Abstract
The current clinical assessment of fracture risk lacks information about the inherent quality of a person's bone tissue. Working toward an imaging-based approach to quantify both a bone tissue quality marker (tissue hydration as water bound to the matrix) and a bone microstructure marker (porosity as water in pores), we hypothesized that the concentrations of bound water (C(bw)) are lower and concentrations of pore water (C(pw)) are higher in patients with osteoporosis (OP) than in age- and sex-matched adults without the disease. Using recent developments in ultrashort echo time (UTE) magnetic resonance imaging (MRI), maps of C(bw) and C(pw) were acquired from the uninjured distal third radius (Study 1) of 20 patients who experienced a fragility fracture of the distal radius (Fx) and 20 healthy controls (Non-Fx) and from the tibia mid-diaphysis (Study 2) of 30 women with clinical OP (low T-scores) and 15 women without OP (normal T-scores). In Study 1, C(bw) was significantly lower (p = 0.0018) and C(pw) was higher (p = 0.0022) in the Fx than in the Non-Fx group. In forward stepwise, logistic regression models using Bayesian Information Criterion for selecting the best set of predictors (from imaging parameters, age, BMI, and DXA scanner type), the area-under-the-receiver operator characteristics-curve (AUC with 95 % confidence intervals) was 0.73 (0.56, 0.86) for hip aBMD (best predictors without MRI) and 0.86 (0.70, 0.95) for the combination of C(bw) and C(pw) (best predictors overall). In Study 2, C(bw) was significantly lower (p = 0.0005) in women with OP (23.8 ± 4.3 (1)H mol/L) than in women without OP (29.9 ± 6.4 (1)H mol/L); C(pw) was significantly higher by estimate of 2.9 (1)H mol/L (p = 0.0298) with clinical OP, but only when accounting for the type of UTE-MRI scan with 3D providing higher values than 2D (p < 0.0001). Lastly, C(bw), but not C(pw), was sensitive to bone forming osteoporosis medications over 12-months. UTE-MRI-derived measurements of bound and pore water concentrations are potential, aBMD-independent predictors of fracture risk.