Several studies [4, 5, 8, 11-16] have shown that misoprostol, a synthetic analog of prostaglandin E₁, affords significant protection against NSAID-induced gastric and duodenal ulcers; the drug has also been shown to reduce the incidence of ulcer complications [17]. However, the use of misoprostol at the currently recommended dosage of 200 µg four times daily is associated with frequent side effects, which primarily affect the gastrointestinal tract. These side effects are generally mild but may lead to poor patient compliance. A lower dosage of misoprostol, such as 200 µg twice or three times daily, might result in a lower incidence of adverse events (and, hence, better compliance) without adversely affecting efficacy. We sought to determine the effectiveness and tolerability of three different misoprostol regimens.

Methods

Top Methods Results Discussion Author & Article Info References

The study was a 12-week, randomized, parallel, placebo-controlled, double-blind comparison of three regimens of misoprostol (Cytotec, G.D. Searle & Co., Chicago, Illinois) and was done at 135 centers. A single study protocol was approved by the institutional review boards for all study sites. Informed consent was obtained from all patients. Patients qualifying for study entry had a clinical diagnosis of osteoarthritis, rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, or the Reiter syndrome and were receiving NSAID therapy that was expected to continue uninterrupted for at least 3 additional months at a fixed dose. The following minimum daily NSAID dosages were required: ibuprofen, 1200 mg; piroxicam, 20 mg; naproxen, 750 mg; sulindac, 200 mg; tolmetin, 1200 mg; indomethacin, 75 mg; flurbiprofen, 200 mg; ketoprofen, 150 mg; or diclofenac, 150 mg.

Qualifying patients had to be having upper gastrointestinal symptoms, such as pain, cramps, bloating, or heartburn. Before study entry, patients supplied a medical history and had a physical examination, upper gastrointestinal symptom assessment, routine laboratory tests, and upper gastrointestinal endoscopy. Patients with a gastric or duodenal mucosal defect 0.3 cm or less in diameter, a mucosal defect of any size with perceptible depth, or any esophageal erosions or ulcers were excluded from the study. Also excluded were patients who had had upper gastrointestinal surgery within 30 days of anticipated entry into the study and patients with upper gastrointestinal malignancy, pyloric obstruction, acute hepatitis, pancreatitis, inflammatory bowel disease, or a bleeding diathesis.

Patients were assigned to their regimens according to a centralized, computer-generated randomization schedule. Each center was assigned with one or more randomization blocks of seven in sealed envelopes. Patients were assigned sequentially to receive one of four regimens: placebo four times daily; 200 µg of misoprostol twice daily plus placebo twice daily; 200 µg of misoprostol three times daily plus placebo once daily; or 200 µg of misoprostol four times daily. One patient was assigned to the group receiving misoprostol four times daily for every two patients assigned to the other groups. These assignment ratios were chosen in light of the known therapeutic effectiveness of four-times-daily dosing. They were calculated to demonstrate a reduction in the rate of gastric ulcer development from 13% in the placebo group to 4% in each of the active treatment arms (twice daily, three times daily, and four times daily), and to show a similar reduction in duodenal ulcer development from 6.3% to 1%, with an overall error rate of 0.05% and a power of 80%. Blister cards of scored tablets were supplied by the sponsor and contained tablets of misoprostol, 200 µg, or identical tablets composed of inert excipient. Labels on the blister packs indicated the breakfast, lunch, dinner, and bedtime doses.

For the first 3 days of the study, patients were instructed to take one half of one tablet four times daily and to discard the other halves of the tablets. Subsequently, full tablets were taken. Patients who missed a dose were instructed to skip that dose. Each patient was provided with antacid tablets, 84 600 mg aluminum hydroxide (Amphogel, Wyeth-Ayerst Laboratories, Philadelphia, Pennsylvania), and instructed to take as many as four tablets per day as required for relief of upper abdominal symptoms during the first 3 weeks of the study. Study medication dispensed at one visit was collected and inventoried at the next visit to ensure at least 60% compliance in the dosing of misoprostol. Patients had endoscopic and upper gastrointestinal symptom evaluations after 4, 8, and 12 weeks of therapy. Endoscopic examinations included assessment of the gastric and duodenal mucosa.

Patient Groups Analyzed

Of the 1623 patients enrolled in the study, 1259 finished the trial. Of these, 1197 met major accession, study drug compliance, and endoscopic evaluation criteria and composed the evaluable group of patients. Primary analyses were done on this evaluable group. Confirmatory analyses were also done on the intention-to-treat group, which comprised patients who had received at least one dose of study medication.

