Background/Purpose: High Mobility Group Box 1 Protein (HMGB1) is a nuclear protein that stabilizes DNA and modulates gene expression.  In sepsis and in certain other systemic inflammatory conditions HMGB1 is released extracellularly to mediate an array of cell signaling pathways that promote downstream release of pro-inflammatory cytokines.  Elevated blood levels of HMGB1 are found in association with sepsis and a variety of conditions characterized by systemic inflammation. The purpose of this study was to determine if HMGB1 levels could serve as a discriminating biomarker in early Kawasaki Disease (KD) compared to new onset systemic juvenile idiopathic arthritis (sJIA).

Methods: Children (age ≤ 16 years) with KD and, for comparison, those with sepsis (systemic inflammatory response syndrome or bacteremia) and other non-infectious/non-inflammatory conditions were prospectively enrolled at the time of hospital admission. Prospectively collected data from new onset, treatment naive sJIA participants were derived from the BBOP Study (Biologically-based Outcome Predictors in JIA; www.bbop.ca). Blood was collected in P100 vacutainers (BD) and plasma stored at -80⁰C until assayed. HMGB1 was measured in duplicate (1:100 dilution) by enzyme immunoassay (Biomatrix). Inter-group differences were compared by independent samples t tests and correlations by Pearson correlation coefficients.

Results:   The study population comprised 70 participants (Table).  HMGB1 levels were significantly higher in KD compared to sJIA (t= 4.19; p<.001; CI= 310-892) and to patients with other conditions (t=2.89; p=.009; CI=140-875) but were not different from HMGB1 levels in sepsis (t=.82; p=0.42; CI=-246-582).  Participants with sJIA had significantly lower levels of HMGB1 compared to those with sepsis (t=2.42; p=.019; CI=73-792) and did not differ from those with other conditions.   HMGB1 levels were significantly higher in younger children (r=.26); children 3 years of age or younger had higher HMGB1 levels than those older than age 3 (p=.028; CI=34-600). There was no correlation between HMGB1 levels and c-reactive protein in any of the groups.  No sex differences in HMGB1levels were observed.

Conclusion:  Results of this study show that HMGB1 levels are significantly higher in children with KD than in sJIA and comparable to levels seen in children with sepsis. HMGB1 levels in sJIA are not different than in children with other non-infectious/non-inflammatory conditions. Further prospective studies in larger cohorts are required to determine if HMGB1 could serve as a biomarker to predict disease course and outcomes in KD and as an early biomarker useful for distinguishing KD from sJIA. Results suggest that HMGB1 has a role in mediating KD pathogenesis and that HMGB1, as a pro-inflammatory mediator, might be a potential target for biologically-based therapy in acute KD.