Accuracy of CT Colonography for Detection of Large Adenomas and Cancers

• Published in: Obstetrical & gynecological survey Vol. 64; no. 1; pp. 35 - 37
• Main Authors: Johnson, C Daniel, Chen, Mei-Hsiu, Toledano, Alicia Y, Heiken, Jay P, Dachman, Abraham, Kuo, Mark D, Menias, Christine O, Siewert, Betina, Cheema, Jugesh I, Obregon, Richard G, Fidler, Jeff L, Zimmerman, Peter, Horton, Karen M, Coakley, Kevin, Iyer, Revathy B, Hara, Amy K, Halvorsen, Robert A, Casola, Giovanna, Yee, Judy, Herman, Benjamin A, Burgart, Lawrence J, Limburg, Paul J
• Format: Journal Article
• Language: English
• Published: Lippincott Williams & Wilkins, Inc 01-01-2009
• ISSN: 0029-7828; 1533-9866
• DOI: 10.1097/01.ogx.0000340778.07423.1f

Although recognition and curative treatment of both precursor adenomas and localized cancers are possible through early-detection, and regular screening reduces mortality from colorectal cancer, screening is largely underused. Reasons for underuse include performance drawbacks and issues of comfort, availability, and expense. Computed tomographic (CT) colonography is an alternative that provides minimally invasive, structural evaluation of the entire colorectum, and permits rapid imaging of the entire colorectum without sedation, and with few procedure-related complications. Its effectiveness in detecting asymptomatic colorectal lesions in previous studies is controversial, in large part because of differences in patient populations, imaging techniques, and radiologists’ qualifications.This large prospective trial compared the accuracy of CT colonography in detecting histologically confirmed, large colorectal adenomas and cancers (≥10 mm in diameter), with findings on screening colonoscopy and histologic review, the reference standard.A total of 2600 asymptomatic patients 50 years of age or older scheduled to undergo routine colonoscopy were recruited at 15 clinical sites between 2005 and 2006. Lesions measuring 5 mm or more in diameter were identified by radiologists trained in CT colonography. The preparation for CT colonography included stool and liquid tagging, laxative purgation, and colonic insufflation. Examinations were performed with multidetector-row CT scanners with a minimum of 16 rows.Complete data were available for 2512 participants examined with CT colonography and colonoscopy on the same day. For adenomas and cancers ≥10 mm, the mean (±SE) per-patient estimates of several measured parameters were as followssensitivity (0.90 ± .06); specificity (0.86 ± .05), positive (0.23 ± .04) and negative (0.99 ± <.01) predictive values, and area under the receiver-operating-characteristic curve for CT colonography (0.89 ± .04). The estimated sensitivity of 0.90 for identifying patients with adenomas or cancers ≥10 mm indicates failure of CT colonography to identify large lesion(s) detected by colonoscopy in 10% of patients. The per-patient sensitivity for detecting lesions 6 to 9 mm in diameter was 0.78 ± .07. The per-polyp sensitivity for detecting lesions ≥10 mm was 0.84 ± .04.These data show that CT colonographic screening identified 90% of asymptomatic adults with adenomas and cancers ≥10 mm in diameter that were detected by screening colonoscopy. For smaller lesions (6 to 9 mm), the sensitivity was lower (78%).