Background/Purpose: Psoriatic Arthritis (PsA) is an inflammatory arthritis associated with psoriasis that leads to progressive joint damage. Biomarkers in addition to clinical features may help stratify patients with PsA at high risk of joint damage so that they may be aggressively treated. Our purpose was to determine the predictive value of clinical features to predict joint damage in PsA.

Methods: The study was conducted in a large prospective cohort of patients with PsA who are assessed according to a standard protocol that includes detailed clinical evaluation every 6 months and radiographic evaluation every 2 years. Patients who had at least 2 radiographs were included in this study. Radiographic joint damage progression was defined as increase in the count of radiographically damaged joints between 1st and 2nd radiographic assessment in any of the following joints- wrists, MCP, PIP, DIP, MTPs and 1^st IP of toes (total 42 joints). The following clinical variables at the time of first radiographic assessment were evaluated as predictors- age, sex, age at diagnosis of psoriasis and PsA, duration of psoriasis and PsA, PASI, active, swollen, clinically damaged and radiographically damaged joint count, dactylitis, presence of axial disease, ESR, SF-36, HAQ, NSAID, DMARD and biologic use at baseline, HLA-B*27 and HLA-C*6. Univariate followed by multivariate models were developed to determine association of clinical variables with radiographic damage progression. ROC curves were constructed using variables found to be significantly associated with radiographic progression.

Results: 656 (380 males) with a mean age of 43 years, age at onset of psoriasis 28 years and of PsA 37 years were included. At the time of the first radiograph these patients had 11active, 5 swollen, 9 clinically damage, 4 radiographically damaged joints. The ESR was 24 mm/hr and PASI score 5. 33% had dactylitis and 45% had evidence of axial disease. 17% were HLA-B*27 and 32% HLA-C*06 positive. Their SF-36 PCS was 36.5 and HAQ 0.8. In univariate analyses older age, age at diagnosis of psoriasis, duration of psoriasis and PsA, active, swollen, clinically damaged and radiographically damaged joint counts, dactylitis, axial arthritis, ESR, HAQ and treatment with DMARDs were associated with radiographic progression (p<0.05). The predictors independently associated with radiographic progression in a multivariate reduced model are shown in Table 1. The area under the curve of the ROC curves was 0.71.

Table 1: Predictors of radiographic damage progression in a multivariate reduced model

Conclusion: Clinical features have only a fair ability to discriminate patients who have radiographic joint damage progression in PsA. Therefore, identifying biomarkers that predict joint damage is an important research question.