Association of lower eosinophil-related Thelper 2 (Th2) cytokines with coronary arterylesions in Kawasaki disease

Kawasaki disease (KD) is an acute febrile multi-systemic vasculitis of unknown etiology, whichwas first reported by Kawasaki et al. (1) fromJapan in 1974 in the English language literature.It occurs worldwide and mainly affects childrenless than 5 yr old. The clinical characteristics ofKD include prolonged fever, bilateral non-puru-lent conjunctivitis, diffuse mucosal inflammation,polymorphous skin rashes, indurative angioe-dema of the hands and feet, and non-suppurativecervical lymphadenopathy (1–4). The most seri-ous complication of KD is coronary arterylesions (CAL) including myocardial infarction,coronary artery dilatation, or coronary arteryaneurysm (2, 3). In the developed countries, KDis the leading cause of acquired heart diseases inchildren (2–5).

A complication of coronary artery aneurysmdevelops in 20% of untreated KD affectedchildren (3). A single high-dose (2 g/kg) ofintravenous immunoglobulin (IVIG) plus aspirincan diminish the incidence of aneurysm from20% to 3–5% (6). The IVIG treatment can alsoshorten the fever duration when given within10 days after the disease onset (6, 7). As CALoccurs at a mean of 9.5 days after the onset ofKD, it is important to treat and prevent pro-gression of coronary artery injury within 10 days(2, 3). Kawasaki et al. (1) were the first todescribe eosinophilia in the peripheral blood ofKD patients. Later, eosinophils were also foundin coronary artery autopsies (8). We haverecently shown that eosinophilia was positivelycorrelated to IVIG treatment success (9), and

Kuo H-C, Wang C-L, Liang C-D, Yu H-R, Huang C-F, Wang L,Hwang K-P, Yang KD. Association of lower eosinophil-related Thelper 2 (Th2) cytokines with coronary artery lesions in Kawasakidisease.Pediatr Allergy Immunol 2009: 20: 266–272.� 2008 The AuthorsJournal compilation � 2008 Blackwell Munksgaard

Kawasaki disease (KD) is a systemic febrile vasculitis particular coro-nary artery involvement. Eosinophilia has been found in our and otherstudies in KD. This study further investigates whether eosinophil-related T helper 2 (Th2) cytokines or the activation marker (eosinophilcationic protein – ECP) is involved in KD with coronary artery lesions(CAL). A total of 95 KD patients were enrolled for this study. Plasmasamples were subjected to the measurement of interleukin (IL)-4, IL-5,and eotaxin by Luminex-Bedalyte multiplex beadmates system and tothe measurement of ECP by fluoroimmunoassay. Patients with KD hadhigher eosinophils than controls. Eosinophil-related mediators: IL-4,IL-5, eotaxin, and ECP levels were also higher in KD patients thancontrols before intravenous immunoglobulin (IVIG) treatment. AfterIVIG treatment, ECP decreased but IL-4, IL-5, and eotaxin increasedsignificantly. The higher the IL-5 and eosinophil levels after IVIGtreatment, the lower rate of CAL was found. Changes of eosinophilsafter IVIG treatment were positively correlated to changes of IL-5 levelsbut not ECP levels. An increase of eosinophils and IL-5, but not ECPlevels after IVIG treatment, was inversely correlated with CAL for-mation in KD.

Ho-Chang Kuo1,2, Chih-Lu Wang3, Chi-Di Liang4, Hong-Ren Yu1,2, Chien-FuHuang4, Lin Wang1,2, Kao-Pin Hwang5

and Kuender D. Yang1,2

1Division of Allergy, Immunology and Rheumatology;Department of Pediatrics, Chang Gung MemorialHospital-Kaohsiung Medical Center; 2GraduateInstitute of Clinical Medical Sciences, Chang GungUniversity College of Medicine; 3Department ofPediatrics, Po-Jen Hospital; Divisions of 4Cardiologyand 5Infectious Disease, Department of Pediatrics,Chang Gung Memorial Hospital-Kaohsiung MedicalCenter, Kaohsiung, Taiwan

