The effect of antibiotic resistance on the outcome of three 1-week triple therapies against Helicobacter pylori

Transcription

1 Aliment Pharmacol Ther 1999; 13: 667±673. The effect of antibiotic resistance on the outcome of three 1-week triple therapies against Helicobacter pylori A. PILOTTO*, G. LEANDRO, M. FRANCESCHI*, M. RASSUà, L. BOZZOLA, F. FURLANà,F.DIMARIO±&G.VALERIO* *Department of Geriatrics, S. Bortolo Hospital, Vicenza; Department of Gastroenterology, IRCCS, Castellana Grotte, Bari; àmicrobiology Unit, S. Bortolo Hospital, Vicenza; Clinical Pathology Unit, S. Bortolo Hospital, Vicenza; and ±Department of Gastroenterology, University of Padova, Italy Accepted for publication 21 December 1998 SUMMARY Background: Resistance of Helicobacter pylori to antibiotics may be a major reason for treatment failure. Aim: To evaluate the effect of primary H. pylori resistance to antibiotics on the cure rates of three anti- H. pylori 1-week triple therapies. Methods: One hundred and sixteen consecutive patients diagnosed H. pylori-positive by gastric histology, rapid urease test and culture were enrolled. Activity of tested antibiotics was determined by means of the E-test. Patients were treated for 7 days with: (i) pantoprazole 40 mg o.d. plus amoxycillin 1 g b.d. and metronidazole 250 mg q.d.s. (PAM); (ii) pantoprazole 40 mg o.d. plus clarithromycin 250 mg b.d. and metronidazole 250 mg q.d.s. (PCM); or (iii) pantoprazole 40 mg o.d. plus amoxycillin 1 g b.d. and clarithromycin 250 mg b.d. (PAC). Two months after completion of therapy, endoscopy and gastric biopsies were repeated. Results: Primary resistance rates to metronidazole, clarithromycin and amoxycillin were 17.2, 6.9 and 0%, respectively. Overall H. pylori cure rates expressed as intention-to-treat and per protocol analyses were, respectively, 79% and 86% with PAM, 82% and 89% with PCM, and 85% and 85% with PAC. Signi cantly lower cure rates were observed in metronidazoleresistant patients treated with PAM (56% vs. 96%, P ˆ 0.01) or PCM (50% vs. 97%, P ˆ 0.01). A trend towards lower H. pylori cure rates was observed in clarithromycin-resistant patients treated with PCM (67% vs. 91%, P ˆ 0.74) or PAC (50% vs. 87%, P ˆ 0.68). Conclusion: Primary resistance to metronidazole in uences the H. pylori cure rate of anti-h. pylori proton pump inhibitor-based triple therapies which include this antibiotic. A similar trend exists for primary clarithromycin resistance. INTRODUCTION The cure of Helicobacter pylori infection results in healing of chronic active gastritis and dramatically reduces peptic ulcer relapse rate. The treatment regimens for H. pylori infection currently recommended Correspondence to: Dr A. Pilotto, Divisione Geriatria, Ospedale Civile `S. Bortolo', Via Rodol 37, Vicenza, Italy. pilotto@mbox.avnet.it involve a 1-week triple therapy consisting of a proton pump inhibitor plus two of the following antibiotics: amoxycillin, a nitroimidazole (metronidazole or tinidazole) and clarithromycin. 1 Resistance of H. pylori to the antibiotics included in current therapeutic regimens has been reported as a major reason for treatment failure. 2 Therefore, it has been suggested that the selection of antibiotics used in association with proton pump inhibitors for primary treatment of H. pylori infection should be based on the Ó 1999 Blackwell Science Ltd 667

