Ankylosing spondylitis (AS) is characterized by a unique and complex "bone paradox," in which systemic trabecular bone loss coexists with pathological new bone formation and syndesmophyte growth. This "bone paradox"—the coexistence of systemic osteoporosis and local syndesmophyte formation—poses a significant challenge in clinical management, as traditional imaging often fails to accurately reflect trabecular integrity due to overlapping pathological new bone. Whether the biologic control of systemic inflammation directly translates into the recovery of vertebral trabecular bone remains a subject of intense debate. This study aims to evaluate the changes in thoracolumbar trabecular volumetric bone mineral density (vBMD) in biologic-naïve AS patients treated with secukinumab or adalimumab, utilizing quantitative computed tomography (QCT) to avoid the diagnostic interference typically associated with dual-energy X-ray absorptiometry (DXA) in the presence of syndesmophytes.

We performed a retrospective cohort study of biologic-naïve adults with AS who received secukinumab (n=30) or adalimumab (n=22) continuously for 12 months. Systemic inflammation was assessed using C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and composite blood-derived indices including systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and platelet-to-albumin ratio (PAR). Trabecular volumetric bone mineral density (vBMD) from T1 to L1 was quantified by QCT at baseline and at 1 year. The primary outcome was change in mean T1–L1 vBMD; secondary outcomes included changes in inflammatory indices and their associations with vBMD changes.

After 12 months, both secukinumab and adalimumab significantly reduced CRP and ESR (all P<0.05). Secukinumab was additionally associated with significant decreases in SII, PLR, and PAR (all P≤0.01), whereas adalimumab significantly reduced PAR (P=0.006). Despite inflammatory improvement, mean T1–L1 trabecular vBMD did not change significantly in either group (secukinumab: −1.08 mg/cm³, P=0.56; adalimumab: +1.16 mg/cm³, P=0.68), and the between-group difference in vBMD change was not significant (P=0.50). Changes in CRP, ESR, SII, NLR, PLR, or PAR were not associated with L1 vBMD change in correlation or linear regression analyses (all P>0.05).

In biologic-naïve AS patients, 1 year of secukinumab or adalimumab markedly suppressed systemic inflammation but did not yield measurable improvement in thoracolumbar trabecular vBMD by QCT, indicating a short-term dissociation between inflammatory control and trabecular bone restoration. Bone health should be assessed and managed as an independent treatment target in AS.