Incomplete Intestinal Metaplasia as an Indicator for Early Detection of Gastric Carcinoma in The Events of Helicobacter Pylori Positive Chronic Atrophic Gastritis

Th e aim of the study was to ascertain the existence of intestinal metaplasia in gastric mucosa of patients with gastric carcinoma coupled with H. pylori positive chronic atrophic gastritis and possible connection of IM with the development of gastric carcinoma. Th e paper presents prospective study that included  patients with gastric carcinoma and  patients with chronic atrophic H. pylori positive gastritis. All the patients were subjected to gastroscopy as well as biopsy targeted at antrum, lesser curvature and corpus and at the area - cm removed from tumor lesion. Biopsy samples were sliced by microtome and stained. We analyzed presence, frequency and severity of infl ammatory-regenerative, metaplastic and dysplastic changes in the mucosa and evaluated their prognostic value. We typed IM immunohistochemically. Th is study confi rmed responsibility of H. pylori for infl ammatory events in gastric mucosa in patients with gastric carcinoma. According to our fi ndings incomplete IM of types IIa and IIb as precancerous lesion is responsible for the development of gastric carcinoma and is associated with chronic atrophic gastritis grade I and II ( of subjects, p=,, h=, p=,). Th us, the fi nding of incomplete intestinal metaplasia may be used as an indicator for early gastric carcinoma detection. Patients with patho-histologicaly verifi ed incomplete intestinal metaplasia associated with active chronic atrophic gastritis of levels I and II represent risk group for the development of gastric carcinoma of intestinal type.


Introduction
H. pylori induced chronic atrophic gastritis (CAG) is a precancerous condition accompanied by gradual loss in epithelia diff erentiation that eventually leads towards carcinoma.Intestinal metaplasia (IM) as precancerous lesion is either partial or complete transformation of gastric glands epithelial cells into the intestinal type.It may appear either in antrum or pylorus mucosa or diff usely.Various classifi cations are based on morphological, histochemical, enzymehistochemical, immunohistochemical and ultrastructural characteristics (,).Complete type of mature IM is characterized by histology, mucine histochemistry, enzymes and endocrine cells identical to those found in small intestine.Pathohistological features of this type are columnar absorptive cells, multiplied goblet cells that contain predominantly acid mucines, Paneth's non-diff erentiated cells and Moendocrine cells.Incomplete type of mature IM retains curved structure in the antrum glands.Goblet cells are less frequent and contain acid mucines while columnar cells contain traces of neutral mucines.Paneth's and entero-chromophilic are either absent or extremely rare.Immature IM type is encountered in severe atrophic forms of gastritis due to rapid desquamation and regeneration as well as delayed cell diff erentiation.Hyperplastic glands laid with juvenile non-diff erentiated cells with hyper-chromic nuclei and increased mitotic index develop in the process.Th e non-diff erentiated cells may diff erentiate into various cell types in the presence of various promoters.IM is believed to develop due to the absence of antioxidative activities in the mucosa (,,).An issue is raised of whether Helicobacter pylori (H.pylori) presence is necessary for the development of IM.It is known that H. pylori does not dwell in the area of advanced IM as it is not found in biopsic samples collected from antrum where the atrophy is pronounced.H. pylori is found in the mucosa in B gastritis and IM.Positive correlation between H. pylori infection and IM type II is noted in patients with gastric carcinoma.

