Diagnostic accuracy of colorectal cancer staging with whole-body PET/CT colonography

Context: Staging of patients with colorectal cancer often requires a multimodality, multistep imaging approach. Colonography composed of a combined modality of positron emission tomography (PET) and computed tomography (CT) provides whole-body tumor staging in a single session. Objectives: To determine the staging accuracy of whole-body PET/CT colonography compared with the staging accuracies of CT followed by PET (CT + PET) and CT alone and to evaluate the effect of PET/CT colonography on therapy planning compared with conventional staging (CT of the abdomen and thorax and optical colonoscopy). Design, Setting, and Patients: Prospective study of 47 patients enrolled between May 2004 and June 2006 with clinical findings and optical colonoscopy that suggested primary colorectal cancer (mean [SD] age, 71 [11] years; range, 47-92 years). Patients underwent whole-body PET/CT colonography 1 day after colonoscopy. The study was conducted at a university hospital with a mean (SD) follow-up of 447 (140) days (range, 232-653 days). Main Outcome Measures: Correct classification of overall TNM stage using PET/CT colonography compared with CT + PET and CT alone. Secondary outcome measures were the accurate assessment of T-stage, N-stage, and M-stage by PET/CT colonography compared with CT + PET and CT alone and the effect of PET/CT colonography on therapy planning. Results: Of the 47 patients with a total of 50 lesions, the overall TNM stage was correctly determined for 37 lesions with PET/CT colonography (74%; 95% confidence interval [CI], 60%-85%), 32 lesions with CT + PET (64%; 95% CI, 49%-77%), and 26 lesions with CT alone with a 0.7-cm node threshold (52%; 95% CI, 37%-66%). Compared with optimized abdominal CT staging alone, PET/CT colonography was significantly more accurate in defining TNM stage (difference, 22%; 95% CI, 9%-36%; P=.003), which was mainly based on a more accurate definition of the T-stage. Differences were not detected for defining N-stage between PET/CT colonography and CT alone with a threshold of 0.7 cm for malignant nodes but were detected with a threshold of 1 cm. Differences were not detected in defining M-stage separately or when comparing the accuracies of PET/CT colonography with CT + PET. PET/CT colonography affected consecutive therapy decisions in 4 patients (9%; 95% CI, 2.4%-20.4%) compared with conventional staging (CT alone and colonoscopy). Conclusions: In this preliminary study, PET/CT colonography is at least equivalent to CT + PET for tumor staging in patients with colorectal cancer. Thus, PET/CT colonography in conjunction with optical colonoscopy may be a suitable concept of tumor staging for patients with colorectal cancer.