Nonclinical model for assessing gastric effects of bisphosphonates

Gastrointestinal intolerance has been associated with amino bisphosphonate therapy in the clinic. The objective of this study was to develop a model for assessing bisphosphonate-induced gastric damage that may aid in the development of future bisphosphonate therapies. Rats were dosed concomitantly with indomethacin (40 mg/kg, subcutaneously) and an amino or pyridinyl bisphosphonate (orally at 150, 225 or 300 mg/kg). The bisphosphonates studied were pamidronate and alendronate (primary amino bisphosphonates) and risedronate and NE-97221 (pyridinyl bisphosphonates). Macroscopically, alendronate induced significantly (P < 0.05) more antral damage (both lesion length and number) than pamidronate and risedronate at 225 and 300 mg/kg, and more than NE-97221 at 300 mg/kg. NE-97221 induced significantly more antral damage (lesion length) than risedronate at 225 mg/kg and a greater number of lesions compared to pamidronate and risedronate at 225 and 300 mg/kg. The model was validated histologically, and macroscopic findings correlated with histologic evidence of antral mucosal necrosis and inflammatory infiltration of the lamina propria. The calcium chelators EGTA and EDTA did not induce gastric damage in this model when dosed according to the same protocol as the nitrogen-containing bisphosphonates. This suggests that calcium chelation does not account for the gastric effects in this model. The fasted, indomethacin-treated rat provides a novel nonclinical model to assess gastric effects of bisphosphonates, which may aid in the development of future bisphosphonate therapies. These data suggest that when expressed on an actual or anticipated clinical dose basis for osteoporosis (pamidronate, 150 mg; alendronate, 5-10 mg; risedronate and NE-97221, 5 mg), primary amino bisphosphonates may have a greater potential for inducing gastric damage than do pyridinyl bisphosphonates.