Subjects and Methods
Th is paper presents prospective study, which includes three viewpoints at the possibility of gastric adenocarcinoma development: clinical, pathohistological and microbiological.Th e study included  patients with gastric carcinoma that are treated at the Gastroenterohepatology Clinic.All the patients were subjected to endoscopic examination where stomach was macroscopically divided into sections: corpus, antrum, antropyloric region and pylorus.Th e patients with cardiac and distal esophagus carcinoma were not included in the study.Following the macroscopic verifi cation of the tumor we performed biopsy of the tumoral formation, as well as biopsy of the mucosa in the region - cm removed from the tumor.We also collected mucosa samples from corpus, antrum and pyloric region.Biopsy mucosa samples were conserved in   buff ered neutral formalin, paraffi n embedded and sliced with microtome into  μm thick sections.Th e sections were stained with Heamatoxylin and Eosin (HE staining), immunohistochemically for p antibodies.IM typing was performed.All the patients were examined for the presence of infl ammatory-regenerative changes in gastric mucosa.Based on the defi ned histological criteria we graded inflammatory-regenerative changes as chronic superfi cial or chronic atrophic gastritis of the degree I, II and III.Th e degree of infl ammation activity was determined based on leukocyte infi ltration.It was defi ned either as active or dormant phase.When the changes were graded as infl ammatory-regenerative or dysplastic the tissue was prepared for immunohistochemical analysis.The tissue was HE stained prior to the processing.IM was assigned to one of the three possible categories based on the detailed histochemical analysis of mucine type, PAS and HID/AB staining and morphological changes.Th e degree of IM dispersal was also analyzed.IM was considered focal when found in / of the biopsic sample, medium when found in / of the sample and severe when dispersed over the entire sample.Th e mucine type was examined within goblet and absorptive cells.Mucines may be either neutral or acid.HID/AB staining enables diff erentiation of acid mucines into sulfo-or sialomucin (sulfomucin -brown, sialomucin -blue).In PAS method acid mucines are stained blue while neutral mucines are stained red.We registered brush border on the surface of absorptive cells that was either well developed, underdeveloped or completely missing.Paneth's cells were also identifi ed and graded as numerous, reduced or missing.Villi development was also graded as: well developed, underdeveloped and undeveloped.In IM type I, mucosa architecture is mainly well preserved as well as Paneth's cells abundance.Goblet cells mostly secrete sialomucines, some sulfomucines, while absorptive cells have no secretionary properties.Brush border at their surface is preserved.Mucosa surface is not villous although it may appear slightly villous in small number of cases.In IM type II mucosa architecture is signifi cantly degraded, villi are well developed, glandular crypts are elongated and curved and coated in tall columnar cells with either underdeveloped or no brush border.Paneth's cells are ei-

Results
The study included  subjects divided into two groups.The first group consisted of  patients with stomach carcinoma,  male and  female.Average age of patients was  for males and  for females.Anatomically,   of carcinoma formations were endoscopically identifi ed in the distant part of the stomach.Intestinal metaplasia was diff erentiated according to regions.
IM in antral region was identifi ed in  subjects ().,  of the subjects were suffering from superficial chronic gastritis in active phase.,  of the subjects had intestinal metaplasia associated with chronic atrophic gastritis grade I. ,  of those were in active phase.  had intestinal metaplasia associated with chronic atrophic gastritis grade II in active phase.IM was identifi ed in  patients with epithelial dysplasia.IM was associated with slight epithelial dysplasia in  cases and with mild epithelial dysplasia in  cases. women,  and  years average age respectively.
In the control group  subjects had IM in the region of antrum ( ).In ,  subjects IM was associated with CAG grade I in active phase.In  subjects it was associated with epithelial dysplasia.Correlation between IM and epithelial dysplasia was signifi cantly high at p=, signifi cance threshold.Intestinal metaplasia in the region of lesser curvature was found in  subjects ().Of those,   had CAG grade I. Active phase was ascertained in ,  subjects.Rank correlation was significant between the degree of epithelial dysplasia and IM presence at p=, signifi cance threshold.IM in the corpus region was found in one patient along with CAG grade I in active phase.Correlation between IM and epithelial dysplasia was not signifi cant.

Discussion
Progression of chronic active gastritis is accompanied by changes in phenotype of gastric mucosa epithelial cells.
