TARGETED VERSUS STANDARD NON-TARGETED BIOPSIES IN BARRETT’S ESOPHAGUS SURVEILLANCE

Aims Barrett’s Esophagus survaillance is performed with random 4-quadrant biopsy every-2cm to detect dysplasia complemented with targeted biopsies of suspicious lesions - Seattle Protocol. The main limitations are sampling error and costs associated with histology. The main aim was compare diagnostic yield, defined by dysplasia in the histology, of random biopsies with targeted biopsies.

Methods Cross-sectional-study, with consecutive inclusion of patients undergoing Barrett’s esophagus surveillance, between 2015-2018. High-resolution scopes were used and Barrett’s esophagus evaluation was performed under white-light endoscopy and in some cases complemented with Narrow-band-imaging. Random biopsies were performed in all patients. Additionally targeted biopsies were taken when focal lesions were suspected.

Results A total of 127 surveillance exams were included. High-definition- white-light endoscopy evaluation was complemented with Narrow-band-imaging in 40.2%(n=51). Random biopsies alone were performed in 112(88.2%) and complemented with targeted biopsies in 15(11.8%). diagnostic yield of random biopsies was significantly lower (2.7%;n=3) than targeted biopsies (40%;n=6)

(p< 0.001). There was a significant association between targeted biopsies and dysplasia detection, both in white-light (p< 0.001) and Narrow-band-imaging assessment (p=0.002). Targeted biopsies reached a sensitivity of 85.7%.

Conclusions Targeted biopsies had a significantly higher diagnostic yield, either under white-light or with Narrow-band-imaging, compared to random biopsies.