Incidence and distribution of hybrid goblet cells in complete type intestinal metaplasia of the stomach

Previous reports suggest that hybrid goblet cells (HGCs) sharing both gastric and intestinal mucin phenotypes are rarely observed in complete intestinal metaplasia (cIM) of the stomach. However, we have made a different observation. Thus, we compared the incidence and distribution of HGCs within the tubules of gastric cIM and the duodenum in order to define the significance of HGCs. Fifteen antral sections and 16 fundic sections from tissue with cIM and gastric cancer, as well as 19 sections from duodenal tissue with cancer of the Papilla of Vater, were stained for human gastric mucin (HGM), Con A, MUC2, CD10, and Ki-67. Multivariate analysis showed that antral location, a distance of 5mm or less from the tumor margin, and the presence of underlying pyloric glands were significant predictive factors for tubules containing >50% HGCs as part of their goblet cell population. The incidence of tubules with HGCs differed significantly in tissue samples from the antrum, body and duodenum. HGCs did not stain for Ki-67 and were not surrounded by gastric foveolar-type epithelium within the tubules of cIM foci. These findings indicate that alterations in the proportion of HGCs may occur under some circumstances, and that HGCs are not precursors to gastric foveolar-type cells in the stomach and duodenum.     

Morphometric definition and grading of gastric intestinal metaplasia

Type I and type III intestinal metaplasia in gastric mucosa have been examined using morphometric methods. Tissue (volume per cent gland, lumen, epithelium, goblet cell vacuoles) and nuclear parameters (area, with related standard deviation, and form factors) were used as indicators of gland crowding, nuclear-cytoplasmic ratio, nuclear atypia, and pleomorphism. In type III intestinal metaplasia, there is significantly (i) greater nuclear pleomorphism, (ii) a higher nuclear-cytoplasmic ratio, and (iii) smaller and less numerous goblet cell vacuoles in both the upper and the lower parts of the crypts. These two parameters have significantly higher values in the lower than in the upper part of individual crypts. No cell population with large pleomorphic nuclei characterized type III metaplasia, though there was more variation in nuclear size.     

Coexistence of gastric- and intestinal-type endocrine cells in gastric and intestinal mixed intestinal metaplasia of the human stomach

Intestinal metaplasia (IM) in the human stomach has previously been classified into a gastric and intestinal mixed (GI-IM) and a solely intestinal phenotype (I-IM). The phenotypes of mucous and endocrine cells were evaluated in 3034 glandular ducts associated with chronic gastritis. In the pyloric region, the relative expression of gastric endocrine cell markers, such as gastrin and somatostatin, decreased gradually from glandular ducts with only gastric mucous cell phenotype (G type) to GI-IM toward I-IM, while that of the intestinal endocrine cell markers, glicentin, gastric inhibitory polypeptide (GIP), and glucagon-like peptide-1 (GLP-1) was inversely correlated. In the fundic region, gastrin-positive cells emerged in the pseudo-pyloric and GI-IM glands, whereas I-IM glands did not possess any gastrin-positive cells, suggesting the presence of a distinct pathway of intestinalization. Double staining revealed coexistence of gastrin- and GLP-1-positive cells in the same gland and occasionally in the same cell in GI-IM glands. These results suggest that the phenotypes of endocrine cells are in line with those for mucous counterparts and support the concept that all of the different types of mucous and endocrine cells in normal and IM glands might be derived from a single progenitor cell in each gland.     

Gastric and intestinal mixed and solely intestinal types of intestinal metaplasia in the human stomach

