Misoprostol Compared with Sucralfate in the Prevention of Nonsteroidal Anti-inflammatory Drug-induced Gastric Ulcer A Randomized, Controlled Trial Naurang M. Agrawal, MD; Sanford Roth, MD; David Y. Graham, MD; Richard H. White, MD; Bernard Germain, MD; Jeffry A. Brown, MD; and Scott C. Stromatt, MD
• Objectives: To compare the efficacy and frequency of adverse experiences of misoprostol and sucralfate in the prevention of gastric ulcers in patients receiving nonsteroidal anti-inflammatory drug (NSAID) therapy. • Design: A prospective, randomized, single-blind, multicenter trial. • Patients: Patients with osteoarthritis receiving treatment with ibuprofen, piroxicam, or naproxen and experiencing abdominal pain were eligible. • Interventions: Patients who were expected to receive at least 3 months of NSAID therapy and who did not have a gastric ulcer at the time of the initial screening endoscopy were randomized to receive misoprostol, 200 |xg four times a day, or sucralfate, 1 g four times a day. A gastric ulcer was defined as a lesion of the gastric mucosa 0.3 cm or greater in diameter. Patients were followed clinically, and repeat endoscopies were performed after 4, 8, and 12 weeks. • Main Measurement: The development of a gastric ulcer, which was regarded as a prophylaxis failure. • Results: Two hundred fifty-three patients were evaluable for efficacy analysis. A gastric ulcer developed in 2 of the 122 (1.6%, 95% CI, 0.3% to 6.4%) patients on misoprostol, compared with 21 of 131 patients on sucralfate (16%, CI, 10.4% to 23.7%). The difference in ulcer rates was 14.4% (CI, 10.4% to 19.5%; P < 0.001). • Conclusion: In patients receiving chronic NSAID therapy for osteoarthritis, treatment with misoprostol for 3 months was associated with a significantly lower frequency of gastric ulcer formation, compared with treatment with sucralfate (P< 0.001).
Annals of Internal Medicine. 1991;115:195-200. From Tulane University School of Medicine, New Orleans, Louisiana; Arizona Arthritis Research & Education, Ltd., Phoenix, Arizona; Baylor College of Medicine/Veterans Affairs Medical Center, Houston, Texas; University of California, Davis and Sacramento, California; University of South Florida, Tampa, Florida; University of Illinois, Chicago, Illinois; and University of Health Sciences/Chicago Medical School, Chicago, Illinois. For a list of additional study investigators and for current author addresses, see end of text.
W onsteroidal anti-inflammatory drugs (NSAIDs) are an integral part of the therapy of rheumatic diseases. These drugs can damage the gastrointestinal tract and have been implicated as a cause of peptic ulceration and life-threatening bleeding (1-4). The inhibitory effect of NSAIDs on the endogenous biosynthesis of prostaglandins is thought to be the major mechanism for the therapeutic properties of NSAIDs in the treatment of inflammatory arthritis, although other mechanisms have recently been proposed (5). Prostaglandins play a significant role in the defense of gastrointestinal mucosa (6, 7). Prostaglandins of the E series have gastric antisecretory and mucosal protective properties, and NSAID-induced inhibition of gastric mucosal prostaglandin synthesis is thought to be responsible for much of the gastrointestinal tract toxicity associated with these agents (3, 6-8). Misoprostol, a synthetic prostaglandin E, analog, has been shown to be effective in the prevention of NSAIDinduced gastric ulcers in patients with arthritis (9). This agent has also been shown to prevent the development of NSAID-induced duodenal lesions both in normal subjects (10, 11) and in arthritic patients (12). The beneficial effects of misoprostol on the gastrointestinal tract are not accompanied by a compromise of the antirheumatic effects of the NSAID (13, 14). Sucralfate is an effective anti-ulcer drug that has been shown to increase the release of endogenous prostaglandins from the gastric mucosa (15, 16), suggesting that stimulation of prostaglandin synthesis may be a mechanism for its action. Despite the fact that sucralfate is widely used to prevent NSAID-induced gastrointestinal mucosal damage, no study has evaluated the efficacy of sucralfate in the prevention of NSAID-induced mucosal damage in chronic NSAID users. We compared misoprostol and sucralfate in the prevention of NSAID-induced gastric ulcer in a large cohort of patients with osteoarthritis who were experiencing upper gastrointestinal pain in association with the use of NSAID therapy. Methods Patients were recruited from private practice offices and clinics, Veterans Affairs clinics, health maintenance organizations, and academic institutions. Patients were eligible to enter the study if they had osteoarthritis and were experiencing upper abdominal pain that was thought to be caused by one of three NSAIDs: ibuprofen, piroxicam, or naproxen. In addition,
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
195
it was required that all patients be expected to receive NSAIDs for at least 3 additional months. Women had to be postmenopausal, surgically sterilized, or practicing adequate contraception. Women of child-bearing potential were informed of the possibility that use of misoprostol could "result in a miscarriage. A pregnancy test was done 72 hours before receiving the first dose of study medication; if the pregnancy test was positive, the patient was excluded. Exclusion criteria included a history of recurrent peptic ulcer disease, active bleeding ulcer, upper gastrointestinal malignancy or metastasis to the upper gastrointestinal tract, pyloric or duodenal obstruction, acute hepatitis, pancreatitis, inflammatory bowel disease, bleeding diathesis, upper gastrointestinal surgery within 30 days, or severe renal impairment. Patients taking antineoplastic drugs, anticoagulants, or anti-ulcer drugs, other than the study drugs, were excluded. Study Design The study was a prospective, randomized, single-blind, multicenter comparison of misoprostol, 200 /xg given four times a day (with meals and at bedtime) and sucralfate, 1 g four times a day (30 minutes before meals and at bedtime). Patients were referred for abdominal pain that was thought by the investigators to be due to NSAID therapy. Patients were entered consecutively into the study. Eligible patients underwent an initial screening upper gastrointestinal endoscopy. Patients in whom a gastric or duodenal ulcer was found were excluded from the study. Patients without a gastric ulcer were randomized and started on a study drug within 72 hours of the entry endoscopic examination. Patients continued to take NSAIDs at the same dose that was administered before the study. Although the patients in the study could have possibly recognized their study medication, the endoscopists were blinded to the medications being taken by the patients. In addition, patients were instructed not to discuss their study medications or symptoms with the endoscopists. The gastric mucosa was examined by fiberoptic endoscopy after 4 weeks (± 3 days), 8 weeks (± 5 days), and 12 weeks (± 5 days) after randomization. Patients were allowed to take up to four aluminum hydroxide antacid tablets per day, during the first week only, for relief of upper gastrointestinal pain. Patients were instructed to take the study medication the night before the endoscopy and not to take the next dose until after the procedure was completed. Noncompliance was defined as failure to take at least 75% of prescribed medication and was determined at 4, 8, and 12 weeks by counting pills not taken. The protocol was approved by an institutional review board, and each patient gave written, informed consent.
End Point The primary end point of the study was the development of a gastric ulcer at the time of any of the follow-up endoscopic procedures. A gastric ulcer was defined as a circumscribed break in the gastric mucosa of 0.3 cm in diameter or greater. Additionally, the frequency of adverse experiences was assessed by the clinical investigator. Statistics Treatment group comparability at baseline was assessed by evaluating the hypothesis of equivalent treatment group means using analysis of variance for the following factors: age, height, weight, temperature, heart rate, and blood pressure. This method was also used to compare the mean change from baseline at the end of the trial between the treatment groups, with respect to each of the continuous laboratory variables. The Pearson chi-square test or Fisher exact test was used to compare the proportion of patients in the two treatment groups who developed a gastric ulcer during trial participation, with respect to both intent-to-treat and evaluable patient groups. The gastric ulceration rate was analyzed using survival methods. Life-table estimates of the ulcer-free rate were computed and the log rank test was used to compare treatments. Additionally, ulcer rates were analyzed by size at the 12-week 196
Table 1. Demographic and Clinical Characteristics of the Intent-to-Treat Cohort (All Randomized Patients) Characteristic
Misoprostol Group
Sucralfate Group
Patients, n Median age, (range) y Women:men, % Race, % White Nonwhite NSAID* used, % Ibuprofen Naproxen Piroxicam Smokers (> 10 cigarettes/d), % Alcohol use, % Previous antacid use, % Duration of osteoarthritis, y Duration of NSAID use, v History of ulcer disease, %
179 60 (30 to 82) 56:44
177 60 (29 to 82) 59:41
89 11
84 16
35 40 25 14 18 84 9.4 1.7 22.4
41 31 28 15 23 80 10.1 1.4 26.6
* NSAID = nonsteroidal anti-inflammatory drug.
