Skip to main content

Predictive model based on gene and laboratory data for intravenous immunoglobulin resistance in Kawasaki disease in a Chinese population

Abstract

Background

Here, we investigated the predictive efficiency of a newly developed model based on single nucleotide polymorphisms (SNPs) and laboratory data for intravenous immunoglobulin (IVIG) resistance in Kawasaki disease (KD) in a Chinese population.

Methods

Data relating to children with KD were acquired from a single center between December 2015 and August 2019 and used to screen target SNPs. We then developed a predictive model of IVIG resistance using previous laboratory parameters. We then validated our model using data acquired from children with KD attending a second center between January and December 2019.

Results

Analysis showed that rs10056474 GG, rs746994GG, rs76863441GT, rs16944 (CT/TT), and rs1143627 (CT/CC), increased the risk of IVIG-resistance in KD patients (odds ratio, OR > 1). The new predictive model, which combined SNP data with a previous model derived from laboratory data, significantly increased the area under the receiver-operator-characteristic curves (AUC) (0.832, 95% CI: 0.776-0.878 vs 0.793, 95%CI:0.734-0.844, P < 0.05) in the development dataset, and (0.820, 95% CI: 0.730-0.889 vs 0.749, 95% CI: 0.652-0.830, P < 0.05) in the validation dataset. The sensitivity and specificity of the new assay were 65.33% (95% CI: 53.5-76.0%) and 86.67% (95% CI: 80.2-91.7%) in the development dataset and 77.14% (95% CI: 59.9-89.6%) and 86.15% (95% CI: 75.3-93.5%) in the validation dataset.

Conclusion

Analysis showed that rs10056474 and rs746994 in the SMAD5 gene, rs76863441 in the PLA2G7 gene, and rs16944 or rs1143627 in the interleukin (IL)-1B gene, were associated with IVIG resistant KD in a Chinese population. The new model combined SNPs with laboratory data and improved the predictve efficiency of IVIG-resistant KD.

Background

Kawasaki disease (KD) is an acute vasculitis of unknown etiology that predominantly occurs in children under 5 years of age. The major complication of KD is coronary artery lesions (CALs) [1], including coronary artery dilatation and coronary aneurysm (CAA); these may subsequently result in long-term sequelae such as coronary stenosis, coronary obstruction, and myocardial infarction. KD has become the leading cause of acquired childhood heart disease in developed countries [2].

A single high dose of intravenous immunoglobulin (IVIG), together with a high dose of acetylsalicylic acid (ASA) is the current standard treatment for acute KD and can effectively reduce CAA from 20 to 25% to 3-5% [3]. However, approximately 10-20% of patients still experience persistent or recurrent fever after standard treatment,which defines IVIG-resistance [4, 5]; Patients not responding to IVIG are at heightened risk for CALs [6]. Therefore, there is an urgent need to develop methods to predict the IVIG therapeutic response before initial treatment in order to improve prognosis by individualized treatment. Several scoring systems comprising clinical features and laboratory data have been proposed to predict IVIG resistant in KD patients representing various geographic locations and ethnicities with good sensitivity and specificity for the respective source populations [7,8,9,10,11]. However, these systems are not sufficient for other populations [12], thus indicating that IVIG resistance might be related to genetic and ethnic factors. Previous genome-wide association studies (GWASs) have demonstrated a correlation between genetic factors and IVIG resistant KD in different ethnic populations [13, 14]. Therefore, the present study was designed to investigate the genes associated with IVIG resistance in a Chinese population and investigate whether combining this genetic information with the existing scoring system could improve the predictive efficacy.

Methods

Patients

We recruited children with KD who had been hospitalized in the cardiology department at the Children’s Hospital Capital Institute of Pediatrics between December 2015 and August 2019. Data acquired from this cohort (the development dataset) were then used to screen target genes and construct the predictive model. Data from a second cohort (the validation dataset) were acquired from children with KD who had been hospitalized at the Beijing Children’s Hospital were then used to validate the model. Our enrollment time was after receiving the first dose of IVIG and treatment failure.

