Impact factors for refractory KD and comparison of different treatment protocols

Intravenous immunoglobulin (IVIG) has been widely accepted as a standard treatment against Kawasaki disease (KD). About 20% of patients are nonresponsive after the rst IVIG treatment.This study was performed to reveal the responsiveness, mechanism of action, and incidence rate of coronary artery lesions in the long-term follow-up visit of action of IVIG combined with glucocorticoids in the treatment of refractory KD. Method: A total of 633 pediatric patients with KD (experimental group) and 200 pediatric patients with upper respiratory infection hospitalized during the same period (control group) were retrospective selected. KD was treated with IVIG combined with aspirin.IVIG-nonresponsive group was given the second treatment.Protocol 1 group was administered 1 g/kg IVIG; Protocol 2 group was administered 2 g/kg IVIG; and Protocol 3 group was administered 1 g/kg IVIG + glucocorticoid (GC). Following the second treatment, the responsive group (Group A) and nonresponsive group (Group B) were classied based on whether the temperature was reduced to normal.

second IVIG treatment, ulinastatin (UTI) treatment, methotrexate (MTX) treatment, cyclosporin treatment, in iximab (IFX) treatment, and plasmapheresis [2,3]. However, no agreement has been reached on the preferred second-line drugs for such pediatric patients in the community. In US clinical practice, about 7.7-18% of pediatric patients with KD nonresponsive to the rst IVIG treatment were given GC [4]. In another Spanish KD cohort study, about 14.5% of pediatric patients with KD were given GC in the rst or second treatment [4]. In China, GC was also used in the rst and second treatment of KD [2,5]. Hence, the treatment outcome of GC in KD has aroused immense attention at home and abroad. However, its responsiveness, mechanism of action, and incidence rate of CAL in the long-term follow-up visit remain unclear. This study was performed to reveal the application value and mechanism of action of IVIG combined with glucocorticoids in the treatment of refractory KD.

Participants
This retrospective study included 633 pediatric patients with KD who underwent standard IVIG treatment in the hospital from January 2016 to March 2019. The inclusion criterion was pediatric patients with KD having a de nitive diagnosis and complete data, referring to the diagnostic criteria proposed by American Heart Association 2017 [6]. Patients complicated with other autoimmune diseases, infectious diseases, and tumors were excluded. Further, 200 pediatric patients with an upper respiratory infection and hospitalized in the hospital during the same period were considered as the control group. Two groups of pediatric patients were not given hormones, immunosuppressants, or IVIG within 2 weeks. The experimental group received standard IVIG treatment combined with aspirin and was subdivided into IVIG-responsive group and IVIG-nonresponsive group based on the response to IVIG treatment (IVIG nonresponsive was de ned as persistent fever of >38℃ for 36 h after the rst-dose IVIG treatment, abatement of fever for 2-7 days after drug administration, fever recurring after 2 weeks, with at least one of KD symptoms). The IVIG-nonresponsive group received the second treatment and then was distributed into Protocol 1 group (20 participants), Protocol 2 group (36 participants), and Protocol 3 group (30 participants). Following the second treatment, the IVIG-nonresponsive group was further classi ed into responsive group (Group A) and nonresponsive group (Group B), depending on whether the fever was abated. The family members of all pediatric patients included in this study signed the informed consent,

