LAUSR

Background ACPA +individuals with non-specific MSK symptoms are at risk of inflammatory arthritis (IA) and may benefit from early intervention. Clinical, serological and US markers have previously been assessed to determine risk of progression.1 Objectives Evaluate the value of MR and US imaging in characterising and quantifying risk in a large ACPA +cohort. Methods Eligible ACPA +individuals without clinical synovitis had gadolinium enhanced 3.0 T MRI of the dominant hand and wrist. Images were scored by 2 radiologists for synovitis, bone marrow oedema (BME), erosions and tenosynovitis (TSV) according to OMERACT RAMRIS. Joint counts for each abnormality at each joint were corrected for age using a healthy controls reference range.2 US of the same regions were scored using OMERACT definitions. Maximum MRI and US abnormality scores observed per patient across all joints scored were dichotomised <2,≥2. Potential associations between baseline US (greyscale (GS) and powerDoppler (PD)) and MRI findings and i) progression to IA and ii) development of clinical synovitis within a joint were identified using Cox and penalised regression. Results Imaging of 98 individuals (mean age 47, 69% female) was available. 30% (29/98) progressed to IA. Median time to progression was 31 weeks (IQR 24, 67). BME and erosions scores≥2 were reported in 10%, preferential location to the carpal bones/wrist joints. Synovitis score ≥2 was present in 9%, preferential location at MCP5 and radial carpal/intercarpal joints. TSV was the most frequent reported abnormality with 22% scoring ≥2, 40% scoring 1. US GS and PD scores≥2 were reported in 23% and 9% respectively. The unadjusted analysis HRs for all imaging abnormalities were high, indicating potential association with risk of progression. Controlling for variables, MRI TSV was associated with time to IA with an increased HR. US GS and PD were also independently associated with time to progression and confirmed on penalised regression, table 1. At the joint level MRI TSV, BME and US GS were associated with the risk of progression to clinical synovitis, HR=7.03 p<0.001, HR 4.22 p=0.076 and HR 8.04 p<0.001 respectively.Abstract OP0042 – Table 1 Patient-level Cox regression proportional hazard modelling of associations between maximum observed score per patient for baseline MRI abnormalities and time to IA (n=95) Abnormality No IA% (n=66) IA% (n=29) Unadjusted HR (90% CI),P value Adjusted HR (90% CI),P value Penalised HR Small joint tenderness 44(29) 55(16) 1.60 (0.87,2.97), p=0.207 1.20 (0.58,2.46), p=0.679 1 RF and/or ACPA>3ULN 83(55) 93(27) 2.14 (0.64,7.16), p=0.301 1.02 (0.27,3.80), p=0.981 1 EMS≥30 min 27(18) 45(13) 2.00 (1.08,3.71), p=0.064 1.52 (0.69,3.37), p=0.384 1 US PD≥2 2 (1) 28(8) 7.21 (3.62,14.36), p<0.001 5.09 (1.93,13.44), p=0.006 4.37 US GS≥2 11(7) 52(15) 4.97 (2.69,9.19), p<0.001 2.69 (1.14,6.34), p=0.059 2.17 MRI erosion≥2 6 (4) 17(5) 2.15 (0.96,4.82), p=0.120 0.59 (0.20,1.76), p=0.431 1 MRI BME≥2 8 (5) 17(5) 2.30 (1.01,5.23), p=0.097 1.55 (0.56,4.27), p=0.482 1 MRI synovitis≥2 3 (2) 24(7) 4.22 (2.03,8.75), p=0.001 1.08 (0.46,2.54), p=0.881 1 MRI TSV≥2 14(9) 41(12) 3.51 (1.89,6.55), p=0.001 4.02 (1.91,8.44), p=0.002 3.16 Conclusions ACPA +at risk individuals have features on imaging which assists prediction of development to IA. MRI TSV provides additional predictive ability over and above the clinical and US variables. References [1] Rakieh C. 2014. [2] Mangnus L. 2016. Disclosure of Interest None declared