Comparison of PET images obtained with PET-CT vs PET-MR: A pilot study

551 Objectives Compare PET images obtained using PET-CT vs PET-MR for lesion detection, artifacts, altered biological distribution(BD), and standard uptake values (SUV’s). Methods A prospective IRB-approved study was conducted at NIH for patients referred for PET-CT and giving informed consent for further imaging by PET-MR, if considered of potential diagnostic significance. PET-MR was performed within 90 minutes after PET-CT, using same initial radioisotope injection. The reconstruction technique was different between the 2 scanners (post-filtering and resolution recovery), but consistent. SUV calculation was identical in all studies. Administered activity was 9.4 -16.2 mCi for F18-fluorodeoxyglucose (F18FDG) and 12.3 mCi for F18DOPA. Three physicians evaluated all images, with side-by-side comparison of PET images from PET-CT and PET-MR. Results 15 paired studies were performed following a single F18-FDG administration (14 patients). 1 patient had F18-DOPA administration. All FDG-avid lesions (>80) were identified on both PET-CT and PET-MR at the same sites. On the F-DOPA study two lesions seen on PET/CT were not seen on PET/MR. Artifacts causing poor image quality were observed in 2 PET-MR studies due to urinary bladder activity and metallic implants. The BD was identical in 11 paired studies; 5 studies showed some difference in BD. The individual SUVs for PET-CT vs. PET-MR were 23.0 vs. 21.0 for average max SUV, and 2.7-121.0 vs. 2.7-90.7 for max SUV range. These variations are tentatively attributed to differences in attenuation correction methods and uptake time. Conclusions This pilot study indicates that F18FDG-PET performed on PET-MR detected the same number of lesions as on the PET-CT. Overall PET-MR image quality was good, except in a few pts with reduced quality of PET images from PET-MR, mostly related to artifacts. The BD of the 18F-FDG was similar, while the F-DOPA study showed differences, possibly related to lesions biology