Methods
General data: The clinical and experimental data of all the pediatric patients with KD during hospitalization and follow-up visit were recorded. The blood drawn from the experimental and control groups when admitted to the hospital was retained for testing the serum levels of Meprin A, interleukin-1β (IL-1β), and interferon-induced protein-10 (IP-10). White blood cell (WBC) count, erythrocyte sedimentation rate (ESR), and levels of C-reactive protein (CRP), interleukin-6 (IL-6), and interleukin-17A (IL-17A) were tested during hospitalization. The time of fever abatement (the time from the day of second treatment to the resumption of normal body temperature without a recurrent fever) and the experimental data before and after the second treatment (pretreatment: prior to the second IVIG treatment, i.e., within 3-5 days after the rst IVIG treatment; post-treatment: within 3-5 days after the second IVIG treatment), including WBC count, ESR, and levels of CRP,Meprin A, IL-1β, IL-6, IL-17A, and IP-10, were recorded for pediatric patients with KD in the IVIG-nonresponsive group. All the pediatric patients with KD in these groups were subjected to an ultrasonic cardiogram test from the onset to the 1-year follow-up visit (1, 2, 4, 8, and 12 weeks, and 6 and 12 months after the onset). The ndings suggested coronary artery dilation or coronary aneurysm, de ned as CAL. The diagnostic criterion for coronary artery dilation was as follows: coronary artery diameter >3 mm if participants aged <5 years; and coronary artery diameter >4 mm if participants aged >5 years. The diagnostic criterion for coronary artery aneurysm was as follows: coronary artery localized or diffused dilation, and its diameter 1.5 times higher compared with the adjacent normal coronary artery. Small coronary aneurysm was identi ed if the inner diameter was 4.1-4.9 mm; medium coronary aneurysm was identi ed if the inner diameter was 5.0-7.9 mm; and huge coronary aneurysm was identi ed if the inner diameter was ≥8 mm. Post-treatment CALs included CALs appearing during the follow-up visit after the second treatment, excluding those found before the second treatment.
For serum meprin A, IL-1β, and IP-10 test, 4 mL of fasting venous blood was drawn from all pediatric patients with KD during the acute phase (course of disease: 3-7 days), 4 mL from the IVIG-nonresponsive group within 3-5 days after the rst IVIG treatment, and 4 mL from the IVIG-nonresponsive group within 3-5 days after the second IVIG treatment. The blood was then centrifuged at 3000 rpm for 10 min. The serum was separated from the residue and stored in a refrigerator at -80℃ for testing. Then, 4 mL of fasting venous blood was drawn from all pediatric patients in the control group and processed in the same manner as for the KD group. The serum was collected, and the levels of Meprin A, IL-1β, and IP-10 s were detected using Meprin A ELISA kit, IL-1β ELISA kit, and IP-10 kit, respectively (all supplied by BD Biosciences .

Treatment protocols
Standard protocol: All pediatric patients with KD were given IVIG 2 g/kg once, combined with 30-50 mg/(kg × day) aspirin. The dosage of aspirin was reduced to 3-5 mg/kg 3 days after the abatement of fever.
Second treatment protocol: IVIG-nonresponsive group: Protocol 1 group was given IVIG 1 g/kg once again; Protocol 2 group was given IVIG 2 g/kg once again; Protocol 3 group was given IVIG 1 g/kg + Solu-Medrol injection 2 mg/kg (intravenous ×3 days, changed to oral administration of prednisone tablets, and dosage gradually reduced until withdrawn, with a cycle of 1-2 weeks).
Third treatment protocol: If Protocol 1 group and Protocol 2 group were nonresponsive after the second treatment: Solu-Medrol injection 2 mg/kg was given (intravenous ×3 days, changed to oral administration of prednisone tablets, and dosage gradually reduced until withdrawn, with a cycle of 1-2 weeks; if Protocol 3 was nonresponsive, IVIG 1g/kg was administered.
The body temperatures of all the pediatric patients with KD were under control after the third treatment.
All patients were given immunoglobulins produced by the same company (Chengdu Rongsheng Pharmaceuticals Co., Ltd., under State Food and Drug Administration (SFDA) Approval No.: S19993042, 50 mL: 2.5 g/vial), aspirin enteric-coated tablets (Cisen Pharmaceutical Co., Ltd., under SFDA Approval Statistical analysis SPSS19.0 was used for statistical analysis. The normal distribution data were expressed as mean ± standard deviation and checked using the independent-samples t test. The data with non-normal distribution were expressed as median (interquartile range) and compared using the Wilcoxon rank-sum test. The counting data were expressed as a percentage (%) and processed using the chi-square test. P value less than 0.05 was considered to be statistically signi cant.

