Incidental Findings with Dual-Energy X-Ray Absorptiometry: Spectrum of Possible Diagnoses

New dual-energy X-ray absorptiometry (DXA) technologies provide improved spatial resolution and high image quality. Our purpose was to review DXA examinations to detect collateral findings and to understand their potential impact on patient healthcare. We retrospectively and randomly reviewed 739 DXA examinations (191 of 739, 25.8 %, whole body; 96 of 739, 13.0 %, vertebral fracture assessment; 231 of 739, 31.3 %, lumbar spine; 221 of 739, 29.9 %, femur) that were performed in our institution with a new DXA equipment. Whenever an extra finding was discovered, the physician's report was read and the clinical history of the patient was investigated to understand whether that finding was already known, as well as to check the diagnosis. The population included 208 male and 531 female subjects (58 +/- A 14 years old). Incidental findings were detected in 117 (15.8 %) of 739 DXA examinations (17 of 117, 14.5 %, whole body; 41 of 117, 35.0 %, vertebral fracture assessment; 32 of 117, 27.4 %, lumbar spine; 27 of 117, 23.1 %, femur): biliary and urinary stones (4.8 %), vascular calcifications (33.7 %), other soft tissue calcifications (25.3 %-e.g., tendons, lymph nodes, intraparenchymatous calcifications), vertebral abnormalities (14.5 %), other bone abnormalities (12.1 %), and morphovolumetric alterations or abnormal anatomical structures (9.6 %). Among all these findings, 50 (42.7 %) of 117 could be verified by other imaging modalities. Forty-nine (98.0 %) of 50 incidental findings were identified as true findings, and DXA was able to orient the diagnosis (exact diagnosis in 37 of 50, 74.0 %); however, none of them was mentioned on available DXA reports. An interpreting physician should treat the DXA image with the same attention given to any other X-ray image. Sometimes DXA may allow a qualitative diagnosis of collateral findings. However, potential negative effects on healthcare economy should be considered for false-positive or insignificant findings.