period using the Fisher exact test. The time until gastric ulcer occurrence was analyzed by fitting a Cox proportional hazards regression model with the following factors included as covariates: treatment group, aluminum hydroxide use, NSAID type, baseline erosion, baseline gastritis, baseline abnormal mucosa, history of ulcer disease including gastritis, history of ulcer disease excluding gastritis, and history of gastritis. The frequency of endoscopically abnormal mucosa, "gastritis," and erosions was analyzed at each time point using the chi-square test. Statistical significance was assessed at the 0.05 level. All reported P values are two-sided. Results Four hundred three patients underwent baseline endoscopy screening, of whom 47 (11.7%) were found to have an active ulcer and were excluded from the study. However, the screening data possibly under-report the point prevalence of gastric ulcers because several investigators were screening patients for other trials. Of the remaining 356 patients, 179 (50%) were randomized to receive misoprostol and 177 (50%) were randomized to receive sucralfate. A total of 352 patients took at least one dose of the study medication (176 in each group) and therefore qualified for analysis of the frequency of adverse experiences. Three patients in the misoprostol group and one patient in the sucralfate group did not take any medication after randomization. One hundred twenty-two of the 179 patients randomized to receive misoprostol and 131 of the 177 randomized to the sucralfate group were evaluable for efficacy analysis. Among the patients who were randomized (the intentto-treat cohort), there were no statistically significant differences between the study groups with respect to demographic or clinical characteristics (Table 1). The median age for all study participants was 60 (range 29 to 82 years), with two thirds of the patients over 55 years of age. Fifty-eight percent were women and 86% were white. Approximately 20% of the patients reported consumption of alcohol, and 15% smoked more than 10 cigarettes per day. The mean reported duration of osteoarthritis was 9.7 years and of NSAID use, 1.6 years. The proportion of patients using each of the three
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
NSAIDs was comparable: 38% ibuprofen, 26% piroxicam, and 36% naproxen. The proportions within each treatment group were also comparable. In the misoprostol group, 35% were using ibuprofen; 40%, naproxen; and 25%, piroxicam. In the sucralfate group, 41% were using ibuprofen; 31%, naproxen; and 28%, piroxicam. A history of ulcer disease was reported by 25% of patients randomized. Of the 356 patients, 253 could be evaluated using life-table analysis. One hundred and three patients were not evaluable for efficacy: 57 in the misoprostol group and 46 in sucralfate group (P = 0.2). In the misoprostol group, 43 of these 57 patients (75%) were noncompliant, 7 were lost to follow-up, and 7 had a protocol violation. In the sucralfate group, 41 of the 46 patients (89%) were noncompliant, 4 were lost to follow-up, and 1 had a protocol violation. There were no clinically significant differences in clinical or demographic characteristics between the groups in the evaluable cohort or in the patients who dropped out of the study. Prevention of Nonsteroidal Anti-inflammatory Drug-induced Ulcers Analyzing the 253 evaluable patients who completed the study, 2 of 122 (1.6%; CI, 0.3% to 6.4%) patients randomized to receive misoprostol developed a gastric ulcer compared with 21 of 131 (16%; CI, 10.4% to 23.7%) patients randomized to receive sucralfate over the 3-month study period (P < 0.001; Figure 1). The difference between groups in the frequency of gastric ulcer development (14.4%; CI, 10.4% to 19.5%) remained significant when the analysis was restricted to ulcers equal to or greater than 0.5 cm in diameter (8.4%; 95% CI, 5.4% to 12.7%). The frequency of gastric ulcers of 0.5 cm or greater in patients randomized to receive misoprostol was 0.8% (CI, 0.04% to 5.1%) compared with 9.2% (CI, 5.1% to 15.8%) in patients randomized to receive sucralfate (P = 0.003; Figure 1). Figure 2 shows the probability of being free of a gastric ulcer as a function of the treatment interval using lifetable analysis of the 253 evaluable patients. When results were analyzed based on the principle of intention-to-treat, 5 of 179 (3%) patients randomized to misoprostol and 25 of 177 (14%) of patients randomized to sucralfate developed a gastric ulcer (P < 0.005). Of the 356 patients who were randomized, 292 (82%) used some aluminum hydroxide tablets during the first week for treatment of ongoing upper gastrointestinal pain. There was no statistical difference between treatment groups in the proportion of patients taking aluminum hydroxide (P > 0.2). Further, the use of aluminum hydroxide was not predictive of the subsequent development of a gastric ulcer (P > 0.2). In patients randomized to receive misoprostol, 5 of 151 (3%) patients who took aluminum hydroxide developed a gastric ulcer, whereas none of the 28 patients who did not take aluminum hydroxide developed a gastric ulcer. In the sucralfate group, 19 of the 141 (14%) patients who took aluminum hydroxide during the first week developed a gastric ulcer, compared to 6 of the 36 patients (17%) who did not take aluminum hydroxide during the first week.
A Cox proportional hazards model (stepwise-forward) was fit to the time to gastric ulceration using peptic ulcer disease history and the presence of gastric erosions at the time of initial screening endoscopy among the covariates. Treatment group entered the model at step 1 (improvement chi-square P < 0.001; global chisquare P < 0.001) and baseline erosion entered at step 2 (improvement chi-square P = 0.02; global chi-square P < 0.001). The coefficient associated with the treatment term, comparing sucralfate to misoprostol, is 2.4 (CI, 0.9 to 3.8). The adjusted hazard (risk) ratio is then 10.9 (CI, 2.6 to 44.7). The graph of the log negative-log survival function stratified by treatment group is roughly parallel. Therefore, the proportional hazards assumption is satisfied. A history of peptic ulcer disease was not associated with an increased risk of gastric ulcer formation among the evaluable cohort (P > 0.2). Twenty-six of the 122 (21%) evaluable misoprostol-treated patients and 35 of the 131 (27%) sucralfate-treated patients reported one previous episode of ulcer disease. One of the two patients who developed a gastric ulcer while taking misoprostol had a history of a previous ulcer, and 3 of the 21 patients who developed a gastric ulcer while taking sucralfate had a history of a previous ulcer. Baseline gastric erosions in the evaluable cohort were present in 47 of 122 (39%) patients randomized to misoprostol and in 50 of 131 (38%) patients randomized to sucralfate. The frequency of erosions decreased over the 3-month period in the misoprostol group and remained unchanged in the sucralfate-treated patients. Erosions at baseline were found to be significantly related to ulcer development, regardless of the treatment
Figure 1. Percentage of patients with gastric ulcers as a function of ulcer diameter. Among sucralfate-treated patients, 16.0% (95% CI, 10.4% to 23.7%) had an ulcer >0.3 cm in diameter compared with 1.6% (CI, 0.3% to 6.4%) in the misoprostol group (P < 0.001). Ulcers > 0.5 cm in diameter developed in 9.2% (CI, 5.1% to 15.8%) of the sucralfate group compared with 0.8% (CI, 0.04% to 5.1%) of the misoprostol group (P = 0.003. P values are derived from the methods of Fleiss).
