Erosive progression by high-resolution peripheral quantitative computed tomography and conventional radiography in rheumatoid arthritis

Abstract Objectives To investigate the diagnostic accuracy of high-resolution peripheral quantitative computed tomography (HR-pQCT) to assess erosive progression during 1 year compared with conventional radiography (CR) in rheumatoid arthritis (RA). Methods This prospective study included 359 patients with RA (disease duration ≥5 years) between March 2018 and October 2020. HR-pQCT and CR were obtained at inclusion and after 1 year. Erosive assessment was performed at two metacarpophalangeal joints of the dominant hand using HR-pQCT, and progression was defined as an increase in erosion number ≥1 or an increase in erosive volume > least significant change. CR of hands, wrists and feet was evaluated using Sharp/van der Heijde scores and erosive progression was defined as a 1.1-point increase in erosion score according to the smallest detectable change. Results In paired analyses (n = 310), erosive progression was identified in 30 patients using CR and in 40 patients using HR-pQCT. In the 40 patients with erosive progression on HR-pQCT, progression was not identified by CR in 33 patients. Adding HR-pQCT to CR doubled the proportion of patients identified with progression from 30 (10%) to 63 (20%) patients. Using CR as the reference, the sensitivity of HR-pQCT for identifying erosive progression was 23.3% (95% CI: 9.9, 42.3%) and the specificity was 88.2% (95% CI: 83.8, 91.7%). Conclusion A substantial proportion of patients with erosive progression are overlooked using CR only to monitor erosive progression. Adding high-resolution peripheral CT to CR doubles the proportion of patients who may benefit from individualized therapy targeting erosive progression in RA.