Results
Comparison of pretreatment clinical and lab indexes between the IVIG-responsive group, IVIG-nonresponsive group, and control group Comparison between the three groups: The IVIG-responsive group comprised 574 out of 633 patients with KD (86.4%). The IVIG-nonresponsive group comprised 86 patients (13.6%). The control group comprised 200 participants. No statistically signi cant difference in age and sex was found between the three groups (P>0.05). The WBC count, ESR, platelet (PLT) count, and serum levels of CRP, Meprin A, IL-1β, IL-6, IL-17A, and IP-10 were higher in the IVIG-responsive and nonresponsive groups than in the control group, and the difference was statistically signi cant (P < 0.001). More details are shown in Table 1.
Comparison between the IVIG-nonresponsive and IVIG-responsive groups: The CAL incidence rate was higher in the IVIG-nonresponsive group (58.1%) than in the IVIG-responsive group (12.2%; P<0.05). The IVIG-nonresponsive group was found with longer time of rst application of IVIG (P<0.001) and higher WBC count, ESR, and levels of CRP, Meprin A, IL-1β, IL-6, IL-17A, and IP-10 in the serum, with statistically signi cant differences (P<0.001). More details are shown in Table 1.
Comparison of clinical and lab indexes before and after treatment between three protocols Comparison of Protocol 3 group with Protocol 1 and 2 groups: The differences in time of rst application of IVIG and lab indexes in every acute phase were not statistically signi cantly different between the three groups (P>0.05). In Protocol 3 group, the average time of fever abatement after the second treatment was 1 day, 96.7% of patients had fever abatement (responsive rate), and the CAL incidence rate after treatment was 10%; these differences were statistically signi cant compared with Protocol 1 and 2 groups (P<0.05). The differences in WBC count, ESR, and serum levels of CRP, Meprin A, IL-1β, IL-6, IL-17A, and IP-10 were statistically signi cant between Protocol 3 and 1 groups (P<0.05). The differences in WBC count, ESR, and levels of CRP and Meprin A were statistically signi cant between Protocol 3 and 2 groups. More details are given in Table 2.
Comparison between Protocol 1 and 2 groups: The responsive rate (77.8%) was signi cantly higher in Protocol 2 group than in Protocol 1 group (50.0%). The CAL incidence rate (22.2%) after treatment was signi cantly lower in Protocol 2 group than that in Protocol 1 group (55.0%). The extent of decrease in serum levels of Meprin A, IL-1β, and IL-6 after treatment in Protocol 2 group was more obvious, and the differences were statistically signi cant (P < 0.05). More details are shown in Table 2.
Comparison of clinical and lab indexes after the second treatment between responsive (Group A) and nonresponsive groups (Group B) The 86 IVIG-nonresponsive patients with KD included 67 (77.9%) in the second treatment responsive group(Group A) and 19 (22.1%) in the second treatment nonresponsive group(Group B) . In Group B, 18 (94.7%) were given IVIG-alone treatment, and 1 (5.3%) was given IVIG combined with glucocorticoids. The difference in age and sex was not statistically signi cant between the two groups (P >0.05). In Group B after the second treatment, the CAL incidence rate was up to 78.9%, which was signi cantly higher than that in Group A (P<0.001). The differences in serum levels of Meprin A, IL-1β, and IL-17A were statistically signi cant between the two groups (P<0.05), but the differences in WBC count, Hb, PLT count, ESR, and levels of CRP, IL-6, and IP-10 were not statistically different (P>0.05). All lab indexes were tested 3-5 days after the second treatment. More details are shown in Table 3.

