Prognostic Significance of QT Interval Dispersion in the Response to Intravenous Immunoglobulin Therapy in Kawasaki Disease

Abstract

Background: Kawasaki disease (KD) is classified as a systemic vasculitis syndrome and QT interval dispersion (QTD) has been associated with cardiac involvement and disease activity in patients with cardiovasculitis. We examined whether baseline QTD could predict a response to intravenous immunoglobulin (IVIG) in KD.

Methods and Results: QTD was recorded in 86 patients with KD before IVIG, who were separated into IVIG responders (R group; n=62) and nonresponders (N group; n=24). The association between baseline QTD and response to IVIG was investigated, and the predictive response value was compared with conventional risk scores from Gunma and Kurume universities. Baseline-corrected QTDs with Bazett’s (QTbcD) and Fridericia’s (QTfcD) formulae were significantly increased in the N group (R group vs. N group: 31.6 [28.3, 44.0] ms vs. 66.6 [50.5, 76.3] ms and 27.4 [25.2, 39.1] ms vs. 55.2 [42.4, 66.3] ms, respectively, both P<0.001). Multiple logistic regression analysis revealed QTfcD as an independent predictor of a response to IVIG after adjustment for conventional scores (odds ratio: 1.133, 95% confidence interval: 1.061–1.210, P<0.001). Moreover, QTfcD provided incremental predictive value for IVIG nonresponders over Gunma score (increment in global χ²=25.46, P<0.001).

Conclusions: QTD was significantly associated with a response to IVIG in KD patients and may represent a useful identifier of IVIG nonresponders with high risk of coronary aneurysm.

QT interval dispersion (QTD) is the difference between the maximum and minimum QT intervals on the 12-lead ECG and reflects the degree of myocardial repolarization inhomogeneity. In adults, QTD has been described as a predictive factor of heart failure, fatal arrhythmia, and cardiac sudden death.^1–3 Several reports have also demonstrated QTD as a prognosticator in children with congenital heart disease or cardiomyopathy.^4–8 However, the predictive value of QTD in cardiovascular diseases remains controversial.^9,10

Kawasaki disease (KD) is an acute childhood febrile illness that is classified as a systemic vasculitis syndrome. Together with aspirin, high-dose intravenous immunoglobulin (IVIG) therapy is generally effective in resolving the inflammation in KD and reducing the occurrence of coronary artery abnormalities (CAA). However, 10–20% of KD patients experience persistent or recurrent fever (i.e., severe KD) after IVIG therapy.^11,12 Being able to predict the response to IVIG therapy would enable selection of alternative therapy for patients with KD who have a high incidence of CAA. Predictive scores of the response to IVIG therapy, on the basis of demographic and laboratory variables before initial treatment, have been reported previously.^13,14 These scores have been used to make decisions about alternative or additional treatment. But in fact some nonresponders to IVIG treatment have not been predicted by these conventional risk scores.

In patients with cardiovasculitis, such as systemic lupus erythematosus (SLE) and Behçet’s disease, QTD has been associated with cardiovascular involvement and disease activity grade.^15,16 Prolonged QTD is a potentially useful method of early detection of subclinical cardiac involvement, such as pericarditis, myocarditis, and coronary arteritis, similar to that in KD. We hypothesized that QTD could enable assessment of disease severity of KD as well as of SLE and Behçet’s disease. Therefore, we aimed to elucidate whether baseline QTD could predict a response to IVIG therapy in patients with KD.

Methods

Study Population and Design

We retrospectively reviewed the clinical records of 86 patients who were diagnosed as having KD between 2010 and 2015 at 13 medical institutions in Nagano Prefecture in Japan (55 boys; median age, 2.2 years, range, 0.3–10.1 years). Criteria for KD included fever >38℃ accompanied by at least 4 of the following 5 findings: bilateral conjunctival injection, changes in the lips and oral cavity, non-purulent cervical lymphadenopathy, polymorphous exanthema, and changes in the extremities. These diagnostic criteria were compliant with the Diagnostic Guidelines for Kawasaki Disease.¹⁷