2 668 A. PILOTTO et al. prevalence of antibiotic resistance of H. pylori strains in speci c populations. 3 The aim of this study was to evaluate the effect of primary H. pylori resistance to antibiotics, i.e. amoxycillin, metronidazole and clarithromycin, on the cure rates of three anti-h. pylori 1-week triple therapies consisting of a proton pump inhibitor (pantoprazole) plus two antibiotics, i.e. amoxycillin and metronidazole (PAM), clarithromycin and metronidazole (PCM), and amoxycillin and clarithromycin (PAC). MATERIALS AND METHODS Patients and treatments This was a prospective, single-centre study performed in patients who consecutively underwent an upper gastrointestinal endoscopy at the Digestive Pathophysiology Center of the Geriatric Department in Vicenza, Italy, from June to December The inclusion criterion was the presence of gastric H. pylori infection, as documented by gastric histology, rapid urease test and a positive culture for H. pylori. Exclusion criteria were ulcer complications, i.e bleeding and perforation; chronic therapy with non-steroidal anti-in ammatory drugs; concomitant or previous (within the preceding 4 weeks) treatments with antibiotics, proton pump inhibitors (omeprazole, lansoprazole, pantoprazole), H 2 -blockers (cimetidine, ranitidine, nizatidine, famotidine, roxatidine) or H. pylori eradication therapy of any kind; severe concomitant diseases; gastric cancer or other neoplasms; previous major gastrointestinal surgery; and predictable low compliance for anti-h. pylori treatment. After diagnosis, all patients received a clear explanation of the purpose of the study and those who gave their informed consent were consecutively assigned to one of the following regimens, according to a computergenerated randomization list: 1 Pantoprazole 40 mg o.d. plus amoxycillin 1 g b.d. and metronidazole 250 mg q.d.s. (PAM). 2 Pantoprazole 40 mg o.d. plus clarithromycin 250 mg b.d. and metronidazole 250 mg q.d.s. (PCM). 3 Pantoprazole 40 mg o.d. plus amoxycillin 1 g b.d. and clarithromycin 250 mg b.d. (PAC). Pantoprazole therapy was prescribed for 1 week in patients with gastritis and for 4 weeks in gastric or duodenal ulcer patients. The antibiotic combination was administered for 7 days in all patients. Two months after completion of the drug regimen, endoscopy with gastric biopsies for histology and the rapid urease test were repeated. Patients were considered H. pylori-negative after therapy if both histology and rapid urease test were H. pylori-negative. Cultures were not performed in post-treatment endoscopies. Patients were evaluated during the 1-week antibiotic therapy and 4 weeks after baseline to record side-effects and count tablets. Compliance was de ned as `good' if more than 90% of the tablets had been taken by the patient. Histology and rapid urease test During endoscopy, gastric biopsies were taken from both the antrum, 5 cm proximal to the pylorus (four biopsies), and from the body, halfway along the greater curvature (four biopsies). Two antral and two body biopsies were used for histological analysis, while one from each site was used for culture and the rapid urease test (CLO-test; Delta West Pty Ltd, Western Australia). For histological examination, biopsy specimens were immediately xed in buffered neutral formalin and embedded in paraf n. Sections were stained with haematoxylin±eosin and modi ed Giemsa for the detection of H. pylori. Chronic gastritis was de ned histologically according to the Sydney system. 4 Microbiology The gastric bioptic specimens taken during endoscopy were transported to the laboratory within 1 h. Biopsies were cultured onto selective media (Helicobacter selective agar plus 7% de brinated horse blood; Becton Dickinson, Cockesville, MD). The plates were incubated for 5±7 days at 37 C in microaerobic conditions (Campy Pak Plus, BBL; Becton Dickinson, Cockesville, MD) and 100% relative humidity. H. pylori strains grown were re-suspended in 2 ml of Dulbecco's modi ed Eagle medium for cell culture. From this suspension, adjusted to a no. 3 McFarland turbidity standard, 100 ll were ooded on Columbia agar plates containing 7% horse blood. The E-test strip was placed on the plate when the surface of the plate was dry. All plates were incubated for 3 days at 37 C under microaerobic conditions. The minimum inhibitory concentration (MIC) for metronidazole, clarithromycin and amoxycillin were determined by the E-test (AB Biodisk, Uppsala,