Discussion
KD is acute, systemic, nonspeci c, self-limiting vasculitis. Its adverse effects depend mainly on the incidence of CALs, including coronary artery dilation, coronary aneurysm, thrombosis, myocardial ischemia, and even sudden death [7,8]. Intravenous immunoglobulin (IVIG) is one of the rst-line drugs used to treat KD. However, the nonresponsive rate for KD treatment with IVIG continued to increase every year, and the exposure rate to CALs was high [9,10]. In this study, the nonresponsive rate for the rst IVIG treatment was 13.6% and the CAL rate was up to 58.1%, which were signi cantly higher than those in the rst IVIG-responsive group. In the second treatment, the nonresponsive rate for the treatment with IVIG 1 g/kg alone was up to 50.0%, and the CAL rate during the follow-up visit after the second treatment was 55.0%. The nonresponsive rate for the treatment with IVIG 2 g/kg alone was still up to 22.2%, and the CAL rate during the follow-up visit after the second treatment remained 22.2%. In summary, the response of IVIG 2 g/kg to refractory KD was con rmed compared with IVIG 1 g/kg, but with a certain limitation. However, the disease mechanism of refractory KD remains unclear to date. Compared with the IVIGresponsive and control groups, the serum levels of Meprin A, IL-1β, IL-6, IL-17A, and IP-10 signi cantly increased in the IVIG-nonresponsive group (P<0.05). After the second treatment, the serum levels of Meprin A, IL-1β, and IL-17A were still signi cantly higher in the second nonresponsive group than in the second responsive group (P<0.05). These ndings suggested that the serum levels of Meprin A, IL-1β, IL-17A, and other factors might in uence the response to IVIG, and hence serve as a predictive index for refractory KD.
IL-1β plays a crucial role in the disease mechanism of KD by inhibiting the IL-1β pathway and blocking IL-1β signaling transduction, thus inhibiting the onset and progress of KD. IL-1β is also closely associated with the resistance to IVIG [11,12]. Recent studies also suggested that IFN-γ and IL-6 were independent risk factors for pediatric refractory KD [13]. Meprins are zinciferous metallopeptidases with α and β subunits [14]. Meprin A is a homodimer of α subunit (α-α) or a heterodimer of α and β subunits (α-β), which can transform the precursor form of IL-1β (pro-IL-1β) into the biologically active IL-1β. It not only hydrolyzes and inactivates IL-6 but also produces sIL-6R through the lysis of IL-6R on the cell membrane to mediate IL-6 countersignaling [15,16]. In addition, an experiment on mice showed that the effect of overexpressed Meprin-α on the formation of atherosclerosis might be closely associated with the in ammatory response mediated by active type 1 helper T cells, which facilitated the secretion of IFN-g [17]. Consequently, it was presumed that Meprin A might be involved in the onset and progress of refractory KD by regulating the imbalance of in ammatory mediators (e.g., IL-6 pathway) and the further release of IL-1β and IFN-g. The present study found that the serum level of IFN-g of patients with KD in the rst IVIG-nonresponsive group induced a signi cant increase in the expression of IP-10. IP-10 was secreted from a variety of IFN-g-induced tissue cells and immune cells [18]. It exerted a recruiting and chemotactic effect on T cells in autoimmune diseases and aggregated the corresponding T cells to target organs, causing tissue and organ damage [19]. IP-10/IL-17 secreted from the plasma of pediatric patients with KD acted on the calci cation of human coronary artery smooth muscle cells (HCASMCs); IL-17 might also interact with IFN-g to enhance the production of IP-10 [8]. Thus, it was inferred that IP-10 and IL-17A cooperatively in uenced the CALs in KD. IL-17A is a member of the IL-17 family. The proin ammatory cytokines with one or more biological functions are produced from immune cells, such as Th17 cells, natural killer cells, and mast cells. They upregulate the expression of pro-in ammatory genes and induce the production of IL-6, IL-8, and INF-α in some cells (e.g., endothelial cells and macrophages) [20]. Japanese researchers also found that IL-1beta and IL-17A are involved in IVIG resistance through activation of C/EBPβandδ in a coronary artery model of Kawasaki disease [1].
Glucocorticoids (GCs), as an alternative to in ammation suppression, have gained increasing attention in the treatment of refractory KD. However, their mechanism of action remains unclear. GCs can upregulate or downregulate up to 20% of gene expression [21]. They bind to cytoplasmic receptors to inhibit the transcription of in ammatory proteins but promote the transcription of anti-in ammatory proteins, thus ful lling the anti-in ammatory role at the genome level. These proteins include cytokines such as TNF-α, IL-6, IFN-g, chemokines, and cell adhesion molecules [22]. They can also reduce the secretion of in ammatory cytokines by in uencing the activity of NF-kB [5]. In this study, the extent of decrease in the serum levels of Meprin A, IL-1β, and IL-6 after the second treatment was higher in the IVIG 2 g /kg group compared with the IVIG 1 g/kg group. In the IVIG 1 g/kg + GC group, the serum levels of Meprin A, IL-1β, and IL-6 decreased signi cantly, and the extent of decrease in the IL-17A and IP-10 levels was also extremely obvious (P < 0.05). The extent of decrease in serum levels of Meprin A after the second treatment was even higher in the IVIG 1 g/kg +GC group compared with the IVIG 2 g/kg group (P<0.05). Moreover, 94.7% of the patients in the second treatment nonresponsive group were given IVIG alone; serum levels of Meprin A, IL-1β, and IL-17A were signi cantly higher than that in the second treatment responsive group. The responsive rate in the IVIG-1 g/kg + GC group was up to 96.7%, and the CAL incidence during the follow-up visit after the second treatment was 10%, which was 2.2 times lower than that in the IVIG-2 g/kg group and 5.5 times lower than that in the IVIG-1 g/kg group. This suggested that GCs showed a better anti-in ammatory effect and a long-term prognosis. This might be associated with a more rapid function of GCs. Previous studies showed that IVIG combined with a GC protocol rapidly decreased the levels of in ammatory cytokines such as IL-2, IL-6, and IL-8 in the peripheral circulation within 24 h [23], when used to treat KD cases not responsive to IVIG. Reduced duration of systemic in ammation would signi cantly reduce the time of fever abatement and decrease the exposure rate of CAL.

