# Opportunistic osteoporosis screening in primary hyperparathyroidism using routine CT: validation of a standardized vertebral cBMD index

## TL;DR

This study shows that a new CT-based bone density index can accurately detect osteoporosis in patients with primary hyperparathyroidism and is linked to parathyroid hormone levels.

## Contribution

The study introduces a phantom-less CT-derived bone density index that offers high diagnostic accuracy for osteoporosis in PHPT patients.

## Key findings

- Vertebral cBMD strongly correlates with QCT-vBMD (r > 0.95, P < 0.001).
- L2-cBMD detects osteoporosis with high sensitivity (100%) and specificity (92.5%).
- cBMD is independently associated with serum PTH levels (P = 0.010), unlike QCT-vBMD.

## Abstract

To evaluate the diagnostic performance of a routine CT-derived vertebral bone density index, the standardized percentage change in bone mineral density (cBMD), for detecting osteoporosis in patients with primary hyperparathyroidism (PHPT), and to explore its association with parathyroid hormone (PTH) compared to Quantitative CT (QCT).

This retrospective study included 175 consecutive patients with biochemically confirmed PHPT. QCT–derived volumetric bone mineral density (vBMD), obtained using asynchronous calibration, served as the reference standard. cBMD was calculated from routine non-contrast CT images using a phantom-less normalization method. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. Multivariable logistic regression was used to assess the incremental diagnostic value of cBMD beyond demographic and biochemical variables, and multivariable linear regression was performed to compare independent determinants of cBMD and QCT-vBMD.

Vertebral cBMD showed a strong linear correlation with QCT-vBMD across T12–L3 (r > 0.95, P< 0.001). For osteoporosis detection, L2-cBMD achieved an area under the ROC curve (AUC) of 0.992, with 100.0% sensitivity and 92.5% specificity. Incorporation of L2-cBMD into multivariable models significantly improved diagnostic performance compared with models including demographic and biochemical variables alone (AUC increase from 0.937 to 0.996, P< 0.001). In linear regression analyses adjusted for age and body mass index, cBMD showed an independent association with serum PTH levels (P = 0.010), whereas QCT-vBMD did not (P = 0.398).

cBMD is a robust, phantom-less tool derived from routine CT images that demonstrates high diagnostic accuracy for osteoporosis in PHPT. Its association with serum PTH highlights its potential utility as a complementary imaging biomarker for opportunistic bone health screening in this population.