Diagnostic performance of HScore for macrophage activation syndrome complicating Still's disease in adults: a multicentre case-control study.

Abstract

To validate the diagnostic performance of the HScore for Still's disease (SD)-associated macrophage activation syndrome (MAS; SD-MAS) in adults compared with those of the HLH-2004 diagnostic criteria used with modifications (HLH-04) and the 2016 EULAR/ACR/PRINTO classification criteria for systemic juvenile idiopathic arthritis-associated MAS (MAS-2016). This multicentre case-control study, conducted at three medical centres in Japan between 2004 and 2023, enrolled patients ≥16 years with active SD and allocated them into SD without MAS and SD-MAS groups by an expert panel. For each patient, we calculated the HScore, HLH-04, MAS-2016, and modified HScore, which excluded haemophagocytosis on bone marrow aspirate. Receiver operating characteristic curve analysis was used to assess the discriminative ability of each criterion. Logistic regression analysis was performed separately for 'model-1' (HScore and HLH-04) and 'model-2' (MAS-2016 and modified HScore). This study included 86 patients with SD, 25 of whom had SD-MAS. The HScore showed the best area under the curve at 0.991 (95% confidence interval [CI]: 0.977-1.000), with a cut-off value of 192 (sensitivity, 96.0%; specificity, 96.7%). The optimal cut-off for the modified HScore was 171 (sensitivity, 92.0%; specificity, 83.6%). Multivariate analysis identified only the HScore in model-1 (odds ratio [OR]: 1.271; 95% CI: 1.006-1.606), whereas both MAS-2016 (OR: 4.249; 95% CI: 1.636-11.041) and modified HScore (OR: 1.066; 95% CI: 1.017-1.118) were identified in model-2. The HScore and modified HScore could be useful tools for diagnosing SD-MAS in adults with and without bone marrow aspiration, respectively.