To generate a novel understanding of Intestinal metaplasia (IM) on the basis of cellular differentiation status, a total of 132 gastric surgical specimens were studied using gastric and small intestinal cell markers by much histochemical and Immunohistochemical techniques. The cases were divided into two types: (i) gastric and intestinal (GI) mixed type; and (ii) solely intestinal (I) type, with the reference to the presence of gastric and/or intestinal cell markers. The GI mixed type was subdivided into six subtypes: (i) a subtype consisting of surface mucous (Su), pyloric gland (Py), Intestinal absorptive (Ab), and goblet (Go) cells, but lacking Paneth (Pa) cells, GI(Pa-); (ii) a GI(Pa-) subtype without Py cells, GI(Py-, Pa-); (iii) a GI(Pa-) subtype without Su cells, GI(Su-, Pa-); (iv) a GI(Su-, Pa-) subtype with Pa cells, GI(Su-, Pa+); (v) a Gi(Pa-) subtype with Pa cells, GI(Pa+); and (vi) a GI(Pa+) subtype without Py cells, GI(–, Pa+).The I type was subdivided Into: (I) a subtype consisting of cells with Ab and Go cells, I(Pa-); and (ii) a I(Pa-) subtype with Paneth cells, I(Pa+). The GI mixed subtypes, except for the GI(Py-, Pa-) and GI(Py-, Pa+), were characterized by Intestinalized gastric plts connected with underlying pyloric glands. Immunohistochemical staining of proliferating cell nuclear antigen (PCNA) revealed a common prolifemtive cell zone between the two. The GI mixed type, especially the GI(Pa-) subtype, predominated in the pyloric mucose, while the I type was most frequent In the fundle region, suggesting that the pathogenesis of IM differs between these two locations. The results of the study confirm that IM is an abnormal and unstable differentiation status between the stomach and small Intestine.     

Pathology of the gastric antrum and body associated with Helicobacter pylori infection in non-ulcerous patients: is the bacterium a promoter of intestinal metaplasia?

A series of 115 consecutive, non-ulcerous, dyspeptic patients were examined for Helicobacter pylori (H. pylori) colonization in the gastric antral and/or body mucosa using Giemsa staining. Findings were correlated with the presence and degree of activity of superficial gastritis, deep gastritis, atrophic gastritis and with the presence of intestinal metaplasia. The prevalence of H. pylori positivity was 61.7%. In 59 of the 71 positive patients (83%), H. pylori was detected in the antrum or in both the antral and oxyntic mucosa. In the remaining 12 positive patients, H. pylori was detected only in the oxyntic mucosa and in all these cases, the antrum showed intestinal metaplasia associated with atrophic gastritis (25%). In both antral and oxyntic mucosa, the activity of the gastritis was significantly correlated with H. pylori colonization. Linear logistic regression analysis showed that in patients with intestinal metaplasia the presence of H. pylori infection was significant in predicting the presence of more extensive intestinal metaplasia after adjusting for age. The prevalence of intestinal metaplasia types II and III was 65.5% in the H. pylori positive and 25% in the H. pylori negative patients. The antral mucosa is thought to be the elective site for H. pylori related histological lesions. At a later stage, H. pylori can be detected only in the oxyntic area while the antral mucosa shows extensive metaplastic or atrophic lesions. We would suggest that H. pylori plays a promotional role in the morphogenesis of intestinal metaplasia.     

Immunohistochemical detection of O-acetylated sialomucins in intestinal metaplasia and carcinoma of the stomach

Eighty-two selected gastric mucosal biopsy or resection specimens were stained both conventionally, to classify subtypes of intestinal metaplasia and carcinoma, and immunohistochemically with a mouse monoclonal antibody (MMM-17), raised against normal human colonic mucin, which has an affinity for di- and/or tri-O-acetylated sialomucin. The aims of the study were to reassess the prevalence of O-acetylated sialomucins in normal, metaplastic and carcinomatous gastric mucosa and to investigate whether the production of these mucins by intestinal metaplasia is related to its associated mucosal pathology. O-acetylated sialomucins were not seen in normal mucosa. They were, however, prevalent in all sub-types of metaplastic (64.8%) and carcinomatous (42.9%) mucosa. Type 1 intestinal metaplasia was significantly more likely to contain this type of mucin if Helicobacter pylori infection was identifiable in the adjacent gastric mucosa (81.0% v. 38.5%, P < 0.025). Type 3 showed a similar, albeit nonsignificant, relationship (100% v. 62.5%). O-acetylated sialomucins are, therefore, much more prevalent in gastric intestinal metaplasia and carcinoma than previously recognized by conventional staining techniques. The production of this type of mucin by intestinal metaplasia may reflect an adaptive response to alterations in the luminal environment such as an increase in bacterial content.     

