Artificial Intelligence For Outcomes Research in Crohn's Disease: Digital Pathology Assessment of Fibrosis and Association With Patient History and Clinical Outcomes

Overview

The goal of this retrospective observational study is to learn whether artificial intelligence-based digital pathology can measure intestinal fibrosis and help identify patients with Crohn's disease who may be at greater risk of disease progression and surgery. The main questions the study aims to answer are: Does the FibroNest Phenotypic Fibrosis Composite Score \[Ph-FCS(Muc)\], measured from routinely collected ileal biopsies, reflect the severity of fibrosis assessed by a pathologist? * Is Ph-FCS(Muc) associated with Crohn's disease characteristics and history? * Can Ph-FCS help predict whether a patient will subsequently require a first ileal surgical resection for fibrostenotic Crohn's disease? Researchers will retrospectively analyze previously collected ileal biopsies and medical information from adults with Crohn's disease. Biopsy slides will be digitized and analyzed using FibroNest digital pathology, and the resulting fibrosis measurements will be compared with pathologist assessments, clinical characteristics, disease history, and subsequent clinical outcomes.

Crohn's disease is a chronic inflammatory bowel disease that can lead to progressive intestinal fibrosis. Fibrosis contributes to structural bowel damage, strictures and obstruction and is an important cause of surgical intervention. However, intestinal fibrosis remains difficult to quantify objectively, particularly from routinely collected endoscopic biopsies, and there are limited validated tools to identify tissue fibrosis phenotypes associated with subsequent disease progression. AIFOR-CD (Artificial Intelligence for Outcomes Research in Crohn's Disease) is a retrospective observational study evaluating the use of artificial intelligence-based digital pathology to quantify and characterize fibrosis in routinely collected ileal biopsies from patients with Crohn's disease. The study uses the FibroNest Phenotypic Fibrosis Composite Score \[Ph-FCS(Muc)\], a continuous quantitative digital pathology biomarker derived from histological features of fibrosis measured in the mucosa of ileal biopsies. The study is organized around three complementary objectives: to validate Ph-FCS(Muc) against pathologist-assessed mucosal fibrosis severity; to investigate its relationship with Crohn's disease characteristics and history; and to explore whether quantitative fibrosis phenotypes in ileal biopsies can identify patients at increased risk of subsequent first ileal surgical resection for fibrostenotic Crohn's disease. The study will use previously collected histological specimens and retrospective clinical information. Digital pathology analyses will be performed using pseudonymized whole-slide images of ileal biopsies, with appropriate blinding between biomarker measurements and the reference assessments or clinical outcomes when applicable. AIFOR-CD is intended to determine whether quantitative assessment of fibrosis from routinely obtained endoscopic biopsies can provide objective information on intestinal fibrotic burden and contribute to the characterization and risk stratification of patients with Crohn's disease.