Investigation of T regulator and T helper cells in urticaria immunopathogenesis: Investigation of RORγt, mTOR, STAT3, FOXP3 transcription factors and related cytokine levels in serum

A cross-sectional study was conducted to elucidate the contribution of Th17/Treg imbalance and the associated cytokine environment in the pathogenesis of chronic spontaneous urticaria (CSU), to evaluate the diagnostic performance of an IL-23-centered biomarker panel and its relationship with disease activity, and to develop machine learning-based predictive models. The cross-sectional study included 45 CSU patients and age- and gender-matched healthy controls. Serum levels of FOXP3, IL-17A, IL-22, IL-23, IL-6, IL-13, IL-21, IL-10, IL-35, mTOR, STAT3, and RORγt were measured by ELISA. Demographic and clinical data were recorded. Statistical analyses were performed using SPSS v26.0; ROC analysis and logistic regression, as well as Random Forest, SVM, and XGBoost algorithms, were applied using the Python scikit learn library. In the CSU group, IL-23, IL-6, IL-13, IL-17A, IL-22, and RORγt were significantly elevated, while mTOR was markedly reduced (all p<0.001). A threshold of >1.34 pg/mL for IL-23 provided 87% sensitivity and 100% specificity. The best diagnostic performance (AUC=0.975) was achieved with the Random Forest model, and the strongest predictors were IL-23, IL-6, mTOR, and IL-13. Disease activity (UAS7) showed a positive correlation with D-dimer (ρ=0.37) and a negative correlation with IL-21 (ρ=−0.38). High IL-23 was an independent predictor for distinguishing the Type 2B subtype (OR = 13.6). The findings support the altered Th17/Treg axis in CSU and reveal IL-23 as a strong diagnostic biomarker; the Random Forest-based approach integrated with routine laboratory parameters offers high accuracy in CSU classification. Multicenter validation and investigation of the clinical impact of IL-23-targeted strategies are recommended. Keywords: Chronic spontaneous urticaria; Th17; Treg; IL 23; Biomarker; Machine learning