Diagnostic Performance of Primary 3-Dimensional Computed Tomography Colonography in the Setting of Colonic Diverticular Disease

Background & Aims: Colonic diverticular disease (CDD) is a leading cause of nondiagnostic segmental evaluation at computed tomography colonography (CTC). The primary goal of this study was to evaluate the impact of CDD on polyp detection at primary 3-dimensional (3D) CTC. Methods: The study group consisted of 280 asymptomatic average-risk adults (mean age, 58.1 y; 167 men, 113 women). All patients underwent CTC followed by same-day optical colonoscopy with segmental unblinding (reference standard). Primary 3D endoluminal evaluation with 2-dimensional correlation was used for initial polyp detection at CTC. Without knowledge of polyp findings, all colonic segments were reviewed for the presence of CDD and graded as absent, minimal, moderate, or extensive disease. Results: Moderate or extensive CDD was present in 271 (12.1%) of 2240 colonic segments and 142 (50.7%) of 280 patients. CTC performance for polyps 6 mm or larger in the presence and absence of moderate-extensive CDD was as follows: by-segment sensitivity of 90.3% (28/31) and 79.4% (123/155); by-patient sensitivity of 86.5% (64/74) and 83.1% (54/65); by-segment specificity of 97.1% (233/240) and 97.7% (1772/1814); by-patient specificity of 83.8% (57/68) and 83.6% (61/73); by-segment positive predictive value of 80.0% (28/35) and 74.5% (123/165); by-patient positive predictive value of 85.3% (64/75) and 81.8% (64/66); by-segment negative predictive value of 98.7% (233/236) and 98.2% (1172/1804); and by-patient negative predictive value of 85.1% (57/67) and 84.7% (61/72), respectively (not significant, P ≥ .15). Conclusions: CDD was common in this asymptomatic screening population, but its presence did not degrade the diagnostic performance of primary 3D CTC for polyp detection.