The diagnosis of KD was based on the guidelines proposed by the American Heart Association in 2004 [15], and parents or guardians provided informed consent to participate in the study. Children were diagnosed with KD if they had experienced fever for at least 5 days and fulfilled four or more of the five major clinical features of KD (erythema of the mucosa with a strawberry tongue and cracked lips, bilateral non-purulent conjunctivitis, cervical lymphadenopathy, polymorphous rash, and swelling or redness of the extremities in the acute phase or periungual desquamation in the subacute phase). However, patients who had experienced fever for ≥5 days and had < 4 of the principal features were also diagnosed with KD if coronary aneurysm or dilatation was evident on 2-dimensional echocardiography or coronary angiography. After the diagnosis of KD, all patients were treated with a high dose of IVIG (2 g/kg) and a high dose of aspirin (30-50 mg/kg.d, Tid); these drugs were taken orally within 10 days of disease onset. Patients were excluded from the study if they had comorbidity with another rheumatic or infectious vasculitis, or contraindication to IVIG (i.e., an allergy to IVIG), or if they had an incomplete clinical dataset for the predictive model.

Patients who had experienced fever for at least 5 days and fulfilled the criteria for four or five principal clinical features, or had four principal clinical features along with coronary artery abnormalities on echocardiography, were diagnosed with complete KD (cKD) [15]. Patients who had three principal clinical features, with coronary artery abnormalities on echocardiography, but were devoid of other febrile illnesses were defined as incomplete KD (iKD). Patients who had experienced persistent fever for 36 h (an axillary temperature > 38 °C) after the initial IVIG infusion, or had experienced recrudescent fever within 7 days of IVIG treatment accompanied by at least one of the clinical manifestations of KD except for other reasons, were defined as IVIG-resistant [15]. Due to the costs involved in genetic data analyses, we utilized matching methods in this study. Patients who met the criteria were IVIG-resistance were recruited as an IVIG-resistance group. We also recruited an age- and sex-matched IVIG-response group during the same period of hospitalization (in a 1:2 ratio).

This study was approved by the Ethics Committee of the Capital Institute of Pediatrics (Reference: SHERLL 2015040). Informed consent was obtained from a parent or guardian of each patient prior to the study.

Target gene selection and genotype detection

First, we retrieved literature published before 2015 from the human genome retrieval website (https://www.ncbi.nlm.nih.gov/) provided by the NCBI, and selected 18 SNPs that may be associated with IVIG resistance in patients with KD [16,17,18,19,20,21,22,23,24]. (Supplementary Table 1) We also evaluated the distribution of these SNPs in healthy East Asian populations using the human genome retrieval website. (Supplementary Table 2).

From each patient, we acquired a 2 ml sample of peripheral blood in an EDTA anticoagulant tube following early morning fasting and after the diagnosis of KD but before the initiation of treatment (during the early stages of disease; 6-10 days from the onset of fever). Serum was separated by centrifugation and 200-500 μg of blood cells were stored in a refrigerator at − 80°c for DNA extraction. The genotypes of the 18 SNPs were determined by Sanger sequencing.

DNA extraction and PCR procedures followed the standard operating procedures for one generation sequencing (ABI 3730xl, USA). Sequencing results were compared with a reference sequence using Mutation Surveyor v4.0 software (Softgenetics, USA).

The quality control procedures were carried out in the following steps. A sample of each genotype was randomly selected for DNA sequencing to verify the allele sequence; results were 100% concordant to the initial analysis. All images acquired from the agarose gel electrophoresis of PCR amplicons,and absolute quantification curves for fluorescence data derived from Taqman assays, were conducted by an independent researcher. Furthermore, 5% of samples were randomly selected and run in duplicate to ensure the accuracy of the genotyping.

Clinical data

We collected a range of clinical data, including age, sex, and the laboratory indicators involved in the predictive model established previously, such as the percentage of neutrophils (N%), C reactive protein (CRP), albumin (ALB), sodium ion concentration (Na), and total bilirubin (TBIL). If these laboratory data were performed several times before treatment, the highest value was adopted for N%, CRP, and TBIL, while the lowest value was adopted for ALB and Na.

Statistical analysis

Our research team included a pediatrician and a data analyst. When the patients were admitted, the pediatrician made a diagnosis according to specific diagnostic criteria; the data were collected and analyzed by the data analyst. The outcome assessors were blinded to the groups that the patients had been assigned to.

We used a range of statistical software packages to analyze our data, including SPSS (Version 25.0), R (Version 3.6.0), Haploview (Version 4.2), and MedCalc 11.4.2.0. Data were expressed as Mean ± standard deviation (SD) for continuous variables or a percentage of the total number of patients for categorical variables. The Student’s t-test was used to compare continuous variables and the chi-square test was used to compare categorical data. P < 0.05 was considered to be statistically significant.

