Automated bone age calculation is an independent predictor of vertebral fracture burden in Duchenne muscular dystrophy.

Abstract

Bone age (BA) assessments by BoneXpert were similar to BA evaluations by experienced radiologists in males with Duchenne muscular dystrophy (DMD). Furthermore, BA delay as determined by both radiologists and BoneXpert, and the BoneXpert-derived Bone Health Index Z-score (a surrogate for bone strength), were associated with spinal fragility. To compare BA assessed by experienced radiologists versus BoneXpert in males with DMD and to evaluate the extent to which BA delay, and the BoneXpert-derived Bone Health Index (BHI, a proposed surrogate for bone strength), are associated with the Spinal Deformity Index (a measure of vertebral fracture [VF] burden). Results were compared for BA evaluated on 54 left-hand radiographs from patients with DMD (chronological age [CA] 11.6 ± 3.8 years; mean ± SD) by experienced radiologists versus BoneXpert. The relationships between BA delay (CA-BA), BHI, and the SDI were assessed. The mean BA delays reported by radiologists (-1.9 ± 2.4 years) and BoneXpert (-1.8 ± 2.4 years) were similar (p = 0.138) with an intraclass correlation coefficient of 0.978 (95% confidence interval [CI], 0.961-0.987). Multivariable Poisson regression showed that every 1-year BA delay by BoneXpert was associated with a 1.5-fold (95% CI, 1.1-2.1; p = 0.009) increase in incidence rate ratio (IRR) for SDI, similar to radiologists (1.8; 95% CI, 1.3-2.5; p = 0.001). In a separate model, every 1.0 unit decrease in the BHI Z-score by BoneXpert was associated with a 2.9-fold (95% CI, 2.0-4.2; p < 0.001) increase in IRR for SDI. BoneXpert provided BA results that were similar to those generated by radiologists in individuals with DMD and that retained the clinically relevant association between BA delay and VF. We also demonstrated that lower BHI Z-scores in individuals with DMD were associated with spinal fragility.