Gastric intestinal metaplasia type III cases are classified as low-grade dysplasia on the basis of morphometry

The aim of this study was to try to place gastric intestinal metaplasia, type III (type III IM) in the stepwise chain of events from atrophic gastritis to cancer. A number of dysplastic, periulcer regenerative, and type III IM lesions were qualitatively diagnosed (and graded) blindly and independently by several pathologists. These lesions were further analysed by means of quantitative parameters, with the aim of differentiating dysplastic from regenerative changes. Inconsistencies between the qualitative and quantitative classification (about 7 per cent of cases) were eliminated and homogeneous groups (low-grade dysplasia, high-grade dysplasia, regenerative changes) were obtained. These cases were taken as the gold standard against which type III IM was compared. The results indicate that the great majority (91·4 per cent) of cases of type III IM fulfil the nuclear and architectural criteria for low-grade dysplasia.     

Endocrine cells in intestinal metaplasia of the stomach

In this study we have investigated the mucin profile and the endocrine cell population in gastric endoscopic biopsies from 22 patients affected by chronic gastritis and intestinal metaplasia and in five surgical specimens of stomachs removed because of intestinal-type carcinoma (4) or peptic ulcer (1). High iron diamine-Alcian blue (HID-Ab) staining and peptide immunocytochemistry (peroxidase anti-peroxidase technique) were used. Forty-one foci of intestinal metaplasia were detected, 15 produced sulphomucins and 26 sialomucins. Of the endocrine cells investigated, gastrin and somatostatin cells were the most frequently observed, while cholecystokinin, glucose-dependent insulinotropic peptide-, secretin- and enteroglucagon-containing cells were also found in the metaplastic areas, but less frequently. No significant correlation was found between the type of mucin and the types of endocrine cells present, the latter usually resembling those normally found in the small intestine. On the basis of these results we conclude that intestinal metaplasia involves mucin- and peptide-producing cells of the stomach in a variable manner, with no correlation between the two.     

Tissue microarray analysis of multiple gene expression in intestinal metaplasia, dysplasia and carcinoma of the stomach

AIMS: To study multiple gene expression patterns and their roles in the process of gastric carcinogenesis. METHODS AND RESULTS: Using a high-throughput tissue microarray technique, 169 specimens from gastric carcinomas, precursor lesions and normal mucosa were immunostained on a series of tissue chips for p53, p21(WAF1/CIP1) cyclin E, Bcl-2, c-met and mucin 5AC expression. The overexpression of p53 was observed in 10.7% of low-grade dysplasia (LGD), 38.1% of high-grade dysplasia (HGD) and 39.6% of intestinal type gastric carcinoma (IGC). Expression of p21(WAF1/CIP1) was found in 47.6% of incomplete intestinal metaplasia (IM), 36.7% of dysplasia (Dys) and 29.5% of IGC. The overexpression of cyclin E was more frequently present in carcinomas than in Dys (P < 0.05); moreover, high-level expression (> 25% in extent) of cyclin E was observed only among IGC. Abnormal Bcl-2 expression was present in 81.0% of incomplete IM, 69.4% of Dys and 22.9% of IGC. Along with progression of the lesion, the expression of c-met increased; in contrast, mucin 5AC decreased gradually. CONCLUSIONS: The specific expression pattern in incomplete IM was mucin 5AC+/Bcl-2+/p53-/cyclin E-, while mucin 5AC-/cyclin E+ was specific for IGC. p53 was useful for distinguishing LGD from HGD. High-level expression of cyclin E might be an indicator for malignant transformation of dysplasia.     

CDX2 co-localizes with liver-intestine cadherin in intestinal metaplasia and adenocarcinoma of the stomach

CDX2 and liver-intestine (LI)-cadherin are intestine-specific markers and both are physiologically expressed in the small intestine and colon. Recent studies have demonstrated that CDX2 regulates LI-cadherin gene (CDH17) expression in colorectal cancer. The present study investigated the relationship of CDX2 and LI-cadherin expression in gastric cancer. One hundred and nine pairs of tumour and non-cancerous gastric mucosa were collected from gastrectomy specimens. Protein expression levels of CDX2 and LI-cadherin were determined by immunohistochemical staining. Semi-quantitative RT-PCR showed that the mRNAs of both CDX2 and CDH17 were highly expressed in tumour compared with non-cancerous mucosa. Overexpression of CDX2 was significantly associated with CDH17 in gastric adenocarcinoma. Furthermore, the expression of CDX2 and LI-cadherin proteins was strongly coupled in intestinal metaplasia. In conclusion, overexpression of CDH17 is significantly associated with CDX2.     