SNP association analysis refers to the non-random association of different alleles that are combined as a small module to influence expression; LD gene association analysis is the predominant method used at present [25]. Hardy-Weinberg equilibrium (HWE) analysis was performed for each SNP to determine deviations of the observed genotypic distribution from the expected distribution; this information was analyzed by Haploview software. The genotypic frequency of SNPs was compared using univariable and multivariable logistic regression analysis. SNPs that exhibited a P < 0.01 in the univariate analysis were then included in multivariable analysis as covariates to output SNP loci that could influence the responsiveness of IVIG. In addition, multivariable logistic regression was used to analyze correlations between the genotypes of the SNPs and the risk of IVIG-resistance by adjusting confounders, such as gender or age. We also determined the crude odds ratio (Crude OR), adjusted OR (AOR), and 95% confidence interval (CI).

The established model featured several key laboratory variables [26]. Numerical variables used in the previous model were turned into dichotomous variables by adopting appropriate cut-off points for logistic regression. The SNPs used in the Nomogram were then turned into dichotomous variables according to the type of mutation (wild type vs mutant). Each variable was then assigned score points according to the Nomogram, and the total number of points was calculated for each patient. The final risk score was divided into two risk strata (low-risk or high-risk group). The predictive power of different models was evaluated by creating receiver operator characteristic curves (ROC). The method described by DeLong et al. [27] was then used to compare the area under the ROC curves (AUC) between the new model and the previous models. The goodness-of-fit of the model was evaluated by the Hosmer-Lemeshow test [28]; P > 0.05 indicated a lack of deviation between the model and the observed event rate.

Results

Demographic data

A total of 331 children with KD were enrolled into this study (Fig. 1). In total, 231 cases were enrolled in the development dataset: 77 patients (male/female = 61/16, mean age: 2.28 ± 1.56 years) were placed in the IVIG-resistance group and 154 patients (male/female = 120/34, mean age: 2.31 ± 1.54 years) were placed in the IVIG-response group. A further 100 patients were enrolled in the validation dataset: 35 patients (male/female = 25/10, mean age: 2.51 ± 2.06 years) were placed in the IVIG-resistance group and 65 patients (male/female = 45/20, mean age:1.95 ± 1.65 years) were placed in the IVIG-response group (Table 1).

Fig. 1
figure1

Flow chart of enrolled participants

Table 1 General clinical data of KD children

LD and HWE analysis

Haploview was used to analyze the 18 identified SNPs, as shown in Fig. 2. This analysis revealed that the rs16944 and rs1143627 loci in the interleukin (IL)-1B gene were in complete LD (D′ = 1.0, r2 = 0.974) and that the rs10056474 and rs746994 loci in the SMAD5 gene were in incomplete LD (D′ = 1.0, r2 = 0.222). Further analysis revealed that the rs403016 and rs447536 loci were not in HWE (P < 0.001). Some mutations were detected in rs3219018, rs780467580, and rs333, with a minor allele frequency (MAF) < 0.01 [29]).

Fig. 2
figure2

Results of linkage disequilibrium analysis

Logistic regression analysis of high-risk SNPs for IVIG-resistance

According to LD and HWE analysis, rs3219018, rs780467580, and rs333, loci were excluded due to the presence or absence of rare mutations (MAF < 0.01). We also excluded rs403016 and rs447536 loci owing to the fact that these were not in HWE. After adjusting for these factors, a total of 13 SNPs were used for univariate analysis.

Furthermore, a comparison of genotype frequencies showed that cases of KD with a SMAD5-rs10056474 GG genotype had a 2.459-fold higher risk of IVIG-resistance than those with CC/CG genotypes (OR, 2.459; 95%CI, 1.185-5.101) (Table 2).

Table 2 Comparison of distribution of 13 SNPs genotypes between IVIG-resistance and IVIG-response group

SNPs that exhibited a P < 0.01 in the univariate analysis (rs10056474, rs76863441, rs16944, rs114362, and rs396991) were then included in multivariable analysis. rs746944 was included as an interactive variable for rs10056474 due to incomplete LD (D′ = 1.0, r2 = 0.222). Furthermore, because there was a complete LD between rs16944 and rs1143627 (D′ = 1.0, r2 = 0.974), neither rs16944 or rs1143627 was included. Finally, 5 SNPs (rs10056474, rs746994, rs76863441, rs16944 (or rs1143627), and rs396991) were included in the multivariable logistic analysis. Results showed that the rs10056474 and rs746994 loci of the SMAD5 gene, the rs76863441 locus of the PLA2G7 gene, and the rs16944 or rs1143627 locus of the IL-1B gene, were correlated with IVIG-resistance in children with KD. (Table 3).

