Eur J Pediatr. 2026 Sep 8;185(10):729. doi: 10.1007/s00431-026-07390-8.
ABSTRACT
Kawasaki disease shock syndrome (KDSS) is a severe phenotype of Kawasaki disease (KD) characterized by hemodynamic instability and systemic inflammation. Early identification of children at risk for KDSS remains challenging. N-terminal pro-B-type natriuretic peptide (NT-proBNP) reflects cardiovascular stress and has been associated with severe manifestations of KD; however, its interpretation in children is complicated by age dependency. This study aimed to evaluate the predictive utility of baseline NT-proBNP, compare raw NT-proBNP concentrations with age-adjusted NT-proBNP z-scores, and assess their performance relative to other routinely available laboratory biomarkers for KDSS prediction. A total of 588 KD patients were finally enrolled in this prospective cohort study between January 2015 and March 2021. The demographic/clinical characteristics and laboratory data were compared between the patients with KDSS (KDSS group, n = 35) and those without (KD group, n = 553). NT-proBNP concentrations were analyzed as both raw values and age-adjusted z-scores calculated using established pediatric reference ranges. Multivariable logistic regression models were constructed using baseline variables, with log10-transformed NT-proBNP and NT-proBNP z-score evaluated in separate models. Receiver operating characteristic (ROC) analyses were performed to compare predictive performance among NT-proBNP-related parameters and other candidate biomarkers. Children in the KDSS group were older than those in the non-KDSS group (66.11 ± 47.06 vs. 31.36 ± 23.57 months, P < 0.001). KDSS patients had significantly higher rates of coronary artery lesions (40.0% vs. 11.6%, P < 0.001) and intravenous immunoglobulin (IVIG) resistance (51.5% vs. 12.5%, P < 0.001), while classic clinical manifestations such as bilateral conjunctival injection (74.3% vs. 92.8%, P = 0.001) and oral mucosal erythema (77.1% vs. 93.1%, P = 0.004) were less frequent. Regarding laboratory parameters, KDSS patients exhibited significantly higher neutrophil-to-lymphocyte ratio (NLR), C-reactive protein-to-albumin ratio (CAR), creatinine, and NT-proBNP, but significantly lower lymphocyte-to-C-reactive protein ratio (LCR), hemoglobin, platelet count, hematocrit, sodium, and potassium levels compared with non-KDSS patients. Multivariate logistic regression identified older age, higher NT-proBNP (both raw and age-standardized Z-score), lower platelet, hemoglobin, and rates of erythema of oral and pharyngeal mucosa as independent predictors of KDSS. ROC analysis showed that the raw NT-proBNP cutoff of 1354 pg/mL gave an AUC of 0.71 (95% CI 0.61-0.81), with sensitivity of 74.3% and specificity of 64.2%, and age-standardized NT-proBNP Z-score had an AUC of 0.73 (95% CI 0.63-0.83), with an optimal cutoff of ≥ 3.3 yielding a sensitivity of 74.3% and specificity of 65.8%.
CONCLUSION: NT-proBNP might be a complementary laboratory marker for the prediction of KDSS in KD.
WHAT IS KNOWN: • NT-proBNP concentrations are age-dependent in children, but most previous studies of KDSS have relied on absolute NT-proBNP concentrations without age-standardized interpretation.
WHAT IS NEW: • Baseline NT-proBNP was independently associated with KDSS, and age-adjusted NT-proBNP z-scores showed comparable predictive performance, supporting NT-proBNP as an adjunctive marker for early KDSS risk stratification after consideration of age dependency.
PMID:42711435 | DOI:10.1007/s00431-026-07390-8