Benefits from a computer-aided detection device in colonoscopy (ACCENDO-COLO) – an interim analysis of an italian multicenter randomized clinical trial

Aims Adenoma detection rate (ADR) has been inversely associated with the risk of interval colorectal cancer. The aim of the study is to evaluate the benefit of a computer-aided detection device (CADe Endo-AID) in a consecutive, non-selected series of patients undergoing colonoscopy.

Methods This is an interim analysis of a prospective multicenter randomized control trial (RCT) that included patients aged 40-85 scheduled for screening, surveillance or diagnostic colonoscopy randomly assigned to CADe (Endo-AID, Olympus, Japan) or standard colonoscopy (SC). Patients with a Boston Bowel Preparation Scale (BBPS) < 2 in any segment were excluded from the analysis. Polyp Detection Rate (PDR), Polyp Per Colonoscopy (PPC), ADR, Adenoma Per Colonoscopy (APC), Serrated Detection Rate (SDR), Advanced adenoma, and the presence of ≥ 3 precancerous lesions were evaluated in the two groups (Table 1)

Results 901 patients were enrolled of whom 60 were excluded for inadequate bowel cleansing. Therefore, 839 subjects (428 CADe vs 411 SC) were included in the final analysis. ADR was significantly higher in CADe than in the control group [50.2% (215/428) vs 41.6% (171/411) p= 0.01]. CADe also significantly increased PDR, PPC and APC [67.5% (289/428) vs 55.5% (228/411) p< 0.01; 1.68 vs 1.34 p< 0.01; 1.17 vs 0.86 p< 0.01, respectively]. SDR, advanced adenomas, and ≥ 3 precancerous lesions were higher in the intervention group without reaching statistical significance.

Conclusions The preliminary results of the present RCT show an increase in the diagnostic yield of AI-assisted colonoscopy in all the calculated indicators. ([Table 1]).

Table 1  Not statistically significant measurements among the Full Study Population. SDR: Serrated Detection Rate, ADV: Advanced, CADe Computer-Aided Detection