O64 Cumulative endoscopic inflammation does not predict colorectal neoplasia in PSC-IBD: a TAILOR-IBD study

Résumé fourni par la source

Background Patients with primary sclerosing cholangitis and inflammatory bowel disease (PSC-IBD) are at high risk of colorectal neoplasia (CRN). While cumulative inflammation is known to drive neoplastic risk in IBD, the relationship between inflammation and CRN in PSC-IBD remains poorly defined. Methods We conducted a national retrospective multicentre cohort study of PSC-IBD patients. Those with CRN or colectomy prior to first assessment were excluded. Follow-up continued until first CRN, colectomy, or last follow-up, with analyses restricted to available electronic colonoscopy data. CRN was classified as low-risk (indefinite for dysplasia (IND), low-risk low-grade dysplasia [lrLGD]), high-risk (hrLGD [≥1 cm, flat, or irregular], high-grade dysplasia [HGD], or colorectal cancer [CRC]). Endoscopic cumulative inflammatory burden (eCIB) was derived from serial colonoscopies using a time-weighted severity score, assigning mild activity where inflammation was recorded without grading. A time-dependent Cox model assessed associations with first CRN. Results Among 439 patients (64% male), median age at IBD diagnosis was 22 years (15-33) and PSC diagnosis of 28 years (16-40). Over a median follow-up of 7 years (4-12), 77 patients (17.5%) developed CRN: 3 CRC, 5 HGD, 40 hrLGD, 25 lrLGD, 4 IND. Patients with high- and low-risk CRN were older at CRN detection than those with no CRN at last follow-up (54 years [41-69] and 49 [44-62] vs 38 [29-52]; p<0.001) and diagnosed at an older age (31 years [22-49], 28 [23–42], 21 [15-33] respectively, p<0.001). Surveillance intensity was higher in high- and low-risk CRN groups (0.8/year and 0.9/year vs. 0.6; p<0.001). All CRN were predominantly right-sided, with 68% of high-risk and 73% of low-risk lesions arising in the right colon. Active inflammation was frequently observed around the time of high-risk CRN detection; seen in 60% at index or preceding colonoscopy with at least a 40% concordance of lesion-inflammation segment, compared to 21% in low-risk CRN with 8% concordance (p=0.005). However, median eCIB over all available procedures was not higher in the high-risk (1.6, 0–3.4), compared to the low-risk and no CRN groups (0 [0–3.3] and 2.1 [0–5.0]; p=0.09). Cox regression analysis did not show separation between the groups based on eCIB (aHR 1.06 per unit, 95% CI 0.97–1.15). Crohn’s disease (aHR 0.42, 95% CI 0.19-0.95; p=0.04) reduced risk but PSC duct type did not impact CRN risk (aHR 0.71 95% CI 0.3-1.67; p=0.43). Conclusions While a temporal and spatial association between high-risk colorectal neoplasia and active inflammation was observed, cumulative inflammatory burden measured by endoscopic activity did not associate with colorectal neoplasia in PSC-IBD. This contrasts with conventional ulcerative colitis, in which cumulative inflammatory burden is a key determinant of neoplasia risk, and suggests that surveillance strategies in PSC-IBD should not rely on cumulative endoscopic activity alone.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.