A randomized crossover study comparing light-induced fluorescence endoscopy with standard videoendoscopy for the detection of early neoplasia in Barrett's esophagus.

Abstract

BACKGROUND: Light-induced fluorescence endoscopy (LIFE) may improve the detection of high-grade dysplasia (HGD) and early stage cancer (EC) in Barrett's esophagus (BE). The aim of this study was to compare LIFE with standard endoscopy (SE) in a randomized crossover study. METHODS: Fifty patients with BE underwent SE and LIFE in a randomized sequence (4 to 6-week interval between procedures). The two procedures were performed by two different endoscopists who were blinded to the findings of the other examination. Targeted biopsy specimens were taken from detected lesions, followed by random biopsy specimens with a 2-cm interval, 4-quadrant protocol. Biopsy specimens were routinely evaluated and subsequently reviewed by a single, blinded expert GI pathologist. RESULTS: Targeted biopsy specimens had a sensitivity for the diagnosis of HGD/EC of 62% (8/13) for both techniques. The overall sensitivity (all biopsy specimens) was 85% for SE and 69% for LIFE (p = 0.69). All targeted biopsy specimens had a positive predictive value (PPV) for HGD/EC of 41% for SE and 28% for LIFE (p = 0.40); autofluorescence-targeted biopsy specimens had a PPV of 13%. False-positive lesions had a significantly higher rate of acute inflammation than random biopsy specimens. CONCLUSIONS: In this study, LIFE did not improve the detection of HGD or EC in patients with BE compared with SE.