The prognostic value of sulphomucin positive intestinal metaplasia in the development of gastric cancer

Gastric biopsy specimens from 230 patients with chronic atrophic gastritis were investigated with the use of mucin stains for the presence and type of intestinal metaplasia. Metaplasia was not shown in 59 cases; 92 of the 171 cases with metaplasia were sulphomucin positive and 79 were negative. The patients were followed-up from 1976 to 1985. Eight patients were registered as having gastric cancer over this period. However, five of them had to be eliminated from the study because on careful review of all the clinical data it was clear that they had gastric cancer at the time of the biopsy. Two of the three remaining patients had sulphomucin negative biopsies. Thus, only one patient out of 90 with chronic atrophic gastritis and sulphomucin positive intestinal metaplasia developed gastric cancer when followed-up for 8-9 years. None of the patients with unequivocal type IIb metaplasia developed gastric cancer. We conclude that sulphomucin positive intestinal metaplasia does not identify a high risk group and its recognition is thus of no value in surveillance for gastric cancer.     

Distinction between intestinal metaplasia in the cardia and in Barrett's esophagus: the role of histology and immunohistochemistry

Intestinal metaplasia in Barrett's esophagus (BIM) is a precancerous condition, whereas the carcinogenic potential of intestinal metaplasia of the cardia (CIM) is uncertain. Although clinically important, histological distinction between both conditions by endoscopic biopsies is considered problematic. In the present study, 4-mm samples of BIM (n=31) and CIM (n=9) were selected from esophagectomy specimens that had been resected for esophageal cancer. Slides were coded and stained with hematoxylin and eosin (H&E), Alcian blue-periodic acid-Schiff (PAS), cytokeratins (CK) 7 and 20, and CD10, which labels the intestinal brush border. The predictive value of these stains for the recognition of BIM and CIM was evaluated independently by two senior pathologists. With the use of H&E-stained slides exclusively, BIM samples were categorized correctly in 93.5% and 83.9% of cases (pathologists 1 and 2, respectively), and CIM samples, in 100% and 88.9% of cases. Alcian blue-PAS-positive goblet cells were identified by both investigators in all BIM and CIM samples. BIM-typical CK 7 and 20 immunostaining pattern was identified in 90.3%/83.9% of BIM but only in 11.1%/11.1% of CIM. CD10-positive brush border was present in 32.3%/25.8% of BIM and in 88.9%/88.9% of CIM. When HE-stained slides and immunohistologically stained slides were used together for tissue recognition, BIM were categorized correctly in 90.3%/80.6% of cases, and CIM, in 88.9%/88.9% of cases. In conclusion, BIM and CIM can be usually distinguished on the basis of HE sections. CK 7 and CK 20 expression pattern analysis discriminates correctly between BIM and CIM in the majority of cases. CD10-positive intestinal brush border is present in the majority of CIM but only in a minority of BIM. However, immunohistochemical investigations could not improve the diagnostic accuracy of HE histology alone.     

Fatty acid synthase is highly expressed in carcinoma, adenoma and in regenerative epithelium and intestinal metaplasia of the stomach

AIMS: To investigate the relation of fatty acid synthase (FAS) expression to the clinicopathological characteristics of gastric cancers and gastric tumorigenesis. METHODS AND RESULTS: FAS expression was examined immunohistochemically in 626 gastric cancers, 51 gastric adenomas, and non-neoplastic epithelium contiguous with cancer tissue including normal foveolae, intestinal metaplasia, regenerative epithelium, and gastric glands. FAS expression was found in more than 70% of gastric cancers. Interestingly, it occurred preferentially in differentiated carcinomas, male cases, and in patients over 51 years old. Although previous reports showed that FAS expression is closely related to cancer progression, in gastric cancers FAS expression had no relationship with prognosis, cancer progression as indicated by depth of invasion, venous and lymphatic permeation, and distant metastasis. Gastric tubular adenoma and intestinal metaplasia, which are thought to be associated with well-differentiated gastric carcinomas, showed a frequency of FAS expression similar to that of differentiated carcinomas; however, normal foveolae and gastric glands showed no or less FAS expression. CONCLUSIONS: FAS expression occurs at the early stage of tumorigenesis and plays important roles in the formation of precancerous foci rather than in the progression of carcinoma of the stomach.     

