Background TNF inhibitors (TNFi) are effective treatment in radiographic and non-radiographic axial spondyloarthritis (r-axSpA and nr-axSpA). Nr-axSpA have lower response rate in case of absence of sacroiliitis on magnetic resonance imaging (MRI) or elevated C-reactive protein (CRP) but TNFi efficacy in the absence of clinical, biological or radiological inflammatory signs is unknown due to the low number of patients included in previous studies.
Objectives To compare TNFi response rate in axSpA depending on the presence or not of objective signs of inflammation.
Methods Nx-axSpA patients fulfilling ASAS 2009 criteria, without objective signs of inflammation defined as the absence at the time of TNFi initiation of radiographic and MRI sacroiliitis, elevated CRP level, dactylitis, anterior uveitis and inflammatory bowel disease, treated for at least 3 months between January 2001 and September 2015 were included in this retrospective multicentric study. They were matched 1:1, by age, sex and type of TNFi, with axSpA patients with at least one objective signs of inflammation. Efficacy was defined as an achievement of 50% improvement of the initial Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) at 3, 6 or 12 months and both groups were compared. Predictive factors of response were investigated in the two groups.
Results We included 84 nx-axSpA patients without objective signs of inflammation and 84 ax-SpA patients with objective signs. Patients were mainly women (76.2%) with a mean ± standard deviation age of 42 ± 11 years. BASDAI 50 achievement rate, at 3, 6 or 12 months, was significantly higher in patients with objective signs than in patients without (61.9% versus 21.4% considering all time point for evaluation, p<0,0001). The BASDAI mean improvement was also significantly higher in patients with objective signs at 3 (-24.3 ± 19.8 versus -9.7 ± 15.9; p<0.0001), 6 (-27.7 ± 22.6 versus -8.8 ± 18.3; p<0.0001) and 12 months (-26.5 ± 22.5 versus -13.7 ± 24.3; p=0.01). The only significant predictive factor of non-response considering the two groups was a high mean body mass index (BMI) (26.2 ± 5.7 kg/m2 in non-responders versus 24.2± 4.6 kg/m2 in responders; p=0.04). This factor was especially predictive of non-response among nr-axSpA without objective signs (26.5 ± 5.6 kg/m2 in non-responders versus 22.3 ± 3.0 kg/m2; p=0.008).
Conclusions TNFi are efficient only in 20% of axSpA patients when objective signs of inflammation are absent. A high BMI is a predictive factor of non-response in axSpA in the presence or not of objective signs of inflammation.
Disclosure of Interest None declared