Background/Purpose: Axial joint involvement commonly links psoriatic arthritis (PsA) to spondyloarthritis with widely variable prevalence (25-70%). Patients with axial PsA (AxPsA) are older than patients with ankylosing spondylitis (AS) and identifying radiographic changes related to the spondylitis may be more challenging leading to disagreement among readers. Currently, there is no universally accepted method to score AxPsA. In AS, the mSASSS system is the most widely used tool to assess the severity of spondylitis and define progression. Based on previous studies in AxPsA, the mSASSS had the highest sensitivity, specificity, and odd ratios of determining disease progression. We aimed to assess the reliability of the mSASSS scoring system between two rheumatologists and various confounders using the PsA Scoring Module for Axial Radiograph Toronto (SMART) tool developed by Biagioni et al.

Methods: A retrospective study was conducted in which radiographs from 2015-2018 in our Psoriatic Arthritis Clinic database were retrieved by an assessor (DIP) via convenience sampling and scored with the SMART tool by two rheumatologists (SET, AW). According to the SMART tool, patients are scored for the presence of osteophytes, degenerative disc disease, osteopenia, and diffuse proliferative skeletal hyperostosis (DISH). We then removed patients who had DISH recorded in our database. Discordant radiographs were adjudicated by a third independent reader (DDG, N = 11). The intra-rater reliability was reported in Intraclass Correlation Coefficient (ICC) with the balanced one-fold random model in all patients as scored by two readers (N = 119), and after adjudication and removing the DISH patients (N = 109).

Results: Patients (N = 119) with well-established PsA with and without axial disease were included. The mean age was 58.19 ± 14.15 years, the mean age at PsA diagnosis was 36.74 ± 12.59 years. 17 patients had inflammatory back pain and 34 had mechanical back pain. The intra-rater ICC between readers of the total mSASSS, cervical mSASSS, and lumbar mSASSS were 0.25, 0.23, and 0.20 respectively. With DISH exclusion, the ICC were 0.89, 0.93, and 0.40 respectively.

Conclusion: Initial results demonstrated poor agreement between readers in the presence of DISH and other confounders. With removal of radiographs with DISH and adjudication of discordant readings, the total mSASSS demonstrated moderate intra-reader reliability. Therefore, the presence of DISH leads to misinterpretation of mSASSS scores and should be considered when reading radiographs in AxPsA.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/validation-of-the-modified-stokes-ankylosing-spondylitis-spinal-score-msasss-as-a-tool-to-assess-axial-psa/