POS0896 PREDICTORS OF RESPONSE IN PATIENTS WITH NON-RADIOGRAPHIC AXIAL SPONDYLOARTHRITIS RECEIVING CERTOLIZUMAB PEGOL IN THE C-AXSPAND STUDY

Abstract

Background/Purpose: Identification of predictive clinical factors of long‐term treatment response in non‐radiographic axial spondyloarthritis (nr‐axSpA) may contribute to improved management of patients with this chronic disease. Certolizumab pegol (CZP) is currently the only FDA‐approved tumor necrosis factor inhibitor (TNFi) for treatment of nr‐axSpA.1 We aim to identify whether any demographic or baseline characteristics of nr‐axSpA patients from the C‐axSpAnd study2 are predictive of achieving a clinical response after 1 year of CZP treatment. Methods: C‐axSpAnd (NCT02552212) is a phase 3, interventional multicenter study including a completed 52‐week double‐blind, placebo‐controlled period. Full study design is reported elsewhere.2 Multivariate stepwise logistic regression analysis was used to identify predictors of response for the primary efficacy variable (ASDAS ‐ major improvement [ASDAS‐MI] at Week 52) and the main secondary efficacy variable (ASAS40 at Week 52) in patients randomized to CZP 200 mg every 2 weeks (Q2W). Predictive factors used in the model included demographic and baseline characteristics, and clinical outcomes at Week 12. A p value ≤0.05 was required for forward selection into the model and p=0.1 for backward elimination from the model. Non‐responder imputation was used to account for missing data or values collected after switching to open‐label treatment. A sensitivity analysis was conducted to account for patients who had changes in their non‐biologic background medication during the 52‐week placebo‐controlled phase. Results: 159/317 patients were randomized to CZP 200 mg Q2W and 158/317 to placebo. Predictive factors identified for Week 52 ASDAS‐MI in the CZP‐treated patients included those who were positive for both presence of sacroiliitis on MRI (MRI+) and human leukocyte antigen (HLA)‐B27 (HLA+), those with a higher BASDAI at baseline, and those with a larger Week 12 improvement in ASDAS (Table 1). For ASAS40 response, MRI+/HLA‐B27+ was also identified as a predictor of Week 52 response, along with a lower baseline BASMI and larger Week 12 improvements in PtGADA and ASQoL and (Table 1). Sensitivity analysis identified the same predictors for ASDAS‐MI and ASAS40, with the exception of change from baseline in PtGADA as a predictor of ASAS40. Sensitivity analysis also identified achievement of Week 12 ASAS40 as a predictor of Week 52 ASAS40. In placebo‐treated patients, no meaningful predictors of response at Week 52 were identified. Conclusion: Presence of sacroiliitis on MRI and HLA‐B27 positivity were identified as consistent predictors of Week 52 response (ASDAS‐MI and ASAS40) in nr‐axSpA patients treated with CZP. To our knowledge, this is the first report from an interventional 52‐week placebo‐controlled study in nr‐axSpA to identify objective clinical features, particularly the presence of sacroiliac joint inflammation, as being predictive of response.