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
197
group. The overall relative odds ratio, which includes all evaluable patients in both the misoprostol and the sucralfate groups, was 2.75 (odds of ulcer with erosions divided by odds of ulcer without erosions; CI, 1.14 to 6.63). The presence of gastric erosions at the time of initial endoscopy was associated with an increased risk of gastric ulcer development (P = 0.02). Thirteen of 50 (26%) patients in the sucralfate group, who had erosions at baseline endoscopy, developed a gastric ulcer compared with 8 of 81 (10%) patients without initial erosions. A gastric ulcer developed in 1 of the 47 misoprostol-treated patients who had gastric erosions at baseline endoscopy compared with 1 of 75 misoprostoltreated patients who had no baseline gastric erosions. The assessment of duodenal ulcer development was not a primary objective of this prospective study. However, a review of the endoscopy results from the 253 evaluable patients revealed only three duodenal ulcers: two in the misoprostol-treated group and one in the sucralfate-treated group. Frequency of Adverse Experiences The most frequent adverse effect was dyspepsia, which was reported by 55 (31%) patients in the misoprostol group and 42 (24%) in the sucralfate group (P = 0.1). Overall, diarrhea developed in 45 (26%) patients randomized to receive misoprostol and in 9 (5%) patients randomized to receive sucralfate (P < 0.01). Sixty-eight percent of the diarrheal episodes occurred in the first 2 weeks of misoprostol therapy. The episodes of diarrhea in the misoprostol group were mild or moderate in 79% of patients. Twenty-one percent of the
Figure 2. Percentage of patients without a nonsteroidal antiinflammatory drug-induced gastric ulcer as determined by lifetable analysis. At 3 months, the probability of being free of gastric ulcer was 97.2% (95% CI, 93.3% to 100%) in the misoprostol-treated group and 80.8% (CI, 73.0 to 88.5%) in the sucralfate-treated group. * = significantly shorter time to ulcer development compared with misoprostol (log rank P < 0.001). 198
episodes were classified as severe, but the drop-out rate attributed to diarrhea was only 3% (Table 2). Thirty-one of the 176 (18%) patients randomized to receive misoprostol and 16 of the 176 (9%) patients randomized to receive sucralfate, who took at least one dose of study medication, terminated the study because of an adverse event (P = 0.02). Among patients randomized to receive misoprostol, 15 (9%) patients stopped the study because of dyspepsia and 5 (3%) patients terminated the study because of diarrhea. In patients randomized to sucralfate, 10 (6%) patients terminated the study because of ongoing dyspepsia and one patient terminated the study because of diarrhea. A total of 14 patients (seven in each treatment group) reported adverse experiences that were regarded as potentially serious. However, none of these events resulted in death. Chest pain was the most frequently occurring serious event (two patients in the misoprostol group and one in the sucralfate group). There were no clinically significant differences between treatment groups in changes in any of the laboratory values during the study (P > 0.2 for all comparisons). Discussion Two anti-ulcer drugs were compared for their ability to prevent gastric ulcers in patients with osteoarthritis who were receiving chronic NSAID therapy. Misoprostol is a synthetic prostaglandin E, analog that has been proved to prevent NSAID-induced gastric ulcer in arthritic patients (9). Although the mechanism is still unclear, evidence is mounting that one major effect is through the replacement of mucosal prostaglandins (1720). Misoprostol decreases gastric acid secretion, stimulates the production and release of bicarbonate and mucus, and maintains mucosal blood flow. It had been suggested that sucralfate might be useful for this indication because of its ability to reduce mucosal damage from various irritants, including ethanol (21) and NSAIDs (15, 22, 23). The current study indicates that in patients with osteoarthritis and abdominal pain who are receiving NSAIDs, use of misoprostol for 3 months is associated with a significantly lower frequency of gastric ulcer compared with use of sucralfate. Larger studies are needed to determine whether the lower frequency of gastric ulcer development associated with the use of misoprostol leads to a lower frequency of complications, such as upper gastrointestinal hemorrhage or perforation. This study did not address the issue of which NSAID users are at risk for ulcer development and should receive prophylaxis. However, Fries and coworkers have examined patients at risk of NSAID-induced gastric ulcer and the patients within that group who are at particularly high risk (4). They reported that patients at high risk for hospitalization and death from gastrointestinal complications are characteristically older, have had previous upper abdominal pain, have previously stopped taking NSAIDs because of adverse gastrointestinal side effects, have previously used antacids or H 2 -receptor antagonists for gastrointestinal side effects, and are often taking corticosteroids. Fries and associates (4) noted that a patient with any two major
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
Table 2. Frequency of the Most Commonly Reported Adverse Experiences and the Frequency with Which They Led to Withdrawal from the Trial Event Misoprostol (n = 176)
Frequency of Event* Sucralfate (n = 176)
P Value
Misoprostol (n = 176)
n(%) Dyspepsia Diarrhea Nausea Constipation
55(31.3) 45 (25.6) 9(5.1) 5 (2.8)
Frequency of Termination Sucralfate (n = 176)
P Value
n(%) 42 (23.9) 9(5.1) 18(10.2) 8 (4.5)
0.1 < 0.01 0.07 >0.2
15 (8.5) 5 (2.8) 3(1.7) 1 (0.6)
10 (5.7) 1 (0.6) 1 (0.6) 1 (0.6)
>0.2 0.2 >0.2 >0.2
* Includes all patients who took at least one dose of study medication. Three patients randomized to misoprostol and one patient randomized to sucralfate never took study medication. Patients who experienced more than one episode of an event were counted only once.
risk factors may reasonably be expected to have at least two to three times the risk for serious complications compared with a patient with no risk factors, whereas a patient with four or more risk factors may be at extremely high risk. In short-term studies done in animals, sucralfate has been shown to increase mucosal prostaglandin synthesis (24, 25). This effect is believed to be nonspecific and due to minor epithelial damage caused by the aluminum present in sucralfate. Although this mechanism may prevent damage in animals pretreated with sucralfate, it would be unlikely to occur during chronic NSAID administration, because the enhanced mucosal prostaglandin synthesis induced by sucralfate is effectively abolished by NSAID pretreatment (15, 16). In a clinical study involving arthritic patients, sucralfate was found to be no better than placebo in the treatment of NSAID-induced gastrointestinal injury (26). In another short-term (2-week) study, sucralfate provided no gastric protection against aspirin-induced mucosal damage in healthy human volunteers (27). A clinical pharmacologic study performed in healthy subjects receiving aspirin (1 g four times daily for 6 days) found misoprostol to be significantly more effective than sucralfate or placebo in protecting the gastroduodenal mucosa (28). The influence of certain risk factors on gastrointestinal toxicity associated with the use of NSAIDs has been well established. A history of peptic ulcer disease has been shown in some studies to be associated with an increased incidence of NSAID-induced gastrointestinal mucosal injury (4, 29). Because our study excluded patients with a recurrent history of peptic ulcer, the effect of this risk factor on the frequency of ulcer formation could not be determined. However, 24.5% of the evaluable patients did report one previous episode of ulcer disease. One of the two patients on misoprostol and 3 of the 21 patients on sucralfate who developed a gastric ulcer had a history of a previous ulcer. There was no statistically significant association between ulcer history and the development of gastric ulcer during the study in the evaluable cohort (P > 0.2). The presence of gastric erosions has never been analyzed as a potential risk factor for the development of a gastric ulcer in chronic NSAID users. In our study, the presence of endoscopically documented gastric mucosal erosions at the time of baseline endoscopic evaluation was associated with an increased risk of the subsequent development of a gastric ulcer.
Neither form of treatment ameliorated the problem of dyspepsia. Dyspepsia occurred in 31% and 24% of the patients randomized to receive misoprostol and sucralfate, respectively. Ongoing dyspepsia was the most important reason for patient withdrawal in both treatment groups. Dyspepsia, however, may not have represented a true adverse effect because it was also an entry criterion. It is thus impossible to distinguish dyspepsia related to NSAID use from that possibly related to use of the study medication. Further, our study was not designed to evaluate the effects of the study medications on the relief of symptoms because it was singleblind. In NSAID-induced ulcer prevention studies, misoprostol was found to be associated with diarrhea and abdominal cramping early in the course of therapy (9). In our study, diarrhea occurred with greater frequency in patients receiving misoprostol compared with sucralfate (26% and 5%, respectively). In most instances, however, the diarrhea was transient and disappeared despite continued drug administration. Diarrhea is a relatively common side effect associated with initiation of misoprostol therapy. Nausea and constipation occurred at a slightly higher frequency with sucralfate than with misoprostol. There was no statistically significant difference (P > 0.2) between the treatment groups in the proportion of patients withdrawing due to dyspepsia, diarrhea, nausea, or constipation (Table 2). In conclusion, the administration of misoprostol, 200 jug four times a day for 3 months is more effective than sucralfate, 1 g four times a day, in the prevention of gastric ulceration in patients with osteoarthritis receiving chronic NSAID therapy. This study represents the work of a nationwide study group, which, in addition to the authors, includes the following investigators: Richard Aaronson, MD, Chicago Heights, Illinois; Alphonso Belsito, MD, Bradenton, Florida; Jacques R. Caldwell, MD, Gainesville, Florida; Don E. Cheatum, MD, Dallas, Texas; Robert E. Ettlinger, MD, Tacoma, Washington; Edward Fudman, MD, Austin, Texas; Oren B. Gum, MD, New Orleans, Louisiana; Richard Jaszewski, MD, Allen Park, Michigan; Abraham Kolodny, MD, Baltimore, Maryland; Pamela Prete, MD, Long Beach, California; Martin Lidsky, MD, Houston, Texas; Jeffrey Lisse, MD, Galveston, Texas; Maren Mahowald, MD, Minneapolis, Minnesota; R. K. Marwah, MD, El Paso, Texas; Ronald Messner, MD, Minneapolis, Minnesota; Jehangir Rao, MD, Wayne, Michigan; William Tatum, MD, Oklahoma City, Oklahoma; Elizabeth Tindall, MD, Portland, Oregon; Robert Trapp, MD, Springfield, Illinois; J. P. Waring, MD, Phoenix, Arizona. Grant Support: By a grant from G. D. Searle & Company, which provided the study design and data analyses. Current Author Addresses: Dr. Agrawal: Tulane University School of Medicine, Section of Gastroenterology, 1430 Tulane Avenue, New Orleans, LA 70112.
