Abstract
OBJECTIVE: To determine the performance of FDG-PET/CT in the detection of relevant colorectal neoplasms (adenomas > or =10 mm, with high-grade dysplasia, cancer) in relation to CT dose and contrast administration and to find a PET cut-off. METHODS: 84 patients, who underwent PET/CT and colonoscopy (n = 79)/sigmoidoscopy (n = 5) for (79 x 6 + 5 x 2) = 484 colonic segments, were included in a retrospective study. The accuracy of low-dose PET/CT in detecting mass-positive segments was evaluated by ROC analysis by two blinded independent reviewers relative to contrast-enhanced PET/CT. On a per-lesion basis characteristic PET values were tested as cut-offs. RESULTS: Low-dose PET/CT and contrast-enhanced PET/CT provide similar accuracies (area under the curve for the average ROC ratings 0.925 vs. 0.929, respectively). PET demonstrated all carcinomas (n = 23) and 83% (30/36) of relevant adenomas. In all carcinomas and adenomas with high-grade dysplasia (n = 10) the SUV(max) was > or =5. This cut-off resulted in a better per-segment sensitivity and negative predictive value (NPV) than the average PET/CT reviews (sensitivity: 89% vs. 82%; NPV: 99% vs. 98%). All other tested cut-offs were inferior to the SUV(max). CONCLUSION: FDG-PET/CT provides promising accuracy for colorectal mass detection. Low dose and lack of iodine contrast in the CT component do not impact the accuracy. The PET cut-off SUV(max) > or = 5 improves the accuracy.