CT colonography. Technique, diagnostic accuracy and patient acceptance

Sammanfattning: The aim was to evaluate diagnostic potential, technical performance, and patient acceptance of CT colonography when used for detection of colorectal polyps and cancer in symptomatic patients with a high risk of colorectal neoplasia.Methods: After bowel preparation, 111 symptomatic patients underwent CT colonography followed by conventional colonoscopy. Image review for colorectal lesions was performed, and three-dimensional endoluminal images (virtual colonoscopy) were used as a complement to axial images. Colonic CT findings were compared with colonoscopy results and with other available data. Patient acceptance was evaluated by questionnaires addressing the CT colonography and the colonoscopy, respectively, and the preference for either examination was evaluated by comparative questions after completion of both examinations. Air distension, degree of fluid redistribution with change in body position, influence of residual stool on bowel wall assessability, and quality of overall colon visualisation were evaluated using scales. Image review for extra-intestinal findings was performed, and findings were classified according to their potential clinical importance. A follow-up review of patient records was also performed. Results: At CT colonography, 72 (67%) of 108 lesions were identified. The sensitivity increased with lesion size and was 91% for lesions ³10 mm. All 10 carcinomas were detected as were 91% of all patients with carcinoma or adenoma ³5 mm. Unconfirmed or false positive CT findings were generally small and/or reported with low diagnostic certainty. The CT colonography was generally well accepted by the patients, and a majority (82%) preferred CT colonography to colonoscopy. A larger proportion of patients rated pain higher during colonoscopy than during CT colonography (95% CI, 30% to 56%), and discomfort from air-filling of the colon was the major complaint about CT colonography. Complete overall visualisation of the colon wall was found in 27% of patients, while 47% had subtotal visualisation restricted to a limited part. Examination in both the supine and prone positions was necessary for an adequate visualisation and largely compensated for residual fluid and suboptimal air-filling. Previously unknown extra-colonic pathology of major clinical importance was identified in 14 (13%) of the patients, and 26 (23%) patients had findings of major clinical importance. Conclusions: The sensitivity for detection of colorectal lesions of 10 mm or larger size was 91%. CT colonography was considered less painful and less difficult overall than colonoscopy, and was the preferred examination. The colon wall was completely, or almost completely, visualised in 75% of the patients, and examination in the supine and prone positions was necessary for complete visualisation. Extra-colonic findings were common.