Opportunistic Osteoporosis Assessment and Fracture Risk Determination Using Cancellous Density Measurement in Hounsfield Units of Native Lumbar Computed Tomography Images-A Comparative Study with Conventional Bone Density Evaluation.

Osteoporosis is a global problem that will increase as the population increases and ages, requiring prevention, early detection, and appropriate treatment. An increasing loss in bone mineral density (BMD) is the hallmark of osteoporosis, leading to an increased risk for insufficiency fractures. We aimed to investigate and analyze the applicability of native lumbar spine computed tomography (CT) scans for the evaluation of bone density compared with standard bone density measurements with quantitative computed tomography (QCT) and computed tomography X-ray absorptiometry of the hip (CTXA). Patients who were referred to our institution for diagnostic investigations and underwent CT imaging of the lumbar spine, as well as standard osteoporosis assessments including QCT and CTXA, were included in the study, resulting in a total of 240 patients (mean age: 65.9 years, range: 24-91). An ANOVA test was used to compare patient groups without a fracture, with one fracture, with more than one fracture, and with additional sacral fractures. An ROC analysis was performed to assess the predictive power of fracture risk estimation considering HU, QCT, and CTXA values. At least one fracture was detected in 42.9% of these patients. For the lumbar spine, the median HU was 89.9 (range 67.9-126.9) and the median BMD was 73.7 (range 57.1-104.2) mg/cm. With a correlation coefficient of 0.98 ( < 0.001), the HU values obtained from native lumbar CT scans can be calculated using the following formula: BMDspine = 0.84 + (0.81 × HU). With HU values < 80 and a BMD of the lumbar spine < 66 mg/cm, a significantly increased number of osteoporotic vertebral fractures were found in the mid-thoracic, thoracolumbar, and sacral regions with an effect size of 0.89. In 32 patients (13.3%), additional sacral fractures were found; these patients showed the lowest density values with a median HU value of 31.8 (12.7-58.2). An ROC analysis of HU revealed a 93% sensitivity for the coincidence of a vertebral fracture. There was no significant difference compared with the AUC of QCT ( = 0.395) for concomitant vertebral body fractures. CTXA values also allowed for risk assessment but showed a significantly lower AUC. We found a negative correlation of BMD with age and a positive correlation of BMD with body mass index. Cancellous density measurements in HU values can be effectively converted into quantitative BMD values in mg/cm, enabling a reliable assessment of osteoporosis severity and fracture risk prediction. Further quantitative density evaluation of the hip does not add value to fracture risk assessment for the axial skeleton. Based on this study's findings, using HU values in native CT of the lumbar spine alone offers a viable, opportunistic approach towards fracture risk evaluation of the spine.