All patients initially received IVIG therapy of 2 g/kg over 24 h, together with aspirin (30–50 mg/kg/day) or flurbiprofen (2 mg/kg/day). The aspirin dose was decreased to 5 mg/kg/day after normalization of C-reactive protein (CRP). Patients were considered as afebrile when their body temperature remained at <37.5℃ for more than 24 h. Additional IVIG therapy was administered when patients had a persistent fever that lasted >48 h from the start of the first IVIG treatment or a recrudescent fever associated with KD symptoms after an afebrile period. Patients exhibiting persistent or recurrent fever despite a second course of IVIG therapy more than 48 h after the second IVIG start were defined as IVIG nonresponders who received additional rescue therapy that included infliximab, steroids, and/or plasma exchange. Coronary arteries were assessed by 2D echocardiography performed before the initial treatment and again between 2 and 7 days, 2 weeks, 1 month, 3 months, 6 months, and 1 year after the initial treatment. CAA were diagnosed when either of those examinations showed an internal lumen diameter ≥3 mm in children aged <5 years or ≥4 mm in children aged ≥5 years, an internal segment diameter ≥1.5-fold as large as that of an adjacent segment, or an irregular lumen appearance.¹⁸

The patients were divided into 2 study groups: R group comprised patients who were responders to the first or second IVIG therapy, and N group patients who were refractory to the first and second IVIG treatments and required additional therapy.

A standard 12-lead ECG taken at rest was obtained during the period between hospitalization and immediately prior to initial IVIG therapy (2–8 days of illness; median, 4.5 days). Two observers blinded to the diagnosis measured the QT intervals in every lead from the initial deflection of the QRS complex to the point at which a tangent crossed the isoelectric line by means of a digitizer.¹⁹ When a large U wave was present, the end of the T wave was measured from the intersection of the tangent to the repolarization slope with the isoelectric line.²⁰ The QT interval was not measured when the end of the T wave could not be identified. Because none of the subjects had a U wave that crossed the terminal portion of the T wave, QT intervals were recorded in 9–12 leads for each subject and thereafter corrected (QTc) with Bazett’s (QTbc) or Fridericia’s (QTfc) formula.^21,22 The QTD, QTbcD, and QTfcD were calculated by subtracting the shortest QT and QTc values from the longest ones across all 12 ECG leads at the same time (Figure 1). The mean values of QTD and QTcD that the 2 observers measured were evaluated.

Figure 1.

Measurement of QT dispersion in a responder (A) and a non-responder (B) to intravenous immunoglobulin therapy for Kawasaki disease. The QT intervals were measured in every lead. The QT dispersion (QTD) and corrected QTD (QTcD) were calculated by subtracting the shortest QT and QTc values from the longest ones across all 12 ECG leads at the same time.

The laboratory variables (% neutrophils, platelet count, aspartate aminotransferase (AST), alanine aminotransferase (ALT), serum sodium, and CRP) immediately prior to the first IVIG therapy were analyzed for associations with IVIG unresponsiveness. If a laboratory test was performed twice or more before the first IVIG, the highest value was chosen for analysis in the case of % neutrophils, AST, ALT, and CRP; or the lowest value in the case of platelet count and serum sodium. All data were compared with the conventional risk scores reported by Gunma¹³ and Kurume¹⁴ universities.

This study was approved by the Institutional Review Board of Shinshu University School of Medicine (authorization no. 2696).

Statistical Analysis

Statistical analyses were carried out using SPSS statistical software, version 18.0J. Normality of distribution was assessed through the use of the Shapiro-Wilk test. Data are expressed as the mean±SD or the median (interquartile range). To determine the significance of differences between 2 independent groups, the unpaired t-test was used when the data were normally distributed, and the Mann-Whitney U test was used for other variables. Univariate Spearman’s rank test was used to find correlations between QTfcD and other valuables. In the univariate analysis, logistic regression analysis was adopted to evaluate risk factors for IVIG nonresponders. Multiple logistic regression analysis was used to analyze the simultaneous effects of QTcD after adjustment for conventional risk scores from Gunma and Kurume universities. The incremental value of QTfcD over the Gunma score to assess the response to IVIG therapy was studied by calculating the improvement in global Chi-square. Differences identified throughout the study period were assessed with one-way repeated measures of ANOVA, followed by post-hoc (Bonferroni) testing. P<0.05 was considered statistically significant.

