P382 Novel Image Enhanced Endoscopy (IEE) technologies for surveillance of colonic Inflammatory Bowel Disease: A pilot study evaluating Texture and Colour Enhancement Imaging (TXI) and Linked Colour Imaging (LCI) as compared with chromoendoscopy.

Picardo, S.(1);Menon, S.(1);So, K.(1);Venugopal, K.(1);Cheng, W.(1);Ragunath, K.(1);

(1)Royal Perth Hospital, Department of Gastroenterology and Hepatology, Perth, Australia;

Background

Colorectal cancer is a serious complication associated with long-standing inflammatory bowel disease (IBD). Current guidelines recommend routine surveillance with dye-spray chromoendoscopy (CE) to facilitate early detection of premalignant lesions. Novel image enhanced endoscopy (IEE) technologies, including Linked Colour Imaging (LCI) (Fujifilm) and Texture and Colour Enhancement Imaging (TXI) (Olympus), alter the way light interacts with tissue and have been demonstrated to improve adenoma detection in screening colonoscopy. We evaluated these technologies in the surveillance of IBD.

Methods

Patients that met Australian criteria for IBD surveillance, at a tertiary IBD centre, between January and October 2021, were enrolled to undergo colonoscopy with either LCI or TXI, followed immediately by CE. Each colonic segment was examined with the respective IEE and all lesions identified were photographed and classified according to Paris and JNET classifications. Lesions were endoscopically characterised as neoplastic or non-neoplastic with the aid of additional IEE characterisation modes, Narrow Band Imaging (NBI)(Olympus) and Blue Light Imaging (BLI)(Fujifilm). Each segment was then examined with CE (methylene blue 0.1%) and all lesions identified were classified and resected. Diagnostic yield, missed lesions and accuracy in characterising neoplastic lesions were calculated for each modality.

Results

25 patients, median age 49 years, 52% male, were included. 16 underwent surveillance with TXI and 9 with LCI. Patient and procedural characteristics are summarised in Table 1. TXI identified 22 lesions (8 neoplastic) and LCI 11 lesions (3 neoplastic) (Table 2). 3 non-neoplastic lesions were missed by IEE and subsequently identified with CE (2 TXI and 1 LCI). These were all small (<5mm) non-polypoid lesions. Neoplastic lesions were identified in 5 patients with an overall detection rate of 20%. 2 lesions (1 TXI/NBI, 1 LCI/BLI) were characterised endoscopically as non-neoplastic but subsequently confirmed as neoplastic (both sessile serrated lesions). 5 lesions (4 TXI/NBI and 1 LCI/BLI) were characterised as neoplastic but subsequently confirmed as non-neoplastic (4 inflammatory and 1 hyperplastic) (Table 3). IEE had a sensitivity, specificity, and accuracy of 82%, 77% and 79% respectively, in prediction of neoplastic lesions.

Conclusion

The novel IEE modalities TXI and LCI are effective in colonic IBD surveillance with no neoplastic lesions missed as compared to CE. Overall neoplastic lesion detection rate was 20%. The utility of IEE to accurately characterise neoplastic potential was 79%. A larger randomised controlled trial comparing these novel technologies to CE is required and may eliminate the need for CE in the future.