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3 Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
199
Dr. Brown: University of Illinois Medical Center, 1740 W. Taylor, Chicago, IL 60612. Dr. Germain: University of South Florida Medical Center, Division of Rheumatology, 12901 Bruce B. Downs Boulevard, Box 19, Tampa, FL 33612. Dr. Graham: Baylor Health Science Center/Veterans Affairs Medical Center, 2002 Holcombe Boulevard (HID), Houston, TX 77211. Dr. Roth: Arizona Arthritis Research & Education, Ltd., 3330 North 2nd Street, Phoenix, AZ 85012. Dr. Stromatt: Chicago Medical School, 3333 Green Bay Road, North Chicago, IL 60064. Dr. White: University of California, Davis, Department of Internal Medicine, 2221 Stockton Boulevard, Sacramento, CA 95817.
14.
15.
16.
17. 18.
References 1. Griffin MR, Ray WA, Schaffner W. Nonsteroidal anti-inflammatory drug use and death from peptic ulcer in elderly persons. Ann Intern Med. 1988;109:359-63. 2. Armstrong CP, Blower AL. Nonsteroidal anti-inflammatory drugs and life threatening complications of peptic ulceration. Gut. 1987; 28:527-32. 3. Graham DY, Smith JL. Gastroduodenal complications of chronic NSAID therapy. Am J Gastroenterol. 1988;83:1081-4. 4. Fries JF, Miller SR, Spitz PW, Williams CA, Hubert HB, Bloch DA. Toward an epidemiology of gastropathy associated with nonsteroidal anti-inflammatory drug use. Gastroenterology 1989;96:647-55. 5. Abramson SB, Weissmann G. The mechanisms of action of nonsteroidal antiinflammatory drugs. Arthritis Rheum. 1989;32:1-9. 6. Dajani EZ. Overview of the mucosal protective effects of misoprostol in man. Prostaglandins. 1987;33(Suppl): 117-29. 7. Miller TA. Protective effects of prostaglandins against gastric mucosal damage: current knowledge and proposed mechanisms. Am J Physiol. 1983;245:G601-23. 8. Agrawal NM, Dajani EZ. Treatment and prevention of nonsteroidal anti-inflammatory drug-induced gastrointestinal mucosal injury. Clin Pharmacol Res. 1989;9:347-57. 9. Graham DY, Agrawal NM, Roth SH. Prevention of NSAID-induced gastric ulcer with misoprostol: multicentre, double-blind, placebocontrolled trial. Lancet. 1988;2:1277-80. 10. Lanza FL, Fakouhi D, Rubin A, et al. A double-blind, placebocontrolled comparison of the efficacy and safety of 50, 100, and 200 fig of misoprostol QID in the prevention of ibuprofen-induced gastric and duodenal mucosal lesion and symptoms. Am J Gastroenterol. 1989;84:633-6. 11. Lanza FL, Aspinall R, Swabb EA, Davis RE, Rack MF, Rubin A. Double-blind, placebo-controlled endoscopic comparison of the mucosal protective effects of misoprostol versus cimetidine on tolmetin-induced mucosal injury to the stomach and duodenum. Gastroenterology. 1988;95:289-94. 12. Geis S. Prevention of diclofenac-induced gastroduodenal mucosal damage by Cytotec [Abstract]. Clin Exp Rheumatol. 1990;S:4. 13. Roth S, Agrawal N, Mahowald M, et al. Misoprostol heals gas-
200
1 August 1991 • Annals
of Internal
Medicine
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
troduodenal injury in patients with rheumatoid arthritis receiving aspirin. Arch Intern Med. 1989;146:775-9. Bussiere JL, Naudin R, Etherenaux C, Caprun MN. Effect of misoprostol on the anti-inflammatory activity of NSAIDs. [Abstract]. Clin Exp Rheumatol. 1990;8:S-4. Ligumsky M, Karmeli F, Rachmilewitz D. Sucralfate protection against gastrointestinal damage: possible role of prostanoids. Isr J Med Sci. 1986;22:801-6. Hollander D, Tarnawski A, Gergely H, Zipser RD. Sucralfate protection of the gastric mucosa against ethanol-induced injury: a prostaglandin-mediated process? Scand J Gastroenterol. 1984;19(Suppl 101):97-102. Collins PW. Misoprostol: discovery, development, and clinical applications Med Res Rev. 1990;10:149-72. Wilson DE, Quadros E, Rajapaska T, Adams A, Noar M. Effects of misoprostol on gastric acid and mucus secretion in man. Dig Dis Sci. 1986;31(Suppl 2):126S-9S. Sato N, Kawano S, Fukuda M, Tsuji S, Kamada T. Misoprostolinduced changes in gastric mucosal hemodynamics. A double-blind parallel study in human volunteers. Am J Med. 1987;83(Suppl A): 15-21. Isenberg JI, Hogan DL, Selling JA, Koss MA. Duodenal bicarbonate secretion in humans: role of prostaglandins. Dig Dis Sci. 1986; 31(Suppl 2):130S. Cohen MM, Bowdler R, Gervais P, Morris GP, and Wang HR. Sucralfate protection of human gastric mucosa against acute ethanol injury. Gastroenterology. 1989;86:292-8. Konturek SJ, Kwiecien N, Obtulowicz W, Kopp B, Oleksy J. Double blind controlled study on the effect of sucralfate on gastric prostaglandin formation and microbleeding in normal and aspirin treated man. Gut. 1986;27:1450-6. Waisman Y, Zahavi I, Marcus H, Ligumsky M, Rosenback Y, Dinari G. Sucralfate is protective against indomethacin-induced intestinal ulceration in the rat. Digestion. 1988;41:78-82. Tarnawski A, Hollander D, Krause WJ, Zipser RD, Stachura J, Gergely H. Does sucralfate affect the normal gastric mucosa? Histologic, ultrastructural, and functional assessment in the rat. Gastroenterology. 1986;90:893-905. Wallace JL, Morris GP, Beck PL, Williamson TE, Gingras GR. Effects of sucralfate on gastric prostaglandin and leukotriene synthesis: relationship to protective actions. Can J Physiol Pharmacol. 1988;66:666-70. Caldwell JR, Roth SH, Wu WC, et al. Sucralfate treatment of nonsteroidal anti-inflammatory drug-induced gastrointestinal symptoms and mucosal damage. Am J Med. 1987;83(Suppl 3B):74-82. Stem Al, Ward F, Sievert W. Lack of gastric mucosal protection by sucralfate during long-term aspirin ingestion in humans. Am J Med. 1986;86(Suppl 6A):66-9. Lanza F, Peace K, Gustitus L, Rack MF, Dickson B. A blinded endoscopic comparative study of misoprostol versus sucralfate and placebo in the prevention of aspirin-induced gastric and duodenal ulceration. Am J Gastroenterol. 1988;83:143-6. Roth SH. NSAIDs and gastropathy: a rheumatologist's review. J Rheumatol. 1988;15:912-9.
• Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
• Objectives: To compare the efficacy and frequency of adverse experiences of misoprostol and sucralfate in the prevention of gastric ulcers in patients receiving nonsteroidal anti-inflammatory drug (NSAID) therapy. • Design: A prospective, randomized, single-blind, multicenter trial. • Patients: Patients with osteoarthritis receiving treatment with ibuprofen, piroxicam, or naproxen and experiencing abdominal pain were eligible. • Interventions: Patients who were expected to receive at least 3 months of NSAID therapy and who did not have a gastric ulcer at the time of the initial screening endoscopy were randomized to receive misoprostol, 200 |xg four times a day, or sucralfate, 1 g four times a day. A gastric ulcer was defined as a lesion of the gastric mucosa 0.3 cm or greater in diameter. Patients were followed clinically, and repeat endoscopies were performed after 4, 8, and 12 weeks. • Main Measurement: The development of a gastric ulcer, which was regarded as a prophylaxis failure. • Results: Two hundred fifty-three patients were evaluable for efficacy analysis. A gastric ulcer developed in 2 of the 122 (1.6%, 95% CI, 0.3% to 6.4%) patients on misoprostol, compared with 21 of 131 patients on sucralfate (16%, CI, 10.4% to 23.7%). The difference in ulcer rates was 14.4% (CI, 10.4% to 19.5%; P < 0.001). • Conclusion: In patients receiving chronic NSAID therapy for osteoarthritis, treatment with misoprostol for 3 months was associated with a significantly lower frequency of gastric ulcer formation, compared with treatment with sucralfate (P< 0.001).
Annals of Internal Medicine. 1991;115:195-200. From Tulane University School of Medicine, New Orleans, Louisiana; Arizona Arthritis Research & Education, Ltd., Phoenix, Arizona; Baylor College of Medicine/Veterans Affairs Medical Center, Houston, Texas; University of California, Davis and Sacramento, California; University of South Florida, Tampa, Florida; University of Illinois, Chicago, Illinois; and University of Health Sciences/Chicago Medical School, Chicago, Illinois. For a list of additional study investigators and for current author addresses, see end of text.
W onsteroidal anti-inflammatory drugs (NSAIDs) are an integral part of the therapy of rheumatic diseases. These drugs can damage the gastrointestinal tract and have been implicated as a cause of peptic ulceration and life-threatening bleeding (1-4). The inhibitory effect of NSAIDs on the endogenous biosynthesis of prostaglandins is thought to be the major mechanism for the therapeutic properties of NSAIDs in the treatment of inflammatory arthritis, although other mechanisms have recently been proposed (5). Prostaglandins play a significant role in the defense of gastrointestinal mucosa (6, 7). Prostaglandins of the E series have gastric antisecretory and mucosal protective properties, and NSAID-induced inhibition of gastric mucosal prostaglandin synthesis is thought to be responsible for much of the gastrointestinal tract toxicity associated with these agents (3, 6-8). Misoprostol, a synthetic prostaglandin E, analog, has been shown to be effective in the prevention of NSAIDinduced gastric ulcers in patients with arthritis (9). This agent has also been shown to prevent the development of NSAID-induced duodenal lesions both in normal subjects (10, 11) and in arthritic patients (12). The beneficial effects of misoprostol on the gastrointestinal tract are not accompanied by a compromise of the antirheumatic effects of the NSAID (13, 14). Sucralfate is an effective anti-ulcer drug that has been shown to increase the release of endogenous prostaglandins from the gastric mucosa (15, 16), suggesting that stimulation of prostaglandin synthesis may be a mechanism for its action. Despite the fact that sucralfate is widely used to prevent NSAID-induced gastrointestinal mucosal damage, no study has evaluated the efficacy of sucralfate in the prevention of NSAID-induced mucosal damage in chronic NSAID users. We compared misoprostol and sucralfate in the prevention of NSAID-induced gastric ulcer in a large cohort of patients with osteoarthritis who were experiencing upper gastrointestinal pain in association with the use of NSAID therapy. Methods Patients were recruited from private practice offices and clinics, Veterans Affairs clinics, health maintenance organizations, and academic institutions. Patients were eligible to enter the study if they had osteoarthritis and were experiencing upper abdominal pain that was thought to be caused by one of three NSAIDs: ibuprofen, piroxicam, or naproxen. In addition,
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
195
it was required that all patients be expected to receive NSAIDs for at least 3 additional months. Women had to be postmenopausal, surgically sterilized, or practicing adequate contraception. Women of child-bearing potential were informed of the possibility that use of misoprostol could "result in a miscarriage. A pregnancy test was done 72 hours before receiving the first dose of study medication; if the pregnancy test was positive, the patient was excluded. Exclusion criteria included a history of recurrent peptic ulcer disease, active bleeding ulcer, upper gastrointestinal malignancy or metastasis to the upper gastrointestinal tract, pyloric or duodenal obstruction, acute hepatitis, pancreatitis, inflammatory bowel disease, bleeding diathesis, upper gastrointestinal surgery within 30 days, or severe renal impairment. Patients taking antineoplastic drugs, anticoagulants, or anti-ulcer drugs, other than the study drugs, were excluded. Study Design The study was a prospective, randomized, single-blind, multicenter comparison of misoprostol, 200 /xg given four times a day (with meals and at bedtime) and sucralfate, 1 g four times a day (30 minutes before meals and at bedtime). Patients were referred for abdominal pain that was thought by the investigators to be due to NSAID therapy. Patients were entered consecutively into the study. Eligible patients underwent an initial screening upper gastrointestinal endoscopy. Patients in whom a gastric or duodenal ulcer was found were excluded from the study. Patients without a gastric ulcer were randomized and started on a study drug within 72 hours of the entry endoscopic examination. Patients continued to take NSAIDs at the same dose that was administered before the study. Although the patients in the study could have possibly recognized their study medication, the endoscopists were blinded to the medications being taken by the patients. In addition, patients were instructed not to discuss their study medications or symptoms with the endoscopists. The gastric mucosa was examined by fiberoptic endoscopy after 4 weeks (± 3 days), 8 weeks (± 5 days), and 12 weeks (± 5 days) after randomization. Patients were allowed to take up to four aluminum hydroxide antacid tablets per day, during the first week only, for relief of upper gastrointestinal pain. Patients were instructed to take the study medication the night before the endoscopy and not to take the next dose until after the procedure was completed. Noncompliance was defined as failure to take at least 75% of prescribed medication and was determined at 4, 8, and 12 weeks by counting pills not taken. The protocol was approved by an institutional review board, and each patient gave written, informed consent.
End Point The primary end point of the study was the development of a gastric ulcer at the time of any of the follow-up endoscopic procedures. A gastric ulcer was defined as a circumscribed break in the gastric mucosa of 0.3 cm in diameter or greater. Additionally, the frequency of adverse experiences was assessed by the clinical investigator. Statistics Treatment group comparability at baseline was assessed by evaluating the hypothesis of equivalent treatment group means using analysis of variance for the following factors: age, height, weight, temperature, heart rate, and blood pressure. This method was also used to compare the mean change from baseline at the end of the trial between the treatment groups, with respect to each of the continuous laboratory variables. The Pearson chi-square test or Fisher exact test was used to compare the proportion of patients in the two treatment groups who developed a gastric ulcer during trial participation, with respect to both intent-to-treat and evaluable patient groups. The gastric ulceration rate was analyzed using survival methods. Life-table estimates of the ulcer-free rate were computed and the log rank test was used to compare treatments. Additionally, ulcer rates were analyzed by size at the 12-week 196
Table 1. Demographic and Clinical Characteristics of the Intent-to-Treat Cohort (All Randomized Patients) Characteristic
Misoprostol Group
Sucralfate Group
Patients, n Median age, (range) y Women:men, % Race, % White Nonwhite NSAID* used, % Ibuprofen Naproxen Piroxicam Smokers (> 10 cigarettes/d), % Alcohol use, % Previous antacid use, % Duration of osteoarthritis, y Duration of NSAID use, v History of ulcer disease, %
179 60 (30 to 82) 56:44
177 60 (29 to 82) 59:41
89 11
84 16
35 40 25 14 18 84 9.4 1.7 22.4
41 31 28 15 23 80 10.1 1.4 26.6
* NSAID = nonsteroidal anti-inflammatory drug.