Results

Patients’ Characteristics and Clinical Outcomes

The inter and intraobserver intraclass correlation coefficients for QTD were 0.904 and 0.921, respectively. Table 1 shows the baseline characteristics and clinical outcomes of the patients. IVIG nonresponders were significantly older than the responders. The day of illness at initial IVIG was significantly earlier in nonresponders. Both the Gunma and Kurume scores were significantly increased in the N group at baseline, as were QTD, QTbcD, and QTfcD.

Table 1. Clinical Characteristics and Comparison Between Responders (R Group) and Nonresponders (N Group) to IVIG

	R group (N=62)	N group (N=24)	P value
Age, years	2.1 (1.2, 3.0)	3.3 (2.0, 4.9)	0.007
Male sex, n (%)	42 (68%)	13 (54%)	0.24
Days of illness at time of ECG	5 (4, 5)	4 (3, 5)	0.06
Days of illness at 1 st IVIG	5 (4, 6)	4 (3, 5)	0.003
Sodium (mmol/L)	136.0±2.1	133.0±2.2	<0.001
AST (IU/L)	38.0 (27.3, 83.0)	112.0 (38.5, 352.0)	0.008
ALT (IU/L)	26.0 (15.0, 122.8)	126.5 (36.0, 272.0)	0.012
Plt (104/μL)	29.4 (26.0, 42.0)	26.0 (21.3, 33.2)	0.05
Neutrophils (%)	66.3 (57.0, 76.7)	84.1 (80.7, 87.5)	<0.001
CRP (mg/dL)	6.0 (4.2, 8.0)	9.1 (7.0, 12.0)	0.001
Gunma score	2.0 (1.0, 3.0)	6.0 (4.8, 8.0)	<0.001
Kurume score	1.0 (1.0, 3.0)	3.0 (2.0, 4.0)	<0.001
Heart rate (beats/min)	138.5±24.4	146.5±16.6	0.084
QT maxmum (ms)	276.0 (250.0, 299.0)	280.0 (255.0, 292.5)	0.87
QT minimum (ms)	245.0 (224.0, 278.0)	240.0 (220.0, 240.0)	0.01
QTbc maxmum (ms)	411.7±24.4	430.2±28.6	0.004
QTbc minimum (ms)	376.5±24.2	365.9±19.0	0.058
QTfc maximum (ms)	359.2±25.7	370.6±26.9	0.072
QTfc minimum (ms)	328.4±26.3	315.2±18.1	0.026
QTD (ms)	20.0 (20.0, 30.0)	40.0 (30.0, 50.0)	<0.001
QTbcD (ms)	31.6 (28.3, 44.0)	66.6 (50.5, 76.3)	<0.001
QTfcD (ms)	27.4 (25.2, 39.1)	55.2 (42.4, 66.3)	<0.001

ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; IVIG, intravenous immunoglobulin; Plt, platelet count; QTD, QT interval dispersion; QTbcD, corrected QT interval dispersion by Bazett’s formula; QTfcD, corrected QT interval dispersion by Fridericia’s formula.

In the univariate Spearman’s rank correlations, although the correlation coefficient was very low, QTfcD was associated with sodium (R=0.28, P=0.011), AST (R=0.23, P=0.034) and Gunma score (R=0.33, P=0.002, Table 2).

Table 2. Univariate Spearman’s Rank Correlations Between QTfcD and Laboratory Data and Conventional Risk Scores for Kawasaki Disease

	QTfcD (ms)
	R	P value
Sodium (mmol/L)	0.28	0.011
AST (IU/L)	0.23	0.034
ALT (IU/L)	0.18	0.098
Plt (104/μL)	0.024	0.83
Neutrophils (%)	0.17	0.12
CRP (mg/dL)	0.17	0.12
Gunma score	0.33	0.002
Kurume score	0.20	0.06

Abbreviations as in Table 1.

The incidence of CAA during and after therapy was higher in the IVIG nonresponders, although there was no significant statistical difference (R group vs. N group: 6 [9.7%] vs. 5 [20.8%], P=0.17 by Chi-square test).

