- Email: [email protected]
The effect of antibiotic therapy on bleeding from duodenal ulcer
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 1999 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 94, No. 4, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00047-7
The Effect of Antibiotic Therapy on Bleeding From Duodenal Ulcer Amnon Sonnenberg, M.D., M.Sc., Carol A. Olson, M.D., and Jie Zhang, Ph.D. Department of Veterans Affairs Medical Center and The University of New Mexico, Albuquerque, New Mexico; and Abbott Laboratories, Abbott Park, Illinois
OBJECTIVE: We conducted this study to test whether eradication of Helicobacter pylori (H. pylori) infection prevents hemorrhage related to duodenal ulcer. METHODS: Patients with H. pylori infection and endoscopically proven duodenal ulcers without ulcer-related hemorrhage were enrolled into four randomized, double-blind, multicenter studies using the same study protocol. They were treated with clarithromycin plus omeprazole (441 patients), omeprazole alone (447 patients), or ranitidine alone (263 patients). Success of H. pylori eradication was evaluated by the 13C-urea breath test 4 – 6 wk after the last dose of study drug. Follow-up continued at monthly intervals up to 1 yr after the last dose of study drug. RESULTS: Bleeding due to duodenal ulcer was not observed in any patients who received clarithromycin plus omeprazole, whereas five patients in the omeprazole treatment group and six patients in the ranitidine treatment group experienced an episode of ulcer-related hemorrhage during follow-up. All patients who experienced ulcer-related bleeding were male. When compared by bleeding, there were no significant differences with respect to ethnicity, alcohol consumption, or tobacco use. H. pylori infection was no longer detectable in 68% of patients after treatment with clarithromycin plus omeprazole, compared with 5% after treatment with omeprazole alone or 4% after treatment with ranitidine alone. CONCLUSION: In a population of duodenal ulcer patients without predisposing risk factors for ulcer bleeding, antibiotic eradication or suppression of H. pylori infection prevented the occurrence of ulcer-related hemorrhage for up to 1 yr after therapy. (Am J Gastroenterol 1999;94:950 –954. © 1999 by Am. Coll. of Gastroenterology)
have all been reported to be significant risk factors for ulcer-related hemorrhage (3, 4). The likelihood of recurrent hemorrhage is also increased in subjects with a previous history of ulcer bleeding (5). There is now clear evidence that H. pylori infection is present in .90% of the patients with duodenal ulcer (6), and that cure of the infection leads to dramatic reduction in ulcer recurrence (7). As ulcer recurrence is requisite for bleeding to occur, one is inclined to hypothesize that effective treatment of H. pylori infection should provide protection against the morbidity and mortality associated with ulcerrelated hemorrhage. Along with the alternatives of maintenance treatment with antisecretory agents or surgery, antibiotic treatment of H. pylori infection may decrease the frequency of ulcer recurrence and thus decrease the chances of ulcer-related hemorrhage in ulcer patients. Several previous studies suggested that successful eradication of the H. pylori infection decreased the incidence of gastrointestinal hemorrhage in peptic ulcer patients, compared with unsuccessful antibiotic therapy of the H. pylori infection or maintenance therapy with antisecretory agents (8 –13). However, there have been no reports of large, randomized clinical trials evaluating the effect of curing H. pylori infection on subsequent ulcer-related hemorrhage. This article reports results compiled from four large, randomized, multicenter studies carried out in the United States. It includes patients with proven duodenal ulcer but without a history of gastrointestinal hemorrhage. In this respect, the results of our study may be more reflective of the outcome to be expected in a general population of ulcer patients who receive primary treatment for an uncomplicated duodenal ulcer.
MATERIALS AND METHODS INTRODUCTION Peptic ulcer is the most common cause of acute hemorrhage in the upper gastrointestinal tract, accounting for about 50% of all cases (1). The overall mortality for patients with bleeding ulcers is approximately 6 –7% (2). Many factors appear to contribute to the occurrence of bleeding ulcers: Helicobacter pylori (H. pylori) infection, use of nonsteroidal antiinflammatory drugs (NSAIDs), and increased age
Between 1993 and 1995, four studies were carried out by Abbott Laboratories to assess the efficacy of a dual therapy of clarithromycin plus omeprazole in eradicating H. pylori and preventing the recurrence of duodenal ulcer. Although the four studies were conducted in sequence, an identical study protocol was used for all four studies. The same investigators and study coordinators were involved in monitoring the study sites, accumulating patient records, and
AJG – April, 1999
Bleeding Duodenal Ulcer
951
Table 1. Summary of Individual Clinical Trials Number of Patients per Treatment Arm Study Number
Number of Sites
Clarithromycin 1 Omeprazole
Omeprazole
1 2 3 4
36 37 64 68
82 82 137 140
80 88 135 144
analyzing the data. For these reasons, the cumulative data may be regarded as if it were generated in one large study. Bleeding from duodenal ulcer is a relatively rare event, especially in patients without a previous bleeding history. To achieve statistical power in comparing the influence of different treatment types on ulcer bleeding, for the present analysis we have compiled these previously unpublished data from the four studies and analyzed them together. Patients were enrolled into a randomized double-blind multicenter clinical trial at 205 sites distributed throughout the United States. The data, which form the basis of the present analysis, are summarized in Table 1. A dual combination of clarithromycin plus omeprazole was compared with omeprazole given alone in all four studies. The third and fourth study included a third treatment arm of ranitidine given alone. Patients received clarithromycin 500 mg t.i.d. plus omeprazole 40 mg q.d. for 14 days, followed by omeprazole 20 mg q.d. for an additional 14 days; or omeprazole 40 mg q.d. for 14 days, followed by 20 mg q.d. for an additional 14 days; or ranitidine 150 mg b.i.d. for 28 days. A total