3 ANTIBIOTIC RESISTANCE AND H. PYLORI ERADICATION 669 Table 1. Prevalence of resistant H. pylori strains to the different antibiotics evaluated in a population of 116 patients who had tested positive for H. pylori gastric infection No. of patients Mean age H. pylori isolates (rate) Males Females (range) Peptic ulcer (%) Chronic gastritis (%) Antibiotic susceptible 89/116 (76.3%) 37 (41.5%) 52 (58.4%) 70.4 (33±89) 43/89 (48.3%) 46/89 (51.7%) Metronidazole resistant 20/116 (17.2%)* 10 (50%) 10* (50%) 61.6 (31±90) 8/20* (40%) 12/20 (60%) Clarithromycin resistant 8/116 (6.9%)* 7 (87.5%) 1* (12.5%) 71.3 (25±90) 5/8* (62.5%) 3/8 (37.5%) Amoxycillin resistant 0/ Ð Ð Ð Total resistant 27/116 (23.3%) 17 (62.9%) 10 (40.7%) 64.8 (25±90) 12/27 (45%) 15/27 (55%) *One patient was resistant to both metronidazole and clarithromycin. Sweden) following the manufacturer's instructions. Strains were considered resistant when the MIC was > 8 lg/ml for metronidazole, > 2 lg/ml for clarithromycin and > 1 lg/ml for amoxycillin according to NCCLS recommendations. 5 Statistics Results were evaluated using both `per protocol' and `intention-to-treat' analyses; the 95% con dence intervals (95% CI) were also calculated. Statistical analysis was performed by means of the chi-squared test (comparison of eradication rates obtained with the three treatments and comparison of cure rates among patients with H. pylori strains resistant and sensitive to antibiotics) and the Mann±Whitney test (comparison of baseline characteristics). The predicted cure rates presented in Figure 2 were estimated using a linear model. Power of the tests was calculated by the standard formula. All P-values were two-tailed with statistical signi cance indicated by a value of P < RESULTS Prevalence of H. pylori antibiotic resistance One hundred and sixteen patients ful lling the inclusion criteria were enrolled in the study: 48 males and 68 females. The mean age was 67.7 years with a range of 25±90 years. Active gastric or duodenal ulcer was diagnosed in 55 patients (47.4%), while the remainding 61 patients (52.5%) had chronic gastritis. Table 1 shows the overall primary resistance of H. pylori strains to the different antibiotics tested in this population. The prevalence of H. pylori resistance to metronidazole was 17.2% (20/116) and to clarithromycin was 6.9% (8/116); no H. pylori strains resistant to amoxycillin were found. In one case, H. pylori was found to be resistant to both metronidazole and clarithromycin. No differences in prevalence of antibiotic resistance were found as regards age, sex and the presence of peptic ulcer. Table 2. Clinical characteristics of the 116 patients treated with one of three different anti-h. pylori regimens PAM PCM PAC Total Number of patients Mean age, years (range) 71.7 (31±89) 65.7 (25±90) 70.1 (39±88) 67.7 (25±90) Males (%) 16 (42.1%) 20 (52.6%) 12 (30%) 48 (41.37%) Active ulcer 19 (50%) 19 (50%) 17 (42.2%) 55 (47.4%) Chronic gastritis 19 (50%) 19 (50%) 23 (57.5%) 61 (52.6%) Antibiotic sensitive (%) 26 (68.4%) 28 (73.7%) 35 (87.5%) 89 (76.7%) Metronidazole resistant (%) 10* (26.3%) 7 (18.4%) 3 (7.5%) 20 (17.2%) Clarithromycin resistant (%) 3* (7.9%) 3 (7.9%) 2 (5.0%) 8 (6.9%) Drop outs (no. of patients) 3* * One patient was resistant to both metronidazole and clarithromycin. PAM, pantoprazole, amoxycillin, metronidazole; PCM, pantoprazole, clarithromycin, metronidazole; PAC, pantoprazole, amoxycillin, clarithromycin.

