Effect of Triple Therapy (Antibiotics plus Bismuth) on Duodenal Ulcer Healing A Randomized Controlled Trial David Y. Graham, MD; Ginger M. Lew, PA-C; Dolores G. Evans, PhD; Doyle J. Evans, Jr., PhD; and Peter D. Klein, PhD

• Objective: To determine whether antimicrobial therapy for Helicobacter pylori infection accelerates the healing of duodenal ulcers. • Design: Single-blind, randomized, controlled trial. • Setting: Veterans Affairs hospital. • Participants: One hundred and five patients with endoscopically verified duodenal ulcers. • Intervention: Patients received either ranitidine, 300 mg/d, or ranitidine, 300 mg/d, plus "triple therapy" (2 g/d of tetracycline, 750 mg/d of metronidazole, and 5 or 8 bismuth subsalicylate tablets per day). Triple therapy was administered for only the first 2 weeks of ulcer treatment. • Measurements: Videoendoscopic assessment of ulcer status was done until ulcer healing was complete. Evaluations were done after 2,4,8,12, and 16 weeks of therapy. • Main Results: Ulcer healing was more rapid in patients receiving ranitidine plus triple therapy than in patients receiving ranitidine alone ( P < 0.01). The cumulative percentages of patients with healed ulcers in the group receiving ranitidine plus triple therapy and in the group receiving ranitidine alone were as follows: 37% and 18% after week 2; 74% and 53% after week 4; 84% and 68% after week 8; 96% and 80% after week 12; and 98% and 84% after week 16. • Conclusion: Combined therapy with anti-H. pylori agents and ranitidine was superior to ranitidine alone for duodenal ulcer healing. Our results indicate that H. pylori plays a role in duodenal ulcer disease.

Annals of Internal Medicine. 1991;115:266-269. From Baylor College of Medicine, the Veterans Affairs Medical Center, and the U.S. Department of Agriculture/Agricultural Research Center Children's Nutrition Center, Houston, Texas. For current author addresses, see end of text. 266

A n etiologic association between gastroduodenal infection with Helicobacter pylori and duodenal ulcer disease has been strongly suspected (1). Chronic duodenal ulcer disease is nearly always associated with H. pylori infection, and the eradication of H. pylori infection markedly reduces the frequency of ulcer relapse (2-7). We did a study to determine whether duodenal ulcer healing with H 2 -receptor antagonists was further accelerated by the addition of therapy to eradicate H. pylori infection. Methods Patients presenting with dyspepsia or upper gastrointestinal bleeding to the gastroenterology service at the Houston Veterans Affairs Medical Center between June 1988 and May 1990 underwent upper gastrointestinal endoscopy. Consecutive patients with active duodenal ulcer were entered into our singlecenter study unless they were receiving active anti-ulcer therapy, were scheduled for surgery, were either unable or unwilling to participate, or had severe comorbid disease that made their completing the study doubtful. Ulcers were identified by endoscopy using Fujinon videoendoscopes (Fujinon Inc., Wayne, New Jersey). A video still (ProMavica, Sony, Sony Park Ridge, New Jersey) of each ulcer was made so that the site and characteristics of the ulcer could be reviewed before subsequent endoscopic procedures. One videodisk was assigned to each patient. Endoscopy was done at entry and endoscopic evaluations were scheduled after 2, 4, 8, 12, and 16 weeks of therapy. The ulcer status (healed or unhealed) was evaluated either until ulcer healing was achieved or until the end of the study at 16 weeks. An ulcer was defined as a circumscribed break in the duodenal mucosa that measured at least 5 mm in diameter with apparent depth and that was covered with an exudate. Pyloric channel ulcers were included as duodenal ulcers. Ulcer healing was defined as complete re-epithelialization. Patients were randomly assigned (in blocks of 10) to receive either ranitidine or ranitidine plus *'triple therapy," with the endoscopist remaining blinded to treatment assignment. Ranitidine, 300 mg, was given once daily in the evening. Triple therapy consisted of bismuth and two antibiotics: tetracycline hydrochloride, 500 mg four times a day, and metronidazole, 250 mg three times a day. Bismuth was administered in the form of bismuth subsalicylate tablets (Pepto-Bismol, Procter & Gamble, Cincinnati, Ohio); the first 30 patients received 1 tablet with each meal and 2 at bedtime, and the last 23 patients received 2 tablets with each meal and 2 at bedtime. PeptoBismol tablets contain 151 mg bismuth per tablet and De-Nol (Gist Brocades Pharmaceuticals, Leiderdorp, Holland), which has been used in other studies, contains 108 mg bismuth per tablet. Triple therapy was administered for only the first 2 weeks (14 ± 2 days) of ulcer treatment. Patients regularly receiving therapy with nonsteroidal anti-inflammatory drugs were allowed to continue it if they wished; no medications except antiulcer medications were prohibited during the study. Assessment of symptoms and compliance with medication (pill count) was done by one of the investigators, and the results

