|
A correction has been published: N Engl J Med 1995;333(3):190.
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background There is a strong association between infection with Helicobacter pylori and gastric ulcers that are unrelated to the use of nonsteroidal antiinflammatory medications. We studied the efficacy of antibacterial therapy without medication to suppress gastric acid for the treatment of patients with H. pylori infection and gastric ulcers unrelated to the use of nonsteroidal agents.
Methods Patients with gastric ulcers seen on endoscopy and with H. pylori infection confirmed by smear or culture were randomly assigned to receive either a one-week course of antibacterial agents (120 mg of bismuth subcitrate, 500 mg of tetracycline, and 400 mg of metronidazole, each given orally four times a day) or a four-week course of omeprazole (20 mg orally per day). Follow-up endoscopies were performed after five and nine weeks. The patients and their physicians were aware of the treatment assignments, but the endoscopists were not.
Results A total of 100 patients were randomly assigned to treatment, and 85 completed the trial. At five weeks, H. pylori had been eradicated in 41 of the 45 patients in the antibacterial-treatment group (91.1 percent; 95 percent confidence interval, 82.9 to 99.3) and in 5 of the 40 in the omeprazole group (12.5 percent; 95 percent confidence interval, 2.3 to 22.7; P<0.001). The gastric ulcers were healed in 38 of the patients treated with antibacterial drugs (84.4 percent; 95 percent confidence interval, 73.9 to 95.0) and in 29 of those treated with omeprazole (72.5 percent; 95 percent confidence interval, 58.6 to 86.4; P = 0.28). At nine weeks, ulcer healing was confirmed in 43 of the patients receiving antibacterial therapy and in 37 of those receiving omeprazole (P = 1.0). The mean (±SD) duration of pain during the first week of treatment was 1.9±2.6 days in the omeprazole group, as compared with 3.6±3.0 days in the antibacterial-treatment group (P = 0.004). One year after treatment, recurrent gastric ulcers were detected in 1 of 22 patients (4.5 percent) in the antibacterial-treatment group and in 12 of 23 (52.2 percent) in the omeprazole group (P = 0.001). H. pylori was detected in the 1 patient with a recurrent ulcer who had received antibacterial treatment and in 10 of the 12 patients with recurrent ulcers who had received omeprazole.
Conclusions In patients with H. pylori infection and gastric ulcers unrelated to the use of nonsteroidal antiinflammatory drugs, one week of antibacterial therapy without acid suppression heals the ulcers as well as omeprazole and reduces the rate of their recurrence.
In a previous study, we found that duodenal ulcers healed as well after the eradication of H. pylori with antibacterial therapy as they did after treatment with both antibacterial therapy and medication to suppress gastric acid.7 In this study, we investigated the efficacy of antibacterial therapy for the treatment of H. pylori–associated gastric ulcers after excluding cases related to the use of nonsteroidal antiinflammatory drugs. Our hypotheses were that gastric ulcers not associated with these medications are causally related to H. pylori infection and that antibacterial therapy without additional medications can heal these ulcers and reduce the likelihood of their recurrence.
Methods
All patients presenting to the Prince of Wales Hospital with dyspepsia or epigastric pain who were found to have gastric ulcers by endoscopy were eligible for the study. Patients were excluded if they were under 16 years of age, had used nonsteroidal antiinflammatory drugs in the previous three months, had had gastrointestinal bleeding within the previous four weeks, had previously undergone surgery to reduce gastric acid, or had received antibacterial therapy in the past. A gastric ulcer was defined endoscopically as a breach of the mucosa with a well-defined ulcer crater. Antral biopsies as well as biopsies around the gastric ulcers were performed to test for urease activity (Campylobacter-Like Organism [CLO] test, Delta West, Western Australia), with Gram's-stain smears and cultures to detect H. pylori. Patients were considered to be positive for H. pylori if either the culture or the smear from either the antral biopsy or the biopsy around the ulcer was positive. A positive result on the CLO test alone was not considered a sufficient criterion for a patient to be classified as positive for H. pylori. Patients initially classified as positive for H. pylori on the basis of the CLO test were reclassified as negative if the result of the CLO test was not confirmed by positive smears or cultures. The size of the ulcer was measured with an endoscopic ruler during endoscopy.
Patients were randomly assigned to one of two treatment groups with the use of sealed envelopes that contained the assignments. Treatment assignments were determined with a list of random numbers generated by computer. Both the patients and their physicians were aware of the treatment assignments, but the endoscopists were not.
Patients were assigned to receive either a course of bismuth subcitrate (120 mg), tetracycline (500 mg), and metronidazole (400 mg), with each medication given orally four times per day for one week, or omeprazole (20 mg per day given orally) for four weeks. In all cases, treatment was started on the day of randomization. A patient was assigned to treatment when the CLO test was positive. If infection was not confirmed by a positive smear or culture or both, the patient was classified as negative for H. pylori and excluded from the study.
