Article Text


Recurrent skin peeling following Kawasaki disease


Long term follow up of 259 cases of Kawasaki disease led to the observation that 11% of children have episodes of recurrent peeling of the skin for several years after their recovery. These events were usually associated with an upper respiratory tract infection and were distinct from a recurrence of Kawasaki disease. Repeeling was significantly less frequent in children who had suffered coronary artery dilatation and was more frequently seen in those with nasal staphylococcal colonisation. The mechanism for this phenomenon is unclear, but it has been observed in a number of other conditions caused by infectious agents and their toxins. Paediatricians need to be aware of this phenomenon which is distinct from recurrence of Kawasaki disease.

Statistics from

Children with Kawasaki disease (KD) usually have skin erythema which may involve the perineum or peripheries, one of the defining conditions of this condition. This is followed by peeling, keratolysis, or desquamation one to three weeks after the resolution of fever. A proportion of children may then develop recurrent episodes of skin peeling months or years after their initial illness. Similar desquamation has been reported in patients recovering from toxic shock syndrome and illness involving streptococcal toxins such as scarlet fever.1 We set out to document the phenomenon of repeeling in a cohort of children with KD.


The frequency of repeeling was investigated in a cohort of 259 British patients who had KD after 1986; the study therefore covered 1239 patient years in total (table 1). Patients met five or six of the six criteria for the diagnosis of Kawasaki syndrome as defined by the American Centres for Disease Control. All patients had peeling as a component of their initial illness. The cohort was compiled from two sources. One hundred and seventeen successive patients and their parents attending a long term Kawasaki disease follow up clinic between 1996 and 1999 were asked prospectively as part of their routine clinic visits about the health of their skin and episodes of peeling. These patients had full documentation of their original illness available in the clinical notes. Nasal staphylococcal colonisation was checked in 35 of these patients during follow up clinic visits; all positive cultures were checked for toxin production using reverse passive latex agglutination kits (Unipath, Basingstoke, UK). Four patients had skin peeling observed during an admission to hospital for other reasons; clinical records of these admissions were reviewed carefully.

Table 1

Comparison between the total cohort and those who suffered from repeeling

The remainder of the cohort was compiled using a retrospective questionnaire applied by a national patient help group for KD over the same time period. Of 195 questionnaires sent out, 142 fully completed answers were received. The original medical records were sought in 70 cases in which there were queries about aspects of the original diagnosis. The questionnaire covered the same areas as were reviewed in the prospective cohort. Data were analysed using a χ2test.


This patient group (table 1) was representative of the type of cohort reported by other authors in a previous British study.2 No significant differences were detected between data collected in the prospective outpatient group and the retrospective questionnaire; in particular the proportions of repeeling detected by these two methods were the same. None of the patients suffered from skin peeling prior to their episode of KD. This phenomenon would therefore appear to be linked in some way to their illness. None of the 190 siblings of the total patient cohort suffered from recurrent peeling, an observation supporting this statement. Repeeling episodes were most numerous in the first two years after diagnosis (139 episodes in the 29 patients); none was documented seven years after diagnosis. In the cohort 29 children suffered from recurrent peeling of their skin and two suffered from a recurrence of the full spectrum of Kawasaki disease. Recurrent Kawasaki disease in both cases was accompanied by prolonged high fever, enlarged lymph nodes, and by erythema of the skin for some seven to 14 days prior to peeling (as described in a Japanese cohort3).

Using the admission data available for 187 cases, there was no correlation between acute phase measures of KD, including platelet count (in all cases), white count (in all cases), eosinophil count (in 185 cases), or IgE concentration (in 42 cases) and the subsequent development of repeeling. All four children with repeeling during a hospital admission were admitted for an upper respiratory tract infection. In none of these cases was streptococci or staphylococci cultured from throat swabs, but on all occasions staphylococci were isolated from the nose (in the one inpatient tested, toxin screening of this organism showed no toxin production). Desquamation was observed on the second day of fever in one case, and between the third and fifth day of fever in the other three cases. None of the admissions showed evidence of erythema, conjuctivitis, or widespread lymphadenopathy and all responded rapidly to conservative measures or oral antibiotic for their infection. None had an increased eosinophil count or increase in urinary histamine during the period when their skin peeled.


The recurrence of skin peeling in patients who have previously suffered from KD appeared to be a relatively common entity seen in 11% of patients, documented approximately 188 times in 1259 patient years in this cohort. These data suggested that this condition is distinct from recurrence of the disease. Desquamation in the initial illness followed inflammatory changes in the skin of the hands and feet whereas repeeling developed within a few days of the beginning of a minor illness, or simultaneous with it, and was not associated with a recurrence of skin inflammation. Repeeling may persist for several years after the episode of Kawasaki disease before resolving with no long term sequelae. The areas of skin which repeeled were seldom the most severely affected areas during the first episode of KD.

With few exceptions the patients with repeeling were children who had not developed measurable coronary artery dilatation and therefore were thought by clinicians to have a more mild illness. A tendency to repeel did not relate to the timing of the commencement of therapy with immunoglobulin, the use of aspirin, or the timing of hospitalisation. The majority of children with repeeling were otherwise healthy. However, it was interesting to observe that children with notable behavioural sequelae or hearing problems following Kawasaki disease were observed at a higher frequency in the cohort which had recurrent desquamation.4-6 Distinctive clinical patterns such as this presumably represents a heterogeneity in the disease process, the host response, or both.

The mechanism of desquamation in KD is not clear. Following the observation that there was a disproportionately high rate of staphylococcal carriage in children with recurrent desquamation one may propose that release of superantigens or toxins from resident or colonising bacteria may play some role. A similar process has been proposed in two cases of recurrent perineal erythema, in whom the recurrence of erythema in conjunction with pharyngitis was attributed to toxin producing streptococci colonising the pharynx.7Challenging this rather convenient model is the observation that the majority of staphylococcal cultures in this cohort collected showed no toxin production. Alternatively recurrent peeling may represent abnormal apoptosis of epithelial cells,8 or abnormal cytokine release from resident immune cells in the skin. This latter process could include T cells rendered anergic by the initial illness (as has been proposed for other illnesses involving bacterial toxins and superantigens9). The similarity of this peeling rash to that seen in scarlet fever or other streptococcal infections and toxic shock syndrome, or the systemic administration of interferon alfa suggests a more generalised mechanism rather than a specific one.

In conclusion paediatricians and dermatologists should be aware of the phenomenon of repeeling following Kawasaki disease, and the importance of distinguishing between recurrent disease and this milder, and seemingly benign complication. Children with Kawasaki disease should have some element of long term follow up or clear information sheets in order to reassure parents about this non-cardiac complication of this disorder.


Ealing Paediatric Courses funding supported this study. We should like to thank Cathy Marriott and Mari Satou for their assistance in reviewing the Japanese literature, and for a review of the script by Professor Mike Dillon. We should also like to thank the families of patients who have helped with this study, and their clinicians for referring patients to the Kawasaki Clinic.

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.