Introduction Eighty percent of paediatric patients with Primary Sclerosing Cholangitis (PSC) have underlying Inflammatory Bowel Disease (IBD). Delineating the unique macroscopic intestinal phenotype of PSC-IBD has revealed a predilection for pancolitis, often more severe in the right colon, backwash ileitis and rectal sparing. However, the histologic characteristics of PSC-IBD have not been critically examined. Furthermore, histologic activity indices developed and used in conventional UC, such as the Nancy Index (NI), haven’t been evaluated in PSC-IBD.

Aim The primary aim of this study was to delineate the histologic phenotype of paediatric PSC-IBD. Secondary aims were to assess the inter-rater reliability and construct validity of the NI in paediatric PSC-IBD.

Subjects and Methods We included children with PSC-IBD and controls with UC with baseline pre-treatment colonoscopy between 2000 and 2018, matched on sex, and age and year at diagnosis. Disease activity colonoscopy was assessed by PUCAI score and physician global assessment of endoscopic severity (ES). Two paediatric GI pathologists independently reviewed diagnostic mucosal biopsies. NI was determined for both the right and left colon. The prevalence of individual histologic features in PSC-IBD versus UC was compared by Chi squared test. Inter-rater reliability was assessed using Fleiss’ Kappa and intra-class correlation coefficients (ICC), as appropriate. Construct validity of the Nancy Index was assessed against ES, using Spearman correlations. Logistic regression was used to examine the association between histologic features and ES and clinical outcomes.

Results 50 PSC-IBD patients (median age 13.4 years, 68% male) and 81 UC controls (median age 12.3 years, 65% male) were included. Histopathologically, pancolitis, more severe inflammation in the right colon, and backwash ileitis were all significantly more common in PSC-IBD patients vs. colitis controls (table 1, all p<0.05). Basal plasmacytosis, lamina propria predominant neutrophils and eosinophilia were more common in PSC-IBD (all p<0.05). NI inter-rater reliability was very high (ICC >0.9). Right and left Nancy index correlated with corresponding ES in both PSC-IBD and UC controls (for right colon, respectively r=0.31 and 0.34, p<0.05; for left colon, respectively r=0.52 and 0.26, p<0.05). Both the right and left Nancy Index correlated moderately with PUCAI in controls (r=0.31–0.42, p<0.01) but not in PSC-IBD (r=0.1, p>0.05).

The right NI was predictive of a need for steroids in PSC-IBD (OR 2.92, 95% CI 0.996 – 8.62). In controls, crypt abscess (p=0.034), ulceration (p=0.05) and chronic inflammation (p< 0.01) (all components of the NI) accounted for most of the variance in ES, as shown in table 2. While these features were also associated with ES in PSC-IBD, additional features including surface villiform change, basal plasmacytosis and eosinophilia, were significantly (all p<0.05) correlated with ES as well.

Summary and Conclusion The distinct endoscopic features of PSC-IBD are confirmed on histologic examination of mucosal biopsies. While our findings support good construct validity for the Nancy Index in PSC-IBD, they also demonstrate that additional histologic features are unique to PSC-IBD. This may warrant development of a PSC-IBD-specific histologic activity index. Clinical activity correlates poorly with NI in PSC-IBD, consistent with past studies showing subclinical inflammation.