Empirical rescue therapy after Helicobacter pylori treatment failure: a 10-year single-centre study of 500 patients.

BACKGROUND: Several 'rescue' therapies have been recommended to eradicate Helicobacter pylori, but they still fail in >20% of the cases, and these patients constitute a therapeutic dilemma. AIM: To evaluate the efficacy of different 'rescue' therapies empirically prescribed during 10 years to 500 patients in whom at least one eradication regimen had failed to cure H. pylori infection. DESIGN: Prospective single-centre study. PATIENTS: Consecutive patients in whom at least one eradication regimen had failed. INTERVENTION: Rescue regimens included: (i) quadruple therapy with omeprazole-bismuth-tetracycline-metronidazole; (ii) ranitidine bismuth citrate-tetracycline-metronidazole; (iii) omeprazole-amoxicillin-levofloxacin; and (iv) omeprazole-amoxicillin-rifabutin. Antibiotic susceptibility was unknown (rescue regimens were chosen empirically). OUTCOME: Eradication was defined as a negative (13)C-urea breath test 4-8 weeks after completing therapy. RESULTS: Five hundred patients were included (76% functional dyspepsia, 24% peptic ulcer). Compliance rates with first-, second- and third-line regimens were 92%, 92%, and 95%, respectively. Adverse effects were reported by 30%, 37%, and 55% of the patients receiving second-, third-, and fourth-line regimens. Overall, H. pylori cure rates with the second-, third-, and fourth-line rescue regimens were 70%, 74%, and 76%, respectively. Cumulative H. pylori eradication rate with four successive treatments was 99.5%. CONCLUSION: It is possible to construct an overall treatment strategy to maximize H. pylori eradication, on the basis of administration of four consecutive empirical regimens; thus, performing bacterial culture even after a second or third eradication failure may not be necessary.

Helicobacter pylori: its discovery and relevance for medicine.

In 1979, Warren was first to observe these bacteria in an inflamed gastric epithelium, and subsequently in peptic ulcer-associated gastritis. In 1981, Marshall starts his research and cultivates the bacteria, performs prospective studies, and administers therapy regimens using antibiotics and bismuth salts. The organism was designated Campylobacter-like after its similarity with said genus. In 1984, Marshall manages to meet Koch s postulates using self-inoculation by ingesting a bacterial culture, which resulted in gastritis that was then treated and cured with bismuth salts and metronidazole. The Gastroenterological Society of Australia rejected the abstract with preliminary data. In 1983, "The Lancet" published such data in two brief letters with the same heading, with each being signed by one of these two researchers. Cooperating with microbiologists they classify the new bacteria within the Helicobacter genus as the pylori species. Gastritis and peptic ulcer are currently considered infectious diseases. Their diagnosis and therapy include bacterial detection methods and antibiotics, respectively. In addition, a causal relationship between Helicobacter pylori and both gastric adenocarcinoma and gastric lymphoma has been established, and its association with some extra-digestive conditions has been suggested.

Third-line rescue therapy with levofloxacin is more effective than rifabutin rescue regimen after two Helicobacter pylori treatment failures.

