Abstract
1692
Objectives Reproducible quantitative measurement of vessel wall inflammation is important for monitoring disease state among patients as well as between scan instances for individual patients. We aim to describe the effect that slice position has on measurements of the aorta.
Methods 10 prospectively enrolled participants (65yrs±4.5, 70% male, 83kg±18) underwent two 18FDG PET/CTs (Dose=321MBq±56, Uptake Time=139min±5) 6-months apart on a single Siemens Biograph mCT. Scanning was caudocranial for 6-7 bed positions (3min/bed). Iterative reconstruction (2 iterations, 24 subsets, 2.0 FWHM Gaussian filter) was utilized to produce 3mm slice, 256x256 PET data fused to CT. Spherical regions of interest were drawn on the ascending (AAo) and descending (DAo) aorta at the carina (S0). The DAo was measured in 50mm increments for 150 slices below S0 (S50, S100, S150). Linear regression was used to measure slice-dependent changes in blood pool (BLP) activity, with correction for ROI area. Axial slice position related to max SUV and Target to Background Ratios (TBR = same slice max SUV of whole aorta/mean SUV of lumen) used simple linear regression.
Results Regression is reported in terms of average% caudocranial change over the 150-slice interval with 95%CI. There was a -16%(-2% to -19%) linear decrease in BLP uptake with more cranial slice location. After adjustment for a 93% increase in ROI size (avg 1.5cm2 at S150 and 2.9cm2 at S0), there was an -11%(-2% to -20%) change. TBR increased by 15%(7% to %23) over the interval while max SUV of the whole aorta was unchanged (6.6%, -2.7% to 16.0%).
Conclusions Progressive blood clearance of radiotracer due to time delays between bed positions may be a significant source of changes in background BLP activity within the aorta, which may affect the accuracy and comparability of TBR measurements.
Research Support Supported by the NIH intramural research program and the Imaging Sciences Training Program (ISTP).