Edgar Corona, MPH1, Roshan Bastani, PhD2, Liu Yang, MD, MPH3, Paul Shao, MD, MS1, Didi Mwengela, MD1, Michelle Didero, BS1, Ishan Asokan, MD, MSc1, Alex Bui, PhD3, William Hsu, PhD3, Folasade May, MD, PhD3
1David Geffen School of Medicine, University of California Los Angeles, Los Angeles, CA; 2UCLA Center for Cancer Prevention and Control Research, Fielding School of Public Health and Jonsson Comprehensive Cancer Center, Los Angeles, CA; 3University of California Los Angeles, Los Angeles, CA
Introduction: National guidelines recommend that patients with colorectal adenomas ≥1cm undergo surveillance colonoscopy to prevent colorectal cancer. Polyp size is often assessed twice, once during colonoscopy and again during pathology examination. It remains unclear whether these measurements vary and which should govern surveillance intervals.
Methods: We retrospectively analyzed a random sample of primary care patients who underwent polypectomy between January 1, 2013 and January 1, 2016 in a large academic healthcare center. Chart abstractors reviewed electronic health record data for colonoscopic polyp descriptors and measurements, corresponding pathology measurements, and histology. We excluded cases with normal histology, missing pathology, fragments, unreliable polyp matches between reports, and >5 polypectomies. We performed a paired t-test to assess polyp size discordance between reports and calculated Cohen’s kappa coefficient to assess agreement whether the size was ≥1cm.
Results: We reviewed 1,528 patient charts, and 831 (54.4%) met inclusion criteria, representing 1,532 polyps. Sizes ranged from 0.1-3.0cm on colonoscopy and 0.1-2.7cm on pathology reports. On colonoscopy report, the majority (881, 57.5%) used only a descriptive term rather than a numeric value to describe size (Table). Descriptive size terms were diminutive (513), tiny (39), small (328), and large (1). Only 2.8% of polyps described as diminutive, small, or tiny on the colonoscopy report were ≥1cm on the corresponding pathology report. Among the 367 (24.0%) polyps with discrete numeric sizes on both reports, size discordance varied by >0.2cm for 50% of polyps (Figure 1). More of these polyps were ≥1cm on colonoscopy (70) than on pathology (51) report, and there was remarkable clustering at 1cm by colonoscopy report estimation (Figure 2). There was moderate agreement (kappa=0.52) on size ≥1cm. Among tubular adenomas (TA) ≥1cm on colonoscopy or pathology report (36), 10 (27.8%) were ≥1cm on both, 18 (50.0%) were ≥1cm on colonoscopy only, and 8 (22.2%) were ≥1cm on pathology only.
Discussion: There is considerable discordance between colonoscopic and pathology polyp size measurements, with more polyps sized as ≥1cm by colonoscopic estimation compared to pathology, which has implications for surveillance frequency. Future studies should investigate the implications of discordant measurements on clinical outcomes and inform guidelines on surveillance intervals.
Citation: Edgar Corona, MPH; Roshan Bastani, PhD; Liu Yang, MD, MPH; Paul Shao, MD, MS; Didi Mwengela, MD; Michelle Didero, BS; Ishan Asokan, MD, MSc; Alex Bui, PhD; William Hsu, PhD; Folasade May, MD, PhD. P1277 - DISCORDANCE IN COLORECTAL POLYP SIZE DETERMINATION BETWEEN COLONOSCOPY AND PATHOLOGY REPORTING. Program No. P1277. ACG 2019 Annual Scientific Meeting Abstracts. San Antonio, Texas: American College of Gastroenterology.