Key words: Kawasaki disease; coronary arterylesions; eosinophils; T helper 2; interleukin-5;eosinophil cationic protein

Kuender D. Yang, Department of Medical Researchand Pediatric Allergy, Immunology and Rheumatology,Chang Gung Memorial Hospital-Kaohsiung MedicalCenter, 123 Ta-Pei Road, Niaosung Hsiang,Kaohsiung, TaiwanTel.: +886-7-7317123Fax: +886-7-7312867E-mail: yangkd@adm.cgmh.org.tw

Accepted 28 May 2008

Pediatr Allergy Immunol 2009: 20: 266–272

DOI: 10.1111/j.1399-3038.2008.00779.x

� 2008 The AuthorsJournal compilation � 2008 Blackwell Munksgaard

PEDIATRIC ALLERGY AND

IMMUNOLOGY

266

patients� characteristics as well as IVIG prepara-tions might affect eosinophilia (10).To explore the mechanism of eosinophilia in

KD, this study was conducted to investigatewhether eosinophil-related T helper 2 (Th2)cytokines (IL-4, IL-5, and eotaxin) or the eosin-ophil activation mediator (eosinophil cationicprotein – ECP) were associated with CALformation in KD patients.

Patients and methodsPatients studied

All subjects studied were children who fulfilledthe criteria for KD (4) and were admitted forIVIG treatment at Chang Gung Memorial Hos-pital-Kaohsiung Medical Center from 2001 to2006. All patients were initially treated with asingle dose of IVIG (2 g/kg) during a 12-hperiod. Aspirin (3–5 mg/kg/day) was given untilall signs of inflammation were resolved orregression of CAL under two-dimensional (2D)echocardiography was seen. This study wasapproved by the Institutional Review Board ofChang Gung Memorial Hospital. Blood sampleswere collected after informed consent wasobtained from the parents or guardians. Bloodsamples collected both before (within 24 h beforeIVIG treatment, pre-IVIG) and after IVIGtreatment (within 3 days after IVIG treatment,post-IVIG) were subjected to this study. Patientswhose symptoms did not fit the KD criteria, hadan acute fever for less than 5 days, or incompletecollection of pre- and post-IVIG blood sampleswere excluded. CAL was defined as the internaldiameter being at least 3 mm of the coronaryartery (4 mm if the subject was over the age of5 yr) or the internal diameter of a segment atleast 1.5 times as large as that of an adjacentsegment by echocardiogram (11, 12). We ana-lyzed the complete blood counts (CBC)/differen-tial counts (DC) and C-reactive protein (CRP)from 95 KD patients and 30 age-matched febrilecontrols (male/female = 16/14). Blood samplesfrom the febrile control patients who wereadmitted for upper and/or lower respiratorytract infections (including acute bronchiolitis,acute pharyngitis, acute bronchitis, croup, andacute tonsillitis) without a past history of allergicdisease were also included for comparison.

Measurement of eosinophil-related Th2 chemokines andcytokines by the Luminex-100 system

Plasma concentrations of IL-4, IL-5, and eotaxinwere assessed by the Upstate Beadlyte Human

Cytokine Beadmates system (Upstate Group,Inc., Billerica, MA, USA) for the 95 KD patientsand 30 control samples. The study method wasmodified from a previous report (13). In brief,50 ll samples were mixed with multiplexedantibody-conjugated beads before being sub-jected to multi-channel detection of the bead-array. Acquired fluorescence data were assessedby the MasterPlexTM QT software (Ver. 1.2;MiraiBio, Inc., South San Francisco, CA, USA).Calibration of cytokine concentrations wasdetermined by interpolation of a series of well-known standard samples following the manufac-turer�s recommendation. In order to make surethat the effect seen is not an IVIG effect itself, wealso measured the levels of IL-4, IL-5, andeotaxin in IVIG production (Gamimune N10%, Bayer Corporation, Elkhart, IN, USA)from eight different bottles of four differentbatches based on the normalization to averageblood immunoglobulin (Ig) G concentration(2000 mg/dl) (14). The assay sensitivities of thesecytokines were 1.8 pg/ml of IL-4, 0.2 pg/ml ofIL-5, and 1.4 pg/ml of eotaxin, respectively. Toavoid inter-assay bias of immunoassays, thecytokines in paired samples before and afterIVIG therapy were measured at the same time.

