Clinical manifestations and risk factors of coronary artery lesions in children with Kawasaki disease

Abstract

To analyze the clinical manifestations of children with Kawasaki disease (KD), and risk factors of coronary artery lesion (CAL). A total of 223 patients admitted to Anhui Children Hospital from January 2017 to December 2019 were enrolled. According to the clinical data, the children with KD were divided into complete Kawasaki disease (CKD) and incomplete Kawasaki disease (IKD) groups. According to the results of the cardiac color ultrasound, the children were divided into the CAL and nCAL groups. The clinical symptoms of children with KD were compared between the CKD and IKD groups. The risk factors of CAL were analyzed by univariate and binary logistic regression analyses. The incidence constituent ratio of KD increased annually from 2017 to 2019 (P < .05). The proportion of fever duration no longer than 10 days, chapped lips, fingertip decrustation, perianal desquamation, and fever combined with rash in the CKD group was significantly higher compared to the IKD group (P < .05), while intravenous immunoglobulin non-response and CAL were significantly lower than those in the IKD group (P < .05). The proportion of males, age <1 year, fever duration longer than 10 days, and IKD in the CAL group were significantly higher compared to the nCAL group, while hemoglobin levels were significantly lower than that in the nCAL group (P < .05). Sex, age, fever duration, atypical KD, and hemoglobin levels were risk factors for CAL in children with KD. Persistent fever, conjunctival hyperemia, chapped lips, and rash were common clinical symptoms in children with KD. The risk of CAL was relatively higher in children with low hemoglobin levels and IKD, whose ages were <1 year old and whose fever time was more than 10 days, which requires high clinical vigilance.

1. Introduction

Kawasaki disease (KD) is an acute febrile rash disease with systemic vasculitis as the main lesion. It usually occurs in children under the age of 5 years, with symptoms including fever, rash, and non-suppurative lymph node enlargement in the neck. The pathogenesis of this disease remains unclear, and the lesions often involve coronary arteries, which seriously threaten the life and health of these children.^[1] At present, the clinical diagnosis of KD lacks specific laboratory indicators, and preliminary judgment is often made based on the clinical symptoms presented by these children.^[2] It is reported that 15% to 36.2% of the children with KD have atypical clinical symptoms.^[3] Therefore, summarizing and analyzing the clinical symptoms of children with KD is of significance for the accurate assessment of the disease. Coronary artery lesion (CAL) is one of the most serious complications in children with KD,^[4] but there is no consensus on the risk factors for CAL in these cases. Based on this background, clinical data of children were collected and retrospectively analyzed. The results have been described below.

2. Data and methods

2.1. General data

A total of 227 cases of KD admitted to our hospital from January 1, 2017, to December 31, 2019, were enrolled.

Inclusion criteria were as follows: complete clinical data and those conforming to KD diagnostic criteria established by revision of diagnostic guidelines for Kawasaki disease (6th revised edition).^[5]

The diagnostic criteria of complete Kawasaki disease (CKD):

• (1)Fever, and have at least 4 of the following 5 major clinical features.
• (2)Fever, at least 3 of the following 5 major clinical features, and coronary artery abnormality by echocardiography.
• (3)Five major clinical features:

  • (i) Bilateral bulbar conjunctival congestion;

  • (ii) Oral and labial Changes: dry red lips, strawberry tongue, diffuse congestion of oropharyngeal mucosa;

  • (iii) Rash, (including redness at the site of Bacille Calmette Guerin inoculation).

  • (iv) Changes of peripheral extremities: (Initial stage) reddening of palms and soles, edema. (Convalescent stage) periungual desquamation.

  • (v) Non-suppurative cervical lymph node enlargement.

The diagnostic criteria of incomplete Kawasaki disease (IKD): Fever ≥5 days, but <4 major clinical features.

• (1)Fulfilling 3 major clinical features without coronary artery dilatation, but with some other significant laboratory features and other diseases are ruled out;
• (2)Only 1 or 2 major clinical features, excluding other diagnoses, may also be considered IKD;
• (3)Other significant laboratory features:

  • (i) Elevation of hepatic transaminases early in the course of the disease.

  • (ii) Increased leukocytes in the urine sediment of an infant.

  • (iii) Thrombocytosis in the convalescent phase.

  • (iv) Elevation of BNP or NT-pro BNP.

  • (v) Mitral valve regurgitation or pericardial effusion by echocardiography.

  • (vi) Enlargement of the gallbladder (hydrops of gallbladder).

  • (vii) Hypoalbuminemia or hyponatremia.

Exclusion criteria were as follows: malignant tumors (including solid and hematological), or congenital malformations; dysfunction of important organs; primary cardiomyopathy, and incomplete case records.