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/12/2017

were not shared with the endoscopist. The protocol was approved by the Institutional Review Board at the Houston Veterans Affairs Medical Center and Baylor College of Medicine. Written informed consent was obtained before patient entry. All patients were assessed for H. pylori infection with the 13 C-urea breath test (8, 9) and a sensitive, specific enzymelinked immunosorbent assay (ELISA) for IgG antibody against the high-molecular-weight cell-associated proteins of H. pylori (10); most patients were also assessed by culture and histologic evaluation of antral mucosal biopsy specimens. Eradication was defined as no evidence of H. pylori infection by any test 1 or more months after stopping triple therapy. The frequency of ulcer healing was calculated as the cumulative proportion of ulcers healed at each time point. Tests of categorical data were done using the chi-square or Fisher exact test. Numerical data were analyzed using Wilcoxon signedrank test. The time to healing was also assessed using the Mantel-Haenszel method and other survival analysis methods (for example, log-rank). All P values < 0.05 (two-tailed) were considered significant.

of urea breath test and culture or histologic evaluation was as follows: Helicobacter pylori infection was eradicated in 43 patients, no infection was present in 2 patients, antimicrobial therapy failed in 2 patients, and status was unknown in 6 patients. In both patients who failed therapy, infection had been suppressed (negative urea breath test) by the end of therapy. Of the 6 patients whose status was unknown, 4 failed to return after ulcer healing and 2 were lost to follow-up before ulcer healing was documented. All patients assigned to receive ranitidine alone were infected with H. pylori and remained so during the study. Side effects of triple therapy were experienced by six patients. Queasiness occurred in two patients, nausea and vomiting in two, fatigue in one, and diarrhea in one. Only one patient had symptoms severe enough to discontinue therapy. No patient receiving ranitidine alone experienced a side effect.

Results One hundred and five patients with endoscopically active duodenal ulcers were entered into the study; 52 were randomly assigned to receive ranitidine and 53 to receive ranitidine plus triple therapy. The two treatment groups were well matched; there were no significant differences between them regarding any variable (Table 1). Helicobacter pylori infection was documented in 51 of the 53 patients receiving ranitidine plus triple therapy and in all 52 patients receiving ranitidine alone. Healing of duodenal ulcers was more rapid in patients receiving ranitidine plus triple therapy than in those receiving ranitidine alone (Figure 1) (P < 0.01). The difference between the two groups was similar whether the analyses were done on cumulative healing or by lifetable analysis. Five patients who received ranitidine alone and one patient who received ranitidine plus triple therapy had poor compliance (defined as < 70% of prescribed drug taken) with ranitidine. Inclusion or exclusion of these patients in the analyses did not change the results. Ulcers that were 2 cm or more in diameter healed slowly; none healed in less than 8 weeks, and the median time to healing of these larger ulcers was 12 weeks. Fifteen patients elected to continue therapy with nonsteroidal anti-inflammatory drugs throughout the study: In the group assigned to ranitidine alone, 3 patients were taking ibuprofen; 2, aspirin; 1, naproxen; and 1, tolmetin. In the group assigned to ranitidine plus triple therapy, 4 patients were taking ibuprofen; 2, aspirin; and 2, piroxicam. Duodenal ulcer healing was achieved at a median of 4 weeks in both groups for patients who continued treatment with nonsteroidal anti-inflammatory drugs. None of the patients in the ranitidine alone group who had unhealed ulcers at 16 weeks had been taking nonsteroidal anti-inflammatory drugs. The single patient in the triple therapy group who had an unhealed ulcer at 16 weeks had a large ulcer (2 cm in diameter) and had continued to receive ibuprofen. This patient's ulcer healed after stopping ibuprofen and receiving treatment with omeprazole, 40 mg daily for 4 weeks; the H. pylori infection had been eradicated. The final status of H. pylori infection in patients receiving triple therapy as determined by the combination