of 1151 patients were enrolled in the three treatment groups shown in Table 1. Male or female adults aged $18 yr with a single, active duodenal ulcer and H. pylori infection were eligible for enrollment. The duodenal ulcer was diagnosed by upper gastrointestinal endoscopy. H. pylori infection was diagnosed by a positive CLOtest, histology, culture, 13C-urea breath test, or a combination of these tests. Results were considered undefined if subjects underwent the test but no definitive result was reported by the laboratory. Such instances occurred, for instance, if biopsies were too small for a definitive histological evaluation or no 13 C was detectable in the air sample. Patients with any of the following characteristics were excluded from the studies: bleeding or perforated ulcer at the time of enrollment; evidence of gastric ulcer or gastric malignancy; any history of gastric surgery, gastric outlet obstruction, esophageal stricture, or chronic NSAID consumption; evidence of hepatic and renal impairment; evidence of drug or alcohol abuse; and hospitalization at the time of enrollment. The studies were carried out in accordance with regulations by the Food and Drug Administrations and local regulations and statutes governing study conduct at individual study sites. Approval by institutional review boards was obtained at each participating study site. All patients gave written informed consent before inclusion in the studies. Evaluations were carried out within 3 days before first
Ranitidine
Follow-up Time 6 6 12 12
133 130
months months months months
drug administration and within 5 days after the last dose of study drug. Follow-up continued at monthly intervals up to 1 yr after the last dose of study drug. Duodenal ulcer and H. pylori status were determined at the pretreatment, posttreatment, and final follow-up visit. During follow-up, ulcer relapse was assessed by the recurrence or new onset of symptoms. All episodes of ulcer-related hemorrhage were confirmed by upper gastrointestinal endoscopy. Success of H. pylori eradication was evaluated by the 13C-urea breath test 4 – 6 wk after the last dose of study drug. If both 13 C-urea breath test and histology results were available, both were required to be negative to consider the response to therapy as successful. For the purposes of the present analysis, clinical failure was defined as an endoscopically proven ulcer-related hemorrhage. All subjects enrolled into the study were included in the analysis. The statistical differences among the three treatment groups were compared by x2 tests with Yates correction. Pairwise comparisons for treatment effect on ulcer-related hemorrhage were carried out using Fisher’s exact test. The time until ulcer-related hemorrhage was analyzed using Kaplan-Meier survival analysis.
RESULTS A total of 1151 patients were enrolled into the study comparing three treatments for cure of duodenal ulcer disease. Premature discontinuation of the study drug occurred in 35 (7.9%) patients taking clarithromycin plus omeprazole, 25 (5.6%) patients taking omeprazole alone, and 16 (6.3%) patients taking ranitidine alone (Table 2). All subjects initially enrolled into the study were considered in the present Table 2. Reasons for Premature Discontinuation of the Study Total Enrolled Total enrolled Adverse event Noncompliance Lost to follow-up Withdrawn from study* Sum of premature discontinuation
Clarithromycin 1 Omeprazole Omeprazole Ranitidine 441 13 8 7 7
447 8 5 6 6
253 6 0 3 7
35
25
16
* Patients were withdrawn by their physicians from the study because of worsening of signs and symptoms, lack of response, or because patients had moved to a new location.
952
Sonnenberg et al.
AJG – Vol. 94, No. 4, 1999
Table 3. Bleeding by Treatment Group
Treatment
No DU Bleeding (n 5 1140)
Clarithromycin 1 Omeprazole Omeprazole Ranitidine
441 442 257
DU Bleeding p (n 5 11) Value 0 5 6
0.0013
DU 5 duodenal ulcer.
intention-to-treat analysis, and premature discontinuation of study drug was not a reason for exclusion. Patients enrolled in the three treatment groups were comparable with respect to age, gender, ethnicity, alcohol consumption, and tobacco use. No statistical differences were observed among the patient characteristics in the three treatment groups. Eleven of 1151 duodenal ulcer patients developed ulcer-related bleeding. When patient characteristics were compared according to bleeding outcome, there were no significant differences with respect to age, ethnicity, alcohol consumption, or tobacco use. However, all 11 patients who experienced ulcer-related bleeding were male, as compared with 790 of 1140 patients without bleeding (p 5 0.0610). As shown in Table 3, episodes of ulcer-related hemorrhage occurred during the follow-up period only in patients who had been treated initially with omeprazole or ranitidine alone. The statistical comparison revealed a significant treatment effect. Pairwise comparison between each two treatment groups, using Fisher exact test, showed less ulcerrelated bleeding after clarithromycin plus omeprazole than after ranitidine alone (p 5 0.003) or omeprazole alone (p 5 0.062). Figure 1 compares the time until ulcer-related hemorrhage in the three treatment groups. The p-values from the log-rank test are both 0.002 for comparing clarithromycin plus omeprazole versus omeprazole alone or versus ranitidine alone. As shown by the figure, all ulcer-related hemorrhages occurred within the first 6 months after treatment of the acute duodenal ulcer. This pattern excludes a potential bias that might have been introduced by the longer observation periods of patients followed-up after treatment with ranitidine alone (Table 1). To assess the effect of treatment on H. pylori status, we excluded patients from the analysis whose H. pylori status was not defined before or after treatment (Table 4). Among patients with undefined H. pylori status, three taking omeprazole and two taking ranitidine experienced ulcer-related bleeding. The other patients with ulcer-related hemorrhage
Figure 1. Kaplan-Meier analysis of ulcer bleeding. The x-axis depicts the time until ulcer-related bleeding, the y-axis the probability of nonbleeding. Each curve represents one of the three treatment arms. The differences between the bleeding probabilities of clarithromycin plus omeprazole vs omeprazole alone or vs ranitidine alone were both significant at p 5 0.002.
