Introduction Coeliac disease (CD) is common yet under-diagnosed at endoscopy. The optical zoom technology of Pentax Magniview endoscopes can magnify images by 136 times. We aimed to assess the accuracy of Magniview in the detection of duodenal villous atrophy (VA), and the ease of interpretation of live zoomed-in images.

Method Patients attending for a gastroscopy from 2014–17 were recruited. Endomysial (EMA) and tissue transglutaminase (TTG) antibodies were taken at endoscopy. Appearances of the first and second part of the duodenum (D1 and D2) observed with high definition white light endoscopy (HDWLE) and Magniview were graded on a 3-point scale validated by Cammarota et al.: normal, partial or marked villous atrophy (PVA, MVA). In all patients, quadrantic biopsies were taken from D2 and 1 biopsy from D1. Endoscopic findings of the duodenum were compared to histology as the reference standard. A short training video with 12 clips of normal duodenum and various grades of VA with Magniview was used to train medical staff to identify VA. They were assessed immediately after the training with 22 video clips for the ability to detect normal duodenum, PVA or MVA.

Results A total of 250 patients were recruited (145 females, 58%; age range 16–84, median age 50.5). Ninety-six patients had VA on histology (38.4%), of which 80 were newly diagnosed CD and 16 were persistent VA in known CD patients, leaving the remaining 154 as controls. Magniview had a significantly higher sensitivity in detecting VA compared to HDWLE (p=0.0005). (Table 1) Ten individuals (5 endoscopists and 5 non-endoscopists) undertook the Magniview training, and 9/10 correctly identified all cases of VA. (Table 2)

Conclusion Magniview had superior diagnostic accuracy in detecting VA compared to HDWLE. A short training on the identification of VA with Magniview demonstrated its ease of use and suitability for Magniview to be utilised by general endoscopists in their daily practice to increase the detection of CD.

Disclosure of Interest W. White: None Declared, M Lau: None Declared, P Mooney: None Declared, M Rees: None Declared, M Burden: None Declared, D Sanders Conflict with: Professor Sanders has received educational research grants from Dr Schaer (a gluten-free food manufacturer) and Tillotts Pharma (producer of a point of care test for celiac disease) for investigator led studies. Dr Shaer and Tillott’s Pharma did not have any input in the study design, access to study data, interpretation of the findings or drafting of the manuscript.