Computed tomography colonography performs poorly in detection of sessile serrated lesions

Computed tomography colonography (CTC) is an alternative to colonoscopy for the detection of polyps and colorectal cancer (CRC). One-third of CRCs arise via the sessile serrated pathway. Evidence supports using CTC to detect adenomas and CRC; however, its accuracy for sessile serrated lesions (SSLs) remains uncertain. This study aimed to determine the accuracy of CTC in detecting SSLs compared with colonoscopy.

Electronic records identified all colonoscopy procedures where a histologically validated SSL was excised over a 11-month period. In those patients who had a CTC within 1 year prior to colonoscopy, the presence, size and location of SSLs were compared to determine the accuracy of CTC in SSL identification.

A total of 4,346 procedures were performed (2,548 people, 2,082 [47.9%] male, mean age 59.6). A total of 2,204 SSLs were removed, representing 24% of all polypectomies. SSLs were predominantly located in the right colon (65.1%) and were typically (85%) <10mm in size. A total of 110 SSLs were obtained from 39 procedures with a prior CTC. Of these procedures, 12 (30.8%) had lesions identified on CTC; however, CTC only accurately identified 14.5% of the total SSLs. Five of 16 (32%) SSLs ≥10mm were correctly identified compared with 11 of 94 (11%) SSLs 1-9mm, (odds ratio 3.42, p=0.0495).

This study demonstrated that CTC has poor efficacy in detecting SSLs, irrespective of polyp size and location. Based on these findings, CTC as a substitute for colonoscopy is not advisable in patients at risk of SSLs.