A total of 4 cases were excluded and 223 cases were included. There were 223 KD children, including 135 males and 88 females, with a male-to-female ratio of 1.53:1. The average age was 23.30 (10.43, 34.00) months ranging from 2 months to 10 year-and-six month-old. This study was reviewed and approved by the Medical Ethics Review Committee of the Anhui Provincial Children Hospital.

2.2. Research methods

1. Data collection: sex, age, course of the disease, the season of onset, duration of fever, treatment, laboratory indicators (including hemoglobin, white blood cell count, neutrophil count, platelet count, etc), and clinical manifestations.

2. CAL judgment standards^[6]: normal coronary artery diameter <2.5 mm in children <3 years old; <3.0 mm in children ≤3 years old ~ <9 years old, and <3.5 mm in children ≤9 years old ~ <14 years old. Coronary artery diameter > normal value but <4.0 mm was defined as coronary artery dilation. A coronary aneurysm was defined at 4.0 to 8.0 mm was defined as, with >8.0 mm suggesting a large coronary aneurysm.

3. Grouping method of KD children: according to the corresponding clinical data, children with KD were divided into CKD and IKD groups. According to the results of the cardiac color ultrasound, children with KD were divided into the CAL and nCAL groups.

2.3. Analysis indexes

Comparison of KD incidence proportion from 2017 to 2019; Comparison of the clinical symptoms of CKD and IKD; Univariate analysis of CAL in children with KD; and Binary Logistic regression analysis of CAL in children with KD.

2.4. Statistics

The SPSS23.0 software was used for data processing. Age and other measurement data were presented as (x ± s), and an independent sample t test was performed for comparison between groups. Count data were described by n (%), and the comparison between the 2 groups was performed using the χ² test. The trend of the chi-square test was used to compare the KD component ratio from 2017 to 2019, and the risk factors of CAL in children with KD were analyzed by binary logistic regression. P < .05 was considered a statistically significant result.

3. Results

3.1. Comparison of KD incidence rates from 2017 to 2019

The detection rate of KD increased annually from 2017 to 2019 (P < .05), as shown in Table 1.

Table 1.

Comparison of KD incidence rates from 2017 to 2019.

Time	KD (n)	Constituent ratio (%)
2017	68	30.49
2018	72	32.29
2019	83	37.22
合计	223	100.00
Trend chi-square value	32.621
P	<.001

KD = Kawasaki disease.

3.2. Comparison of the clinical symptoms of CKD and IKD

Clinical data showed that 223 children with KD were diagnosed, of whom 207 had CKD and 16 presented with IKD. The proportion of fever duration ≤10 days, chapped lips, peeling fingers, perianal skin peeling, and rash cases in the CKD group was significantly higher compared to the IKD group (P < .05). The proportion of intravenous immune globulin (IVIG) non-response and CAL in the IKD group was significantly higher compared to the CKD group (P < .05), with no significant differences in other clinical data between the 2 groups (P > .05), as shown in Table 2.

Table 2.

Comparison of the clinical manifestations between the CKD and IKD groups.

Clinical manifestations	–	CKD (n = 207)	IKD (n = 16)	χ2/continuous correctionχ2	P
Sex [n, %]	Male	128 (61.84)	10 (62.50)	0.003	.958
	Female	79 (38.16)	6 (37.50)
Age (yr) [n, %]	<1	45 (21.74)	2 (12.50)	0.308*	.579
	1~12	162 (78.26)	14 (87.50)
Duration of fever (d) [n, %]	≤10	168 (81.16)	7 (43.75)	12.305	<.001
	>10	39 (18.84)	9 (56.25)
Onset seasons [n,%]	Autumn and winter	113 (54.59)	6 (37.50)	1.743	.187
	Spring and summer	94 (45.41)	10 (62.50)
Clinical symptoms [n, %]	Bulbar conjunctival hyperemia	122 (58.94)	9 (56.25)	0.028	.868
	Chapped lips	160 (77.29)	4 (25.00)	18.273*	<.001
	Peeling of the fingers	98 (47.34)	3 (18.75)	4.900	.027
	Perianal skin peeling	86 (41.55)	1 (6.25)	7.777	.005
	Rash	96 (46.38)	3 (18.75)	4.592	.032
	Cervical mass	17 (8.21)	2 (12.50)	0.015*	.904
	Strawberry-like tongue	152 (73.43)	10 (62.50)	0.428*	.513
	IVIG non-response	0 (0.00)	3 (18.75)	26.342*	.000
	Digestive tract symptom	7 (3.38)	1 (6.25)	0.011*	.918
CAL [n, %]	Yes	60 (28.99)	10 (62.5)	5.058	.025
	No	147 (71.01)	6 (37.5)

CAL = coronary artery lesion, CKD = complete Kawasaki disease, IKD = incomplete Kawasaki disease, IVIG = intravenous immune globulin.