Table 3 Multivariable regression analysis of IVIG-resistance of KD

A predictive model for IVIG-resistance

In the present study, five SNPs were found to be associated with IVIG-resistance. These SNPs, along with the laboratory indicators used in the previous model [26] (Table 4) were used as variables to create a new model. In total, data from 225 KD patients (75 IVIG-resistance, 150 IVIG-response) were used to construct a new predictive model; 6 patients had been excluded due to incomplete laboratory data. Each variable was scored using a Nomogram (Fig. 3). A total score > 12.5 points was considered high-risk for IVIG resistance; the AUC was 0.832 (95%CI: 0.776-0.878) and the sensitivity and specificity were 65.33% (95%CI:53.5-76.0%) and 86.67% (95%CI:80.2-91.7%), respectively. There was no significance with regards to the Hosmer-Lemeshow statistic (P = 0.585 > 0.05).

Table 4 Old model used to predict IVIG resistance in patients with KD [26]
Fig. 3
figure3

Nomogram. (rs16944 CT/TT: 2 points (rs1143627 CC/CT: 2 points); rs746994 GG: 2 points; rs76863441 GT: 3 points; rs10056474 GG: 5 points; CRP ≥ 90 mg/L: 1.5 points; TBIL> 20 μmol/L: 2 points; ALB< 35 g/L: 4.5 points; Na < 135 mmol/L: 4.5points; N% ≥70%: 5 points)

Validation and comparison of the model

Data from patients hospitalized at Beijing Children’s Hospital were then used to validate the newly constructed model. The AUC was 0.820 (95% CI: 0.730-0.889) while the sensitivity and specificity were 77.14% (95% CI: 59.9-89.6%) and 86.15% (95% CI: 75.3-93.5%), respectively.

Comparisons of the new and previously established. Model showed that the AUC was higher than the previous model in the development dataset (0.832, 95% CI: 0.776-0.878 vs 0.793, 95% CI: 0.734-0.844, Z = 2.316, P = 0.021) and in the validation dataset (0.820, 95% CI: 0.730-0.889 vs 0.749, 95% CI: 0.652-0.830; Z = 2.145, P = 0.032) (Fig. 4).

Fig. 4
figure4

ROC curves of the two models. (A: ROC curves in development dataset; B: ROC curves in validation dataset)

Discussion

The present study showed that SNPs in the rs10056474 and rs746994 loci of the SMAD5 gene, in the rs76863441 loci of the PLA2G7 gene, and the rs16944 or rs1143627 loci in the IL-1B gene, were all associated with IVIG-resistance in children with KD in a Chinese population. The SNPs of these genes can be sequenced quickly by the Sanger method, which is convenient and cheap for clinical application. The results of Sanger sequencing to identify the SNPs in the text can be completed in our laboratory, and it takes about 4 h from specimen collection to experimental results. Moreover, the new predictive model, based on gene polymorphisms and laboratory data, showed improved predictive efficacy for IVIG-resistance and may help pediatricians to identify high-risk IVIG-resistant patients prior to the initiation of treatment in order to improve the prognosis of patients with KD [30]. With regards to the high-risk patients with IVIG resistance that were selected by the predictive model, we preferred to recommend the addition of corticosteroids or infliximab into their treatment plans; this strategy was based on the 2017 AHA scientific statement and clinical practice in our center [4]. Other alternative adjunctive therapies, such as cyclosporine, cyclophosphamide, and plasma exchange, have been recommended in the literature [4].

.The pathogenesis of IVIG-resistance in children with KD has yet to be fully elucidated. Numerous studies have reported that IVIG-resistant KD is associated with genetic background. However, most of the studies that were reported previously focused on one gene or several genes within a particular pathway in different populations; few studies investigated multiple genes within the same population [31]. Studies have also shown that multiple gene associations may be of greater predictive value than a single gene association [32]. In this study, we selected multiple genes that were possibly related to IVIG-resistance that had been reported in different ethnic populations. We used these to evaluate the role of gene polymorphisms in the therapeutic response to IVIG in a Chinese population.

