P379 Surveillance with chromoendoscopy to detect new dysplasia in inflammatory bowel disease

Abstract

Abstract Background Inflammatory Bowel Disease patients (IBD) are in a higher risk of colorectal cancer (CRC) than the general population. Studies show that surveillance with chromoendoscopy (CE) is useful for prevention as it detects a higher number of dysplastic lesions, allowing their endoscopic resection and therefore preventing CRC. We aimed to assess our endoscopic results in IBD patients with previous diagnose of dysplasia. Methods We performed a retrospective data collection from IBD patients who have had dysplastic lesions in screening chromoendoscopy, performed in our hospital from January 2013 to June 2019. Demographic and clinical data at basal time were collected, as well as endoscopic and histological findings in basal chromoendoscopy and in the first follow-up Results 353 chromoendoscopies were performed in our Endoscopy Unit in this period and 50 patients showed dysplastic lesions. 14 (28%) of them had already had dysplasia in previous colonoscopies, which were performed with white light and aleatory biopsies. Basal characteristics are shown in the following table. In the initial chromoendoscopy, an average of 2.57 dysplastic lesions with an average size of 7.59 mm was detected. 70.8% were flat lesions (0-IIa/0-IIb Paris classification) and 53% were located distal to the splenic flexure. 85.7% were tubular adenomas low-grade dysplasia (LGD), 2 cases of high-grade dysplasia (HGD), one serrated polyp and one adenocarcinoma. Twenty-nine follow-up chromoendoscopy have been performed and new dysplasia lesions were found in 41.37% (average of 1.86 lesions, median 8 mm). 55.6% of the lesions were tubular adenomas LGD and 11.1% serrated adenomas LGD. Our analysis shows that patients with new lesions in the second chromoendoscopy had had worse bowel preparation on the first one (7.5 vs. 8.9, p = 0,006). We also found that these patients with new lesions had more flat lesions than sessile ones (70.6% Paris IIa, 23.5% IIb, 5.9% Is, p < 0,005). Conclusion Surveillance with chromoendoscopy in IBD patients with previous dysplastic lesions detects new dysplasia in 28% of them and LGD is the most frequent histology. In our sample, finding flat lesions or worse bowel cleaning in the first chromoendoscopy predicts having a higher number of lesions on the follow-up.