Screening for Colorectal Neoplasia with Virtual Colonoscopy: Results of a Prospective Multicenter Trial in 1233 Asymptomatic Adults

PURPOSE: The performance characteristics of CT virtual colonoscopy (VC) for detection of colorectal neoplasia have not been established in an average-risk screening population. We conducted a prospective multicenter VC screening trial using optical colonoscopy (OC) as the reference standard. METHODS: 1233 asymptomatic adults (mean age, 57.8 years; 728 men, 505 women) underwent VC examination followed by same-day OC. 1201 patients were classified as average risk. Prospective VC interpretation immediately preceded OC, with segmental unblinding of VC results at OC. Radiologists relied primarily on the three-dimensional (3D) endoluminal display for polyp detection. RESULTS: VC sensitivity by patient for adenomatous polyps was 93.8% (45/48), 93.9% (77/82), and 88.7% (149/168) for size thresholds of 10, 8, and 6 mm, respectively. At the same size thresholds, OC sensitivity by patient for adenomatous polyps (prior to unblinding) was 87.5%, 91.5%, and 92.3%, respectively. VC specificity by patient for adenomatous polyps was 96.0% (1138/1185), 92.2% (1061/1151), and 79.6% (848/1065), respectively, and for all polyps (regardless of histology) was 97.4% (1131/1161), 95.0% (1050/1105), and 84.5% (826/978), respectively. VC sensitivity by polyp for adenomas was 92.2% (47/51), 92.6% (88/95), and 85.7% (180/210) at 10, 8, and 6 mm thresholds, respectively. Two polyps were malignant; both were detected on VC and one was missed on OC before VC unblinding. CONCLUSIONS: CT virtual colonoscopy with a 3D emphasis is an accurate screening method for detection of colorectal neoplasia in asymptomatic average-risk adults, despite the low prevalence of disease. This minimally invasive total colonic examination compares favorably with OC for detection of clinically relevant lesions. Introduction Colorectal cancer is currently the second leading cause of cancer-related mortality in the United States, accounting for nearly 60,000 deaths each year [1]. The vast majority of colorectal cancers are believed to arise within benign adenomatous polyps that develop slowly over many years [2]. Evidence-based guidelines recommend screening of adults at average risk [3], since detection and removal of adenomas has been shown to reduce both cancer incidence and mortality [4-6]. Unfortunately, about half of the intended U.S. population has not been screened by any means for this largely preventable disease [7]. Virtual colonoscopy (VC), also referred to as “CT colonography”, could provide an attractive alternative for widespread screening since it requires no IV sedation, analgesia, or recovery time [8,9]. Although the performance characteristics of VC have been encouraging in polyp-rich populations [10-11], there is insufficient data among average-risk adults, for whom the prevalence of significant adenomas is much lower [9,12,13]. We conducted a study to evaluate the performance characteristics of VC in an asymptomatic screening population, using optical colonoscopy (OC) as the reference standard.