PET-18F-FDG Pharmacokinetic Modeling without Blood Sampling in Arteries with Atherosclerosis

Pharmacokinetic modeling of 18F-FDG in the arteries is necessary in comparison to the standard uptake value (SUV) and tissue-to-blood ratio (TBR) for two main reasons: the small thickness of the artery wall (around 2.3 mm) and the domination by activity emanating from blood. Consequently, the artery image is subject to partial volume effect (PVE) and it is not appropriate to use an input function IF determined from an artery. In the present work, we used a modified pharmacokinetic model derived from the classical two-tissue compartment model of 18F-FDG. In the classical model, the artery time-activity curve (TAC) is fitted with the model convolved with the IF to calculate the rate constants together with tissue blood volume (TBV) which is a fraction of IF (TVF = k
₅
IF, k
₅
<; ). In the modified model, the images are decomposed with factor analysis (FA) to provide blood image. The same region of interest of the artery is used in the blood image to determine TBV. The model is therefore rewritten to convolve the model with TBV/k
₅
. Such model, called without blood sampling (WOBS) has several advantages mainly, since the same region of interest (ROI) is used for both TBV and the artery, PVE cancels out. We applied these two models and SUV to artery images in patients with atherosclerosis, with and without medication. For the classical model, artery TAC and IF, determined on the sagittal view of the aorta, were corrected for PVE. The metabolic rate of glucose (MRG) was calculated in 710 artery image slice with both models. We also compared MRG as a function of the ratio of CT calcification area (RCA) and Agatston calcification scores (ACS). TAC fits with WOBS were more accurate in almost all artery TACs and MRG had globally less variation with WOBS. MRG values in non-medication group were statistically significantly different in both models. WOBS and the classical model also detected the same differences between RCA and ACS.