ENDOSCOPIC PHENOL RED TEST

The distribution of H pylori in gastric mucosa was examined by the endoscopic phenol red test, developed by Kohli et al, details of which have been reported previously.16-18 Briefly, each patient was given 20 mg of omeprazole (Omepral, Fujisawa Pharmaceutical Company Ltd, Osaka, Japan) orally at bedtime the day before the test or 20 mg of famotidine (Gaster, Yamanouchi Pharmaceutical Company Ltd, Tokyo, Japan) intravenously 30 minutes before the test to reduce gastric acid secretion. Five minutes before endoscopy, 80 mg of dimethylpolysiloxane (Gascon, Kissei Pharmaceutical Company Ltd, Matsumoto, Japan) was given orally to remove adherent gastric mucus and 20 mg of scopolamine butylbromide (Buscopan, Behringer Ingelheim Company Ltd, Kawasaki, Japan) was injected intramuscularly to reduce gastric motility. A videoendoscope was then inserted, gastric juice was aspirated to improve visibility during endoscopy, and the interior of the stomach was inspected. A spray tube was inserted through the biopsy forceps channel, and 0.1% phenol red solution with 5% urea was sprayed over the entire surface of the gastric mucosa. Changes in colour occurred within two to three minutes after the dye was sprayed and continued for at least 15 minutes. Areas where phenol red had turned from yellow to red owing to H pylori infection were designated “coloured” areas, whereas areas where the colour had not changed were designated “non-coloured” areas.

The extent of H pylori infected areas was classified on the basis of the extent of coloured areas as follows: (1) none: no areas in the stomach infected with H pylori; (2) localised: H pylori infected areas detected in parts of the stomach; and (3) diffuse: H pylori infected areas detected from the antrum to the cardia.

ASSESSMENT OF H PYLORIINFECTION

During the endoscopic phenol red test, three biopsy specimens were taken with a sterilised biopsy forceps (FB-25U, Olympus Optical Company Ltd, Tokyo, Japan) from coloured and non-coloured areas in the stomachs of all patients. After each examination, endoscopes were cleaned with detergent, disinfected with 10% ethanol, and rinsed with sterile water.

Two biopsy specimens taken from each area were homogenised in a microhomogeniser (Radnoti Glass Technology Company Ltd, Monrovia, CA, USA) with 100 μl sterile saline and inoculated onto the surface of Brucella agar plates with 7% defibrinated horse blood and helicobacter selective supplement (Skirrow, Oxoid Company Ltd, Basingstoke, UK). Inoculation was performed within two to three hours of sampling. Agar plates were incubated for four days under 10% CO₂ in air. Specimens were considered positive for H pylori if one or more colonies of curved or spiral shaped, Gram negative bacteria were found which were positive for oxidase, catalase, and urease. In such cases, one colony was cultured and identified by these characteristics: no hydrolysis of hippurate; no growth in medium with 1% glycine, 1.5% sodium chloride, and 8% glucose; sensitivity to cephlethin (30 μg/disk); resistance to nalidixic acid (30 μg/disk); and growth in medium with 0.04% tetrazolium chloride. The bacteriologists who carried out the examinations were unaware of the endoscopic findings or grouping of specimens.

The third specimen, which was used for histological examination, was fixed in 10% formalin and processed in a routine manner. Sections were cut at 5 to 7 μm and stained with Giemsa to show H pylori organisms. Patients were considered not to have H pylori infection when the culture was negative, when no H pylori was seen on histological examination, and when no areas coloured with phenol red were detected on chromoendoscopy.

HISTOLOGICAL TYPE OF EARLY GASTRIC CANCERS

The depths of involvement and the histological types of gastric cancer were determined histologically in specimens obtained at gastrectomy. Gastrectomy was performed two to 24 days (average duration: 12.5 days) after the phenol red test. Early gastric cancers were defined as those confined to the mucosa or submucosa with no involvement of the muscularis propria.20

According to the criteria of the Japanese Research Society for Gastric Cancer, early gastric cancers were histologically classified as follows: (1) well differentiated adenocarcinoma, (2) moderately differentiated adenocarcinoma, (3) poorly differentiated adenocarcinoma, and (4) signet ring cell carcinoma.19 The histological classification was based on the predominant pattern of the tumours. Well differentiated and moderately differentiated adenocarcinomas were defined as differentiated adenocarcinomas, and poorly differentiated adenocarcinomas and signet ring cell carcinomas as undifferentiated adenocarcinomas. In this series, mucinous adenocarcinomas and special type of cancers (for example, adenosquamous carcinoma and squamous cell carcinoma) were not found.

HISTOLOGICAL OBSERVATION OF SPECIMENS BY GASTRECTOMY

The location of early gastric cancers in relation to the extent of intestinal metaplasia and gastric mucosa infected with H pylori was examined histologically in 25 specimens obtained at gastrectomy. All patients underwent the phenol red test before surgery.

After the operation, the stomach was opened, pinned flat on a cork mat, and fixed with 10% formalin solution. The fixed stomach was cut into longitudinal strips of 5 mm in width. The specimens were embedded in paraffin wax, and serial sections 5 μm thick were stained with haematoxylin and eosin and immunohistochemically examined for the presence of H pylori with anti-H pylorirabbit serum (Dako, Glostrup, Denmark).21

STATISTICAL ANALYSIS

Data were analysed by the χ² test or Fisher’s exact probability test.22 Significant indicates a calculated p value of less than 0.05.

H PYLORI IN BIOPSY SPECIMENS FROM COLOURED AND NON-COLOURED AREAS USING THE PHENOL RED TEST

Using the endoscopic phenol red test, H pylori was detected in 60 (95%) of 63 specimens obtained from coloured areas, but in only 5 (8%) of 65 specimens obtained from non-coloured areas (table1).

H PYLORI INFECTION IN PATIENTS WITH EARLY GASTRIC CANCERS AND ADENOMAS

With chromoendoscopy, H pylori infection was significantly more frequent in patients with differentiated adenocarcinomas than in those with undifferentiated adenocarcinomas (table 2). H pylori infection was also significantly more frequent in patients with well differentiated and moderately differentiated adenocarcinomas than in those with poorly differentiated adenocarcinomas and signet ring cell carcinomas. H pyloriinfection was also significantly more frequent in patients with gastric adenomas than in those with undifferentiated tumours. Differentiated adenocarcinomas were usually in areas of mucosa infected with H pylori, but undifferentiated adenocarcinomas were frequently in uninfected areas (table 3).

COLOURATION OF PHENOL RED IN RELATION TO INTESTINAL METAPLASIA

Intestinal metaplasia was significantly more frequent in differentiated adenocarcinomas than in undifferentiated adenocarcinomas (table 4). Five differentiated adenocarcinomas (38%) were completely surrounded by intestinal metaplasia. Endoscopically these tumours were all located in the non-coloured areas far from the coloured area. This finding indicates that intestinal metaplasia did not affect the phenol red test.