Background The EMBARK trial of etanercept (ETN) (NCT01258738)1 has demonstrated the long-term efficacy of ETN in patients with non-radiographic axial spondyloarthritis (nr-axSpA) and shown correlations between decreased inflammation and clinical outcomes.2
Objectives To determine which baseline (BL) characteristics are associated with clinical improvements after 48 weeks of treatment with ETN in patients with nr-axSpA.
Methods This post hoc analysis was performed on data from patients with nr-axSpA enrolled in the 92-week open-label phase of the EMBARK trial. The primary analysis population was the EMBARK modified intent-to-treat (mITT) population (patients who had ≥1 dose of ETN, had ≥1 on-therapy evaluation, and met the Assessment in Ankylosing Spondylitis classification criteria for axSpA). Patients' SpondyloArthritis Research Consortium of Canada (SPARCC) sacroiliac joint (SIJ) scores and Ankylosing Spondylitis Disease Activity Scores (ASDAS) were recorded, and patients were divided into 4 quadrants (Q1–4) based on the minimally important change (MIC) for each instrument (SPARCC SIJ score, change from BL [cfb] of ≥2.5; ASDAS, cfb ≥1.1). Patients in Q1 achieved MIC in SPARCC SIJ but did not achieve MIC in ASDAS. Q2 did not achieve MIC in SPARCC SIJ and did not achieve MIC in ASDAS. Q3 did not achieve MIC in SPARCC SIJ but achieved MIC in ASDAS. Q4 achieved MIC in SPARCC SIJ and achieved MIC in ASDAS. Missing data were imputed using last observation carried forward. P-values were obtained using the Chi-squared test (categorical variables) and the Kruskal-Wallis test (continuous variables) in order to compare BL characteristics across improvement status. Multivariable multinomial logistic regression modelling was used to calculate which BL characteristics were associated with clinical improvements when adjusted for all other BL characteristics.
Results At screening, male patients had significantly higher median SPARCC SIJ scores compared with female patients (4.83 [range 0.0–48.0] vs 1.5 [0.0–32.0], P<0.0001). At wk48, 69.1% of 194 patients achieved MIC in either SPARCC SIJ scores, ASDAS or both (Q1/3/4, Figure). There was a statistically significant difference in the number of male versus female patients who achieved MIC in both scores (Q4), compared with patients who achieved MIC in only one or neither (Q1–3, Figure) (P<0.0001). However, multinomial logistic regression modelling showed that age (P=0.0358), BL ASDAS (P<0.0001), and screening SPARCC Spine (P=0.0358) and SPARCC SIJ (P<0.0001) scores, not gender, were significantly associated with achievement of MIC in both SPARCC SIJ and ASDAS at wk48 for the total population (after adjustment for the effects of other variables).
Conclusions Although BL summaries showed a significant difference in gender across quadrants, the BL SPARCC SIJ score was strongly correlated with gender (male patients had more inflammation). When both were included in a multivariable model, gender was non-significant. Only age, BL ASDAS, and screening SPARCC SIJ and Spine scores were associated with MIC after 48 weeks of etanercept treatment.
Dougados M, et al. Arthritis Rheumatol. 2014;66:2091–102.
Maksymowych WP, et al. Ann Rheum Dis. 2016;75:1328–35.
Disclosure of Interest K. de Vlam Consultant for: Abbott, Celgene, Janssen, Pfizer, and UCB, Speakers bureau: Abbott, Celgene, Janssen, Pfizer, and UCB, A. Burden Employee of: Quanticate, M. A. Dilleen Shareholder of: Pfizer, Employee of: Pfizer, C. Boone Shareholder of: Pfizer, Employee of: Pfizer