Bismuth-based quadruple therapy with bismuth subcitrate , metronidazole , tetracycline and omeprazole in the eradication of Helicobacter pylori

1Centre Hospitalier de l’Université de Montréal, Hôpital St-Luc, Montréal; 2Centre de Gastro-Entérologie et d’Endoscopie de Montréal, Montréal; 3Centre Hospitalier St François d’Assisse, Québec; 4Hôpital Maisonneuve-Rosemont, Montréal; and 5McGill University Health Center, Montreal, Quebec; 6London Health Science Centre, Victoria Campus, London; 7McMaster University, Hamilton; and 8Oshawa Clinic, Oshawa, Ontario; 9Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia; 10Wellesley Hospital, Toronto, Ontario; 11Hys Center, Edmonton, Alberta; 12Cirion Biopharma Research Inc, Laval; 13Axcan Pharma Inc, Mont-Saint-Hilaire; and 14Faculé de Médicine, Department de Pharmacologie, Université de Montréal, Montréal, Québec Correspondance and reprints: Dr Raymond Lahaie, Division of Gastroenterology, Centre Hospitalier Université de Montréal, Hôpital St-Luc, 1058 St-Denis, Montréal, Quebec H2X 3J4. Telephone 514-281-2121, fax 514-281-6135, e-mail lahaier@odyssee.net Received for publication February 6, 2001. Accepted July 17, 2001 R Lahaie, A Farley, C Dallaire, et al. Bismuth-based quadruple therapy with bismuth subcitrate, metronidazole, tetracycline and omeprazole in the eradication of Helicobacter pylori. Can J Gastroenterol 2001;15(9):581-585.

R eports from the Canadian Helicobacter pylori Consensus Conference (1) and from the Ad Hoc Committee on Practice Parameters of the American College of Gastroenterology (2) advocate testing and eradicating H pylori in patients presenting with gastroduodenal ulcers.However, results of a survey indicated that 43% to 66% of American physicians also eradicate H pylori in patients with nonulcer dyspepsia (3).Over the years, many treatments have been tested for the eradication of H pylori.Among them, the socalled bismuth-based triple-therapy with colloidal bismuth subcitrate, metronidazole and tetracycline (BMT) has been extensively tested with success (4).
In a previous trial (5), a 14-day treatment with colloidal bismuth subcitrate 120 mg qid plus metronidazole 250 mg qid plus tetracycline 500 mg qid gave very good eradication rates of 82% and 90% by modified intent-to-treat (MITT) and per protocol analyses, respectively.The influence of metronidazole sensitivity on the eradication rate was, however, not assessed.
In an effort to increase the efficacy of the regimen, a proton pump inhibitor was added (6)(7)(8)(9).Also, the treatment duration was reduced to seven days, and the amount of tetracycline was lowered to 1 g daily to increase gastrointestinal tolerance.This study, therefore, was designed to assess the efficacy and safety of a seven-day treatment for the eradication of H pylori with colloidal bismuth subcitrate 120 mg qid plus metronidazole 250 mg qid plus tetracycline 250 mg qid, all given 1 h before meals and at bedtime, in addition to omeprazole 20 mg bid with the morning and evening meals.

