Accuracy of FDG-PET in the Detection of Colonic Neoplasms

Abstract

PURPOSE: There is little published data on the sensitivity and specificity of F-18 fluorodeoxyglucose positron emission tomography (FDG-PET) in the detection of colon polyps and early colon cancer. We performed a prospective evaluation of the utility of FDG-PET for the detection of colonic lesions in patients referred for polypectomy or endoscopic mucosal resection. METHODS: 29 patients who were referred for polypectomy or endoscopic mucosal resection of colon polyps, flat lesions and early cancers underwent FDG-PET. A total of 10 cancers and 27 precancerous lesions (adenoma, carcinoma in situ) were present in these patients. Histological confirmation was available in all cases. In cases where complete removal could not be performed endoscopically, surgery was performed. FDG-PET scans were considered true positives when a hypermetabolic abnormality was detected in the region where a lesion was identified on colonoscopy and false positives when a lesion was detected in an inappropriate location based on the colonoscopy results. RESULTS: FDG-PET detected 100% (3/3) of cancers larger than 1.5 cm but only 29% (2/7) of cancers up to 1.5 cm in diameter. 3 of the 4 polypoid cancers were detected, while a 1 cm polypoid cancer was not. Among the flat cancers, which were all 1.5 cm or smaller, only 2/6 were detected. Of the precancerous lesions, FDG-PET detected 83% (5/6) of lesions 3 cm or larger, 50% (2/4) of lesions between 1.6 and 2.9 cm, and 18% (3/17) of lesions between 1 and 1.5 cm. Only 2/9 flat precancerous lesions were detected. There were no false positive readings. CONCLUSIONS: In this prospective study, FDG-PET was highly specific but had limited sensitivity for precancerous and cancerous lesions smaller than 1.5 cm in diameter. Most polypoid adenomas and all polypoid carcinomas larger than 1.5 cm were detected by FDG-PET, while most flat adenomas and flat carcinomas were missed.