Determination of Heterogeneous Proteomic and Metabolomic Response in anti-TNF and anti-IL-6 Treatment of Patients with Rheumatoid Arthritis

Abstract

Reduction in tumor necrosis factor (αTNF) and interleukin-6 (IL-6) activities is a widely utilized strategy for the treatment of rheumatoid arthritis (RA) with a high success rate. Despite both schemes targeting the deprivation of inflammatory reactions caused by the excessive activity of cytokines, their mechanisms of action and the final output are still unequal. This was a comparative longitudinal study that lasted for 24 weeks and aimed to find the answer to why the two schemes of therapy can pass out of proportion in attitude of their efficiency. What are the differences in metabolic and proteomic responses among patients who were being treated by either the anti-TNF or anti-IL-6 strategy? We found increased levels of immunoglobulins A and G (more than 2-fold in anti-IL-6 and more than 4-5-fold in anti-TNF groups) at the final stage (24 weeks) of monitoring but the most profound increase was determined for µ-chains of immunoglobulins in both groups of study. Metabolomic changes displayed main alterations with regard to arginine metabolism and collagen maintenance, where arginine increased 8.86-fold (p < 0.001) in anti-TNF and 5.71-fold (p < 0.05) in anti-IL-6 groups but patients treated by the anti-TNF scheme suffered a higher depletion of arginine before the start of therapy. Some indicators of matrix and bone tissue degradation also increased 4-hydroxyproline (4-HP) more than 6-fold (p < 0.001) in anti-TNF and more than 2-fold (p < 0.05) in the anti-IL-6 group, but the growth dynamics in the anti-IL6 group was delayed (gradually raised at week 24) compared to the anti-TNF group (raised at week 12) following a smooth reduction. The ELISA analysis of IL-6 and TNFα concentration in the study population supported proteomic and metabolomic data. A positive correlation between ΔCDAI and ΔDAS28 indicators and ESR and CRP was established for the majority of patients after 24 weeks of treatment where ESR and CRP reduced by 20% and 40% finally, respectively. A regression model using the Forest Plot was estimated to elucidate the impact of the most significant clinical, biochemical, and anthropometric indicators for the evaluation of differences between considered anti-TNF and anti-IL-6 schemes of therapy.

Keywords: ELISA; HPLC-MS/MS; inhibitors of IL6; inhibitors of TNF; metabolome; proteome; rheumatoid arthritis.

Figure 1

Dynamics of semiquantitative changes in the proteome of patients in the anti-αTNF (A) group and in the group with anti-IL6 (B) therapy during 24 weeks of treatment to baseline values before the start of therapy. The log2FC values for Ig reflect the fraction of proteins of the immunoglobulin class. Log₂FC is calculated as a ratio of the median of Ig-proteins quantity toward its quantity at the initial point (before the starting treatment).

Figure 2

Change in clinical ΔCDAI in the group of patients with anti-αTNF (A) and anti-IL6 therapy (B) and ΔDAS28-ESR in the group of patients with anti-αTNF (C) and anti-IL6 (D) therapy for 24-x weeks of treatment. DAS28-ESR is used to determine eligibility for biologic therapies. DAS28-ESR ≥ 3.2 can be used as a threshold for classifying active RA.

Figure 3

Correlation between ΔDAS28 and ESR in patients with anti-αTNF (A) and anti-IL6 (B) therapy and between ΔDAS28 and CRP in the group of patients with anti-αTNF (C) and anti-IL6 (D) therapy during 24 weeks of treatment.

Figure 4

Inhibition of CRP production and the dynamics of the decrease in the erythrocyte sedimentation rate in patients of the anti-αTNF (A) and anti-IL6 (B) therapy groups during a 24-week course of treatment.

Figure 5

Dynamics of changes in endogenous metabolites during a 24-week course of treatment among patients in the anti-αTNF (A) and ani-IL6 (B) groups of therapy. The Log₂FC value is calculated as a median quantity of a certain metabolite toward its quantity at the initial (before starting treatment) point of treatment.

Figure 6

Dynamics of changes in the content of TNF (A) and IL6 (B) in blood samples of patients with anti-αTNF therapy. The measurements were performed by the ELISA method (Alfa-TNF-IFA-BEST and IL-6-IFA-BEST, Vector Best, Novosibirsk, Russia) in three technical repetitions.

Figure 7

Forest Plot showing the regression coefficients of the confidence intervals of variables in a linear model testing the association of scores with response to anti-TNF (A) and anti-IL6 (B) therapy at week 24.