30th ASBMR 2008
[M358] Comparative Gastrointestinal Safety of Weekly Oral Bisphosphonates
Suzanne Cadarette, Jeffrey Katz, M Brookhart, Til Stürmer, Margaret Stedman, Daniel Solomon.. Brigham and Women's Hospital, Harvard Medical School.
Weekly bisphosphonates are the primary agents prescribed for osteoporosis. Prior evidence from endoscopic trials and observational studies suggests that daily risedronate has better gastrointestinal safety than daily alendronate. The purpose of our study was to compare the relative gastrointestinal safety of treatment with weekly bisphosphonates among older adults. We studied a population-based cohort of new users of weekly alendronate (70 mg) and weekly risedronate (35 mg) enrolled in the Pennsylvania Pharmaceutical Assistance Contract for the Elderly. Pharmacy and Medicare claims served as the primary data sources. Covariates included risk factors plausibly related to gastrointestinal disease and were assessed using information from the 12 months prior to bisphosphonate initiation. The primary outcome was hospitalization for upper gastrointestinal bleed. Secondary outcomes included outpatient diagnoses for upper gastrointestinal disease, symptoms, endoscopic procedures, and use of gastroprotective agents. We used Cox proportional hazard models to compare outcomes between agents within 120 days of treatment initiation, adjusting for covariates using risedronate propensity score quintiles. In sensitivity analysis, we examined composite outcomes and stratified results by gastrointestinal event history (yes or no) and age group (65-79 years or 80+ years). Alendronate was the reference group in all Cox proportional hazard models.
We identified 10,420 new users of either agent who initiated treatment between June 2002 and August 2005, mean age=79 years (SD=6.9) and 95% female. We observed 31 hospitalizations for upper gastrointestinal bleed (0.91/100 person-years) within 120 days of treatment initiation. Adjusting for covariates, there was no large difference in hospitalization rates for upper gastrointestinal bleed among those treated with risedronate (HR=1.12, 95%CI=0.55-2.28) compared to alendronate. Similarly, no differences were observed for secondary or composite outcomes, or in stratified analyses. However, among those with a history of gastrointestinal disease, rates for upper gastrointestinal endoscopies were lower among risedronate (HR=0.70, 95%CI=0.49-0.99) compared with alendronate recipients. Nonetheless, when endoscopies were combined with upper gastrointestinal diseases and symptoms as the outcome, this finding was attenuated towards the null (HR=0.93; 95%CI=0.83-1.04).
In conclusion, we found no difference in hospitalizations for upper gastrointestinal bleed between weekly alendronate and weekly risedronate. Small differences between agents were observed for upper gastrointestinal endoscopy, but this is of unclear clinical importance.
S.M. Cadarette, None.
Date: Monday, September 15, 2008
Session Info: Poster: Osteoporosis Treatment (Clinical): Bisphosphonates (11:30 AM-2:30 PM)
Presentation Time: 11:30 AM