Reduction of miss rates of colonic adenomas by zoom chromoendoscopy

Abstract

The aim of this study was to determine the detection rate of polyps using zoom chromoendoscopy (ZE) compared with standard video colonoscopy. End-to-end colonoscopies were performed in 50 patients by two different endoscopists blinded for each other's results. Lesions detected during initial standard colonoscopy (C1) were biopsied or removed by snare resection. The second colonoscopy (C2) was done with a zoom colonoscope spraying the whole colon with indigocarmine (0.4%). In addition, detected mucosal lesions were documented prior to ZE and then classified according to the pit pattern classification before biopsy or removal. The retrieval time for each procedure was determined. The average retrieval time for C1 was 13+/-9 min (9-24) and 28+/-11 min (16-38, p<0.05) for ZE. During C1, 56 lesions were detected in 26 of 50 patients (34 hyperplastic and 22 adenomatous). During C2, 19 additional polyps were documented prior to ZE (15% tandem miss rate), and 20 further lesions were detected with ZE (21% additional polyp detection rate compared to C1 and C2 without ZE). Of the 39 additional lesions removed during C2 after ZE, 29 were hyperplastic and 10 were adenomatous. Most adenomas detected during the second investigation were found in patients in whom adenomatous polyps had already been removed during the initial colonoscopy (9 of 26 patients vs 1 of 24 patients, p<0.02). No carcinoma was detected. The pit pattern classification allowed a correct differentiation between hyperplastic and adenomatous polyps (accuracy 93%, sensitivity 90%, specificity 97%). Using zoom chromoendoscopy, the rate of detecting colonic polyps can be increased at the cost of a longer retrieval time.