Trabecular Bone Score and Hip Structural Analysis in Patients With Atypical Femur Fractures.

ABSTRACT

Bisphosphonate use has declined dramatically in recent years, partly because of fear of rare side effects like atypical femur fractures (AFFs). It is therefore desirable to have a diagnostic method to identify those at risk of AFF to prevent this serious complication. We compared trabecular microarchitecture and hip geometry between 30 patients with AFF and 141 controls of similar age and sex, using bisphosphonates. Trabecular bone score (TBS) and hip structural analysis (HSA) were used to assess trabecular microarchitecture and macroscopic hip geometry from dual-energy X-ray absorptiometry images of the lumbar spine and hip, respectively. General characteristics, TBS, and HSA were compared between patients with AFF and controls using Student's t tests and chi-square statistics. Associations between AFF and TBS and femur geometric characteristics by HSA were adjusted for sex, age, height, weight, ethnicity, duration of bisphosphonate use, and glucocorticoid use. Additionally, the analysis of TBS was adjusted for lumbar spine bone mineral density and the time difference between dual-energy X-ray absorptiometry scanning and the diagnosis of AFF. Patients with AFF had significantly higher body mass index than controls, had used bisphosphonates longer, and glucocorticoids and proton pump inhibitors more frequently. Sex-specific T-score was significantly higher in patients with AFF at the lumbar spine (p = 0.004), but not at the femoral neck (p = 0.190) after adjustment for age, height, and weight. TBS did not differ significantly between patients with AFF and controls. Neither neck shaft angle nor any geometric variables at the femoral shaft measured by HSA differed between patients with AFF and controls. At the narrow neck, patients with AFF had lower buckling ratio and higher centroid position, consistent with a lower risk of classical fragility hip fractures. The findings at narrow neck and higher bone mineral density might be explained by the fact that the majority of patients with AFF used bisphosphonates to prevent glucocorticoid-induced osteoporosis. Based on our results, TBS and HSA do not appear to have value in detecting patients at risk of AFF.