The patients were given diary cards to record symptoms during the five weeks after their assignment to treatment. After one week of treatment, all patients were interviewed by a research nurse about compliance with the treatment, side effects, and symptoms related to the ulcer. Patients assigned to the one-week course of antibacterial therapy were then given 60 antacid tablets (Mylanta) to be taken as required.
Five weeks after randomization (i.e., four weeks after the completion of antibacterial therapy), the patients returned their diary cards and those in the antibacterial-treatment group returned any remaining antacid tablets. Follow-up endoscopy was performed to check the healing of the ulcers, and antral biopsies were repeated to obtain specimens for the CLO test, smear, and culture. If the ulcers were not completely healed, patients in the antibacterial-treatment group received antacid as necessary, and patients in the omeprazole group received omeprazole (20 mg daily) for four more weeks. A second follow-up endoscopy was performed at nine weeks. After nine weeks, all patients with persistent gastric ulcers were given omeprazole (20 mg daily) for four more weeks.
The study was approved by the ethics committee of the faculty of medicine at the Chinese University of Hong Kong. Written informed consent was obtained from all patients before they were enrolled in the trial.
Microbiologic Studies
Gram's staining was performed on minced tissue from two antral-biopsy specimens to detect gram-negative spiral organisms. The minced tissue was cultured on Columbia agar (Oxoid, Basingstoke, United Kingdom) supplemented with 5 percent horse blood and incubated for five days under microaerophilic conditions.8 The presence of H. pylori was confirmed by morphologic analysis of the colony, Gram's staining,9,10 and biochemical tests (for oxidase, catalase, and urease).
Statistical Analysis
The results of treatment in the two groups of patients were compared by chi-square analysis. Ulcer healing and eradication of H. pylori in the two groups were compared with Yates' correction or Fisher's exact two-tailed test.11 The duration of symptoms was analyzed with the Mann–Whitney U test. Differences with a P value less than 0.05 were considered statistically significant.
Results
From May 1992 to April 1994, 100 patients with gastric ulcers associated with H. pylori infection were studied; 54 patients were randomly assigned to antibacterial therapy, and 46 to treatment with omeprazole. The two groups were well matched with respect to age, sex, history of smoking, and the site and size of the gastric ulcers (Table 1). Nine patients were excluded from the analysis: four assigned to receive antibacterial therapy (one each because of unconfirmed H. pylori infection, loss to follow-up, a concomitant duodenal ulcer, and previous therapy for helicobacter infection) and five assigned to receive omeprazole (one because of unconfirmed H. pylori infection and four because of loss to follow-up). Thus, the antibacterial-treatment group included 50 patients, and the omeprazole group 41.
|
Eighty-five patients (45 in the antibacterial-treatment group and 40 in the omeprazole group) completed the trial. The dropout rate (including patients lost to follow-up, those who could not tolerate the medication, and those who did not comply with the protocol) did not differ significantly between the antibacterial group (6 of 54 patients) and the omeprazole group (5 of 46, P = 0.78). Since acid-suppressing medications given to the patients who could not tolerate antibacterial therapy affected ulcer healing, only patients who completed the assigned treatment were included in the analysis. We also conducted an intention-to-treat analysis that included the three patients who failed to finish the antibacterial therapy but returned for follow-up and the one patient in the omeprazole group who took an H2-receptor antagonist.
Five weeks after randomization, eradication of H. pylori was documented in 41 of the 45 patients receiving antibacterial therapy (91.1 percent; 95 percent confidence interval, 82.9 to 99.3) and in 5 of the 40 patients treated with omeprazole (12.5 percent; 95 percent confidence interval, 2.3 to 22.7; P<0.001). Endoscopy at five weeks showed complete healing of the ulcers in 38 of the patients in the antibacterial-therapy group (84.4 percent; 95 percent confidence interval, 73.9 to 95.0) and 29 of those in the omeprazole group (72.5 percent; 95 percent confidence interval, 58.6 to 86.4; P = 0.28). One patient in the latter group was lost to follow-up after five weeks.
At nine weeks, the ulcers had completely healed in 43 of the 45 patients who received antibacterial therapy (95.6 percent; 95 percent confidence interval, 89.5 to 100) and in 37 of the 39 who received omeprazole (94.9 percent; 95 percent confidence interval, 88 to 100; P = 1.0). The four patients whose ulcers had not healed at 9 weeks were examined again at 13 weeks. The two patients in the antibacterial-treatment group had been given omeprazole for four weeks, and their ulcers were healed. One of them remained positive for H. pylori infection. One of the two patients in the omeprazole group had complete healing of the ulcer after a total of 13 weeks of omeprazole therapy; the other patient was referred for surgery because the ulcer remained unhealed.
