Background Fatty changes (FC) on T1-weighted magnetic resonance images (MRI) have been described in the spine of patients with ankylosing spondylitis (AS). Such lesions were recently found to be related to syndesmophyte development in AS patients treated with anti-TNF agents. While TNF blockers have been shown to reduce inflammatory changes seen on STIR MRI sequences, their effect on the course of FC has not been studied in detail to date.

Objectives To study the course of FC in patients with AS under anti-TNF therapy and analyse which clinical parameters are associated with the occurrence of FC.

Methods MRI of AS patients participating in the EASIC Registry who had received infliximab for 2 years were blindly for the time point of examination. Presence or absence of FC was recorded for all vertebral edges (VEs) at baseline, after 24 weeks and after 2 years of treatment. Spinal radiographs at baseline and 2 years were scored by the mSASSS. Relative risk (RR) calculations were performed based on generalized estimation equations (GEE model). Poisson variation and X2 test was used to compare MRI findings between time points.

Results Complete sets of MRI were available of 73 patients (mean age 40.5±10.5 years, 86.3% male, HLA-B27+: 83.6%, mean BASDAI 6.5±1.4, mean disease duration 10.0±8.4 years) resulting in a total of 1,948 VEs at each time point. At baseline, FC were found in 619 VEs (31.8%). After 24 weeks of infliximab therapy, regression of FC was seen in 19 VEs (3.1%), while new FC had appeared in 133/1,329 VEs (10.0%). After 2 years, regression of FC was observed in 35/619 VEs (5.7%) and new FC were seen in 215 VEs (16.2%), while 4/1,948 VEs (0.3%) showed both regression and development to FDL at different time points.

Increased inflammatory activity at baseline was significantly associated with development of FC: 34% of VEs with inflammation vs. 12.9% of VEs without (p<0.0001). Baseline mSASSS scores increased the risk of FC: RR 1.03, 95% CI: 1.01-1.04 (p=0.0003), while a major treatment response (50% BASDAI improvement) was associated with lower risk: RR=0.85, 95% CI: 0.76-0.95 (p=0.005). Being female lowered the risk for FC by 22%: RR: 0.88, 95% CI: 0.59-1.32, p=0.544), and younger age also had some influence: RR: 0.97, 95% CI: 0.93-1.00 (p=0.0327).

References This study confirms that FC develop in AS patients treated with TNF blockers and that inflammation may be one source of that. However, the most important finding of this analysis is that a good response to TNF blockers was associated with significantly lower risk to develop FC. Whether this leads to a decrease in new bone formation over time remains to be seen.

Disclosure of Interest None Declared