Discussion Helicobacter pylori infection is now accepted as the most common cause of chronic gastritis in humans (1). The association between chronic antral gastritis and duodenal ulcer has been recognized for more than 40 years (11). Recent studies have shown that eradication of H. pylori infection is associated with healing of gastritis (12) and a markedly reduced rate of recurrence of duodenal ulcers (2-7, 13). In our study, the addition of therapy designed to treat H. pylori infection accelerated the rate of duodenal ulcer healing and resulted in a higher percentage of completely healed ulcers within the 16-week time frame of the study. These results provide further evidence that H. pylori infection plays a role in the pathogenesis of duodenal ulcer.

Table 1. Demographic and Clinical Characteristics of the Study Patients* Variable

Ranitidine Alone (" = 52)

Ranitidine plus Triple Therapy in = 53)

Median age (range), y Male, % Race, % White Black Other Married, % Education, % < 12 years 12 years > 12 years Recent use of NSAIDst, % Ulcer size (range)$, cm

58 (31-86) 96

57 (28-80) 100

Ulcer > 2.0 cm, n Smoker, % Alcohol use§, % Family history of ulcer, % Helicobacter pylori infection, %

57 39 4 63 46 28 26 28 1.01 ±:0.1 (0.5-6) 1 48 26 17 100

68 32 0 57 36 36 28 30 1.08 ± 0.1 (0.5-5) 5 64 34 13 96

* None of the differences in variables was statistically significant. t NSAIDs = nonsteroidal anti-inflammatory drugs. % Mean ± SE. § One or more times per week.

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/12/2017


the concentrations of bismuth subcarbonate or bismuth subnitrate were increased fivefold. Subsequent studies have shown important differences between different formulations of bismuth subcitrate; colloidal bismuth subcitrate is effective in preventing ulcers in Shay rats but noncolloidal bismuth subcitrate is not (19). Thus, it appears that both the bismuth salt and the method by which it is formulated are important considerations, especially when dealing with experimental ulcers. There is no evidence to our knowledge that bismuth subsalicylate has direct antiulcer properties other than its effects on H. pylori. Triple therapy also does not have an effect on gastric antral function, as assessed by bombesinstimulated gastrin release in normal uninfected persons (20). Finally, triple therapy was administered for only the initial 2 weeks of therapy and not for the entire period of ulcer treatment.

Figure 1. Lifetable analysis of the healing of duodenal ulcers in two groups of patients. Healing of duodenal ulcer was significantly more rapid (P < 0.01, log-rank) in patients assigned to receive ranitidine plus triple antimicrobial therapy (triple Rx) (n = 53) than in those assigned to receive ranitidine alone {n = 52).

The percentage of healed ulcers in our patients receiving ranitidine alone may appear to be low when compared with results in previous studies. Nevertheless, our results are consistent with recent studies comparing omeprazole with either cimetidine (14) or with placebo (15), and the low percentage may be attributable to the fact that the centers with a major interest in ulcer disease tend to collect the more severe cases (16). All endoscopic examinations to determine healing were done by a senior endoscopist. We believe that the use of a wide-field magnified image with excellent close resolution and the immediate preprocedure review of the images from previous procedures, coupled with the use of glucagon to temporarily paralyze the duodenal musculature to eliminate motility and muscle spasm, allowed better discrimination of re-epithelization than was previously possible. We cannot be certain that the drugs used did not have effects other than their known bactericidal effect on H. pylori. Colloidal bismuth subcitrate is an effective antiulcer agent whose effectiveness is not limited to ulcers associated with H. pylori infection (17, 18). Colloidal bismuth subcitrate was not used in our study, and other bismuth preparations do not appear to have similar effects. In one study, histochemical staining was used to compare bismuth subnitrate, bismuth subcarbonate, and bismuth subsalicylate with colloidal bismuth subcitrate for the ability to coat experimental gastric ulcers in rats (18). When colloidal bismuth subcitrate was administered, bismuth was deposited in a uniform layer covering the ulcer base. The histochemical staining of the ulcer base was negative with the other bismuth salts and remained negative even when 268

