Abstract
1743
Objectives To determine if MRI was superior to CT portion of PET/CT as mapping images for PET and if PET/MRI fusion was useful in evaluating colorectal cancer(CRC).
Methods We retrospectively reviewed 28 sets of FDG PET/CT and 3-Tesla MRI exams to evaluate CRC in the pelvis, which were performed within a month of each other(23 for recurrence). To assess the ability of mapping images for PET, CT(of PET/CT), T1- and T2-weighted MR images(all non-contrast) were graded regarding the conspicuity of suspected lesion with a 5-point confidence scale(1=definitely absent- 5=definitely present). Fused PET/CT, PET/T1 or T2 MR images, generated with a dedicated software, were evaluated with a 3-point scale regarding additional diagnostic information to PET alone(1=no addition- 3=significantly added) and overall image quality(1=poor- 3=excellent). CT- and MRI-related scales were compared.
Results Lesions of CRC were visualized on CT or MRI in 27 cases, inside the bowel in 5 and outside in 22. FDG uptake was observed in all 27 cases. The scale for the lesion conspicuity was higher on T2 MR(4.8) than CT(4.1, p<0.01), with more frequent point 5 on T2 MR than CT(24 vs 13 of 27). The scales for additional information to PET and for fusion image quality were also higher on T2 MR(2.8 and 2.8) than CT(2.4 and 2.4, p<0.01), respectively. Point 3 was observed more frequently on T2 MR than CT on both evaluations. Same results were obtained if lesions were limited to ≤3cm(n=13) or ≤SUVmax of 3.5(n=7). Scales for T1 MR images were similar to CT in all evaluations.
Conclusions FDG PET/T2-w MRI is likely to be better than PET/CT in evaluating CRC, providing better lesion visualization and characterization.
- © 2009 by Society of Nuclear Medicine