PATIENTS AND METHODS
This was an open-label, multicentre (10) study of H pyloripositive patients with or without a history of peptic ulcer.The local ethics committee of each of the participating centres approved the study.Males and females, aged 18 to 75 years, found positive for H pylori by both 13 carbon urea breath test and histology at entry were eligible after giving informed written consent.
The following main exclusion criteria were applied: macroscopic esophagitis, previous gastric surgery, dysphagia, vomiting, hematemesis, melena, recent documented gastrointestinal bleeding, iron-deficiency anemia, inability to abstain from alcohol, significantly impaired renal or hepatic function, contraindication to the use of bismuth, metronidazole or tetracycline, chronic use of nonsteroidal antiinflammatory drugs, use of antibiotics within 30 days before enrollment, regular use of bismuth compounds in the past 30 days, a previous attempt to eradicate H pylori infection and use of antiulcer drugs (including H 2 receptor antagonists [seven days] or proton pump inhibitors [30 days]) preceding enrollment.
Upon confirmation of willingness to participate, routine assessment of health status and esophagogastroduodenoendoscopy were performed.The patient also underwent a 13 carbon urea breath test (Dia-13 Helico, Dianatec-Iso, Canada).
During endoscopy, three biopsies were taken from the antrum and two from the body.One antrum biopsy was used on the site for rapid urease test.One antrum biopsy and one corpus biopsy were sent to the central study pathologists for the detection of H pylori after staining with Wharthin-Starry solution.One antrum biopsy and one corpus biopsy were placed in Stuart's transportation medium and sent on dry ice to the central microbiologist for assessment of metronidazole sensitivity by E-test.A cutoff value of 8 µg/L was used for the minimum inhibitory concentration to classify sensitivity.
If H pylori was detected at screening by rapid urease test, the patient was allowed to start the treatment immediately.This result had to be later confirmed both by 13 carbon urea breath test and by histology (only four patients were later found to be H pylori negative by 13 carbon urea breath test and excluded from analysis).Patients self-administered the study drugs, provided in a blister pack.They were instructed not to take milk or other dairy products, or antacids within 2 h of taking the study medications.
All medications for dyspepsia were prohibited throughout the study as well as nonsteroidal anti-inflammatory drugs and acetylsalicylic acid.The occasional use of acetaminophen was, however, permitted, and antacids were allowed as rescue medication if the dyspepsia symptoms were severe.
All patients were, a priori, advised to refrain from alcohol during the seven-day treatment period.Female patients using oral contraceptives were informed about the risk of interaction between tetracycline and oral contraceptives, and advised to use an additional means of contraception.Patients were also warned to avoid exposure to direct sunlight and/or ultraviolet light during the seven-day treatment period because of the photosensitizing effect of tetracycline.
The rate of eradication probability was estimated, and the 95% standard confidence interval was computed.The primary analysis was done on the MITT population comprising all patients with positive 13 carbon urea breath test and histology at entry, who were entered in compliance with the inclusion and exclusion criteria, and had taken at least one dose of the study medication.A secondary analysis was done on the per protocol population comprising all patients in the MITT group for whom 13 carbon urea breath test results at one and three months were available, and who did not have any major protocol violation.
Stratification was done a posteriori by sensitivity or resistance to metronidazole.Eradication rate and 95% confidence intervals were computed for each subgroup of patients.These subgroups were compared by the likelihood ratio test.

RESULTS
Two hundred thirty-eight patients were screened for the study.Sixty-eight were found to be H pylori negative at entry, four withdrew before starting the medication, one was included in violation of the upper age limit and was discontinued, and four were included despite the presence of a concomitant condition precluding participation (gastrointestinal bleeding, thyroid disease and elevated transaminase levels at baseline) and were discontinued, leaving 161 patients in the MITT population.Fifteen patients were subsequently withdrawn (six protocol violations, six lost to follow-up, one adverse event and two voluntary withdrawals), leaving 146 patients in the per protocol population.Their demographics are presented in Table 1.
H pylori was successfully eradicated in 128 patients in the MITT population, for a rate of 80%, and in 123 patients in the per protocol population, for a rate of 84% (Table 2).The individual eradication rates by study sites varied between 40% and 100%, and 67% and 100%, in the MITT and per protocol populations, respectively (Figure 1).
A sample size of 140 patients was estimated a priori for the 95% confidence interval based on the assumptions that expected eradication rate of BMT plus omeprazole (OBMT) would be 90%, alpha-error 5% and length of the confidence interval 95%.With an eradication rate of 80%, 161 patients give the power to compute the 94% confidence interval.