The mean (±SD) duration of epigastric pain during the first week of treatment was longer in the group treated with antibacterial therapy (3.6±3.0 days) than in the group treated with omeprazole (1.9±2.6 days, P = 0.004). The duration of pain from the second to the fifth week after randomization was 7.6±9.5 days for those receiving antibacterial therapy and 4.4±8.7 days for those receiving omeprazole (P = 0.089) (Table 2).
|
As of August 1994, 45 patients had returned for the one-year follow-up evaluation. Recurrent gastric ulcers were detected in 1 of 22 patients in the antibacterial-treatment group (4.5 percent) and in 12 of 23 in the omeprazole group (52.2 percent, P = 0.001). H. pylori was detected in the 1 patient in the antibacterial-treatment group with a recurrent ulcer and in 10 of the 12 patients in the omeprazole group with recurrent ulcers (83.3 percent). In addition, duodenal ulcers had developed in two patients in the omeprazole group within one year after treatment; both were positive for H. pylori infection.
Discussion
Most gastric ulcers occur with chronic diffuse gastric inflammation.12 Clinical and histologic studies have suggested that H. pylori–related gastritis and gastric ulcer represent a continuum of progressive disease.13,14 A causal relation between H. pylori infection and the development of gastric ulcers would be supported if the bacterial infection were treated successfully with antibacterial agents alone.
Graham et al. reported that among patients with gastric ulcers who were treated with three antibacterial drugs and ranitidine, the incidence of recurrent ulcers was significantly reduced after the eradication of H. pylori.1 Labenz and Börsch used various combinations of omeprazole and antibiotics (amoxicillin, ciprofloxacin, and roxithromycin) to treat 83 people with H. pylori–associated gastric ulcers and found that the eradication of H. pylori enhanced the healing of the ulcers and reduced their recurrence.15 In both studies, patients received medications that suppressed acid production. The German Gastric Ulcer Study Group randomly assigned patients with gastric ulcers to receive omeprazole or three antibacterial drugs, including eight weeks of bismuth, and found that the eradication of H. pylori resulted in the healing of the ulcers.16 The Finnish Gastric Ulcer Study Group randomly assigned patients with gastric ulcers to receive bismuth (for 8 weeks) plus metronidazole (for 10 days), bismuth alone (for 10 days), or ranitidine alone (for 8 weeks)17 and concluded that the eradication of H. pylori improved the healing of the ulcers and prevented relapses. In both the German and the Finnish studies, however, the healing of ulcers in the groups receiving antibacterial therapy could be attributed in part to the eight-week course of bismuth.
We compared a one-week course of three antibacterial drugs but no medication to suppress acid with a four-week course of omeprazole, a proton-pump inhibitor used as the standard treatment for ulcers. Our findings in the antibacterial-treatment group establish a link between the eradication of H. pylori and ulcer healing, although the possibility of spontaneous healing in some instances cannot be excluded. There was a trend toward an increased rate of healing in the antibacterial-treatment group after five weeks. The rate of healing at 9 and 13 weeks was similar in the two groups. Our findings are very similar to those in our previous study of the use of antibacterial therapy for the treatment of H. pylori–associated duodenal ulcers.7 The two studies provide evidence in support of a causal relation between H. pylori infection and the development of gastric and duodenal ulcers, respectively.
The relief of symptoms was significantly more rapid during the first week of treatment with the use of a medication that suppressed acid production, as compared with antibacterial therapy alone. This difference, which did not persist after the first week of treatment, was probably due to the potent acid-suppressing effect of omeprazole and the gastrointestinal side effects of treatment with three antibacterial drugs. We cannot rule out the possibility of bias in the patients' reports of symptoms, since they were not blinded to treatment. The rate of recurrent ulcers at one year was significantly higher in the omeprazole group than in the antibacterial-treatment group. Most of the patients with recurrent ulcers, including the one patient in the antibacterial-treatment group, were positive for H. pylori infection. Our findings are consistent with those of previous studies in which patients with gastric ulcers were treated with medication directed against H. pylori. These studies found that H. pylori infection was the most important predictor of the recurrence of ulcers.1,13,14,15
We conclude that one week of antibacterial treatment (bismuth subcitrate, tetracycline, and metronidazole) without acid suppression heals H. pylori–associated gastric ulcers that are unrelated to the use of nonsteroidal antiinflammatory drugs and reduces the rate of recurrence of ulcers.
Supported by a grant (CUHK 29/93M) from the Research Grant Committee of Hong Kong.
Source Information
From the Departments of Medicine (J.J.Y.S., V.K.S.L., M.K.K.L.), Surgery (S.C.S.C., M.Y.Y., E.K.W.N., A.K.C.L.), and Microbiology (T.K.W.L., A.F.B.C.), Prince of Wales Hospital and Chinese University of Hong Kong, Hong Kong.
Address reprint requests to Dr. Sung at the Department of Medicine, Prince of Wales Hospital, Shatin, Hong Kong.
References
| |||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Antibacterial Treatment of Gastric Ulcers
Graham D. Y., Go M. F., Faigel D. O., Metz D. C., Salyards H. E., Harris A., Misiewicz J.J., Fagan T., Sung J. J.Y., Chung S.C. S., Yung M. Y.
Extract |
Full Text
N Engl J Med 1995;
333:190-192, Jul 20, 1995.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