Under the infl uence of the modifi ed environment ephithelial metaplasia occurs due to gastric mucosa capacity for regenerative and metaplastic changes.H. pylori is believed to promote hyper-proliferation of epithelial cells by its presence and action upon mucosa (,,).Proliferative properties of germinative cells ensure regeneration of surface layer cells in normal mucosa.Th ese cells, referred to as base or stem cells, are located in lower segments of epithelial glands, which explains absence of mitotic activity in surface layer of mucosa ().Known properties of these cells include auto-regulation, which accounts for adaptive reaction to tissue damage.Th ese cells divide slowly; they are extremely sensitive to radiation and are programmed to self-destroy when their DNA pattern is damaged.Th ey have the ability to preserve genome by selectively maintaining original, paternal DNA copy (,,).Other cell types diff erentiate from those cells since they are multipotent and located in neck parts of glands in the lower crypts segments.Compensation of the lost cells in crypts requires time for regeneration.Cells migrate from the neck part upwards towards the surface.Glandular atrophy occurs in the scope of prolonged chronic atrophic gastritis as infl ammation spreads and aff ects deeper mucosa layers.In atrophic gastritis, gastric mucosa is continuously des-quamated (peeled off ) and regenerated (renewed).Th is process is accompanied by compromised diff erentiation, atypical regeneration, which may lead to carcinoma.Th e measure of gastritis activity is presence of polymorphonucleous infi ltrate, which correlates with H. pylori fi nding.It is widely believed that H. pylori is an important factor in gastritis progression.Along with the progression of atrophy, gastric mucosa suff ers loss of gastric antral and corpus glands.In the process, either partial or complete transformation of gastric glands epithelial elements into intestinal type occurs.In the analysis of the mucosa we registered absorptive enterocytes that alternate with sialomucines secreting goblet cells.Such type of epithelial metaplasia is referred to as complete or intestinal metaplasia type I (,).Further progression is marked with glands being coated with columnar cells, scarce presence of Paneth's cells and sialomucines presence in goblet cells.Absorptive cells are less diff erentiated and secrete sulfomucine and neutral mucine.Th is type is denoted IM type II.In IM type III mucosa surface is extremely villous, goblet cells are scarce and secrete acid, sulfo-and sialomucines, while less diff erentiated absorptive cells secrete sulfomucine.Th us, changes in gastric cells phenotype accompany infl ammation and atrophy of gastric mucosa.Th e normal gastric epithelium is replaced by cells with morphological properties resembling those of the intestinal type.IM progression generates dysplasia which may progress into gastric carcinoma ().Gastric mucosa atrophy introduces marked increase in cell replication along with the expression of fetal antigens, which infl uence loss of cell diff erentiation.
Our research targeted verifi cation of the presence, severeness and frequency of intestinal metaplasia occurrence in various regions of gastric mucosa.We found intestinal metaplasia in antral region in  out of  patients () with gastric carcinoma.In  subjects it was associated with chronic atrophic gastritis grade I, which was in active phase in  subjects.In  subjects we found chronic atrophic gastritis grade II in active phase.We found  cases of IM in the region around carcinoma (). patients had CAG grade I while  had CAG grade II, all of those in active phase.In corpus region we found  IM ().In  cases IM was superfi cial,  had CAG grade I and one had CAG grade II.Four of those were in active phase.In the control group, for compari-

Conclusion
Based on the results of our research we may conclude that IM as precancerous lesion is responsible for the development of gastric carcinoma.When associated with chronic atrophic gastritis grade I and II it can serve as an indicator of the events that lead to intestinal gastric carcinoma development.Special attention is required in patients with pathohistologically verifi ed incomplete IM type IIa associated with slight or mild epithelial dysplasia due to its signifi cantly frequent occurrence in our patients with gastric carcinoma.(,) suggest that metaplastic cell lines ensue as a response of gastric mucosa to chronic damage and precede epithelial dysplasia and gastric adenocarcinoma.In our subjects, we found IM type I in  patients with gastric carcinoma and in  patients with H. pylori positive gastritis. patients had incomplete IM or type II while  had IM type III.In the control group though,  subjects had incomplete IM type II.In our research IM was identifi ed in  () patients with gastric carcinoma.Incomplete IM type IIa was more frequent in patients with gastric carcinoma.According to the severeness of the changes incomplete IM types IIa and IIb were either diff use or spread over / of the biopsic specimen width.In comparison with the control group no signifi cant diff erence in IM frequency in antral region was found.In the region around carcinoma IM was found in  () of the subjects and it was associated with chronic atrophic gastritis grade I and II in active phase.