Measurement of eosinophil activation mediator – ECP

Concentrations of ECP were measured byPharmacia CAP system fluoroimmunoassay(Pharmacia and Upjohn Diagnostics AB,Uppsala, Sweden) according to the instructionsof the manufacturer. In brief, 40 ll plasmasamples from the 95 KD patients and 30controls as well as the eight different IVIGproducts were subjected to the automatic pro-cedure of analysis.

Statistical analysis

Data of CBC/DC andCRP levels between theKDpatients and controls were assessed by theStudent�s t-test. Changes of IL-4, IL-5, eotaxin,and ECP levels before and after IVIG treatmentwere tested by the paired sample t-test. Levels ofIL-4, IL-5, eotaxin, andECPbetweenKDpatientswith and without CAL were tested by the Mann–Whitney U-test. Correlations between groupswere tested by the Pearson�s correlation.A p-value <0.05 was considered as statisticallysignificant. Data are presented as mean andstandard error. All statistical tests were performedusing SPSS 12.0 for Windows XP (SPSS, Inc.,Chicago, IL, USA).

Role of eosinophil-related Th2 mediators in KD

267

ResultsClinical features of the KD patients with and without CAL

A total of 165 KD patients were admitted from2001 to 2006. Ninety-five of the KD patientswhose blood samples were collected both beforeand after the IVIG treatment were enrolled inthis study. There were 31 girls (32.6%) and 64boys (67.4%). There were 20 patients (21.05%)with CAL formation. There were no significantdifference of the rates of CAL formation betweenthe study group (20/95) and the remainder KDpatients (13/70, p = 0.7). The age distribution ofKD patients with and without CAL was 18.36±2.32 (median: 15.0 months) and 22.47 ± 1.55(median: 17.0 months), respectively. The majorclinical features including conjunctivitis, fissuredlips, polymorphous skin rashes, indurative an-gioedema of hands and feet, cervical lymphade-nopathy, and Bacillus Calmette-Guerin (BCG)vaccination scar reaction between KD patientswith and without CAL showed no significantdifference. There were also no difference betweenage distribution and total admission daysbetween the KD patients with and without CAL.

Peripheral leukocytes and eosinophil in the KD patients andcontrols

The KD patients had higher leukocyte countsand platelet counts than controls. There was asignificantly higher eosinophil percentage(2.5 ± 0.2% vs. 0.4 ± 0.1%; p < 0.01) in KDpatients than controls, but no statistical differ-ence in neutrophil, lymphocyte, monocyte, andbasophil percentages between the KD patientsand the controls (Table 1). The KD patients hadlower hemoglobin levels than the controls. AfterIVIG treatment, eosinophils were greatlyincreased (2.5 ± 0.2% vs. 4.7 ± 0.3%; p < 0.0-01, Fig. 1), while the inflammatory marker CRPlevels were significantly decreased (105.7 ± 5.8mg/L vs. 55.4 ± 4.1 mg/L; p < 0.001). Hence,experiments were next performed to assess eosin-ophil-related Th2 cytokines and the eosinophilactivation mediator between the KD patients andthe controls as described below.

