Kohei Yo, Takumi Nagamoto, Yuri Nishida, Mari Hara, Yudai Koshida, K Orita, Hiroyuki Murabe
Objectives This study aimed to evaluate the safety and effectiveness of tocilizumab (TCZ) as induction therapy in Still disease and identify early dynamic biomarkers, including C-reactive protein (CRP) and ferritin changes, for macrophage activation syndrome (MAS) risk stratification. Methods This single-centre retrospective cohort study included patients with Still disease treated from 2004 to 2024. Patients were classified by induction therapy (TCZ vs non-TCZ). MAS was defined using the 2016 European Alliance of Associations for Rheumatology/American College of Rheumatology/Paediatric Rheumatology International Trials Organisation criteria. Early biomarker changes after prednisolone initiation were evaluated. Primary outcomes were week-4 CRP normalisation and clinically inactive disease (CID) at month 3. Associations with induction therapy were analysed using Firth's penalised logistic regression. In TCZ-treated patients, receiver operating characteristic analyses assessed laboratory biomarkers for MAS prediction. Results Among 54 patients, 12 received TCZ as initial induction therapy and 7 for relapse (19 episodes). MAS occurred in 5 episodes (26.3%). No significant differences were observed in week-4 CRP normalisation (75.0% vs 51.2%, P = .193) or month-3 CID (40.0% vs 14.6%, P = .090). After adjusting for the higher initial prednisolone dose, early TCZ exposure was independently associated with month-3 CID (odds ratio: 9.73, 95% CI: 1.25-77.50, P = .030). Week-1/baseline CRP ratio and ferritin ratios predicted MAS after TCZ induction; the week-1/baseline ferritin ratio showed the highest area under the curve (0.914), followed by the CRP ratio (0.877). Conclusions Early TCZ exposure was independently associated with CID at month 3. Insufficient early suppression of inflammatory markers, particularly higher ferritin and CRP ratios, may identify patients at higher risk of subsequent MAS.