Op0148 spatiotemporal dynamics of bone loss before and after the onset of rheumatoid arthritis

Background: Rheumatoid Arthritis (RA) is preceded by a clinically silent pre-phase characterized by autoimmunity against anti-modified protein antibodies including anti-citrullinated protein antibodies (ACPA). At this pre-stage patients already experience significant loss of volumetric peripheral bone mineral density (vBMD) compared to healthy controls measured by high-resolution peripheral quantitative computed tomography (HR-pQCT) (1-2). However, the longitudinal course of vBMD changes during the preclinical phase, after diagnosis, and its association with time to disease onset have not been investigated. Objectives: To longitudinally characterize the changes of metacarpal and radial vBMD before and after the clinical onset of RA and its association with time to onset of arthritis. Methods: To explore the development of arthritis, we initiated a RA-at-risk cohort in 2011. (Ethics 334_16B). This prospective cohort includes adults positive for CCP-AB with or without musculoskeletal symptoms, excluding arthritis. Participants are regularly followed with clinical examination and HR-pQCT imaging of the MCP and radial bone to monitor early bone changes. HR-pQCT images with low motion grade artefacts were analyzed to obtain the total (D100), cortical (DComp) and trabecular (DTrab) vBMD (D100) in mg HA cm3. We descriptively analyzed the vBMD time course in patients who developed RA by fitting regression curves separately for the pre-clinical and clinical periods and estimated time-conditional marginal mean VBMDs for the 5-year peri-RA period. We analyzed time to diagnosis of clinical RA defined by the 2010 ACR/EULAR classification criteria using Cox regression models. Hazard ratios indicate the relative risk of clinical disease onset associated with 1 standard deviation reduction in bone density. Results: 130 subjects (mean [SD] age 47.0 [12.2], 89 female [68%]) between 2011 and 2020 were analyzed. Median (IQR) follow-up duration for the cohort was 18.6 (4.6-47.6) months. Participants underwent 233 HR-pQCT scans and 58 (45%) underwent 2 to 6 scans with a median interval of 16.2 (12.2-21.2) months. 49 (38%) patients who developed RA had a pre-diagnosis follow-up of 4.1 (2.5-13.4) months and post-diagnosis follow-up of 22.0 (8.8-38.9) months. The time course of scaled bone mineral densities depicted in Figure 1A suggest that bone density around the MCP joints deteriorate in the preclinical phase of RA, which is mostly prominent in the trabecular bone. Modelling (Figure 1B) suggests that trabecular bone loss around the MCP joints has a constant pace regardless of the clinical status. Whereas the radial bone densities are relatively stable in the preclinical phase and show a reduction after the clinical onset of RA. Age and sex adjusted hazard ratios (95%CI) for the risk of RA clinical onset were 1.52 (1.03 to 2.25) for radius D100 and 1.66 (1.07 to 2.55) for radius DComp (Table-1). Conclusion: Metacarpal bone showed a constant decline that started already in the pre-phase of RA and continued after its clinical onset. In contrast, bone loss in the radius was not observed in the pre-phase but started at onset of RA. Low radial vBMD in the pre-clinical phase, however, was associated with a higher risk of RA onset. These findings suggest different spatiotemporal dynamics of bone loss before and after RA onset References: [1]Kleyer A. et. al. Ann Rheum Dis. 2014, 73:854-60 [2]Simon D. et. al. Ann Rheum Dis. 2020, doi:10.1002/art.41229 Disclosure of Interests: None declared