In anti-CCP+ at-risk individuals, radiographic bone erosions are uncommon and are not associated with the development of clinical arthritis.

Objectives To investigate the prevalence, distribution and predictive value for the development of inflammatory arthritis (IA) of conventional radiography (CR) bone erosions (BE) in anti-CCP positive (CCP+) at-risk individuals with musculoskeletal (MSK) symptoms but without clinical synovitis. Methods Baseline CR of the hands and feet of 418 CCP+ at-risk individuals were analysed. The presence of US-BE was explored in the anatomical areas in which CR-BE were reported. Hands and feet CR at the time of progression were analysed in a subset of individuals who developed IA (73/123, 59.3%). Logistic regression analyses were performed to calculate the predictive value of baseline CR-BE for the development of IA in 394 CCP+ individuals with ≥1 follow-up visit. Results BE were detected in 17/418 (4.1%) CCP+ at-risk individuals (median Simple Erosions Narrowing Score-BE = 2.0, IQR: 1.0–2.0; median Sharp van der Heijde score-BE = 4.0, IQR: 3.0–8.5), most frequently in the foot joints (11/17, 64.7% individuals). A total of 123/394 (31.2%) CCP+ at-risk individuals developed IA; 7/17 (41.2%) with, and 116/377 (30.8%) without BE on CR (P  = 0.37). US-BE were found in 4/7 (57.1%) individuals with CR-BE who developed IA, but only in 1/10 (10.0%) who did not. At the time of progression, new BE were detected in 4/73 (5.5%) CCP+ individuals on repeated CR. In the regression analyses, baseline CR-BE were not predictive for the development of IA. Conclusions In CCP+ at-risk individuals with MSK symptoms, CR-detected BE are uncommon and do not predict the development of IA. [ABSTRACT FROM AUTHOR]