Risk of fracture among women who lose bone density during treatment with alendronate. The Fracture Intervention Trial

It is commonly believed that the response to treatment in patients on alendronate is proportional to the increase in bone mineral density (BMD), and that those who lose BMD during treatment might not respond to treatment. In the Fracture Intervention Trial 6,459 women were randomly assigned to treatment with alendronate or placebo; BMD was measured annually, and new spine fractures were assessed by lateral spine films, taken at baseline and end of follow-up. Among subjects who took at least 70% of the study drug (5,220 women), we compared reductions in risk of spine fractures at end of follow-up (3 or 4 years) within various levels of change in total hip and spine BMD after 1 and 2 years of treatment, after adjustment for differences in characteristics between the treatment and control groups. Women "losing" BMD at the lumbar spine (0% to 4%) while on alendronate had a reduction of 60% in vertebral fracture risk [OR = 0.40 (0.16, 0.99)] compared to their counterparts in the placebo group. The few women that lost more than 4% did not have a significant benefit [OR = 0.15 (0.02, 1.29)]. Those who "gained" BMD (0% to 4%) during treatment had a reduction in risk of 51% [OR = 0.49 (0.30, 0.78)]. Similarly, women who "lost" total hip BMD (0% to 4%) during the first year on alendronate had a 53% decreased risk of vertebral fracture compared to their controls taking placebo [OR = 0.47 (0.27, 0.81)], whereas those "gaining" BMD (0% to 4%) had a comparable risk reduction [OR = 0.49 (0.34, 0.71)]. This was not observed for the few women who lost more than 4% [OR = 0.61 (0.11, 3.45)]. Patients who lost BMD at both the hip and spine were not protected by alendronate. Among patients who adhere to treatment with alendronate, even those who lose BMD benefit from a substantial reduction in risk of vertebral fracture. So, the reduction in bone turnover induced by alendronate might be more important than BMD changes. The few women who lose the most BMD (more than 4% per year) might not benefit from the treatment.