Background Upper gastrointestinal (UGI) tract mucosal irritation has been reported for a variety of bisphosphonates, including pamidronate, etidronate, clodronate, alendronate (ALN), tiludronate, and risedronate (RIS).

Objectives To use evidence-based medicine to rate and summarise the evidence regarding bisphosphonates and UGI problems.

Methods Review and interpretation of published studies.

Results Although the clinical relevance of short-term, transient endoscopic findings is uncertain, one endoscopic study reported a higher incidence of gastric lesions for ALN vs RIS, and others reported no difference for ALN vs RIS or placebo.^1–4 These endoscopic studies differed in regard to blinding, duration, and control. Endoscopy trials that were appropriately blinded and of longer duration (4–10 wks) found no difference between ALN and RIS or placebo, even at the higher doses used for once-weekly dosing⁴ or to treat Paget’s disease.³Randomised, blinded, controlled trials (RCTs) with clinically relevant GI endpoints represent the highest level of evidence, and tens of thousands of patients have participated in RCTs of ALN and RIS for up to 7 years. UGI complaints were common in these studies (~30% in the first year), and were comparable to placebo. In particular, the incidence of more worrisome UGI events (including esophageal events, and gastroduodenal perforations, ulcers, and bleeds) was low and comparable to the placebo group, even in high risk subgroups, among 6459 women randomised to ALN or placebo for up to 4.5 years in the Fracture Intervention Trial.⁵ Furthermore, two independent RCTs of patients who had previously reported intolerance to bisphosphonate reported that the large majority (~85%) were able to continue ALN or RIS when treatment was resumed, and the permanent discontinuation rates were no different from the placebo group.^6,7

Conclusion Based on extensive data from large RCTs, we conclude that any increase above placebo in clinically relevant UGI problems related to bisphosphonate use is very low – too low to detect in trials involving thousands of participants during several years of follow-up.

References

1. Lowe C, et al. Am J Gastroenterol. 2000;95:634–40

2. Lanza F, et al. Gastroenterology 2000;119:631–8

3. Lanza F, et al. Am J Gastroenterol. 2000;95:3112–17

4. Lanza F, et al. Arthritis Rheum. 2000;43(Suppl 9):S198

5. Bauer DC, et al. Arch Intern Med. 2000;160:517–25

6. Miller PD, et al. Clin Ther. 2000;22(12):1433–42

7. Adachi JD, et al. Arthritis Rheum 2000;43(Suppl 9):S202