A Simple Scoring Model Based on Machine Learning Predicts Intravenous Immunoglobulin Resistance in Kawasaki Disease

In Kawasaki disease (KD), accurate prediction of intravenous immunoglobulin (IVIG) resistance is crucial to reduce a risk for developing coronary artery lesions. To establish a simple and accurate scoring model predicting IVIG resistance, we conducted a retrospective cohort study of 996 KD patients that were diagnosed at 11 facilities for 10 years, in which 108 cases (23.5%) were resistant to initial IVIG treatment. We performed machine learning with random forest model using 30 clinical variables at diagnosis in 796 and 200 cases for training and test datasets, respectively. Random forest model accurately predicted IVIG resistance (AUC; 0.75, sensitivity; 0.54, specificity; 0.80). Next, using top five influential features (days of illness at initial therapy, serum levels of C-reactive protein, sodium, total bilirubin, and total cholesterol) in the random forest model, we designed a simple scoring system. In spite of its simplicity, the scoring system predicted IVIG resistance (AUC; 0.73, sensitivity; 0.55, specificity; 0.83) as accurately as the random forest model itself. Moreover, accuracy of our scoring system with five clinical features was almost identical to that of Gunma score with seven clinical features (AUC; 0.73, sensitivity; 0.53, specificity; 0.83), a well-known logistic regression scoring model, and superior to that of two widely used scores (Kurume score; 0.67, 0.46 and 0.76, respectively, and Osaka score; 0.69, 0.33 and 0.84, respectively). Conclusions: Our simple scoring system based on the findings in machine learning, as well as machine learning itself, seems to be useful to accurately predict IVIG resistance in KD patients.

A Machine Learning Model to Predict Intravenous Immunoglobulin-Resistant Kawasaki Disease Patients: A Retrospective Study Based on the Chongqing Population

Objective: We explored the risk factors for intravenous immunoglobulin (IVIG) resistance in children with Kawasaki disease (KD) and constructed a prediction model based on machine learning algorithms.Methods: A retrospective study including 1,398 KD patients hospitalized in 7 affiliated hospitals of Chongqing Medical University from January 2015 to August 2020 was conducted. All patients were divided into IVIG-responsive and IVIG-resistant groups, which were randomly divided into training and validation sets. The independent risk factors were determined using logistic regression analysis. Logistic regression nomograms, support vector machine (SVM), XGBoost and LightGBM prediction models were constructed and compared with the previous models.Results: In total, 1,240 out of 1,398 patients were IVIG responders, while 158 were resistant to IVIG. According to the results of logistic regression analysis of the training set, four independent risk factors were identified, including total bilirubin (TBIL) (OR = 1.115, 95% CI 1.067–1.165), procalcitonin (PCT) (OR = 1.511, 95% CI 1.270–1.798), alanine aminotransferase (ALT) (OR = 1.013, 95% CI 1.008–1.018) and platelet count (PLT) (OR = 0.998, 95% CI 0.996–1). Logistic regression nomogram, SVM, XGBoost, and LightGBM prediction models were constructed based on the above independent risk factors. The sensitivity was 0.617, 0.681, 0.638, and 0.702, the specificity was 0.712, 0.841, 0.967, and 0.903, and the area under curve (AUC) was 0.731, 0.814, 0.804, and 0.874, respectively. Among the prediction models, the LightGBM model displayed the best ability for comprehensive prediction, with an AUC of 0.874, which surpassed the previous classic models of Egami (AUC = 0.581), Kobayashi (AUC = 0.524), Sano (AUC = 0.519), Fu (AUC = 0.578), and Formosa (AUC = 0.575).Conclusion: The machine learning LightGBM prediction model for IVIG-resistant KD patients was superior to previous models. Our findings may help to accomplish early identification of the risk of IVIG resistance and improve their outcomes.

1025. Prediction of Intravenous Immunoglobulin Resistance and Coronary Artery Dilatation in Kawasaki Disease: a Multicenter Study from Oman