A variant of intestinal metaplasia associated with gastric carcinoma: a histochemical study

Gastrectomy specimens including 32 cases of carcinoma and 35 cases of peptic ulcer were examined in detail histologically and histochemically to show the distribution of intestinal metaplasia (IM) and types of mucin secreted. A variant of IM, with extensive involvement of both antrum and body and excessive secretion of sulphomucin, was found to be associated with carcinomas showing appreciable extracellular sulphomucin secretion. The identification of such a variant of IM may help the interpretation of gastric biopsies.     

Mucin histochemistry and lectin binding sites in intestinal metaplasia of the urinary bladder

We report a case of intestinal metaplasia of the bladder urothelium associated with dysplastic foci and a transitional cell carcinoma. A mixture of sialomucins, O -acetylated sialomucins and sulphomucins was found in the goblet cells. Neuraminidase resistant binding of the lectin peanut agglutinin was demonstrated in the brush border of columnar cells in intestinal metaplasia and diffusely in columnar cells in dysplastic foci. The histochemical findings are compared with those described in normal, dysplastic and neoplastic colonic mucosa.     

Argentaffin cells, intestinal metaplasia and antral metaplasia in carcinoma of the gall bladder

The occurrence of argentaffin cells, intestinal metaplasia, and antral metaplasia has been studied in 20 gall bladders with carcinomas, in a papilloma, and in 20 specimens of cholelithiasis. Argentaffin cells were present in six carcinomas, but in only one specimen were they present in large numbers. Only one adenocarcinoma contained occasional argyrophil cells and no argentaffin or argyrophil cells were seen in the papilloma. Five of the 20 specimens of cholelithiasis contained occasional argentaffin cells. Intestinal and antral metaplasia were found in four carcinomas, in the papilloma, and in seven of the 20 specimens of cholelithiasis. It is suggested that metaplastic changes may play a role in the pathogenesis of carcinoma of the gall bladder.     

Glicentin-containing cells in intestinal metaplasia,adenoma and carcinoma of the stomach

Summary Glicentin-containing cells (Glic. cells) in intestinal metaplasia, adenoma and carcinoma of the stomach were examined using immunohistochemical techniques. Glic. cells first occurred in the gastric mucosa of the transitional area between metaplastic and intact gastric glands. They frequently showed hyperplasia or micronoduli in the budding area of the deeper metaplastic glands, but in completely intestinalized mucosa these endocrine cells decreased remarkably. Gastric adenomas with mild dysplasia had a good number of glicentin-immunoreactive cells which were located in the deeper adenoma glands. Gastrin- and somatostatin-positive cells were also detected in the adenomas. The incidence of glicentin-positive tumor cells was significantly higher in well differentiated adenocarcinoma than in poorly differentiated adenocarcinoma. Among the seven cases of scirrhous argyrophil cell carcinoma, three showed glicentin- and glucagon-immunoreactivity in the same area of the tumor. These findings suggest that the selective increase of Glic. cells in intestinal metaplasia may be closely related to the development of gastric adenoma. Glicentin positive tumor cells in gastric carcinomas can be regarded to be an expression of intestinal or fetal markers.     

Absence of BAT-26 instability in gastric intestinal metaplasia

BAT-26 instability, a sensitive marker for the high-frequency microsatellite instability (MSI-H) phenotype, was analyzed in samples of gastric cancer and in adjacent intestinal metaplastic mucosae. Although all MSI-H gastric cancer samples showed BAT-26 instability, as assessed using 12 dinucleotide microsatellite markers, BAT-26 instability was not found in the adjacent intestinal metaplastic mucosa in any of the samples.