End Points

A patient who developed an endoscopically confirmed gastric or duodenal ulcer ( 0.3 cm in diameter and with perceptible depth) during the study was considered a prophylaxis failure.

Statistical Analysis

The baseline demographic characteristics of the study treatment groups were compared using either the Pearson chi-square test (sex and race) or the Kruskal-Wallis test (age). To evaluate the efficacy of misoprostol therapy in the prevention of NSAID-induced gastric and duodenal ulcers, the following three pairwise comparisons were made: misoprostol twice daily compared with placebo, misoprostol three times daily compared with placebo, and misoprostol four times daily compared with placebo. In addition, comparisons of the incidence of NSAID-induced ulcers in the groups receiving misoprostol twice and three times daily with that in the group receiving misoprostol four times daily were done to determine evidence of similar effectiveness in the prevention of ulcers. Treatment-by-investigator interactions were assessed qualitatively. The dosage-response effect of misoprostol was determined by using a logistic regression analysis with ulcer occurrence as the dependent variable and dosage as the independent variable. The placebo group was not included for these analyses.

Incidence rates for adverse events were calculated for body system, type of event, severity (mild, moderate, and severe), and relation to study drug (none, uncertain, and probable). Interregimen incidence rate comparisons were made using the unadjusted, two-tailed chi-square test. If chi-square assessment of incidence rates for the four regimens for an individual adverse event showed a statistically significant difference, incidence rates for the particular adverse event were further tested without the placebo group to distinguish differences among the three active treatment groups. Adverse events causing withdrawal from the study were compared across treatment groups using the Pearson chi-square test or the Fisher exact test. Dosage-response effect on adverse events was determined by using a logistic regression analysis with adverse event as the dependent variable and dosage as the independent variable. The statistical analysis was done by G.D. Searle & Co.

Results

Top Methods Results Discussion Author & Article Info References

Study Population

The disposition and randomization of patients entered into the study (n = 1623) is presented in Table 1. Five patients received no medication and were excluded from the intention-to-treat group. A total of 421 patients was excluded from the intention-to-treat group (359 were withdrawn before reaching an end point and 62 were excluded for other reasons), leaving 1197 patients in the evaluable group.

Demographic Characteristics

Demographic data for the intention-to-treat group are shown in Table 2. Neither the intention-to-treat group nor the evaluable groups differed significantly with respect to age, sex, race, NSAID use, type of arthritis, prevalence of smoking, or alcohol use.

Gastric Ulcers

Gastric ulcers (evaluable for the gastric ulcer group) were noted in 51 of 325 patients (15.7%) receiving placebo, in 29 of 358 patients (8.1%) receiving misoprostol twice daily, in 13 of 336 patients (3.9%) receiving misoprostol three times daily, and in 6 of 152 patients (4.0%) receiving misoprostol four times daily (Table 3). The incidence of gastric ulcers was significantly lower in groups receiving misoprostol twice daily (difference, 7.6% [95% CI, 2.7% to 12.5%]; P = 0.002), three times daily (difference, 11.8% [CI, 7.4% to 16.3%]; P < 0.001), and four times daily (difference, 11.7% [CI, 6.7% to 16.8%]; P < 0.001) than in the group receiving placebo. Pairwise comparison showed no statistical difference between the group receiving misoprostol four times daily and the groups receiving it three times or twice daily. However, a significant difference (difference, 4.2% [CI, 0.7% to 7.7%]; P = 0.02) was noted between the group receiving misoprostol twice daily and the group receiving it three times daily. A significant dose-response effect across treatment groups was noted (P = 0.02).

Duodenal Ulcers

Duodenal ulcers (evaluable for the duodenal ulcer group) were noted in 23 of 308 patients (7.5%) receiving placebo, in 9 of 344 patients (2.6%) receiving misoprostol twice daily, in 11 of 332 patients (3.3%) receiving misoprostol three times daily, and in 2 of 148 patients (1.4%) receiving misoprostol four times daily (Table 4). Pairwise comparison showed a significantly lower rate in the groups receiving misoprostol twice daily (difference, 4.9% [CI, 1.5% to 8.2%]; P = 0.004), three times daily (difference, 4.2% [CI, 0.6% to 7.7%]; P = 0.019), and four times daily (difference, 6.1% [CI, 2.6% to 9.6%]; P = 0.007) than in the placebo group. No significant difference was seen in rates between the group receiving misoprostol four times daily and the groups receiving it twice and three times daily. In addition, no significant difference was noted between the group receiving misoprostol twice daily and the group receiving it three times daily. No significant dose-response effect in the incidence of duodenal ulcers was noted.