Prediction of IVIG Nonresponders

According to receiver-operating characteristic (ROC) curve analysis, the area under the curve (AUC) of QTfcD (0.89) was similar to those of the conventional risk scores (Gunma: 0.88, Kurume: 0.76). According to a cutoff value of 42.6 ms, the sensitivity and specificity of QTfcD were 75.0% and 88.7%, respectively (Figure 2).

Figure 2.

Receiver-operating characteristic (ROC) curve. The area under the curve (AUC) for corrected QT dispersion (0.89) was similar to that for the conventional risk scores (Gunma=0.88, Kurume=0.76).

In the univariate logistic regression analysis, QTD, QTbcD, QTfcD, Gunma score, and Kurume score significantly correlated with an increased incidence of IVIG non-response. There were no significant differences in AST or platelet count values between responders and nonresponders to IVIG therapy (Table 3).

Table 3. Univariate Logistic Regression Analysis

Variable	Univariate analysis
	OR (95% CI)	P value
Sodium	0.455 (0.309–0.668)	<0.001
AST	1.002 (1.000–1.004)	0.079
ALT	1.003 (1.000–1.005)	0.027
Plt	0.964 (0.919–1.010)	0.13
Neutrophils	1.131 (1.066–1.199)	<0.001
CRP	1.247 (1.089–1.429)	0.001
Gunma score	2.136 (1.547–2.950)	<0.001
Kurume score	2.228 (1.453–3.415)	<0.001
QTD	1.174 (1.095–1.258)	<0.001
QTbcD	1.108 (1.062–1.156)	<0.001
QTfcD	1.129 (1.073–1.187)	<0.001

CI, confidence interval; OR, odds ratio. Other abbreviations as in Table 1.

In the multivariate logistic analysis after adjustment for Gunma score, QTfcD was found to be a significant and independent predictive factor of a response to IVIG (odds ratio (OR): 1.133, 95% confidence interval (CI): 1.061–1.210, P<0.001) (multiple-adjusted model 1, Table 4). Similar findings were seen after adjustment for Kurume score (OR: 1.132, 95% CI: 1.071–1.197, P<0.001) (multiple-adjusted model 2, Table 4). The incremental value of QTfcD to predict the response to IVIG is shown in Figure 3. The QTfcD provided an additional benefit over conventional risk score (Gunma score).

Table 4. Multivariate Logistic Regression Analysis

Variable	Multiple-adjusted
	Model 1		Model 2
	OR (95% CI)	P value	OR (95% CI)	P value
QTfcD	1.133 (1.061–1.210)	<0.001	1.132 (1.071–1.197)	<0.001
Gunma score	2.235 (1.405–3.554)	0.001	–	–
Kurume score	–	–	1.344 (1.344–4.930)	0.004

Abbreviations as in Tables 1,3.

Figure 3.

Incremental value of QT dispersion over conventional risk score. Addition of corrected QT dispersion (QTfcD) to a Cox model of the Gunma score resulted in a significant improvement in the predictive value for unresponsiveness to immunoglobulin therapy in a patient with Kawasaki disease.

QTD as an Indicator of Inflammation

We examined whether QTD was related to the grade of inflammation among nonresponders to IVIG therapy. There was no remarkable difference between baseline QTD and after second IVIG therapy. However, QTD was significantly improved at 2 and 4 weeks after the onset of KD following commencement of additional treatment (Figure 4A). Similar results were obtained by QTbcD and QTfcD (Figure 4B,C).

Figure 4.

Changes of QT interval dispersion (QTD) and corrected QT interval dispersion (QTcD) during the clinical course of nonresponders to intravenous immunoglobulin (IVIG) therapy. QTD (A) and QTcD (B,C) improved significantly at 2 weeks and 4 weeks after the onset of Kaswasaki disease following commencement of additional treatment. *P<0.01 vs. value at baseline; ^†P<0.01 vs. value after 2nd IVIG therapy; ^‡P<0.05 vs. value after 2nd IVIG therapy.

Discussion

QTD is an indicator of the degree of myocardial repolarization inhomogeneity that can prognosticate heart failure and fatal arrhythmia in adults and ventricular arrhythmia, cardiomyopathy, and cardiac sudden death in both children and adults.^1–8 In the present study, QTD and QTcD at baseline were significantly associated with a response to IVIG therapy in KD patients.

