Abstract
BACKGROUND: Morphoea disease state (active vs. inactive) is difficult to ascertain using clinical examination alone, but it is crucial to optimizing treatment. To date, no biomarkers indicative of disease state have been fully validated in morphoea. OBJECTIVES: To better validate CXCL9 as a biomarker of morphoea disease state. METHODS: This was a prospective longitudinal cohort study of 133 people with morphoea. Morphoea activity and damage were determined with the validated Localized Scleroderma Cutaneous Assessment Tool (LoSCAT) and CXCL9 serum concentrations were measured by enzyme-linked immunosorbent assay. Analysis included measures of association of CXCL9 concentrations with clinical activity and damage scores. Published guidelines for biomarker validation were used. RESULTS: At baseline, CXCL9 concentration was elevated in patients with active morphoea (n = 50/81; 62%) vs. those with inactive morphoea [n = 20/52; 38% (P = 0.02)] and healthy control individuals [n = 7/25; 28% (P = 0.02)]. In the morphoea group overall, CXCL9 concentration correlated with validated activity and damage scores [modified Localized Scleroderma Skin Severity Index (mLoSSI) r = 0.43 (P < 0.001); Localized Scleroderma Skin Damage Index r = 0.52 (P < 0.001)]. Correlation of CXCL9 concentrations with damage was driven by the sclerosis subscore (r = 0.53; P < 0.001). Receiver operating characteristic analysis of cutoff calculated from healthy control participants revealed increasing sensitivity (mLoSSI > 3: 66.1%; > 10: 78.9%; > 20: 86.5%) and stable specificity (mLoSSI > 3: 58.3%; > 10: 55.6%; > 20: 51.9%) with more active disease. Baseline CXCL9 concentrations were elevated in patients with the generalized subtype of morphoea vs. all other subtypes and in adults vs. children. Multivariable longitudinal analysis corroborated positive correlations of mLoSSI and older age with CXCL9 concentration and revealed a new, positive association with musculoskeletal extracutaneous manifestations. Increases in CXCL9 concentrations predicted reactivation of morphoea activity in 73% of patients (n = 8/11) within 9 months of the increase. CONCLUSIONS: Our findings suggest that CXCL9 serum concentrations can be used as a biomarker, in addition to clinical assessment of morphoea, to identify and predict morphoea disease state and severity.