Vertebral CT Hounsfield units in postmenopausal women with osteoporotic vertebral compression fracture: identification and validation of reference intervals.

PURPOSE

Based on the population of postmenopausal women with osteoporotic vertebral compression fractures (OVCFs), this study aimed to identify the optimal and alternative levels of vertebral CT Hounsfield units (HU) for osteoporosis and osteoporosis-related complication assessment, establish age-specific reference intervals (RIs) of HU values, and validate its quantitative predictive value for new vertebral fractures (NVFs) after percutaneous kyphoplasty (PKP).

METHODS

Consecutive postmenopausal women diagnosed with OVCFs at our department between January 2016 and August 2024 were retrospectively reviewed. The vertebral HU of T12-L2 was measured on CT images by two independent spine surgeons twice, with a 2-week interval. The segmental average HU was assessed in terms of the representativeness of overall osteoporotic status, reproducibility, and association with clinical outcomes to identify the optimal and alternative levels. Age-specific RIs were built using the indirect Hoffmann method. The associations between HU and NVFs were assessed by correlation, receiver operator characteristic (ROC) curve, and multivariate analyses.

RESULTS

A total of 922 patients were enrolled in the optimal level identification and RI establishment study. Intraclass correlation coefficient (ICC) between segmental and average HU values was the highest at L1 (ICC, 0.970), followed by T12 (ICC, 0.955) and L2 (ICC, 0.955) in the whole population. The age-specific RIs determined in postmenopausal women with primary OVCFs were 39.22-170.92 HU (56-65 years), 21.23-132.48 HU (66-75 years), and 11.15-108.85 HU (> 75 years) at T12; 37.25-156.46 HU (56-65 years), 17.83-123.68 HU (66-75 years), and 10.71-103.59 HU (> 75 years) at L1; and 30.88-148.28 HU (56-65 years), 9.61-121.00 HU (66-75 years), and - 1.67 to 99.65 HU (> 75 years) at L2. Significant weak negative correlations were found between NVFs and average/segmental HU value (Spearman r, - 0.146 to - 0.245, P < 0.05), and risks of nonadjacent NVFs, fracture cascade, and overall NVFs after PKP increased in the individuals with decreases in HU (particularly at L1).

CONCLUSION

This study identified the optimal and alternative levels of CT HU value in postmenopausal women with OVCFs and established its corresponding age-specific RIs. Furthermore, we validated that low HU value posed high risks of NVFs after PKP and quantitatively clarified the dynamic trend of their association. This study may provide inspiration and a novel methodological approach for further research on osteoporotic diseases.