period using the Fisher exact test. The time until gastric ulcer occurrence was analyzed by fitting a Cox proportional hazards regression model with the following factors included as covariates: treatment group, aluminum hydroxide use, NSAID type, baseline erosion, baseline gastritis, baseline abnormal mucosa, history of ulcer disease including gastritis, history of ulcer disease excluding gastritis, and history of gastritis. The frequency of endoscopically abnormal mucosa, "gastritis," and erosions was analyzed at each time point using the chi-square test. Statistical significance was assessed at the 0.05 level. All reported P values are two-sided. Results Four hundred three patients underwent baseline endoscopy screening, of whom 47 (11.7%) were found to have an active ulcer and were excluded from the study. However, the screening data possibly under-report the point prevalence of gastric ulcers because several investigators were screening patients for other trials. Of the remaining 356 patients, 179 (50%) were randomized to receive misoprostol and 177 (50%) were randomized to receive sucralfate. A total of 352 patients took at least one dose of the study medication (176 in each group) and therefore qualified for analysis of the frequency of adverse experiences. Three patients in the misoprostol group and one patient in the sucralfate group did not take any medication after randomization. One hundred twenty-two of the 179 patients randomized to receive misoprostol and 131 of the 177 randomized to the sucralfate group were evaluable for efficacy analysis. Among the patients who were randomized (the intentto-treat cohort), there were no statistically significant differences between the study groups with respect to demographic or clinical characteristics (Table 1). The median age for all study participants was 60 (range 29 to 82 years), with two thirds of the patients over 55 years of age. Fifty-eight percent were women and 86% were white. Approximately 20% of the patients reported consumption of alcohol, and 15% smoked more than 10 cigarettes per day. The mean reported duration of osteoarthritis was 9.7 years and of NSAID use, 1.6 years. The proportion of patients using each of the three
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
NSAIDs was comparable: 38% ibuprofen, 26% piroxicam, and 36% naproxen. The proportions within each treatment group were also comparable. In the misoprostol group, 35% were using ibuprofen; 40%, naproxen; and 25%, piroxicam. In the sucralfate group, 41% were using ibuprofen; 31%, naproxen; and 28%, piroxicam. A history of ulcer disease was reported by 25% of patients randomized. Of the 356 patients, 253 could be evaluated using life-table analysis. One hundred and three patients were not evaluable for efficacy: 57 in the misoprostol group and 46 in sucralfate group (P = 0.2). In the misoprostol group, 43 of these 57 patients (75%) were noncompliant, 7 were lost to follow-up, and 7 had a protocol violation. In the sucralfate group, 41 of the 46 patients (89%) were noncompliant, 4 were lost to follow-up, and 1 had a protocol violation. There were no clinically significant differences in clinical or demographic characteristics between the groups in the evaluable cohort or in the patients who dropped out of the study. Prevention of Nonsteroidal Anti-inflammatory Drug-induced Ulcers Analyzing the 253 evaluable patients who completed the study, 2 of 122 (1.6%; CI, 0.3% to 6.4%) patients randomized to receive misoprostol developed a gastric ulcer compared with 21 of 131 (16%; CI, 10.4% to 23.7%) patients randomized to receive sucralfate over the 3-month study period (P < 0.001; Figure 1). The difference between groups in the frequency of gastric ulcer development (14.4%; CI, 10.4% to 19.5%) remained significant when the analysis was restricted to ulcers equal to or greater than 0.5 cm in diameter (8.4%; 95% CI, 5.4% to 12.7%). The frequency of gastric ulcers of 0.5 cm or greater in patients randomized to receive misoprostol was 0.8% (CI, 0.04% to 5.1%) compared with 9.2% (CI, 5.1% to 15.8%) in patients randomized to receive sucralfate (P = 0.003; Figure 1). Figure 2 shows the probability of being free of a gastric ulcer as a function of the treatment interval using lifetable analysis of the 253 evaluable patients. When results were analyzed based on the principle of intention-to-treat, 5 of 179 (3%) patients randomized to misoprostol and 25 of 177 (14%) of patients randomized to sucralfate developed a gastric ulcer (P < 0.005). Of the 356 patients who were randomized, 292 (82%) used some aluminum hydroxide tablets during the first week for treatment of ongoing upper gastrointestinal pain. There was no statistical difference between treatment groups in the proportion of patients taking aluminum hydroxide (P > 0.2). Further, the use of aluminum hydroxide was not predictive of the subsequent development of a gastric ulcer (P > 0.2). In patients randomized to receive misoprostol, 5 of 151 (3%) patients who took aluminum hydroxide developed a gastric ulcer, whereas none of the 28 patients who did not take aluminum hydroxide developed a gastric ulcer. In the sucralfate group, 19 of the 141 (14%) patients who took aluminum hydroxide during the first week developed a gastric ulcer, compared to 6 of the 36 patients (17%) who did not take aluminum hydroxide during the first week.
A Cox proportional hazards model (stepwise-forward) was fit to the time to gastric ulceration using peptic ulcer disease history and the presence of gastric erosions at the time of initial screening endoscopy among the covariates. Treatment group entered the model at step 1 (improvement chi-square P < 0.001; global chisquare P < 0.001) and baseline erosion entered at step 2 (improvement chi-square P = 0.02; global chi-square P < 0.001). The coefficient associated with the treatment term, comparing sucralfate to misoprostol, is 2.4 (CI, 0.9 to 3.8). The adjusted hazard (risk) ratio is then 10.9 (CI, 2.6 to 44.7). The graph of the log negative-log survival function stratified by treatment group is roughly parallel. Therefore, the proportional hazards assumption is satisfied. A history of peptic ulcer disease was not associated with an increased risk of gastric ulcer formation among the evaluable cohort (P > 0.2). Twenty-six of the 122 (21%) evaluable misoprostol-treated patients and 35 of the 131 (27%) sucralfate-treated patients reported one previous episode of ulcer disease. One of the two patients who developed a gastric ulcer while taking misoprostol had a history of a previous ulcer, and 3 of the 21 patients who developed a gastric ulcer while taking sucralfate had a history of a previous ulcer. Baseline gastric erosions in the evaluable cohort were present in 47 of 122 (39%) patients randomized to misoprostol and in 50 of 131 (38%) patients randomized to sucralfate. The frequency of erosions decreased over the 3-month period in the misoprostol group and remained unchanged in the sucralfate-treated patients. Erosions at baseline were found to be significantly related to ulcer development, regardless of the treatment
Figure 1. Percentage of patients with gastric ulcers as a function of ulcer diameter. Among sucralfate-treated patients, 16.0% (95% CI, 10.4% to 23.7%) had an ulcer >0.3 cm in diameter compared with 1.6% (CI, 0.3% to 6.4%) in the misoprostol group (P < 0.001). Ulcers > 0.5 cm in diameter developed in 9.2% (CI, 5.1% to 15.8%) of the sucralfate group compared with 0.8% (CI, 0.04% to 5.1%) of the misoprostol group (P = 0.003. P values are derived from the methods of Fleiss).
