PATIENTS

The study involved six patients with enlarged fold gastritis (allH pylori positive) (four men and two women, mean age 47.5 (3.3) years, range 35–59 years), sixH pylori positive patients without enlarged folds (three men and three women, mean age 48.3 (3.9) years, range 38–59 years), and six H pylori negative patients with dyspeptic symptoms (three men and three women, mean age 45.2 (3.3) years, range 34–56 years). The patients with enlarged fold gastritis had crowded, tortuous, and enlarged folds in the gastric body shown by barium x ray performed as part of a mass screen for gastric carcinoma. Subsequently, the patients had an endoscopic examination and were conclusively shown to have inflammation and H pylori colonisation of the body mucosa. The patients without enlarged folds had an endoscopic examination to investigate dyspepsia. These patients were divided into two groups on the basis of the presence or absence ofH pylori infection. A bariumx ray photograph was taken of each patient by experienced radiologists, using the same equipment and agents. The width of the gastric fold was measured on double contrast radiographs of the appropriately distended stomach with the patient in the supine position. Gastric body folds were considered to be enlarged when the widest fold was more than 5 mm based on previous studies and the Sydney system.3-7 None of the patients had peptic ulcer disease or gastric cancer and none were receiving any regular medication and were free of H₂ receptor antagonists, proton pump inhibitors, prostaglandin analogues, and non-steroid anti-inflammatory drugs before and during the period of this study. Five minutes before endoscopy the patients received pharyngeal anaesthesia with lidocaine hydrochloride and an intramuscular injection of butylscopolamine bromide (20 mg). Written informed consent was obtained from each patient before endoscopic examination, and the investigation was approved by the research ethical committee of Osaka University Medical School. Two biopsy specimens were taken from the lesser curvature of the antrum (one for light microscopy and the other for culture studies), and three from the greater curvature of the middle corpus (one for light microscopy, one for culture, and the last for electron microscopic studies). The diagnosis of H pylori infection was made based on positive culture results (Department of Chemotherapy, Pharmacological Research Laboratory, Fujisawa Pharmaceutical Co., Osaka, Japan) and/or the result of a urease test (CLO test; Delta West, Bentley, Australia) using biopsy specimens from the antrum and the greater curvature of the upper portion of the gastric body.13 ,14

MEASUREMENT OF MUCOSAL AND FOVEOLAR LENGTH

The mucosal and foveolar length of the body mucosa were measured on sections vertical to the mucosal surface using an image analysis system (Krypton-40; Flovel). Values were expressed as the mean of measurements along three or more different foveolae. Glandular length was calculated by subtracting foveolar length from mucosal length.

HISTOLOGICAL EXAMINATION

Biopsy specimens were fixed with 10% phosphate buffered formalin. Thin sections of paraffin wax embedded tissues were stained with haematoxylin and eosin for evaluation of mononuclear and polymorphonuclear infiltration. Infiltrations of mononuclear and polymorphonuclear cells were graded as follows: 0 = none or minimal; 1 = mild; 2 = moderate; and 3 = severe.3 The assessment of gastric atrophy was scored by two pathologists using a number of selected gastric biopsy specimens on a scale of 0 to 3 on the basis of the proportion of glandular loss.4 ,15 Lymphocytic gastritis was diagnosed when the number of intraepithelial lymphocytes was 30% or more of the number of epithelial cells in the areas of high lymphocyte concentration.16

