Application of machine learning techniques to explore the occurrence of macrophage activation syndrome in Still's disease: results from the GIRRCS AOSD Study Group and the AIDA Network Still's Disease Registry

Creative Commons License

Ruscitti P., Masedu F., Vitale A., Caggiano V., Di Cola I., Atzeni F., ...Daha Fazla

Frontiers in immunology, cilt.17, ss.1811317, 2026 (SCI-Expanded, Scopus) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 17
  • Basım Tarihi: 2026
  • Doi Numarası: 10.3389/fimmu.2026.1811317
  • Dergi Adı: Frontiers in immunology
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, EMBASE, MEDLINE, Directory of Open Access Journals
  • Sayfa Sayıları: ss.1811317
  • Anahtar Kelimeler: adult onset Still’s disease, machine learning, macrophage activation syndrome, Still’s disease, systemic juvenile idiopathic arthritis
  • Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
  • Marmara Üniversitesi Adresli: Evet

Özet

Objectives: This study aims to explore the application of machine learning techniques in assessing macrophage activation syndrome (MAS) in Still's disease. Methods: A multicenter, observational, prospective study was conducted, including patients with Still's disease enrolled in the Gruppo Italiano di Ricerca in Reumatologia Clinica e Sperimentale (GIRRCS) AOSD Study Group and the AutoInflammatory Disease Alliance (AIDA) Network Still's Disease Registry. Results: A total of 737 patients (age: 35.5 ± 17.8, male sex: 44.7%) with Still's disease were assessed; 11.4% were affected by MAS, and 3% had a poor prognosis. First, random forest imputation was applied to the original dataset. Subsequently, a machine-learning-driven assessment was developed to explore MAS occurrence. Collectively, regression models, an exploration decision tree, and a random forest were applied, suggesting the importance of ferritin, age, C-reactive protein (CRP), and systemic score. A logistic regression model accounting for data leakage concerns was then generated using these variables, and missing values were imputed using random forest imputation. This analysis supported the role of the selected variables, which were further combined across different clinical scenarios to estimate the probability of MAS. The highest risk of MAS was estimated for patients simultaneously characterized by age ≥ 45 years, ferritin ≥ 4,178.10 ng/mL, CRP ≥ 27.15 mg/L, and a systemic score ≥ 7, corresponding to a 34.7% probability of MAS, as well as for those characterized by ferritin ≥ 4,178.10 ng/mL, CRP ≥ 27.15 mg/L, and systemic score ≥ 7, corresponding to a 33.5% probability of MAS. Conclusions: A machine-learning-driven prediction of MAS was explored in Still's disease, highlighting the importance of age of onset, hyperferritinaemia, increased CRP, and multiorgan involvement. A combination of these features may suggest a clinician-friendly algorithm for stratifying the probability of MAS during Still's disease.