This study showed that two loci (rs10056474 and rs746994) in the SMAD5 gene were associated with IVIG-resistance in children with KD. rs10056474 GG was assigned the highest score in the predictive model; this may indicate that the SMAD5 gene plays an important role in IVIG-resistance. SMAD5, as an intracellular mediator of the transforming growth factor (TGF)-β signal transduction pathway, is involved in the induction of angiogenesis, cardiomyocyte hypertrophy, calcification, and fibrosis in the cardiovascular system [33]. Several studies have shown that the SMAD protein family is correlated with KD susceptibility and therapeutic response; however, these studies mainly focused on the SMAD3 and SMAD4 genes [23]. Some previous studies on SMAD5 gene polymorphisms in KD demonstrated that these variations were associated with KD susceptibility rather than IVIG therapeutic response [34]. The rs10056474 and rs746994 loci of the SMAD5 gene were related to IVIG-resistance and located in the introns of the SMAD5 gene. This may be associated with the fact that SMAD5 can be activated by bone morphology proteins (BMPs) in the TGF-β pathway to participate in angiogenesis. However, our present study did not find that any of the genes in the TGF-β pathway, except for SMAD5, were associated with IVIG-resistance; these findings are consistent with a previous study carried out in Taiwan [35].

.The rs76863441 loci of the PLA2G7 gene were associated with IVIG-resistant KD in this study. Platelet-activating factor acetylhydrolase (PAF-AH), also known as lipoprotein-associated phospholipase A2 (Lp-PLA2), is encoded by the PLA2G7 gene and maintained stable by PAF (a form of inflammatory factor) at appropriate levels to play an anti-inflammatory role [36]. Previous studies have shown that PLA2G7 gene polymorphisms determine plasma PAF-AH activity [37]. In the present study, we found that IVIG resistance in the rs76863441GT genotype was significantly higher than that in the rs76863441GG genotype; this was consistent with previous observations reported by Minami et al. which stated that IVIG-resistance in the GT + TT genotype was significantly higher than that in the GG genotype [22]. When the G base is replaced by the T base, resulting in a change in the genetic code from GUU to UUU, and substitution of valine (Val) with phenylalanine (Phe) at position 279, there was a clear reduction in PAF-AH activity; subsequent increases in the level of PAF level, and the activation of production for other inflammatory mediators [38], may underlie the poor therapeutic response to IVIG in KD.

The two SNPs (rs16944, rs114362) in the IL-1B gene were found to be related to the therapeutic response to IVIG in this study. The resistance of IVIG in the rs16944CT/TT genotype was significantly higher than that in the CC genotype and the resistance of IVIG in the rs1143627CT/CC genotype was significantly higher than that in the TT genotype; these findings were consistent with those reported previously by Taiwanese scholars, who stated that the rs16944TT and rs1143627CC genotypes were associated with a significant increase in the risk of IVIG resistance [21]. We also found that the rs16944 and rs1143627 SNPs had strong LD, as observed previously [39]; consequently, either rs16944 or rs1143627 could exhibit biological activity at these two loci. IL-1B is an inflammatory cytokine produced by activated macrophages and plays a key role in the pathogenesis of KD. Both the rs16944 and rs11436227 SNPs were located in the promoter region of IL-1B which regulates the transcription of IL-1B. Children with the rs16944CT/TT or rs1143627CT/CC genotype may exhibit increased IL-1B levels; this gives rise to the chemotaxis of leukocytes and the induction of neutrophils to release large amounts of cytokines, thus resulting in a poor response to IVIG treatment [40]. A case report on the efficacy of an IL-1 receptor antagonist (Anakinra) on IVIG-resistant KD patients [41], along with ongoing clinical trials involving the blockade of IL-1 or the treatment of acute KD [42], further confirmed this conjecture.

There are some limitations to this study that need to be considered. First, because of the high costs associated with SNP testing, we used a ratio of 1:2 to match IVIG-resistant versus IVIG-responsive cases. Thus, 85.6% (77/90) of IVIG-resistant cases and 17.6% (154/877) of IVIG-responsive cases were included. The much lower proportion of IVIG responsive patients may have led to selection bias. However, the general characteristics between the matched and non-matched children in the response population were compared and there was no statistical difference (Supplementary Table 3). Second, although the sample size for the development model was small and derived from a single center, the center involved is one of the largest children’s hospitals in the Beijing area and patients attend this hospital from all over the country. In the present study, 90% of the patients were from northern China while 10% were from southern China; they were all Han Chinese. More importantly, the quality of our study was guaranteed by the same brand of IVIG that was used in the center; the treatment response was observed by the same researchers in accordance with consistent standards. Third, although we validated our model at another center with good results, it is possible that there may be issues related to external validity or generalizability. Furthermore, our study population only featured Chinese patients; whether the results of this study are applicable to other ethnicities needs further investigation and validation. Overall, our findings need to be validated in a larger dataset in future. Finally, our patients were recruited from the two largest children’s hospitals in the Beijing area in which the children have a more serious condition. Consequently, our prediction model may only be applicable to medical institutions that are similar to these two children’s hospitals.