Immunohistochemical localisation of H⁺ and K⁺ stimulated adenosine triphosphatase (H⁺,K⁺-ATPase) was performed by the peroxidase-antiperoxidase method as previously described17using biopsy specimens taken from the greater curvature of the body of the stomach in each patient. A polyclonal rabbit antibody against pig gastric vesicles enriched in H⁺,K⁺-ATPase (provided by Dr M Maeda, Department of Organic Chemistry and Biochemistry, Institute of Scientific and Industrial Research, Osaka University) was used as the primary antibody. It was previously confirmed that this antibody is specific for the 94 kDa H⁺,K⁺-ATPase α subunit by western blot analysis and ELISA.18 All procedures were performed at room temperature unless otherwise stated. Biopsy samples were fixed in 10% phosphate buffered formalin and embedded in paraffin wax. Paraffin wax sections (about 5 μm thick) were deparaffinised and dehydrated by passage through xylene and graded concentrations of ethanol. The specimens were then incubated with 3% H₂O₂solution for 15 minutes to block endogenous peroxidase activity. Sections were then rinsed in distilled water for five minutes and in 0.01 M phosphate buffered saline, pH 7.4, for five minutes. After incubation with normal swine serum diluted 1:20 in phosphate buffered saline for 20 minutes at room temperature, a three step immunoperoxidase procedure was applied, in which, first, the specimens were incubated overnight at 4°C with rabbit antibody against pig gastric vesicles enriched in H⁺,K⁺-ATPase which was diluted 1:1000 in phosphate buffered saline containing 1% bovine serum albumin; secondly, the specimens were incubated for 20 minutes with swine anti-rabbit immunoglobulin (Dakopatts, Copenhagen, Denmark) which was diluted 1:200 in phosphate buffered saline containing 1% bovine serum albumin; thirdly, the specimens were incubated for 20 minutes with peroxidase labelled rabbit anti-peroxidase immunoglobulin complexes (Dakopatts). The reaction was visualised using 5’-diaminobenzidine as chromogen. In addition, the presence or absence of anti-gastric autoantibodies reactive with human gastric mucosa was examined by an immunohistochemical method described previously.19

EVALUATION OF ULTRASTRUCTURE

Biopsy specimens for electron microscopy, taken from the greater curvature of the body of the stomach in non-stimulated patients, were immediately fixed in 3% glutaraldehyde solution buffered at pH 7.4 with 0.1 M Millonig’s phosphate buffer for two hours at 4°C. Secondary fixation was carried out in 1% osmium tetroxide solution buffered at pH 7.4 with 0.1 M Millonig’s phosphate buffer for one hour at 4 °C. The specimens were then dehydrated and embedded in epoxy resin (Epon 812). Semithin sections were cut perpendicular to the luminal surface from at least five different tissue blocks for each patient and then stained with 0.5% toluidine blue dissolved in 0.1 M Millonig’s phosphate buffer. Ultrathin sections were cut on a Reichert-Jung Ultracut E Ultramicrotome, followed by double staining with 3.0% aqueous uranyl acetate and Reynolds’ lead citrate, and were then examined using an Hitachi H-7000 electron microscope. The specimens were analysed semiquantitatively for various features characteristic of abnormal parietal cell ultrastructure. Parietal cells were classified into three groups according to their abnormal ultrastructure.20 Those with dilated canaliculi containing a few short microvilli were classified as altered parietal cells with dilated canaliculi; those with two or more large membrane bound structures containing homogeneous fluffy material and a few small vesicles or membrane remnants were classified as altered parietal cells with vacuoles; those with oedematous cytoplasm and general pallor, which was increased in certain areas as the result of apparent dissolution of the cytosol, were classified as degenerated parietal cells. In such cells, tubulovesicles were partially collapsed and disorganised compared with normal parietal cells. Parietal cells with both dilated canaliculi and vacuoles were often observed and classified as altered parietal cells with vacuoles. For each patient, an average of 400 parietal cells were scored in ultrathin sections from at least five tissue blocks.

DETERMINATION OF BASAL ACID OUTPUT, MAXIMAL ACID OUTPUT, AND SERUM GASTRIN CONCENTRATION

Gastric secretions, basal and stimulated (intravenous injection of 4 μg/kg tetragastrin), were collected for 30 and 60 minutes respectively through a gastric tube with the patient in the left recumbent position, for the determination of basal acid output and maximal acid output. Fasting serum gastrin concentrations were determined by radioimmunoassay (Gastrin RIA Kit II; Dainabot, Tokyo, Japan).21

TREATMENT FOR H PYLORIINFECTION

Patients with enlarged fold gastritis (all of which wereH pylori positive) received triple agent treatment with bismuth subnitrate or bismuth subcarboxylate 1 g three times a day, 250 mg of metronidazole twice a day, and tetracycline hydrochloride 250 mg four times a day for three weeks.6 ,10 At 4–23 weeks after the completion of this treatment schedule, the endoscopic examinations, fasting serum gastrin determinations, and gastric secretion measurements were repeated. At 7–25 weeks after the completion of the treatment, the upper gastrointestinal barium x ray examination was repeated for reassessment of the size of the folds. Eradication ofH pylori was successful in all patients.

