Relationship between disease activity status or clinical response and patient-reported outcomes in patients with non-radiographic axial spondyloarthritis: 104-week results from the randomized controlled EMBARK study

Abstract

Background: We assessed the external validity of composite indices Ankylosing Spondylitis Disease Activity Score (ASDAS), Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), and Assessment in SpondyloArthritis international Society (ASAS) 40 response (ASAS40) by evaluating the correlations between the changes in some patient reported outcomes (PROs) for patients with non-radiographic axial spondyloarthritis (nr-axSpA) and the changes in the scores of the composite indices. Methods: This was a post-hoc analysis of data from the EMBARK study in patients with nr-axSpA treated with etanercept. PROs were grouped according to ASDAS status (inactive [< 1.3], low [≥ 1.3 to < 2.1], high [≥ 2.1 to ≤3.5], and very high [> 3.5]), patient achievement of > 50% improvement in BASDAI (BASDAI50 responders), and > 40% improvement in ASAS (ASAS40 responders) at 104 weeks. Analyses were conducted on observed cases available at Week 104. Changes in PROs from Baseline to Week 104 were assessed using analysis of covariance with adjustment for baseline with linear contrast. Results: Higher ASDAS disease activity at 104 weeks was associated with lower long-term improvement from baseline in PROs (e.g., total back pain [visual analog scale, cm (95% confidence interval):-4.58 (-4.95,-4.21),-3.86 (-4.28,-3.43),-2.15 (-2.68,-1.61), and 1.30 (-0.51, 3.12) for inactive, low, high, and very high ASDAS disease activity, respectively; Multidimensional Fatigue Inventory (MFI) general fatigue:-4.77 (-5.70,-3.84),-2.96 (-4.04,-1.87),-1.00 (-2.32, 0.31), and 2.14 (-2.10, 6.38); all p < 0.001)]. BASDAI50 non-responders had less improvement in PROs from Baseline to Week 104 vs. responders (e.g., total back pain:-1.61 (-2.05,-1.18) vs.-4.43 (-4.69,-4.18); MFI general fatigue:-0.01 (-1.12, 1.09) vs.-4.30 (-4.98,-3.62); all p < 0.001). ASAS40 non-responders also had less improvement in PROs from Baseline to Week 104 vs. responders (e.g., total back pain:-1.91 (-2.30,-1.52) vs.-4.75 (-5.05,-4.46); MFI general fatigue:-0.63 (-1.56, 0.30) vs.-4.64 (-5.37,-3.91); all p < 0.001). Conclusion: Composite indices are valid for monitoring treatment response and adequately reflect treatment-related changes experienced by patients with nr-axSpA. Trial registration: ClinicalTrials.gov identifier: NCT01258738. Registered 9 December 2010.