P0980 Virtual chromoendoscopy is comparable to dye spray chromoendoscopy for IBD surveillance: a single centre retrospective study

Abstract

Listen

Translate article

Abstract Background Patients with colonic inflammatory bowel disease (IBD) have an increased risk of developing colorectal cancer (CRC) and interval surveillance endoscopy is recommended 8 years after diagnosis or symptom onset to identify premalignant lesions (considered to be dysplastic adenomas, sessile serrate lesions [SSL] or large hyperplastic polyps). To enhance detection the current gold-standard technique is dye-spray chromoendoscopy (DCE). Advancements in high-definition electronic imaging have facilitated the development of virtual chromoendoscopy (VCE) and the 2019 European Society of Gastrointestinal Endoscopy (ESGE) guidelines advise that either DCE or VCE can be used. However, there is limited data comparing DCE and VCE. Methods We undertook a retrospective review of the 197 IBD surveillance colonoscopies performed at Frimley Health NHS Foundation Trust in a 6-month period from December 2023 to July 2024 using the accepted techniques of DCE, VCE and white light endoscopy (WLE) with segmental biopsies. DCE (with an indigo carmine mix) and WLE were performed using either Olympus 260, 290 or 1500 colonoscopes, and VCE with Olympus 290 or 1500 colonoscopes using TXI (Texture and Color Enhancement Imaging) optical enhancement in addition to EndoCuff Vision, narrow band imaging (NBI) and/or Endo-Aid CADe computer aided detection where appropriate. Results Of the 197 procedures, 125 (64%) were DCE, 44 (22%) VCE and 28 (14%) WLE with segmental biopsies. 43% (n=85/197) of patients were female and 57% (n=112/197) male. The median age was 52 years (range 21-86) and mean disease duration 18 years (range 2-48). Suspected premalignant lesions were identified in 56% (n=70/125) of cases using DCE; 68% (n=30/44) using VCE; and 29% (8/28) using WLE. Both DCE and VCE were superior to WLE at detecting lesions (p= 0.0016 and p=0.0035 respectively). There was no statistical difference in lesion detection between DCE and VCE (p>0.05). Withdrawal time (median) was shorter in VCE (20 minutes, range 6-67) than DCE (26 minutes, range 10-59), but shortest in WLE (15 minutes, range 7-42). Dysplasia, SSL or hyperplasia were histologically confirmed in 49% (n=34/70) of lesions identified with DCE, 60% (n=18/30) with VCE, and 50% (n=4/8) with WLE. Conclusion VCE is comparable to DCE for lesion detection. Both DCE and VCE are statistically superior to WLE for lesion detection. Withdrawal times are shorter in VCE than DCE without detriment to lesion detection. Optical diagnosis improved with the use of VCE. Our findings support VCE as an acceptable alternative to DCE for IBD surveillance and highlight potential advantages.