(two in the omeprazole treatment group and four in the ranitidine treatment group) had remained H. pylori positive. Among the 866 patients available for analysis, H. pylori infection was no longer detectable in 68% after treatment with clarithromycin plus omeprazole, compared with 5% after treatment with omeprazole alone and 4% after treatment with ranitidine alone (Table 5).
DISCUSSION In the present study, we analyzed the occurrence of ulcerrelated bleeding among patients who participated in four randomized, double-blind, multicenter trials, which evaluated the efficacy of clarithromycin plus omeprazole, omeprazole alone, and ranitidine alone in the treatment of H. pylori infection. None of the patients treated with clarithromycin plus omeprazole experienced episodes of ulcer-related bleeding during the follow-up period. Ulcer-related bleeding was significantly more frequent among patients who had received omeprazole alone or ranitidine alone. Previous studies established the presence of H. pylori infection in .90% of patients with duodenal ulcer (1), and showed that there is a dramatic decrease in ulcer recurrence after successful treatment of the infection (2). Other reports suggested that antibiotic treatment is more effective than maintenance therapy with ranitidine in preventing ulcer complications, particularly ulcer-related hemorrhage (8 –13). However, these studies were restricted to relatively few patients who presented initially with bleeding gastric
Table 4. Patients Available for Analysis by H. pylori Status Excluded From Analysis by Helicobacter pylori Status Treatment Group
Total No. of Patients in Treatment Group
Not Defined Pretreatment
Not Defined Posttreatment
Total No. of Patients Excluded
Clarithromycin 1 Omeprazole Omeprazole Ranitidine
441 447 263
28 33 17
83 89 35
111 122 52
AJG – April, 1999
Bleeding Duodenal Ulcer
Table 5. H. pylori Eradication Rates Treatment Group Clarithromycin 1 Omeprazole Omeprazole Ranitidine
H. pylori Negative
H. pylori Positive
Percent Eradication
223
107
68%
15 9
310 202
5% 4%
953
In summary, in a population of duodenal ulcer patients without predisposing risk factors for ulcer bleeding, antibiotic eradication or suppression of H. pylori infection prevented the occurrence of ulcer-related hemorrhage for up to 1 yr after therapy.
ACKNOWLEDGMENTS and duodenal ulcers. The results of the present analysis with .1000 patients from controlled clinical trials support and extend the results of the previous studies. In the majority of ulcer patients, ulcer-related hemorrhage represents a rare complication. Our study yields an incidence rate of 1.55%/yr among the 710 patients treated with antisecretory medication alone. This rate is lower than the rates available from the literature for patients with a previous history of ulcer bleeding (5, 14). Nevertheless, the study has the statistical power to show that even in patients without previous ulcer-related hemorrhage, antibiotic eradication of H. pylori carries the benefit of preventing such rare complication. It is mostly the occurrence of such rare but costly complications that results in the substantial economic burden associated with peptic ulcer disease (15). The H. pylori eradication rate for clarithromycin plus omeprazole reported here compares favorably to the results previously reported for individual clinical trials (16 –19). Dual therapy was considered state-of-the-art treatment at the time the studies were initiated. By adding metronidazole or amoxicillin to the dual therapy, newer triple regimens may achieve eradication rates in excess of 90% (20, 21). Despite the lower eradication rates obtained with dual therapy, the absence of bleeding episodes in patients treated with clarithromycin plus omeprazole suggests that antibiotic therapy directed against H. pylori infection is clinically effective. Over a time period of 6 –12 months, the beneficial effect of antibiotic therapy may apply even to patients in whom eradication was not successful. This suggests that suppression rather than eradication of H. pylori is already sufficient to provide some degree of protection against ulcer-related hemorrhage. In addition to H. pylori infection, other factors have been reported to contribute to the occurrence of ulcer-related hemorrhage in peptic ulcer patients. In particular, age, use of NSAIDs, previous bleeding episodes, and presence of gastric versus duodenal may all be relevant to the occurrence of bleeding episodes (3–5). Because in the present study we excluded patients with NSAID use, gastric ulcer, or ulcerrelated hemorrhage, the influence of these factors could not be assessed. Within our patient population, consumption of alcohol and tobacco was not associated with an increased occurrence of ulcer-related hemorrhage. Based on the results presented here, H. pylori infection appears to be the primary influence on the frequency of ulcer-related bleeding episodes. However, it is possible that such factors as alcohol and tobacco use act in concert with H. pylori infection in increasing the likelihood of ulcer-related hemorrhage.
