Galectin‐9 and CXCL10 as Biomarkers for Disease Activity in Juvenile Dermatomyositis: A Longitudinal Cohort Study and Multicohort Validation

OBJECTIVE: Objective evaluation of disease activity is challenging in patients with juvenile dermatomyositis (DM) due to a lack of reliable biomarkers, but it is crucial to avoid both under‐ and overtreatment of patients. Recently, we identified 2 proteins, galectin‐9 and CXCL10, whose levels are highly correlated with the extent of juvenile DM disease activity. This study was undertaken to validate galectin‐9 and CXCL10 as biomarkers for disease activity in juvenile DM, and to assess their disease specificity and potency in predicting the occurrence of flares. METHODS: Levels of galectin‐9 and CXCL10 were measured by multiplex immunoassay in serum samples from 125 unique patients with juvenile DM in 3 international cross‐sectional cohorts and a local longitudinal cohort. The disease specificity of both proteins was examined in 50 adult patients with DM or nonspecific myositis (NSM) and 61 patients with other systemic autoimmune diseases. RESULTS: Both cross‐sectionally and longitudinally, galectin‐9 and CXCL10 outperformed the currently used laboratory marker, creatine kinase (CK), in distinguishing between juvenile DM patients with active disease and those in remission (area under the receiver operating characteristic curve [AUC] 0.86–0.90 for galectin‐9 and CXCL10; AUC 0.66–0.68 for CK). The sensitivity and specificity for active disease in juvenile DM was 0.84 and 0.92, respectively, for galectin‐9 and 0.87 and 1.00, respectively, for CXCL10. In 10 patients with juvenile DM who experienced a flare and were prospectively followed up, continuously elevated or rising biomarker levels suggested an imminent flare up to several months before the onset of symptoms, even in the absence of elevated CK levels. Galectin‐9 and CXCL10 distinguished between active disease and remission in adult patients with DM or NSM (P = 0.0126 for galectin‐9 and P < 0.0001 for CXCL10) and were suited for measurement in minimally invasive dried blood spots (healthy controls versus juvenile DM, P = 0.0040 for galectin‐9 and P < 0.0001 for CXCL10). CONCLUSION: In this study, galectin‐9 and CXCL10 were validated as sensitive and reliable biomarkers for disease activity in juvenile DM. Implementation of these biomarkers into clinical practice as tools to monitor disease activity and guide treatment might facilitate personalized treatment strategies.