RESUMEN
BACKGROUND: This study aimed to explore the correlation of circulating microRNA (miRNA) expression profile with clinical response to tumor necrosis factor (TNF) inhibitor in treating rheumatoid arthritis (RA) patients. METHODS: Baseline PBMC samples from eight responders and eight non-responders after 24-week TNF inhibitor (etanercept) treatment were subjected to miRNA microarray. Then, top 10 dysregulated miRNAs were selected and further validated by quantitative polymerase chain reaction (qPCR) in baseline PBMC samples from 92 RA patients treated with 24-week TNF inhibitor (etanercept). Responders and non-responders were divided referring to the decline in disease activity score in 28 joints. RESULTS: In microarray assay, total 59 upregulated and 78 downregulated miRNAs were identified in responders compared to non-responders, which were mainly enriched in regulating immune- and inflammation-related biological processes and pathways. The top 10 dysregulated miRNAs were as follows: miR-192-5p, miR-146a-5p, miR-19b-3p, miR-320c, miR-335-5p, miR-149-3p, miR-766-3p, let-7a-5p, miR-24-3p, and miR-1226-5p. In qPCR validation, miR-146a-5p was increased, while let-7a-5p was decreased in responders compared with non-responders. Multivariate logistic analysis illuminated that miR-146a-5p and CRP independently correlated with higher clinical response, while let-7a-5p and biologics history independently associated with lower clinical response. Subsequently, receiver operating characteristic curve showed that combination of these four independent factors presented with a great predictive value for clinical response with area under curve: 0.863, 95% CI 0.781-0.945. CONCLUSION: miRNA expression profile is closely implicated in the treatment efficacy of TNF inhibitor, and combined measurement of miR-146a-5p, let-7a-5p, CRP, and biologics history disclosed a great predictive value for clinical response to TNF inhibitor in RA patients.