Dr. Sonnenberg’s research was funded through a grant by the Centers for Disease Control and Prevention, Atlanta, Georgia. This work was conducted while Dr. Sonnenberg was employed by the Department of Veterans Affairs. The clinical study was supported by Abbott Laboratories, Abbott Park, Illinois. We are indebted to Larry Hancock for his assistance in writing the manuscript. Reprint requests and correspondence: Amnon Sonnenberg, M.D., M.Sc., Department of Veterans Affairs Medical Center 111F, 2100 Ridgecrest Drive SE, Albuquerque, NM 87108. Received July 9, 1998; accepted Nov. 30, 1998.
REFERENCES 1. Laine L, Peterson WL. Bleeding peptic ulcer. N Eng J Med 1996;331:717–27. 2. Silverstein FE, Gilbert DA, Tedesco FJ, et al. The national ASGE survey on upper gastrointestinal bleeding. II. Clinical prognostic factors. Gastrointest Endosc 1981;27:80 –93. 3. Van Deventer GM, Elashoff JD, Reedy TJ, et al. A randomized study of maintenance therapy with ranitidine to prevent the recurrence of duodenal ulcer. N Engl J Med 1989;320: 1113–9. 4. Taha AS, Russell RI. Helicobacter pylori and non-steroidal anti-inflammatory drugs: Uncomfortable partners in peptic ulcer disease. Gut 1993;34:580 –3. 5. Elashoff JD, Van Deventer G, Reedy TJ, et al. Long-term follow-up of duodenal ulcer patients. J Clin Gastroenterol 1983;5:509 –15. 6. Consensus Conference. Helicobacter pylori in peptic ulcer disease. J Am Med Assoc 1994;272:65–9. 7. Hopkins RJ, Girardi LS, Turney EA. Relationship between Helicobacter pylori eradication and reduced duodenal and gastric ulcer recurrence. Gastroenterology 1996;110:1244 –52. 8. Graham DY, Hepps KS, Ramirez FC, et al. Treatment of Helicobacter pylori reduces the rate of rebleeding in peptic ulcer disease. Scand J Gastroenterol 1993;28:939 – 42. 9. Labenz J, Bo¨rsch G. Role of Helicobacter pylori eradication in the prevention of peptic ulcer bleeding relapse. Digestion 1994;55:19 –23. 10. Rokkas T, Karameris A, Mavrogeogis A, et al. Eradication of Helicobacter pylori reduces the possibility of rebleeding in peptic ulcer disease. Gastrointest Endosc 1995;41:1– 4. 11. Jaspersen D, Koerner T, Schorr W, et al. Helicobacter pylori eradication reduces the rate of rebleeding in ulcer haemorrhage. Gastrointest Endosc 1995;41:5–7. 12. Powell KU, Bell GD, Bolton GH, et al. Helicobacter pylori eradication in patients with peptic ulcer disease: Clinical consequences and financial implications. Quat J Med 1994;87:283–90. 13. Santander C, Gravalos RG, Gomez-Cedenilla A, et al. Antimicrobial therapy for Helicobacter pylori infection versus long-term maintenance antisecretion treatment in the prevention of recurrent hemorrhage from peptic ulcer: Prospective
954
14. 15. 16.
17.
Sonnenberg et al.
nonrandomized trial on 125 patients. Am J Gastroenterol 1996;91:1549 –52. Pulvertaft CN. Comments on the incidence and natural history of gastric and duodenal ulcer. Postgrad Med J 1968;44:597–602. Sonnenberg A, Everhart JE. Health impact of peptic ulcer in the United States. Am J Gastroenterol 1997;92:614 –20. Hunt R, Schwartz H, Fitch D, et al. Dual therapy of clarithromycin and omeprazole for treatment of patients with duodenal ulcers associated with H. pylori infection. Gut 1995;37(suppl 1):A5. Logan RPH, Bardhan KD, Celestin LR, et al. Eradication of H. pylori and prevention of recurrence of duodenal ulcer: A randomized, double-blind, multi-centre trial of omeprazole with or without clarithromycin. Aliment Pharmacol Ther 1995;9:417–23.
AJG – Vol. 94, No. 4, 1999
18. O’Morain C, Dettmer A, Rambow A, et al. Double-blind, multicenter, placebo-controlled evaluation of clarithromycin and omeprazole for Helicobacter pylori-associated duodenal ulcer. Helicobacter 1996;1:130 –7. 19. Bardhan KD, Graham DY, Hunt RH, et al. Effects of smoking on cure of H. pylori infection and duodenal ulcer recurrence in patients treated with clarithromycin and omeprazole. Helicobacter 1997;2:27–31. 20. Bazzoli F, Zagari RM, Fossi S, et al. Short-term low-dose triple therapy for the eradication of Helicobacter pylori. Eur J Gastroenterol Hepatol 1994;6:773–7. 21. Lind T, Veldhuyzen van Zanten S, Unge P, et al. Eradication of Helicobacter pylori using one week triple therapies combining omeprazole with two antimicrobials: The MACH I study. Helicobacter 1996;1:138 – 44.
