Abstract
This study evaluated trabecular bone score (TBS) for assessing vertebral fractures and spinal deformity in children and adolescents with osteogenesis imperfecta (OI). TBS showed superior performance in identifying vertebral fractures compared to areal bone mineral density (aBMD), especially in patients without densitometric osteoporosis, suggesting its potential for monitoring vertebral fractures and spinal deformity risk. BACKGROUND: TBS, derived from a textural greyscale analysis of lumbar spine dual-energy X-ray absorptiometry (DXA) images, offers a non-invasive and indirect evaluation of bone microarchitecture. This method potentially enhances the assessment of skeletal phenotypes beyond the scope of aBMD. We aim to explore the utility of TBS in assessing vertebral fractures and spinal deformity in children and adolescents with OI. METHODS: In this cross-sectional study, 153 children and adolescents with OI were enrolled. DXA was used to measure TBS and aBMD, and their Z-scores were calculated based on reference values for BMD and TBS in normal children and adolescents with the same age and sex. Lateral thoracolumbar films were used to evaluate vertebral fractures and calculate the spine deformity index (SDI). The accuracy of TBS and aBMD for identifying vertebral compression fractures (VCFs) was assessed using area under the curve (AUC). RESULTS: TBS Z-score was negatively correlated with the age of children with OI (r = - 0.435, P < 0.001) and was positively correlated to aBMD Z-score at the lumbar spine and femoral neck (both P < 0.01), even after adjusting for confounding factors. TBS Z-score was as effective as lumbar spine aBMD Z-score in discriminating VCFs (AUC, 0.667 vs 0.666, P > 0.05). Notably, in patients without densitometric osteoporosis, TBS Z-score demonstrated superior discriminative power for VCFs compared to lumbar spine aBMD Z-score (AUC, 0.719 vs 0.545, P < 0.05). In this population, only the TBS Z-score (r = - 0.358, P < 0.05), rather than the lumbar spine aBMD Z-score, was negatively correlated with the SDI. CONCLUSION: TBS has a close correlation with bone mineral density in children and adolescents with OI. In patients without densitometric osteoporosis, the Z-score of TBS is more effective than that of bone mineral density in assessing VCFs and spinal deformity, highlighting the potential of TBS in evaluating the risk of VCFs and monitoring the progression of spinal deformity.