Diagnostic accuracy of CT colonography in patients with positive faecal occult blood test: results of the Italian project Legatumori 2003–2006

• Published in: Radiologia medica Vol. 114; no. 4; pp. 586 - 594
• Main Authors: Neri, E., Vagli, P., Turini, F., Cerri, F., Bardine, A., Cecchi, C., Naldini, G., Costa, F., Marchi, S., Bartolozzi, C.
• Format: Journal Article
• Language: English
• Published: Milan Springer Milan 01.06.2009
• Subjects: Abdominal Radiology/Radiologia Addominale; Aged; Aged, 80 and over; Colonography, Computed Tomographic; Colonoscopy; Colorectal Neoplasms - diagnosis; Colorectal Neoplasms - diagnostic imaging; Diagnosis, Differential; Diagnostic Radiology; Humans; Image Processing, Computer-Assisted; Imaging; Imaging, Three-Dimensional; Interventional Radiology; Italy; Mass Screening; Medicine; Medicine & Public Health; Middle Aged; Neuroradiology; Occult Blood; Pilot Projects; Predictive Value of Tests; Radiology; Retrospective Studies; Risk Factors; Sensitivity and Specificity; Time Factors; Ultrasound; Screening; Colonscopia virtuale; Cancro colorettale; CT colonography; Colorectal cancer
• ISSN: 0033-8362; 1826-6983
• DOI: 10.1007/s11547-009-0342-x

Purpose In the framework of the 3-year project of the Italian Legatumori (2003–2006), we evaluated the diagnostic accuracy of computed tomography (CT) colonography in detecting colorectal lesions in a screening population with positive faecal occult blood test (FOBT). Materials and methods Two hundred and thirty asymptomatic subjects (age range 45–80 years) were enrolled in the study. CT colonography was performed with standard patient preparation (no faecal tagging) and a 4-detector-row CT scanner. Image analysis was carried out with primary 2D analysis and the use of 3D endoluminal views to solve difficult cases. Patients were referred for conventional colonoscopy in the following situations: detection of three or more suspected lesions with maximum diameter ≤6 mm; evidence of one or more lesions with maximum diameter >6 mm; presence of colonic masses (maximum diameter >3 cm). Results CT colonography detected colonic masses in 12 out of 135 subjects (8%). It generated 93 false positives and 19 false negatives in the identification of diminutive lesions (≤6 mm), and 70 false positives and six false negatives in lesions >6 mm. Sensitivity was 83% in smaller lesions and 93% in lesions >6 mm; specificity was 45% and 59%, respectively. Conclusions In a screening population with positive FOBT, CT colonography without faecal tagging and no definite size threshold for the reporting of polyps showed very low specificity but high sensitivity in the detection of all colorectal lesions.