Poor efficacy of TNF inhibitors in non-radiographic axial spondyloarthritis in the absence of objective signs: A bicentric retrospective study.

OBJECTIVE: To compare BASDAI 50 response rate to TNFi in axial spondyloarthritis (axSpA) depending on the presence or not of objective signs of axSpA and to look for predictive factors of TNFi efficacy.

METHODS: Patients diagnosed with axSpA according to ASAS criteria "clinical arm" and treated between January 2001 and September 2015 with TNFi were included. First group included patients with at least one objective sign such as arthritis, dactylitis, enthesitis, uveitis, inflammatory bowel disease, elevated C-reactive protein or radiological sacroiliitis, and second group included non-radiographic axSpA (nr-axSpA) patients without any objective sign corresponding to patients with inflammatory back pain and either a good response to NSAID or a SpA family history. The primary outcome was the TNFi efficacy, defined as an achievement of BASDAI 50 at 3 months. The secondary outcomes were BASDAI 50 achievement over 1 year and analysis of predictive factors of TNFi response.

RESULTS: We included 84 nr-axSpA patients without any objective signs and 84 axSpA patients with objective signs (48.2% r-axSpA and 52.8% nr-axSpA). BASDAI 50 achievement rates were significantly higher in patients with objective signs than in patients without, at 3 months (45.1% versus 13.7%, P<0.0001) and at any of the visit-time points over the first year (61.9% versus 21.4%, P<0.0001). In multivariate analysis, overweight/obesity and sacroiliitis on MRI were respectively negative and positive predictive factors of TNFi efficacy in the total population at 3 months (OR=0.32, 95%CI [0.11, 0.96], P=0.041 and OR=6.92, 95% CI (2.41, 19.8), P<0.0001, respectively).

CONCLUSION: TNFi should be used with caution in axSpA when objective signs are absent as only 13.7% of these patients were BASDAI 50 responders at 3 months.