Higham et al reported that a prolonged QTD was caused by regional changes in action potential duration and conduction, local populations of cells with afterdepolarizations, and neurohormonal factors, all of which are affected by myocardial ischemia.²³ Osada et al described increased QTD as correlated with the severity of coronary lesions in KD, and that >60 ms dispersion had a higher sensitivity in detecting severe coronary aneurysms.^24,25 One possible mechanism of the increased QTD in KD is myocardial ischemia caused by coronary involvement. Thus, QTD may have been influenced by the extent of ischemic myocardial damage. In the present study, although nonresponders had more than double the incidence of abnormal coronary lesions, there was not a significant difference, because of the small number of patients. The patients with coronary dilatations, which all transiently regressed, did not have ischemic sign such as chest pain, or abnormal ECG findings.

Meanwhile, others have found no correlation between QT interval and coronary involvement. Moreover, QTD was higher in acute-phase KD patients without overt CAA as compared with a healthy control group.^26,27 Amoozgar et al noted that increased QTD in KD patients might also be caused by cardiovascular inflammation, such as carditis.²⁷ The observed cardiac pathologies in acute-stage KD include early vasculitis with coronary artery involvement, pericarditis, myocarditis, endocarditis, valvulitis, and conduction system inflammation. These changes are followed by coronary artery panvasculitis and aneurysm formation, and in rare cases progress to chronic myocarditis or cardiomyopathy caused by myocarditis.^28–32 A significant increase in cardiac troponin I level has been documented before IVIG therapy in acute-stage KD, suggesting that acute myocarditis or myocardial cell injury begins in the early phase of the disease and that myocyte injury or inflammation is likely to occur before the onset of definitive clinical symptoms and signs of myocarditis.³³

QTD has been associated with cardiovascular involvement and disease activity in systemic cardiovasculitides such as SLE and Behçet’s disease.^13,14 Hence, prolonged QTD is a potentially useful method of early detection of subclinical cardiac involvement, such as pericarditis, myocarditis, and coronary arteritis, similar to that in KD. As KD has been classified as a systemic vasculitis syndrome, we hypothesized that QTD could enable assessment of disease severity. In our study, QTD and QTcD were presumed to be associated with the grade of inflammation during the clinical course because they both decreased after additional therapy (e.g., infliximab, plasma exchange) in nonresponders who might have had severe cardiovasculitis. This finding indicated the possibility that QTD at baseline reflected the severity of myocardial inflammation at the onset of KD. Accordingly, QTD may be predictive of a response to IVIG therapy, in addition to conventional risk scores. QTD could be helpful for making decisions about alternative initial therapy or additional therapy following IVIG (e.g., IVIG plus steroid and additional infliximab, steroid following initial IVIG) because there are IVIG nonresponders who are not identified even by the conventional risk scores.

Measurement of QTD and QTcD is easy, safe, and noninvasive for evaluation of the risk of resistance to IVIG therapy. However, the relatively small number of subjects in this study reflects the difficulty in precisely evaluating the effect of QTD and QTcD values on the response to IVIG in KD. A large, prospective, multicenter study may resolve this limitation and determine the cutoff value of QTD as a predictor of KD prognosis. Further investigations are warranted to clarify the mechanisms underlying QTD and to fine-tune the use of this parameter in routine clinical practice.

Conclusion

QTD and QTcD were significantly associated with a response to IVIG in KD patients. Subclinical myocardial involvement may be indicated by repolarization abnormalities. In patients with KD, this parameter might represent a useful and noninvasive marker for the identification of IVIG nonresponders with severe vasculitis and a high risk of coronary aneurysm.

Acknowledgments

We thank all the physicians at the participating hospitals for their valuable help with data collection: Department of Pediatrics, Hokushin General Hospital; Department of Pediatrics, Nagano Red Cross Hospital; Department of Pediatrics, Nagano Municipal Hospital; Department of Pediatrics, Nagano Prefectural Suzaka Hospital; Department of Pediatrics, Shinonoi General Hospital; Department of Pediatrics, Shinshu Ueda Medical Center; Department of Pediatrics, Saku Central Hospital; Department of Pediatrics, Matsumoto Medical Center; Department of Pediatrics, Nagano Prefectural Kiso Hospital; Department of Pediatrics, Ina Central Hospital; Department of Pediatrics, Suwa Red Cross Hospital, and Department of Pediatrics, Iida Municipal Hospital.

