Endoscopy 2020; 52(S 01): S51
DOI: 10.1055/s-0040-1704157
ESGE Days 2020 oral presentations
Thursday, April 23, 2020 14:30 – 16:00 Polyp forensics: Colon advanced Wicklow Meeting Room 3 Imaging 2
© Georg Thieme Verlag KG Stuttgart · New York

DIAGNOSTIC ACCURACY OF THE JNET CLASSIFICATION FOR ENDOSCOPIC DIAGNOSIS OF COLORECTAL LESIONS: A META-ANALYSIS

EJ Aguila
1   St. Luke’s Medical Center Global City, Institute of Digestive and Liver Diseases, Taguig City, Philippines
,
S Edding
1   St. Luke’s Medical Center Global City, Institute of Digestive and Liver Diseases, Taguig City, Philippines
,
CP Francisco
1   St. Luke’s Medical Center Global City, Institute of Digestive and Liver Diseases, Taguig City, Philippines
,
J Co
1   St. Luke’s Medical Center Global City, Institute of Digestive and Liver Diseases, Taguig City, Philippines
,
J Cervantes
1   St. Luke’s Medical Center Global City, Institute of Digestive and Liver Diseases, Taguig City, Philippines
› Author Affiliations
Further Information

Publication History

Publication Date:
23 April 2020 (online)

 

Aims Magnification using narrow band imaging (NBI) has been established as a reliable method for differentiating neoplastic from non-neoplastic colorectal lesions emphasizing surface and vessel pattern of these lesions. Several classifications have been proposed but recently, a newer classification, the Japan NBI Expert Team (JNET) has been developed.

Methods A comprehensive literature search from the PubMed Central, Embase, Cochrane Library, and Ovid was performed with the following search terms: JNET, NBI and colorectal lesions. Three studies were selected and validated using the QUADAS-2 model. Summary estimates; the pooled sensitivity, specificity, diagnostic odds ratio (DOR) and Areas under the Curve (AUC) of JNET for the diagnosis of colorectal lesions were computed using random-effects models using the R statistical software. The primary outcome of study was the diagnostic accuracy of the JNET classification in the endoscopic diagnosis of colorectal lesions in comparison to histopathologic findings.

Results Three trials comprising of 4,534 reviewed lesions were analyzed. The pooled sensitivity and specificity of the JNET classification in the endoscopic diagnosis of colorectal lesions were computed from the three trials. Type 1 was diagnosed with a pooled sensitivity of 82.74% (95% CI 73.91-89.03) and a pooled specificity of 99.43% (95% CI 96.21-99.91). JNET type 2A has a pooled sensitivity of 91.03% (95% CI 83.93-95.18) and a pooled specificity of 73.67% (95% CI 69.74-77.26). JNET type 2B has a pooled sensitivity of 58.08% (95% CI 43.59-71.29) and a pooled specificity of 90.22% (95% 82.32-94.81). JNET type 3 was diagnosed with a pooled sensitivity of 42.75% (95% CI 28.67-58.11) and a pooled specificity of 99.89% (95% CI 99.55-97).

Conclusions This study shows that as the colorectal lesion becomes endoscopically complex on NBI, the sensitivity of the JNET classification decreases while the specificity remains high. Endoscopists start to have varied endoscopic diagnosis for lesions that are JNET Type 2B and 3.