Sequential disease activity assessment for predicting future treatment trajectories in patients with rheumatoid arthritis treated with JAK inhibitors: a group-based trajectory and landmark analysis.
retrospective_cohort · Level III
Where this comes from
- Record sourced from PubMed, PMID 42722589.
- Also identified by DOI 10.1016/j.ard.2026.08.010.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
This study aimed to identify disease activity trajectories in patients with rheumatoid arthritis (RA) treated with Janus kinase (JAK) inhibitors and to determine whether early disease activity improved prediction of long-term treatment trajectories. We analysed 843 patients treated with JAK inhibitors from a multicentre registry. Group-based trajectory modelling was used to identify distinct Clinical Disease Activity Index (CDAI) trajectories over 52 weeks. Landmark analyses at weeks 4 and 12 were performed using multivariate logistic regression with multiple imputations. Model performance was evaluated using the area under the receiver operating characteristic curve, Brier score, and bootstrap-based internal validation. Four distinct trajectory groups were identified: rapid responders (64.9%), partial responders with persistent residual disease activity (18.1%), responders with high baseline disease activity (12.7%), and persistent nonresponders (4.3%). Week-52 remission rates ranged from 0.0% to 47.7% across the groups. Incorporating the CDAI at landmark time points substantially improved the prediction of a favourable trajectory, whereas the CDAI slope provided minimal additional predictive value. Brier scores improved markedly after incorporating week-12 CDAI, with minimal further improvement after CDAI slope inclusion. Bootstrap validation demonstrated minimal optimism and good calibration. The optimal CDAI thresholds for predicting a favourable trajectory were 12.0 and 9.3 at weeks 4 and 12, respectively. Early disease activity dynamically refined the prediction of future treatment trajectories in patients with RA treated with JAK inhibitors. The absolute CDAI at landmark time points was more informative than change-based metrics, supporting sequential disease activity assessment for time-anchored treatment decision-making.