STATISTICAL ANALYSIS

Data are expressed as mean (SEM). Statistical analyses were performed using the Wilcoxon signed rank test and the Mann-Whitney test. Pearson’s correlation coefficient was also determined. p<0.05 was considered statistically significant.

H PYLORI STATUS

For all six patients with enlarged fold gastritis and in six of the 12 patients without enlarged folds, H pylori was present in both the antrum and the body except in one patient with enlarged fold gastritis who was positive forH pylori only in the gastric body.

MUCOSAL LENGTH

The gastric body fold width in patients with enlarged fold gastritis was significantly reduced from 9.7 (0.8) to 6.0 (0.4) mm after eradication (p<0.01) (table 1). The fold width in theH pylori positive patients without enlarged folds was significantly greater than that in the H pylori negative patients (p<0.01). In patients with enlarged fold gastritis, the mucosal and foveolar length of the body mucosa were significantly greater than those in both the H pylori positive and negative patients without enlarged folds (p<0.01). The foveolar/glandular length ratio in the patients with enlarged fold gastritis was significantly higher than that in theH pylori negative patients without enlarged folds (p<0.01) (table 1). No significant differences in mucosal, foveolar, and glandular length, and foveolar/glandular length ratio between the H pylori positive and negative patients without enlarged folds was observed.

INFLAMMATORY CELL INFILTRATION AND MUCOSAL ATROPHY

Inflammatory cell infiltration was mainly confined to the foveolar and surface epithelial layer. A histological examination showed that mononuclear and polymorphonuclear infiltrates in the body mucosa, as well as in the antral mucosa, of the patients with enlarged fold gastritis were significantly more severe than those in theH pylori negative patients without enlarged folds (p<0.01). Inflammatory cell infiltration in the body mucosa in the patients with enlarged fold gastritis was more severe than that inH pylori positive patients without enlarged folds, but this difference was not statistically significant. TheH pylori positive patients without enlarged folds showed appreciably more extensive infiltration of inflammatory cells than did the H pylori negative patients (table 2). Neither mononuclear nor polymorphonuclear cell infiltrates were observed in H pylorinegative patients. No typical features of lymphocytic gastritis were observed in any of the patients.

No significant difference in gastric body atrophy score was found among the six patients with enlarged fold gastritis, the sixH pylori positive patients without enlarged folds, or the six H pylori negative patients (0.3 (0.2), 0.3 (0.2), and 0.0 respectively). After eradication, there was no significant change in the degree of body atrophy in patients with enlarged fold gastritis.

IMMUNOSTAINING FOR H⁺,K⁺-ATPase IN PARIETAL CELLS

Immunostaining specific for the α subunit of H⁺, K⁺-ATPase was specifically localised to the parietal cells. However, different staining patterns were observed among patients with enlarged fold gastritis. In theH pylori positive patients without enlarged folds and H pylori negative patients, the immunostaining pattern for H⁺, K⁺-ATPase was uniform throughout the cytoplasm (fig 1A), whereas cytoplasmic H⁺, K⁺-ATPase staining in parietal cells was less uniform in patients with enlarged fold gastritis. In some parietal cells, positively stained cytoplasm was interrupted by several clear H⁺, K⁺-ATPase negative vacuole-like areas (fig1B). After H pylori eradication, the vacuole or vacuole-like structures and dilated canaliculi had nearly disappeared. The staining pattern of the H⁺, K⁺-ATPase became homogeneous in the cytoplasm of parietal cells and similar to that in tissues from H pylori negative patients. Also, no anti-gastric autoantibody could be detected in sera of patients with enlarged fold gastritis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Immunohistochemical detection of H⁺,K⁺-ATPase in human gastric mucosa. (A) Gastric mucosal tissue of an H pylori negative patient without enlarged folds. Parietal cell cytoplasm is specifically and uniformly stained (arrow). Original magnification × 210. (B) Gastric mucosal tissue of an H pylori positive patient with enlarged fold gastritis. Some parietal cells are uniformly stained as above. In other parietal cells, stained cytoplasm is less uniform and appears granular or contains several vacuole-like clear areas which are negatively stained for H⁺,K⁺-ATPase (arrows). Original magnification × 210.