Eosinophil-related Th2 cytokines/eosinophils activationmediator between the KD patients and controls

As shown in Table 2, we found that IL-4 (p <0.001), IL-5 (p < 0.001), eotaxin (p = 0.004),ECP (p < 0.001), and CRP (p = 0.003) weresignificantly higher in the KD patients beforeIVIG treatment than the controls. Furtheranalysis found that the plasma levels of IL-4

(12.07 ± 1.36 pg/ml vs. 28.55 ± 3.84 pg/ml,p < 0.001, Fig. 2a), IL-5 (5.17 ± 0.56 pg/mlvs. 10.33 ± 0.92 pg/ml, p < 0.001, Fig. 2b),and eotaxin (129.1 ± 14.3 pg/ml vs. 296.5 ±31.5 pg/ml, p < 0.001, Fig. 2c) were signifi-cantly increased after IVIG treatment. In con-trast, the eosinophil activation mediator (ECP)levels were greatly decreased after IVIG treat-ment (11.57 ± 1.98 pg/ml vs. 7.49 ± 1.22 pg/ml, p = 0.03, Fig. 2d).

Eosinophil and IL-5 but not CRP levels associated with KD withCAL after IVIG treatment

After IVIG treatment, CRP levels decreased andrevealed no significant difference between the KDpatients with and without CAL (63.7 ± 8.4 mg/lvs. 51.4 ± 4.3 mg/l, p = 0.15). In contrast, wefound that eosinophils were significantly higherin the KD patients without CAL (5.09 ± 0.43%vs. 3.33 ± 0.54%; p = 0.03, Fig. 3a) than thosewith CAL after IVIG treatment. The eosinophil-related Th2 cytokine (IL-5) was also significantlyhigher in the KD patients without CAL thanthose with CAL (11.4 ± 1.2 pg/ml vs. 5.5 ±1.6 pg/ml; p = 0.02, Fig. 3b). The plasma levelsof IL-4, eotaxin, and ECP between KD patientswith and without CAL showed no significantdifference. Changes of eosinophils after IVIGtreatment had a positive correlation with changesof IL-5 levels (p = 0.007, R2 = 0.19, Fig. 3c),but had no significant correlation with changesof ECP levels (p = 0.29, Fig. 3d).

Levels of cytokines: IL-4, IL-5, eotaxin, and ECP in the IVIGproducts

In consideration of cytokines in the IVIG prod-ucts, which may contribute to the elevation of

Table 1. Analysis of complete blood counts/differential counts in KD and age-matched controls

KD (n = 95) Control (n = 30) p Values

Total leukocyte/mm3 13792 € 733 9412 € 551 <0.01Hemoglobin (g/dl) 10.8 € 0.1 12.2 € 0.1 <0.01Platelet (·104/mm3) 37.6 € 1.4 26.4 € 1.1 <0.01Neutrophil (%) 65.9 € 1.4 62.4 € 2.7 0.23Lymphocyte (%) 24.6 € 1.2 27.5 € 2.5 0.31Monocyte (%) 5.7 € 0.3 6.8 € 0.7 0.20Eosinophil (%) 2.5 € 0.2 0.4 € 0.1 <0.01Basophil (%) 0.17 € 0.03 0.21 € 0.05 0.61

Data in KD were measured before IVIG treatment (5.97 € 1.96 days afterdisease onset). Data in the control group were also collected in the acutestage of upper or lower respiratory tract infection (including acute bronchio-litis, acute pharyngitis, acute bronchitis, croup, and acute tonsillitis withoutpast history of allergic disease). Values presented as mean € SE. p Valueswere assessed by the Student�s t-test.

Kuo et al.

268

IL-4 and IL-5 levels after IVIG treatment, wemeasured IL-4, IL-5, eotaxin, and ECP levels ineight different bottles of four different batches ofIVIG products. Using a basis of the blood IgGconcentration at 2000 mg/dl, the ECP levels fromeight different bottles of IVIG had a concentra-tion less than 2.0 pg/ml. The levels of IL-5, IL-4,and eotaxin from eight different bottles of IVIGshown a median fluorescence intensity less thanbasal fluorescence intensity detected by Luminex.This suggests that exogenous cytokines in theIVIG preparation may not affect the levels ofblood cytokines that are measured 48–72 h afterIVIG treatment.