Vol. 94, No. 4, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00047-7
The Effect of Antibiotic Therapy on Bleeding From Duodenal Ulcer Amnon Sonnenberg, M.D., M.Sc., Carol A. Olson, M.D., and Jie Zhang, Ph.D. Department of Veterans Affairs Medical Center and The University of New Mexico, Albuquerque, New Mexico; and Abbott Laboratories, Abbott Park, Illinois
OBJECTIVE: We conducted this study to test whether eradication of Helicobacter pylori (H. pylori) infection prevents hemorrhage related to duodenal ulcer. METHODS: Patients with H. pylori infection and endoscopically proven duodenal ulcers without ulcer-related hemorrhage were enrolled into four randomized, double-blind, multicenter studies using the same study protocol. They were treated with clarithromycin plus omeprazole (441 patients), omeprazole alone (447 patients), or ranitidine alone (263 patients). Success of H. pylori eradication was evaluated by the 13C-urea breath test 4 – 6 wk after the last dose of study drug. Follow-up continued at monthly intervals up to 1 yr after the last dose of study drug. RESULTS: Bleeding due to duodenal ulcer was not observed in any patients who received clarithromycin plus omeprazole, whereas five patients in the omeprazole treatment group and six patients in the ranitidine treatment group experienced an episode of ulcer-related hemorrhage during follow-up. All patients who experienced ulcer-related bleeding were male. When compared by bleeding, there were no significant differences with respect to ethnicity, alcohol consumption, or tobacco use. H. pylori infection was no longer detectable in 68% of patients after treatment with clarithromycin plus omeprazole, compared with 5% after treatment with omeprazole alone or 4% after treatment with ranitidine alone. CONCLUSION: In a population of duodenal ulcer patients without predisposing risk factors for ulcer bleeding, antibiotic eradication or suppression of H. pylori infection prevented the occurrence of ulcer-related hemorrhage for up to 1 yr after therapy. (Am J Gastroenterol 1999;94:950 –954. © 1999 by Am. Coll. of Gastroenterology)
have all been reported to be significant risk factors for ulcer-related hemorrhage (3, 4). The likelihood of recurrent hemorrhage is also increased in subjects with a previous history of ulcer bleeding (5). There is now clear evidence that H. pylori infection is present in .90% of the patients with duodenal ulcer (6), and that cure of the infection leads to dramatic reduction in ulcer recurrence (7). As ulcer recurrence is requisite for bleeding to occur, one is inclined to hypothesize that effective treatment of H. pylori infection should provide protection against the morbidity and mortality associated with ulcerrelated hemorrhage. Along with the alternatives of maintenance treatment with antisecretory agents or surgery, antibiotic treatment of H. pylori infection may decrease the frequency of ulcer recurrence and thus decrease the chances of ulcer-related hemorrhage in ulcer patients. Several previous studies suggested that successful eradication of the H. pylori infection decreased the incidence of gastrointestinal hemorrhage in peptic ulcer patients, compared with unsuccessful antibiotic therapy of the H. pylori infection or maintenance therapy with antisecretory agents (8 –13). However, there have been no reports of large, randomized clinical trials evaluating the effect of curing H. pylori infection on subsequent ulcer-related hemorrhage. This article reports results compiled from four large, randomized, multicenter studies carried out in the United States. It includes patients with proven duodenal ulcer but without a history of gastrointestinal hemorrhage. In this respect, the results of our study may be more reflective of the outcome to be expected in a general population of ulcer patients who receive primary treatment for an uncomplicated duodenal ulcer.
MATERIALS AND METHODS INTRODUCTION Peptic ulcer is the most common cause of acute hemorrhage in the upper gastrointestinal tract, accounting for about 50% of all cases (1). The overall mortality for patients with bleeding ulcers is approximately 6 –7% (2). Many factors appear to contribute to the occurrence of bleeding ulcers: Helicobacter pylori (H. pylori) infection, use of nonsteroidal antiinflammatory drugs (NSAIDs), and increased age
Between 1993 and 1995, four studies were carried out by Abbott Laboratories to assess the efficacy of a dual therapy of clarithromycin plus omeprazole in eradicating H. pylori and preventing the recurrence of duodenal ulcer. Although the four studies were conducted in sequence, an identical study protocol was used for all four studies. The same investigators and study coordinators were involved in monitoring the study sites, accumulating patient records, and
AJG – April, 1999
Bleeding Duodenal Ulcer
951
Table 1. Summary of Individual Clinical Trials Number of Patients per Treatment Arm Study Number
Number of Sites
Clarithromycin 1 Omeprazole
Omeprazole
1 2 3 4
36 37 64 68
82 82 137 140
80 88 135 144
analyzing the data. For these reasons, the cumulative data may be regarded as if it were generated in one large study. Bleeding from duodenal ulcer is a relatively rare event, especially in patients without a previous bleeding history. To achieve statistical power in comparing the influence of different treatment types on ulcer bleeding, for the present analysis we have compiled these previously unpublished data from the four studies and analyzed them together. Patients were enrolled into a randomized double-blind multicenter clinical trial at 205 sites distributed throughout the United States. The data, which form the basis of the present analysis, are summarized in Table 1. A dual combination of clarithromycin plus omeprazole was compared with omeprazole given alone in all four studies. The third and fourth study included a third treatment arm of ranitidine given alone. Patients received clarithromycin 500 mg t.i.d. plus omeprazole 40 mg q.d. for 14 days, followed by omeprazole 20 mg q.d. for an additional 14 days; or omeprazole 40 mg q.d. for 14 days, followed by 20 mg q.d. for an additional 14 days; or ranitidine 150 mg b.i.d. for 28 days. A total of 1151 patients were enrolled in the three treatment groups shown in Table 1. Male or female adults