Biological differences between reflux stimulated proliferative stomal lesions and N-methyl-N'-nitro-N-nitrosoguanidine induced carcinomas in Wistar rats.

The morphology and evolution of epithelial lesions that developed at a gastrojejunal stoma due to reflux of duodenal contents were compared with MNNG-induced carcinomas in the pyloric mucosa of rats in a long term experiment. Random bred male Wistar rats were given MNNG in drinking water (100 mg/l) for 12 weeks and then one group was submitted to a gastrojejunal anastomosis at the greater curvature in the oxyntic mucosa. Untreated rats underwent either gastrojejunostomy or gastrotomy. The animals were killed at the 24th and 66th weeks of the experiment. The lesions obtained in the pyloric mucosa and in the mucosa of the gastrojejunal stoma were analyzed histologically using hematoxylin and eosin staining and immunohistochemistry for pepsinogen isoenzyme 1. Duodenal reflux induced proliferative lesions at the gastrojejunal junction that increased in incidence and size with time. Histologically they consisted of benign epithelial proliferation of gastric type. No evidence of malignant transformation within the gastric components of the proliferative lesions at the gastrojejunal stoma was observed even at the 66th week. Adenocarcinomas induced by MNNG in the pyloric mucosa increased in size during the experiment and were morphologically and histochemically distinct from the proliferative lesions at the gastrojejunal junction. In conclusion, proliferative lesions at the gastrojejunal stoma stimulated by duodenal reflux are biologically distinct from adenocarcinomas induced by MNNG in the pyloric mucosa. They do not seem to be precursor lesions of gastric carcinogenesis, as they do not undergo malignant transformation even after long-term, up to 66 weeks, follow-up.

Sialosyl-Tn antigen as a marker of gastric cancer progression: an international study.

Sialosyl-Tn, a mucin-associated carbohydrate antigen, is not expressed by normal mucus-producing cells of the stomach but becomes expressed in metaplastic, pre-malignant and malignant gastric tissues. Reports vary as to the frequency of STn expression and its prognostic role in gastric cancer. To determine whether these differences might be due to inter-country variations in gastric cancer biology, we immunohistochemically analyzed 340 gastric cancers from 2 countries at high-risk (high incidence) for gastric cancer (Japan and Chile), one with intermediate risk (Brazil) and one with low-risk (USA). Expression of STn was correlated with clinico-pathological features of the tumors and with cancer-related survival. Regardless of country, the frequency of STn-positive tumors was lower in non-invasive ("early") than in advanced gastric cancer. Consequently, high-risk countries where early gastric cancer is more common demonstrated a lower overall frequency of STn-positive tumors. In all 4 countries, STn expression directly correlated with depth of invasion, stage, and lymph node involvement. In addition, STn expression correlated with a poor prognosis in all 4 countries, but the effect of STn on survival was not independent of tumor stage. Our findings indicate the need to consider the inherent gastric cancer risk and prevalence of early gastric cancer in the study population when reporting frequency of STn expression in gastric cancer. Regardless of country, however, STn expression is a marker of gastric cancer progression suggesting that cancer-associated mucins play a role in the malignant behavior of this tumor.

Gastric and small intestinal lesions after partial stomach resection with Billroth II or Roux-en-Y reconstruction in the rat.

