Relationship of changes in total hip bone mineral density to vertebral and nonvertebral fracture risk in women with postmenopausal osteoporosis treated with once-yearly zoledronic acid 5 mg: the HORIZON-Pivotal Fracture Trial (PFT).

Measurements of change in bone mineral density (BMD) are thought to be weak predictors of treatment effect on the reduction of fracture risk. In this study we report an alternative year-on-year approach for the estimation of treatment effect explained by BMD in which we examine the relationship between fracture risk and the most recent change in BMD. We studied 7736 postmenopausal women (ages 65 to 89 years) who were participants in the Health Outcomes and Reduced Incidence with Zoledronic Acid Once Yearly-Pivotal Fracture Trial (HORIZON-PFT) and were randomized to either intravenous administration of zoledronic acid or placebo. The percentage of treatment effect explained by change in total hip BMD was estimated using the alternative year-on-year approach and the standard approach of looking at change over 3 years. We also studied a subset of 1132 women in whom procollagen type 1 amino-terminal propeptide (PINP) was measured at baseline and 12 months, to estimate the percentage of treatment effect explained by change in PINP. Regardless of the method used, the change in total hip BMD explained a large percentage of the effect of zoledronic acid in reducing new vertebral fracture risk (40%; 95% CI, 30% to 54%; for the 3-year analysis). The treatment effects for nonvertebral fracture were not statistically significant for the year-on-year analysis but 3-year change in BMD explained 61% (95% CI, 24% to 156%) of treatment effect. Change in PINP explained 58% (95% CI, 15% to 222%) of the effect of zoledronic acid in reducing new vertebral fracture risk. We conclude that our estimates of the percentage of treatment effect explained may be higher than in previous studies because of high compliance with zoledronic acid (due to its once-yearly intravenous administration). Previous studies may have underestimated the relationship between BMD change and the effect of treatment on fracture risk.