Back-to-back comparison of auto-fluorescence imaging (AFI) versus high resolution white light colonoscopy for adenoma detection

Abstract

Background: Some patients under close colonoscopic surveillance still develop colorectal cancer, thus suggesting the overlook of colorectal adenoma by endoscopists. AFI detects colorectal adenoma as a clear magenta, therefore the efficacy of AFI is expected to improve the detection ability of colorectal adenoma. The aim of this study is to determine the efficacy of AFI in detecting colorectal adenoma.

Methods: This study enrolled 88 patients who underwent colonoscopy at Asahikawa Medical University and Kushiro Medical Association Hospital. A randomly selected colonoscopist first observed the sigmoid colon and rectum with conventional high resolution endosopy (HRE). Then the colonoscopist changed the mode to AFI and handed to the scope to another colonoscopist who knew no information about the HRE. Then the second colonoscopist observed the sigmoid colon and rectum. Each colonoscopist separately recorded the findings. The detection rate, miss rate and procedural time were assessed in prospective manner.

Results: The detection rate of flat and depressed adenoma, but not elevated adenoma, by AFI is significantly higher than that by HRE. In less-experienced endoscopists, AFI dramatically increased the detection rate (30.3%) and reduced miss rate (0%) of colorectal adenoma in comparison to those of HRE (7.7%, 50.0%), but not for experienced endoscopists. The procedural time of HRE was significantly shorter than that of AFI.

Conclusions: AFI increased the detection rate and reduced the miss rate of flat and depressed adenomas. These advantages of AFI were limited to less-experienced endoscopists because experienced endoscopists exhibited a substantially high detection rate for colorectal adenoma with HRE.

Figure 1

A flow diagram of this study.

Figure 2

Cases of flat and depressed adenoma and normal mucosa. A faint red area in the sigmoid colon was noted by HRE (A). AFI revealed a strong magenta area at the same site (B). In the normal mucosa, HRE showed no abnormal findings (C) and AFI revealed no magenta area (D).

Figure 3

Procedural time of HRE and AFI. The procedural time of HRE (144.5 ± 8.8 seconds) was significantly shorter than that of AFI in all participants (267.0 ± 20.8 seconds; p < 0.05) (A). An increase in the procedural time for the AFI examination was revealed in either experienced endoscopists (B) or less-experienced endoscopists (C).