Our results are also consistent with those of a recent study that suggested that combination therapy with an H 2 -receptor antagonist and colloidal bismuth subcitrate (De-Nol), two proven active antiulcer agents, may induce more rapid duodenal ulcer healing than ranitidine alone (21). These studies, taken together with our study, suggest that the suppression of H. pylori infection may be sufficient to achieve the accelerated rate of ulcer healing that we demonstrated. In contrast, eradication of H. pylori infection is apparently required to prevent duodenal ulcer recurrence (2-4, 6, 7). Peptic ulcers can be divided into three major categories: H. pylori-related, nonsteroidal anti-inflammatory drug-induced, and hypersecretory (for example, those found in the Zollinger-Ellison syndrome) (22). Two of our patients, who had no evidence of//, pylori infection or the Zollinger-Ellison syndrome, were defined as having chronic duodenal ulcers that were induced by nonsteroidal anti-inflammatory drugs. One of these patients took more than 12 aspirin tablets daily for headache. Initially, his duodenal ulcer healed after 4 weeks of therapy. It recurred and was resistant to therapy (that is, it continued to enlarge despite therapy with cimetidine, 200 mg four times a day for 4 weeks, followed by therapy with omeprazole, 40 mg for 4 weeks). Finally, the patient agreed to stop taking aspirin, and after 4 weeks of therapy with omeprazole, 40 mg, and misoprostol, 300 /Ag four times a day, the ulcer healed and has not recurred. The patient has remained off aspirin. Although we showed that adding anti-//. pylori agents to ranitidine therapy significantly accelerated duodenal ulcer healing, we believe that the overall clinical benefit is small. Other strategies, such as omeprazole therapy, would be just as effective. A higher percentage of the ulcers (at 12 or 16 weeks) healed when triple therapy was combined with ranitidine, suggesting that antiH. pylori therapy may be particularly useful for the patient with a resistant ulcer. We do not recommend routinely adding therapy to eradicate H. pylori to accelerate ulcer healing. There are already several therapies that have been proved to accelerate ulcer healing and that are remarkably safe. The addition of anti-//. pylori therapy is best considered as a method to alter the natural history of the disease, that is, to prevent or reduce ulcer relapse (2-4, 6, 7).

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/12/2017

Acknowledgments: The authors thank Patrice A. Michaletz, Z. A. Seed, Hoda Malaty, Antone Opekun, Gloria Thompson, Valeria Tapscott, and Sara Sekely for their help. Grant Support: In part by the Department of Veterans Affairs, by grant DK 39919 from the National Institute of Diabetes and Digestive and Kidney Diseases, by the U.S. Department of Agriculture/Agricultural Research Service Children's Nutrition Center, and by Hilda Schwartz. The ranitidine was supplied by Glaxo and the Pepto-Bismol tablets by Procter and Gamble. Requests for Reprints: David Y. Graham, MD, Veterans Affairs Medical Center (11 ID), 2002 Holcombe Boulevard, Houston, TX 77030. Current Author Addresses: Drs. Graham, Evans, Evans, and Ms. Lew: Veterans Affairs Medical Center (HID) 2002 Holcombe Boulevard, Houston, TX 77030. Dr. Klein: Children's Nutrition Research Center, 1100 Bates Street, Houston, TX 77030. References 1. Graham DY. Campylobacter pylori and peptic ulcer disease. Gastroenterology. 1989;96:615-25. 2. Coghlan JG, Gilligan D, Humphreys H, McKenna D, Dooley C, Sweeney E, et al. Campylobacter pylori and recurrence of duodenal ulcers—a 12-month follow-up study. Lancet. 1987;2:1109-11. 3. Lambert JR, Borromeo M, Korman MG, Hansky J, Eaves ER. Effect of colloidal bismuth (De-Nol) on healing and relapse of duodenal ulcers—role of Campylobacter pyloridis. Gastroenterology. 1987 ;92:1489. 4. Marshall BJ, Goodwin CS, Warren JR, Murray R, Blincow ED, Blackbourn SJ, et al. Prospective double-blind trial of duodenal ulcer relapse after eradication of Campylobacter pylori. Lancet. 1988;2:1437-42. 5. Smith AC, Price AB, Borriello P, Levi AJ. A comparison of ranitidine and tripotassium dicitrato-bismuth (T.D.B.) in relapse rates of duodenal ulcer: the role of Campylobacter pylori (C.P.) [Abstract]. Gastroenterology. 1988;94:A431. 6. Rauws EA, Tytgat GN. Cure of duodenal ulcer associated with eradication of Helicobacter pylori. Lancet. 1990;335:1233-5. 7. George LJ, Borody TJ, Andrews P, Devine M, Moore-Jones D, Walton M, et al. Cure of duodenal ulcer after eradication of Helicobacter pylori. Med J Aust. 1990;153:145-9. 8. Graham DY, Klein PD, Evans DJ Jr, Evans DG, Alpert LC, Opekun AR, et al. Campylobacter pylori detected noninvasively by the 13C-urea breath test. Lancet. 1987;1:1174-7.