4 670 A. PILOTTO et al. Table 3. Eradication rates, expressed using both intention-to-treat and per protocol analyses, study drop outs and side-effects in patients divided according to the three anti-h. pylori regimens Intention-to-treat analysis Per protocol analysis Regimen No. of patients Cure rate 95% CI Cure rate 95% CI Drop-outs Side-effects PAM 38 79% (30/38) 66±92 86% (30/35) 74± PCM 38 82% (31/38) 69±94 89% (31/35) 78± PAC 40 85% (34/40) 74±96 85% (34/40) 74± PAM, pantoprazole, amoxycillin, metronidazole; PCM, pantoprazole, clarithromycin, metronidazole; PAC, pantoprazole, amoxycillin, clarithromycin. H. pylori cure rates and side-effects Table 2 shows the clinical characteristics of patients, divided according to the different treatments: no signi cant differences were found among the three groups of patients as regards sex, age, endoscopic diagnosis or H. pylori antibiotic resistance rates. Six patients (5.1%) dropped out of the study: two were non-compliant, consuming less than 90% of the prescribed medication and four patients refused the nal endoscopic examination. Nine patients (7.7%) reported side-effects; two were in treatment with PAM (one patient reported diarrhoea and one glossitis), three were in treatment with PCM (two patients reported diarrhoea and one headache) and four patients were in treatment with PAC (two patients reported glossitis and two urticaria). None of these side-effects required suspension of the treatment. The overall H. pylori cure rates, evaluated regardless of the prevalence of antibiotic resistance and expressed as both intention-to-treat and per protocol analyses, were, respectively: 79% and 86% with PAM, 82% and 89% for PCM and 85% and 85% with PAC. No statistical differences in cure rates were found among the three regimens (Table 3). Effect of H. pylori antibiotic resistance on cure rates Figure 1 illustrates the H. pylori cure rates strati ed according to the presence of metronidazole and clarithromycin resistance. In comparison to metronidazolesensitive patients, signi cantly lower cure rates were observed in metronidazole-resistant patients treated with PAM (56% vs. 96%, P ˆ 0.01) and PCM (50% vs. 97%, P ˆ 0.01), but not with PAC (100% vs. 84%, P ˆ N.S.). In comparison to clarithromycin-sensitive patients, lower H. pylori cure rates were observed in clarithromycin-resistant patients treated with PCM (67% vs. 91%, P ˆ 0.74) and PAC (50% vs. 87%, P ˆ 0.68) but not with PAM (100% vs. 85%, P ˆ N.S.). The differences were not statistically signi cant due to the low number of patients with clarithromycin-resistant H. pylori strains (power of test: 1 ± b < 0.70). DISCUSSION Recent recommendations from The European Helicobacter Pylori Study Group 1 suggest that the optimal anti-h. pylori regimen for primary treatment is a combination of a proton pump inhibitor with two of the following antibiotics: amoxycillin, clarithromycin and a nitroimidazole (either tinidazole or metronidazole). The nal choice of which regimen to use should be based on the local prevalence of H. pylori strains resistant to these antibiotics, principally metronidazole and clarithromycin. Until recently, few studies have reported the prevalence rates of antibiotic resistance in Italy and none has evaluated the clinical effect of such resistances on the outcome of three 1-week proton pump inhibitor-based anti-h. pylori regimens. The present study demonstrated the prevalence of antimicrobial resistance in this population in Northern Italy to be 17.2% for metronidazole, 6.9% for clarithromycin and 0% for amoxycillin. The observed rate of metronidazole resistance was similar to that previously reported for Italy by a multicentre European survey, 6 and comparable to the 12% metronidazole resistance observed in Italian dyspeptic patients 7 and the 23% reported in a subset of 13 Italian duodenal ulcer patients included in a multicentre clinical trial. 8 The rate of resistance to clarithromycin observed in this study was in the range of the 5±15% reported in several studies from Europe. 2, 9±11 Regarding the amoxycillin data, the results of this study con rmed that resistance 2, 7, 8 to amoxycillin is virtually absent.

