Identification of a novel transcriptome signature for predicting the response to anti-TNF-α treatment in rheumatoid arthritis patients

Abstract

ObjectivesTo identify and validate a transcriptomic signature capable of predicting the response to antitumour necrosis factor (TNF) therapy in patients with rheumatoid arthritis (RA) before treatment initiation. MethodsWe performed a retrospective transcriptomic analysis using two public datasets, RNA-seq data from peripheral blood mononuclear cells in GSE138746 and microarray data from whole blood in GSE33377, to define a small-scale gene signature predictive of the response to anti-TNF treatment. Two external validations were then conducted using data from the COMBINE and the Reina Sofia Hospital cohorts, resulting in a total of 253 individuals, 155 responders and 98 nonresponders. ResultsInitial RNA-seq analysis (GSE138746) revealed 53 genes that were differentially expressed between responders and nonresponders; however, none of these genes remained significant after p value adjustment with the Benjamini-Hochberg method. A small-scale genetic signature comprising the 18 most discriminatory genes was then developed, achieving a leave-one-out cross-validation (LOOCV) predictive accuracy of 88.75%. We further refined this list to seven genes (COMTD1, MRPL24, DNTTIP1, GLS2, GTPBP2, IL18R1, and KCNK17) that effectively predicted the response to anti-TNF- treatment, with an area under the receiver operating characteristic (ROC) curve (AUC) of 0.84 in the GSE33377 dataset. Internal validation of the GSE138746 dataset yielded an AUC of 0.89. Finally, external validation confirmed the robustness of the seven-gene model (AUC[≥] 0.85). ConclusionsWe identified a transcriptomic signature that aids the prediction of the response to anti-TNF treatment in RA patients. These findings support its potential use as a precision medicine tool to improve therapeutic decision-making and reduce exposure to ineffective treatments in RA patients.