LIGHT MICROSCOPY OF GASTRIC MUCOSA SEMITHIN SECTIONS

To examine the morphological differences in parietal cells between the patients with enlarged fold gastritis and those without enlarged folds (the H pylori negative andH pylori positive), semithin sections of Epon embedded gastric mucosal tissue stained with toluidine blue were examined by light microscopy. In the tissue from H pylori negative patients, normal parietal cells appeared as large pyramidal or spherical cells with a central nucleus, and the cytoplasm was dotted with mitochondria (fig 2A). On the other hand, many parietal cells in the H pylori positive patients with enlarged folds showed morphological changes (fig 2B). They often contained abnormally dilated canaliculi or vacuole-like structures, which appeared to be different from canaliculi. AfterH pylori eradication, the appearance of most parietal cells returned to normal. Altered parietal cell morphology was only rarely observed in H pylori positive patients without enlarged folds.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Semithin plastic sections of gastric mucosa stained with toluidine blue. (A) Gastric mucosal tissue from an H pylori negative patient without enlarged folds. Parietal cells appear large and pyramidal or spherical with central nuclei (arrow). Original magnification × 210. (B) Gastric mucosal tissue from an H pylori positive patient with enlarged fold gastritis. Extensive infiltration of inflammatory cells in the lamina propria is visible. Numerous parietal cells show abnormally dilated canaliculi (arrow) and sometimes vacuole-like structures which are different from canaliculi (arrowhead). Original magnification × 210.

ELECTRON MICROSCOPY OF PARIETAL CELLS

Electron microscopy showed that H pylori was present in the mucous gel layer but was often located close to the surface mucous cells in patients with enlarged fold gastritis (fig 3). Parietal cell ultrastructure was examined in all patients. Parietal cells showed no signs of poor tissue preservation such as swelling of mitochondria and dilatation of nuclear membrane. Most parietal cells in the gastric mucosal tissue of theH pylori negative patients were in the typical resting state, with small canaliculi, short microvilli, and numerous tubulovesicles (data not shown). In contrast, numerous parietal cells in H pylori positive patients with enlarged fold gastritis showed atypical ultrastructures. These parietal cells often exhibited abnormally dilated canaliculi containing few short microvilli (fig 4A), and sometimes spherical light cytoplasmic areas resembling vacuole-like structures were displayed (fig 4B). Some parietal cells appeared pale, probably because of early necrosis. Macrophage infiltration of the lamina propria was greater inH pylori positive patients with enlarged fold gastritis (data not shown). After H pylori eradication in these patients, normal parietal cells with a resting morphology predominated (fig 4C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Low magnification electron micrograph of the gastric mucosal lumen in an H pylori positive patient with enlarged fold gastritis. N, nucleus. Numerous bacteria (arrowheads) are observed in the mucous layer and close to gastric epithelial cells. There is considerable superficial vacuolation (*), evidence of loss of microvilli, and depletion of mucin granules. The organisms are also observed in the foveolar and intercellular spaces (arrows). Original magnification × 2000 (bar = 5 μm).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Electron micrographs of parietal cells in an H pylori positive patient with enlarged fold gastritis before (A, B) and after (C) H pylori eradication. N, nucleus; C, intracellular canaliculus. (A) This parietal cell is classified as an altered parietal cell with dilated canaliculi. It has dilated canaliculi containing few short microvilli (bar = 3 μm). (B) This parietal cell is classified as an altered parietal cell with vacuoles. Three large spherical electron lucent structures interpreted to be a vacuole-like structure (VL) are present in the cytoplasm, and microvilli at the luminal surface have completely disappeared (bar = 3 μm). (C) After eradication of H pylori in patients with enlarged fold gastritis, the parietal cell has almost reverted to a normal appearance as seen in H pylori negative patients. Most of the cytoplasm is occupied by tubulovesicles and mitochondria, and canaliculi with developed microvilli can be recognised (arrows) (bar = 3μm). Original magnification: A, × 4000; B, × 4000; C, × 4000.

