Background It remains to be clarified whether disease changes detected by color Doppler ultrasound (CDUS) and magnetic resonance imaging (MRI) during biological treatment reflect alterations of synovial pathology in rheumatoid arthritis (RA).

Objectives Compare the changes CDUS and MRI scores during biological (bDMARD) treatment of RA patients with the effects of the same bDMARD on inflammatory mediator release from synovial explants obtained at baseline.

Methods RA patients opted for bDMARD treatment with synovial hypertrophy in hand joints were eligible. CDUS and MRI and subsequent synovectomy by a needle athroscopic procedure was performed on the hand joint with most pronounced ultrasound pathology. Patients were re-evaluated with imaging after a minimum three months bDMARD treatment. Synovectomies were performed in ≤6 positions. Explant cultures, were established from the synovectomy specimens at baseline and incubated for 2 weeks with the bDMARD prescribed for the patient and isotype control (10μg/mL). CDUS activity was defined as the systolic color pixel/gray scale ratio (CFmax). MRI outcomes were the RA MRI score (RAMRIS) components for synovitis (0-3), bone marrow edema (BMI, 0-3) and erosion (0-10). Synovial mediator release of interleukin 6 and 8 (IL-6 and IL-8), monocyte chemoattractant protein 1 (MCP-1) and macrophage inflammatory protein 1 beta (MIP-1b) were detected by multiplex immuno assays (Myriad RBM™). Heterophilic antibody interference was blocked. Mixed linear models were applied for the statistical analyses for the associations of changes in imaging and the fold change (72 hours/two weeks culture) in mediator concentration. Approximative correlation coefficients (approx.rho) were calculated using Spearman rank correlation. P-values <0.05 were considered significant.

Results 16 RA patients initiated bDMARD treatment. Median follow up time was 7 months (inter quartile range 7-11 months). Fifteen patients had matching CDUS and explant data while 11 patients had matching MRI and explant data, respectively. The release of IL-6 from bDMARD treated explant samples were statistically significantly associated to changes in CFmax (approx. rho =0.52, P=0.04) and RAMRIS BME score (approx.rho=0.56, P=0.03), while MCP-1 production was only significantly associated to RAMRIS BME (approx.rho=0.49, P=0.01). There were no significant associations to change in RAMRIS synovitis or erosion score. Analysis of IL-8 and MIP-1b was discarded due to a large number of samples outside assay detection limits.

Conclusions In RA, changes in CDUS and MRI pathology during bDMARD treatment reflected the changes in cytokine production induced by the same biologic in synovial biopsy cultures. This substantiates the use of imaging parameters as markers of local disease activity. The lack of associations with explant activity and changes in MRI synovitis and erosions may be due to the synovectomy procedure and relative short follow up, respectively.

Acknowledgements The OAK Foundation, Novo Nordisk A/S and The Danish Agency for Science, Technology and Innovation for unrestricted grants.

Disclosure of Interest M. Andersen Employee of: Novo Nordisk A/S, M. Boesen: None declared, K. Ellegaard: None declared, K. Söderström Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, N. Søe: None declared, P. Spee Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, U. Mørch Employee of: Data collected during employment at Novo Nordisk A/S, S. Torp-Pedersen: None declared, E. Bartels: None declared, B. Danneskiold-Samsøe: None declared, L. Karlsson Shareholder of: Novo Nordisk A/S, Employee of: Novo Nordisk A/S, H. Bliddal: None declared