BACKGROUND: In patients with a first eradication failure, a second (rescue) therapy still fails in > 20% of cases. AIM: To compare rifabutin and levofloxacin rescue regimens in patients with two consecutive Helicobacter pylori eradication failures. METHODS: Patients, in whom first treatment with omeprazole-clarithromycin-amoxicillin and a second trial with omeprazole-bismuth-tetracycline-metronidazole (or ranitidine bismuth citrate with these antibiotics) had failed, received 10 days of treatment with either rifabutin (150 mg b.d.) or levofloxacin (500 mg b.d.), plus amoxicillin (1 g b.d.) and omeprazole (20 mg b.d.). Cure rates were evaluated by the (13)C-urea breath test. RESULTS: Twenty patients received rifabutin, and 20 levofloxacin. All the patients returned for follow-up. Compliance in the rifabutin group was 100%. Four patients in the levofloxacin group did not take the medication correctly (in two cases due to adverse effects: myalgia and rash). Side effects in the rifabutin and levofloxacin groups were reported in 60% and 50% of the cases, respectively. Five patients (25%) treated with rifabutin presented with leucopenia, and six (30%) treated with levofloxacin presented with myalgias. Per-protocol cure rates were 45% (95% confidence interval, 26-66%) in the rifabutin group, and 81% (57-93%) in the levofloxacin group (P < 0.05). Intention-to-treat cure rates were, 45% (26-66%) and 85% (64-95%), respectively (P < 0.01). CONCLUSIONS: After two previous H. pylori eradication failures, a 10-day triple levofloxacin-based rescue regimen is more effective than the same regimen with rifabutin.

Spanish scientific output on Helicobacter pylori. A study through Medline.

OBJECTIVES: to analyze scientific output from Spanish hospitals in relation to Helicobacter pylori infection. METHODS: papers collected from the Medline database between January 1988 and December 2003 were selected. Our search strategy was: "Helicobacter pylori" [MeSH] AND ((Spain [AD] OR Espana [AD] OR Spanien [AD] OR Espagne [AD] OR Espanha [AD]) OR (Spanish [LA]) OR Spain). The following was analyzed: geographic area, Spanish or foreign publication, topic, and year of publication. Output and impact bibliometric markers were evaluated. RESULTS: in all, 691 papers were identified, of which 241 were excluded. Number of papers went from 2 in 1988 to 47 in 2002 and 13 in 2003. There were more reports in Spanish versus foreign journals (58 vs. 42%). In the first 5 years the areas with greater output were associated with diagnosis and microbiology (33 and 20%), whereas therapy was the predominating subject during the last 5 years (27%). Original papers were most common among publications (69%). Hospitals with highest output included La Princesa (24%) and Ramón y Cajal (17.6%) in Madrid, and Parc Taulí in Barcelona (6.4%). Mean impact factor progressively increased from 1.826 in 1988 to 2.142 in 2002 and 2.493 in 2003. CONCLUSIONS: the production and impact of documents published by Spanish scientists regarding H. pylori infection considerably increased during the past two decades.

Recurrence of Helicobacter pylori infection after several eradication therapies: long-term follow-up of 1000 patients.

AIM: To study the incidence of Helicobacter pylori recurrence, its chronological aspects, and the variables that might influence it. METHODS: A total of 1000 patients in whom H. pylori had been eradicated were prospectively studied. Therapies were classified as low and high efficacy regimens. Four to eight weeks after completion of therapy, 13C-urea-breath-test was performed, and it was repeated yearly up to 5 years. In some patients, endoscopy with biopsies was also performed to confirm H. pylori eradication. RESULTS: A total of 1000 patients were included, giving 2744 patient-years of follow-up. Seventy-one H. pylori recurrences were observed (2.6% per patient-year). Probability of being H. pylori-negative at 1 year was 94.7%, and at 5 years 90.7%. In the multivariate analysis, low age (OR: 1.84; 95% CI: 1.04-3.26) and low efficacy therapies (OR: 2.5; 1.23-5.04) correlated with 1-year H. pylori recurrence. Differences were observed when Kaplan-Meier curves were compared depending on age and therapy regimen. CONCLUSION: Risk of posteradication H. pylori recurrence is higher during the first year, which suggests that most recurrences during this period are recrudescence and not true reinfections. H. pylori recurrence is more frequent in younger patients and in those treated with low efficacy therapies, but is exceptional if high efficacy therapies are used, in which case post-therapy eradication can be safely confirmed at 4 weeks with 13C-urea-breath-test.

Long-term follow-up of 13C-urea breath test results after Helicobacter pylori eradication: frequency and significance of borderline delta13CO2 values.

