Optimal duration for triple therapy against non-resistant strains of Helicobacter pylori

A1286 AGA ABSTRACTS

GASTROENTEROLOGY Vol. 118, No.4

5874 ASSESSMENT OF A DISCRIMINAT PRONOSTIC ANALYSIS FOR PEPTIC ULCER HEMORRHAGE. Antoni Obrador, Juan Vidal, Luis Barranco, Amparo Sapina, Juan Riera, Pere Vaquer, Alfred Llompart, Jaume Gaya, Son Dureta Hosp, Palma, Spain. In a previous study, which included more than 300 patients, we accomplished a logistic regression multivariate analysis of the relapsing factors of the gastroduodenal ulcer hemorrhage. The independent variables were: type of ulcer base, size of the ulcer, endoscopic detection of blood in the explored tract, and clinical signs of low cardiac output. Objective. Assessment of a discriminative analysis for predicting rebleeds due to gastroduodenal ulcer in a prospective series of patients. Patients and methods. The patients who arrived at the hospital for an upper gastrointestinal hemorrhage (after performing a diagnostic endoscopy) were evaluated to estimate the risk of relapse according to the discriminant analysis obtained in the same center. The applied analysis is as follows: Probability = 1/(1+e- Z ) , being e = 2,7183 and z = the sum of the value of the variables in a certain patient. The study was carried out from July 1998 to May 1999. The series was of 172 patients. Results: the following contingence table shows the results of the application of the discriminative function. These data agree to obtain the next parameters: sensibility 13,3%; specificity 99,4%; positive predictive value 66,6; negative predictive value 92,3% and a global efficiency of 91,3%. If we had applied the four criteria in the 13 patients in whom the analysis did not predict the relapse, there had not turned out any error in the prevision. Conclusions. I) The discriminative analysis obtained applied to prospective series yields a small percentage of classification errors in the prediction of rebleeding. 2) The application of the four criteria that integrates the analysis permits the correction of the errors in all the cases. 3) This information is basic to set up a program of early discharge for patients suffering from gastroduodenal ulcer hemorrhage. Contingence table ofreal relapses vs predicted relapses

yes no

total

yes

no

total

2 13 15

1

156 157

3 169 172

Ordinates are the predicted relapses and abscisses are real relapses

5875 OPTIMAL DURATION FOR TRIPLE THERAPY AGAINST NON· RESISTANT STRAINS OF HELICOBACTER PYLORI. Keiji Ogura, Haruhiko Yoshida, Shin Maeda, Yutaka Yamaji, Yuzo Mitsuno, Masao Akanuma, Yoshihiro Hirata, Makoto Okamoto, Takao Kawabe. Yasushi Shiratori, Masao Ornata, The Institute for Adult Diseases, Tokyo, Japan; Dept of Gastroenterology. Univ of Tokyo. Tokyo. Japan; Univ of Tokyo, Tokyo, Japan. BACKGROUND: We reported mixed infection with wild-type (sensitive) and mutant (resistant) Helicobacter pylori strains by using a PCR-based preferential homo-duplex formation assay (PCR-PHFA) to detect 23S rRNA gene mutations associated with clarithromycin resistance (J Clin Microbiol, in press). Half cases with the mixed infection were determined as sensitive by conventional MIC assessment and yet failed in clarithromycin-based therapy. AIM: To assess the efficacy of clarithromycin-based triple therapy in patients infected exclusively with wild-type strains as determined by PCR-PHFA. METHODS: Ninety patients who had pure wild-type H. pylori infection were randomly assigned to take clarithromycin (200 mg b.i.d.), amoxicillin (500 mg q.i.d.), and lansoprazole (30 mg b.i.d.) for either 5 days or 7 days (n = 48 and n = 42, respectively). The outcome of eradication was assessed by [13C] urea breath test. RESULTS: Eradication rates were 36/48 (75%) vs. 39/42 (93%) by intention-to-treat analysis (P =0.022), and 36/45 (80%) vs. 39/40 (98%) by per protocol analysis (P =0.012), for the 5-day and 7-day protocols. respectively. Compliance and the incidence of untoward effects were similar in both groups. CONCLUSIONS: Seven-day administration is necessary and sufficient for the triple therapy with clarithrornycin, amoxicillin, and Iansoprazole in patients with pure wild-type (clarithromycin-sensitive) H. pylori infection.

5876 REGRESSION OF ATROPHIC GASTRITIS AND INTESTINAL METAPLASIA IN PATIENTS FROM WHOM HELICOBACTER PY· LORI WAS ERADICATED MONITORED PROSPECTIVELY FOR AT LEAST ONE YEAR. Toshifumi Ohkusa, Kazuhiko Fujiki, Ichizen Takashimizu, Jiro Kumagai, Toru Tanizawa, Yoshinobu Eishi, Tokyo Med and Dental Univ, Tokyo, Japan. Background: There have been few endoscopic studies during long-term follow-up after eradication of Helicobacter pylori. We examined patients repeatedly by endoscopy and biopsy for at least I year after treatment intended to eradicate the bacterium. Methods: We evaluated 163 infected patients given a proton-pump inhibitor and antibiotics who had at least three endoscopic examinations with antral and corporal biopsies before treatment and again 1-3 and 12-15 months after treatment ended. The