Sources of Funding

No external funding.

Disclosures

All authors have no financial relationships relevant to this article disclose.

Conflict of Interests

All authors have no conflicts of interest to disclose.

References

• 1.   Stankovic I, Putnikivic B, Janicijevic A, Jankovic M, Cvjetan R, Pavlovic S, et al. Myocardial mechanical and QTc dispersion for the detection of significant coronary artery disease. Eur Heart J 2015; 16: 1015–1022.
• 2.   Zimarino S, Tatasciore A, Marazia S, Torge G, Di Iorio C, De Caterina R. Defective recovery of QT dispersion predicts late cardiac mortality after percutaneous coronary intervention. Heart 2011; 97: 466–472.
• 3.   Pan KL, Hsu JT, Chang ST, Chung CM, Chen MC. Prognostic value of QT dispersion change following primary percutaneous coronary intervention in acute ST elevation myocardial infarction. Int Heart J 2011; 52: 207–211.
• 4.   Dalient L, Rizzoli G, Menti L, Baratella MC, Turrini P, Nava A, et al. Accuracy of electrocardiographic indices in predicting life threatening ventricular arrhythmias in patients operated for tetralogy of Fallot. Heart 1999; 81: 650–655.
• 5.   Motoki N, Shimizu T, Akazawa Y, Saito S, Tanaka M, Yanagisawa R, et al. Increased pretransplant QT dispersion as a risk factor for development of cardiac complications during and after preparative conditioning for pediatric allogeneic hematopoietic stem cell transplantation. Pediatr Transplant 2010; 14: 986–992.
• 6.   Huh J, Noh CI, Yun YS. The usefulness of surface electrocardiogram as a prognostic predictor in children with idiopathic dilated cardiomyopathy. J Korean Med Sci 2004; 19: 652–655.
• 7.   Yetman AT, Hamilton RM, Benson LN, McCrindle BW. Long-term outcome and prognostic determinants in children with hypertrophic cardiomyopathy. J Am Coll Cardiol 1998; 32: 1943–1950.
• 8.   Das BB, Sharma J. Repolarization abnormalities in children with a structurally normal heart and ventricular ectopy. Pediatr Cardiol 2004; 25: 354–356.
• 9.   Tamaki S, Yamada T, Okuyama Y, Morita T, Sanada S, Tsukamoto Y, et al. Cardiac iodine-123 metaiodobenzylguanidine imaging predicts sudden cardiac death independently of left ventricular ejection fraction in patients with chronic heart failure and left ventricular systolic dysfunction: Results from a comparative study with signal-averaged electrocardiogram, heart rate variability, and QT dispersion. J Am Coll Cardiol 2009; 53: 426–435.
• 10.   Mir TS, Läer S, Eiselt M, Falkenberg J, Gottschalk U, Weil J. Effect of carvedilol on QT duration on paediatric patients with congestive heart failure. Clin Drug Invest 2004; 24: 9–15.
• 11.   Burns JC, Capparelli EV, Brown JA, Newburger JW, Glode MP. Intravenous gamma-globulin treatment and retreatment in Kawasaki disease: US/Canadian Kawasaki Syndrome Study Group. Pediatr Infect Dis J 1998; 17: 1144–1148.
• 12.   Tremoulet AH, Best BM, Song S, Wang S, Corinaldesi E, Eichenfield JR, et al. Resistance to intravenous immunoglobulin in children with Kawasaki disease. J Pediatr 2008; 153: 117–121.
• 13.   Kobayashi T, Inoue Y, Takeuchi K, Okada Y, Tamura K, Tomomasa T, et al. Prediction of intravenous immunoglobulin unresponsiveness in patients with Kawasaki disease. Circulation 2006; 113: 2606–2612.
• 14.   Egami K, Muta H, Ishii M, Suda K, Sugahara Y, Iemura M, et al. Prediction of resistance to immunoglobulin treatment in patients with Kawasaki disease. J Pediatr 2006; 149: 237–240.
