1433: Identifying Predictors of Difficult-to-Manage Axial Spondyloarthritis: Real-World Insights from a Taiwanese Registry Cohort

Background/Purpose: The concept of Difficult-to-Manage Axial Spondyloarthritis (D2M-ax-SpA) was recently introduced by the Assessment of Spondyloarthritis International Society (ASAS). According to this new definition, D2M-ax-SpA is characterized by (a) failure of ≥2 biologic or targeted synthetic DMARDs (b/tsDMARDs) with different mechanisms of action (MOAs), (b) persistent disease activity, and (c) symptoms considered problematic by either the patient or rheumatologist.

This study is among the first to evaluate the real-world prevalence and clinical predictors of D2M-ax-SpA using this newly proposed definition.

Methods: We conducted a retrospective cohort study of patients with axial spondyloarthritis (ax-SpA) enrolled between January 2009 and December 2021 at Taipei Veterans General Hospital. Inclusion criteria included fulfillment of the ASAS classification criteria and ≥6 months of follow-up. Patients were categorized into three groups: D2M-ax-SpA (per ASAS definition), those treated with b/tsDMARDs with only one MOA, and those who never received b/tsDMARDs. Demographic, radiographic, and laboratory data were analyzed. Group comparisons were performed using the Mann–Whitney U and Fisher’s exact tests. Univariate and multivariate logistic regression were conducted to identify predictors of D2M-ax-SpA, and ROC curve analysis was applied to assess the predictive performance of laboratory markers.

Results: Among 560 patients with ax-SpA (all HLA-B27 positive), 20 (3.6%) met the criteria for D2M-ax-SpA. Ninety-four patients received b/tsDMARDs with a single MOA, while 446 were treatment-naïve to b/tsDMARDs. Compared to the other groups, patients with D2M-ax-SpA had significantly longer disease duration (median 3890 vs. 3292 and 2034 days, p< 0.01), more frequent peripheral arthritis (p< 0.01), higher grades of sacroiliitis on plain radiography (p< 0.01), and elevated ESR and CRP levels (p< 0.01). In multivariate analysis, peripheral arthritis emerged as an independent predictor of D2M-ax-SpA (OR: 4.68, 95% CI: 1.74–12.60, p=0.0022). ROC analysis demonstrated that ESR and CRP were strong predictive markers, each with an AUC of 0.785 (p< 0.0001), and optimal thresholds identified as ESR >20 mm/hr and CRP >0.99 mg/dL.

Conclusion: This study provides the first real-world evidence on the clinical characteristics and predictors of D2M-ax-SpA using the newly proposed ASAS criteria. Peripheral arthritis and elevated inflammatory markers (ESR, CRP) were strongly associated with treatment resistance. These findings highlight the importance of early identification of high-risk patients, which may inform treatment selection and improve long-term outcomes. Further prospective studies in larger, multi-ethnic cohorts are warranted to validate these results.