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
197
group. The overall relative odds ratio, which includes all evaluable patients in both the misoprostol and the sucralfate groups, was 2.75 (odds of ulcer with erosions divided by odds of ulcer without erosions; CI, 1.14 to 6.63). The presence of gastric erosions at the time of initial endoscopy was associated with an increased risk of gastric ulcer development (P = 0.02). Thirteen of 50 (26%) patients in the sucralfate group, who had erosions at baseline endoscopy, developed a gastric ulcer compared with 8 of 81 (10%) patients without initial erosions. A gastric ulcer developed in 1 of the 47 misoprostol-treated patients who had gastric erosions at baseline endoscopy compared with 1 of 75 misoprostoltreated patients who had no baseline gastric erosions. The assessment of duodenal ulcer development was not a primary objective of this prospective study. However, a review of the endoscopy results from the 253 evaluable patients revealed only three duodenal ulcers: two in the misoprostol-treated group and one in the sucralfate-treated group. Frequency of Adverse Experiences The most frequent adverse effect was dyspepsia, which was reported by 55 (31%) patients in the misoprostol group and 42 (24%) in the sucralfate group (P = 0.1). Overall, diarrhea developed in 45 (26%) patients randomized to receive misoprostol and in 9 (5%) patients randomized to receive sucralfate (P < 0.01). Sixty-eight percent of the diarrheal episodes occurred in the first 2 weeks of misoprostol therapy. The episodes of diarrhea in the misoprostol group were mild or moderate in 79% of patients. Twenty-one percent of the
Figure 2. Percentage of patients without a nonsteroidal antiinflammatory drug-induced gastric ulcer as determined by lifetable analysis. At 3 months, the probability of being free of gastric ulcer was 97.2% (95% CI, 93.3% to 100%) in the misoprostol-treated group and 80.8% (CI, 73.0 to 88.5%) in the sucralfate-treated group. * = significantly shorter time to ulcer development compared with misoprostol (log rank P < 0.001). 198
episodes were classified as severe, but the drop-out rate attributed to diarrhea was only 3% (Table 2). Thirty-one of the 176 (18%) patients randomized to receive misoprostol and 16 of the 176 (9%) patients randomized to receive sucralfate, who took at least one dose of study medication, terminated the study because of an adverse event (P = 0.02). Among patients randomized to receive misoprostol, 15 (9%) patients stopped the study because of dyspepsia and 5 (3%) patients terminated the study because of diarrhea. In patients randomized to sucralfate, 10 (6%) patients terminated the study because of ongoing dyspepsia and one patient terminated the study because of diarrhea. A total of 14 patients (seven in each treatment group) reported adverse experiences that were regarded as potentially serious. However, none of these events resulted in death. Chest pain was the most frequently occurring serious event (two patients in the misoprostol group and one in the sucralfate group). There were no clinically significant differences between treatment groups in changes in any of the laboratory values during the study (P > 0.2 for all comparisons). Discussion Two anti-ulcer drugs were compared for their ability to prevent gastric ulcers in patients with osteoarthritis who were receiving chronic NSAID therapy. Misoprostol is a synthetic prostaglandin E, analog that has been proved to prevent NSAID-induced gastric ulcer in arthritic patients (9). Although the mechanism is still unclear, evidence is mounting that one major effect is through the replacement of mucosal prostaglandins (1720). Misoprostol decreases gastric acid secretion, stimulates the production and release of bicarbonate and mucus, and maintains mucosal blood flow. It had been suggested that sucralfate might be useful for this indication because of its ability to reduce mucosal damage from various irritants, including ethanol (21) and NSAIDs (15, 22, 23). The current study indicates that in patients with osteoarthritis and abdominal pain who are receiving NSAIDs, use of misoprostol for 3 months is associated with a significantly lower frequency of gastric ulcer compared with use of sucralfate. Larger studies are needed to determine whether the lower frequency of gastric ulcer development associated with the use of misoprostol leads to a lower frequency of complications, such as upper gastrointestinal hemorrhage or perforation. This study did not address the issue of which NSAID users are at risk for ulcer development and should receive prophylaxis. However, Fries and coworkers have examined patients at risk of NSAID-induced gastric ulcer and the patients within that group who are at particularly high risk (4). They reported that patients at high risk for hospitalization and death from gastrointestinal complications are characteristically older, have had previous upper abdominal pain, have previously stopped taking NSAIDs because of adverse gastrointestinal side effects, have previously used antacids or H 2 -receptor antagonists for gastrointestinal side effects, and are often taking corticosteroids. Fries and associates (4) noted that a patient with any two major
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
Table 2. Frequency of the Most Commonly Reported Adverse Experiences and the Frequency with Which They Led to Withdrawal from the Trial Event Misoprostol (n = 176)
Frequency of Event* Sucralfate (n = 176)
P Value
Misoprostol (n = 176)
n(%) Dyspepsia Diarrhea Nausea Constipation
55(31.3) 45 (25.6) 9(5.1) 5 (2.8)
Frequency of Termination Sucralfate (n = 176)
P Value
n(%) 42 (23.9) 9(5.1) 18(10.2) 8 (4.5)
0.1 < 0.01 0.07 >0.2
15 (8.5) 5 (2.8) 3(1.7) 1 (0.6)
10 (5.7) 1 (0.6) 1 (0.6) 1 (0.6)
>0.2 0.2 >0.2 >0.2
* Includes all patients who took at least one dose of study medication. Three patients randomized to misoprostol and one patient randomized to sucralfate never took study medication. Patients who experienced more than one episode of an event were counted only once.
risk factors may reasonably be expected to have at least two to three times the risk for serious complications compared with a patient with no risk factors, whereas a patient with four or more risk factors may be at extremely high risk. In short-term studies done in animals, sucralfate has been shown to increase mucosal prostaglandin synthesis (24, 25). This effect is believed to be nonspecific and due to minor epithelial damage caused by the aluminum present in sucralfate. Although this mechanism may prevent damage in animals pretreated with sucralfate, it would be unlikely to occur during chronic NSAID administration, because the enhanced mucosal prostaglandin synthesis induced by sucralfate is effectively abolished by NSAID pretreatment (15, 16). In a clinical study involving arthritic patients, sucralfate was found to be no better than placebo in the treatment of NSAID-induced gastrointestinal injury (26). In another short-term (2-week) study, sucralfate provided no gastric protection against aspirin-induced mucosal damage in healthy human volunteers (27). A clinical pharmacologic study performed in healthy subjects receiving aspirin (1 g four times daily for 6 days) found misoprostol to be significantly more effective than sucralfate or placebo in protecting the gastroduodenal mucosa (28). The influence of certain risk factors on gastrointestinal toxicity associated with the use of NSAIDs has been well established. A history of peptic ulcer disease has been shown in some studies to be associated with an increased incidence of NSAID-induced gastrointestinal mucosal injury (4, 29). Because our study excluded patients with a recurrent history of peptic ulcer, the effect of this risk factor on the frequency of ulcer formation could not be determined. However, 24.5% of the evaluable patients did report one previous episode of ulcer disease. One of the two patients on misoprostol and 3 of the 21 patients on sucralfate who developed a gastric ulcer had a history of a previous ulcer. There was no statistically significant association between ulcer history and the development of gastric ulcer during the study in the evaluable cohort (P > 0.2). The presence of gastric erosions has never been analyzed as a potential risk factor for the development of a gastric ulcer in chronic NSAID users. In our study, the presence of endoscopically documented gastric mucosal erosions at the time of baseline endoscopic evaluation was associated with an increased risk of the subsequent development of a gastric ulcer.
Neither form of treatment ameliorated the problem of dyspepsia. Dyspepsia occurred in 31% and 24% of the patients randomized to receive misoprostol and sucralfate, respectively. Ongoing dyspepsia was the most important reason for patient withdrawal in both treatment groups. Dyspepsia, however, may not have represented a true adverse effect because it was also an entry criterion. It is thus impossible to distinguish dyspepsia related to NSAID use from that possibly related to use of the study medication. Further, our study was not designed to evaluate the effects of the study medications on the relief of symptoms because it was singleblind. In NSAID-induced ulcer prevention studies, misoprostol was found to be associated with diarrhea and abdominal cramping early in the course of therapy (9). In our study, diarrhea occurred with greater frequency in patients receiving misoprostol compared with sucralfate (26% and 5%, respectively). In most instances, however, the diarrhea was transient and disappeared despite continued drug administration. Diarrhea is a relatively common side effect associated with initiation of misoprostol therapy. Nausea and constipation occurred at a slightly higher frequency with sucralfate than with misoprostol. There was no statistically significant difference (P > 0.2) between the treatment groups in the proportion of patients withdrawing due to dyspepsia, diarrhea, nausea, or constipation (Table 2). In conclusion, the administration of misoprostol, 200 jug four times a day for 3 months is more effective than sucralfate, 1 g four times a day, in the prevention of gastric ulceration in patients with osteoarthritis receiving chronic NSAID therapy. This study represents the work of a nationwide study group, which, in addition to the authors, includes the following investigators: Richard Aaronson, MD, Chicago Heights, Illinois; Alphonso Belsito, MD, Bradenton, Florida; Jacques R. Caldwell, MD, Gainesville, Florida; Don E. Cheatum, MD, Dallas, Texas; Robert E. Ettlinger, MD, Tacoma, Washington; Edward Fudman, MD, Austin, Texas; Oren B. Gum, MD, New Orleans, Louisiana; Richard Jaszewski, MD, Allen Park, Michigan; Abraham Kolodny, MD, Baltimore, Maryland; Pamela Prete, MD, Long Beach, California; Martin Lidsky, MD, Houston, Texas; Jeffrey Lisse, MD, Galveston, Texas; Maren Mahowald, MD, Minneapolis, Minnesota; R. K. Marwah, MD, El Paso, Texas; Ronald Messner, MD, Minneapolis, Minnesota; Jehangir Rao, MD, Wayne, Michigan; William Tatum, MD, Oklahoma City, Oklahoma; Elizabeth Tindall, MD, Portland, Oregon; Robert Trapp, MD, Springfield, Illinois; J. P. Waring, MD, Phoenix, Arizona. Grant Support: By a grant from G. D. Searle & Company, which provided the study design and data analyses. Current Author Addresses: Dr. Agrawal: Tulane University School of Medicine, Section of Gastroenterology, 1430 Tulane Avenue, New Orleans, LA 70112.
