47 - RELIABLE PREDICTION OF CORONARY ARTERY LESIONS DUE TO THE TREATMENT-RESISTANT KAWASAKI DISEASE PRIOR TO INITIAL TREATMENT WITH NEUTROPHILS AND TOTAL BILIRUBIN VALUES
Background Kawasaki disease (KD) is known to form coronary artery lesions (CALs) in about one in four of patients without treatment, which causes the risk for myocardial infarction. Particularly in the case of treatment resistance, CALs occur at a high rate. To prevent CALs, it is important to identify risks as early as possible and establish aggressive treatment strategies. Aim of this study is to investigate whether CALs caused by treatment-resistant KD cases can be predicted before initial treatment. Methods Three Hundred Sixty-Four KD patients who admitted in our hospital were included. We detected independent risk factors that contribute to the prediction of CALs (Z≥2.5) due to treatment-resistant KD (defined as requiring third-line treatment) by multivariate analysis. We also determined the cut-off value for identifying subjects at risk of CALs by receiver operating characteristic analysis, and we developed a new model for predicting CALs (including transient dilatation) due to treatment-resistant KD. Results There were 22 treatment-resistant cases (6%) and 6 cases that had CALs (1.6%). All patients with CALs were in the treatment-resistant group, not in the treatment-response group (p < 0.01). Neutrophils and Total Bilirubin (T-Bil) before initial treatment were detected as independent risk factors. When both Neutrophils ≥20,000 and T-Bil ≥2.4 were present, CALs occurred with a sensitivity of 100%, a specificity of 99%, a positive predictive value of 67%, and a negative predictive value of 100%. Conclusion It is possible to predict CALs caused by treatment-resistant KD cases with high sensitivity and specificity before initial treatment with a new model using neutrophils and T-bil. If the occurrence of CALs due to treatment-resistant KD is expected, it is possible to consider an aggressive treatment strategy from the initial treatment.
