Osteoporotic fractures are one of the main causes of disability worldwide and are associated with increased mortality. Despite the clinical relevance and growing prevalence of the disease due to the aging population, only the minority of affected elderly men and women undergo bone mineral density (BMD) screening with one of the established methods, such as dual-energy x-ray absorptiometry (DXA) or quantitative CT (qCT), at least once in their life. Among the reasons for this, dedicated examinations are associated with additional scan time, organizational efforts, and additional costs. Moreover, these examinations expose patients to ionizing radiation.

Therefore, previous studies have investigated whether BMD-equivalent values can be obtained from already existing imaging data such as routine multidetector-row CT (MDCT), and whether patients benefit from this additional information in the clinical context. Several different approaches exist to date, ranging from easy-to-implement attenuation measurements in asynchronously calibrated CT examinations to more complex analyses such as bone strength assessments based on finite element models, or texture analysis.

In our study, we developed conversion equations for BMD-equivalent values derived from both routine MDCT and MDCT myelography examinations of the lower spine, with qCT values as standard of reference.

Regions of interest (ROIs) were manually placed in the vertebrae, and attenuation values were converted, adding less than 2 minutes of image analysis per examination. With the use of the converted values it was possible to distinguish between patients with versus without existing vertebral fractures as well as to predict incident fractures in all patients. Of note, it was also possible to predict screw loosening in patients who underwent pedicle screw-based spondylodesis over a follow-up period of 15 months.

The latter finding especially is of high clinical relevance because the additional information on the patient’s bone stability obtained by routine MDCT enables appropriate therapeutic planning—eg, the selection of alternative techniques for screw fixation—without causing any additional radiation, scan duration, or costs. At our institution, the assessment of BMD-equivalent values has become an indispensable component of presurgical image analyses, as pedicle screws are cemented in case of low BMD.

Nevertheless, further development of the presented method is necessary in order to improve interexamination reproducibility for reliable follow-up assessment or therapy monitoring. Conversion equations are scanner-specific; however, the method can be applied to any scanner.

In a further clinical study, presented at RSNA 2015, MDCT-derived BMD values, among other imaging parameters, were assessed in patients with painful segmental instability of the lower spine who underwent dynamic posterior stabilization. Interestingly, patients with higher baseline BMD showed significantly higher clinical improvement over 24 months compared with patients with low BMD values.

We believe that these findings underline the clinical relevance of alternative approaches for the assessment of bone stability and that the investigation of methods for obtaining additional quantitative data from clinical scans should be further pursued.

 

Read this article at AJNR.org …