Eur J Pediatr. 2026 May 20;185(6):418. doi: 10.1007/s00431-026-07087-y.
ABSTRACT
Intravenous immunoglobulin (IVIG) at 2 g/kg is the standard first-line therapy for acute Kawasaki disease (KD), yet a subset of patients remains febrile after treatment. Whether IVIG non-response is primarily explained by a low early post-infusion IgG concentration remains uncertain. We retrospectively reviewed consecutive patients with KD at a single center (2009-2019). Patients receiving initial IVIG 2 g/kg without adjunctive therapy were included (N = 109; responders = 86, non-responders = 23). Serum IgG was measured at admission and within 24 h after completion of IVIG. Associations with non-response were examined using multivariable Firth penalized logistic regression, with IgG scaled per 100 mg/dL and adjustment for prespecified clinical covariates. Sensitivity analyses included additional adjustment for the Kobayashi score and exclusion of one non-responder treated without meeting the objective fever criterion. Baseline IgG was correlated with early post-IVIG IgG (Spearman's ρ = 0.481, P < 0.001). In multivariable Firth penalized logistic regression model, higher post-IVIG IgG was independently associated with non-response (OR 1.26 per 100 mg/dL; 95% CI 1.06-1.53), whereas baseline IgG was not significant in the same model. Neutrophil percentage (pre-IVIG) was also independently associated with non-response (OR 1.08 per 1% increase; 95% CI 1.02-1.16). Sensitivity analyses yielded consistent direction and significance of the main associations. Conclusions: Under the standard 2 g/kg regimen, IVIG non-response in Kawasaki disease was not explained by a low early post-infusion serum IgG concentration. Non-response occurred even at early post-IVIG IgG concentrations within-and above-the range observed among responders, suggesting escalation strategies beyond simply repeating IVIG to further increase IgG.
PMID:42159808 | DOI:10.1007/s00431-026-07087-y