Discussion

A number of factors including prolonged fever,young age, male gender, initially high CRP,higher neutrophil, and band form counts in KDpatients have been implicated in prediction ofCAL (15–17). These predictors, however,revealed inconsistent correlation to CAL in

different studies (15–21), while the underlyingmechanism of CAL remains unclear. Kawasakiet al. (22) first observed that 11 of 50 KDpatients (22%) had eosinophilia in the peri-pheral blood. Terai et al. (8) found accumula-tion of eosinophils in the coronary micro-vessellesions and eosinophilia in peripheral blood(PB) and postulated the involvement of eosin-ophils in the pathogenesis of KD vasculitis. Wehave recently shown that eosinophils weresignificantly elevated in KD both before andafter IVIG treatment, and eosinophilia afterIVIG treatment had an inverse correlation tothe KD patients with IVIG-resistance (9). Inthis study, we have further shown that eosino-phil-related Th2 mediators (IL-4, IL-5, andeotaxin) increased, but ECP and CRP levelsdecreased after IVIG treatment in KD.Further analysis found that changes of eosin-

ophil percentage after IVIG treatment werehigher in KD patients without CAL than withCAL. The absolute cell counts of eosinophil werenot significantly different between the KDpatients with and without CAL (367.1 ± 75.1vs. 444.3 ± 48.5/mm3, p = 0.23, respectively).The higher eosinphil percentage in the KDpatients without CAL may result from increasedrecruitment of eosinophils from the bone marrowor from the decrease of other leukocyte subpop-ulations. The changes of eosinophil percentagewere correlated to changes of IL-5 levels but notcorrelated to changes of ECP levels suggesting aTh2 reaction associated with an increase ofeosinophil chemotactic factors but not eosinophilactivation factors associated with the decrease ofCAL in KD patients. The changes of absoluteeosinophil counts did not positively correlate withchanges of IL-5 levels (p = 0.12, R2 = 0.051).This may be the reason that higher Th2 reactionin KD could induce disproposrtion of eosinophilsbut not eosinophil activation or absolute eosino-phil counts. This is compatible to our recentreport that although a decline of total leukocytecount after IVIG treatment was found in bothKD patients with and without CAL, total leuko-cyte count after IVIG treatment remained signif-icant higher in KD patient with CAL (23). Takentogether, a higher total leukocyte with lowereosinophil percentage after IVIG treatment isassociated with CAL formation in KD patients.We have also measured the levels of IL-4, IL-5,

eotaxin, and ECP in IVIG products from eightdifferent bottles. All cytokines were lower thanthose in plasma levels based on the normalizationto average blood IgG concentration (2000 mg/dl). In consideration of the short half-life ofcytokines of about 20 min in blood, the

Table 2. Comparison of eosinophil-related Th2 cytokines/eosinophil mediatorand CRP between the KD patients and age-matched controls

KD (n = 95) Control (n = 30) p value

IL-4 (pg/ml) 12.07 € 1.36 5.96 € 0.54 <0.001IL-5 (pg/ml) 5.17 € 0.56 2.65 € 0.55 <0.001Eotaxin (pg/ml) 129.1 € 14.3 74.5 € 7.4 0.004ECP (pg/ml) 11.57 € 1.98 2.98 € 0.23 <0.001CRP (mg/l) 105.7 € 5.8 39.5 € 14.8 0.003

IL, interleukin; ECP, eosinophil cationic protein; CRP, C-reactive protein. Valuespresented as mean € SE. p Values were assessed by the Student�s t-test.

Fig. 1. Eosinophils were greatly increased after IVIG treat-ment. Bars represent mean and standard error of mean(n = 95, paired sample t test). IVIG, intravenous immuno-globulin.