ZORA VUKOBRATBIJEDIĆ ET AL.: INCOMPLETE INTESTINAL METAPLASIA AS AN INDICATOR FOR EARLY DETECTION OF GASTRIC CARCINOMA IN THE EVENTS OF HELICOBACTER PYLORI POSITIVE CHRONIC ATROPHIC GASTRITIS ther rare or missing.Goblet cells predominantly secrete sialomucines while poorly diff erentiated absorptive cells secrete either sulfomucines or neutral mucines.Absorptive cells are scarce.In IM type III mucosa architecture is completely degraded while the surface is exceptionally villous.Mucosa surface glands are elongated and curved to a higher degree compared to those in the IM type II.Gland branching and distortion is evident.Goblet cells are scarce and secrete mainly acid mucines, sulfo-and sialomucines, while poorly differentiated absorptive cells contain sulfomucines.Brush border is absent from absorptive cells.Paneth's cells are missing.Th e control group included  patients suff ering from chronic H. pylori positive gastritis.We endoscopically aimed biopsy at antrum, lesser curvature and corpus mucosa.We applied the same methodology as in the cases of gastric carcinoma in conserving, embedding and staining of the samples as well as in the analysis of infl ammatoryregenerative and dysplastic changes in gastric mucosa.We used immunohistochemical methods in IM typing.
HUT testing was used for verifi cation of H. pylori presence in gastric mucosa in the areas of antrum and corpus.
Correlation is significant and high, p=,, d=.Rank correlation(Wilcox  test)between the level of dysplasia and IM in antral region was signifi cant and high (signifi cance level p=,).IM in the region around carcinoma was confirmed in  subjects ().In   of cases IM was associated with chronic atrophic gastritis grade I.All of those were in active phase.In  cases IM was associated with chronic atrophic gastritis grade II in active phase.With respect to epithelial dysplasia, it was slight in  IM patients and mild in  cases.In  patients epithelial dysplasia was not found.Rank correlation (Wilcox test) between IM and the degree of epithelial dysplasia was not significant (p=,).IM in the region of corpus was found in  patients ().In  patients IM was associated with superficial and in  patients with CAG grade I and grade II.In ,  patients gastritis was in active phase.In the corpus region correlation between IM and gastritis grade was significantly high at p=, level of significance.Rank correlation (Wilcox test) between IM and the degree of epithelial dysplasia was high and signifi cant (p=,).Control group included  men and ZORA VUKOBRATBIJEDIĆ ET AL.: INCOMPLETE INTESTINAL METAPLASIA AS AN INDICATOR FOR EARLY DETECTION OF GASTRIC CARCINOMA IN THE EVENTS OF HELICOBACTER PYLORI POSITIVE CHRONIC ATROPHIC GASTRITIS ZORA VUKOBRATBIJEDIĆ ET AL.: INCOMPLETE INTESTINAL METAPLASIA AS AN INDICATOR FOR EARLY DETECTION OF GASTRIC CARCINOMA IN THE EVENTS OF HELICOBACTER PYLORI POSITIVE CHRONIC ATROPHIC GASTRITIS son, IM in antral region was found in  out of  examined patients ().Of those  () had CAG grade I,  () had CAG grade II and one had superfi cial gastritis.All of those were in active phase.()cases of IM in lesser curvature region were identifi ed.Five of those were associated with CAG grade I, one with CAG grade II and two with superfi cial gastritis.We found one case of IM in corpus region associated with CAG grade I in active phase.Comparing with the group of patients with gastric carcinoma, no signifi cant diff erence in distribution of intestinal metaplasia in antral region was found in our research.There is significant difference in frequency (p=,, h= for signifi cance threshold p=,).Signifi cant diff erence exists in frequency of IM in corpus region.IM is more frequent in patients with gastric carcinoma.Th e diff erence is signifi cant with p=, at signifi cance threshold p=,.We used immunohistochemical staining to type IM. () subjects had complete IM type I.  () of them had incomplete IM type IIa.Four () of patients had incomplete metaplasia type IIb.According to the above, IM was found in  out of  patients with gastric carcinoma.In  () patients it was associated with CAG grade I, of those,  were in active phase.In  () patients we found CAG grade II in active phase.In ()complete IM was focal.Incomplete IM type IIa was of focal appear-Th ere is signifi cant diff erence between groups in IM occurrence in antral region with p=,, h= at the signifi cance threshold p=,.No signifi cant diff erence in gastritis presence in corpus region was found.Analyzing earlier published studies we found that Reis () studied IM and concluded that classic sequential paths from IM I to IM III lead through IM II.In our research we found incomplete IM type IIb (type III according to certain authors) in only  () patients with gastric carcinoma.No IM type III cases were found in the control group.Safele-Ribiero stated IM type I as the most frequent in their studies.It is suggested