University of Texas Health Science Center at Houston Houston, TX, United States
Faisal Ali, MD1, Nirav Thosani, MD, MHA2 1University of Texas Health Science Center at Houston, Chicago, IL; 2University of Texas Health Science Center, Houston, TX
Introduction: I-Scan image post-processing technology may improve the detection of adenomatous polyps. The incremental yield of I-Scan compared to standard high-definition white light endoscopy (WLE) has not been summarized thoroughly.
Methods: A systematic search of Medline and Embase from inception-May 2021 was conducted to identify prospective studies and randomized-trials comparing I-Scan to WLE for detection of adenomas < 10mm in size. The primary measure of incremental yield of adenoma detection was extrapolated from the difference in proportion of adenomatous polyps/patients detected by I-Scan versus WLE with their corresponding 95% confidence intervals. Each polyp was regarded as a patient for the purpose for our analyses. Diagnostic test performance was analyzed and used to compute number needed to detect as a secondary measure of incremental yield of I-Scan.
Results: Screening of 1508 articles yielded 10 studies (2505 patients/polyps in I-Scan, 2091 patients/polyps in WLE arm). The pooled adenoma detection rate was 54% (31-78%) with I-Scan and 42% (24-59%) with WLE use. I-Scan led to a 10% increase in yield of adenoma detection (risk difference (RD): 0.10; 0.05-0.16; I2 72.63%), with 11% increase in subgroup analysis of prospective studies (2-20%; I2 63.61%) and 10% in randomized-trials (2-17%; I2 71.01%). Upon stratifying studies by year of publication, reports published ≥2015 showed a 13% increase in yield of adenoma detection with I-Scan use, though with a wide confidence interval (RD 0.13; -0.01-0.26; I2 72.99%), compared to 9% in reports published < 2015 (RD 0.09; 0.03-0.15; I2 72.06%). Influence analysis revealed that none of the studies had an overdue impact on the final analysis. Subgroup analysis of diminutive polyps (< 5mm) showed a 7% increase in yield of adenoma detection with I-Scan use (RD 0.07; 0.01-0.13; I2 76.88%).
The pooled sensitivity and specificity of I-Scan in characterizing polyps was 87% (71-95%) and 86% (80-90%) and that of WLE was 77% (58-90%) and 79% (56-91%), respectively. The number needed to detect one additional adenomatous polyp with I-Scan use was 1.4.
Discussion: I-Scan increases the yield of adenoma detection by 10% (7% in diminutive polyps) compared to WLE based on a meta-analysis of prospective studies and randomized-trials, with a low number needed to detect one additional adenomatous polyp, albeit with notable heterogeneity in the pooled data. Our findings may be used to advocate for the use of I-SCAN to enhance detection of adenomas < 10 mm in size.
Figure: Forrest plots of difference in detection of adenomatous polyps <10 mm in size and diminutive (<5 mm) adenomatous polyps.
Faisal Ali indicated no relevant financial relationships.
Nirav Thosani: Boston scientific – Consultant. Pentax America – Consultant.
Faisal Ali, MD1, Nirav Thosani, MD, MHA2. P1303 - Incremental Yield of Adenoma Detection with High-Definition I-Scan Colonoscopy Compared to White Light Endoscopy – A Meta-Analysis of Prospective Studies and Randomized Controlled Trials, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.