Tolerability and Safety

Table 5 shows the most common side effects encountered during the study. No patient died while enrolled in the study. Patients receiving placebo had significantly fewer total adverse events than did patients receiving misoprostol four times daily. Gastrointestinal adverse events were the most commonly encountered events. The placebo group had significantly (P = 0.002) fewer gastrointestinal adverse events than did the group receiving misoprostol four times daily. However, the frequency of gastrointestinal adverse events in the groups receiving misoprostol twice and three times daily was not statistically different from the frequency in the placebo group. Of specific gastrointestinal symptoms, only flatulence (P = 0.02) was significantly related to misoprostol dosage.

The rate of early withdrawal from the study prompted by total adverse events was significantly lower in the placebo group (P < 0.001) and in the groups receiving misoprostol twice (P = 0.003) and three times daily (P = 0.002) than in the group receiving misoprostol four times daily. Similarly, early withdrawal because of any gastrointestinal adverse event was significantly lower in the placebo group (P < 0.001) and in the groups receiving misoprostol twice (P < 0.001) and three times daily (P = 0.006) than in the group receiving misoprostol four times daily. Significant dosage-related trends were noted in the incidence of early withdrawal due to any adverse event (P = 0.01) and to any gastrointestinal adverse event (P = 0.002). With regard to early withdrawal because of specific gastrointestinal symptoms, statistically significant dosage-related trends were noted in the incidences of abdominal pain (P = 0.005), nausea (P = 0.04), and flatulence (P = 0.03).

Discussion

Top Methods Results Discussion Author & Article Info References

When prescribing any drug with potentially substantial side effects, physicians must balance anticipated therapeutic benefit with the risk for unwanted, potentially therapy-limiting effects. The most serious unwanted side effects of NSAID therapy involve the upper gastrointestinal tract and include gastric and duodenal ulcerations and their complications. Although previous studies have shown that misoprostol, 200 µg four times daily, reduces the incidence of ulcer formation associated with NSAID therapy, the use of the drug at this currently recommended dosage is associated with side effects that often limit the drug's utility [18].

Our results show that in patients with upper gastrointestinal symptoms, all dosages of misoprostol significantly reduced the incidence of NSAID-induced gastric and duodenal ulcers compared with placebo. This confirms the findings of two other preliminary reports [15, 16]. Pairwise comparisons of ulcer rates in the active treatment groups constituted a secondary analysis. The results of these comparisons indicate that rates of gastric ulcer in the groups receiving misoprostol twice and three times daily were different. No difference, however, was noted between the group receiving misoprostol twice daily and the group receiving misoprostol four times daily. This may be because the number of patients in the group receiving misoprostol four times daily was only about 50% of that in the groups receiving misoprostol twice or three times daily, thus reducing the power of the test. In addition, no statistical difference was noted in rates of gastric ulcer between the groups receiving misoprostol three (3.9%) and four times daily (4%). Similar comparisons on duodenal ulcer rates demonstrated no statistical difference among the active treatment groups.

Dose-response analysis showed that the gastric ulcer-preventing effect of misoprostol was dose-related in the active treatment groups: This protective effect appears to plateau between 200 µg twice daily and 200 µg three times daily. Thus, 200 µg four times daily does not offer any additional protection. Studies [18, 19] suggest that depletion of endogenous prostaglandins plays a critical role in the development of NSAID-induced mucosal injury. Thus, our trend analysis helps to support the conclusion that the protective effect of misoprostol is related, in part, to the restoration of prostaglandin levels in the gastric mucosa. On the other hand, no misoprostol dose-response effect was noted regarding the prevention of duodenal ulcers. All doses of misoprostol tested were equally efficacious in the prevention of duodenal ulcers.

Analysis of adverse events and study withdrawals showed that misoprostol was better tolerated when received twice and three times daily than when received four times daily. Gastrointestinal symptoms are among the side effects most frequently reported with misoprostol. The use of misoprostol twice or three times daily reduced the rates of adverse events and withdrawals caused by adverse events for many of the individually recorded gastrointestinal effects. Furthermore, a significant dosage-related effect was seen: Higher dosages of misoprostol were more likely to cause higher rates of adverse events and withdrawals.

