Abstract
Background: Adult-onset Still's disease (AOSD) is a rare systemic inflammatory disorder of unknown etiology and pathogenesis. Some patients fail to respond to conventional glucocorticoids and immunosuppressant therapies, a condition known as refractory AOSD. The prognosis for patients with refractory AOSD is typically poor, significantly impacting their quality of life and overall health. This study retrospectively analyzes the predictive factors for refractory AOSD to provide new strategies and insights for clinical diagnosis and treatment. Methods: Overall, 105 AOSD patients hospitalized between January 2008 and October 2024 were selected, 41 of whom were classified as refractory. Multivariate logistic regression analysis was conducted to identify risk factors for refractory AOSD, and receiver operating characteristic (ROC) curves were used to evaluate the predictive power of these indicators. Results: Patients with refractory AOSD were more likely to develop splenomegaly and MAS. Additionally, the neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase, serum ferritin (SF) levels, and AOSD system score were higher in refractory cases than in nonrefractory cases, while lymphocyte count and platelet (PLT) count were lower in the refractory AOSD group (p < 0.05). Multivariate logistic regression analysis identified PLT, NLR, and AOSD system scores as independent risk factors for predicting refractory AOSD. ROC curve analysis revealed that the area under the curve for PLT, NLR, and AOSD system scores were 0.659, 0.661, and 0.660, respectively. The optimal cutoff values for PLT, NLR, and AOSD system score in predicting refractory AOSD were 314.5 × 10(9)/L, 10.555, and 5.5, respectively, with sensitivities of 80.5%, 53.7%, and 75.6% and specificities of 46.9%, 75.0%, and 50.0%, respectively. Conclusion: PLT < 314.5 × 10(9)/L, NLR > 10.555, or an AOSD system score of > 5.5 before treatment may serve as independent risk factors for predicting refractory AOSD, providing clinicians with an early warning to identify disease progression.