Accuracy of trabecular structure by HR-pQCT compared to gold standard μCT in the radius and tibia of patients with osteoporosis and long-term bisphosphonate therapy

Abstract

Despite an increasing use of high-resolution peripheral quantitative computed tomography (HR-pQCT) to evaluate bone morphology in vivo, there are reservations about its applicability in patients with osteoporosis and antiresorptive therapy. This study shows that HR-pQCT provides acceptable in vivo accuracy for bone volume fraction (BV/TV) in patients with osteoporosis and bisphosphonate (BP) treatment. The primary aim was to analyze agreement of trabecular structure between HR-pQCT and gold standard microtomography (μCT) in patients with osteoporosis and long-term BP therapy. In the BioAsset study, we analyzed cadaver radii and tibiae of 34 postmenopausal females (81.1 ± 7.1 years) with osteoporosis (no BP n = 22, 1-5 years BP n = 5, >5 years BP n = 7). Two HR-pQCT protocols (patient-mode and μCT-mode) were compared with gold standard μCT after image registration. Undecalcified histological sections were obtained to quantify nonmineralized bone matrix. Bland-Altman plots illustrated methodological agreement. Multiple regression analysis was used to test for variables associated with method agreement. In the radius and tibia, patient-mode HR-pQCT derived indices including bone volume fraction, trabecular number, and trabecular separation correlated well with gold standard μCT (R(2) = 0.78 - 0.88) except for trabecular thickness (R(2) = 0.11). Bland-Altman plots illustrated adequate agreement for bone volume fraction. Lower agreement of trabecular number and trabecular separation improved with decreasing structural impairment at the tibia only. Trabecular thickness was not appropriately assessed with HR-pQCT at both skeletal sites. Higher agreement for bone volume fraction was associated with increasing tissue mineral density in the tibia. HR-pQCT provides acceptable in vivo accuracy for BV/TV in patients with osteoporosis and BP treatment. Higher TMD was associated with higher BV/TV accuracy in vivo. Overall, methodological agreement got less accurate with increasing structural impairment in the tibia.