Role of eosinophil-related Th2 mediators in KD

269

exogenous cytokines in IVIG preparation maynot affect the blood cytokines levels that aremeasured 48–72 h after IVIG treatment. Therewere four out of the 95 study KD patients hadallergy history before the diagnosis of KD. TheTh2 cytokines measured in this study showed nosignificant difference between KD patients withand without allergic disease history (p > 0.1).Therefore, the higher Th2 cytokines found in KDpatients may be the natural course of KD.There are several lines of evidence pointing out

an abnormal Th1/Th2 balance in KD patients.Brosius et al. (24) reported that the incidence ofatopic dermatitis among children with KD wasnine times greater than that of controls. SerumIgE and IL-4 levels were also significantly higherin KD patients than age-matched children (24–26). Matsubara et al. (27) found a decrease ofIFN-c expression in KD patients. These resultssuggest a skewed imbalance toward Th2 reactionin KD. There are certain subpopulations of Th2s

that release different cytokine profiles in responseto different stimuli. They reciprocally constitutean immunoregulatory loop between Th1 and Th2cytokines from Th1 cells inhibiting Th2 cells andvice versa (28). Th2 cells produce IL-4, IL-5, andother cytokines, which promote humeral immu-nity, allergic inflammation, and stimulate B cellsto produce IgE as well as other Igs (29). Incontrast, Th1 cells secrete IFN-c and IL-2, whichinitiate the killing of intracellular organisms andviruses through activating cytotoxic T cells andmacrophages. In this study, we found that Th2cytokines (IL-4, IL-5, and eotaxin) significantlyincreased in KD patients before and after IVIGtreatment. It is controversial whether Th1 cyto-kines such as IFN-c increase or decrease in KDpatients (30–32). It remains unclear why Th2mediators are increased but Th1 mediators donot decrease in KD. We are the first to demon-strate that IL-5 and eosinophil levels wereassociated with the CAL in KD. Further studies

(a) (b)

(c) (d)

Fig. 2. Levels of eosinophil chemotactic factors and activation factor before and after IVIG treatment. Eosinophil-relatedTh2 mediators (IL-4, IL-5, and eotaxin) significantly increased after IVIG treatment (n = 95, p < 0.001) (a–c). Eosinophilactivation mediator (ECP) significantly decreased after IVIG treatment (n = 95, p = 0.03) (d). Bars represent mean andstandard error of mean. Values were tested by paired samples t-test. IVIG, intravenous immunoglobulin.

Kuo et al.

270

to clarify whether genetic variants of Th2 genesare responsible for the susceptibility and mor-bidity of KD are needed.Certain studies have shown that atheroscle-

rosis is related to a skewed Th1-like response(33), and autoimmune disease such as Behcet�sdisease is related to Th1 cytokine and pro-inflammatory cytokine (34, 35). We postulatethat the prominent Th2 reaction may be devel-oped to combat the Th1-mediated vasculitis inKD after IVIG treatment. Th2-related cytokinessuch as IL-4 and IL-5 may not only suppressTh1 reaction, but also promote B cell develop-ment and eosinophil mobilization (28). We havepreviously shown that eosinophilia was associ-ated with the less IVIG treatment failure (9); in

this study, we found an increase in the Th2mediators (IL-4 and IL-5), but a decrease ofECP, was associated with eosinophilia and lessCAL formation in KD, suggesting eosinophiliain KD is a bystander of Th2 response, but notan effector of KD.

AcknowledgmentThis study was in part supported by a grant CCF07-01 fromthe Foundation of Taiwanese Childhood Heart Diseases.

References

1. Kawasaki T, Kosaki F, Osawa S, Shigemitsu I,Yanagawa S. A new infantile acute febrile mucocuta-neous lymph node syndrome (MLNS) prevailing inJapan. Pediatrics 1974: 54: 271–6.

(a) (b)

(c)(d)

Fig. 3. Correlation of eosinophilia with IL-5 but not eosinophil cationic protein (ECP) levels in the KD patients before andafter IVIG treatment. Eosinophils (p = 0.03) and IL-5 (p = 0.02) levels after IVIG treatment were significantly higher in theKD patients without CAL than with CAL (a, b) (Mann–Whitney U-test). Changes of eosinophils after IVIG treatment had asignificant correlation with changes of IL-5 levels (p = 0.007, Pearson�s correlation test, R2 = 0.19) but not significantlycorrelated with changes of ECP levels (p = 0.29) (c, d). Bars represent mean and standard error of mean. IVIG, intravenousimmunoglobulin; CAL, coronary artery lesions; ECP, eosinophil cationic protein.