BACKGROUND: The precise choice of cut-off point for the 13C-urea breath test to define whether it is positive or negative represents a controversial issue. AIM: To quantify the 13C-urea breath test result for several years following Helicobacter pylori eradication, and to evaluate the frequency and the significance of borderline delta13CO2 values. METHODS: Two-hundred H. pylori eradicated patients confirmed by 13C-urea breath test (100 mg of urea, citric acid), and having had repeated this test yearly up to 5 years, were studied. Delta13CO2 values between 2 and 5/1000 were considered as borderline results. RESULTS: Eight H. pylori recurrences were observed during 406 patient-years of follow-up (1.97% yearly). In two of eight reinfected patients, the reinfection was preceded by a negative delta13CO2 value >2/1000. Borderline delta13CO2 values were detected in 4% of the 606 urea breath tests performed, and in 25% when only patients in whom H. pylori recurrence was detected in subsequent urea breath tests were included (P < 0.05). The negative-predictive value of a post-treatment delta13CO2 >2/1000 for the diagnosis of H. pylori recurrence was 99%. CONCLUSIONS: Positive and negative urea breath test results tend to cluster outside the range between 2/1000 and 5/1000. Nevertheless, a borderline urea breath test delta value (e.g. very close to the selected cut-off point) should be interpreted cautiously, and the result should probably be confirmed either by repeating the urea breath test or by other diagnostic methods. On the contrary, a delta13CO2 value <2/1000 very confidently confirms H. pylori eradication.

Is there any correlation between 13C-urea breath test values and response to first-line and rescue Helicobacter pylori eradication therapies?

AIM: To study if there is a correlation between 13C-urea breath test values prior to treatment and the response to first-line and rescue Helicobacter pylori eradication therapies. METHODS: Six-hundred patients with peptic ulcer or functional dyspepsia infected by H. pylori were prospectively studied. Pre-treatment H. pylori infection was established by 13C-urea breath test. Three-hundred and twelve patients were treated with first-line eradication regimen, and 288 received a rescue regimen. H. pylori eradication was defined as a negative 13C-urea breath test, 8 weeks after completion of treatment. RESULTS: H. pylori eradication was achieved in 444 patients. No statistically significant differences were demonstrated when mean delta 13C-urea breath test values were compared between patients with eradication success and failure (49.4+/-33 versus 49.2+/-31). Differences in mean pre-treatment delta 13CO2 between patients with eradication success/failure were not demonstrated either when first-line or rescue regimens were prescribed. With the cut-off point of pre-treatment delta 13CO2 set at 35 units, sensitivity and specificity for the prediction of H. pylori eradication success was 43 and 60%. The area under the receiver operating characteristic curve evaluating all the cut-off points of the pre-treatment delta 13CO2 for the diagnosis of H. pylori eradication was 0.5. Finally, delta 13CO2 values did not influence the eradication in the logistic regression model. CONCLUSION: No correlation was observed between 13C-urea breath test values before treatment and the response to first-line and rescue H. pylori eradication therapies. Therefore, we conclude that the quantification of delta 13CO2 prior to treatment is not useful to predict the success or failure of eradicating therapy.

Evaluation of 3 different tests for the detection of stool antigens to confirm Helicobacter pylori eradication after treatment. A pilot study.