infection was diagnosed if at least two of four tests (culture, histology. rapid urease test. and urea breath test) gave positive results. The endoscopic and histological severity of gastritis was graded by the Sydney system. Statistical analysis was done by the unpaired t-test, Fisher exact test. Wilcoxon rank-sum test, and Wilcoxon signed-rank test. Results: For the liS patients (mean age, 54 years; 84 men and 31 women) treated successfully. the proportions with edema, erythema, friability. exudate, and rugal hypertrophy had decreased by 1-3 months, and the proportions with nodularity had decreased by 12-15 months. Mean neutrophil activity and inflammation decreased by 1-3 months. and both atrophy in the corpus and intestinal metaplasia in the antrum had regressed by 12-15 months. The mean extent of atrophy in terms of Kimura-Takemoto atrophic patterns had decreased by 12-15 months. In the 48 other patients (mean age. 59 years; 30 men and 18 women) in whom the bacterium was not eradicated. there were no clear changes in findings. Conclusion: In long-term follow-up but not short-term follow-up. atrophy and intestinal metaplasia regressed in patients in whom the bacterium was eradicated. Regression of the endoscopic features of edema, erythema. friability, exudate. rugal hypertrophy, and nodularity was related to regression of the histological findings of neutrophil activity and inflammation.

5877 ACTIVATED PROTEIN C INHIBITS THE SECRETION OF TUMOR NECROSIS FACTOR-A INDUCED BY CAG A FROM HEL-

ICOBACTER PYLORI. Satoko Oka, Esteban Cesar Gabazza, Ichiro Imoto, Nagahito Toyoda, Shigehito Nakashima, Michihiko Yamaguchi, Yukiko Taguchi, Yukihiko Adachi, Koji Suzuki. Mie Univ Sch of Medicine. Tsu, Japan. Background/Aim: Activated protein C (APC). besides regulating the activation of the coagulation system, has recently been suggested to modulate the inflammatory response. We previously reported that APC is present at high concentration in the gastric mucosa of patients with Helicobacter pylori (H. pylori) -associated gastritis, and that lipopolysaccharide (LPS), cytotoxin-associated antigen (CagA) and vacuolating cytotoxin from this bacterium increase the formation of APC on the surface of monocytic cells . The secretion of cytokines induced by CagA has been recently reported to play an important role in the mechanism of the gastric mucosal injury. In the present study, we evaluated whether APC inhibits the CagA-induced secretion of cytokines in THP-I cells. Method: The monocytic THP-I cells were culutured using RPMI 1640 medium containing 2.5% fetal bovine serum. For assessing the stimulatory activity of CagA on TNF-asecretion from THP-l cells, the cells (lxlOOOOOO/well) were stimulated with varying concentrations of CagA. The time course of CagA-mediated secretion of TNF-awas also evaluated. For evaluating the inhibitory activity of APC, THP-l cells were cultured in the presence of defined concentrations of CagA and of varying concentrations of APC (0-15/-Lg/ml). Results: CagA significantly (p<0.05) stimulated the secretion of TNF-afrom THP-I cells in a concentration-dependent manner; CagA also time-dependently stimulated the secretion of TNF-afrom these monocytic cells, the secretion reaching a peak at 12 h. Culture of THP-I cells in the presence of APC, significantly inhibited the secretion of CagA-inducedsecretion of TNF-ain a dose-dependent manner. Conclusion: CagA was found to increase the secretion of TNF-afrom monocytic cells. suggesting an important mechanism by which CagA injures the gastric mucosa. In addition. APC was found to inhibit significantly the CagA-mediated stimulation of TNFosecretion. suggesting that APC may play an important role in the protection of the gastric mucosa during H. pylori -associated gastritis.

5878 MECHANISM BY WHICH BEER AND WINE STIMULATES GASTRIC ACID SECRETION IN DOGS: ROLE OF GASTRIN. Susumu Okabe, Tasuku Murai, Kuniharu Matsuno, Atsuko Ono, Kyoto Pharm Univ, Kyoto, Japan. Background/Aim: Oral administration of beer and wine stimulates gastric acid secretion in man and animals, most probably by inducing an increased gastrin release. To confirm the role of circulating gastrin, we examined whether or not (S-0509, a CCK-B/gastrin-receptor antagonist) inhibits the gastric acid secretion stimulated by beer, wine and ethanol in dogs. Materials & Methods: Animals: Ten beagle dogs (10-14 kg) of both sexes, either with a vagally-denervated Heidenhain pouch (n=7) and innervated fistula (n=3), were used in the study. Gastric samples were collected every IS min and analyzed for volume and acid output. Acid secretion was stimulated by beer (50-200 ml), red or white wine (25-100 ml), or 5% ethanol (contained in the beer, 200 ml). S-0509, famotidine or atropine (as reference drugs) was given iv 15 min prior to secretagogues. Results: Beer, wine (red and white) and 5% ethanol significantly stimulated gastric acid secretion in dogs with a Heidenhain pouch and gastric fistula in a volumerelated mannter. Special beer (Root beer), without ethanol, also stimulated the gastric acid secretion which was not different from the normal beer with ethanol. The degree of acid stimulation by 50 ml of red and white wine was much the same. The peak acid outputs were observed 30 min after administration, approx. 0.4 mEq/15 min for beer, 0.2 mEqll5 min for wine and 0.1 mEq/15 min for 5% ethanol. These stimulated secretion returned to the basal level within 60 min. Serum gastrin level significantly increased after the administration of beer and wine, but not after 5% ethanol. S-0509 (0.3, I and 3 mg/kg) significantly inhibited both beer and wine-stimulated acid secretion in a dose-related manner. The duration of antisecretory effect of S-0509 on wine-stimulated secretion was shorter