Background: The presence of anti-cytokine autoantibodies (ACAAs) seems to be a physiologic mechanism to control the immune response and regulate cytokine activity. Biological therapies also regulate cytokine activities and have greatly improved the quality of life of RA patients. However, alteration of the cytokine network by the use of these treatments may lead to a disbalance in the regulatory system of ACAAs. We hypothesize the ACAAs network may influence the course of immune response in RA patients and may be useful to predict the therapy efficacy.

Objectives: We aimed to explore the potential of circulating ACAAs to predict flare in a cohort of RA patients treated with biological therapy.

Methods: We employed sera at baseline from 194 RA patients of the clinical trial OPTIBIO ¹ (A Coruña), whose primary endpoint is to evaluate the usefulness of standardized protocol strategies of dose reduction in patients with RA in clinical remission treated with biologics. These patients were treated with TNF inhibitors (Etanercept, N=47; Infliximab, N=12; Adalimumab, N=35; Certolizumab (CTZ), N=17; Golimumab, N=5), Tocilizumab (TCZ, N=60) and Abatacept (ABA, N=18). Patients were in clinical remission (DAS 28 <2.6 or SDAI <5 or ACR/EULAR 2011 criteria) at least from 6 months. Patients were followed during a minimum period of one year and maximum period of 3 years. Flare was considered when remission criteria were not fulfilled. The bead-based antigen array MILLIPLEX MAP Human Cytokine Autoantibody Magnetic Bead Panel was used for the simultaneous detection and quantification in sera of anti-BAFF, anti-G-CSF, anti-IFNβ, anti-IFNγ, anti-IL-1α, anti-IL-6, anti-IL-8, anti-IL-10, anti-IL-12 (p40), anti-IL-15, anti-IL-17A, anti-IL-17F, anti-IL-18, anti-IL-22, and anti-TNFα. Non-parametrical tests, ROC curves and logistic regressions were performed for the statistical data analysis using SPSS. P-value < 0.05 was considered statistically significant.

Results: The levels of anti-17A and anti-IL-1α were increased in the sera from patients who suffered a flare during the follow-up period (N= 76), compared to those who remained in remission (N= 118), showing an area under the curve (AUC) of 0.586 and 0.594, respectively. Segregating by treatment, the levels of anti-17A were specifically increased in those relapsing patients under CTZ (N=6), ABA (N=12) and TCZ (N=20) treatment. The AUC of anti-17A within these three therapies was 0.867, 0.903 and 0.682, respectively. Logistic regression analysis also associated the levels of anti-17A with the risk of suffering a flare in TCZ-treated patients (OR=1.11; p =0.015, for 100 MFI increase). In addition, the TCZ-treated patients who suffered a flare also showed higher levels of anti-IL17F, anti-IL-1α, and anti-IL-18 compared to those that remained in remission, showing AUCs of 0.689, 0.657 and 0.698, respectively. Anti-IL-18 was also associated with the risk of flare in these patients (OR=1.65; p =0.028, for 100 MFI increase). The presence of these three ACAAs was also higher in the TCZ-treated patients who suffered a flare compared to those in remission.

Conclusion: Although further validation of our results is needed, we present a ground-breaking study showing the potential of anti-IL17A, anti-IL-1α, and anti-IL18 to predict flare in RA patients under biological therapies.

REFERENCES:

[1]Bejerano C, et al. Clinical evaluation usefulness of standardized protocol strategies of dose reduction in patients with RA in clinical remission treated with biologic therapies. The Optibio Study. Arthritis Rheumatol. 2016; 68 (suppl 10): 649.

Disclosure of Interests: None declared.