INTRODUCTION: Recently, several new diagnostic methods aimed to detect Helicobacter pylori stool antigens have been developed. Our aim was to evaluate the accuracy of 3 different stool tests to confirm H. pylori eradication. PATIENTS AND METHODS: Twenty-six patients received H. pylori eradication treatment. Eradication was confirmed with 13C-urea breath test 6-8 weeks later, when stool samples were analyzed by polyclonal (Premier-Platinum-HpSATM), monoclonal (Amplified-IDEIATM-HpStARTM), and rapid test (ImmunoCard-STAT-HpSATM). RESULTS: H. pylori was eradicated in 85% of the cases. Sensitivity, specificity, positive predictive value and negative predictive value with the polyclonal test were: 25%, 91%, 33% and 87%. Corresponding results with the monoclonal test, using the cut-off point recommended by the manufacturer, were 100%, 46%, 25% and 100%. However, the best cut-off point in our study had 100% sensitivity and 91% specificity. The area under ROC curve for the polyclonal and the monoclonal tests was 0.65 and 0.95. Diagnostic accuracy with the rapid test was 75%, 90%, 60% and 95%. CONCLUSION: Neither the polyclonal stool antigen test nor the rapid stool antigen test can be recommended to confirm H. pylori eradication after treatment. The monoclonal test has better diagnostic accuracy, although more studies are necessary to definitively recommend its use for the confirmation of H. pylori eradication success.

Helicobacter pylori first-line treatment and rescue options in patients allergic to penicillin.

BACKGROUND: Helicobacter pylori eradication is a challenge in patients allergic to penicillin, especially those who have failed a first-eradication trial. AIM: To assess the efficacy and tolerability of H. pylori first-line treatment and rescue options in patients allergic to penicillin. METHODS: Prospective single centre study including 40 consecutive treatments administered to patients allergic to penicillin. Therapy regimens: First-line (12 patients) omeprazole, clarithromycin and metronidazole for 7 days; second-line (17 patients) ranitidine bismuth citrate, tetracycline and metronidazole for 7 days; third-line (nine patients) rifabutin, clarithromycin and omeprazole for 10 days; and fourth-line (two patients) levofloxacin, clarithromycin and omeprazole for 10 days. OUTCOME VARIABLE: a negative (13)C-urea breath test 8 weeks after completion of treatment. RESULTS: Per-protocol/intention-to-treat eradication rates were: first-line (64/58%); second-line (ranitidine bismuth citrate; 53/47%); third-line (rifabutin; 17/11%) and fourth-line regimen (levofloxacin; 100/100%). Compliance with treatment was generally good, except with the rifabutin-based regimen, which presented adverse effects in 89% of the patients, including four cases of myelotoxicity. CONCLUSIONS: H. pylori-infected patients who are allergic to penicillin may be treated with a first-line treatment combining a proton-pump inhibitor, clarithromycin and metronidazole. Rescue options may include a regimen with ranitidine bismuth citrate, tetracycline and metronidazole. A levofloxacin-based rescue regimen (with proton-pump inhibitor and clarithromycin) may also represent an alternative, even when two or more consecutive eradication treatments have previously failed. However, rifabutin + clarithromycin + proton-pump inhibitor regimen is ineffective and poorly tolerated.

13C-urea breath test in the management of Helicobacterpylori infection.

The urea breath test is a noninvasive and very accurate test for the diagnosis of Helicobacter pylori infection. However, false negative urea breath test results have been reported to occur in a considerable percentage of the individuals taking proton pump inhibitors; the interval needed to be completely confident that false negative tests had been excluded has varied among the different studies between 6 and 14 days. The impact of H(2)-receptor antagonists on the accuracy of urea breath test remains controversial, although, in contrast with proton pump inhibitors, the data suggest that H(2)-receptor antagonists, for the most part, have little effect on the result of the urea breath test. The urea breath test does not represent a suitable tool for estimating the density of H. pylori colonization. The only quantitative information to be obtained from the urea breath test is that the higher the delta value, the lower the probability of a false-positive urea breath test result. Although some authors have demonstrated a correlation between urea breath test values and histological lesions of the gastric mucosa, the practical utility of this relationship remains unclear. It has been suggested that the pretreatment urea breath test has the potential to identify patients who require modification of the standard therapeutic regimen (for example, prolonging the duration of treatment or increasing the pharmacological dose when bacterial density is high), but other studies could not confirm this relationship. Some studies have shown that the urea breath test is less accurate in patients who have undergone partial gastrectomy. Finally, in contrast with biopsy-based methods, which are responsible for a high number of false-negative results when used to diagnose H. pylori infection in patients with upper gastrointestinal bleeding, urea breath test seems not to be negatively influenced by the presence of this complication.