^*Represents continuous correction χ² test.

3.3. Univariate analysis for CAL in children with KD

The proportion of male sex, age <1 year, fever duration >10 days, and IKD in the CAL group was significantly higher compared to the local group, while the hemoglobin levels were significantly lower compared to the CAL group, with statistical significance (P < .05), as shown in Table 3.

Table 3.

Univariate analysis for CAL in children with KD ($\overline{x}\pm s$).

Clinical manifestations		CAL (n = 70)	nCAL (n = 153)	χ2/t	P
Sex	Male	66	72	45.413	.000
	Female	4	81
Age (yr)	<1	29	18	25.406	.000
	1~12	41	135
Duration of fever (d)	≤10 d	40	135	27.487	.000
	>10 d	30	18
Initiation time of IVIG use (d)	<7 d	30	65	0.003	.958
	≥7 d	40	88
IKD	Yes	10	6	7.746	.005
	No	60	147
Mycoplasma infection	Yes	18	29	1.319	.251
	No	52	124
Red blood cell count	(×1012/L)	4.24 ± 1.03	3.99 ± 1.04	1.671	.096
White blood cell count	(×109/L)	12.25 ± 1.06	11.98 ± 1.03	1.800	.073
Neutrophil count	(×109/L)	4.03 ± 1.01	3.86 ± 0.99	1.183	.238
Hemoglobin	(g/L)	101.24 ± 10.65	127.24 ± 9.99	17.664	.000
PCT*	(ng/mL)	1.11 ± 0.12	1.09 ± 0.10	1.300	.195
CRP	(mg/L)	81.25 ± 6.03	80.09 ± 6.21	0.161	.872
ESR	(mm/h)	65.62 ± 4.57	64.92 ± 4.69	1.043	.298
PLT	(×109/L)	550.66 ± 232.78	512.16 ± 165.62	1.247	.215

CAL = coronary artery lesion, CRP = C-reactive protein, ESR = erythrocyte sedimentation rate, IKD = incomplete Kawasaki disease, IVIG = intravenous immune globulin, KD = Kawasaki disease, PCT = procalcitonin, PLT = platelet.

3.4. Binary logistic regression analysis of CAL in children with KD

With the incidence of CAL in KD children as the dependent variable (CAL = 1, non-CAL = 0), sex (male = 1, female = 0), age (<1 year = 1, ≥1 year = 0), duration of fever (>10 days = 1, >10 days = 0), IKD (yes = 1, no = 0), and hemoglobin levels (≤100 g/L = 1100 g/L = 0) were the independent variables, and a binary logistic regression analysis was performed. The results showed that sex, age, duration of fever, IKD, and hemoglobin levels were risk factors for CAL in children with KD, as shown in Table 4.

Table 4.

Binary logistic regression analysis for CAL in children with KD.

Factor	B	SE	Wald	P	OR	95%CI
Sex	−1.029	0.376	7.477	.006	0.367	(0.171, 0.743)
Age	0.921	0.363	6.422	.011	2.511	(1.232, 5.118)
Duration of fever	1.381	0.560	6.088	.014	3.980	(1.329, 11.925)
IKD	−1.488	0.532	7.831	.005	0.226	(0.080, 0.640)
Hemoglobin	1.219	0.371	10.775	.001	3.384	(1.634, 7.006)
Constant	−0.030	0.564	0.003	.957

CAL = coronary artery lesion, IKD = incomplete Kawasaki disease, KD = Kawasaki disease.

4. Discussion

The incidence of cardiovascular damage caused by delayed treatment in children with KD is about 10% to 45%,^[7] among which CAL is one of the common complications endangering the life and health of these children.^[8] At present, the diagnosis of KD in China is mainly based on clinical manifestations, and a lack of objective laboratory diagnostic indicators increases the difficulty in the clinical diagnosis and treatment of KD; misdiagnosis and missed diagnosis are frequent. Studies suggest that the under-diagnosis and treatment of CAL gradually progress to coronary artery obstruction or calcification, causing the formation of thrombosis and increasing the risk of death among these children.^[9] At present, factors influencing CAL in children with KD are largely unclear. This study analyzed the clinical symptoms and risk factors of CAL in children with KD to provide a reference for the effective diagnosis and treatment of these children.

