Background mucosal changes of primary advanced large intestinal cancer in patients without familial polyposis coli

First observation of microadenomas in the ileal mucosa of patients with familial adenomatous polyposis and colectomies

BACKGROUND & AIMS

Little data are available on the type and prevalence of mucosal changes involved in the development of ileal adenomas in patients with familial adenomatous polyposis who have undergone colectomy. However, colonic metaplasia of the ileal epithelium is thought to be an important step in the development of such adenomas.

METHODS

Retrograde endoscopy and biopsy of the distal ileum were performed in 17 affected patients who underwent total colectomy or proctocolectomy 3-184.1 months (mean, 80.3 +/- 13.9 months) before the study.

RESULTS

Macroscopic ileal polyps were identified in 14 (82.4%) patients. All polyps were sessile and 1-5 mm in size. Histological analysis showed adenomas in 9 (52.9%) patients and lymphoid hyperplasia or inflammation in the others. In 1 patient, an area of colonic-type metaplasia of the ileal mucosa was found close to an adenoma. However, in 5 (29.4%) patients, random biopsy specimens of the normal-appearing mucosa showed foci of abnormal crypts in the absence of metaplasia, with histological appearance similar to the findings described in dysplastic aberrant crypt foci of the colon. Such lesions, previously observed only in colorectal mucosa and referred to as microadenomas or oligocryptal adenomas, are considered putative preneoplastic abnormalities.

CONCLUSIONS

Although the hypothesized sequence normal ileal mucosa leading to colonic-type metaplasia leading to adenoma cannot be excluded, our findings support the sequence normal ileal mucosa leading to microadenoma leading to gross adenoma and possibly cancer as the main histogenetic pathway, as already suggested for the large bowel.

Identification and quantification of aberrant crypt foci and microadenomas in the human colon

The objective of the present study was to determine whether aberrant crypt foci (ACF) similar to those observed in the colons of experimental animals exposed to colon carcinogens could be identified and quantified in the human colon. Twenty-seven colon resections from patients affected by familial adenomatous polyposis (FAP, five cases), colorectal cancer (CRC, 12 cases), and benign diseases of the large bowel (BD, 10 cases) were collected from a pathology repository or immediately after operation. Ten or more 1-cm2 formalin-fixed, methylene-blue--stained samples of colonic mucosa from each colon were scored under light microscopy for ACF. The number of ACF per cm2 and the number of crypts per ACF for each colon were calculated. The average number of ACF per cm2 in the FAP group (20 +/- 19, mean +/- SD) was significantly higher (P less than 0.01) than those of the CRC (0.37 +/- 0.41) and BD (0.18 +/- 0.35) groups. At least one ACF was found in every colon resection from CRC patients and in six out of 10 colon resections from the BD group. The average number of crypts per ACF ranged from five to 35 with absolute values from 1 to over 100. Fifty-five histologic specimens, 43 with ACF of various size and 12 without, were prepared by sectioning the colon parallel to the mucosal surface. There was a close correlation between the number of crypts per ACF in each specimen as scored by methylene-blue and histologic examination. Twenty-six aberrant crypt foci displayed dysplasia as evident by histologic analysis. In these instances we feel the term microadenoma is appropriate and, using this unique approach of examining the human colon, they can be easily identified and quantified. These lesions may well be precursors for adenomatous polyps and colorectal cancer.

A comparative case-control study of colorectal cancer and adenoma

We conducted a comparative case-control study of colorectal cancer and adenoma involving 221 cases with colorectal cancer, 525 cases with colorectal adenoma and 578 neighborhood controls. Daily vegetables intake was associated with lower risks of distal colon adenoma (relative risks (RR) = 0.59, 95% confidence interval (CI): 0.39-0.89) and rectal cancer (RR = 0.46, 95% CI: 0.25-0.84). Daily beans intake was associated with lower risk of colon adenoma (RR = 0.58, 95% CI: 0.37-0.91 for the proximal colon and RR = 0.63, 95% CI: 0.45-0.88 for the distal colon) and daily intake of seaweeds was associated with lower risk of rectal cancer (RR = 0.42, 95% CI: 0.22-0.82). Daily intake of fish and shellfish also showed an inverse association with the risk of colon adenoma (RR = 0.67, 95% CI: 0.45-0.99 for the proximal colon and RR = 0.70, 0.52-0.94 for the distal colon). Generally, intakes of animal or vegetable fat-rich foods, especially meats, were associated with decreases in risks of both adenoma and cancer, though the association of cancer was not statistically significant. Other than dietary factors, daily alcohol drinking was associated with an increased risk of adenoma in the proximal colon (RR = 1.95, 95% CI: 1.15-3.29) and ex-drinkers showed higher risks for colon adenoma and colorectal cancer. Sports or occupational activities and coffee drinking were inversely associated and family history of colorectal cancer was positively associated with the risk of both colorectal adenoma and cancer.