5 ANTIBIOTIC RESISTANCE AND H. PYLORI ERADICATION 671 Figure 1. Cure rates of H. pylori infection according to the presence of metronidazole (above) or clarithromycin (below) resistance. PAM: pantoprazole, amoxycillin, metronidazole; PCM: pantoprazole, clarithromycin, metronidazole. PAC: pantoprazole, amoxycillin, clarithromycin. Patients harbouring H. pylori strains resistant to metronidazole had H. pylori cure rates signi cantly lower than patients with H. pylori metronidazole-sensitive strains when treated with regimens which included metronidazole (PAM and PCM). These data are in agreement with other studies which demonstrated signi cantly lower H. pylori cure rates in metronidazole-resistant patients after treatment with metronidazole in triple therapies which included a proton pump inhibitor and amoxycillin, 12±17 a proton pump inhibitor 14, 18, 19 and clarithromycin, colloidal bismuth subcitrate and amoxycillin, 20±22 as well as in quadruple therapy with omeprazole, colloidal bismuth subcitrate and tetracycline. 23 In con ict with the present results and with those of the above-cited reports, 14, 18, 19 some studies failed to nd a negative in uence of metronidazole resistance on the clinical ef cacy of a proton pump inhibitor-based triple therapy which included metronidazole (or tinidazole) in combination with clarithromycin. 12, 17, 24, 25 A different methodological approach to testing antimicrobial activity 25 or the low number of patients with metronidazole-resistant H. pylori strains 17, 24 may account for this discrepancy. 2 Similarly, as reported in previous studies, 2, 8, 19 clarithromycin resistance was associated with a trend towards lower H. pylori eradication rates in patients treated with triple therapies which included clarithromycin (PCM and PAC); however, these results were not statistically signi cant (60% vs. 89%, P ˆ 0.25). The low prevalence of clarithromycin-resistant H. pylori strains in this population may account for this lack of statistical signi cance. The low power of the test (1 ± b ˆ 0.63) supports this conclusion. Because prevalence of H. pylori antibiotic resistance to metronidazole and/or clarithromycin is populationspeci c, 2, 6 the clinical relevance of this resistance may be different in speci c populations. 26 In patients of the present study, the in uence of H. pylori antibiotic resistance on clinical outcome seemed to be low. Indeed, when results were evaluated in the overall population, regardless of antibiotic resistance, the cure rates observed with the three pantoprazole-based anti-h. pylori regimens were very similar (from 85% to 89% using the per protocol analysis) to each other and to those reported in studies which used the same therapeutical 27, 28 regimens. Using data from the present study for extrapolation, we can calculate estimated cure rates in populations with different prevalences of antibiotic resistance. For example, in an area with a resistance rate to metronidazole of 40%, the predicted eradication with PAM and PCM is 80% and 78%, respectively. In an area with a resistance rate of 80%, this predicted eradication rate falls to 64±59%. The same information can be extrapolated for the clarithomycin resistance: the cure rates with PCM and PAC decrease with increasing prevalence of clarithromycin-resistant H. pylori strains (Figure 2).

