Background/Purpose: discovery of biomarkers that can identify pretreatment patients who will respond to biologic DMARDs therapy is one of the major interests in RA. We proposed to test the possible predictive role of RF isotypes IgM and Ig A, anti-CCP, anti-mutated citrullinated vimentin (anti-MCV), 14-3-3 eta protein and COMP on a group of patients treated with anti-TNF α agents.We have also assessed the status of these biomarkers and response to biologic therapy.

Methods: prospective and observational study including 64 patients followed 12 months with active RA, uncontrolled by conventional synthetic DMARDs or declared nonresponders to one of the biologic DMARDs and required switch to another biologic DMARDs. Clinical assessment was performed at 0, 6 and 12 months according to ACR criteria approved by OMERACT and evaluation of treatment response according to EULAR criteria (good /moderate /nonresponder).

Results: mean age was 57.55 ± 9.427 years, of the 64 patients included in the study 59 (92.2%) were women and 5 (7.9%) men. At 6 months, 7 patients were declared nonresponders, 38 achieved a moderate response and 19 good response. Following baseline immunological parameters titres and the response at 6 months, general tests have identified significant differences between groups.Tests for identifying differences between the groups showed that lower titres of both RF isotypes , anti-CCP, 14-3-3 eta protein and COMP had predictive value on achieving a good EULAR response at 6 months. Grouping patients in 2 categories (responders/nonresponders), just 14-3-3 eta protein and anti-CCP maintained their predictive value for the response at 6 months.

After 12 months, 1 patient was declared nonresponder, 11 achieved moderate response and 14 good response. For this visit, lower baseline titres for RF type Ig M (92.93±120.22U/ml, p=0.01032) and Ig A (49.96±98.08 U/ml,p=0.00247) had predictive value for achieving a good response at 12 months . We didn’t obtain other informations grouping patients in 2 categories . Regarding the status of biomarkers and treatment response, the only differences were obtained for COMP (p = 0.0001) at 6 months for good response and RF type IgA (p=0.0041) for good response at 12 months. Using multivariate logistic regression methods we obtained a statistical model for predicting the response at 6 months including normal values for 14-3-3 eta protein, anti-CCP and COMP. The model is a good approximation for the situation analyzed (Hosmer and Lemeshow according test λ² = 5.795, p = 0.670 ≥0.05) with a predictive response accuracy of 89.1%.

Conclusion: because until now has not been discovered a single biomarker capable of distinguishing pretreatment responders versus nonresponder, in the future a version using multiple biomarkers could increase accuracy for identifying pretreatment these patients.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictive-role-of-biomarkers-for-the-response-to-biologic-therapy-in-rheumatoid-arthritis-old-new-and-how-future-looks-like/