Desquamation in Kawasaki Disease and its Relationship with Coronary Artery Abnormalities

Background Desquamation of hands and feet is a common characteristic of Kawasaki disease (KD), and occurs in 68-98% of patients diagnosed with KD. In this study, we analyzed patients’ varying desquamation levels in their hands or feet, in correlation with clinical presentation, to assess the relationship. Method We retrospectively reviewed children with KD who had follow-up visits in Kaohsiung Chang Gung Memorial Hospital’s pediatric outpatient department (OPD) during 2018 and 2019. We analyzed their age, laboratory data before intravenous immunoglobulin (IVIG) treatment, and coronary artery abnormalities (CAA) based on the desquamation level of their hands and feet. Results We enrolled a total 112 patients in the study, and 92 patients (82.1%) had hand or feet desquamation. We found the hands’ high-grade desquamation was positively associated with age and segmented neutrophil percentage (p=0.047 and 0.029, respectively) but negatively associated with lymphocyte and monocyte percentage (p=0.03 and 0.006, respectively). Meanwhile, the feet’s high-grade desquamation was positively associated with total white blood cell count (p=0.033). Furthermore, we found that high-grade hand desquamation had less probability of CAA formation compared with that of a low grade (7.1% vs. 40.8%, p=0.016). We observed no signicant difference in IVIG resistance between the groups. Conclusion This report is the rst to demonstrate that the desquamation level of hands or feet in Kawasaki disease is associated with different coronary artery abnormalities and laboratory ndings. Physicians should be cautious of CAA formation, particularly in patients with low-grade hand peeling.

desquamation, from ngertips (toes) to palm (sole) and compare them with their clinical characteristics in case studies.

Method
We retrospectively reviewed children with KD who had follow-up visits in Kaohsiung Chang Gung Memorial Hospital's pediatric outpatient department (OPD) during 2018 to 2019. Patients who were treated in another hospital during the acute phase or cases with incomplete medical records were excluded. KD was diagnosed in accordance with the American Heart Association (AHA) criteria, which includes fever for more than 5 days and typically having four of ve clinical features (bilateral nonexudative conjunctivitis, oral changes, cervical lymphadenopathy (>1.5cm in diameter), hand and feet edema or desquamation, and polymorphism skin rash), while incomplete KD was diagnosed as prolonged fever and two or three clinical features, plus CRP ≥ 30 mg/L and/or ESR ≥ 40 mm/hour and supplemental laboratory data (anemia by age, platelet count ≥450000/mm 3 after 7 th day of fever, albumin ≤ 3g/dL, elevated ALT level, WBC count ≥ 15000/mm 3 , urine ≥ 10 WBC/high power eld) or echocardiography ndings. Coronary artery abnormalities (CAA) are de ned as: (1) left anterior descending (LAD) coronary artery or right coronary artery (RCA) Z score of ≥2.5; (2) coronary artery aneurysm; (3) the presence of ≥3 other suggestive features, including decreased left ventricular function, mitral regurgitation, pericardial effusion, Z scores in LAD coronary artery, or RCA of 2 to 2.5 [1]. IVIG resistance was de ned as persistent fever at least 48 hours after initial IVIG therapy where a second IVIG infusion had to be administered to patients.
We classi ed groups into desquamation positive or not and then further divided into "hand desquamation" and "foot desquamation". We provided an illustration ( Figure 1) to identify the desquamation grade. In hands and feet, a classi ed number of 0 meant no desquamation, while 1 to 3 represented ngertip (toe tip), between ngertip (toe tip) and palm (sole) (Figure 2 and 3). High-grade desquamation was de ned as levels 2 and 3, while low-grade referred to level 1 (limited to ngertips or toe tips). Patients were divided into groups according to "desquamation or not" and "desquamation grades". We analyzed patients' age and laboratory data within 3 days prior to the administration of intravenous immunoglobulin (IVIG) and follow-up echocardiography series from admission to 1 to 2 years after disease onset. This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital (201600714A3).
We performed the Kolmogorov-Smirnova test to examine data distribution, the Kruskal-Wallis and Mann-Whitney tests for continuous variable analysis, and the Chi-square test for categorical variables. To determine the cut-off value, we used the receiver operating characteristic (ROC) curve, Youden index, and minimal distance. Data are presented as percentage and median with interquartile range (IQR). Results were considered statically signi cant when p-values <0.05. All statistical analysis was performed using SPSS 22.0 (SPSS, Inc., Chicago, Illinois).
This study included a total of 112 patients, and 92 of them (82.1%) had hand or foot desquamation. The laboratory data and IVIG resistant rate demonstrated no signi cant difference between patients with and those without desquamation (Table 1). We further analyzed data based on "hand desquamation" and "foot desquamation". In both categories, we further classi ed into sub-groups according to "Desquamation positive or negative" and "High-grade or low-grade" (Tables 2 and 3).
In the "hand desquamation" group (Table 2), 91 patients were positive for hand desquamation, and the distribution of grades 1, and 3 were 77, 7, and 7 cases, respectively. We found older age to be related to a positive desquamation nding (p=0.046). Furthermore, older age (p=0.047), elevated segmented white blood cell (WBC) percentage (p=0.029), lower lymphocyte (p=0.03), and lower monocyte (p=0.006) percentage were all associated with high-grade desquamation. Regarding age, the area under the curve (AUC) of hand desquamation positive is 0.64 ( Figure 4), with an optimal cut-off value of 2 years old. The desquamation positive rate was 97%(≧2 years old) and 74.7%(<2 years old), respectively (p=0.006).
In the "foot desquamation" group (Table 3), 63 patients were positive for foot desquamation, and the distribution of grades 1, 2, and 3 were 54, 2, and 7 cases, respectively. The desquamation positive group had higher C-reactive protein (CRP) but lower aspartate transaminase (AST) levels (p=0.023 and 0.045, respectively), while the high-grade desquamation group had a higher WBC count (p=0.033). We observed no signi cant difference in the IVIG-resistant rate between hand and foot desquamation.
With regard to coronary artery abnormalities (CAA), the total desquamation and hand and foot desquamation incidences were similar in patients with CAA and those without (Table 4). We further analyzed the subgroups and found that patients with CAA formation had a signi cantly lower incidence of only 3% accompanied with high-grade hand desquamation (p=0.016). Furthermore, high-grade hand peeling had a decreased incidence of CAA formation than low-grade hand peeling (7.1% vs. 40.8%, p=0.016) in two-year series echocardiography, while we had no signi cant nding in the feet subgroups.