The evolution and phenotypic expression of mucosal lesions of the gastric stump were investigated in male rats submitted to gastric resection with reconstruction by the Billroth II technique (BII with biliopancreatic reflux, BPR) or by the Roux-en-Y procedure (without BPR). Animals were studied at 24, 36, 54 and 64 weeks after surgery and the phenotypic expression of lesions analysed using routine hematoxylin and eosin staining, immunohistochemical staining for pepsinogen isoenzyme 1 and histochemical procedures for mucins (paradoxical concanavalin A, galactose oxidase Schiff (GOS) and sialidase GOS reactions). BPR was found to be responsible for the formation of adenomatous hyperplasia (AH), increasing in incidence and size with time, since the Roux-en-Y procedure failed to induce the gastric stump lesions observed after BII reconstruction. AHs always occurred in the transition of the gastrojejunal junction, a site offering special conditions for BPR influence, and were classified as gastric (G), intestinal (I) and G+I types according to their phenotypic expression. No pure I type AH was diagnosed at any time point. The G and G+I types developed at approximately equal incidences (i.e., G type 7/17, G+I type 10/17 at the 64th week). It was suggested that both gastric and intestinal mucosal elements were stimulated to proliferate by BPR, with the gastric mucosa tending to demonstrate AH. Intestinal type components of AH were found adjacent to the jejunum and not at the stomach margin, indicating an origin from intestinal mucosa. No metaplasia of the gastric mucosa was observed in any animal after partial gastric resection. In 101 rats submitted to the BII procedure, 5 mucinous adenocarcinomas were eventually diagnosed, mostly located in the subserosa of the gastrojejunal junction. All carcinomas expressed the phenotype of cells of the small intestine. Evidence of malignant transformation within the gastric components of AH was not observed even at the 64th week. In conclusion, all lesions induced by BPR in the rat remnant stomach are benign, and the few true cancers that arise in association are derived from the small intestine.

Reversibility of adenomatous hyperplasia in the gastric stump after diversion of bile reflux in rats.

Morphological and phenotypical patterns of proliferative epithelial lesions induced in the gastric stump mucosa by duodenal content reflux after Billroth II partial gastrectomy (BII) were evaluated in rats. Control animals were either sham-operated or submitted at different times after BII to Roux-en-Y (RY) surgical procedure which prevents duodenal reflux. The lesions were analysed using routine haematoxylin and eosin staining, immunohistochemical staining for pepsinogen isoenzyme 1 and histochemical procedures for mucins (paradoxical Concanavalin A, galactose oxidase Schiff and sialidase galactose oxidase Schiff reactions). Mucosal hyperplasia (H) was observed in the group submitted to BII procedure 6 weeks after surgery. Adenomatous hyperplasia (AH) also appeared 6 weeks after induction of the reflux and its incidence and size increased until the 54th week of the experiment. RY procedure performed in the normal animals at the beginning of the experiment or at the 24th week after BII gastrectomy led to a significantly lower incidence of AH which was related to the moment of surgery. Most of H was due to pyloric mucosal hyperplasia. AH consisted mainly of gastric type glands but in some animals glands of the intestinal type were present probably originating from the intestinal mucosa. Six mucinous adenocarcinomas were observed, all of them of intestinal type. This study demonstrates that AH induced by BII procedure is a reversible lesion and that the anomalous epithelial proliferation in the stoma may lead to adenocarcinomas.

Placental form of glutathione S-transferase in normal and diseased human uterine cervical mucosa.

Expression of the human placental form of glutathione S-transferase (GST-pi), an enzyme proposed as a marker for human and experimental neoplasia, was immunohistochemically evaluated in 51 samples of 'normal' and diseased adult human uterine cervix. Five fetal uteri were also studied. GST-pi positivity was detected in 54, 92, 95 and 83% of the 'normal', non-neoplastic, cervical intra-epithelial neoplasia (CIN) and cancer cases respectively. All five fetal uteri and the positive 'normal' adult cases presented cells immunostained for GST-pi throughout the thickness of the mucosa, including the basal layer. Some non-neoplastic conditions like inflammation, repair and metaplasia and some dysplastic and neoplastic lesions showed areas of positively stained cells within an otherwise negative tissue, indicating a phenotypic heterogeneity regarding the enzyme expression. Our results confirm that GST-pi has a fetal character and indicate that it may appear in the adult cervical squamous epithelia under 'normal' or pathological conditions not necessarily linked to the process of carcinogenesis. Therefore it cannot be used as a marker for cervical epithelial neoplasia.