Infectious trigger

The aetiology of KD remains unknown. Pronounced seasonality and clustering of KD cases have led to the hunt for infectious agents as a cause. So far, however, no single agent has been consistently identified.19–21 Many published reports implicate a number of bacterial and viral pathogens, including retroviruses, Epstein–Barr virus, coronavirus, propionibacterium acnes, staphylococcal and streptococcal superantigens, and unidentified virus particles as infectious triggers of KD.2 ,5 To date, no single pathogen has been confirmed in subsequent studies. The debate regarding the infectious cause of KD has centred around the mechanism of immune activation: conventional antigen versus superantigen (SAg).22 ,23 Superantigens seemed a plausible cause of the disorder in view of the clinical and immunological similarity between KD and staphylococcal and streptococcal superantigen-mediated disorders. Several studies have presented evidence supporting superantigen-triggered process, but others have not confirmed the association.5 ,22 ,24 One explanation for the inconsistent finding is that KD is triggered by a range of different superantigens, and animal studies suggest the possibility that both superantigen, and Toll-like receptor agonists may need to act synergistically.5 ,25 Against a superantigen (SAg)-mediated process is the report by Rowley et al26 of three fatal cases of KD in which IgA plasma cell infiltration into the vascular wall during the acute phase of the illness was observed. By examining the clonality of this IgA response using reverse transcriptase (RT)-PCR in lesional vascular tissue, these researchers observed that the IgA response was oligoclonal, suggesting a conventional antigenic process rather than a superantigenic-driven one.26 Although uncertainty remains regarding the mechanism(s) of initial immune activation, most authorities believe that one or more potentially ubiquitous infectious agents produces a deleterious host response in a genetically susceptible subject.2 ,5 ,9

Genetics

A genetic contribution to the risk of KD is suggested by the much higher risk of the disease in Asian children, particularly the Japanese and Koreans, which persists when patients of these ethnicities migrate to other countries27; from the observed increased relative risk to siblings of index cases compared with the general population, from twin studies and from well documented multicase families.28

Many candidate genes have previously been suggested, either as susceptibility genes for developing KD; or increasing risk of CAA.29–31 These are summarised elsewhere.31 Most of these earlier studies, however, failed to identify definitive genetic associations, emphasising the difficulties of the candidate gene approach for a disease where the pathogenesis is poorly understood.32 By contrast with the candidate gene approach, genome-wide association studies (GWAS) have the advantage of identifying disease-associated genes without requiring prior knowledge of the mechanisms involved.32 A number of GWAS of KD have been published so far.33–39 From these studies, several single nucleotide polymorphisms (SNPs) associated with susceptibility to KD, including ITPKC, ABCC4 and FCGR2A, CD40 and a gene region near FAM167A-BLK (table 1). Furthermore, other genes have been associated with non-response to intravenous immunoglobulin (IVIG) and risk of CAA, including CASP3, FCGR3B and genes of the TGF-β signalling pathway.40 ,41 It is likely that many other genetic factors have yet to be identified, as the GWAS methodology has so far only identified the most significant associations.32 Whole exome sequencing studies applied to individual KD cases of extreme phenotype, such as those with giant CAA, could clarify the contribution of rarer genetic variants in these extreme cases. Thus, the study of the genetic contribution to KD remains an intense area of research worldwide, and still very much a work in progress.

Vascular involvement

The main sites of clinically important vascular involvement are the coronary arteries, although other vessels such as the axillary arteries can be involved. CAA occur in 15–25% of untreated cases, with additional cardiac features in a significant proportion of these including pericardial effusion, electrocardiographic abnormalities, pericarditis, myocarditis, valvular incompetence, cardiac failure and myocardial infarction.6 When systemic arterial injury (major limb arteries, renal and other visceral vessels) occurs, it is rarely seen in the absence of CAA.2

