ABSTRACT
BACKGROUND: To validate the feasibility of water enema PET/CT (WE-PET/CT) in incidental colorectal 18F-FDG uptake and improve the accuracy of diagnosing colorectal neoplastic lesions. METHODS: We retrospectively analysed the electronic records of 338 patients undergoing common PET/CT and WE-PET/CT at our hospital. PET/CT results were correlated with colonoscopy pathology and follow-up results. The ROC contrast curve was plotted to evaluate the accuracy of SUVmax on common PET/CT and WE-PET/CT for detecting neoplastic lesions. SUVmax and the median retention indexes (RIs) of cancerous, precancerous, and benign lesions and physiologic uptake were compared. RESULTS: The sensitivity, specificity and accuracy of diagnosing neoplastic lesions with common PET/CT were 84.0%, 78.3% and 80.2%, respectively. The corresponding results with WE-PET/CT were 95.8%, 96.5% and 96.2%. The AUC of SUVmax on WE-PET/CT was significantly higher than that on common PET/CT (0.935 vs. 0.524, p < 0.001). The median SUVmax on WE-PET/CT was significantly higher than that on common PET/CT in cancerous and precancerous lesions, and significantly decreased in benign lesions and physiologic uptake (p < 0.001). The RI was significantly different between cancerous lesions and physiologic uptake, between precancerous lesions and physiologic uptake, between benign lesions and physiologic uptake, and between cancerous and benign lesions (p < 0.05). CONCLUSIONS: WE-PET/CT is a noninvasive, well-tolerated and effective technique for diagnosing incidental colorectal 18F-FDG uptake. It is helpful for a timely colonoscopy and can effectively avoid an unnecessary colonoscopy for incidental colorectal 18F-FDG uptake.