Autofluorescence Imaging Videoendoscope System for the Diagnosis of Suquamous Cell Carcinoma of Esophagus and Pharynx
Haruhisa Suzuki,Yutaka Saito,Ichiro Oda,Takahisa Matsuda,Hisatomo Ikehara,Yumi Mashimo,Chizu Yokoi,Kohei Takizawa,Takuji Gotoda,Daizo Saito
DOI: https://doi.org/10.1016/j.gie.2007.03.846
IF: 10.396
2007-01-01
Gastrointestinal Endoscopy
Abstract:Background: Autofluorescence imaging (AFI; Olympus Optical Co, Ltd, Tokyo, Japan) videoendoscope system produces real-time, pseudo-color images that combine green and red fluorescence. It can distinguish neoplastic tissue from non-neoplastic one based on differences in the intensity of autofluorescence. The benefit of this system for the diagnosis of bronchial squamous cell carcinoma (SCC) has been confirmed, therefore, AFI system may be useful for the diagnosis of SCC of esophagus and pharynx. Aim: To evaluate the usefulness of AFI system for the diagnosis of SCC of esophagus and pharynx compared with white light endoscopy (WLE). Methods: In a prospective study, we enrolled 32 patients with superficial esophageal carcinoma and 7 patients with superficial pharyngeal carcinoma diagnosed previously in other hospitals. We observed the esophagus and pharynx with WLE followed by AFI and both procedures were documented with endoscopic images. For the lesions diagnosed as SCC histologically, three experienced endoscopists evaluated the quality of visualization of AFI and WLE images. The visualization was evaluated on three-tier scale: visible, iilegible and invisible. Results: In total, 43 lesions with superficial esophageal carcinoma were diagnosed (estimated depth of invasion; m:sm = 28:15, macroscopic type; flat:elevated = 36:7, circumference of the lumen; less than half:more than half = 21:22). There were 28, 11 and 4 lesions considered as visible, illegible and invisible with WLE respectively, compared to 34, 5 and 4 lesions with AFI. Therefore, esophageal carcinoma was significantly better visualized by AFI compared with WLE (proportion of visible lesions were 79% with AFI and 65% with WLE; p < 0.0001). In addition, AFI could reveal the 6 lesions that were not well visualized by WLE and these lesions had a tendency to affect less than half circumference of the lumen and to be flat. On the other hand, 8 lesions with superficial pharyngeal carcinoma were diagnosed (estimated depth of invasion; epithelial:sub-epithelial invasion = 4:4). With WLE images, there were 3, 3 and 2 lesions considered as visible, illegible and invisible respectively, in contrast 6, 2 and 0 lesions with AFI. Therefore, pharyngeal carcinoma had a tendency to be better visualized by AFI compared with WLE (proportion of visible lesions were 75% with AFI and 38% with WLE; not significant). Conclusions: AFI system should be used as an endoscopic screening method in a population at high risk for SCC of esophagus and pharynx because these lesions can be better visualized by AFI compared with conventional WLE.