IBUPROFEN SHOWS COMPARABLY LOWER G-I BLEEDING RISK THAN OTHER NSAIDS

IBUPROFEN SHOWS COMPARABLY LOWER G-I BLEEDING RISK THAN OTHER NSAIDS in two pharmacoepidemiology studies presented at the Ninth International Conference on Pharmacoepidemiology in Washington, D.C. on Sept. 1. The studies, involving a total of 1,490 cases of upper gastrointestinal bleeding (UGIB) and 3,307 controls, found that ibuprofen usage was associated with a lesser risk of upper gastrointestinal bleeding than aspirin and other nonsteroidal anti-inflammatories. In a multicenter, case-control study evaluating 1,004 cases of UGIB in Hungary, Sweden, and the U.S., Judith Kelly, Boston University, et al. found that the relative risk of recent ibuprofen users was 1.6 (at a 95% confidence interval, 0.9-3.1), compared with 4.9 (4.0-6.1) for aspirin users, 6.5 (3.8-11) for naproxen users, and 6.6 (2.3-19) for piroxicam users. Naproxen and piroxicam currently are prescription-only in the U.S. However, Kelly et al. concluded that those findings were dependent on usage frequency and dose. Regular dosage, defined as at least "every other day for more than one week" prior to the index day, was universally associated with a greater risk of UGIB episodes than occasional usage, defined as at least once during the week prior to the index day, which was determined by the incidence of UGIB. Using a multiple logistic regression, the Kelly group found that neither "regular" nor "occasional" use of ibuprofen at lower doses was associated with increased risk of stomach bleeds. "Regular" usage of aspirin was associated with a relative risk of 5, while "occasional" use was associated with a relative risk of 2.4; "regular" use of naproxen put a patient at greater risk (RR = 6.8) than "occasional" usage (RR = 4.9). In the piroxicam cases, the Kelly study indicated that regular usage (RR = 20) was considerably more dangerous than "occasional" (RR = 6.6). Kelly noted that the study results suggest that the UGIB risk posed by ibuprofen and aspirin varies according to dose, while data on naproxen and piroxicam proved insufficient to support any dose differential. In patients taking "regular" doses of ibuprofen of 1,200 mg or greater, the study found a three-fold increase in risk for UGIB episodes compared to the lower dose, Kelly reported. Ibuprofen taken "occasionally" at doses of 600 mg or greater also showed a statistically significant increased risk. For aspirin doses of less than 250 mg per day, the study found a relative risk of 3.4; for doses higher than 1,000 mg a day, the relative risk "jumps" to 7.4, and for doses higher than 2,000 mg, the relative risk reaches 24, Kelly reported. The "occasionally" use of aspirin at doses less than 325 mg slightly increased the risk of gastrointestinal bleeds (RR = 1.9), but Kelly concluded that those findings were not significant. For "occasionally" use at larger doses (1,000 mg or greater), the increased risk became statistically significant (RR = 3). Patients occasionally taking 2,000 mg doses of aspirin a day increased their relative risk of bleeding to 7.3, which Kelly noted was comparable to the risk of "regular" aspirin use of 1,000 mg a day. Kelly et al. also found that patients using NSAIDs for shorter durations -- less than one month -- were at a greater risk for a UGIB episode, regardless of the particular drug taken. For the purposes of determining patient risk, Kelly directed the audience to also look at figures for excess, or attributable, risk. In her study, Kelly found that excess risk translated into 49 cases per 100,000 users per year. NSAIDs that were "significantly associated" with excess risk included aspirin, naproxen, and piroxicam. Ibuprofen displayed an excess risk only at doses higher than 1,200 mg, when its risk was "comparable to low dose aspirin," Kelly noted. Data for the studies was culled from information gathered from regular hospital visits in eastern Massachusetts, Stockholm, and Budapest. All cases were patients older than 18 years, 97% of whom had their UGIB diagnoses confirmed through endoscopy, surgery, or radiology. Cases were limited to first-time sufferers whose symptoms lasted less 30 days. The 2,446 controls were matched to UGIB patients by sex and age (+/- 5 years); American and Swedish controls were found through a town census and a computerized population registry, while