aged $18 yr with a single, active duodenal ulcer and H. pylori infection were eligible for enrollment. The duodenal ulcer was diagnosed by upper gastrointestinal endoscopy. H. pylori infection was diagnosed by a positive CLOtest, histology, culture, 13C-urea breath test, or a combination of these tests. Results were considered undefined if subjects underwent the test but no definitive result was reported by the laboratory. Such instances occurred, for instance, if biopsies were too small for a definitive histological evaluation or no 13 C was detectable in the air sample. Patients with any of the following characteristics were excluded from the studies: bleeding or perforated ulcer at the time of enrollment; evidence of gastric ulcer or gastric malignancy; any history of gastric surgery, gastric outlet obstruction, esophageal stricture, or chronic NSAID consumption; evidence of hepatic and renal impairment; evidence of drug or alcohol abuse; and hospitalization at the time of enrollment. The studies were carried out in accordance with regulations by the Food and Drug Administrations and local regulations and statutes governing study conduct at individual study sites. Approval by institutional review boards was obtained at each participating study site. All patients gave written informed consent before inclusion in the studies. Evaluations were carried out within 3 days before first
Ranitidine
Follow-up Time 6 6 12 12
133 130
months months months months
drug administration and within 5 days after the last dose of study drug. Follow-up continued at monthly intervals up to 1 yr after the last dose of study drug. Duodenal ulcer and H. pylori status were determined at the pretreatment, posttreatment, and final follow-up visit. During follow-up, ulcer relapse was assessed by the recurrence or new onset of symptoms. All episodes of ulcer-related hemorrhage were confirmed by upper gastrointestinal endoscopy. Success of H. pylori eradication was evaluated by the 13C-urea breath test 4 – 6 wk after the last dose of study drug. If both 13 C-urea breath test and histology results were available, both were required to be negative to consider the response to therapy as successful. For the purposes of the present analysis, clinical failure was defined as an endoscopically proven ulcer-related hemorrhage. All subjects enrolled into the study were included in the analysis. The statistical differences among the three treatment groups were compared by x2 tests with Yates correction. Pairwise comparisons for treatment effect on ulcer-related hemorrhage were carried out using Fisher’s exact test. The time until ulcer-related hemorrhage was analyzed using Kaplan-Meier survival analysis.
RESULTS A total of 1151 patients were enrolled into the study comparing three treatments for cure of duodenal ulcer disease. Premature discontinuation of the study drug occurred in 35 (7.9%) patients taking clarithromycin plus omeprazole, 25 (5.6%) patients taking omeprazole alone, and 16 (6.3%) patients taking ranitidine alone (Table 2). All subjects initially enrolled into the study were considered in the present Table 2. Reasons for Premature Discontinuation of the Study Total Enrolled Total enrolled Adverse event Noncompliance Lost to follow-up Withdrawn from study* Sum of premature discontinuation
Clarithromycin 1 Omeprazole Omeprazole Ranitidine 441 13 8 7 7
447 8 5 6 6
253 6 0 3 7
35
25
16
* Patients were withdrawn by their physicians from the study because of worsening of signs and symptoms, lack of response, or because patients had moved to a new location.
952
Sonnenberg et al.
AJG – Vol. 94, No. 4, 1999
Table 3. Bleeding by Treatment Group
Treatment
No DU Bleeding (n 5 1140)
Clarithromycin 1 Omeprazole Omeprazole Ranitidine
441 442 257
DU Bleeding p (n 5 11) Value 0 5 6
0.0013
DU 5 duodenal ulcer.
intention-to-treat analysis, and premature discontinuation of study drug was not a reason for exclusion. Patients enrolled in the three treatment groups were comparable with respect to age, gender, ethnicity, alcohol consumption, and tobacco use. No statistical differences were observed among the patient characteristics in the three treatment groups. Eleven of 1151 duodenal ulcer patients developed ulcer-related bleeding. When patient characteristics were compared according to bleeding outcome, there were no significant differences with respect to age, ethnicity, alcohol consumption, or tobacco use. However, all 11 patients who experienced ulcer-related bleeding were male, as compared with 790 of 1140 patients without bleeding (p 5 0.0610). As shown in Table 3, episodes of ulcer-related hemorrhage occurred during the follow-up period only in patients who had been treated initially with omeprazole or ranitidine alone. The statistical comparison revealed a significant treatment effect. Pairwise comparison between each two treatment groups, using Fisher exact test, showed less ulcerrelated bleeding after clarithromycin plus omeprazole than after ranitidine alone (p 5 0.003) or omeprazole alone (p 5 0.062). Figure 1 compares the time until ulcer-related hemorrhage in the three treatment groups. The p-values from the log-rank test are both 0.002 for comparing clarithromycin plus omeprazole versus omeprazole alone or versus ranitidine alone. As shown by the figure, all ulcer-related hemorrhages occurred within the first 6 months after treatment of the acute duodenal ulcer. This pattern excludes a potential bias that might have been introduced by the longer observation periods of patients followed-up after treatment with ranitidine alone (Table 1). To assess the effect of treatment on H. pylori status, we excluded patients from the analysis whose H. pylori status was not defined before or after treatment (Table 4). Among patients with undefined H. pylori status, three taking omeprazole and two taking ranitidine experienced ulcer-related bleeding. The other patients with ulcer-related hemorrhage
Figure 1. Kaplan-Meier analysis of ulcer bleeding. The x-axis depicts the time until ulcer-related bleeding, the y-axis the probability of nonbleeding. Each curve represents one of the three treatment arms. The differences between the bleeding probabilities of clarithromycin plus omeprazole vs omeprazole alone or vs ranitidine alone were both significant at p 5 0.002.
