AI310 AGA ABSTRACTS GASTROENTEROLOGY Vol. 118, No.4

Allvalues are means±SD, " p<0.01 vs. wtIwt.TP<0.05 VS. wtlmut

interval of at least 2 weeks between the 2 trials . Twenty-four-hour intra­gastric pH measurements were performed on day 4 of each dosin g period .Results : When lansoprazole was admini stered, the median intragastri c pHof poor metabolizer genot ype (mUl/mut) group was sign ificantly higherthan that of the other genotype groups. However, with rabeprazole, therewere no significant differences in median intragastric pH among 3 geno­type group s. Conclusion : CYP2CI9 genotype status has a major influenceon the effect of lansoprazole but not on that of rabeprazole.

5981

A NOVEL PROTON-PUMP INHIBITOR, RABEPRAZOLE, AT·TENUATES THE EXTENSION OF H.PYWRI-ASSOCIATED GAS·TRIC MUCOSAL OXIDATIVE LESION IN THE MONGOLIANGERBIL THROUGH ANTI·OXIDANT PATHWAYS.Hidekazu Suzuki, Chikako Watanabe , Shoichi Nagahashi, MasaharuMiyazawa, Masaru Sato, Motoak i Bessho , Chizuko Kawaguchi, HiroshiNagata, Soichiro Miura, Hiromasa Ishii, Keio Univ Sch of Medicine,Tokyo, Japan; Eisai Research Lab . Tokyo, Japan ; National Defense MedColi . Tokorozawa, Japan .

Background and Aim . We have previously reported that Il.p ylori (Hp)colonization evoked the gastric mucosal inflammation with an extensiveincrease in lipid peroxides in Mong olian gerbils (Free Radic. Bioi. Med.26:679, 1999). Rabeprazole (RZ) , a novel proton pump inhibitor, has beenreported to have a mild toxicity to Hp and to be the inducer of anti-oxidant,glutathione, in gastri c mucosa. The present study was designed to examinethe effect of RZ on gastric mucosal lesion formation in Hp-colonizedgerbils. Methods . Fourty-one male Mongolian gerbils (MGS/Sea, 50g)were orally inoculated with Hp (ATCC43504, 5 X 108CFU/ml) (Hp group)and 40 gerbils were inoculated with the culture media (control ). Twenty­one gerbils in the Hp group and 20 gerbils in the controls were injec tedwith RZ (1.0 mglkg body weight, s.c.) for a week at 5 weeks after thebacterial inoculation. The remainin g gerbils were injected with saline.Fourty two gerbils were evaluated 12 weeks after the inoculation and theremaining 39 gerbils were evaluated 24 weeks after the inoculation. Tissueneutrophil infiltration was assessed by the activity of myeloperoxidase(MPO) . Mucosal oxidative stress was eva luated by the contents of thio­barbituri c acid-reactive substances (TBARS) and total glutathione. Result s.

5980

COX·2 EXPRESSION, CELLULAR APOPTOSIS AND PROLIFER·ATION IN INTESTINAL METAPLASIA AT ONE YEAR AFTERERADICATION OF H, PYLORI,Joseph Jy Sung, Minnie Yy Go, Alfred SI Cheng, Wai K. Leung , FrancisKl Chan, Sr Lin, Chinese Univ of Hong Kong, Hong Kong, Hong Kong ;Beijing Med Univ, Beijing. P. R. China.

BACKGROUND H pylori infecti on is known to increase apoptosis andproliferation of gastric epithelial cell s and up-regulate COX-2 expre ssion.The relationship of COX -2 expres sion and cellular proliferation/apoptosisin H pylori infection is unclear . Controversies also exist in whetherchanges in apoptosis/proliferation would lead to regression of prem alignantgastri c lesions. AIM To study the effects of H pylori eradication onCOX-2 expression and cellular apoptosis and proliferation in patien ts withintestinal metaplasia (1M). PATIENTS AND METHODS Endoscopic bi­opsies were taken from the body of the stomach in patients with H. pyloriinfection and 1M before and at one year after eradication of the bacterium .The degree of 1M, apopt osis, proliferation indices and COX-2 express ionwere determined by H&E stain (Updated Sydne y Classification), term inaluridine deoxynucleotidyl nick end-l abel ing (TUNEL), immuno-histochem­istry using anti-Ki67 monoclonal antibody and anti-COX-2 antibody.COX-2 expre ssion was scored by a semi-quantitative scale of 0-4: O= ab­sence of staining, 1= <5%, 2=5-25%, 3=25-50% and 4= >50% cellstaining positive. All assessments were done by 2 independent investiga­tors blinded to the treatment received. RESULTS 27 paired biops y sampleswere exam ined . A significant drop in apoptotic, proliferative indices aswell as COX-2 expre ssion was documented after eradication of H. pylori .Regression of COX -2 expression was associated with reducti on in apopto­tic index but not with proliferative index. No significant change was foundin the grading of 1M at one year after treatment. CONCLUSION Despitereduction of apoptosis and proliferation of epithelial cell s and down­regulation of COX-2 after eradication of H. pylori, there was no significantchange in the 1M at one year.

4.4±1.25.3±055.B±1.0

3.2±1.24.3±O.7

6.2±1.1 ..'t

Median intragastric pHlansoprazole 30 mg rabeprazole 20mg

COX·2 Apoptosls (%) Ki·67(%) 1M grading

2.59±062 7392±7.22 71.04±14.02 1.71±0.771.88±0.93 17.11±15.63 17.04±14.24 182±0.88

0.048 <0.001 <0.001 0.70

665

wtIwtwt'mulmutlmut

Genotype

Before RxAfterRxp

5978

BLEEDING ESOPHAGEAL ULCERS CAUSED BY NSAIDS.Choic hi Sugawa, Daisuke Higuchi , Cha rles E. Lucas, Kunio Ukawa,Wayne State Univ, Detroit , MI ; Showa Univ, Yokohama, Japan .

