Bryan Dustin Pooler MD, Presenter: Nothing to Disclose

David H. Kim MD, Abstract Co-Author: Consultant, Viatronix, Inc Co-founder, VirtuoCTC, LLC Medical Advisory Board, Digital ArtForms, Inc

Perry J. Pickhardt MD, Abstract Co-Author: Co-founder, VirtuoCTC, LLC Stockholder, Cellectar Biosciences, Inc

The diagnostic performance of optical colonoscopy (OC) for colorectal polyp detection has been estimated in previous CT colonography (CTC) trials using segmental unblinding of CTC findings. However, these estimates do not account for lesions missed by OC after unblinding, which have been unavoidably labeled as CTC false positives. Our purpose was to determine how many discordant lesions in our clinical practice actually prove to be OC false negatives on subsequent examination.

During a 113 month period, 9,336 patients (mean age 57.1±8.0 years, M:F 4,210:5,126) underwent CTC at a single center, yielding 2,606 non-diminutive polyps. Of 1,731/2,606 polyps that underwent follow-up OC, 1,550 (90%) were concordant and 181 (10%) were discordant. CTC results (size, location, morphology) were revealed to colonoscopists prior to OC. After independent consensus review by at least two radiologists, 115 discordant findings were felt to be possible OC false negatives, and were further evaluated at repeat CTC and/or OC.

Of the 115 possible OC false negatives, 37 were either lost to follow-up or still awaiting follow-up at the time of study. Of the remaining 78 polyps 31 (40%) were confirmed to be OC false negatives at follow-up evaluation (26 by OC, 5 by CTC), and 47 (60%) were again not found, and remain CTC false positives. Compared with CTC false positives, OC false negatives were more likely to be larger (10.6 ± 5.3 mm vs 8.5 ± 3.3 mm, p=0.034) and to have higher diagnostic reader confidence at initial CTC (mean 2.8/3 vs 2.3/3, p=0.001). OC false negatives were more likely than OC/CTC concordant polyps to be located in the right colon (71% vs 47%, p=0.010). Of OC false negatives confirmed at subsequent OC, 17/26 (65%) had adenomatous histology (1 tubulovillous adenoma, 11 tubular adenoma, 5 serrated adenoma), of which 6 were advanced lesions.

Among discordant polyps at OC following positive CTC, OC false negatives are a common occurrence even when CTC findings are known prior to colonoscopy. Proven OC false negatives were ≥10 mm on average, more likely to be located in the right colon, and called with higher diagnostic confidence on CTC. Most ultimately resected OC false negatives proved to be adenomatous histology, including a substantial fraction of advanced lesions.

An understanding of missed polyps at colon cancer screening is vital to improving detection and patient care.

Pooler, B, Kim, D, Pickhardt, P, Missed Colorectal Polyps at Optical Colonoscopy Despite Prospectively Known Positive CT Colonography Findings.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14011925.html