Background Prediction of intravenous immunoglobulin (IVIG) resistance and coronary artery dilatation continues to be a challenge in the management of Kawasaki disease. Significant differences exist among different populations. Methods Children < 13 years of age who presented to the two main tertiary care hospitals in Oman (Royal Hospital and Sultan Qaboos University Hospital) between 2008 and 2019 with a diagnosis of Kawasaki disease were included. Diagnosis was confirmed and clinical, laboratory and echocardiography data was systematically collected and checked for accuracy. The primary outcome was the presence of IVIG resistance or coronary artery dilatation at the 6-week follow-up. Bivariate analysis was used to identify significant predictors of the primary outcome, followed by multivariable logistic regression to determine independent predictors. The Muscat Index of Kawasaki disease Severity (MIKS) score was created based on the results. Results 156 children with Kawasaki disease were included. Median age was 2.1 years (IQR 0.9-3.8), and 64% were males. All patients received IVIG, 26 (17%) received steroids, and one received infliximab. Coronary dilatation was identified in 41 (26%) patients on initial echocardiogram, and 26 (18%) at the 6-week follow-up visit. Variables significantly associated with the primary outcome were age ≤15 months (P=0.031), hemoglobin (P=0.009), WBC count (P=0.002), absolute neutrophil count (P=0.006), and CRP ≥150 mg/L (P=0.015). These variables in addition male gender (P=0.058), ALT >80 IU/L (P=0.10) and serum sodium (P=0.10), were entered into multivariable logistic regression. A predictive model based on CRP ≥150 mg/L (LR=2.2, P=0.049), male gender (LR=2.1, P=0.095) and WBC (LR=1.1, P=0.017) resulted, and it was used as basis for the MIKS score (Table 1). The MIKS score performed favorably to the Kobayashi score in its sensitivity to predict the primary outcome and its separate components (Table 2). Combining the MIKS score with other high-risk criteria had a sensitivity of 95% in predicting the primary outcome and a specificity of 56%. Table 1. Calculation of the Muscat Index of Kawasaki disease Severity (MIKS) score Table 2. Sensitivity, specificity and P value for the Kobayashi, MIKS, and combined high risk criteria in predicting IVIG resistance, coronary dilatation at 6 weeks, separately or in combination, among patients with Kawasaki disease. MIKS: Muscat Index of Kawasaki disease Severity. *High risk: presence of coronary artery dilatation on initial echocardiogram or age <1> Conclusion The MIKS score predicts IVIG resistance and coronary artery dilatation in Kawasaki disease in our setting, with favorable performance compared to the Kobayashi score. Disclosures All Authors: No reported disclosures

Predictive value of Albumin-Bilirubin grade for intravenous immunoglobulin resistance in a large cohort of patients with Kawasaki disease: a prospective study

Background Intravenous immunoglobulin (IVIG) resistance prediction is one of the primary clinical issues and study hotspots in KD. This study aimed to prospectively investigate the value of albumin-bilirubin grade (ALBI) in predicting IVIG resistance in KD and to assess whether ALBI has more predictive value or accuracy than either ALB or TBil alone in predicting IVIG resistance. Methods A total of 823 patients with KD were prospectively enrolled. The clinical and laboratory data were compared between the IVIG-response group (n = 708) and the IVIG-resistance group (n = 115). Multivariate logistic regression analysis was performed to identify the independent risk factors for IVIG resistance. Receiver operating characteristic (ROC) curves analysis was applied to assess the validity of ALBI, ALB, and TBil in predicting IVIG resistance. Results ALBI was significantly higher in patients with IVIG resistance and was identified as an independent risk factor for IVIG resistance in KD. The parameter of ALBI ≥ − 2.57 (AUC: 0.705, 95 %CI: 0.672–0.736), ALB ≤ 33.0 g/L (AUC: 0.659, 95 %CI: 0.626–0.692), and TBil ≥ 16.0µmol/L (AUC: 0.626, 95 %CI: 0.592–0.659), produced a sensitivity, specificity, PPV, and NPV of 0.617, 0.657, 0.226 and 0.914; 0.374, 0.850, 0.289 and 0.893; 0.269, 0.941, 0.425 and 0.888, respectively. Conclusions A higher ALBI was an independent risk factor for IVIG resistance in KD. It yielded better predictive ability than ALB and TBil alone for initial IVIG resistance.

Prediction of repeated intravenous immunoglobulin resistance in children with Kawasaki disease

Background Repeated intravenous immunoglobulin (IVIG) resistance prediction is one of the pivotal topics in Kawasaki disease (KD). Those non-responders of repeated IVIG treatment might be improved by an early-intensified therapy to reduce coronary artery lesion and medical costs. This study investigated predictors of resistance to repeated IVIG treatment in KD. Methods A total of 94 children with IVIG-resistant KD treated at our hospital between January 2016 and August 2020 were retrospectively analyzed. According to the therapeutic effect of a second dose IVIG treatment, the children were divided into repeated IVIG-responsive group and repeated IVIG-resistant group, and the clinical and laboratory data were compared. Predictors of repeated IVIG resistance and the optimal cut-off value were determined by multiple logistic regression analysis and receiver operating characteristic (ROC) curve analysis. Results The Pre-IVIG laboratory data showed the percentage of neutrophils (N%) and levels of serum procalcitonin (PCT), N-terminal pro-brain natriuretic peptide (NT-proBNP) were significantly higher in repeated IVIG-resistant group compared with repeated IVIG-responsive group, while levels of serum sodium and albumin (ALB) were significantly lower (P < 0.05). The post-IVIG laboratory values of N% and C-reactive protein (CRP) were significantly higher in the repeated IVIG-resistant group compared with repeated IVIG-responsive group, while hemoglobin and ALB were lower (P < 0.05). Pre-IVIG PCT and post-IVIG CRP exhibited AUC of 0.751 and 0.778 respectively in predicting repeated IVIG resistance in KD. Pre-IVIG PCT > 1.81ng/ml (OR 4.1, 95 % CI 1.4 ~ 12.0, P < 0.05) and post-IVIG CRP > 45 mg/L (OR 4.6, 95 % CI 1.3 ~ 16.2, P < 0.05) were independent predictors of repeated IVIG resistance in KD. Conclusions Our study illustrates the serum PCT level before initial IVIG treatment and CRP after initial IVIG could be used to predict repeated IVIG resistance in KD.

