Comparison of computed virtual chromoendoscopy and conventional chromoendoscopy with acetic acid for detection of neoplasia in Barrett's esophagus.

BACKGROUND AND STUDY AIMS: Computed virtual chromoendoscopy (CVC) is a new imaging technique that enhances mucosal surface contrast and highlights the vascular pattern without the need for dye-spraying as in conventional chromoendoscopy. The aim of this prospective randomized pilot study with a crossover design was to compare CVC with conventional chromoendoscopy with acetic acid (CAA) for the detection of high grade intraepithelial neoplasia (HGIN) or early cancer in patients with Barrett's esophagus.

PATIENTS AND METHODS: 57 patients with Barrett's esophagus (mean length 3.8 cm) and a history of HGIN/early cancer or suspected HGIN/early cancer were randomly allocated to undergo either CAA or CVC. All patients were re-examined with the alternative procedure at 4 - 6 weeks after the initial endoscopy. The two procedures were performed by five endoscopists, who were blinded to the findings of the other examination. At each examination, targeted biopsies were taken from all detected lesions, followed by random four-quadrant biopsies.

RESULTS: In 24/57 patients, 30 lesions with HGIN/early cancer were detected. The sensitivity of targeted biopsies for HGIN/early cancer on a 'per lesion' basis was 87 % (26/30) for both CAA and CVC. The positive predictive value was 39 % (26/66) for CAA and 37 % (26/70) for CVC. In the 'per patient' analysis, sensitivity was 83 % (20/24) and 92 % (22/24) for CAA and CVC, respectively ( P = 0.617). Stepwise random four-quadrant biopsies identified only one patient with HGIN/early cancer that was missed by both, CAA and CVC.

CONCLUSIONS: Computed virtual chromoendoscopy is a helpful adjunct for surveillance of Barrett's esophagus and appears to be as accurate as conventional chromoendoscopy in the detection of HGIN/early cancer.