Less strict intervention thresholds for the FRAX and TBS-adjusted FRAX predict clinical fractures in osteopenic postmenopausal women with no prior fractures

Martin Kužma,Didier Hans,Tomáš Koller,Eva Némethová,Peter Jackuliak,Zdenko Killinger,Heinrich Resch,Juraj Payer

DOI: https://doi.org/10.1007/s00774-017-0864-1

Abstract:Little is known about the clinical relevance of treating post-menopausal women with no prior history of fragility fracture and bone mineral densities (BMD) within the osteopenic range. In recent years, in addition to BMD and FRAX fracture probability assessments, a surrogate measure of bone micro-architecture quality, called the trabecular bone score (TBS), has been proven to predict future fragility fractures independently of both BMD and the FRAX. In this retrospective analysis of a follow-up study, we compared three risk assessment instruments-the FRAX, the TBS, and a TBS-adjusted FRAX score-in their ability, to predict future fragility fractures over a minimum of five years of follow-up among post-menopausal osteopenic women with no prior fragility fractures. We also sought to determine if more- versus less-stringent criteria were better when stratifying patients into higher-risk patients warranting osteoporosis-targeted intervention versus lower-risk patients in whom intervention would usually be deemed unnecessary. Over a mean 5.2 years follow-up, 18 clinical fragility fractures were documented among 127 women in the age 50 years and older (mean age = 66.1). On multivariate analysis utilizing regression models and Kaplan-Meier curve analysis, less-stringent criteria for the FRAX and TBS-adjusted FRAX were capable of predicting future fractures (with sensitivity/specificity of 83/31; 39/77 and 78/50% for TBS, FRAX and TBS-adjusted FRAX, respectively), while more-stringent criteria were incapable of doing so (with sensitivity/specificity of 56/60; 39/77 and 39/74 for TBS, FRAX and TBS-adjusted FRAX, respectively). Neither TBS threshold alone was a significant predictor of future fracture in our study. However, hazard ratio analysis revealed slight superiority of the TBS-adjusted FRAX over the FRAX alone (HR = 3.09 vs. 2.79). Adjusting the FRAX tool by incorporating the TBS may be useful to optimize the detection of post-menopausal osteopenic women with no prior fractures who warrant osteoporosis-targeted therapy.