Colin G. Miller, PhD • Hui Jing Yu, PhD • Blaine Horvath • Jessie Libber • Diane Krueger • Neil Binkley, MD
BioClinica, Inc., Newtown, PA 18940, 2University of Wisconsin–Madison, Madison, WI 53705
The purpose of this study was to assess the use of Dual Energy X-Ray Absorptiometry (DXA) phantoms as part of the validation process of DXA instruments in clinical trials.
An in vitro calibration was carried out using three DXA spine phantoms: the Bona Fide Phantom (BFP, BioClinica, Inc., Newtown, PA), GE (Lunar, Madison WI) encapsulated spine phantom, the Hologic (Bedford MA) encapsulated spine phantom, and one whole body phantom: the BioClinica Body Composition Phantom (BBCP, BioClinica, Inc., Newtown, PA). The BBCP was developed specifically for quality control and monitoring machine drift of body composition the the major DXA manufacturers. The unique phantom design measures 36 cm in width and 61 cm in length and weighs 29 kg. The phantom contains high-density polyethylene, polyvinyl chloride, and aluminum plates for simulation of bone.
Each phantom was scanned 30 times consecutively on an existing iDXA instrument, with 15 minutes wait between each scan series. Total Bone Mineral Density (BMD) values for L1-L4 were calculated for the Lunar phantom, Hologic phantom and BFP. Total BMD values, total fat mass and total lean mass for whole body were calculated for the BBCP. This was repeated on a newly installed iDXA instrument. The coefficient of variation (%CV) was calculated for each of the scan series. The mean difference between the two instruments was also calculated.
There was a systematic increase in %CV observed for the new iDXA instrument when compared to the existing iDXA instrument. The %CVs for the new iDXA were 0.19% for Lunar phantom, 0.33% for Hologic phantom, 0.21% for BFP, and 0.46% for BBCP, whereas the %CV for the existing iDXA were 0.12% for Lunar phantom, 0.20% for Hologic phantom, 0.18% for BFP, and 0.27% for BBCP. BMD measured on the new iDXA instrument was approximately 1% higher than the existing instrument for L1-L4 and about 2% higher for whole body measure, although the correlations were not statistically significant between the two instruments. Moreover, similar findings were observed using the BBCP for total fat mass and total lean mass. These results were sent to GE for evaluation and it was confirmed the new iDXA instrument required adjustment, and that the tube head was malfunctioning. Adjustments were made to the new iDXA and the %CVs improved in alignment to the other instrument. In addition, the %CVs from iDXA instruments were lower when compared to those obtained from a Prodigy scanner, where the %CVs from 30 consecutive scans on the Prodigy scanner were 0.88% for total BMD, 1.52% for total fat mass, and 0.96% for total lean mass.
This study demonstrated the importance of performing repeated phantom measurements for calibration and validation with new DXA scanners before using in the clinic or clinical trials. Understanding the values and %CV of each phantom tis critical to then understand the instrument performance. However, since this is only a sample of 1, further work would be needed to see if this "phantom signal" is reproducible. This is the first study where 4 phantoms were used on a single instrument which was not calibrated correctly. The BBCP is the first total body DXA scans of this new phantom showed that this semi-portable phantom may prove to be appropriate in clinical trials for the evaluation of body composition.
Presented at: The American Society for Bone and Mineral Research Annual Meeting, Baltimore, Maryland, USA, October 4 - 7, 2013