Laboratory findings

KD is invariably associated with an inflammatory process, with elevation of ESR, CRP and white cell count.2 ,6 ,9 In the absence of significant inflammation, the diagnosis of KD is unlikely.2 ,6 ,9 Not all the inflammatory markers may be abnormal at first presentation, and repeat blood testing should be undertaken if there is diagnostic uncertainty. Thrombocytosis occurs towards the end of the second week of the illness and, therefore, may not be helpful, diagnostically, in the early stages.2 ,6 ,9 Acute thrombocytopenia or low/normal platelet count may occur and may be associated with a poorer prognosis (see below).2 ,6 ,9 ,47 Liver function may be deranged and some patients present with jaundice and elevation of transaminases.2 ,6 ,9 Hypoalbuminaemia is common; sterile pyuria and cerebrospinal fluid (CSF) pleocytosis (predominantly lymphocytes) representing aseptic meningitis also occur.2 ,6 ,9 ,48

Predicting IVIG resistance/high CAA risk

Several scoring systems have been developed to identify children at highest risk of IVIG resistance and, hence, highest risk of developing CAA (table 3).47 ,49 ,50 Kobayashi et al47 developed a model to predict unresponsiveness to IVIG in Japanese children with KD. This was used to define severe cases in a pivotal clinical trial of corticosteroids(see below), because IVIG resistance is known to be a strong risk factor for the development of CAA. The Kobayashi, Egami and Sano scores when tested in a USA study demonstrated comparably high specificity for predicting IVIG non-response in non-Japanese patients, but had low sensitivity.51 The clinical implication of this is that if the risk score is ‘positive’ in a non-Japanese patient (eg, ≥5 for the Kobayashi score) then IVIG resistance is likely; however, a negative score does not reliably exclude IVIG resistance. Attempts to develop a more sensitive and specific score for patients outside of Japan have thus far been unsuccessful.52

IVIG

Early recognition and treatment of KD with aspirin and IVIG has been shown unequivocally by randomised controlled trials and meta-analysis to reduce the occurrence of CAA.6 ,53 Two g/kg of IVIG is the optimal dose, usually given as a single infusion.6 ,53 Meta-analysis of randomised controlled trials comparing divided lower doses of IVIG (400 mg/kg/day for four consecutive days) versus a single infusion of high-dose IVIG (2 g/kg over 10 h) has shown that even though the 4-day regimen has some benefit, the single-dose regimen has a greater therapeutic effect in the prevention of CAA.54

Aspirin

Meta-analysis comparing anti-inflammatory doses of aspirin (30–50 mg/kg/day) with high-dose aspirin (80–120 mg/kg/day) combined with IVIG found no significant difference in the incidence of CAA between the groups.55 Currently, aspirin at a dose of 30–50 mg/kg/day is recommended during the acute phase of the illness, as this may be better tolerated than higher doses in terms of gastrointestinal and other side effects. The dose should be reduced to an antiplatelet dose of 3–5 mg/Kg once fever and inflammation have subsided.

Corticosteroids for the primary treatment of severe KD

IVIG resistance occurs in up to 20% of cases, and these patients are at increased risk of developing CAA unless they receive additional treatment.2 ,6 ,9 Recently, the findings of the published RAISE study that selected patients at high risk of IVIG non-response emphasised this point as treatment of the control arm with IVIG/ aspirin was still associated with a CAA complication rate of 23%.14

Corticosteroids are effective treatment for other forms of vasculitis, but early retrospective analyses suggested that corticosteroids were associated with increased risk of CAA.56 However, this almost certainly reflected selection bias as the sickest patients received steroids.

Clinical trials evaluating the use of corticosteroids plus IVIG have produced seemingly confusing results.10 ,11 ,14 ,57–59 Chen et al60 recently reported a meta-analysis comparing the frequency of CAA in patients treated with IVIG plus corticosteroids or IVIG alone for the primary treatment of KD. They followed standard guidelines for conduct of meta-analyses, and used defined criteria for trial quality assessment, including evaluation of criteria for diagnosis; study design; follow up and blinding.60 They identified nine clinical studies meeting their quality criteria, involving 1011 patients. Six out of nine were prospective RCTs; two were non-randomised controlled studies; and one was a retrospective report.60 Of the 1011 patients included, 536 received IVIG+corticosteroids and 475 IVIG alone.60 They found that significantly fewer patients receiving IVIG+corticosteroids developed CAA than those receiving IVIG alone (7.6% vs 18.9%; OR: 0.3; 95% CI 0.20 to 0.46).60 The benefit was found in several subgroup analyses including the six prospective RCTs, and studies using prednisolone or intravenous methylprednisolone.60 They found no significant differences in frequency of severe adverse events between the steroid and non-steroid treatment groups.60 Chen's meta-analysis provides convincing evidence that steroids combined with IVIG as initial treatment reduces overall risk of CAA in severe KD.60 However, neither the meta analysis nor the RAISE study provides clear answers as to whether all children in the UK should be treated with corticosteroids, and what dose, duration and route of corticosteroids should be used.

