Non-Invasive Estimation of Subcutaneous and Visceral Adipose Tissue

Subcutaneous adipose tissue blood flow is finely regulated, and tuned with fat metabolism; little is known about visceral fat, which is less accessible in humans. In estimating blood flow with positron emission tomography (PET) and oxygen-15-labelled water ([ 15 O]H2O), the input function is obtained invasively from arterial blood samples. The aim of the current study was to validate a non-invasive method to measure blood flow in adipose depots, by extracting input function curves from [ 15 O]H2O-PET images. Data of twenty subjects undergoing abdominal [ 15 O]H2O-PET were used. Images were reconstructed with filtered backprojection (FBP). Location, diameter, and inner radioactivity levels of the abdominal aorta were automatically determined. Image derived arterial curves (IDI) were compared to measured arterial blood data, as obtained by an online blood sampler (OSI). Blood flow in three adipose tissue depots was estimated using the autoradiographic method with OSI vs the FBP image derived input (F-IDI) function. Correlations between blood flow results obtained with OSI and IDI were significant (r 0.87, p<0.0001) in all regions. Estimates of the aortic diameter ranged between 10.7-17.2 mm. A good agreement was found between area under the curve (AUC) values of F-IDI and OSI curves; the AUCF-IDI/AUCOSI ratio was 0.97±0.10. Our results support the implementation of the current method for the non-invasive detection of the abdominal aorta input function from a dynamic [ 15 O]H2O PET image in the quantification of regional blood flow in low flow tissues. This method allows simultaneously examine subcutaneous and intraabdominal fat depots.