Temporal Clinical Ultrasound Asynchrony in Psoriatic Arthritis Enthesitis: Implications for Personalized Monitoring

Background: In psoriatic arthritis (PsA), clinical tenderness and ultrasound (US) capture distinct yet related aspects of entheseal disease activity. However, their longitudinal relationship after initiation of biologic disease-modifying antirheumatic drugs (bDMARDs), and the clinical significance of early discordance during follow-up remain unclear. Methods: In this retrospective observational cohort study based on routinely collected medical records, patients with CASPAR-defined PsA and clinically and ultrasonographically active enthesitis at baseline (Clin+/US+) who initiated bDMARD therapy underwent paired, same-day, blinded clinical and US assessments at approximately 6 and 12 months. Agreement between clinical and US findings was quantified using Cohen's kappa. Discordant states (Clin-/US+ and Clin+/US-) were prespecified, and predictors of Clin-/US+ status at 6 months were analyzed using models that accounted for within-patient clustering. Results: Thirty-nine patients contributed 82 entheses and were treated with either tumour necrosis factor inhibitors (53.8%) or interleukin-17 inhibitors (46.2%). At 6 months, agreement between clinical and US assessments was fair (κ = 0.286; 95% confidence interval [CI], 0.080 to 0.492), with 23.2% of entheses classified as Clin-/US+ and 52.4% as concordantly inactive. At 12 months, agreement improved to substantial-to-almost-perfect levels (κ = 0.779; 95% CI, 0.595 to 0.963), with only 1.2% of entheses remaining Clin-/US+ and 80.5% achieving concordant remission. NSAID exposure was the only significant predictor of Clin-/US+ status at 6 months in univariable analysis (odds ratio [OR], 3.82; 95% CI, 1.27 to 11.47; p = 0.017) and remained associated after multivariable adjustment (OR, 6.16; 95% CI, 1.14 to 33.2; p = 0.03). Conclusions: In PsA patients starting bDMARD therapy, clinical and US assessments of enthesitis showed partial discordance at 6 months, followed by greater convergence at 12 months. These findings suggest that clinical and imaging abnormalities may resolve asynchronously during follow-up and should therefore be interpreted in an integrated, time-aware manner. Residual US abnormalities in the setting of clinical improvement should be interpreted cautiously and within the broader clinical context.