Parietal cells were scored for abnormal ultrastructure as normal, altered, or degenerated on the basis of the criteria described in Materials and methods; about 400 parietal cells per patient were scored from the electron micrographs. In the H pylori negative patients, 3.5 (0.2)% of parietal cells were altered, consistent with previously reported data in mice.22 Altered or degenerated parietal cells were rare at the isthmus and neck, and 6.7 (0.4)% of parietal cells in the base of the gastric gland were altered or degenerated.

It is noteworthy that, in the H pyloripositive patients with enlarged fold gastritis, 56.6 (1.2)% of the parietal cells were altered. Of these altered cells, 31.5 (4.2)% contained dilated canaliculi, 19.5 (3.7)% contained vacuole-like structures, and 5.6 (1.4)% were degenerated. Altered and degenerated parietal cells were consistently encountered at the isthmus, neck, and base of the gastric gland. Thus the percentage of altered and degenerated parietal cells was greatly increased inH pylori positive patients with enlarged fold gastritis compared with that in both H pylori positive patients without enlarged folds and H pylori negative patients. AfterH pylori eradication inH pylori positive patients with enlarged fold gastritis, parietal cell ultrastructure was generally similar to that in H pylori negative patients (fig 4C). Consequently, the percentage of normal parietal cells significantly increased to 90.2 (2.3), and the percentage of parietal cells with dilated canaliculi, vacuole-like structures, and degeneration significantly decreased to 5.8 (1.2), 3.8 (0.6) and 1.6 (0.9) respectively (fig 5).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Semiquantitative analysis of parietal cell morphology before and after eradication of H pylori in six patients with enlarged fold gastritis and 12 patients without enlarged folds (six H pylori positive and six H pylori negative). Electron micrographs of gastric mucosa were scored according to parietal cell morphology as being normal, altered with dilated canaliculi, altered with vacuoles, or degenerated. The mean of the percentage of each class is represented by bars. Standard errors are represented by vertical lines.

BASAL ACID OUTPUT, MAXIMAL ACID OUTPUT, AND SERUM GASTRIN CONCENTRATION

Figure 6 shows basal acid output, tetragastrin stimulated maximal acid output, and serum gastrin concentrations before and after eradication of H pylori in the patients with enlarged fold gastritis (all H pyloripositive), the H pylori positive patients without enlarged folds, and the H pylorinegative patients. Basal acid output in the patients with enlarged fold gastritis before H pylori eradication was much lower than in both the H pyloripositive patients without enlarged folds and H pylori negative patients (0.5 (0.5) v3.5 (0.2) v 5.0 (0.1) mmol/h respectively; p<0.01) (fig 6A). Likewise, tetragastrin stimulated maximal acid output in patients with enlarged fold gastritis beforeH pylori eradication was much lower than in both the H pylori positive patients without enlarged folds and H pylori negative patients (2.5 (1.2) v 11.7 (1.5)v 15.5 (1.4) mmol/h respectively; p<0.01) (fig 6B). After H pylori eradication, basal acid output and tetragastrin stimulated maximal acid output significantly increased in patients with enlarged fold gastritis to 4.1 (1.5) mmol/h and to 13.8 (0.7) mmol/h (p<0.01) respectively. Serum gastrin concentration in the patients with enlarged fold gastritis was much higher than in both the H pyloripositive patients without enlarged folds and H pylori negative patients (213.5 (31.5)v 77.5 (10.9) v50.7 (7.6) pg/ml respectively; p<0.01) (fig 6C). After eradication ofH pylori in patients with enlarged fold gastritis, the serum gastrin concentration significantly decreased to 70.2 (7.5) pg/ml (p<0.01).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Basal acid output (A), 4 μg/kg tetragastrin stimulated maximal acid output (B), and fasting serum gastrin concentrations (C) before and after eradication of H pylori in six patients with enlarged fold gastritis, in six H pylori positive patients without enlarged folds, and six H pylori negative patients.