Background In the present study, we evaluated relationships between serum biomarkers and clinical/magnetic resonance imaging (MRI) findings in golimumab-treated patients with ankylosing spondylitis. Methods In the GO- RAISE study, 356 patients with ankylosing spondylitis randomly received either placebo (n = 78) or golimumab 50 mg or 100 mg (n = 278) injections every 4 weeks through week 24 (placebo-controlled); patients continuing GO-RAISE received golimumab through week 252. Up to 139/125 patients had sera collected for biomarkers/serial spine MRI scans (sagittal plane, 1.5-T scanner). Two blinded readers employed modified ankylosing spondylitis spine magnetic resonance imaging score for activity (ASspiMRI-a) and ankylosing spondylitis spine magnetic resonance imaging score for chronicity. Spearman correlations (r s) were assessed between serum biomarkers (n = 73) and Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), C-reactive-protein (CRP)-based Ankylosing Spondylitis Disease Activity Score (ASDAS), modified Stokes Ankylosing Spondylitis Spine Score (mSASSS), and ASspiMRI scores. Serum biomarkers predicting postbaseline spinal fatty lesion development and inflammation were analyzed by logistic regression. Results Significant, moderately strong correlations were observed between baseline inflammatory markers interleukin (IL)-6, intracellular adhesion molecule-1, complement component 3 (C3), CRP, haptoglobin, and serum amyloid-P and baseline ASDAS (r s = 0.39–0.66, p ≤ 0.01). Only baseline leptin significantly correlated with ASDAS improvement at week 104 (r s = 0.55, p = 0.040), and only baseline IL-6 significantly predicted mSASSS week 104 change (β = 0.236, SE = 0.073, p = 0.002, model R 2 = 0.093). By logistic regression, baseline leptin, C3, and tissue inhibitor of metalloproteinase (TIMP)-1 correlated with new fatty lesions per spinal MRI at week 14 and week 104 (both p < 0.01). Changes in serum C3 levels at week 4 (r s = 0.55, p = 0.001) and week 14 (r s = 0.49, p = 0.040) significantly correlated with BASDAI improvement at week 14. Baseline IL-6 and TIMP-1 (r s = −0.63, −0.67; p < 0.05) and reductions at week 4 in IL-6 (r s = 0.61, p < 0.05) and C3 (r s = 0.72; p < 0.05) significantly correlated with week 14 ASspiMRI-a improvement. Conclusions Extensive serum biomarker multiparametric analyses in golimumab-treated patients with ankylosing spondylitis demonstrated few correlations with disease activity or MRI changes; IL-6 weakly correlated with radiographic progression.
