Real‑world assessment of the effectiveness and response profiles of biologic DMARDs in psoriatic arthritis: findings from an emulated multicentre clinical trial.
Abstract
Objectives
To evaluate 6-month effectiveness and long-term drug-retention rates of biologic disease-modifying antirheumatic drugs (bDMARDs) in psoriatic arthritis (PsA) using real-world data analysed within an emulated multicentre clinical trial framework. Secondary objectives included the characterisation of response profiles and the identification of possible predictors of clinical response across bDMARD classes.
Methods
Between January 2022 and June 2025, consecutive patients with moderate-to-high PsA disease activity initiating a bDMARD were prospectively enrolled in a multicentre Gruppo Italiano di Ricerca in Reumatologia Clinica e Sperimentale study. Real-world data were analysed within a target trial emulation framework comparing tumour necrosis factor (TNF) inhibitors, interleukin (IL)-17 inhibitors, and IL-23-pathway inhibitors. Propensity score weighting was used to improve comparability in measured baseline characteristics across treatment groups. Effectiveness was assessed through longitudinal changes in the Disease Activity Index for Psoriatic Arthritis (DAPSA) over 6 months. Subsequent long-term drug retention was evaluated using a 6-month landmark analysis restricted to patients who remained on the index bDMARD at month 6. Potential clinical factors associated with treatment response and discontinuation were also explored.
Results
Overall, 990 patients were included: 342 initiated TNF inhibitors, 313 IL-17 inhibitors, and 335 IL-23-pathway inhibitors. DAPSA decreased significantly over 6 months in all 3 treatment groups, with no statistically significant between-class differences in adjusted longitudinal trajectories. Among patients who remained on the index bDMARD through month 6, the adjusted analyses did not identify statistically significant between-class differences in subsequent long-term drug retention. Shorter disease duration was associated with more favourable outcomes in selected exploratory analyses, whereas associations with previous bDMARD exposure were inconsistent. Additional class-specific associations were identified but should be considered exploratory and hypothesis-generating. Adverse-event-related discontinuations documented in routine clinical records were described; however, the study was not designed to support formal comparative safety conclusions.
Conclusions
In this real-world emulated multicentre clinical trial, bDMARDs showed comparable overall trajectories of short-term effectiveness and long-term treatment persistence in PsA. The identification of partially distinct response patterns may support a more individualised approach to bDMARD selection.