Pediatr Gastroenterol Nutr. 2024 Dec 17. doi: 10.1002/jpn3.12434. Online ahead of print.

Rebecca Little ¹, Juan Putra ², Binita M Kamath ^1 3, Anne M Griffiths ^1 3, Amanda Ricciuto ^1 3, Iram Siddiqui ^2 4

Author information

¹Division of Gastroenterology, Hepatology and Nutrition, The Hospital for Sick Children, Toronto, Ontario, Canada.

²Division of Pathology, The Hospital for Sick Children, Toronto, Ontario, Canada.

³Department of Paediatrics, University of Toronto, Toronto, Ontario, Canada.

⁴Department of Laboratory Medicine & Pathobiology - Anatomic Pathology, University of Toronto, Toronto, Ontario, Canada.

Abstract

Background: We aimed to characterize the histologic gut phenotype of pediatric primary sclerosing cholangitis (PSC)-associated inflammatory bowel disease (IBD) against non-PSC colitis, and to assess Nancy Index (NI) performance in pediatric PSC-IBD.

Methods: Single-center retrospective cohort study including children diagnosed with PSC-IBD or non-PSC colitis (ulcerative colitis [UC] or IBD-unclassified) from 2000 to 2018, with diagnostic intestinal biopsies. Biopsies were re-reviewed by two independent pathologists who assessed microscopic disease distribution, NI scores, and specific histological features in the right and left colons, overall and stratified by endoscopic severity (moderate-severe vs. no more than mild). We examined NI inter-rater reliability with Fleiss' weighted (quadratic) kappa and NI construct validity against global endoscopic severity (Spearman correlation) and clinical outcomes (logistic regression).

Results: Fifty children with PSC-IBD and 81 colitis controls were included. Histologically, pancolitis (84% vs. 55%), right colon-predominant colitis (48% vs. 3%), and backwash ileitis (53% vs. 12%) (all p < 0.01) were significantly more common in PSC-IBD; histologic rectal sparing occurred at similar rates (6% vs. 10%, p = 0.54). Lamina propria-predominant neutrophils, prominent eosinophilic infiltration (left colon), and surface villiform change (right colon) were more common in PSC-IBD than colitis controls (p < 0.01). NI showed excellent inter-rater reliability (kappa > 0.9) and correlated moderately with global endoscopic severity but poorly with clinical activity in PSC-IBD.

Conclusions: Pediatric PSC-IBD has a distinct histologic phenotype that largely mirrors the endoscopic phenotype in distribution and includes a greater frequency of features not included in conventional UC histologic activity indices. Future work should investigate whether a PSC-IBD-specific index incorporating these features is warranted.