THU0272 Which scoring method depicts spinal radiographic damage in (EARLY) axial spondyloarthritis best? five-year results from the desir cohort

Background Scores capturing spinal radiographic damage have been proposed and compared in r-axSpA. In early phases of the disease, it is still unknown how these perform. Objectives To compare the performance of different radiographic scores of the spine in patients with early axial spondyloarthritis(axSpA). Methods Five-year follow-up data (baseline, 2 and 5 years) from the DESIR cohort, including patients with early axSpA, have been used. Spine (cervical, thoracic and lumbar), sacro-iliac joints(SI), and hips were scored on radiographs centrally and independently by 3 readers (scores averaged) for the calculation of different radiographic methods (table 1). Following the OMERACT filter, scores were compared with regard to truth, discrimination (sensitivity to change and reliability) and feasibility. Baseline status scores, and 2- and 5 year change scores were calculated for each of the methods, as well as the proportion of patients with a net change (number of patients with a positive change minus number of patients with a negative change divided by all patients) above the smallest detectable change (SDC). The proportion of total variance explained by the patient (‘true variance’) was calculated for the change scores of the different instruments and their components using ANOVA, as a measure of reliability. Results In total, 699 patients (mean age 34 (SD 9) years, 47% males) had at least one radiograph available. Mean baseline and 5 year change scores were: mSASSS 0.4(SD 1.5) and 0.4 (1.8), RASSS 0.5 (1.6) and 0.6 (2.2), SASSS 0.2 (0. 7) and 0.3 (1.1), BASRI spine 1.0 (1.2) and 0.2 (0.6), BASRI spine with thoracic spine: 1.1 (1.4) and 0.3 (0.7), BASRI total 1.0 (1.3) and 0.3 (0.6) and BASRI total with thoracic spine 1.2 (1.4) and 0.3 (0.7), respectively. SDCs and proportion of 2- and 5 year change, including net change, are presented in the table 1. The mSASSS and the RASSS performed the best in terms of capturing the signal (i.e. positive change) despite the noise (i.e. negative change), which is taken into account in the net change calculation. The proportion of variance explained by the patient for the radiographic scores was highest for the mSASSS and RASSS, both for status and progression scores (e.g. 70% for mSASSS 69% for RASSS 2 year progression). However, the proportion of patient variance in the thoracic segment of the RASSS was unsatisfactory (36% for 2 year progression, compared to 54% lumbar segment and 73% cervical segment). In what concerns feasibility, all scores seemed feasible, but the thoracic segment was missing in up to 7% of the cases, thus not allowing computation of BASRI modifications to include that segment. Conclusions The existing scoring methods to assess spinal radiographic damage performed well in early phases of axSpA. The mSASSS and RASSS captured most change. There was no clear gain in additionally scoring the thoracic spine for the RASSS while an increased noise was introduced. The mSASSS remains the most sensitive and valid scoring method in axSpA, including early phases of the disease. Disclosure of Interest None declared