The results of this study showed that the ratio of male-to-female KD patients was 1.53:1, and the incidence of KD in males was significantly higher than that in females. This may be related to the fact that boys are naturally active and often play in crowded places, easily causing cross-infection; infection is a possible mechanism of KD pathogenesis. From 2017 to 2019, the proportion of KD incidence in our hospital showed an increasing trend annually, unlike the results reported by Liu Hui et al^[10] whereby the detection rate of children with KD increased first and then decreased from 2007 to 2016. This may be because the detection rate of children with KD increased significantly due to the obvious improvement in imaging technology and pediatrician diagnosis and treatment experience in recent years. It may also be related to the small sample size of this study but clinical attention should nonetheless be paid to the clinical diagnosis and treatment of children with KD. Due to the lack of typical symptoms in children with IKD, clinical diagnosis is difficult, and delayed treatment can increase the risk of coronary artery injury. Currently, no specific laboratory indicators for the differential diagnosis of CKD and IKD cases are available. Fever duration ≤10 days, chapped lips, peeling fingers, perianal skin peeling, and rash cases in the CKD group were significantly higher than those in the IKD group, and IVIG non-response and CAL in the IKD group were significantly higher compared to the CKD group, indicating that clinical symptoms like persistent fever, chapped lips, rash, and peeling fingers are more common in children with KD. Among them, the duration of fever in children with IKD is longer as compared to children with CKD, and the risk of children with IKD being non-response to IVIG and CAL is significantly higher relative to children with CKD. In clinical settings, children with long-duration of fever and those IVIG non-responders should be paid high attention. There was no significant difference in the age of children with CKD and IKD, inconsistent with the conclusion of the obvious difference in the age distribution of children with CKD and IKD reported by Liu Ding et al,^[11] and the conclusion of foreign studies,^[12] whereby the age spectrum of children with IKD is more common in <1 year and >9 years. This may be because of the following reasons: in this study, the age range of children with KD was large and our center was concentrated on cases of children with KD, leading to a possible selection bias in the results of this study. Further studies are needed to confirm whether there is a significant difference in the age distribution between CKD and IKD.

Many factors affect the occurrence of CAL in children with KD. The results of a study conducted by Zhang Yanhan et al^[13] on the risk factors of CALs in children with KD showed a higher risk of CAL in boys reflected before IVIG treatment. This study also suggests that the male sex is a risk factor for CAL. As for overweight or CAL in boys, another study indicated that an increase in fever duration, C-reactive protein levels, platelet count, alanine aminotransferase levels, and aspartate aminotransferase levels are high-risk factors for KD complicated with CAL.^[14] Age <1 year was a risk factor for KD complicated with CAL, consistent with many domestic and foreign studies. This may be related to the fact that the clinical symptoms of small infants are not typical and easily missed or misdiagnosed, so they fail to receive IVIG treatment in time, leading to a significant increase in the incidence of CAL in this group. The youngest child with KD in this study was only 2 month-old.

The results of this study showed that sex, age, duration of fever, IKD, and hemoglobin levels were risk factors for CAL in children with KD, suggesting that boys, children aged <1 year, with a duration of fever >10 days, and reduced hemoglobin levels were at a significantly higher risk of CAL. Among them, the conclusion that boys and children with KD fever for a long duration are at a higher risk of CAL is consistent with those of domestic and foreign reports.^[15,16] Fever is the most common clinical manifestation in children with KD and is also an important factor for predicting coronary artery injury.^[17] When the fever lasts longer, the risk of CAL is higher. The more inflammatory mediators produced by the body, the more serious the vascular inflammation damage. Therefore, clinical attention should be paid to males and children with KD with a long fever duration. Early and effective diagnosis and treatment is the key to preventing the occurrence of CAL in children with KD. Children with KD often present with anemia, which may be caused by the decrease in hemoglobin levels, leading to tissue hypoxia, triggering sympathetic nerve excitation, over-activating the renin-angiotensin system, and damaging the coronary artery wall, thus promoting the occurrence and development of CALs. Therefore, a decrease in hemoglobin level suggests a high risk of CAL in children with KD.^[18]

5. Conclusion

In conclusion, the binary logistic regression analysis in this study suggested that clinical cases of KD in boys and young children, atypical symptoms, long fever duration, and reduced hemoglobin level should be paid great attention to owing to the possibility of CAL. Early and effective treatment is the key to preventing the occurrence of CAL. However, the limitations of this study lie in that the included children in this analysis were all from a single center, and there were few laboratory indicators affecting the incidence of CAL in the included children. Further research is needed to address these shortcomings in the future.

Acknowledgments

We thank all the doctors and nurses at Anhui Provincial Children Hospital for their help in conducting this study and all the parents and children who made it possible. We would like to thank KetengEdit (www.ketengedit.com) for its linguistic assistance during the preparation of this manuscript.

Author contributions

Conceptualization: Ling Wang, Biquan Chen, Xiaoyan Zeng.

Data curation: Ling Wang, Xiaoyan Zeng.

Formal analysis: Ling Wang.

Methodology: Ling Wang.

Software: Ling Wang.

Validation: Biquan Chen.

Writing – original draft: Ling Wang.

Writing – review & editing: Biquan Chen.