Low- versus high-concentration intravenous immunoglobulin for children with Kawasaki disease in the acute phase

Purpose: Few studies have compared the effects of low-concentration (5%) and high-concentration (10%) intravenous immunoglobulin (IVIG) preparations for patients with Kawasaki disease (KD) in the acute phase. The purpose of this study was to compare outcomes between low- and high-concentration IVIG preparations in children with KD, using a national inpatient database in Japan.Method: We used the Diagnostic Procedure Combination database to identify patients with KD treated with IVIG from April 2012 to March 2020. We identified those receiving high- and low-concentration IVIG preparations as an initial treatment. The outcomes included the proportions of patients with coronary artery abnormalities (CAAs) and IVIG resistance, length of stay, and medical costs. Propensity score-matched analyses were conducted to compare the outcomes between the two groups. Instrumental variable analyses were performed to confirm the results.Result: We identified 48,046 patients with KD and created 4:1 propensity score-matched pairs between the low- and high-concentration IVIG groups. There was a significant difference in the percentage with IVIG resistance between the two groups (20.6% vs 24.1%; risk difference, 3.5% [95% confidence interval, 2.3-4.7]; p < 0.001). However, there was no significant difference in CAAs (1.6% vs 1.6%; risk difference, 0.013% [95% confidence interval, −0.34 to 0.37]; p = 0.953). The instrumental variable analyses showed similar results.Conclusions: The proportion of CAAs did not differ significantly between those receiving low- and high-concentration IVIG. To confirm the results of this study, prospective studies adjusting for duration of IVIG administration and duration of observation are needed.

Predictive Value of the Systemic Immune-Inflammation Index for Intravenous Immunoglobulin Resistance and Cardiovascular Complications in Kawasaki Disease

Background: The prediction of intravenous immunoglobulin (IVIG) resistance and cardiovascular complications are critically clinical issues in Kawasaki disease (KD). This prospective study firstly aimed to determine the predictive ability of the systemic immune inflammation index (SII) for IVIG resistance and cardiovascular complications and compare the prognostic accuracy of SII with that of neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR).Methods: Patients with KD were divided into different groups according to the presence of IVIG resistance or cardiovascular complications (coronary artery lesions, valve regurgitation, myocarditis, pericardial effusion, and Kawasaki disease shock syndrome [KDSS]). The clinical and laboratory parameters were compared. Further analysis stratified by platelet level was performed. Multivariate logistic regression analysis was used to identify predictors for IVIG resistance and cardiovascular complications. The receiver operating characteristic (ROC) curve was applied to assess and compare the ability of SII, NLR, and PLR for predicting IVIG resistance and cardiovascular complications.Results: SII was significantly higher in KD patients with IVIG-resistance, myocarditis, valve regurgitation, and KDSS. It was identified as an independent risk factor for IVIG resistance, myocarditis, and valve regurgitation. For KD patients with thrombocytopenia, there were no significant differences in SII between KD patients with IVIG resistance/cardiovascular complications and those without. The best cutoff values of SII for IVIG resistance, myocarditis, valve regurgitation, and KDSS prediction in the whole cohort were 1331.4 × 109, 1368.6 × 109, 1002.4 × 109, and 1485.4 × 109, with a corresponding sensitivity of 0.525, 0.614, 0.754, and 0.670, a specificity of 0.711, 0.723, 0.584, and 0.730, respectively. The predictive value of SII for both IVIG resistance and cardiovascular complications were not superior to that of NLR.Conclusion: Although the parameter of SII may predict IVIG resistance, myocarditis, valve regurgitation, and KDSS in KD as a single parameter, its predictive ability was not good enough and not superior to NLR. SII might not be applicable in patients with KD having thrombocytopenia.

