Search

DOP03 Early change in epithelial neutrophilic infiltrate predicts long-term response to biologics in Ulcerative Colitis

Narula, N.(1);Wong, E.(1);Colombel, J.F.(2);Riddell, R.(3);Marshall, J.(4);Reinisch, W.(5);Dulai, P.(6)

(1)McMaster University, Department of Medicine- Division of Gastroenterology, Hamilton, Canada;(2)Icahn School of Medicine at Mount Sinai, Division of Gastroenterology, New York, United States;(3)Mount Sinai Hospital, Department of Pathology and Laboratory Medicine, Toronto, Canada;(4)McMaster University, Department of Gastroenterology, Hamilton, Canada;(5)Medical University of Vienna, Department of Internal Medicine III- Division of Gastroenterology and Hepatology, Vienna, Austria;(6)University of California San Diego, Division of Gastroenterology, La Jolla, United States

Background

The prognostic value of histologic scores, grades, and individual histologic sub-components, alone or in combination with endoscopy, for predicting endoscopic improvement (EI) and histo-endoscopic mucosal improvement (HEMI) during maintenance therapy in ulcerative colitis (UC) remains uncertain.

Methods

Post-hoc analysis of participants from the VARSITY trial (n=734 with histology). Receiver operating characteristic and multi-variate logistic regression analyses were performed to assess whether baseline and/or week 14 assessments for the Robarts Histopathology Index (RHI), Geboes score, or individual histologic sub-components, could predict the achievement of EI and HEMI at week 52.

Results

RHI and Geboes scores at baseline, and changes in scores at week 14, had fair to poor accuracy for predicting week 52 EI or HEMI (Area Under the Curve [AUC] 0.58-0.67). Among individual sub-components, changes in epithelial neutrophil involvement from baseline to week 14 had the best performance for predicting week 52 EI (AUC 0.83, 95% confidence interval (CI) 0.74-0.91) and HEMI (AUC 0.85, 95% CI 0.76-0.94). On multivariate analyses, improvement of neutrophils in the epithelium was associated with increased odds of week 52 EI (Odds Ratio [OR] 3.63, 95% CI 1.45-9.08, p=0.0059) and HEMI (OR 6.88, 95% CI 3.29-14.36, p<0.0001), and lack of improvement of neutrophils in the epithelium was associated with lack of week 52 EI (OR 0.24, 95% CI 0.12-0.48, p<0.0001) and lack of HEMI (OR 0.03, 95% CI 0.01-0.014, p<0.0001). No other parameters significant on univariate analysis were significant within the multivariate model.  Among patients with endoscopic Mayo scores of 2 or 3 at week 14, those who had >50% crypts involved with neutrophils at week 14 were significantly less likely to achieve week 52 EI (18.0%; 31/172 vs. 34.3%; 72/210, p<0.001), HEMI (9.9%; 17/172 vs. 22.4%; 47/210, p=0.001) (Figure 1) and clinical remission (11.0%; 19/172 vs. 23.3%; 49/210, p=0.002). Among patients with endoscopic Mayo scores of 0 or 1 at week 14, those who had >50% crypts involved with neutrophils were significantly less likely to achieve HEMI at week 52 (33%; 4/12 vs. 62.4%; 141/226, p=0.044) (Figure 1).


Figure 1: Histo-endoscopic improvement at week 52 stratified by week 14 Mayo endoscopic score and neutrophil presence in the epithelium

Conclusion

Our results on epithelial neutrophilic infiltrate following induction therapy as the only independent predictor for achievement of maintenance EI or HEMI helps clarify the clinical relevance of measuring histologic disease activity in UC. Epithelial neutrophilic infiltrate poses a means to stratify patients according to their likelihood of response to biologic treatment.

Portal