Admin_Adham
TOPLINE
Lumbar CT-derived Hounsfield unit (HU) measurements outperformed MRI-derived vertebral bone quality (VBQ) scores in diagnostic accuracy for identifying osteopenia and osteoporosis, supporting opportunistic bone density screening, a systematic review and meta-analysis found.
METHODOLOGY
- Researchers conducted a systematic review and meta-analysis of 42 studies involving 9214 participants across Asia, Europe, North America, and South America, with all studies evaluating HU or VBQ as diagnostic markers of osteoporosis or osteopenia.
- Investigators evaluated HU measurements of lumbar CT and MRI-derived VBQ scores referenced against DEXA or quantitative CT measurements.
- Continuous outcomes were pooled as standardised mean differences, and diagnostic accuracy was synthesised using bivariate hierarchical summary area under the receiver operating characteristic curve models to derive pooled areas under the curve (AUCs) and diagnostic odds ratios (DORs).
- Mean diagnostic threshold values were calculated across modalities and validated using random-effects models to determine pooled risk ratios for low bone density.
- Methodological quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 tool, with most domains rated as a low concern for bias.
TAKEAWAY
- For osteopenia, HU achieved higher diagnostic accuracy than VBQ, with a better AUC value (0.87 vs 0.75), sensitivity (0.79 vs 0.67), specificity (0.81 vs 0.74), and DOR (16.56 vs 5.79).
- For osteoporosis, HU outperformed VBQ, with a higher AUC value (0.86 vs 0.79), sensitivity (0.78 vs 0.73), specificity (0.81 vs 0.74), and DOR (14.58 vs 7.55).
- Pooled diagnostic cutoffs were 138.9 HU for osteopenia (AUC, 0.92; DOR, 36.74) and 104.6 HU for osteoporosis (AUC, 0.86; DOR, 16.37), whereas MRI-derived VBQ cutoffs of 2.83 and 3.14 showed lower discrimination (AUC, 0.70 and 0.68), respectively.
- Bootstrap validation showed greater diagnostic stability for HU than for VBQ, with corrected AUC reductions of -0.08 for osteopenia (P = .047) and -0.09 for osteoporosis (P = .006), confirming MRI-derived VBQ score attenuation.
IN PRACTICE
"HU and VBQ provide valuable opportunistic imaging biomarkers for assessing vertebral bone quality, yet HU consistently demonstrates superior diagnostic accuracy and stability across analyses," the authors wrote.
"These findings support the potential integration of HU into routine clinical workflows while encouraging further optimisation of MRI-based metrics for comprehensive bone-quality assessment," they added.
SOURCE
The study was led by Omar Lubbad, University of Sussex, Brighton and Hove, England. It was published online on August 18, 2026, in European Radiology.
LIMITATIONS
The study was limited by predominantly indirect comparisons between CT-derived HU and MRI-derived VBQ measurements; substantial interstudy heterogeneity in imaging protocols, scanner types, acquisition parameters, sampling methods, and reference standards; the mostly retrospective design of included studies; smaller sample sizes and greater methodological variability in VBQ evidence; the reliance on reconstructed analyses rather than patient-level data; and the evaluation of trabecular bone quality alone without assessing cortical microarchitecture, bone turnover, or the risk for systemic fracture.
DISCLOSURES
This study did not receive any funding. The authors reported having no conflicts of interest.
This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.