1 August 1991 • Annals of Internal Medicine • Volume 115 • Number 3 Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
199
Dr. Brown: University of Illinois Medical Center, 1740 W. Taylor, Chicago, IL 60612. Dr. Germain: University of South Florida Medical Center, Division of Rheumatology, 12901 Bruce B. Downs Boulevard, Box 19, Tampa, FL 33612. Dr. Graham: Baylor Health Science Center/Veterans Affairs Medical Center, 2002 Holcombe Boulevard (HID), Houston, TX 77211. Dr. Roth: Arizona Arthritis Research & Education, Ltd., 3330 North 2nd Street, Phoenix, AZ 85012. Dr. Stromatt: Chicago Medical School, 3333 Green Bay Road, North Chicago, IL 60064. Dr. White: University of California, Davis, Department of Internal Medicine, 2221 Stockton Boulevard, Sacramento, CA 95817.
14.
15.
16.
17. 18.
References 1. Griffin MR, Ray WA, Schaffner W. Nonsteroidal anti-inflammatory drug use and death from peptic ulcer in elderly persons. Ann Intern Med. 1988;109:359-63. 2. Armstrong CP, Blower AL. Nonsteroidal anti-inflammatory drugs and life threatening complications of peptic ulceration. Gut. 1987; 28:527-32. 3. Graham DY, Smith JL. Gastroduodenal complications of chronic NSAID therapy. Am J Gastroenterol. 1988;83:1081-4. 4. Fries JF, Miller SR, Spitz PW, Williams CA, Hubert HB, Bloch DA. Toward an epidemiology of gastropathy associated with nonsteroidal anti-inflammatory drug use. Gastroenterology 1989;96:647-55. 5. Abramson SB, Weissmann G. The mechanisms of action of nonsteroidal antiinflammatory drugs. Arthritis Rheum. 1989;32:1-9. 6. Dajani EZ. Overview of the mucosal protective effects of misoprostol in man. Prostaglandins. 1987;33(Suppl): 117-29. 7. Miller TA. Protective effects of prostaglandins against gastric mucosal damage: current knowledge and proposed mechanisms. Am J Physiol. 1983;245:G601-23. 8. Agrawal NM, Dajani EZ. Treatment and prevention of nonsteroidal anti-inflammatory drug-induced gastrointestinal mucosal injury. Clin Pharmacol Res. 1989;9:347-57. 9. Graham DY, Agrawal NM, Roth SH. Prevention of NSAID-induced gastric ulcer with misoprostol: multicentre, double-blind, placebocontrolled trial. Lancet. 1988;2:1277-80. 10. Lanza FL, Fakouhi D, Rubin A, et al. A double-blind, placebocontrolled comparison of the efficacy and safety of 50, 100, and 200 fig of misoprostol QID in the prevention of ibuprofen-induced gastric and duodenal mucosal lesion and symptoms. Am J Gastroenterol. 1989;84:633-6. 11. Lanza FL, Aspinall R, Swabb EA, Davis RE, Rack MF, Rubin A. Double-blind, placebo-controlled endoscopic comparison of the mucosal protective effects of misoprostol versus cimetidine on tolmetin-induced mucosal injury to the stomach and duodenum. Gastroenterology. 1988;95:289-94. 12. Geis S. Prevention of diclofenac-induced gastroduodenal mucosal damage by Cytotec [Abstract]. Clin Exp Rheumatol. 1990;S:4. 13. Roth S, Agrawal N, Mahowald M, et al. Misoprostol heals gas-
200
1 August 1991 • Annals
of Internal
Medicine
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
troduodenal injury in patients with rheumatoid arthritis receiving aspirin. Arch Intern Med. 1989;146:775-9. Bussiere JL, Naudin R, Etherenaux C, Caprun MN. Effect of misoprostol on the anti-inflammatory activity of NSAIDs. [Abstract]. Clin Exp Rheumatol. 1990;8:S-4. Ligumsky M, Karmeli F, Rachmilewitz D. Sucralfate protection against gastrointestinal damage: possible role of prostanoids. Isr J Med Sci. 1986;22:801-6. Hollander D, Tarnawski A, Gergely H, Zipser RD. Sucralfate protection of the gastric mucosa against ethanol-induced injury: a prostaglandin-mediated process? Scand J Gastroenterol. 1984;19(Suppl 101):97-102. Collins PW. Misoprostol: discovery, development, and clinical applications Med Res Rev. 1990;10:149-72. Wilson DE, Quadros E, Rajapaska T, Adams A, Noar M. Effects of misoprostol on gastric acid and mucus secretion in man. Dig Dis Sci. 1986;31(Suppl 2):126S-9S. Sato N, Kawano S, Fukuda M, Tsuji S, Kamada T. Misoprostolinduced changes in gastric mucosal hemodynamics. A double-blind parallel study in human volunteers. Am J Med. 1987;83(Suppl A): 15-21. Isenberg JI, Hogan DL, Selling JA, Koss MA. Duodenal bicarbonate secretion in humans: role of prostaglandins. Dig Dis Sci. 1986; 31(Suppl 2):130S. Cohen MM, Bowdler R, Gervais P, Morris GP, and Wang HR. Sucralfate protection of human gastric mucosa against acute ethanol injury. Gastroenterology. 1989;86:292-8. Konturek SJ, Kwiecien N, Obtulowicz W, Kopp B, Oleksy J. Double blind controlled study on the effect of sucralfate on gastric prostaglandin formation and microbleeding in normal and aspirin treated man. Gut. 1986;27:1450-6. Waisman Y, Zahavi I, Marcus H, Ligumsky M, Rosenback Y, Dinari G. Sucralfate is protective against indomethacin-induced intestinal ulceration in the rat. Digestion. 1988;41:78-82. Tarnawski A, Hollander D, Krause WJ, Zipser RD, Stachura J, Gergely H. Does sucralfate affect the normal gastric mucosa? Histologic, ultrastructural, and functional assessment in the rat. Gastroenterology. 1986;90:893-905. Wallace JL, Morris GP, Beck PL, Williamson TE, Gingras GR. Effects of sucralfate on gastric prostaglandin and leukotriene synthesis: relationship to protective actions. Can J Physiol Pharmacol. 1988;66:666-70. Caldwell JR, Roth SH, Wu WC, et al. Sucralfate treatment of nonsteroidal anti-inflammatory drug-induced gastrointestinal symptoms and mucosal damage. Am J Med. 1987;83(Suppl 3B):74-82. Stem Al, Ward F, Sievert W. Lack of gastric mucosal protection by sucralfate during long-term aspirin ingestion in humans. Am J Med. 1986;86(Suppl 6A):66-9. Lanza F, Peace K, Gustitus L, Rack MF, Dickson B. A blinded endoscopic comparative study of misoprostol versus sucralfate and placebo in the prevention of aspirin-induced gastric and duodenal ulceration. Am J Gastroenterol. 1988;83:143-6. Roth SH. NSAIDs and gastropathy: a rheumatologist's review. J Rheumatol. 1988;15:912-9.
• Volume 115 • Number 3
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19731/ by a University of California San Diego User on 04/03/2017
