Study on the predictive model of response of patients with inflammatory bowel disease to infliximab treatment

This study aimed to develop a predictive model for how patients with inflammatory bowel disease (IBD) respond to infliximab (IFX) treatment. One hundred adult IBD patients admitted to Shulan (Hangzhou) Hospital from August 2023 to November 2024 were included and divided into response and non-response groups based on their reaction to IFX. The response group consisted of 57 patients (57.0%), while the non-response group had 43 patients (43.0%). Clinical data, including gender, age, BMI, disease type (Crohn’s disease/ulcer-ative colitis), disease activity indices (CDAI/UCAI), history of IFX treatment, and infusion reactions, were collected and compared between the two groups. Additionally, biomarker levels, such as TNF-α, CRP, calprotectin, anti-infliximab antibody (ATI), IL -6, and IL -8, were measured during the midcourse of IFX treatment. Single-factor analysis identified variables that differed, and logistic regression showed that calprotectin level (OR=1.099, 95%CI=1.039-1.163), ATI (OR=3.756, 95%CI=1.222-11.546), IL -6 (OR=1.261, 95%CI=1.069-1.488), and IL -8 (OR=1.014, 95%CI=1.004-1.024) were key factors influenc-ing treatment response (p < 0.05). A nomogram was created using these fac-tors to predict treatment response in IBD patients. ROC analysis showed AUC values of 0.809, 0.762, 0.850, and 0.775 for calprotectin, ATI, IL -6, and IL -8, respectively, with corresponding 95% confidence intervals. The calibration curve indicated good model fit. These findings underscore the important roles of these cytokines in IBD pathogenesis and the action of IFX, as well as the high predictive power of the nomogram model.