(two in the omeprazole treatment group and four in the ranitidine treatment group) had remained H. pylori positive. Among the 866 patients available for analysis, H. pylori infection was no longer detectable in 68% after treatment with clarithromycin plus omeprazole, compared with 5% after treatment with omeprazole alone and 4% after treatment with ranitidine alone (Table 5).
DISCUSSION In the present study, we analyzed the occurrence of ulcerrelated bleeding among patients who participated in four randomized, double-blind, multicenter trials, which evaluated the efficacy of clarithromycin plus omeprazole, omeprazole alone, and ranitidine alone in the treatment of H. pylori infection. None of the patients treated with clarithromycin plus omeprazole experienced episodes of ulcer-related bleeding during the follow-up period. Ulcer-related bleeding was significantly more frequent among patients who had received omeprazole alone or ranitidine alone. Previous studies established the presence of H. pylori infection in .90% of patients with duodenal ulcer (1), and showed that there is a dramatic decrease in ulcer recurrence after successful treatment of the infection (2). Other reports suggested that antibiotic treatment is more effective than maintenance therapy with ranitidine in preventing ulcer complications, particularly ulcer-related hemorrhage (8 –13). However, these studies were restricted to relatively few patients who presented initially with bleeding gastric
Table 4. Patients Available for Analysis by H. pylori Status Excluded From Analysis by Helicobacter pylori Status Treatment Group
Total No. of Patients in Treatment Group
Not Defined Pretreatment
Not Defined Posttreatment
Total No. of Patients Excluded
Clarithromycin 1 Omeprazole Omeprazole Ranitidine
441 447 263
28 33 17
83 89 35
111 122 52
AJG – April, 1999
Bleeding Duodenal Ulcer
Table 5. H. pylori Eradication Rates Treatment Group Clarithromycin 1 Omeprazole Omeprazole Ranitidine
H. pylori Negative
H. pylori Positive
Percent Eradication
223
107
68%
15 9
310 202
5% 4%
953
In summary, in a population of duodenal ulcer patients without predisposing risk factors for ulcer bleeding, antibiotic eradication or suppression of H. pylori infection prevented the occurrence of ulcer-related hemorrhage for up to 1 yr after therapy.
ACKNOWLEDGMENTS and duodenal ulcers. The results of the present analysis with .1000 patients from controlled clinical trials support and extend the results of the previous studies. In the majority of ulcer patients, ulcer-related hemorrhage represents a rare complication. Our study yields an incidence rate of 1.55%/yr among the 710 patients treated with antisecretory medication alone. This rate is lower than the rates available from the literature for patients with a previous history of ulcer bleeding (5, 14). Nevertheless, the study has the statistical power to show that even in patients without previous ulcer-related hemorrhage, antibiotic eradication of H. pylori carries the benefit of preventing such rare complication. It is mostly the occurrence of such rare but costly complications that results in the substantial economic burden associated with peptic ulcer disease (15). The H. pylori eradication rate for clarithromycin plus omeprazole reported here compares favorably to the results previously reported for individual clinical trials (16 –19). Dual therapy was considered state-of-the-art treatment at the time the studies were initiated. By adding metronidazole or amoxicillin to the dual therapy, newer triple regimens may achieve eradication rates in excess of 90% (20, 21). Despite the lower eradication rates obtained with dual therapy, the absence of bleeding episodes in patients treated with clarithromycin plus omeprazole suggests that antibiotic therapy directed against H. pylori infection is clinically effective. Over a time period of 6 –12 months, the beneficial effect of antibiotic therapy may apply even to patients in whom eradication was not successful. This suggests that suppression rather than eradication of H. pylori is already sufficient to provide some degree of protection against ulcer-related hemorrhage. In addition to H. pylori infection, other factors have been reported to contribute to the occurrence of ulcer-related hemorrhage in peptic ulcer patients. In particular, age, use of NSAIDs, previous bleeding episodes, and presence of gastric versus duodenal may all be relevant to the occurrence of bleeding episodes (3–5). Because in the present study we excluded patients with NSAID use, gastric ulcer, or ulcerrelated hemorrhage, the influence of these factors could not be assessed. Within our patient population, consumption of alcohol and tobacco was not associated with an increased occurrence of ulcer-related hemorrhage. Based on the results presented here, H. pylori infection appears to be the primary influence on the frequency of ulcer-related bleeding episodes. However, it is possible that such factors as alcohol and tobacco use act in concert with H. pylori infection in increasing the likelihood of ulcer-related hemorrhage.
