the progenitor lesions to the grossly observed polyps of DD, termed polypoid prolapsing mucosal folds and eroded polypoid hyperplasia of the rectosigmoid by different authors. 6,29-31 The prolapsed polypoid mucosal folds can be red and friable at colonoscopy. 29,31 ' 32 Our results suggest that microscopic prolapse-type mucosal changes are commonly present in DD and usually do not form a grossly observable lesion. We believe that DD is one cause of focal prolapse-type mucosal changes that can be occasionally observed on a biopsy specimen procured from the sigmoid colon. These changes are most likely an incidental finding and not clinically significant. The sigmoid colonic mucosa between and around DD ostia is usually normal. A minority of DD cases have patchy mild inflammation at the bases of prominent mucosal folds or immediately around DD ostia. This inflammation may be one cause of mild patchy inflammation that is occasionally observed in sigmoid colon biopsy specimens. Sometimes, the histologic changes of mucosal prolapse can be microscopically present on the surfaces of prominent mucosal folds in diverticular disease, and sometimes they can form polypoid lesions. Diverticular disease should be considered as the cause of mucosal prolapse changes when the latter are observed in a biopsy specimen from the sigmoid colon. Discrimination between a DD-related incidental finding and significant pathologic abnormality can frequently be made with the procurement of multiple biopsy specimens. A distinction may not be possible if only one tissue fragment is procured or if the specific location from which the biopsy specimens Vol. 107 No. 4

444 ANATOMIC PATHOLOGY Original Article were procured or a description of the appearance of the colon at colonoscopy is not provided for the pathologist. Pathologists should be cognizant of these findings when evaluating biopsy specimens from sigmoid colons that have diverticular disease. Acknowledgment: We thank J. Watts, MD, J. Neill, MD, and R. Goldstein, PhD, for their editorial comments. REFERENCES 1. Almy TP, Howell DA. Diverticular disease of the colon. N Engl ] Med. 1980;302:324-331. 2. Hughers LE. Postmortem survey of diverticular disease of the colon. Gut. 1969;10:344-351. 3. Parks TG. Natural history of diverticular disease of the colon. Clin Gastroenterol. 1975;4:53-69. 4. Makapugay LM, Dean PJ. Diverticular disease-associated chronic colitis. Am j Surg Pathol. 1996;20:94-102. 5. Hart J, Baert F, Hanauer S. Sigmoiditis: a clinical syndrome with a spectrum of pathologic features, including a distinctive form of 1BD. Mod Pathol. 1995;8:62. Abstract. 6. Rhatigan RM, Saffos RO. Mucosal hyperplasia in colonic diverticula. Histopathology. 1979;3:153-160. 7. Seldenrijk CA, Morson BC, Meuwissen SGM, et al. Histopathological evaluation of colonic mucosal biopsy specimens in chronic inflammatory bowel disease: diagnostic implications. Gut. 1991;32:1514-1520. 8. Arfwidsson S. Pathogenesis of multiple diverticula of the sigmoid colon in diverticular disease. Acta Chir Scand. 1964;342(suppl):5-68. 9. Schmidt GT, Lennard-Jones JE, Morson BC, Young AC. Crohn's disease of the colon and its distinction from diverticulitis. Gut. 1968;9:7-16. 10. Schmidt GT. Crohn's disease of the colon and diverticulitis. Med j Aust. 1969;(December 20 suppl):1253-1258. 11. du Boulay CEH. Diverticular disease, solitary ulcer syndrome, anorectal prolapse, and incontinence. In: Whitehead R, ed. Gastrointestinal and Oesophageal Pathology. New York, NY: Churchill Livingstone; 1989:355-360. 12. Grimm IS, Friedman LS. Inflammatory bowel disease in the elderly. Gastroenterol Clin North Am. 1990;19:361-389. 13. O'Leary AD, Sweeney EC. Lymphoglandular complexes of the colon: structure and distribution. Histopathology. 1986;10:267-283. 14. Shapiro PA, Peppercorn MA, Antonioli DA, loffe N, Goldman H. Crohn's disease in the elderly. Am J Gastroenterol. 1981;76:132-137. 15. Tchircow G, Lavery IC, Fazio VW. Crohn's disease in the elderly. Dis Colon Rectum. 1983;26:177-181. 16. Carr N, Schofield PF. Inflammatory bowel disease in the elderly patient. Br]Surg. 1982;69:223-225. 17. Fabricius PJ, Gyde SN, Keighley MRB, Alexander-Williams J, Allan RN. Crohn's disease in the elderly. Gut. 1985;26:461^65. 18. Berman IR, Corman ML, Coller JA, Veidenheimer MC. Late onset Crohn's disease in patients with colonic diverticulitis. Dis Colon Rectum. 1979;22:524-529. 19. Klein S, Mayer L, Present DH, et al. Extraintestinal manifestations in patients with diverticulitis. Ann Intern Med. 1988;108:700-702. 20. Morgenstern L, Weiner R, Michel SL. 'Malignant' diverticulitis. Arch Surg. 1979;114:1112-1116. 21. Harper PC, McAuliffe TL, Beeken WL. Crohn's disease in the elderly: a statistical comparison with younger patients matched for sex and duration of disease. Arch Intern Med. 1986;146:753-755. 22. Ettinghausen SE. Collagenous colitis, eosinophilic colitis, and neutropenic colitis. Surg Clin North Am. 1993;73:993-1016. 23. Yardley JH, Lazenby AJ, Giardiello FM, Bayless TM. Collagenous 'microscopic' colitis, lymphocytic colitis, and other gentler and more subtle forms of colitis. Hum Pathol. 1990;21:1089-1091. 24. Stern RS, Carpenter SL, Barnett JL, Greenson JK. The clinical significance of focal active colitis. Mod Pathol. 1996;9:66. Abstract. 25. du Boulay CEH, Fairbrother J, Isaacson PG. Mucosal prolapse syndrome-unifying concept for solitary ulcer syndrome and related disorders. / Clin Pathol. 1983;36:1264-1268. 26. Levey JM, Banner B, Darrah J, Bonkovsky HL. Inflammatory cloacogenic polyp: three cases and literature review. Am } Gastroenterol. 1994;89:438-441. 27. Saul SH. Inflammatory cloacogenic polyp: relationship to solitary rectal ulcer syndrome/mucosal prolapse and other bowel disorders. Hum Pathol. 1987;18:1120-1125. 28. Chetty R, Bhathal PS, Slavin JL. Prolapse-induced inflammatory polyps of the colorectum and anal transition zone. Histopathology. 1993;23:63-67. 29. Franzin G, Fratton A, Manfrini C. Polypoid lesions associated with diverticular disease of the sigmoid colon. Gastrointest Endosc. 1985;31:196-199. 30. Kelly JK. Polypoid prolapsing mucosal folds in diverticular disease. Am ] Surg Pathol. 1991;15:871-878. 31. Burke AP, Sobin LH. Eroded polypoid hyperplasia of the rectosigmoid. Am j Gastroenterol. 1990;85:975-980. 32. Mathus-Vliegen EM, Tytgat GN. Polyp-simulating mucosal prolapse syndrome in (pre-) diverticular disease. Endoscopy. 1986;18:84-86. AJCP April 1997