Association of baseline cytokine levels with anti-tumour necrosis factor treatment response in rheumatoid arthritis

Background Clinical trials of anti-tumour necrosis factor alpha (TNF) have shown efficacy in 60–70% of rheumatoid arthritis patients (RA). Predicting response to anti-TNF drugs at baseline remains an elusive goal in RA management. The purpose of this study was to determine if baseline levels of circulating cytokines, soluble receptors, adhesion molecules and metabolic factors differentiate future responders. Methods RA patients (n=29) with active disease were recruited, who had failed on disease modifying drugs and were recommended for anti-TNF treatment. Peripheral blood samples were collected at baseline. Responders were identified by a ≥1.2 reduction in disease activity score (DAS28-ESR) at 6 months. Five protein arrays quantified 33 proteins in pre-treatment plasma (Cytokine I, III and IV; Metabolic I; Adhesion molecule; Randox Laboratories Ltd., UK). Data was analysed by Pearson ranked correlation and logistic regression to clinical measures of disease activity at baseline and six months anti-TNF treatment. Results Elevated levels of interleukin 8 (p=0.043), monocyte chemoattractant protein (p=0.027), granulocyte-macrophage colony-stimulating factor (p=0.020) and soluble interleukin 2α receptor (p=0.01) were associated with high TJC at baseline. No significant relationship with the protein array panels was recorded relative to baseline DAS28-ESR. Significant inverse correlations were observed between absolute change in DAS28-ESR after 6 months of anti-TNF therapy and baseline levels of interleukin 6 (p=0.026) and resistin (p=0.044). Conclusions These findings suggest that baseline levels of specific circulating proteins may help to differentiate RA patient responders to anti-TNF therapies.