Heterogeneity in corticosteroid dosing in the published trials is an important consideration when translating the results of Chen's meta-analysis into clinical practice, as different corticosteroid regimens were used in each of the trials. This is illustrated when two of the methodologically strongest studies, the American Paediatric Heart Network study10 and the recently reported Japanese RAISE study14 are considered in more detail. The American study evaluated the use of intravenous methylprednisolone (30 mg/kg) given as a single dose in unselected patients with KD.10 By contrast, RAISE evaluated lower-dose (2 mg/kg) intravenous prednisolone given for 5 days; if fever settled, this was then converted to oral prednisolone which was subsequently tapered over 15 days only after the CRP normalised.14 Moreover, patients were included in RAISE only if they were at high risk of IVIG resistance based on the Kobayashi score.14 Perhaps unsurprisingly then, these two studies have produced conflicting results, with steroids conferring significant benefit in the Japanese RAISE study, but a lack of overall benefit in the American study.

Second, as seven of the nine studies in the meta-analysis were undertaken in Japan and only two in the USA, the results may not be applicable outside of Japan. There is some concern that KD in Japan may behave differently from the disease in non-Japanese populations, highlighted by the failure of the Kobayashi and other scores used to predict IVIG non-response in Japan to accurately predict non-response to IVIG in the USA.14 ,47 ,51 The meta-analysis may say more about the beneficial effect of steroids in Japanese patients meeting the criteria for predicted IVIG resistance than about an overall benefit in unselected patients.

Third, while the meta-analysis has not revealed any increase in severe adverse events associated with corticosteroid use, the total number of patients receiving steroids was only about 500.60 Most paediatricians who routinely use corticosteroids for the treatment of other systemic inflammatory diseases will probably be comfortable with a modest dosing regimen of prednisolone weaning over 3 or 4 weeks, if this truly will spare some children from potentially lifelong cardiac sequelae, although vigilance during follow-up for the presence of corticosteroid-related complications, such as hypertension, behavioural changes, secondary infection, hyperglycaemia and bone necrosis is required.

As 80% of patients with KD respond to aspirin and IVIG, and CAA are most commonly seen in those who fail to respond to IVIG, we urgently need a means of predicting IVIG resistance so that corticosteroids can be included in primary treatment of selected patients at risk of IVIG non-response. However, the Kobayashi score, used by Japanese investigators had low sensitivity (but good specificity, table 3) for the prediction of IVIG non-response in studies in the USA.14 ,51 Therefore, a low Kobayashi score does not reliably exclude IVIG resistance. With these caveats in mind, we propose a pragmatic treatment approach (below) based on the current data, while advocating further research to provide the evidence base, which is currently lacking.

Recommended indication for corticosteroids in KD

We suggest that corticosteroids should be considered for

1. Patients who have already declared themselves as IVIG-resistant, that is, with ongoing fever, and/or persistent inflammation or clinical signs ≥48 h after receiving IVIG as a single dose of 2 g/kg.

2. Patients with features of the most severe disease (and therefore the greatest likelihood of developing CAA). In the absence of validated risk scores outside of Japan, we suggest that such patients include the very young (<1-year-old); those with markers of severe inflammation, including: persistently elevated CRP despite IVIG, liver dysfunction, hypoalbuminaemia, and anaemia; and the small group who develop features of haemophagocytic lymphohistiocytosis (HLH)61 and/or shock.

3. Patients who already have evolving coronary and/or peripheral aneurysms with ongoing inflammation at presentation. It is increasingly recognised that echocardiographic studies performed in the first week of KD may already show vessel abnormality, including brightness (suggesting inflammation) or dilatation when compared with age-related normal ranges and/or extracoronary manifestations, including mitral regurgitation and pericardial effusion.62 Patients with these features may also be at greater risk of CAA and, therefore, may require corticosteroids.62

What corticosteroid regimen to use?

