Category: Bone Densitometry or Body Composition

5 - Incidental Screening for Osteoporosis in the Presence of Contrast Agents using No Dose Computed Tomography

Background: Incidental screening as systematic dose-free extraction of bone status information with Asynchronous QCT from existing CT scans may help reduce under-diagnosis and under- treatment of osteoporosis. Spine and femur measurement bias induced by IV-contrast agent and possible bias compensation require further investigation.

Patients were recruited receiving routine native, arterial, and portal phase (Imeron 350, Bracco Imaging Deutschland GmbH, Konstanz, Germany) CT scans (GE Revolution EVO, Waukesha, WI, USA). From 45 patients (age 69.9 +/- 12.8 years) bone mineral density from three vertebrae and the adjacent aorta was obtained (QCT Pro Asynchronous 3D Spine Version 6.1, Mindways Software, Inc., Austin, TX, USA), abdominal area was determined at level L1 (QCT Pro Tissue Composition, Research Module Version 0.1.6). From 33 patients (age 67.7 +/- 12.5 years) bone mineral density at total hip and femoral neck regions was retrieved (QCT Pro Asynchronous CTXA Hip). Statistical Analysis was performed using GNU-R 3.3.2 (R-Foundation for Statistical Computing, Vienna, Austria).

Generalized Additive Models (GAM) of non-linear basis functions (splines) were used to predict native vertebral bone mineral density (BMD). For the arterial phase, a model based on gender, vertebral and aortical BMD, and the interactions of age with vertebral and aortical BMD as well as abdominal area predicted native BMD best (R2=97.9%, residual error = 3.69 mg/ccm). For the portal phase, a model including gender, vertebral and aortical BMD and the interaction of age with aortical BMD predicted native BMD comparably well (R2=97.9, residual error=4.14 mg/ccm).
Density at the total hip (arterial and portal phase) and the femoral neck (arterial phase) was not significantly different between arterial and native scans, significant differences were observed in the portal phase at the femoral neck (aBMD difference: mean 0.013 g/cm2 (11% of a standard deviation)), p < 0.01; vBMD difference: mean 3.9 mg/ccm, p < 0.001).
For the spine, regression results allow for the estimation of native BMD values from both arterial and portal scans. With the exception of a clinically non-significant difference in femoral neck density within the portal phase, hip measurements were not significantly different between contrast enhanced and native phases.

Summarizing, this supports the application of asynchronous QCT on contrast enhanced CT scans for the purpose of incidental screening. Patients identified to have low BMD without exposure to additional radiation dose using existing native or contrast enhanced CT scans can be forwarded to standard diagnostic procedures for the determination of the osteoporotic status.

Reimer Andresen

Head of Department
Institute of Diagnostic and Interventional Radiology/Neuroradiology, Westkuestenklinikum Heide, Academic Teaching Hospital of the Universities of Kiel, Luebeck and Hamburg, Heide, Germany
Heide, Schleswig-Holstein, Germany

J. Keenan Brown

President, Director of R&D
Mindways Software, Inc.
Austin, Texas

Claus-C. Glüer

Professor of Medical Physics
Section Biomedical Imaging, Department of Radiology, University Hospital Schleswig-Holstein, Campus Kiel
Kiel, Schleswig-Holstein, Germany

Wolfram Timm

VP Scientific and Engineering Services
Mindways Software, Inc.
Kiel, Schleswig-Holstein, Germany