Colonoscopy and multidetector-array computed-tomographic colonography: detection rates and feasibility

Computed-tomographic (CT) colonography has been introduced as a minimally invasive colon examination for the detection of colorectal neoplasms. The aim of this study was to compare the performance characteristics of multidetector-array CT colonography (MDCTC) and conventional colonoscopy in a prospective, blinded design.

Sixty-six symptomatic patients, 75 patients undergoing polyp and cancer surveillance, and seven patients undergoing preoperative colonoscopy due to colorectal cancer (CRC) were examined with MDCTC and subsequent colonoscopy. The gold standard was colonoscopy. If MDCTC was positive and the first-pass colonoscopy was negative, a second-pass colonoscopy served as the gold standard.

Complete colonoscopy was achieved in 91% of the patients, while technically satisfying MDCTC was obtained in 76% of the patients (P < 0.01), insufficient air distension in the sigmoid colon being the main problem. MDCTC and colonoscopy both detected all 11 carcinomas. Overall detection rates for polypoid lesions 6 mm or larger in size were 81% (95% CI, 70% to 90%) for MDCTC and 87% (95% CI, 77% to 94%) for colonoscopy (P = 0.52), with a significant difference with regard to the detection of polyps 6-9 mm in size in favor of colonoscopy (P = 0.008). The specificity of MDCTC at a 6-mm level was 97% (95% CI, 92% to 99%).

MDCTC and colonoscopy show equal overall sensitivity for the detection of polypoid lesions 6 mm or larger in size, but more patients are inadequately examined when MDCTC is used.