Conclusion
A systematic analysis found that IL-1β, Meprin A, IL-17A, and other cytokines might be the factors in uencing refractory KD and CAL, which are potentially good predictors. IVIG combined with GCs responsively controlled the levels of IL-1β, Meprin A, IL-17, and other in ammatory factors in patients with refractory KD, reduced the incidence rate of CAL, and improved the responsive rate. This suggested that IVIG combined with GCs would have more advantages and prospects in KD treatment compared with IVIG alone. IVIG combined with a GC protocol should be given in the event of a signi cant increase in the levels of certain in ammatory factors such as IL-1β, Meprin A, and IL-17. However, this study still had limitations. The GC treatment group should be added in subsequent studies on refractory KD for the comparison of response and safety.

Consent for publication
All patients and their parents provided written informed consent for their data to be used in analyses and reported.

No Competing interests
The authors declare that they have no competing interests.

Funding
Medical youth talent program of Zhejiang Province (2019RC267) Ningbo leading and top talent cultivation project (NBLJ201801035); Ningbo science and technology bureau natural fund project(2016A610172).
Authors' contributions LW and FFH was responsible of the collection of clinical information on KD patients, statistical analyses, gures, data interpretation and manuscript preparation. LW and ZJC were responsible for critical review of the statistical analyses and the manuscript. LW and SLZ contributed in the writing of the manuscript and its scienti c content. YZD is the project leader of the study; she is involved in the conceptualization of the project, the study design and preparation of the manuscript. All authors read and approved the nal manuscript.