Association Between Serum miR-221-3p and Intravenous Immunoglobulin Resistance in Children With Kawasaki Disease

ObjectivesIntravenous immunoglobulin (IVIG) resistance was a major cause of coronary artery lesions in children with Kawasaki disease (KD). However, the cause of IVIG resistance in KD remains unknown. miR-221-3p has been confirmed involved in cardiovascular diseases and rheumatoid arthritis. The purpose of this study was to investigate the association between miR-221-3p and IVIG resistance in children with KD.Methods55 KD patients and 29 healthy controls (HCs) were enrolled in this study. KD patients were divided into group of sensitive to IVIG (IVIG-response, n=42) and group of resistant to IVIG (IVIG-resistance, n=13). Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) were used to detect the levels of miR-221-3p.ResultsCompared with the HCs group, miR-221-3p were significantly increased in the KD group (p < 0.05), and the IVIG-resistance group had higher levels of miR-221-3p than those in the IVIG-response group (p < 0.05). CRP (C-reactive protein), PCT (procalcitonin), NLR (neutrophil-lymphocyte ratio) were positively correlated with miR-221-3p in KD patients. In addition, the group of IVIG resistance had a higher level of Kobayashi Score (p <0.001). The receiver operating characteristic (ROC) curve showed that miR-221-3p had a better value for diagnosis IVIG resistance in children with KD than Kobayashi Score and the combination of both with the AUC of 0.811 (95% CI, 0.672-0.951), 0.793 (95% CI, 0.618-0.968) and 0.797 (95% CI, 0.619-0.974), respectively.ConclusionsmiR-221-3p might be involved in the pathogenesis of KD and IVIG resistance and miR-221-3p can be used as a new potential biomarker to predict of IVIG resistance in children with KD.

Perspective of Immunopathogenesis and Immunotherapies for Kawasaki Disease

Kawasaki Disease (KD) is an acute inflammatory illness that mostly occurs in children below 5 years of age, with intractable fever, mucocutaneous lesions, lymphadenopathy, and lesions of the coronary artery (CAL). KD is sharing clinical symptoms with systemic inflammatory syndrome in children (MIS-C) which is related to COVID-19. Certain genes are identified to be associated with KD, but the findings usually differ between countries and races. Human Leukocyte Antigen (HLA) allele types and toll-like receptor (TLR) expression are also correlated to KD. The acute hyperinflammation in KD is mediated by an imbalance between augmented T helper 17 (Th17)/Th1 responses with high levels of interleukin (IL)-6, IL-10, IL-17A, IFN-γ, and IP-10, in contrast to reduced Th2/Treg responses with lower IL-4, IL-5, FoxP3, and TGF-β expression. KD has varying phenotypic variations regarding age, gender, intravenous immunoglobulin (IVIG) resistance, macrophage activation and shock syndrome. The signs of macrophage activation syndrome (MAS) can be interpreted as hyperferritinemia and thrombocytopenia contradictory to thrombocytosis in typical KD; the signs of KD with shock syndrome (KDSS) can be interpreted as overproduction of nitric oxide (NO) and coagulopathy. For over five decades, IVIG and aspirin are the standard treatment for KD. However, some KD patients are refractory to IVIG required additional medications against inflammation. Further studies are proposed to delineate the immunopathogenesis of IVIG-resistance and KDSS, to identify high risk patients with genetic susceptibility, and to develop an ideal treatment regimen, such as by providing idiotypic immunoglobulins to curb cytokine storms, NO overproduction, and the epigenetic induction of Treg function.

Useful predictors of Kawasaki disease without complications before initial acute-phase treatment

Kawasaki disease (KD) is an acute febrile systemic vasculitis that primarily affects children younger than 5 years, with coronary artery lesions (CALs) as its severe complications. Intravenous immunoglobulin (IVIG) therapy resistance has been implicated in CAL development, and its known predictors are as follows: Egami score, Kobayashi score, C-reactive protein (CRP), albumin, CRP-to-albumin ratio, and neutrophil-to-lymphocyte ratio (NLR). However, the most useful predictor for IVIG resistance in patients with KD without complications before initial acute-phase treatment remains unclear. Therefore, this study aimed to determine the most useful predictor for IVIG resistance in such patients. This retrospective study included data from 202 patients with KD who underwent acute-phase treatment from January 2009 to March 2021. Among 46 IVIG-resistant patients, 22 patients required rescue therapy (rescued patients), while the remaining 24 received no rescue therapy for resistance and had no CALs. Among the 6 indices, NLR had the highest sensitivity and specificity for the detection of all IVIG-resistant patients and rescued patients (0.724 and 0.728, respectively), and logistic regression analysis showed that the NLR was the sole independent predictor both for the IVIG-resistant patients and for the rescued patients (P < 0.001 and = 0.002, Odds ratio = 5.797 and 5.814, 95% confidence interval = 2.687–12.504 and 1.954–17.299, respectively). NLR was the useful predictor for all IVIG-resistant patients and rescued patients among those with KD without complications before initial acute-phase treatment.