Abstract
BACKGROUND: Psoriatic arthritis with inflammatory axial involvement (PsA-ax) and axial spondyloarthritis (axSpA) are distinct entities within the spectrum of spondyloarthritides. Despite overlapping clinical and imaging features, the extent and pattern of radiographic spinal progression in PsA-ax relative to axSpA remain insufficiently characterized. OBJECTIVE: To describe and analyse differences in radiographic spinal progression between axSpA and PsA-ax. METHODS: In this retrospective cohort study, 246 patients (axSpA: n=172; PsA-ax: n=74) with lateral radiographs of the cervical, thoracic, and/or lumbar spine were included. Radiographic progression was quantified using the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS). An adjusted linear mixed-effects model was used to analyse differences in the longitudinal association with mSASSS between axSpA and PsA-ax. The estimated β-coefficient along with the 95% confidence interval (CI) is reported. RESULTS: The mSASSS differed substantially at baseline, averaging 6 (SD 14) in axSpA and 0.96 (SD 2.12) in PsA-ax. The patient-level mean MSASSS change over 2 years was small overall, but higher in patients with axSpA than in those with PsA-ax, averaging 0.39 (SD 1.83) and 0.07 (SD 0.29), respectively. A total of 27% of axSpA and 5% of PsA-ax had a patient-level mean MSASSS change over 2 years >0 units. In the adjusted linear mixed-effects model, the estimated mean progression in mSASSS over two years was slightly lower in PsA-ax compared to axSpA (adjusted β = -0.088, 95% CI: -0.446 to 0.271). CONCLUSION: Patients with PsA-ax were found to have slightly lower mean progression in mSASSS over two years than patients with axSpA. However, the wide CI crossing zero indicates large uncertainty and compatibility with no difference in mSASSS progression between PsA with axial involvement and axSpA, warranting further studies.