POS1007 MRI CHANGE SCORE BUT NOT THE STATUS SCORE IS RELATED TO DISEASE ACTIVITY AND CLINICAL RESPONSE OVER TIME IN PATIENTS WITH AXIAL SPONDYLOARTHRITIS: RESULTS FROM ESTHER STUDY.

Abstract

BackgroundMagnetic resonance imaging (MRI) of the spine and sacroiliac joints (SIJ) are important tools for the diagnosis of axial SpA. However, there is very little data investigating how clinical symptoms and response to treatment relate to the degree of MRI inflammation and to its change over time.ObjectivesTo evaluate the relationship between clinical response in axial SpA patients and inflammation in MRI of spine and SIJ after 6 months of treatment.MethodsIn the ESTHER study1, a total of 76 patients with early axial SpA with a symptom duration less than 5 years, and with active inflammation on MRI in the spine and/or SIJ at baseline were randomized to be treated with etanercept (n=40) or sulfasalazine (n=36) for one year.Clinical and laboratory outcome assessments included C-reactive protein (CRP), Bath ankylosing spondylitis disease activity index (BASDAI), Bath ankylosing spondylitis functional index (BASFI) and ankylosing spondylitis disease activity score (ASDAS). The following efficacy parameters were calculated: change in CRP, BASDAI, BASFI and ASDAS between baseline and Week 24; response of the ASAS response criteria for 20 and 40% improvement in disease activity (ASAS20 and ASAS40); and ASAS criteria for partial remission at Week 24.MRI of spine and SIJ were performed at weeks 0, 24, and 48; and were scored by two radiologists, who were blinded for all clinical data, treatment arm and time point, according to the Berlin scoring system. The final osteitis score for the SIJ and for the spine was calculated as the mean score of both readers. Residual inflammation in MRI at Week 24 was defined as an osteitis score of ≥5 for SIJ and >0 for the spine (median values at Week 24).To understand the relationship between MRI and clinical outcomes, two types of models were performed: 1) a model with status scores at baseline and Week 24 for MRI and clinical outcome parameters; and 2) a model with change scores between baseline and Week 24 for MRI and clinical outcome parameters.ResultsA total of 67 patients with axial SpA were included in this analysis due to availability of MRIs at baseline and Week 24. The characteristics of patients included for this analysis were similar to the whole group of the ESTHER study. SIJ and spine osteitis score on MRI at baseline (mean±SD) was 6.9±6.1 and 1.5±2.5, respectively.The status scores showed no association at baseline, neither at Week 24, with the only exception of ASAS20 at Week 24 (Figure 1). However, changes in osteitis score of SIJ and spine were associated to clinical response outcomes between baseline and Week 24 (Figure 1). Further, we compared patients with and without residual inflammation on MRI of the spine and SIJ as defined above. There were no differences between the groups regarding clinical response and disease activity at week 24.Figure 1.Association between disease activity parameters and MRI osteitis score in spine and SIJ in patients with axial spondyloarthritis over 6 months of treatment. Each line in the figure refers to an outcome (dependent variable) in the multivariable regression analysis, which were adjusted for age, sex, symptom duration and HLA-B27 positivity. MRI, magnetic resonance imaging; SIJ, sacroiliac joints.ConclusionChange of score for osteitis in MRI of spine and SIJ was associated to disease activity in patients with axial SpA during 6 months of treatment. Presence of residual inflammation on MRI after 6 months of treatment seems to be irrelevant regarding clinical response and clinical disease activity.