Background Fatty lesions (FLs) of the bone marrow in the axial skeleton (sacroiliac joints – SIJ, and spine) on magnetic resonance imaging (MRI) are considered nowadays as earliest post-inflammatory changes preceding new bone formation in axial spondyloarthritis (axSpA). It has been shown in several trials with tumour necrosis factor (TNF) α inhibitors that resolution of inflammation under anti-TNF therapy is associated with an increase of a FL score [1, 2]. This raised concerns that TNFα blockers might therefore promote the process of new bone formation in axSpA.
Objectives To investigate the difference in FL formation rates in patients treated with the TNF inhibitor infliximab (IFX) – Remicade – added to naproxen (NPX) as compared to NPX alone given over 28 weeks in patients with early axSpA in the Part I of the Infliximab As First Line Therapy in Patients with Early Active Axial Spondyloarthritis Trial (INFAST).
Methods Part I of the INFAST study was a double-blind, randomized controlled trial of IFX in biologic-naïve patients 18–48 years of age with early (<3 years symptom duration) active axSpA with signs of active sacroiliitis on MRI. A total of 158 patients were randomized (2:1) to receive 28 weeks of treatment with either intravenous IFX 5 mg/kg (weeks 0, 2, 6, 12, 18, and 24) + NPX 1000 mg/d (n=106) or intravenous PBO+NPX 1000 mg/d (n=52). MRIs of the SIJ and of the spine were performed at baseline and week 28. Images were scored according to the Berlin scoring system for active inflammation and for FL.
Results MRI data were available for 156 patients (n=105 IFX+NPX, n=51 NPX alone) that corresponds to 3588 potentially assessable vertebral units (VU) and 1248 potentially assessable SIJ quadrants in total. At baseline, a higher number of SIJ quadrants (52.5% vs. 43.5%, p=0.003) and VUs (17.1% vs. 13.7%, p=0.007) were affected by osteitis in the NPX group as compared to IFX+NPX subgroup. A complete resolution of osteitis in both SIJ and in the spine at week 28 was more frequent in the IFX+NPX group as compared to the NPX group (table).
There was nearly equal prevalence of FLs at baseline in both study groups: 62.4% vs. 65.3% of SIJ quadrants and 18.6% vs. 21.1% of VUs were affected in the IFX+NPX and NPX groups, respectively. The majority of new FLs at week 28 occurred in the areas affected by inflammation at baseline following resolution of inflammation (82.7% in the SIJ and 69.8% in the spine) as compared to persisting inflammation (14.3% and 11.5%, respectively) or absence of inflammation (3.1% and 18.8%, respectively). Importantly, there were no difference in the rates of new FLs occurred after resolution of inflammation in both treatment groups (table).
Conclusions Effective anti-inflammatory treatment of axSpA in this study was associated with resolution of inflammation with subsequent formation of new FLs in the SIJ and in the spine. Formation of new FLs after resolution of inflammation was independent of the treatment arm suggesting that FL formation after resolution of inflammation is possibly a universal pathogenetic mechanism in axSpA and not a direct effect of anti-TNF therapy.
Song IH, et al. Ann Rheum Dis 2011;70:1257-63.
Maksymowych WP, et al. Ann Rheum Dis 2013;72:23-8.
Acknowledgements The MRI analysis substudy was supported by an unrestricted research grant from MSD.
Disclosure of Interest D. Poddubnyy Consultant for: AbbVie, MSD, Pfizer, UCB, Speakers bureau: AbbVie, MSD, Pfizer, UCB, J. Sieper Grant/research support from: Abbvie, MSD, Pfizer, Consultant for: AbbVie, MSD, Pfizer, UCB, Speakers bureau: AbbVie, MSD, Pfizer, UCB