The end point investigated in our study was the development of endoscopically detectable ulcer. However, endoscopic mucosal changes do not always correlate with symptoms, nor do they always lead to the development of clinically detectable ulcer disease. Thus, the use of endoscopic end points may not be a sufficient marker for clinically significant ulcer complications. Numerous studies have confirmed that the use of NSAIDs is associated with a high degree of endoscopically detectable mucosal changes, including erosions and ulcers, and that the use of misoprostol significantly reduces the development of NSAID-induced ulcers. Furthermore, the results of a large, multicenter, placebo-controlled study [17] show that misoprostol significantly reduces the incidence of clinically detectable gastrointestinal events, thus confirming the clinical utility of misoprostol.

Our study was designed to monitor all adverse events encountered during its course. Clinically, most of the adverse events were mild and did not require medical intervention. The clinical significance of these events was determined by the physician or the patient, or both. Serious adverse events were reported in 17 patients receiving placebo (3.7%), 15 patients receiving misoprostol twice daily (3.2%), 9 patients receiving misoprostol three times daily (1.9%), and 4 patients receiving misoprostol four times daily (1.8%). These events comprised a broad range of disorders, and no specific clustering was seen. In the intention-to-treat analysis, overall gastrointestinal assessment done by physicians at weeks 4, 8, and 12 did not differ statistically among the four groups. Symptoms leading to study withdrawal were also recorded. Most of these adverse events only required withdrawal of study medication.

In large invasive clinical studies, early withdrawal may introduce bias into the study results. However, the number of patients withdrawn from our study did not differ statistically across the four treatment groups. Adverse events were the most common cause of study withdrawal. The study protocol also required endoscopy at study termination if the patient was withdrawn for any reason. The percentage of patients not having an end-of-study endoscopy (for any reason) was similar in all four groups. To further address the issue of withdrawal bias, statistical analyses were done on the intention-to-treat group, which confirmed the findings in the evaluable groups. In addition, for the statistical analyses, multiple pairwise comparisons were made and thus should be interpreted with some caution.

We evaluated only patients with arthritis. No attempt was made to confirm the rheumatologic diagnoses using classification criteria. The study was designed to address the issue of ulcer prevention in patients with arthritis receiving long-term NSAID therapy. Patients with all forms of arthritis probably constitute a substantial proportion of long-term NSAID users [20]. Some studies [21] suggest that certain forms of arthritis may predispose to a greater incidence of NSAID-induced ulcers. Thus, the inclusion of different forms of arthritis in one study may mask differences in susceptibility to NSAID-induced ulcers among the groups studied. However, for both gastric and duodenal ulcers, the presence of rheumatoid arthritis (an inflammatory arthritis) or osteoarthritis (noninflammatory arthritis) was not a significant prognostic variable in our study (data not shown).

In addition, only patients with upper gastrointestinal symptoms were included in the study. This bias was thought to be necessary to recruit patients and physicians into an invasive (three endoscopies) study such as this. Patients receiving NSAID therapy frequently have silent peptic ulcer disease. There is little indication that symptomatic and asymptomatic patients differ with regard to the frequency or severity of NSAID-induced gastroduodenopathy [22]. However, some data [20, 23] suggest that patients with previous NSAID-induced gastrointestinal side effects or those receiving antacids or histamine-2-receptor antagonists are at greater risk for serious gastrointestinal events. Although this trial studied only patients with existing upper gastrointestinal symptoms, patients without such symptoms have also shown endoscopic reduction in gastroduodenal ulcers with misoprostol [14].

An important issue in the use of misoprostol relates to the appropriate selection of patients for treatment. Because of its abortifacient properties, misoprostol should not be used in pregnant women. Women of childbearing potential should receive appropriate contraception. Furthermore, misoprostol may not be indicated for all patients receiving NSAID therapy, especially in light of the drug's potential cost and side effects. Several studies [22-24] have identified certain patient groups as having a high risk for the development of NSAID-induced ulcer disease. These groups include elderly persons, persons with comorbid diseases, persons with a history of peptic ulcer disease or gastrointestinal bleeding, and persons with a history of NSAID intolerance. These high-risk patients are most likely to benefit from the gastrointestinal protective effects of misoprostol.