Role of eosinophil-related Th2 mediators in KD

271

2. Wang CL, Wu YT, Liu CA, Kuo HC, Yang KD.Kawasaki disease: infection, immunity and genetics.Pediatr Infect Dis J 2005: 24: 998–1004.

3. Burns JC, Glode MP. Kawasaki syndrome. Lancet2004: 364: 533–44.

4. Newburger JW, Takahashi M, Gerber MA, et al.Diagnosis, treatment, and long-term management ofKawasaki disease: a statement for health professionalsfrom the Committee on Rheumatic Fever, Endocarditisand Kawasaki Disease, Council on CardiovascularDisease in the Young, American Heart Association.Circulation 2004: 110: 2747–71.

5. Hsu YH, Wang YH, Hsu WY, Lee YP. Kawasaki dis-ease characterized by erythema and induration at theBacillus Calmette–Guerin and purified protein derivativeinoculation sites. Pediatr Infect Dis J 1987: 6: 576–8.

6. Newburger JW, Takahashi M, Beiser AS, et al.A single infusion of intravenous gamma globulin ascompared with four infusions in the treatment of acuteKawasaki syndrome. N Engl J Med 1991: 324: 1633–9.

7. Newburger JW, Takahashi M, Burns JC, et al. Thetreatment of Kawasaki syndrome with intravenousgamma globulin. N Engl J Med 1986: 315: 341–7.

8. Terai M, Yasukawa K, Honda T, et al. Peripheralblood eosinophilia and eosinophil accumulation incoronary microvessels in acute Kawasaki disease.Pediatr Infect Dis J 2002: 21: 777–81.

9. Kuo HC, Yang KD, Liang CD, et al. The relationshipof eosinophilia to intravenous immunoglobulin treat-ment failure in Kawasaki disease. Pediatr AllergyImmunol 2007: 18: 354–9.

10. Kuo HC, Wang CL, Wang L, Yu HR, Yang KD.Patient characteristics and intravenous immunoglobulinproduct may affect eosinophils in Kawasaki disease.Pediatr Allergy Immunol 2008: 19: 184–5.

11. Akagi T, Rose V, Benson LN, Newman A, Freedom

RM. Outcome of coronary artery aneurysms afterKawasaki disease. J Pediatr 1992: 121: 689–94.

12. Shulman ST, Inocencio J, Hirsch R. Kawasaki dis-ease. Pediatr Clin North Am 1995: 42: 1205–22.

13. Pickering JW, Martins TB, Schroder MC, Hill HR.Comparison of a multiplex flow cytometric assay withenzyme-linked immunosorbent assay for auantitation ofantibodies to tetanus, diphtheria, and Haemophilusinfluenzae Type b. Clin Diagn Lab Immunol 2002: 9:872–6.

14. Baba R, Shibata A, Tsurusawa M. Single high-doseintravenous immunoglobulin therapy for Kawasaki dis-ease increases plasma viscosity. Circ J 2005: 69: 962–4.

15. Mori M, Imagawa T, Yasui K, Kanaya A, Yokota S.Predictors of coronary artery lesions after intravenousc-globulin treatment in Kawasaki disease. J Pediatr2000: 137: 177–80.

16. Burns JC, Capparelli EV, Brown JA, Newburger

JW, Glode MP. Intravenous gamma-globulin treat-ment and re-treatment in Kawasaki disease. PediatrInfect Dis J 1998: 17: 1144–8.

17. Beiser AS, Takahashi M, Baker AL, Sundel RP,Newburger JW. A predictive instrument for coronaryartery aneurysms in Kawasaki disease. Am J Cardiol1998: 81: 1116–20.

18. Kim T, Choi W, Woo CW, et al. Predictive riskfactors for coronary artery abnormalities in Kawasakidisease. Eur J Pediatr 2007: 11: 166: 421–5.