Systematic review and meta-analysis: is 1-week proton pump inhibitor-based triple therapy sufficient to heal peptic ulcer?

AIMS: To systematically review the efficacy on ulcer healing of 1-week combination of a proton pump inhibitor plus two antibiotics and to perform a meta-analysis of randomized clinical trials to evaluate whether 7 days of proton pump inhibitor-based triple therapy is sufficient to heal peptic ulcer. METHODS: Studies where 1-week proton pump inhibitor-based triple therapy was administered to heal peptic ulcer were included. Randomized clinical trials comparing the efficacy on ulcer healing of 7-day proton pump inhibitor-based triple therapy versus this same regimen but prolonging the proton pump inhibitor for several more weeks were included in the meta-analysis. Electronic and manual bibliographical searches were conducted. Meta-analysis was performed combining the odds ratios of the individual studies. RESULTS: Twenty-four studies (2342 patients) assessed ulcer healing with 1-week proton pump inhibitor-based triple therapy. Mean healing rate was 86%, and 95% in Helicobacter pylori-eradicated patients. Six studies (862 patients), were included in the meta-analysis. Mean ulcer healing rate with a 7-day treatment was 91% versus 92% when proton pump inhibitor was prolonged for 2-4 more weeks (odds ratio = 1.11; 95% confidence interval = 0.71-1.74). CONCLUSION: In patients with peptic ulcer and H. pylori infection, prolonging therapy with proton pump inhibitor after a triple therapy for 7 days with a proton pump inhibitor and two antibiotics is not necessary to induce ulcer healing.

Systematic review: regression of lymphoproliferative disorders after treatment for hepatitis C infection.

AIM: To systematically review the experience of therapeutic studies where alpha-interferon with or without ribavirin was administered to patients with lymphoproliferative disorders, in order to evaluate whether eradication of hepatitis C virus may induce regression of lymphoproliferative disorders. METHODS: We used bibliographical searches in electronic databases and in the Cochrane Library to determine our results. RESULTS: Sixteen studies where an anti-viral regimen was administered to 65 hepatitis C virus-infected patients with lymphoproliferative disorders were identified. Complete remission of the lymphoproliferative disorder was achieved in 75% of the cases. In contrast, hepatitis C virus-negative subjects did not respond to interferon, indicating that the response in the hepatitis C virus-infected patients is not merely due to the antiproliferative effect of interferon. Remission after HCV eradication was maintained, provided that infection did not reappear. In hepatitis C virus-infected patients with non-Hodgkin's lymphoma treated with corticosteroids/chemotherapy liver function tests deterioration did not occur. The addition of interferon to standard chemotherapy may decrease hepatic side-effects of chemotherapy. CONCLUSIONS: Although it is evident that larger therapeutical trials of anti-viral therapy are needed to determine the role of this strategy in hepatitis C virus-infected patients with lymphoproliferative disorders, encouraging data emerge from recent studies showing that interferon (plus ribavirin) is an attractive therapeutic option for some hepatitis C virus-related low-grade lymphomas.

Review article: 13C-urea breath test in the diagnosis of Helicobacter pylori infection -- a critical review.

The urea breath test is a non-invasive, simple and safe test which provides excellent accuracy both for the initial diagnosis of Helicobacter pylori infection and for the confirmation of its eradication after treatment. Some studies have found no differences between urea breath test performed under non-fasting conditions. The simplicity, good tolerance and economy of the citric acid test meal probably make its systematic use advisable. The urea breath test protocol may be performed with relatively low doses (<100 mg) of urea: 75 mg or even 50 mg seem to be sufficient. With the most widely used protocol (with citric acid and 75 mg of urea), excellent accuracy is obtained when breath samples are collected as early as 10-15 min after urea ingestion. A unique and generally proposed cut-off level is not possible because it has to be adapted to different factors, such as the test meal, the dose and type of urea, or the pre-/post-treatment setting. Fortunately, because positive and negative urea breath test results tend to cluster outside of the range between 2 and 5 per thousand, a change in cut-off value within this range would be expected to have little effect on clinical accuracy of the test.