In clinical practice, several medications are widely used to prevent NSAID-induced ulcers; these include histamine-2-receptor antagonists (such as cimetidine and ranitidine), omeprazole, and sucralfate. These agents are generally helpful in alleviating dyspeptic symptoms and in treating active ulcer disease. They are also generally well tolerated, but they are not approved by the FDA for the prevention of NSAID-induced ulcers. Ranitidine has been shown to prevent the formation of duodenal ulcers in patients receiving long-term NSAID therapy [25, 26]. However, acid suppression has not been shown to be effective in comparative trials for the prevention of gastric ulcers, the most common type of NSAID-induced ulcers [25-27]. Similarly, omeprazole does not appear to protect the stomach against NSAID-induced damage. However, the role of proton pump inhibitors in the prevention of NSAID-induced gastroduodenal ulcerations has not been established [28, 29]. Sucralfate appears to exert its antiulcer effect locally, through the formation of surface complexes. In a direct comparative trial with misoprostol [13], this mechanism was not shown to prevent gastric ulcers in patients receiving long-term NSAID therapy. Misoprostol is an important therapeutic agent because of its unique ability to protect against both gastric and duodenal ulcers associated with NSAID use. However, its use at the currently recommended dosage of 200 µg four times daily has been limited by gastrointestinal side effects. In patients receiving long-term NSAID therapy who are being considered for misoprostol therapy, dosages of 200 µg twice or three times daily are effective and better tolerated alternatives to the 200 µg four times daily regimen. Protection against NSAID-induced gastric ulcers increases with the dose of misoprostol, but maximum protection appears to be achieved with doses of 400 to 600 µg daily. Maximum protection against NSAID-induced duodenal ulcers can be achieved with doses as low as 400 µg daily. Physicians prescribing misoprostol should choose a dosage that best balances the drug's mucosal protective effects with its side effects.

Appendix

The following investigators each enrolled 20 or more patients into the study: J.B. Raskin, MD, University of Miami School of Medicine, Jackson Memorial Hospital, Miami, Florida; R.A. Erickson, MD, Veterans Affairs Medical Center, Long Beach, California; G. Triadafilopoulos, MD, Veterans Affairs Medical Center, Oakland, California; A.L. Weaver, MD, Arthritis Center of Nebraska, Lincoln, Nebraska; N. Patel, MD, Future Healthcare Research Center, Dayton, Ohio; R.E. Pruitt, MD, Medical Research Institute, Nashville, Tennessee; J.E. Jackson, MD, Walker Clinical Evaluations, Inc., Indianapolis, Indiana; E.A. Tindall, MD, Portland Adventist Medical Center, Portland, Oregon; R.M. Fleischmann, MD, Metroplex Clinical Research Center, Dallas, Texas; R.D. Aaronson, MD, Suburban Heights Medical Center, Chicago Heights, Illinois; P.D. Toth, MD, Midwest Institute for Clinical Research, Indianapolis, Indiana; M.A. Korsten, MD, Veterans Affairs Medical Center, Bronx, New York; R. Jaszewski, MD, Veterans Affairs Medical Center, Allen Park, Michigan; C.A. Birbara, MD, Worcester City Hospital, Worcester, Massachusetts; D.H. Sikes, MD, East Pasco Specialty Care, Zephyrhills, Florida; R.K. Marwah, MD, Texas Technical University, El Paso, Texas; J.L. Kaine, MD, Sarasota Arthritis Center, Sarasota, Florida; R.B. Lies, MD, Wichita Clinic, Wichita, Kansas; D. Lehman, MD, Sacramento Sierra Medical Group, Sacramento, California; K. Unnoppet, MD, Future Healthcare Research Center, Birmingham, Alabama; R.A. Sanowski, MD, Carl T. Hayden Veterans Affairs Medical Center, Phoenix, Arizona; R.H. White, MD, University of California Davis Medical Center, Sacramento, California; D.S. Stanton, MD, Western Medical Center, Santa Ana, California; M.L. Brando, MD, Allergy Medical Group of San Diego, Inc., San Diego, California; J.D. Smucker, MD, Future Healthcare Research Center, Columbus, Ohio; J.L. Skosey, MD, PhD, University of Illinois at Chicago College of Medicine, Chicago, Illinois; F.R. Schmid, MD, Northwestern Memorial Hospital, Chicago, Illinois; T.R. Liebermann, MD, Austin Diagnostic Clinic, Austin, Texas; and J.R. Caldwell, MD, Halifax Clinical Research Center, Daytona Beach, Florida.

Dr. White: 2221 Stockton Boulevard, Sacramento, CA 95817.

Dr. Jackson: 11180 Templin Road, Zionsville, IN 46077.

Dr. Weaver: 2121 South 56th Street, Lincoln, NE 68506.

Dr. Tindall: 10000 SE Main Street #207, Portland, OR 97216.

Dr. Lies: 3311 East Murdock Street, Wichita, KS 67209.

Dr. Stanton: 1140 West Leveta Avenue, Suite 550, Orange, CA 92668.