6 672 A. PILOTTO et al. pump inhibitor-based triple therapies which include this antibiotic. The same trend is observed for clarithromycin. The low prevalence of H. pylori antibiotic resistance observed in this population may account for the lack of clinical differences in H. pylori cure rates among PAM, PCM and PAC regimens. ACKNOWLEDGEMENTS Financial support for this study was provided by the host institutions. The authors wish to thank Prof. Francis Megraud for comments and suggestions in reviewing the manuscript. REFERENCES Figure 2. The predicted effect of H. pylori resistance to metronidazole (above) or clarithromycin (below) on the ef cacy of anti- H. pylori regimens which include metronidazole (PAM, PCM) or clarithromycin (PCM, PAC) as estimated from data of the present study using a linear model for extrapolation. The postulated eradication rates were calculated based on the ndings that 96% of patients with metronidazole-susceptible strains were cured with PAM and 97% with PCM, while 56% of patients with metronidazole-resistant strains were cured with PAM and 50% with PCM. Similarly, 91% of patients with clarithromycin-susceptible strains were cured with PCM and 87% with PAC, while 67% of patients with clarithromycin-resistant strains were cured with PCM and 50% with PAC. Because clinical 8, 17, 19, 29 and epidemiological 30±32 studies have reported that H. pylori antibiotic resistances may change with time even within the same population, principally due to the emergence of secondary resistances to either metronidazole and clarithromycin after treating H. pylori or other infective diseases, a continuous surveillance of H. pylori susceptibility to antibiotics at the national or regional level would be needed to make successful recommendations as to the most effective primary treatment in a speci c geographical area. 2 In conclusion, primary resistance to metronidazole in uences the H. pylori cure rate of anti-h. pylori proton 1 The European Helicobacter pylori Study Group (EHPSG). Current European concepts in the management of Helicobacter pylori infection. The Maastricht Consensus report. Gut 1997; 41: 8±13. 2 Megraud F. Resistance of Helicobacter pylori to antibiotics. Aliment Pharmacol Ther 1997; 11(Suppl. 1): 43±53. 3 Megraud F. A growing demand for Helicobacter pylori culture in the near future? Ital J Gastroenterol Hepatol 1997; 29: 574±6. 4 Misiewicz JJ, Tytgat GNJ, Goodwin CS, et al. The Sydney System: a new classi cation of gastritis. J Hepatol Gastroenterol 1991; 6: 209±22. 5 National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standards. NCCLS Publication M7-A3. Villanova, PA: NCCLS, Glupczynski Y and the European Study Group on Antibiotic Susceptibility of H. pylori. Results of a multicentre European survey in 1991 of metronidazole resistance in Helicobacter pylori. Eur J Clin Microbiol Infect Dis 1992; 11: 777±81. 7 Tucci A, Varoli O, Corinaldesi R, et al. Evaluation of Helicobacter pylori sensitivity to amoxycillin and metronidazole in dyspeptic patients. Ital J Gastroenterol 1993; 25: 65±7. 8 Wurzer H, Rodrigo I, Stamler D, et al. for the ACT-10 Study Group. Short-course therapy with amoxycillin±clarithromycin triple therapy for 10 days (ACT-10) eradicates Helicobacter pylori and heals duodenal ulcer. Aliment Pharmacol Ther 1997; 11: 943±52. 9 Xia HX, Buckley M, Keane CT, O'Morain CA. Clarithromycin resistance in Helicobacter pylori: prevalence in untreated dyspeptic patients and stability in vitro. J Antimicrob Chemother 1996; 37: 473± Lopez-Brea M, Domingo D, Sanchez I, Alarcon T. Evolution of resistance to metronidazole and clarithromycin in Helicobacter pylori clinical isolates from Spain. J Antimicrob Chemother 1997; 40: 279± Jaup BH, Brandebeg A, Stenquist B, Norrby A. Prevalence of antibiotic resistant H. pylori strains among Swedish and immigrant patients with peptic ulcer disease and dyspepsia in

