POS0526 HIGH-RESOLUTION PERIPHERAL QUANTITATIVE COMPUTED TOMOGRAPHY AND PREDICTION OF EROSIVE PROGRESSION AS ASSESSED BY THE GOLD STANDARD CONVENTIONAL RADIOGRAPHY IN ESTABLISHED RHEUMATOID ARTHRITIS: A 1-YEAR COHORT STUDY

Abstract

BackgroundHigh-Resolution peripheral Quantitative Computed Tomography (HR-pQCT) is a promising imaging technique for assessing erosive disease in rheumatoid arthritis (RA). We have shown that the diagnostic accuracy in detecting erosive disease of HR-pQCT of two joints corresponds to conventional radiography (CR) of 44 joints in established RA (1). No data assessing HR-pQCT as a tool for predicting erosive radiographic progression in a large cohort of established RA has been published.ObjectivesTo assess the association between erosive disease identified using either CR or HR-pQCT at baseline and erosive progression using CR during 1-year follow-up in established RA.MethodsThis observational cohort study included 220 patients with RA (disease duration ≥ 5 yrs) at the Department of Rheumatology, Aarhus University Hospital, between Mar. 2018 and Oct. 2020. All participants had demographic information collected, together with HR-pQCT at baseline and CR made at baseline and 1-year follow-up. Erosive assessment using HR-pQCT was performed at the second and third metacarpophalangeal (MCP) joint (2). The Sharp/van der Heijde score (SHS) method was used to assess erosive status and erosive progression using CR of hands, wrists and feet (3).ResultsIn total, 220 participants with RA (median age 66 yrs, interquartile range (IQR) 57-72; 71% women) were included and 212 completed their follow-up visit in Dec. 2021 (dropout n=8 (4 %)). During follow-up (median 1 yrs, range 0.7-1.4), erosive progression on CR was detected in 24 of 211 (11%) (missing baseline CR, n=1). Characteristics of the study population by erosive progression are shown in Table 1. In unadjusted regression analyses, number of erosions on HR-pQCT at baseline and erosion score (per 10) using CR at baseline were associated with erosive progression (yes/no) on CR at 1-year (Odds ratio (OR) 1.1; 95%CI 1.0-1.1; p<.05 and OR 1.1; 95%CI 1.0-1.2; p<.01). After adjusting for age and sex, erosion score remained associated with erosive progression (OR 1.1; 95%CI 1.1-1.2; p<.01), but number of erosions did not (OR 1.0; 95%CI 1.0-1.1; p=.07).Table 1.Demographics at baselineProgression on CRNo progression on CRP valueNo24/211 (11)187/211 (89)Age, yrs66 (57-73)66 (57-71).80Women15/24 (63)134/187 (72).35Disease Duration, yrs20 (8-28)15 (9-23).44DAS28-CRP1.7 (1.5-2.5)1.8 (1.5-2.5).53Erosive on CR24/24 (100)141/187 (75)< .01Erosive 2nd/3rdMCP on CR*17/23 (74)62/185 (34)< .001Erosive on HR-pQCT22/23 (96)153/185 (83).14Erosion score (per 10)4 (2-7)1 (0-3)< .0001Erosion number10 (5-17)3 (1-10)< .01Total erosion volume, mm3128 (19-201)26 (2-111)< .05Data are no. (%) or median with interquartile range. P-values: Mann-Whitney U or Fisher’s exact test. Missing HR-pQCT, n=3. CR = conventional radiography, DAS28-CRP = disease activity score of 28 joints based on C-reactive protein, HR-pQCT = high-resolution peripheral quantitative computed tomography* Same hand as HR-pQCT scannedConclusionIn established RA, baseline CR of 44 joints is superior to baseline HR-pQCT of two joints in identifying individuals at risk of erosive progression on CR. This underlines the need for a head-to-head comparison between progression identified by HR-pQCT and CR to fully assess the clinical utility of HR-pQCT in predicting erosive progression compared to CR.