Esomeprazole-based therapy in Helicobacter pylori eradication: a meta-analysis.

AIM: To perform a systematic review on the efficacy of esomeprazole-based therapies in Helicobacter pylori eradication, and to conduct a meta-analysis comparing the efficacy of esomeprazole and other proton pump inhibitors when co-prescribed with antibiotics. METHODS: Studies evaluating esomeprazole plus antibiotics were considered. Only randomised trials comparing esomeprazole and other proton pump inhibitors with antibiotics, and differing only in the proton pump inhibitor, were included in the meta-analysis. Electronic and manual bibliographical searches were conducted. The percentage (weighted mean) of eradication success was calculated. Meta-analysis was performed combining the odd ratios of the individual studies. RESULTS: Mean cure rates with dual regimens (esomeprazole plus clarithromycin) were 51 and 54%, respectively, by intention-to-treat and by per-protocol. Corresponding figures with triple regimens (esomeprazole plus clarithromycin and either amoxicillin or metronidazole) were 82% (intention-to-treat) and 86% (per-protocol). Four studies were included in the meta-analysis: mean H. pylori eradication rates (intention-to-treat) with esomeprazole plus antibiotics was 85 and 82% when omeprazole was used (odds ratio = 1.19; 95% confidence interval = 0.81-1.74), results being statistically homogeneous. When subanalysis included only high quality studies, the odds ratio for this comparison was closer to one (1.08; 95% confidence interval = 0.4-1.47) and results were more homogeneous. CONCLUSIONS: Esomeprazole-based triple therapy is highly effective for the eradication of H. pylori infection and offers comparable efficacy to omeprazole-based therapy.

Evaluation of three different tests for the detection of stool antigens to diagnose Helicobacter pylori infection in patients with upper gastrointestinal bleeding.

AIM: To evaluate the accuracy of several methods aimed to detect Helicobacter pylori stool antigens in patients with upper gastrointestinal bleeding. METHODS: Thirty-four patients with upper gastrointestinal bleeding because of peptic ulcer were included. The first stool sample during hospitalization was collected, and stool antigens were determined with: polyclonal enzyme-linked immunosorbent assay (Premier-Platinum-HpSA); monoclonal enzyme-linked immunosorbent assay (Amplified-IDEIA-HpStAR); and rapid monoclonal immunochromatographic test (ImmunoCard-STAT HpSA). A patient was considered infected when H. pylori was diagnosed with invasive tests (rapid urease test or histology) or with (13)C-urea breath test. When all tests were negative, a new breath test was repeated after stopping proton pump inhibitors. RESULTS: All patients were infected and, therefore, only sensitivity of the tests could be calculated: polyclonal enzyme-linked immunosorbent assay (74%), monoclonal enzyme-linked immunosorbent assay (94%), and rapid monoclonal immunochromatographic test (60%; concordance between the two observers was high, kappa = 0.9). Neither the presence of maelena nor the delay in obtaining stool samples explained false negatives. CONCLUSIONS: Neither the polyclonal enzyme-linked immunosorbent assay stool antigen test nor the rapid immunochromatographic stool antigen test can be recommended to diagnose H. pylori infection in patients with upper gastrointestinal bleeding. However, the monoclonal enzyme-linked immunosorbent assay stool antigen test is highly sensitive for detecting the infection in patients with this complication, although more studies are necessary to evaluate the specificity of the method.