In summary, rs10056474 and rs746994 SNPs in the SMAD5 gene, the rs76863441 SNP in the PLA2G7 gene, and the rs16944 or rs1143627 SNPs in the IL-1B gene, were all associated with IVIG resistant KD in the Chinese population. Our new model combines SNPs with laboratory data and improved the prediction efficiency of IVIG-resistant KD. However, whether our model can be generalized to other ethnic populations requires further study.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and its supplementary files.

Abbreviations

AHA:

American Heart Association

ALB:

Albumin

AOR:

Adjusted odds ratio

ASA:

Acetysalicylic acid

AUC:

Area under the curve

BMPs:

Bone morphology proteins

CALs:

Coronary artery lesions

CAA:

Coronary aneurysm

CI:

Confidence interval

c-KD:

Complete Kawasaki disease

CRP:

C-Reactive protein

GWAS:

Genome-wide association study

HWE:

Hardy-Weinberg equilibrium

IL:

Interleukin

iKD:

Incomplete Kawasaki disease

IVIG:

Intravenous immunoglobulin

KD:

Kawasaki disease

LD:

Linkage disequilibrium

Lp –PLA:

Lipoprotein-associated phospholipase A

Na:

Sodium ion concentration

OR:

Odds ratio

PAF-AH:

Platelet-activating factor acetylhydrolase

ROC:

Receiver operator characteristic

SNP:

Single nucleotide polymorphism

TBIL:

Total bilirubin

TGF:

Transforming Growth Factor

References

  1. 1.

    Ae R, et al. Epidemiology, treatments, and cardiac complications in patients with Kawasaki disease: the Nationwide survey in Japan, 2017-2018. J Pediatr. 2020;225:23.

    Article  Google Scholar 

  2. 2.

    Burns JC, Glode MP. Kawasaki syndrome. Lancet. 2004;364(9433):533–44.

    Article  Google Scholar 

  3. 3.

    Bar-Meir M, et al. Prediction of resistance to intravenous immunoglobulin in children with Kawasaki disease. J Pediatric Infect Dis Soc. 2018;7(1):25–9.

    PubMed  Google Scholar 

  4. 4.

    McCrindle BW, et al. Diagnosis, treatment, and long-term Management of Kawasaki Disease: a scientific statement for health professionals from the American Heart Association. Circulation. 2017;135(17):e927–99.

    Article  Google Scholar 

  5. 5.

    Son MB, et al. Treatment of Kawasaki disease: analysis of 27 US pediatric hospitals from 2001 to 2006. Pediatrics. 2009;124(1):1–8.

    Article  Google Scholar 

  6. 6.

    Tremoulet AH, et al. Resistance to intravenous immunoglobulin in children with Kawasaki disease. J Pediatr. 2008;153(1):117–21.

    CAS  Article  Google Scholar 

  7. 7.

    Egami K, et al. Prediction of resistance to intravenous immunoglobulin treatment in patients with Kawasaki disease. J Pediatr. 2006;149(2):237–40.

    CAS  Article  Google Scholar 

  8. 8.

    Kobayashi T, et al. Prediction of intravenous immunoglobulin unresponsiveness in patients with Kawasaki disease. Circulation. 2006;113(22):2606–12.

    Article  Google Scholar 

  9. 9.

    Uehara R, et al. Analysis of potential risk factors associated with nonresponse to initial intravenous immunoglobulin treatment among Kawasaki disease patients in Japan. Pediatr Infect Dis J. 2008;27(2):155–60.

    Article  Google Scholar 

  10. 10.

    Lin MT, et al. Risk factors and derived Formosa score for intravenous immunoglobulin unresponsiveness in Taiwanese children with Kawasaki disease. J Formos Med Assoc. 2016;115(5):350–5.

    CAS  Article  Google Scholar 

  11. 11.

    Fu PP, Du ZD, Pan YS. Novel predictors of intravenous immunoglobulin resistance in Chinese children with Kawasaki disease. Pediatr Infect Dis J. 2013;32(8):e319–23.

    Article  Google Scholar 

  12. 12.

