Detection and evaluation of colorectal neoplasms with F-18 FDG PET/CT in asymptomatic subjects

Objectives To evaluate the performance of F-18 FDG PET/CT for the detection and characterization of colorectal neoplasms in asymptomatic subjects.

Methods In a prospective study, 10,166 asymptomatic subjects underwent F-18 FDG PET/CT as a part of a screening program. Colorectal FDG focality on initial PET with intensity exceeding background FDG accumulation that existed on delayed PET/CT, following administration of a laxative-augmented contrast medium, was considered an abnormal lesion. Colonoscopic histopathology and clinical follow-up served as the reference standard. On a per-lesion basis, each colorectal neoplasm was characterized by two PET/CT parameters including standardized uptake value (SUV) and size. The performance of each parameter in identifying colorectal cancer was determined by the area under the receiver operating characteristic curve (AUC).

Results Of 78 colorectal lesions identified on PET/CT, 28 were identified as colorectal cancers and the remaining 50 were adenomas. Additionally, two false negatives (rectal adenocarcinoma and carcinoid tumor) and 19 false positives (hemorrhoid, diverticulitis, colitis, erosion, and chronic granulomatous inflammation) were noted. The AUCs of neoplastic SUV and size to detect colorectal cancer were poor and non-significantly (0.695 ± 0.067 vs. 0.613 ± 0.079, p = 0.43). Combining neoplastic SUV and size to predict colorectal cancer, AUC significantly improved to 0.825 ± 0.058 (p < 0.001) with the best simultaneous sensitivity of 87.8% and specificity of 75.0%. The optimal cutoff values of neoplastic SUV and size were 3.54 and 2.15 cm, respectively.

Conclusions F-18 FDG PET/CT provides promising performance for the detection and evaluation of colorectal neoplasms in asymptomatic subjects. The combination of neoplastic SUV and size can improve the accuracy of differentiating benign and malignant neoplasms in colorectum