Discussion
To the best of our knowledge, this study is the rst to discuss different levels of desquamation associated with the clinical symptoms and signs of KD. Epidermal turnover and desquamation are natural processes in mammals [8], but some diseases may progress these processes. For example, the toxin secreted by Staphylococcus aureus is related to the cleavage of keratinocyte junction and cell-to-cell adhesion in the epidermis [9], which may cause Staphylococcus aureus scalded skin syndrome. The activation of cutaneous T cells induced by the superantigen of streptococcus results in upregulating skinhoming capacity, and the rash of scarlet fever is followed by peeling in response to cytokine release and activated T-cell in ltration [10]. Hand-foot-mouth disease caused by the enterovirus also contributes to skin desquamation, but the related pathophysiology remains poorly understood [11,12]. The characteristics of common pediatric diseases involving fever and desquamation are compared in Table 5 [7, 11, 13-20].
Hand and foot desquamation have distinct characteristics in this study. The desquamation in KD mostly occurred during the convalescent phase of the disease [21]. We did not observe any signi cant difference in the laboratory ndings between patients with and those without desquamation. However, we further classi ed patients into hand or foot desquamation and began to observe differences. Hand desquamation was associated with older age, and high-grade hand peeling was related to older age, elevated segmented WBC percentage, and lower lymphocyte and monocyte percentages. Meanwhile, foot desquamation was associated with elevated CRP and AST, and high-grade foot desquamation was related to increased WBC count. In the AHA diagnosis criteria of atypical KD, supplemental laboratory tests include WBC count, CRP, and ALT level, all of which indicate the in ammation status of KD. Wang et al. revealed desquamation positive KD patients had lower WBC counts, higher band-form percentage, and elevated ALT level, but the age, gender, percentage neutrophils, hemoglobulin, CRP, and platelet count were similar [6]. In another study analyzing desquamation in KD, Kim et al. revealed that patients with desquamation had higher platelet counts and elevated AST and ALT [7]. In our study, the WBC counts were only associated with high-level foot desquamation; in contrast, higher AST levels were negatively related to foot desquamation. Both the band-form percentage and ALT level were similar in our subgroups. The two aforementioned studies did not further discuss differences in laboratory ndings in peeling severity in either the hands or the feet.
Interestingly, we found that high-grade hand peeling was related to a decreased incidence of CAA formation. Kim et al. showed no difference between groups with or without desquamation regarding coronary artery lesions [7]. However, Wang et al. also indicated that cases who did not peel had a higher probability of aneurysm formation [6], and another study revealed that recurrent skin peeling in KD was signi cantly less frequent in patients with coronary dilatation or aneurysm formation [22]. Desquamation seems to have a protective effect on CAA formation. Although the mechanism is not clear, we suppose that the more cytokines released to facilitate skin peeling, the fewer in ammatory factors in coronary arteries. Further research is required to identify the pathophysiology.
This study had certain limitations. We conducted a retrospective review of medical records, and some cases of desquamation were identi ed based on the reporting of family history, both of which may have recall bias and variable accuracy. Furthermore, this study was a single center experience, which is not su cient for re ecting the general condition nationwide. As for CAA formation, some patients who were diagnosed as having KD under 2 years of age, a longer echocardiography is needed to observe coronary artery changes. Both multi-center studies and longer follow-up times are needed to analyze the various grades of desquamation in KD with clinical features.

Conclusion
This report is the rst to study various severities of hand and foot periungual desquamation in KD associated with coronary artery abnormalities. Pediatricians should be aware of prolonged fever and periungual desquamation with unknown origin. Furthermore, the risk of coronary artery changes cannot be overlooked, and we should be particularly cautious with those who have low-grade hand desquamation in KD.

Declarations Ethics approval and consent to participate
This study was approved the Institutional Review Board of Chang Gung Memorial Hospital (201600714A3)

Consent for publication Not applicable
Availability of data and materials The datasets generated and analyzed during the current study are not publicly available due to strict ethical regulation of information privacy in Taiwan. The consent to participate does not include permission to make the data available to a third party

Competing interests
The authors declare that they have no competing interests Funding This study received funding from the following grants: MOST 108-2314-B-182 -037 -MY3 and MOST 103-2410-H-264-004 from the Ministry of Science and Technology of Taiwan and 8E0212 from Chang Gung Memorial Hospital in Taiwan. Although these institutes provided nancial support, they had no in uence on the way in which we collected, analyzed, or interpreted the data or wrote this manuscript.

Authors' contributions
All authors were involved in drafting the article or revising it critically and all authors approved the nal version to be published. MG, YH and WL contributed to conceptualization and study design of the study; JY and WL interpreted and statistically analyzed the data; JY was the major contributor in writing the manuscript. LC and HK were responsible for monitoring and data management, reviewing and editing the manuscript.
1999, 12(2):224-242.        The illustration to identify the grade of desquamation The desquamation of hands; the black arrows point out the margin of peeling; a: grade 1; b: grade 2; c: grade 3