Helicobacter pylori and perforated peptic ulcer prevalence of the infection and role of non-steroidal anti-inflammatory drugs.

AIMS: To study the prevalence of Helicobacter pylori infection in patients with perforated peptic ulcer, to compare it with the prevalence in patients with uncomplicated ulcer, and to assess the role of non-steroidal anti-inflammatory drugs in this prevalence. METHODS: Consecutive patients with perforated peptic ulcer were included in this retrospective study. As a control group, patients undergoing elective outpatient evaluation for the investigation of dyspepsia during the same time period and found to have a peptic ulcer at endoscopy were included. A 13C-urea breath test was carried out in all patients to diagnose H. pylori infection. RESULTS: Sixteen patients with perforated peptic ulcer and 160 with non-complicated peptic ulcer were included. Sixty-two percent of the patients with perforated peptic ulcer were infected by H. pylori, while the microorganism was detected in 87% of the patients without this complication (P = 0.01). Non-steroidal anti-inflammatory drugs intake was more frequent (P = 0.012) in patients with perforated peptic ulcers (56%) than in those without perforation (26%). H. pylori prevalence in perforated peptic ulcers was of 44% in patients with non-steroidal anti-inflammatory drugs intake, but this figure increased up to 86% when only patients not taking non-steroidal anti-inflammatory drugs were considered (P = 0.09). In the multivariate analysis, non-steroidal anti-inflammatory drugs intake was the only variable that correlated with peptic ulcer perforation [odds ratio, 3.6 (95% confidence interval, 1.3-10); P = 0.016]. CONCLUSION: The mean prevalence of H. pylori infection in patients with perforated peptic ulcer is, overall, of only about 60%, which contrasts with the 90-100% figure usually reported in non-complicated ulcer disease. However, the most important factor associated with H. pylori-negative perforated peptic ulcer is non-steroidal anti-inflammatory drugs use, and if this factor is excluded, prevalence of infection is almost 90%, similar to that found in patients with non-perforating ulcer disease.

Empirical Helicobacter pylori "rescue" therapy after failure of two eradication treatments.

AIM: Even with the current most effective Helicobacter pylori treatment regimens, approximately 20% of patients do not eradicate the infection. Several "rescue" therapies have been recommended, but they still fail to eradicate H. pylori in approximately 20-30% of the cases. Our aim was to evaluate the efficacy of different rescue therapies prescribed to patients in whom two consecutive H. pylori eradication regimens had failed. METHODS: Design. Prospective single-centre study. Patients. Consecutive patients in whom two eradication regimens had failed to eradicate H. pylori. Intervention. Third eradication regimens included: (1) omeprazole-amoxicillin-clarithromycin for 7 days; (2) quadruple therapy with omeprazole-bismuth-tetracycline-metronidazole for 7 days; (3) omeprazole-amoxicillin-clarithromycin-bismuth for 14 days; and (4) omeprazole-amoxicillin-rifabutin for 14 days. H. pylori antibiotic susceptibility was unknown and, therefore, rescue regimens were chosen empirically. In no case, was the same regimen repeated. Outcome. H. pylori eradication was defined as a negative in 13C-urea breath test 8 weeks after completing the therapy. RESULTS: Forty-eight patients were included (mean age 45 years, 44% males, 82% with peptic ulcer and 18% with functional dyspepsia). No patient was lost from follow-up. Adverse effects were described in 21% of the patients. One patient receiving omeprazole, amoxicillin and rifabutin was removed from medication due to adverse effects (vomiting). Overall, mean H. pylori eradication with third therapy after failure of two eradication treatments was 34/48 (71%; 95% confidence interval 57-82%) by intention-to-treat and 34/47 (72%; 95% confidence interval 58-83%) by per-protocol. CONCLUSION: It seems that performing culture even after a second eradication failure may not be necessary, as it is possible to construct an overall strategy to maximise H. pylori eradication, based on the different possibilities of empirical treatment.