Nonsteroidal anti-inflammatory drugs (NSAIDs) can damage each level ofthe gastrointestinal tract from the esoph agus through the large intestine.NSAID-induced esophageal ulceration causes pain, difficulty swallowingand anemia. Overt gastrointestinal bleeding from NSAID-related esopha­geal injury has rarely been reported. This report describes seven patients(pts) with NSAID-indu ced bleeding esophageal ulcers. There were threemales and four females , with NSAID-induced esophageal ulcers docu­mented by endo scopy from August 1991 to March 1999. The mean age was48.5 years (range, 20 to 64 years ). The cause of injury was aspirin in threepts, Ibupr ofen in two pts and a combination of aspirin and Ibuprofen in twopts. Six of seven pts required blood transfusion. Endoscopic hemo stasi swas successful in two pts. Three pts had hiatal hernia . One pt with a hiatalhernia devel oped stricture due to reflu x two year s after a bleeding episode.One pt with an aorto-esophageal fistula died . Six pts responded to medicaltreatment and their ulcers healed in three to four weeks. The most commonsympt oms were hematemesis (six pts) , melena, anemia, retrosternal painand dysphagia. NSAID-induced bleed ing ulcers in our seven pts had adistin ct endoscopic feature: a large shallow, discrete ulcer in the mid­esoph agus. In addition to the large ulcer, a few small shallow ulcers werefound in two pts. These findings suggest that the injury resulted frommucosal contact with NSAIDs. Esophageal transit of medications may bedelay ed in pts who have anatomic impediments to the passage of pills or inpts with abnorma l peristalsis . Prec ise history and immediate endoscopicexamination are most important in establishin g a diagno sis of esophagealulcer. If drug-induced injury is recogn ized early and treated appropriately,it is fully reversible. Offending medicat ions should be discontinued com­pletely. Pts should be counseled to take pills in an upright posture withliberal amounts of fluids well before retiring for the night-

5977

CHANGES IN THE METABOLISM OF THE EXTRACELLULARMATRIX (ECM) INFLUENCE THE DEVELOPMENT OF CRYPTHYPERPLASIA AND VILLOUS ATROPHY IN MICE.Eckhard R. Stub er, Dana Richter, Charlotte Kopitz, Ulrich R. Foel sch, IMedicine Uni-K1inik, Kiel, Germany; I Meidzin ische Uni-K1inik, Kiel ,Germany.

Introduction : Intestinal crypt hyperplasia and villous atrophy develop in aT-cell depend ent way during the course of a semi-allogenic Graft- versus­Host reaction (GvHR) in mice. The exact immune mechanisms that resultin these intestinal manifestations are unknown, so far . Previous studiessuggest an important role of certa in matrix metalloproteases (MMP s) andof specific ECM components for this intestinal transformation. Thus, thepresent study investigates changes in the composition of the ECM , alter­ations in the expression of certain MMP s and the effect of a pharmaco­logical inhib ition of MMP s during the development of villous atrophy andcrypt hyperpl asia. Methods : The intestinal expres sion of fibronectin , col­lagen type IV, laminin and tena scin as well as of MMP-2, MMP-3 andMMP- 9 was determined using specific semi-quantitative RT-PCR. Further­more, to investigate the relevance of a putative MMP activation for theintestinal manifestations of the GvHR in mice, the MMP- inhibitor Bati­rnastat was applied. Results: As measured by RT·PCR the mRNA contentof tenascin is reduced in GvHR-mice by approximately 50% as comparedto contro l animals. However, the expression of collagen type IV, fibronec­tin and laminin seems to be grossly unch anged. In addition, the expressionof MMP -2, MMP-3 and MMP-9 is markedly upregulated. Furthermore, theapplication of the MMP-inhibitor Batimastat prevents the development ofcrypt hyperplasia and villous atrophy in GvHR-mice. Conclusions:Chan ges in the metabolism of the ECM occur during the establishment ofcrypt hyperplasia and villous atrophy in GvHR-mice. An inhibition ofECM degrading enzymes (MMP s) prevents this intestinal transformation.

5979

EFFECT OF LANSOPRAZOLE AND RABEPRAZOLE ON INTRA·GASTRIC ACIDITY IN RELATION TO CYP2CI9 GENOTYPESTATUS.Mak oto Suke gawa, Kaku Hokari, Hiroshi Takeda. To shiro Sugiyama,Mototsugu Kato, Masahiro Asaka, Hokkaido Univ Sch of Medic ine, Sap­poro, Japan.

Background and aim: Proton pump inhibitors are metabolized by polymor­phic cytochrome P450 2CI9 enzyme (CYP2C I9) in the liver. The contri­bution of CYP2CI9 on the metabolism of rabeprazole is much less com­pared with that of the other proton pump inhibitors, such as omeprazole andlansoprazole . Several clinical stud ies have been indicated that CYP2CI9genotype status is associated with therap eutic effect of omeprazole, includ­ing the effect on intragastric pH and cure rates for H pylori infection. Theaim of this study was to investigate whether the effect of lansoprazole andthat of rabeprazole on intragastric pH arc modified by CYP2C 19 genotypestatus. Methods: Seventeen healthy male volunteers with 3 different geno­types of CYP2CI9 were studied. Two mutated alleles responsible for poormetab olizer phenotype, CYP2C19*2 and CYP2C19*3, were detected by aPCR-RFLP method. All were serologically negative for H pylori. Eachvolunteer was administered lansoprazole 30 mg or rabeprazole 20 mg oncedaily for 4 days in a randomized crossover manner. There was a washout