Spatial clustering of multiple hyperplastic, adenomatous, and malignant colonic polyps in individual patients

Analysis of relative polyp locations in 426 consecutive patients with multiple colonic polyps found on colonoscopy showed novel findings. First, synchronous and metachronous neoplastic polyps showed spatial clustering in individual patients. For example, patients with their largest neoplasm in the cecum or proximal ascending colon, had 34.3 percent +/- 4.6 percent (standard error) of their other colonic neoplasms in the same location. Second, hyperplastic polyps showed spatial clustering in individuals that was statistically significantly greater than expected from the increased hyperplastic polyp concentration in the rectum and sigmoid. Third, hyperplastic polyps showed spatial clustering with neoplastic polyps; this clustering was similar in magnitude to clustering for exclusively hyperplastic or neoplastic polyps. In contrast, lipomas were not spatially clustered with hyperplastic and neoplastic polyps. The magnitude of clustering between hyperplasia and neoplasia showed a closer association between these histologic types than previously appreciated. Because of clustering, regions with prior polyps appear to merit closer surveillance. These findings suggest clinical study, using a randomized controlled clinical trial, of whether a patient who had only rectal and sigmoid adenomas on initial and follow-up colonoscopy should have surveillance with flexible sigmoidoscopy alternating annually with colonoscopy. A patient with a prior cecal adenoma should have surveillance only with a complete colonoscopy or adequate cecal views on barium enema.

Colorectal adenomas: morphologic features and the risk of developing metachronous adenomas and carcinomas in the colorectum

The morphologic features of 307 colorectal adenomas among 159 patients are reviewed. Most adenomas (66.4%) were located in the sigmoid colon and the rectum, and the percentage decreased proximally to the right colon. The 307 adenomas comprised 244 (79.5%) tubular, 41 (13.3%) tubulovillous, and 22 (7.2%) villous adenomas. The epithelial dysplasia was graded as mild in 260 (84.7%) adenomas, moderate in 33 (10.7%), and severe in 14 (4.6%). The percentage of severe dysplasia was greater in villous adenomas than in tubular adenomas (p less than 0.05) and correlated with the increasing size (greater than 5 mm) of the adenomas (p less than 0.01). The risk of metachronous adenomas could be evaluated among 56 patients, 34 men and 22 women with a history of removed adenoma(s). Fourteen of 56 patients (25%) with 6:1 male to female ratio developed 18 new adenomas, after an average of 34.3 months (range, from 12 to 88 months). Eleven of the 14 patients had multiple adenomas at the initial examination. In addition, a carcinoma of the rectum was found in one male patient. Of the 48 patients, 17 men and 31 women, operated on for colorectal cancer 16 patients (34%) with 1.3:1 male to female ratio had 40 new adenomas after an average of 51.8 months (range, 12 to 252 months) after the surgical excision of their carcinomas. One patient had a recurrent carcinoma at the site of the anastomosis 22 months after anterior resection of his carcinoma. Our data suggest that a history of colorectal carcinoma, multiple adenomas, and male sex predict a higher risk of having future colorectal tumours.

Lectin-binding sites in normal, hyperplastic, adenomatous and carcinomatous human colorectal mucosa

The carbohydrate structures of cellular glycoconjugates in normal, hyperplastic, adenomatous and carcinomatous human colorectal mucosa were analysed with six fluorescein isothiocyanate-conjugated lectins. In normal, hyperplastic and adenomatous colorectal mucosa showing mild or moderate dysplasia Concanavalin A (Con A), Lens culinaris (LCA), and wheat germ (WGA) agglutinins stained goblet cell glycoconjugates (actual mucin goblet itself) while peanut (PNA), Vicia villosa (VVA), and Griffonia simplicifolia-II (GSA-II) agglutinins showed a supranuclear staining of goblet cell glycoconjugates. After neuraminidase treatment of tissue sections PNA and VVA stained mucin goblets of mature cells in normal mucosa, while less differentiated cells in the lower crypt displayed a supranuclear staining with VVA. The mucin goblets in adenomatous mucosa with mild or moderate dysplasia did not stain with PNA and VVA, neither before nor after neuraminidase treatment. Areas of in situ cancer in adenomas and carcinomas displayed a strong and direct binding of Con A, LCA, WGA and PNA in an apical linear distribution, while the binding of VVA and GSA-II was heterogeneous. We conclude that there are alterations in the carbohydrate structures of cellular glycoconjugates, which can be related to goblet cell differentiation in normal colorectal mucosa and to the degree of dysplasia in adenomas. Heterogeneous and incompletely glycosylated glycoconjugates appear to be synthesized by the majority of colorectal carcinomas.