    Song R, Yao W, Li X. Efficacy of four scoring Systems in Predicting Intravenous Immunoglobulin Resistance in children with Kawasaki disease in a Children's Hospital in Beijing, North China. J Pediatr. 2017;184:120–4.

    Article  Google Scholar 

  13. 13.

    Lou J, et al. Systematic confirmation study of GWAS-identified genetic variants for Kawasaki disease in a Chinese population. Sci Rep. 2015;5:8194.

    CAS  Article  Google Scholar 

  14. 14.

    Kuo HC, et al. Prediction for Intravenous Immunoglobulin Resistance by Using Weighted Genetic Risk Score Identified From Genome-Wide Association Study in Kawasaki Disease. Circ Cardiovasc Genet. 2017;10(5):e001625.

    CAS  Article  Google Scholar 

  15. 15.

    Newburger JW, et al. Diagnosis, treatment, and long-term management of Kawasaki disease: a statement for health professionals from the committee on rheumatic fever, endocarditis and Kawasaki disease, council on cardiovascular disease in the youn, American Heart Association. Circulation. 2004;110(17):2747–71.

    Article  Google Scholar 

  16. 16.

    Mamtani M, et al. Association of CCR2-CCR5 haplotypes and CCL3L1 copy number with Kawasaki disease, coronary artery lesions, and IVIG responses in Japanese children. PLoS One. 2010;5(7):e11458.

    Article  Google Scholar 

  17. 17.

    Onouchi Y, et al. ITPKC and CASP3 polymorphisms and risks for IVIG unresponsiveness and coronary artery lesion formation in Kawasaki disease. Pharmacogen J. 2013;13(1):52–9.

    CAS  Article  Google Scholar 

  18. 18.

    Shrestha S, et al. Functional FCGR2B gene variants influence intravenous immunoglobulin response in patients with Kawasaki disease. J Allergy Clin Immunol. 2011;128(3):677–80.

    CAS  Article  Google Scholar 

  19. 19.

    Khor CC, et al. Genome-wide association study identifies FCGR2A as a susceptibility locus for Kawasaki disease. Nat Genet. 2011;43(12):1241–6.

    CAS  Article  Google Scholar 

  20. 20.

    Shrestha S, et al. Role of activating FcγR gene polymorphisms in Kawasaki disease susceptibility and intravenous immunoglobulin response. Circ Cardiovasc Genet. 2012;5(3):309–16.

    CAS  Article  Google Scholar 

  21. 21.

    Weng KP, et al. IL-1B polymorphism in association with initial intravenous immunoglobulin treatment failure in Taiwanese children with Kawasaki disease. Circ J. 2010;74(3):544–51.

    CAS  Article  Google Scholar 

  22. 22.

    Minami T, et al. A polymorphism in plasma platelet-activating factor Acetylhydrolase is involved in resistance to immunoglobulin treatment in Kawasaki disease. J Pediatr. 2005;147(1):78–83.

    CAS  Article  Google Scholar 

  23. 23.

    Shimizu C, et al. Transforming growth factor-beta signaling pathway in patients with Kawasaki disease. Circulation. Cardiovasc Genet. 2011;4(1):16–U99.

    CAS  Article  Google Scholar 

  24. 24.

    Yang J, et al. The correlation between Kawasaki disease and polymorphisms of tumor necrosis factor alpha and interleukin-10 gene promoter. Zhonghua Er Ke Za Zhi. 2003;41(8):598–602.

    PubMed  Google Scholar 

  25. 25.

    Slatkin M. Linkage disequilibrium — understanding the evolutionary past and mapping the medical future. Nat Rev Genet. 2008;9(6):477–85.

    CAS  Article  Google Scholar 

  26. 26.

    Yang S, et al. Predictive tool for intravenous immunoglobulin resistance of Kawasaki disease in Beijing. Arch Dis Child. 2019;104(3):262–7.

    Article  Google Scholar 

  27. 27.

    DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837–45.

    CAS  Article  Google Scholar 

  28. 28.

    Hosmer DW, et al. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med. 1997;16(9):965–80.

    CAS  Article  Google Scholar 

  29. 29.

    Auer PL, Lettre G. Rare variant association studies: considerations, challenges and opportunities. Genome Med. 2015;7(1):16.

    Article  Google Scholar 

  30. 30.

    Chen L, et al. Prediction for intravenous immunoglobulin resistance combining genetic risk loci identified from next generation sequencing and laboratory data in Kawasaki disease. Front Pediatr. 2020;8:462367.