Chen's meta-analysis does not provide us with an evidence base for optimal corticosteroid dose/duration.60 From the studies included, an intravenous preparation equivalent to 2 mg/kg prednisolone for 5–7 days, or until CRP normalises, followed by oral prednisolone weaning over 2–3 weeks seems logical. However, in the absence of a robust evidence base, flexibility of the steroid regimen ultimately used for individual patients is recommended, and should ultimately be determined by treating clinicians: the updated guideline provides two suggested regimens (figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Recommended clinical guideline for the management of Kawasaki disease in the UK. Since risk scores for IVIG resistance perform sub-optimally in non-Japanese patients (Table 3), we cannot recommend their use to define high risk definitively; clinicians may, however, choose to consider the clinical and laboratory parameters listed to identify “high risk” patients. If the Kobayashi risk score is “positive” in a non-Japanese patient (eg,≥ 5) then IVIG resistance is likely; however a score <5 does not reliably exclude IVIG resistance. The aim of treatment is to switch off the inflammatory process that is damaging the coronary arteries as rapidly as possible. In the absence of a strong evidence base favouring a specific corticosteroid regimen, two suggested corticosteroid regimens for high-risk cases are provided for clinicians to choose from. For those on low dose aspirin, we also recommend avoiding the concomitant use of non-steroidal anti-inflammatory drugs (NSAIDs) as these interfere with the anti-platelet effect of low dose aspirin. *Treatment can be commenced before 5 days of fever if sepsis excluded; treatment should also be given if the presentation is > 10 days from fever onset if there are signs of persistent inflammation; **Kobayashi risk score ≥5 points ^aRefer to paediatric cardiologist; ¶ Other specific interventions such as positron emission tomography (PET) scanning, addition of calcium channel blocker therapy, and coronary angioplasty at discretion of paediatric cardiologist. + Other immunomodulators may include ciclosporin. ♥For infants, Z score for internal coronary artery diameter >7 based on Montreal normative data: http://parameterz.blogspot.co.uk/2010/11/montreal-coronary-artery-z-scores.html.

Should a second dose of IVIG be given to patients who fail to respond to the initial dose?

In patients who have shown some but not complete response, we suggest that a second dose of IVIG is given at the same time as commencing steroids if they have not already been commenced for signs of severe disease (see above). A second dose may not be beneficial if there was little response to the first dose. Vigilance during follow-up for the presence of corticosteroid-related complications, such as hypertension, secondary infection, hyperglycaemia and bone necrosis is required.

Role of antitumour necrosis factor-α

There are emerging animal data case reports suggesting a role for anti-tumour necrosis factor (TNF)-α therapy for the treatment of KD.63–66 Serum TNF-α is elevated in KD patients, and higher levels correlate with the development of CAA.67 The most commonly used agent is infliximab, a chimeric murine/human IgG1 monoclonal antibody specifically binding TNF-α.63–66 In one retrospective study of 17 children with IVIG-resistant KD, infliximab was used successfully with abrupt defervescence in 13/16 febrile patients, with no infusion reactions.12 Burns et al68 reported a phase 2 clinical trial including 16 patients receiving infliximab that demonstrated that this treatment was safe and well tolerated in patients resistant to IVIG. Response to therapy with cessation of fever occurred in 13 of 16 patients. CRP level was elevated in all but one patient before infliximab infusion, and the level was lower following infusion in all 10 patients in whom it was remeasured within 48 h of treatment.68 There were no infusion reactions to infliximab, and no complications attributed to infliximab administration in any of the patients.68 A more recent US retrospective review of IVIG-resistant patients treated with either IVIG (n=86) or infliximab (n=20) demonstrated that patients treated with infliximab had fewer days of fever and shorter hospitalisation, but with similar coronary artery outcomes.69

Etanercept (soluble TNF-α receptor) is an alternative TNF antagonist, and has been reported to be safe and well tolerated in 15 children with KD when given at a dose of 0.8 mg per kilogram weekly for three doses, and may be beneficial.70 A recent multicentre study in the USA comparing IVIG and aspirin to IVIG/aspirin plus infliximab as initial therapy in an unselected group of KD patients has been completed and showed no significant reduction in CAA, although the trial was underpowered for this endpoint. There was, however, faster resolution of the acute-phase response in the infliximab group. (Burns JC et al, personal communication and manuscript submitted). Thus, anti-TNFα should be considered in patients with IVIG-resistant KD, and further study is required of its role as first-line therapy.

