Tu1088 SKILL UP STUDY OF OBSERVATION TECHNIQUE USING NARROW BAND IMAGING FOR HEAD AND NECK REGION IN PATIENTS WITH ESOPHAGEAL SQUAMOUS CELL CARCINOMA – A PROSPECTIVE MULTICENTER STUDY

Patients with esophageal squamous cell carcinoma (ESCC) are at a high risk of developing synchronous and metachronous second primary head and neck squamous cell carcinoma (HNSCC). Narrow-band imaging (NBI) is effective in detecting superficial HNSCC. Peroral endoscopic resection of superficial HNSCC is a feasible and effective treatment with curative intent. However, skilled endoscopic technique is necessary to survey the head and neck region to detect the superficial cancer. To evaluate the usefulness of the training program on endoscopic head and neck surveillance for beginner endoscopists. This study was designed as a prospective multicenter study of 13 beginner endoscopists from 13 institutions. All endoscopists underwent training regarding systematic observation technique in the head and neck region, the diagnostic criteria of head and neck cancers, and hands-on lectures. Between May 2016 and February 2017, patients with current or previous diagnosis of ESCC were enrolled. The patients underwent head and neck surveillance with NBI, and the following variables were comparatively analyzed before (Group A) and after (Group B) the training program: detection rate of histologically-confirmed HNSCC, quality of endoscopic images, and head and neck examination times. The primary endpoint was the detection rate of histologically-confirmed HNSCC. Fifteen points in the head and neck region were determined to be captured. Each filed image was evaluated and scored by an independent committee comprising three endoscopists as follows: 2 points: the region was evaluated clearly, 1 point: the region was not evaluated clearly, 0 point: the region is not recorded. The maximum possible score was 30 points. A total of 330 patients (group A: 181, group B: 149) were enrolled. There were 294 males (group A: 164, Group B: 130), with a median age of 72 years. The proportion of patients who had already undergone treatment for ESCC was significantly higher in group A compared with B (group A: 88%, and group B: 79% (p=0.025)). Three patients with HNSCC were each detected in groups A (1.7%) and B (2.0%) (p=1.000). The mean (± standard deviation) examination times were 157 ± 71 seconds in group A, and 174 ± 109 seconds in group B (p=0.073). The mean (± standard deviation) scores of the endoscopic images were 25.04 ± 5.47 points in group A, and 27.01 ± 4.35 points in group B (p<0.001). The detection rate of HNSCC using NBI for patients with ESCC did not improve after the training program. However, the quality of endoscopic images was significantly improved after the training program during examination by beginner endoscopists.