    Article  Google Scholar 

  31. 31.

    Wright VJ, et al. Diagnosis of Kawasaki disease using a minimal whole-blood gene expression signature. JAMA Pediatr. 2018;172(10):e182293.

    Article  Google Scholar 

  32. 32.

    Kuo HC, et al. A replication study for association of ITPKC and CASP3 two-locus analysis in IVIG unresponsiveness and coronary artery lesion in Kawasaki disease. PLoS One. 2013;8(7):e69685.

    CAS  Article  Google Scholar 

  33. 33.

    Euler-Taimor G, Heger J. The complex pattern of SMAD signaling in the cardiovascular system. Cardiovasc Res. 2006;69(1):15–25.

    CAS  Article  Google Scholar 

  34. 34.

    Cho JH, et al. Genetic polymorphism of SMAD5 is associated with Kawasaki disease. Pediatr Cardiol. 2014;35(4):601–7.

    Article  Google Scholar 

  35. 35.

    Kuo HC, et al. Polymorphisms of transforming growth factor-beta signaling pathway and Kawasaki disease in the Taiwanese population. J Hum Genet. 2011;56(12):840–5.

    CAS  Article  Google Scholar 

  36. 36.

    Dohi T, et al. Higher lipoprotein-associated phospholipase A2 levels are associated with coronary atherosclerosis documented by coronary angiography. Ann Clin Biochem. 2012;49(Pt 6):527–33.

    CAS  Article  Google Scholar 

  37. 37.

    Stafforini DM, et al. Platelet-activating factor acetylhydrolase deficiency. A missense mutation near the active site of an anti-inflammatory phospholipase. J Clin Invest. 1996;97(12):2784–91.

    CAS  Article  Google Scholar 

  38. 38.

    Stafforini DM. PAF acetylhydrolase gene polymorphisms and asthma severity. Pharmacogenomics. 2001;2(3):163–75.

    CAS  Article  Google Scholar 

  39. 39.

    Wang X, et al. Interleukin-1beta-31C/T and -511T/C polymorphisms were associated with preeclampsia in Chinese Han population. Plos One. 2014;9(9):e106919.

  40. 40.

    Fu LY, et al. The IL-1B gene polymorphisms rs16944 and rs1143627 contribute to an increased risk of coronary artery lesions in southern Chinese children with Kawasaki disease. J Immunol Res. 2019;2019:4730507.

  41. 41.

    Sánchez-Manubens J, et al. A child with resistant Kawasaki disease successfully treated with anakinra: a case report. Bmc Pediatr. 2017;17(1).

  42. 42.

    Burns JC, et al. Review: Found in Translation: International Initiatives Pursuing Interleukin-1 Blockade for Treatment of Acute Kawasaki Disease. Arthritis Rheumatol. 2017;69(2):268–76.

Download references

Acknowledgments

We thanks to Ms. Han Bao and Mr. Shuai Yang collected the data.

Funding

1.”Peak Climbing” Talents Development Program of Beijing Hospital Authority (DFL20181301); 2. Science Foundation for Clinical Technical Innovation Project of Beijing Municipal Administration of Hospital (XMLX201612); 3. Key Project of Capital Clinical Characteristic Application Research (Z181100001718189).

Author information

Affiliations

Authors

Contributions

Xiao-hui Li conceptualized and designed the study, contributed to the interpretation of the results, reviewed and revised the manuscript. Yue Yuan contributed to the interpretation of the results and critical revision of the manuscript. Qian Jiang conceptualized and designed the study, contributed to the interpretation of the results. Li Meng and Zhen Zhen completed the data analyses and drafted the manuscript. Wei Yao, Ming-ming Zhang, Ai-jie Li collected and supervised data. Lin Shi contributed important intellectual content. Each author has agreed with the submission of this version of the manuscript.

Corresponding author

Correspondence to Xiao-hui Li.

Ethics declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the Capital Institute of Pediatrics (No. SHERLL 2015040). Informed consent was obtained from a parent or guardian of each patient before the study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Meng, L., Zhen, Z., Jiang, Q. et al. Predictive model based on gene and laboratory data for intravenous immunoglobulin resistance in Kawasaki disease in a Chinese population. Pediatr Rheumatol 19, 95 (2021). https://doi.org/10.1186/s12969-021-00582-6

Download citation

Keywords

  • Kawasaki disease
  • Intravenous immunoglobulin resistance
  • Single nucleotide polymorphism