Identifying Biomarkers for Early Prediction of Coronary Artery Lesions and Intravenous Immunoglobulin Non-response in Kawasaki Disease Based on Urine Multi-omics Analysis

<title>Abstract</title> <p> To identify potential biomarkers for predicting coronary artery lesions (CAL) and intravenous immunoglobulin (IVIG) resistance in Kawasaki disease (KD) using integrated metabolomic and proteomic profiling of urine samples. <bold>Methods</bold> : Urine samples from 27 pediatric KD patients (13 CAL, 14 NCAL; 5 IVIG non-responders, 5 responders) were analyzed via LC-MS/MS-based non-targeted metabolomics and data-independent acquisition (DIA) proteomics. Differential metabolites (DEMs) and proteins (DEPs) were identified based on fold change &gt;1.5 or &lt;1/1.5 and p &lt; 0.05. Functional enrichment and correlation analyses were performed using KEGG and Spearman methods. <bold>Results:</bold> In the CAL group, 83 DEMs and 360 DEPs were identified. Integrated analysis revealed three core proteins (APRT, PNP, HPRT1) and three metabolites (guanosine, hypoxanthine, cGMP) linked to purine metabolism and ferroptosis. In IVIG non-responders, guanine was highlighted as a key metabolite, with tryptophan metabolism being the most relevant pathway. <bold>Conclusion:</bold> APRT, PNP, HPRT1, guanosine, hypoxanthine, and cGMP may serve as predictive biomarkers for CAL, while guanine shows potential for early detection of IVIG non-response. </p>