Other therapies

Other immunosuppressive agents such as ciclosporin, cyclophosphamide, methotrexate, and plasma exchange, have occasionally been used to treat patients who do not respond to IVIG, steroids and anti-TNFα.2 ,6 Genome-wide association studies have identified polymorphisms of ITPKC, a negative regulator of T-cell activation, and other T cell signalling pathway genes associated with KD susceptibility, IVIG resistance, and increased risk of CAA in Asian and US children.36 ,40 This suggests that calcineurin inhibitors (such as ciclosporin) may be applicable for the treatment of KD.71 ,72 These initial reports need to be tempered by the knowledge that calcineurin inhibitors can also be toxic to the endothelium; for example, in Behçet's disease it is recommended that these are avoided in those with cerebral vasculitis since there is a concern that this class of drug may exacerbate vasculitic complications in some scenarios.6 ,73 The use of these agents cannot be recommended routinely but can be considered on a case-by-case basis after consultation with specialist units.

Management of KD in the convalescent phase

In the convalescent phase of the condition, if aneurysms persist, antiplatelet therapy in the form of low-dose aspirin (2–5 mg/kg) should be continued long-term until the aneurysms resolve.6 Clopidogrel is an alternative antiplatelet agent that could be considered.6 In the presence of giant aneurysms (>8 mm) warfarin is recommended in addition to aspirin.74 Heparin should be administered initially for at least 48 h and only stopped when warfarin has been commenced and the international normalised ratio (INR) is stable between two and three to avoid paradoxical thrombosis due to protein C and S depletion that may occur when warfarin treatment is started.9 If thrombosis does occur, thrombolytic therapy may be indicated, but expert advice must be sought.6 ,9 Some patients may require coronary angioplasty or a revascularisation procedure should ischaemic symptoms arise or evidence of obstruction occur.6 If formal catheter coronary arteriography is to be considered, if possible, this should be deferred for the first 6 months from the acute illness to avoid procedural-related myocardial infarction, of particular concern, while the coronary endothelium is still actively inflamed.6

Immunisation following KD

The recommendation regarding timing of immunisations after KD remains unchanged from our 2002 guideline.9 Immunisation with all vaccines should be deferred for at least 3 months following an episode of KD treated with IVIG, mainly due to the potential lack of effectiveness of live vaccines following IVIG75 and due to the potential for any vaccine to induce potentially detrimental immune activation during the convalescent phase of KD. The evidence for this latter recommendation is anecdotal. Thereafter, all vaccines should be administered as recommended by national schedules.

Patients who require long-term aspirin for persistent CAA should be considered for immunisation with varicella zoster virus (VZV) vaccine in view of the association of VZV and aspirin with Reye's syndrome.

Recommendations for long-term cardiac management

A detailed review of this area is beyond the scope of this article. A statement of the American Heart Association committee on rheumatic fever, endocarditis and KD published in 2004 provided detailed guidance on the stratification of KD patients according to their relative risk of myocardial ischaemia.6 Risk-level categories are summarised in table 4. This stratification allows for patient management to be individualised with respect to medical therapy to reduce the risk of thrombosis, physical activity, frequency of clinical follow-up and diagnostic testing, and indications for cardiac catheterisation and coronary, CT and MR angiography.6 Stress echocardiography when considered safe by cardiologists, should be performed on all patients with persistent structural abnormalities of the coronary arteries.6 If there is evidence of inducible ischaemia, then invasive angiography is indicated.6 For these patients, bi-annual follow-up and aggressive management of traditional cardiovascular risk factors is also recommended for all patients.6 CT coronary artery calcium scores are linked to mortality in atherosclerosis; it has been suggested that this imaging modality could have an important role in the late follow-up of KD.87 Adolescents and young adults may develop coronary calcification that is detectable only more than 10 years after the acute KD episode; while promising as a potential means of stratifying patients for a long-term follow-up, it is not yet known in KD exactly how coronary artery scores relate to late morbidity or mortality.

At the time of this writing, a long-term follow-up study of premature atherosclerosis/late KD vasculopathy is ongoing in the UK; it is anticipated that this study will help guide clinicians regarding long-term follow-up after KD.