BMT plus omeprazole for H pylori eradication
Can J Gastroenterol Vol 15 No 9 September 2001 583

Figure 1) Individual Helicobacter pylori eradication rates in the modified intent-to-treat (MITT) and per protocol (PP) study populations
The metronidazole sensitivity was successfully documented in strains from 90 patients and showed an average resistance rate of 30.0%.The results by city and site are presented in Table 3. Eighteen of 27 resistant strains (66.7%) and 52 of 63 sensitive strains (82.5%) were eradicated in the MITT population (P=0.105).Seventeen of 24 resistant strains (70.8%) and 51 of 57 sensitive strains (89.55%) were eradicated in the per protocol population (P=0.045).Detailed results are shown in Table 2.
One patient withdrew due to adverse events after four days of treatment.She complained of severe epigastric pain and moderate diarrhea that were classified as "possibly related" to study drugs by the investigator.She also complained of mild peridental bleeding that was classified as "unlikely related" to study drugs by the investigators.All events resolved.

DISCUSSION
The overall eradication rate by MITT analysis was 80%.The rates varied greatly from site to site, and this difference is likely explained, in part, by the small sample size seen in some investigative sites.If the site where only three patients were included in the per protocol analysis is not considered, the range becomes 66% to 100%.
In the MITT population, the combination of colloidal bismuth subcitrate 120 mg plus metronidazole 250 mg plus tetracycline 250 mg qid given with omeprazole 20 mg bid successfully eradicated H pylori in 82.5% and 66.7% of metronidazole-sensitive and metronidazole-resistant strains, respectively.Compared with the previously reported values of 88.5% and 69% in a survey of 1639 patients treated with a bismuth-based regimen (11), the actual rates were a little lower in sensitive strains and as low as those in resistant ones.The regimen used in the present study is, therefore, effective and safe for the eradication of metronidazole-sensitive H pylori strains, but is less effective than the 14-day triple therapy with 2 g daily of tetracycline given without a proton pump inhibitor (5).
In the present study, the overall prevalence of resistance to metronidazole in Canada was 30%.It varies greatly from site to site, and part of this variation is likely explained by the small sample sizes in some of the investigative sites.Nonetheless, this 30% figure compares favourably with the previously reported prevalence of 33% (10).The differences between the eradication rates in metronidazole-sensitive and metronidazole-resistant strains were statistically significant for the per protocol population but not for the MITT protocol; this is likely due to a type 2 error in the latter group.Actually, these tests were exploratory, and the sample size was not estimated based on these comparisons.
Three of the most often reported adverse events were expected, because taste disturbance and nausea are known adverse effects of metronidazole, and diarrhea is a known adverse effect of tetracycline.Adverse event was the reason to discontinue treatment in only one patient.They were limited in time to the period of treatment.Moreover, no serious adverse event or clinically significant laboratory abnormalities were reported in the trial.
The efficacy of this regimen in metronidazole-resistant strains remains a limiting step to its use and should be improved.Published preliminary results have shown more promising efficacy with a new 10-day treatment with a single triple-capsule containing colloidal bismuth subcitrate 120 mg plus metronidazole 375 mg plus tetracycline 375 mg given qid with omeprazole 20 mg given with the morning and evening meals (12).Updated results (data on file) with this capsule have shown, in 125 patients with active or history of duodenal ulcer, an MITT eradication rate of 91.9% in metronidazole-sensitive and 80.4% in metronidazoleresistant strains respectively.

CONCLUSIONS
We conclude that the OBMT regimen used in the present study is effective against metronidazole-sensitive H pylori strains.However, longer duration of treatment (10 days) with higher doses of metronidazole and tetracycline (1.5 g daily each) seems better as suggested by the preliminary results of a more recently completed study (12).

TABLE 1 Demographics of the modified intent-to-treat (MITT) and per protocol study populations
*A given patient may be in both groups