The presence of Okadaella gastrococcus in the gastric erosive ulcers

122 COX-2 INHIBITION SUPPRESSES HUMAN GASTRIC CANCER BY ALTERING IMMUNE RESPONSE AND APOPTOSIS: CLINICAL RELEVANCE OF DNA ARRAY ANALYSIS Clarke Hilbig, M.D., J. Chai, Ph.D., A. Tarnawski, M.D.*. UC, Irvine, Orange, CA and Veterens Administration, Long Beach, CA. Purpose: Using cDNA technology, we studied the effect of selective COX-2 inhibition on genes involved in tumor suppression and cellular signaling. NSAIDS, including COX-2 selective agents, inhibit the growth of numerous cancers, e.g. colon, breast, prostate, liver and lung. To date, most research has focused on apoptosis, yet the mechanisms of action, the genes involved, and the signaling pathways are not fully explained. Little is known regarding the affects of COX-2 inhibition on human gastric cancer. cDNA microarray (gene chip) technology allows potentially ge- nome-wide evaluation of differential gene expression in response to envi- ronmental stressors. Methods: Human gastric cancer AGS cell line was treated with either placebo or COX-2 selective inhibitor NS-398 (0.01mM) for 1 and 5 hours. Cell death was estimated by the TUNEL method. mRNA expression was determined using the AFFYMETRIX Human Focus gene array containing cDNA from approximately 8700 genes and analyzed via SILICON GE- NETICS’ GENE SPRING Sequence Database. Results: TUNEL staining demonstrated NS-398 (0.1mM) induced apopto- sis in approximately 24% and 71% of AGS cells at 1 and 5 hours respec- tively. NS-398 treatment significantly up-regulated mRNA of tumor sup- pression and cellular signaling genes such as: ILT3/ immunoglobulin- superfamily, IL-11/ cutaneous T-cell attracting chemokine, GAS-2/ growth arrest-specific, IL1R2/ interleukin-1 receptor, type 2, LOC51336/ mesen- chymal stem cell protein, IL-21/ interleukin-21 T-cell-derived inducible factor, CCR6/ chemokine receptor, and LIBC/ lost in inflammatory breast cancer tumor suppressor protein. Conclusions: 1) COX-2 selective inhibition triggers apoptosis in human gastric cancer in vitro. 2) Cell death of human gastric cancer may involve activation of growth arrest genes (apoptosis). 3) Cell death of human gastric cancer may also result from activation of pre-apoptotic genes such as those involved in immune-mediated extracellular mechanisms as shown here. If this is confirmed in vivo, drugs targeting COX-2 independent cell death pathways could be developed and the side-effects of current COX-2 in- hibitors avoided. Therapies targeted at immune response would theoreti- cally result in systemic cell death of malignant cells. Metastatic disease would be as treatable as localized disease. Delineation of the COX-2 independent pathways involved in cell death will have profound implica- tions in adenocarcinoma therapy. 123 THE PRESENCE OF OKADAELLA GASTROCOCCUS IN THE GASTRIC EROSIVE ULCERS Takayuki Okada, M.B.B.S.*, Graham Adkins, F.R.C.P.A., Malcolm Jones, Ph.D., Jeffery Searle, F.R.C.P.A., Rodney McDougall, B.Sc., Jenny Robson, F.R.C.P., John Bradley, F.R.C.P. Okada Medical Clinic, Brisbane, Queensland, Australia; Sullivan & Nicolaides Pathology, Brisbane, Queensland, Australia; Queensland Institute of Medical Research, Brisbane, Queensland, Australia; Queensland Medical Laboratory, Brisbane, Queensland, Australia and Flinders Medical Centre, Adelaide, SA, Australia. Purpose: We have reported the pathogenic virulence of Okadaella gas- trococcus (Og) in vitro and in vivo, and suggested the possible involvement of Og in the gastric erosive ulcers. Although the incidence of Hp-negative and NSAID-negative gastric ulcers is increasing, the pathogenesis of gas- tric ulcer and erosions still remains controversial. The aims of this study were (1) to investigate the presence of Og in the gastric erosive ulcers, by histology and transmission electron microscopy, and by culture and (2) to see if intracellular Og can be seen in the tissue culture. Methods: 30 Hp-negative patients (age: 17–72, female:male � 14:16) including two patients with haematemesis underwent upper endoscopy (EGD) and biopsies. NSAID users were not included in this study. Exam- inations