2686: Impact of autoimmunity, disease activity and treatment on the time course of bone density, microstructure and biomechanics in inflammatory arthritis.

Background/Purpose: RA and PsA follow different pathomechanisms and incur disease-specific erosive and osteoproliferative bone changes. We used data from a large cohort of RA and PsA patients who underwent high-resolution peripheral quantitative computed tomography (HR-pQCT) imaging and analyzed the impact of the primary disease, disease activity and exposure to targeted treatments on bone density, structure and biomechanics over time.

Methods: The study included RA patients fulfilling the 2010 ACR/EULAR criteria and PsA patients fulfilling the CASPAR criteria in University Hospital Erlangen with ethics approval (324_16B) and written informed consent of the participants. HR-pQCT of the distal radius and 2^nd and 3^rd metacarpal heads (MCP) was performed bi/annually measuring volumetric bone density (vBMD), trabecular and cortical microstructure, bone stiffness and failure load. Cohort follow-up started at the earliest HR-pQCT and ended at the latest. Imaging episodes between consecutive imaging dates were classified based on the type of coinciding targeted treatments and using time-weighted average DAS-28 as the disease activity exposed. HR-pQCT measurements at the beginning and end of an episode were the episode baseline and outcome values. We used mixed-effects/additive regression to model episode outcome as a function of the episode baseline, episode duration, gender, age and weight of the patients and calculated adjusted marginal means at yearly intervals by diagnosis, and at fixed DAS-28 ESR values representing continuous remission (2.0) and continuous high disease activity (6.0). btsDMARD effects were summarized using their regression coefficients comparing to no use.

Results: We included 946 radius and 916 MCP scans from 482 patients with PsA (n=175), RA- (n=86) and RA+ (n=221) over a median (IQR) follow-up of 28 months (12 to 50). Glucocorticoid exposure was similar in RA+ and RA- patients (Table-1). Measurements at radius showed a total vBMD deterioration in a similar fashion for both RA and PsA while RA+ patients show lower total vBMD at all timepoints. This separation was prominent in trabecular vBMD (Fig1A) trabecular microstructure (Fig1B) and biomechanics (Fig1C). Patients with PsA showed higher cortical vBMD and a numerically slower rate of cortical bone loss compared to RA (Fig1A) and a stable trabecular count (Fig1B). Patterns at the MCP region were grossly similar. Disease activity comparisons indicate deteriorating radial trabecular vBMD in PsA patients after 1 year under high disease activity and in all vBMDs in RA+ patients after 1 year including stiffness and failure load (Fig2A). TNF blockers, baricitinib and rituximab use improved bone density at the MCP in RA+ patients (Fig2B) , no particular pattern was discernible at the radius and for RA- and PsA.

Conclusion: Seropositive RA patients show worse bone density and microstructure across the board, which is also affected by disease activity, segregating from seronegative RA despite similar glucocorticoid use. Intervention with TNF-blockers, rituximab and baricitinib seem beneficial for bone density at the MCP, possibly indicating interception of local inflammation.