• 15.   Kojuri J, Nazarinia MA, Ghahartars M, Mahmoody Y, Rezaian Gr, Liaghat L. QT dispersion in patients with systemic lupus erythematosus: The impact of disease activity. BMC Cardiovasc Disord 2012; 12: 11.
• 16.   Göldeli O, Ural D, Komsuoğlu B, Ağaçdiken A, Dursun E, Cetinarslan B. Abnormal QT dispersion in Behçet’s disease. Int J Cardiol 1997; 61: 55–59.
• 17.   Ayusawa M, Sonobe T, Uemura S, Ogawa S, Nakamura Y, Kiyosawa N, et al. Kawasaki Disease Research Committee. Revision of diagnostic guidelines for Kawasaki disease (the 5th revised edition). Pediatr Int 2005; 47: 232–234.
• 18.   Research Committee on Kawasaki Disease. Report of subcommittee on standardization of diagnostic criteria and reporting of coronary artery lesions in Kawasaki disease. Tokyo, Japan: Ministry of Health and Welfare, 1984 (in Japanese).
• 19.   Yan GX, Antzelevitch C. Cellular basis for the normal T wave and the electrocardiographic manifestations of the long-QT syndrome. Circulation 1998; 98: 1928–1936.
• 20.   Karpanou EA, Vyssoulis GP, Psichogios A, Malakou C, Kyrozi EA, Cokkinos DV, et al. Regression of left ventricular hypertrophy results in improvement of QT dispersion in patients with hypertension. Am Heart J 1998; 136: 765–768.
• 21.   Bazett HC. An analysis of the time-relations of electrocardiograms. Heart 1920; 7: 353–370.
• 22.   Sagie A, Larson MG, Goldberg RJ, Bengtson JR, Levy D. An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study). Am J Cardiol 1992; 70: 797–801.
• 23.   Higham PD, Furniss SS, Campbell RW. QT dispersion and components of the QT interval in ischaemia and infarction. Br Heart J 1995; 73: 32–36.
• 24.   Osada M, Tanaka Y, Komai T, Maeda Y, Oishi M, Sugiyama H, et al. QT dispersion and Kawasaki disease after coronary bypass surgery. Intensive Care Med 2000; 26: 1009.
• 25.   Osada M, Tanaka Y, Komai T, Maeda Y, Kitano M, Komori S, et al. Coronary arterial involvement and QT dispersion in Kawasaki disease. Am J Cardiol 1999; 84: 466–468.
• 26.   Ghelani SJ, Singh S, Manojkumar R. QT interval dispersion in North Indian children with Kawasaki disease without overt coronary artery abnormalities. Rheumatol Int 2011; 31: 301–305.
• 27.   Amoozgar H, Ahmadipour M, Amirhakimi A. QT dispersion and T wave peak-to-end interval in dispersion in children with Kawasaki disease. Int Cardiovasc Res J 2013; 7: 99–103.
• 28.   Fugiwara T, Fugiwara H, Nakano H. Pathologic features of coronary arteries in children with Kawasaki disease in which coronary arterial aneurysm was absent at autopsy: Quantitative analysis. Circulation 1988; 78: 345–350.
• 29.   Amano S, Hazama F, Kubakawa H, Tasaka K, Haebara H, Hamashima Y. General pathology of Kawasaki disease: On the morphological alterations corresponding to the clinical manifestations. Acta Pathol Jpn 1980; 30: 681–694.
• 30.   Yutani C, Go S, Kamiya T, Hirose O, Misawa H, Maeda H, et al. Cardiac biopsy of Kawasaki disease. Arch Pathol Lab Med 1981; 105: 470–473.
• 31.   Yonesaka S, Takahashi T, Furukawa H, Matubara T, Tomimoto K, Oura H, et al. Histopathological analysis of myocardial damages following Kawasaki disease with repeated endomyocardial biopsy. Kokyu To Junkan 1992; 40: 375–381.
• 32.   Sumitomo N, Karasawa K, Taniguchi K, Ichikawa R, Fukuhara J, Abe O, et al. Association of sinus node dysfunction, atrioventricular node conduction abnormality and ventricular arrhythmia in patients with Kawasaki disease and coronary involvement. Circ J 2008; 72: 274–280.
• 33.   Kim M, Kim K. Elevation of cardiac troponin I in the acute stage of Kawasaki disease. Pediatr Cardiol 1999; 20: 184–188.