A Novel Circulating miRNA-Based Model Predicts the Response to Tripterysium Glycosides Tablets: Moving Toward Model-Based Precision Medicine in Rheumatoid Arthritis

Yanqiong Zhang, Hailong Wang, Xia Mao, Qiuyan Guo, Weijie Li, Xiaoyue Wang, Guangyao Li, Quan Jiang, Na Lin

Accumulating clinical evidence show that not all rheumatoid arthritis (RA) patients benefit to the same extent from a Tripterygium wilfordii Hook F (TwHF)-based therapy-Tripterysium glycosides tablets (TG tablets), which emphasizes the need of predictive biomarkers and tools for drug response. Herein, we integrated TG tablets' response-related miRNA and mRNA expression profiles obtained from the clinical cohort-based microarray, miRNA target prediction, miRNA-target gene coexpression, as well as gene-gene interactions, to identify four candidate circulating miRNA biomarkers that were predictive of response to TG tablets. Moreover, we applied the support vector machines (SVM) algorithm to construct the prediction model for the treatment outcome of TG tablets based on the levels of the candidate miRNA biomarkers, and also confirmed its good performance via both 5-fold cross-validation and the independent clinical cohort validations. Collectively, this circulating miRNA-based biomarker model may assist in screening the responsive RA patients to TG tablets and thus potentially benefit individualized therapy of RA in a daily clinical setting.