Biopsy pathology predicts patients with ulcerative colitis subsequently requiring surgery

Area of Science:

• Gastroenterology
• Pathology
• Predictive Medicine

Background:

• Ulcerative colitis (UC) management is challenging due to unpredictable clinical courses.
• 17%-38% of UC patients eventually require surgical intervention.
• Histological differences may distinguish patients needing surgery from those managed medically.

Purpose of the Study:

• To identify specific histological criteria for predicting medical treatment failure in UC.
• To develop a predictive model for surgical intervention in UC patients.

Main Methods:

• Colorectal biopsy specimens from 67 UC patients requiring surgery (UC-S) and 90 UC patients on medication alone (UC-M) were analyzed.
• Multiple logistic regression analysis was performed on 70 histological features, disease extent, and age.
• A predictive equation for UC-S probability (P(UC-S)) was developed and validated using ROC curve analysis.

Main Results:

• Deep ulceration, frequent crypt abscesses, focal/segmental mononuclear cell infiltration, paucity of eosinophils, and wide disease extent were significant predictors of UC-S.
• The derived regression equation demonstrated high predictive accuracy.
• Risk categories (highest to lowest) were established with high sensitivity (>91.0%) and specificity (>98.5%) for higher-risk and lower-risk groups.

Conclusions:

• A novel set of criteria integrating histological features and endoscopic disease extent reliably predicts UC patient outcomes.
• These criteria are expected to assist in determining the necessity of surgical intervention for UC.
• This predictive model offers a valuable tool for personalized UC management strategies.