Abstract
1691
Objectives For individuals treated for hyperlipidemia with statins, FDG vascular imaging may help guide further cardiovascular risk stratification for more intensified or varied treatment. However, the methods of vascular FDG quantification are currently under question. Division or subtraction with blood pool FDG activity is assumed to normalize for blood activity. However, glomerular filtration rate (GFR) is known to modify blood FDG activity; which is potentially also related to glucose at the time of imaging and statin treatment. These relationships will be tested with blood, liver, and target activity along with alternate methods of quantification.
Methods 74 PET/CTs from 34 subjects with hyperlipidemia > 55 y/o were prospectively imaged under standard conditions (uptake 136 min, iterative reconstruction with 24 iterations and 3 subsets, 256x256 matrix, 1.5mm slice). Lean body mass (LBM) was used for dosing and for SUV. The descending aorta (DAo) distal to the arch to the bifurcation was segmented (~300 slices) for average SUV max DAo activity. The left jugular and superior/inferior vena cava were segmented for mean blood SUV. Mean liver activity was also calculated. Target to background ratios (TBR) using blood and liver and target-blood subtraction was calculated. The generalized estimating equation was used for multivariate modeling of measurement methods relative to glucose, statin equivalent dose (normalized to simvastatin), and mean blood activity (or GFR, where appropriate). The model was also corrected for gender, age, and body mass index. Normalized beta is reported with p-values.
Results Shown in the table, TBR or subtraction using blood retains significant relationships to blood activity while TBR using liver does not. TBR-Blood has a negative relationship to blood while subtraction has a positive relationship. Glucose and statin treatment does not significantly modify blood, liver, or target activity in this model.
Conclusions Correction for blood activity is important because maximum vessel SUV is most related to it compared to all other variables considered. TBR or subtraction with blood does not sufficiently correct for blood activity, which is not explained by variation in glucose or statin treatment. Despite the close relationship of liver and blood activity, TBR using liver does not retain any significant relationship to blood activity, glucose or statin treatment; which may make it preferable for FDG vascular imaging. Further analysis is needed in statin naïve or more heterogeneous populations.