OP0117 LONGITUDINAL CHANGE IN THE CENTRAL NERVOUS SYSTEM PAIN RESPONSE AFTER TREATMENT WITH CERTOLIZUMAB OR PLACEBO. A POST-HOC ANALYSIS FROM THE PRECEPRA TRIAL

Abstract

Background: Tumor necrosis factor inhibitors have revolutionized the treatment of rheumatoid arthritis (RA). However, only about 50% of the patients respond well to TNF inhibitors. Therefore, markers that predict response to TNF inhibitors are valuable. Previously we demonstrated that central nervous system (CNS) response to nociceptive stimuli, measured by fMRI of the brain as blood oxygen level dependent (BOLD) signals, decreases already after 24 hours of anti-TNF administration a higher pre-treatment BOLD signal volume seems to predict clinical response to treatment with certolizumabpegol (CZP) 1,2. We therefore hypothesized that the baseline volume of BOLD signal in the CNS could predict anti-TNF treatment response. Objectives: To perform a randomized placebo controlled trial in active RA patients to test the effect of TNF inhibition on arthritis induced pain activity in the brain and to test whether patients with high-level RA-related brain activation react differently to TNF-inhibitors than patients with low-level brain activation. Methods: Adult RA patients fulfilling the 2010 ACR/EULAR classification criteria with a DAS28>3.2 receiving stable DMARD treatment for at least 3 months were eligible. Patients underwent the first fMRI at screening measuring BOLD signal upon MCP joint compression and were stratified into low (< 700 units) and high (>700 units) voxel counts. Then patients were randomized to CZP or placebo with a 2:1 ratio. The second and third fMRI were performed after 12 and 24 weeks, respectively. Control stimulation was done by measuring brain activation after non-painful finger tapping. Results: 156 RA patients with moderate-to-high disease activity participated in the study. In the finger tapping control, fMRI showed no significant changes in BOLD signal in the CZP-L and CZP-H arms, but a slight but significant decrease (p=0.043) was observed. After joint compression, the CZP-L group showed significant increase in the BOLD signal volume (p=0.043) in fMRI-2 as compared to fMRI-1 with no further significant changes. In contrast, in the CZP-H group, the BOLD signal volume significantly decreased (p=0.037) in fMRI-2 and continued to decrease further, p=0.007. No significant changes were observed in the placebo arm over time. Conclusion: TNF inhibition improves arthritis-related brain activity in the subgroup of RA patients with high baseline BOLD activity in the fMRI. (Figure Presented).