Thoracolumbar vertebral bone quality (TLVBQ) score as a possible screening tool for osteoporosis in patients with severe degenerative lumbar scoliosis: a retrospective cohort study.

PURPOSE: The conventional vertebral bone quality (VBQ) score may be difficult to apply in patients with severe degenerative lumbar scoliosis (DLS) because vertebral rotation, deformity, and degenerative changes can affect lumbar ROI placement. This study aimed to introduce a modified thoracolumbar vertebral bone quality (TLVBQ) score and evaluate its association with QCT-derived thoracolumbar vBMD and hip-defined osteoporosis in patients with DLS. METHODS: Patients with DLS treated at our institution from January 2018 to December 2025 were retrospectively analyzed. The study population was divided into a QCT-based cohort and a DEXA-based severe DLS cohort. In the QCT-based cohort, osteoporosis was defined using QCT-derived volumetric bone mineral density (vBMD). In the DEXA-based severe DLS cohort, osteoporosis was defined using hip-based DEXA criteria, namely a femoral neck or total hip T-score ≤ -2.5. Correlation analyses were performed to evaluate the relationships between TLVBQ and bone assessment parameters. Receiver operating characteristic (ROC) analysis was used to assess the screening performance of TLVBQ. A combined TLVBQ and T12-L2 HU model was further constructed using binary logistic regression. RESULTS: A total of 193 patients with DLS were included. TLVBQ was significantly higher in the osteoporosis group than in the non-osteoporosis group in both cohorts (p < 0.001). In the QCT-based cohort, TLVBQ showed a strong negative correlation with T12-L2 vBMD (r = -0.724, p < 0.001). In the DEXA-based severe DLS cohort, TLVBQ showed weak-to-moderate correlations with DEXA-derived BMD and T-scores (r = -0.304 to -0.408, p < 0.005). For identifying hip-defined osteoporosis in patients with severe DLS, TLVBQ achieved an AUC of 0.774 (95% CI 0.680-0.868). At the optimal threshold of 3.649, TLVBQ showed a sensitivity of 73.0%, specificity of 71.7%, PPV of 61.4%, and NPV of 81.1%. The combined TLVBQ and T12-L2 HU model yielded an AUC of 0.850 (95% CI 0.769-0.932). CONCLUSIONS: The TLVBQ score was associated with QCT-derived thoracolumbar vBMD and hip-defined osteoporosis in patients with DLS. TLVBQ may serve as a convenient MRI-based vertebral bone quality indicator for osteoporosis risk screening in patients with severe DLS, particularly when conventional VBQ measurement is challenging. However, TLVBQ should be interpreted as complementary to, rather than a replacement for, established osteoporosis assessment methods. A TLVBQ score > 3.649 should prompt further bone assessment with DEXA or QCT, particularly in patients with additional clinical risk factors.