the Hungarian controls were found among other patients in the Budapest hospital with diagnoses unrelated to NSAIDs. Drug usage information was compiled by interviewers trained to conduct the detailed questionnaire in the respective languages. However, Kelly noted that naproxen and ibuprofen were not used in Hungary. Exclusions were made on the basis of prior UGIB episodes, ulcers, gastritis, gastric cancer or other G-I conditions; histamine antagonist therapy or current anticoagulant therapy; and any bleeding disorders. Other information considered was age, sex, region, marital status, education, caffeine and alcohol consumption, and smoking. The second study, a case-control investigation of UGIB episodes in elderly Canadian patients (68 years or older) conducted by Lucien Abenhaim, MD, McGill University, and Yola Moride, calculated that of the patients exposed to NSAIDs within a month of their UGIB episode, ibuprofen users had the lowest risk of incidence (RR = .8), compared with naproxen users (RR = 2.8) and piroxicam users (RR = 5.2). Overall, the Abenheim study found that NSAID use increased the risk of G-I bleeding by magnitude "of the order of 3," Moride told the conference. Individual risk for G-I bleeding in the elderly, however, hinged on a number of baseline factors other than the specific NSAID used, including duration and pattern of use and past experience with NSAIDs. Abenheim et al. conducted their study in seven Montreal hospitals, using 486 elderly patients who were being treated for acute gastrointestinal episodes -- present for not more than one month -- that were characterized by severe UGIB or peptic, gastric or duodenal ulcers marked by hemorrhage and/or perforation. The 861 controls were patients hospitalized for acute episodes not related to NSAID use or UGIB; they were matched to the cases by age (+/- 5 years), sex, and hospital admission. Patient histories were determined through a database analysis of medical records for three years prior to the episode; drug information was determined through the prescription history of the patients. Patients with chronic conditions, gastric histories, cancer, and drug allergies were ineligible for the study. All the patient cases were validated. Past NSAID use by the elderly was associated with a lower risk of UGIB from usage (RR = .66), while first time elderly users appeared to be at a greater risk for bleeding than those with prior experience with NSAIDs, Moride reported. Long-term usage was associated with a decreased risk for UGIB, but Moride pointed out that these results might be due to a self-selecting sample. In addition, the study found that patients who switched NSAIDs had a greater relative risk for UGIB episodes than patients who did not, Moride said. Recent exposure was defined as exposure to an NSAID within 30 days prior to hospital admission, as found by a database search of a patient's prescription history. Past history encompassed recorded NSAID use between 31 days and three years prior to hospitalization. Overlapping prescriptions were considered to be in two groups. The study defined "long-term use" as more than 30 days, while "short term" use was defined as less than 30 days. The study accounted for such concomitant variables as the recent use of gastro-protective agents, steroids, anti-coagulants, anti- infectives, and aspirin, as well as tobacco use. Another study looking at the comparable safety of NSAIDs and aspirin was also presented at the Sept. 1 meeting. "A Reassessment of Aspirin Therapy in Rheumatoid Arthritis," conducted by Gurkirpal Singh, MD, Stanford University, et al., suggested that aspirin was less toxic than non-salicylate NSAIDs. In 1,521 consecutive courses of aspirin therapy and 4,860 courses of NSAID therapy, the study arrived at a toxicity index for aspirin of 1.77, while the toxicity for NSAIDs ranged from 2.04 to 5.94, Singh reported. During the Q&A period, FDA Office of Drug Evaluation I Director Robert Temple, MD, suggested to Singh that his findings may have been influenced by the fact that aspirin is an older drug with known G-I toxicities, and that patients would stop using it at the first sign of trouble. Consequently, the aspirin group in the Singh study may have included a group of people who already knew that they would not experience toxicity with the drug, Temple observed.