Dr. Sonnenberg’s research was funded through a grant by the Centers for Disease Control and Prevention, Atlanta, Georgia. This work was conducted while Dr. Sonnenberg was employed by the Department of Veterans Affairs. The clinical study was supported by Abbott Laboratories, Abbott Park, Illinois. We are indebted to Larry Hancock for his assistance in writing the manuscript. Reprint requests and correspondence: Amnon Sonnenberg, M.D., M.Sc., Department of Veterans Affairs Medical Center 111F, 2100 Ridgecrest Drive SE, Albuquerque, NM 87108. Received July 9, 1998; accepted Nov. 30, 1998.
REFERENCES 1. Laine L, Peterson WL. Bleeding peptic ulcer. N Eng J Med 1996;331:717–27. 2. Silverstein FE, Gilbert DA, Tedesco FJ, et al. The national ASGE survey on upper gastrointestinal bleeding. II. Clinical prognostic factors. Gastrointest Endosc 1981;27:80 –93. 3. Van Deventer GM, Elashoff JD, Reedy TJ, et al. A randomized study of maintenance therapy with ranitidine to prevent the recurrence of duodenal ulcer. N Engl J Med 1989;320: 1113–9. 4. Taha AS, Russell RI. Helicobacter pylori and non-steroidal anti-inflammatory drugs: Uncomfortable partners in peptic ulcer disease. Gut 1993;34:580 –3. 5. Elashoff JD, Van Deventer G, Reedy TJ, et al. Long-term follow-up of duodenal ulcer patients. J Clin Gastroenterol 1983;5:509 –15. 6. Consensus Conference. Helicobacter pylori in peptic ulcer disease. J Am Med Assoc 1994;272:65–9. 7. Hopkins RJ, Girardi LS, Turney EA. Relationship between Helicobacter pylori eradication and reduced duodenal and gastric ulcer recurrence. Gastroenterology 1996;110:1244 –52. 8. Graham DY, Hepps KS, Ramirez FC, et al. Treatment of Helicobacter pylori reduces the rate of rebleeding in peptic ulcer disease. Scand J Gastroenterol 1993;28:939 – 42. 9. Labenz J, Bo¨rsch G. Role of Helicobacter pylori eradication in the prevention of peptic ulcer bleeding relapse. Digestion 1994;55:19 –23. 10. Rokkas T, Karameris A, Mavrogeogis A, et al. Eradication of Helicobacter pylori reduces the possibility of rebleeding in peptic ulcer disease. Gastrointest Endosc 1995;41:1– 4. 11. Jaspersen D, Koerner T, Schorr W, et al. Helicobacter pylori eradication reduces the rate of rebleeding in ulcer haemorrhage. Gastrointest Endosc 1995;41:5–7. 12. Powell KU, Bell GD, Bolton GH, et al. Helicobacter pylori eradication in patients with peptic ulcer disease: Clinical consequences and financial implications. Quat J Med 1994;87:283–90. 13. Santander C, Gravalos RG, Gomez-Cedenilla A, et al. Antimicrobial therapy for Helicobacter pylori infection versus long-term maintenance antisecretion treatment in the prevention of recurrent hemorrhage from peptic ulcer: Prospective
954
14. 15. 16.
17.
Sonnenberg et al.
nonrandomized trial on 125 patients. Am J Gastroenterol 1996;91:1549 –52. Pulvertaft CN. Comments on the incidence and natural history of gastric and duodenal ulcer. Postgrad Med J 1968;44:597–602. Sonnenberg A, Everhart JE. Health impact of peptic ulcer in the United States. Am J Gastroenterol 1997;92:614 –20. Hunt R, Schwartz H, Fitch D, et al. Dual therapy of clarithromycin and omeprazole for treatment of patients with duodenal ulcers associated with H. pylori infection. Gut 1995;37(suppl 1):A5. Logan RPH, Bardhan KD, Celestin LR, et al. Eradication of H. pylori and prevention of recurrence of duodenal ulcer: A randomized, double-blind, multi-centre trial of omeprazole with or without clarithromycin. Aliment Pharmacol Ther 1995;9:417–23.
AJG – Vol. 94, No. 4, 1999
18. O’Morain C, Dettmer A, Rambow A, et al. Double-blind, multicenter, placebo-controlled evaluation of clarithromycin and omeprazole for Helicobacter pylori-associated duodenal ulcer. Helicobacter 1996;1:130 –7. 19. Bardhan KD, Graham DY, Hunt RH, et al. Effects of smoking on cure of H. pylori infection and duodenal ulcer recurrence in patients treated with clarithromycin and omeprazole. Helicobacter 1997;2:27–31. 20. Bazzoli F, Zagari RM, Fossi S, et al. Short-term low-dose triple therapy for the eradication of Helicobacter pylori. Eur J Gastroenterol Hepatol 1994;6:773–7. 21. Lind T, Veldhuyzen van Zanten S, Unge P, et al. Eradication of Helicobacter pylori using one week triple therapies combining omeprazole with two antimicrobials: The MACH I study. Helicobacter 1996;1:138 – 44.