Abstract
OBJECTIVES: To describe diagnostic framework classification, treatment patterns and outcomes in Kawasaki disease-associated macrophage activation syndrome (KD-MAS) using a single-centre cohort and a structured descriptive synthesis of published cases to inform hypothesis generation and future refinement of diagnostic and management strategies for KD-MAS. METHODS: We performed a retrospective single-centre cohort study. MAS was classified by haemophagocytic lymphohistiocytosis (HLH)-2004, HLH-2009 or 2016 systemic juvenile idiopathic arthritis (sJIA)-MAS criteria. Data were abstracted around a prespecified MAS window; severity was indexed by haemophagocytic syndrome diagnostic score (HScore) and the association between HScore and treatment escalation was assessed using Firth's logistic regression. In parallel, we conducted a literature review. RESULTS: In our centre, incidence was 0.6% (22/3786); mean age was 3.72 years; coronary involvement was 77.2%. The proportions of clinician-diagnosed KD-MAS cases fulfilling each framework were HLH-2004 14/22, HLH-2009 18/22 and 2016 sJIA-MAS 20/22; 11 met all three. Cytopenias, liver dysfunction and coagulopathy were frequent; management followed stepwise escalation from intravenous immunoglobulin (IVIG) or corticosteroid monotherapy to IVIG plus corticosteroids and, when required, to adjunct immunosuppressive/biologic therapy; similar patterns appeared in the literature. HScore was higher with intensified therapy than with standard therapy (median 274 vs 199; p=0.020). Each 10-point HScore increase was associated with higher odds of escalation (OR 1.44 univariable; 1.37 adjusted). CONCLUSION: The 2016 sJIA-MAS criteria demonstrate the highest proportion fulfilling criteria in our cohort for KD-MAS. Treatment in our cohort and in published cases generally reflected a stepwise, typically progressing from IVIG±glucocorticoids to adjunct immunosuppressants/biologics. Higher HScores co-occurred with escalation, suggesting they capture baseline severity; prospective multicentre studies are needed to test incremental value and risk-stratification thresholds.