Incremental Predictive Value of Dynamic Platelet‐to‐Lymphocyte Ratio Beyond ESR and CRP for Glucocorticoid Sparing in Rheumatoid Arthritis: A Multi‐Center Retrospective Cohort Study

ABSTRACT Objective To identify independent predictors of sustained glucocorticoid (GC) sparing in patients with rheumatoid arthritis (RA) initiating biological or targeted synthetic disease‐modifying antirheumatic drugs (b/tsDMARDs), specifically we evaluated whether the longitudinal platelet‐to‐lymphocyte ratio (ΔPLR) trajectory provides superior prognostic value compared to conventional measurements. Methods This multicenter retrospective study analyzed 264 patients from the Taipei Medical University Clinical Research Database. Sparing success was defined as achieving predefined dose reduction targets at month 6 and maintaining stability for ≥ 12 weeks. Candidate predictors were screened using least absolute shrinkage and selection operator (LASSO) regression, followed by multivariable logistic and Cox proportional hazards modeling. Clinical utility was assessed using “rescue rate” analysis. Results Sustained GC sparing was achieved by 139 (52.7%) patients. First‐line b/tsDMARD initiation was the strongest predictor (adjusted odds ratio [aOR] 4.23, p < 0.001) of GC‐sparing success. Although static baseline inflammatory markers were non‐predictive, every 10% reduction in 6‐month ΔPLR independently increased success odds (aOR 1.05 [inverted from 0.95 per 10% increase], p = 0.009) and significantly accelerated the sparing rate (hazard ratio [HR] 1.02, p = 0.021). Conversely, concomitant nonsteroidal antiinflammatory drug use hindered success. Notably, incorporating ΔPLR into the predictive model “rescued” 16.5% of unrecognized candidates for de‐escalation—patients capable of successful GC sparing who were misclassified by standard clinical models. Conclusion Although static baseline inflammatory markers are nonpredictive, 6‐month ΔPLR offers significant incremental prognostic value beyond conventional erythrocyte sedimentation rate (ESR) and C‐reactive protein (CRP). By identifying 16.5% of viable candidates overlooked by standard assessments, dynamic PLR monitoring offers a potential strategy to optimize precision steroid de‐escalation in RA.