9. Klein PD, Graham DY. Campylobacter pylori detection by the 13Curea breath test. In: Rathbone B, Heatley V, eds. Campylobacter pylori and Gastroduodenal Disease. Oxford: Blackwell Scientific Publications; 1989:94-106. 10. Evans DJ Jr, Evans DG, Graham DY, Klein PD. A sensitive and specific serologic test for detection of Campylobacter pylori infection. Gastroenterology. 1989;96:1004-8. 11. Magnus HA. Gastritis. In: Avery Jones F, ed. Modern Trends in Gastroenterology. London: Butterworth & Company; 1952:323-51. 12. Rauws EA, Langenberg W, Houthoff HJ, Zanen HC, Tytgat GN. Campylobacter pyloridis-associalzd chronic active antral gastritis: a prospective study of its prevalence and the effects of antibacterial and antiulcer treatment. Gastroenterology. 1988;94:33-40. 13. Graham DY, Lew GM, Michaletz PA. Randomized controlled trial of the effect of eradication of C. pylori on ulcer healing and relapse [Abstract]. Gastroenterology. 1989;96:A181. 14. Archambault AP, Pare P, Bailey RJ, Navert H, Williams CN, Freeman HJ, et al. Omeprazole (20 mg daily) versus cimetidine (1200 mg daily) in duodenal ulcer healing and pain relief. Gastroenterology. 1988;94:1130-4. 15. Graham DY, McCullough A, Sklar M, Sontag SJ, Roufail WM, Stone RC, et al. Omeprazole verses placebo in duodenal ulcer healing. The United States experience. Dig Dis Sci. 1990;35:6672. 16. Soil AH. Duodenal ulcer and drug therapy. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal Disease: Pathophysiology, Diagnosis and Management. Philadelphia: W.B. Saunders Co.; 1989:81479. 17. Hall DW. Review of the modes of action of colloidal bismuth subcitrate. Scand J Gastroenterol. 1989;24:3-6. 18. Koo J, Ho J, Lam SK, Wong J, Ong GB. Selective coating of gastric ulcer by tripotassium dicitrato bismuthate in the rat. Gastroenterology. 1982;82:864-70. 19. Lavy UI, Koekkoek PH, Jaitly KD. Anti-ulcer activity of colloidal bismuth subcitrate in Shay-rats. Arch Int Pharmacodyn. 1976,224: 291-8. 20. Graham DY, Opekun AR, Lew GM, Klein PD, Walsh JH. Helicobacter pylori exaggerated gastrin release in man: effect of bombesin infusion and urea ingestion. Gastroenterology. 1991 [In press]. 21. Salmon PR. Combination treatment: colloidal bismuth subcitrate with H2-antagonists. Digestion. 1987;37:42-6. 22. Graham DY, Lidsky MD, Cox AM, Evans DJ Jr, Evans DG, Alpert L, et al. Long-term nonsteroidial anti-inflammatory drug use and Helicobacter pylori infection. Gastroenterology. 1991;100:1653-7.

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/12/2017


Effect of triple therapy (antibiotics plus bismuth) on duodenal ulcer healing. A randomized controlled trial.

To determine whether antimicrobial therapy for Helicobacter pylori infection accelerates the healing of duodenal ulcers...
724KB Sizes 0 Downloads 0 Views