A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care.

Human body composition is important in numerous cancer research domains. Our objective was to evaluate clinically accessible methods to achieve practical and precise measures of body composition in cancer patients. Dual-energy X-ray absorptiometry (DXA)-based analysis of fat and fat-free mass was performed in 50 cancer patients and compared with bioelectrical impedance analysis (BIA) and with regional computed tomography (CT) images available in the patients' medical records. BIA overestimated or underestimated fat-free mass substantially compared with DXA as the method of reference (up to 9.3 kg difference). Significant changes in fat-free mass over time detected with DXA in a subset of 21 patients (+2.2 +/- 3.2%/100 days, p = 0.003), was beyond the limits of detection of BIA. Regional analysis of fat and fat-free tissue at the 3rd lumbar vertebra with either DXA or CT strongly predicted whole-body fat and fat-free mass (r = 0.86-0.94; p < 0.001). CT images provided detail on specific muscles, adipose tissues and organs, not provided by DXA or BIA. CT presents great practical significance due to the prevalence of these images in patient diagnosis and follow-up, thus marrying clinical accessibility with high precision to quantify specific tissues and to predict whole-body composition.