7 ANTIBIOTIC RESISTANCE AND H. PYLORI ERADICATION 673 Gothenburg. A consecutive study from an open access endoscopy center. Gut 1997; 41(Suppl. 1): A39(Abstract). 12 Lerang F, Moum B, Haug JB, et al. Highly effective twice-daily triple therapies for Helicobacter pylori infection and peptic ulcer disease: does in vitro metronidazole resistance have any clinical relevance? Am J Gastroenterol 1997; 92: 248± Thijs JC, Van Zwet AA, Thijs WJ, Van Der Wouden EJ, Kooy A. One-week triple therapy with omeprazole, amoxycillin and tinidazole for Helicobacter pylori infection: the signi cance of imidazole resistance. Aliment Pharmacol Ther 1997; 11: 305±9. 14 Misiewicz JJ, Harris AW, Bardhan KD, et al. One week triple therapy for Helicobacter pylori: a multicentre comparative study. Gut 1997; 41: 735±9. 15 Veldhuyzen van Zanten S, Hunt RH, Cockeram A, et al. Adding once-daily omeprazole 20 mg to metronidazole/ amoxicillin treatment for Helicobacter pylori gastritis: a randomized, double-blind trial showing the importance of metronidazole resistance. Am J Gastroenterol 1998; 93: 5± Forbes GM, Collins BJ, McCollough CA, Coombs GW, Robins PD. Short duration therapy for Helicobacter pylori in Western Australia: the impact of metronidazole resistance. Aust NZ J Med 1998; 28: 13± Adamek RJ, Suerbaum S, Pfaffenbach B, Opferkuch W. Primary and acquired Helicobacter pylori resistance to clarithromycin, metronidazole and amoxicillin: in uence on treatment outcome. Am J Gastroenterol 1998; 93: 386±9. 18 Megraud F, Lehn N, Lind T, et al. The Mach 2 study. Helicobacter pylori resistance to antimicrobial agents and its in- uence on clinical outcome. Gastroenterology 1997; 112: A216(Abstract). 19 Buckley MJM, Xia HX, Hyde DM, Keane CT, O'Morain CA. Metronidazole resistance reduces ef cacy of triple therapy and leads to secondary clarithromycin resistance. Dig Dis Sci 1997; 42: 2111± Rautelin H, Seppala K, Renkonen OV, Vainio U, Kosunen TU. Role of metronidazole resistance in therapy of Helicobacter pylori infections. Antimicrob Agents Chemother 1992; 36: 163±6. 21 Noach LA, Langenberg WL, Bertola MA, Dankert J, Tytgat GN. Impact of metronidazole resistance on the eradication of Helicobacter pylori. Scand J Infect Dis 1994; 26: 321±7. 22 Midolo PD, Lambert JR, Turnidge J. Metronidazole resistance: a predictor of failure of Helicobacter pylori eradication by triple therapy. J Gastroenterol Hepatol 1996; 11: 290±2. 23 van der Hulst RW, van der Ende A, Homan A, Roorda P, Dankert J, Tytgat GN. In uence of metronidazole resistance on ef cacy of quadruple therapy for Helicobacter pylori eradication. Gut 1998; 42: 166±9. 24 Harris AW, Pryce DI, Gabe SM, et al. Lansoprazole, clarithromycin and metronidazole for seven days in Helicobacter pylori infection. Aliment Pharmacol Ther 1996; 10: 1005± Moayyedi P, Ragunathan PL, Mapstone N, Axon AT, Tompkins DS. Relevance of antibiotic sensitivities in predicting failure of omeprazole, clarithromycin, and tinidazole to eradicate Helicobacter pylori. J Gastroenterol 1998; 33: 160±3. 26 Megraud F, Doermann HP. Clinical relevance of resistant strains of Helicobacter pylori: a review of current data. Gut 1998; 43(Suppl. 1): S61±5. 27 Pilotto A, Franceschi M, Bierti L, et al. Pantoprazole, amoxycillin and metronidazole in Helicobacter pylori-positive duodenal ulcer patients: effects on symptoms, ulcer healing and eradication rates. Gastroenterology 1998; 114: G Pazzi P, Scagliarini R, Gamberini S, Matarese V, Rizzo C, Gullini S. Short-term low-dose pantoprazole-based triple therapy for cure of Helicobacter pylori infection in duodenal ulcer patients. Aliment Pharmacol Ther 1998; 12: 731±4. 29 Tompkins DS, Perkin J, Smith C. Failed treatment of Helicobacter pylori infection associated with resistance to clarithromycin. Helicobacter 1997; 2: 185±7. 30 Ling TKW, Cheng AFB, Sung JJY, Yiu PYL, Chung SSC. An increase in Helicobacter pylori strains resistant to metronidazole: a ve-year study. Helicobacter 1996; 1: 57± Glupczynski Y, Goutier S, van der Borre C, Butzler JP, Burette A. Surveillance of Helicobacter pylori resistance to antimicrobial agents in Belgium from 1989 to Gut 1995; 37(Suppl. 1): A56(Abstract). 32 van der Wouden EJ, van Zwet AA, Vosmaer GD, Oom JA, de Jong A, Kleibeuker JH. Rapid increase in the prevalence of metronidazole-resistant Helicobacter pylori in the Netherlands. Emerg Infect Dis 1997; 3: 385±9.