19. Takahashi K, Oharaseki T, Naoe S, Wakayama M,Yokouchi Y. Neutrophilic involvement in the damage

to coronary arteries in acute stage of Kawasaki disease.Pediatr Int 2005: 47: 305–10.

20. Hamamichi Y, Ichida F, Yu X, et al. Neutrophils andmononuclear cells express vascular endothelial growth fac-tor in acute Kawasaki disease: its possible role in progres-sion of coronary artery lesions. Pediatr Res 2001: 49: 74–80.

21. Honkanen VE, McCrindle BW, Laxer RM, Feldman

BM, Schneider R, Silverman ED. Clinical relevanceof the risk factors for coronary artery inflammation inKawasaki disease. Pediatr Cardiol 2003: 24: 122–6.

22. Kawasaki T. Acute febrile mucocutaneous syndromewith lymphoid involvement with specific desquamationof the fingers and toes in children: clinical observationsof 50 cases (In Japanese). Arerugi 1967: 16: 178–222.

23. Kuo HC, Wang CL, Ligan CD, et al. Persistentmonocytosis after intravenous immunoglobulin therapycorrelated with the development of coronary arterylesions in patients with Kawasaki disease. J MicrobiolImmunol Infect 2007: 40: 395–400.

24. Brosius CL, Newburger JW, Burns JC, Hojnowski-Diaz P, Zierler S, Leung DY. Increased prevalence ofatopic dermatitis in Kawasaki disease. Pediatr InfectDis J 1988: 7: 863–6.

25. Furukawa S, Matsubara T, Motohashi T, et al.Increased expression of Fc epsilon R2/CD23 onperipheral blood B lymphocytes and serum IgE levels inKawasaki disease. Int Arch Allergy Appl Immunol1991: 95: 7–12.

26. Hirao J, Hibi S, Andoh T, Ichimura T. High levels ofcirculating interleukin-4 and interleukin-10 in Kawasakidisease. Int Arch Allergy Immunol 1997: 112: 152–6.

27. Matsubara T, Katayama K, Matsuoka T, Fujiwara

M,KogaM,Furukawa S.Decreased interferon-gamma(IFN-gamma)-producing T cells in patients with acuteKawasaki disease. Clin Exp Immunol 1999: 116: 554–7.

28. Elias JA, Lee CG, Zheng T, Ma B, Homer RJ, Zhu Z.New insights into the pathogenesis of asthma. J ClinInvest 2003: 111: 291–7.

29. MosmannTR,CoffmanRL.Th1andTh2 cells, differentpatterns of lymphokine secretion lead to different func-tional properties. Annu Rev Immunol 1989: 7: 145–73.

30. Takeshita S, Nakatani K, Tsujimoto H, Kawamura

Y, Kawase H, Sekine I. Increased levels of circulatingsoluble CD14 in Kawasaki disease. Clin Exp Immunol2000: 119: 376–81.

31. Okada Y, Shinohara M, Kobayashi T, et al. Effect ofcorticosteroids in addition to intravenous gammaglobulin therapy on serum cytokine levels in the acutephase of Kawasaki disease in children. J Pediatr 2003:143: 363–7.

32. Abe J, Jibiki T, Noma S, Nakajima T, Saito H, Terai

M. Gene expression profiling of the effect of high-doseintravenous Ig in patients with Kawasaki disease.J Immunol 2005: 174: 5837–45.

33. Mazzolai L, Duchosal MA, Korber M, et al.Endogenous angiotensin II induces atherosclerotic pla-que vulnerability and elicits a Th1 response in ApoE-/-mice. Hypertension 2004: 44: 277–82.

34. Oztas MO, Onder M, Gurer MA, Bukan N, Sancak

B. Serum interleukin 18 and tumour necrosis factor-alpha levels are increased in Behcet�s disease. Clin ExpDermatol 2005: 30: 61–3.

35. Suzuki N, Nara K, Suzuki T. Skewed Th1 responsescaused by excessive expression of Txk, a member of theTec family of tyrosine kinases, in patients with Behcet�sdisease. Clin Med Res 2006: 4: 147–51.

Kuo et al.

272