Background Fat metaplasia in the sacroiliac joints (SIJ) may be an early manifestation of axial SpA. It has been shown to follow resolution of bone marrow edema and treatment group differences in fat metaplasia are evident by 24 weeks in anti-TNF trials. Fat metaplasia in the spine is also associated with development of new syndesmophytes.
Objectives To test the hypothesis that fat metaplasia in the SIJ is a lead indicator of new bone formation in the spine of patients with axial SpA.
Methods We used the SPARCC SIJ Structural Score (SSS) method to assess fat metaplasia (FAT), erosion (ER), backfill (BF), and ankylosis (ANK) in the SIJ. Two readers and an adjudicator independently assessed 169 pairs of MRI scans blinded to time point (baseline, 2 years) from a prospective cohort receiving either standard (n=73) or anti-TNF (n=96) therapies, mean follow up 2.3 years. Two readers and an adjudicator independently scored pairs of radiographs (baseline, 2 years) from the same patients using the mSASSS and calculated ΔmSASSS units/year. Definite radiographic progression was defined as ≥2 mSASSS units/year over ≥2 years. Correlations between radiographic, MRI, and demographic parameters were conducted using Spearman's rho and progressor/non-progressor groups compared using the unpaired t-test. Univariate and multivariate regression analyses focused on identifying significant MRI associations with mSASSS status and progression.
Results Radiographic progression (yes/no) was observed in 78 (53.8%) and definite progression in 27 (16%), mean (SD) change in mSASSS was 2.0 (3.6). The following variables were significantly higher in the progressor group: age (p=0.0007), disease duration (p=0.03), baseline mSASSS (p=0.002), SSS FAT score (p=0.02), SSS BF score (p=0.04). Significant correlations with baseline mSASSS were observed for age (r=0.63, p<0.0001), disease duration (r=0.57, p<0.0001), SSS FAT score (r=0.16, p=0.03), SSS ER (r=-0.46, p<0.0001), SSS ANK (r=0.49, p<0.0001), CRP (r=0.20, p=0.01), ASDAS (r=0.28, p=0.0004). Significant correlations with mSASSS progression were age (r=0.26,p=0.0008), disease duration (r=0.18,p=0.02), baseline mSASSS (r=0.29, p=0.0001), SSS FAT score (r=0.19, p=0.01), SSS score for ER (r=-0.16, p=0.04). Age, sex, disease duration, ASDAS, SSS FAT score, SSS ER score, SSS ANK score were significantly associated with baseline mSASSS in univariate analyses, and in multivariate analysis the following were independently associated with baseline mSASSS: age (β=0.48, p<0.0001), SSS ANK score (β=1.11, p<0.0001), SSS FAT score (β=0.50, p=0.02) (R2=0.46). For progression defined by ΔmSASSS units/year the following were significant in univariate analysis: age (p=0.003), symptom duration (p=0.04), baseline mSASSS (p=0.04), SSS FAT score (p=0.04), SSS ER score (p=0.03). In multivariate analyses with ΔmSASSS units/year as dependent variable, age (β=0.01, p=0.005) and SSS FAT score (β=0.02, p=0.03) were significant. With progression ≥2 mSASSS units/year over ≥2 years (yes/no) as dependent variable, SSS FAT score (OR=1.08 [95%CI 1.01-1.17], p=0.03) was significant.
Conclusions Fat metaplasia in the SIJ may be an important lead indicator of radiographic progression in the spine of patients with axial SpA.
Disclosure of Interest : None declared