EFFECTIVENESS OF THE ALGORITHM OF DEPTH DIAGNOSIS FOR SUPERFICIAL BARRETT'S ADENOCARCINOMA

S Yoshinaga; I Oda; K Furutani; D Hihara; M Koga; T Ito; H Cho; M Yamada; S Abe; S Nonaka; H Suzuki; Y Saito; H Daiko; H Katai; H Taniguchi; S Sekine

doi:10.1055/s-0039-1681719

Endoscopy, Table of Contents

Endoscopy 2019; 51(04): S185-S186
DOI: 10.1055/s-0039-1681719

ESGE Days 2019 ePoster podium presentations

Saturday, April 6, 2019 13:30 – 14:00: ESD 2 ePoster Podium 4

Georg Thieme Verlag KG Stuttgart · New York

EFFECTIVENESS OF THE ALGORITHM OF DEPTH DIAGNOSIS FOR SUPERFICIAL BARRETT'S ADENOCARCINOMA

Authors

S Yoshinaga

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
I Oda

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
K Furutani

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan

²Department of Gastroenterology, Tokyo Teishin Hospital, Tokyo, Japan
D Hihara

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
M Koga

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
T Ito

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
H Cho

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
M Yamada

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
S Abe

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
S Nonaka

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
H Suzuki

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
Y Saito

¹Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan
H Daiko

³Esophageal Surgery Division, National Cancer Center Hospital, Tokyo, Japan
H Katai

⁴Gastric Surgery Division, National Cancer Center Hospital, Tokyo, Japan
H Taniguchi

⁵Pathology Division, National Cancer Center Hospital, Tokyo, Japan
S Sekine

⁵Pathology Division, National Cancer Center Hospital, Tokyo, Japan

Abstract

Full Text

Aims:

Although the depth of invasion is the important factor to make a decision in treatment, little is known about the correlation between endoscopic feature and invasion depth of superficial Barrett's adenocarcinoma (SBA). We aimed to investigate the endoscopic findings indicating deep submucosal invasion of SBA and develop the algorithm of depth diagnosis.

Methods:

First, we investigated endoscopic findings of indicating deep submucosal invasion (SM2; more than 500 µm). This derivation study included 59 SBAs in 57 patients confined to mucosa or submucosa. five expert endoscopists, who were blinded to histology independently, reviewed the endoscopic images. According to previous studies, they selected macroscopic type (0-Ip, 0-Is, mix (0-IIa+IIc or 0-IIc+IIa) or others), estimated lesion size (1˜10 mm, 11˜20 mm, 21˜30 mm or 31 mm˜), and determine present or absent of remarkable redness, uneven surface, margin elevation, ulceration, and enlarged folds. If more than 3 of 5 endoscopists pointed out, such finding was defined as the “positive” finding. Then, we evaluated the relationship between “positive” findings and SM2, and assumed the algorithm to estimate the depth using those “positive” findings. For validation, five novice endoscopists, who were blinded to histology independently as well, reviewed the endoscopic images, and we evaluated their individual diagnosis of SM2, compared with their diagnosis using the assumed algorithm.

Results:

After the derivation study, estimated tumor size (more than 11 mm), uneven surface and margin elevation had significant relationships with SM2 after multivariate analysis. In this derivation results, we assumed the algorithm using these 3 findings. Using this assumed algorithm, specificities and accuracies of SM2 of 3 novice endoscopists were improved compared with their individual diagnosis significantly, and positive predictive values (PPV) of 4 novice endoscopists were increased.

Conclusions:

Using this algorithm, the specificity and accuracy of novice endoscopists in depth estimation of SBAs increased. PPV likely increased as well.