were made with HE, AYTB, and WSS stains in addition to TEM, and culturing under microaerophilic and anaerobic conditions. Christens- en’s medium was used for urease test. Og isolated from a patient with gastric erosive ulcer was inoculated with HeLa cells for 24 hours. TEM was made at 4hours post-infection HeLa cells. HeLa cells were incubated with a saline as a control. Results: All patients were urease negative and free from Hp infection. All patients including two patients with haematemesis were found to have Og in the gastric erosive ulcers by histology and TEM. Og was cultured successfully from 70% of the patients. TEM on 4 hours post-infection HeLa cells identified multiple intracellular Og. Og found in vitro was identical to that of in vivo under TEM. All HeLa cells were destroyed by 24 hours post-infection. All control cells were unaffected. Conclusions: This study raises the possibility that Og might be the initial causative agent of gastric erosive ulcers. As we have reported the co- existence of Hp and Og and the presence of Og following successful eradication of Hp in various gastropathy, Hp-positive and Hp-negative patients with gastric erosions and ulcers need to be followed up carefully. The presence of Og in the gastric erosive ulcers including the patients with haematemesis should be investigated. 124 CORRELATION BETWEEN ENDOSCOPIC AND HISTOLOGIC FINDINGS IN PATIENTS WITH UPPER GASTROINTESTINAL TRACT SYMPTOMS Laticia Mendler, M.D., Parvez S. Mantry, M.D., Uma Sundaram, M.D.*. University of Rochester, Rochester, NY. Purpose: Determine prospectively if upper gastrointestinal tract symptoms correlate with endoscopy and histopathology in patients undergoing upper endoscopy (EGD) using a controlled design. Methods: 221 patients with dyspeptic symptoms (group I) and 146 patients with no dyspeptic symptoms (group II) underwent EGDs by a single operator (US) in 1 year. Upper gastrointestinal tract symptoms included heartburn, dysphagia, abdominal pain, bloating, burping, nausea or vom- iting. The second group consisted of patients with liver disease, GI bleed- ing, anemia, or positive FOBT. Any patients in group II with any of the above symptoms were transferred over to group I. The two groups had similar demographics. 103 of 367 patients used NSAIDs regularly. Results: Group I: Endoscopy: 175/ 221 patients had antral gastritis on visual inspection, 15/221 had ulcers (gastric) and 31/221 patients had a normal antrum. Histopathology: 145/221 patients had an abnormality in the antral mucosa on histopathology (gastritis, foveolar hyperplasia, intestinal meta- plasia). H.Pylori: 46/221 patients were positive on histology. Group II: Endoscopy: 95/146 patients had antral gastritis on visual inspection, 23/146 patients had ulcers (gastric) and 28/146 patients had a normal antrum. Histopathology: 98/ 146 patients had an abnormal antral mucosa H.Pylori: 23/146 patients were positive on histology. Amongst NSAID users, the incidence of endoscopic and histological gastritis was 84% and 77% as compared to 82% and 70% respectively in non NSAID users. Analysis: 86% of patients presenting with dyspepsia had antral gastritis or ulcers endoscopically, 66% of all patients had histological abnormali- ties.31% of all patients with histologic gastritis were positive for H. Pylori. 80% of patients presenting who had endoscopies for reasons other than dyspepsia had antral gastritis or ulcers endoscopically, 67% of all patients had histological abnormalities. 23% of group II patients with histologic gastritis were positive for H. Pylori. Almost 50% of patients who had an endoscopically normal antrum in both groups had histological abnormalities. Conclusions: There is a high incidence of histopathologic (66-67%) gas- tritis in patients presenting with to a tertiary care setting for EGDs with or without upper gastrointestinal tract symptoms, NSAID use or H. Pylori. Further, 50% patients with normal endoscopic appearance have abnormal histopathology. These data, pending larger such studies, suggests routine antral biopsies on EGDs would be beneficial. S43AJG – September, Suppl., 2003 Abstracts