Introduction

Kawasaki disease (KD) is an acute, self-limiting vasculitic illness predominantly affecting infants and young children. It is now a leading cause of acquired heart disease in children in Western countries. The diagnosis is made clinically, and effective treatment is available to reduce the likelihood of potentially fatal coronary vasculitis. KD was first described in 1967 as ‘mucocutaneous lymph node syndrome’ in a series of 50 Japanese children.¹ Although it is most common in Japanese and Korean children (annual incidence 90–150/100 000 children younger than five years), it occurs in all ethnic groups, with an annual incidence in the United States of approximately 10/100 000 children younger than 5 years old. The majority of cases (85%) occur in children aged less than 5 years of age. It is 1.5 times more common in boys than girls.²

Many features of KD suggest an infectious aetiology. These include: seasonal variation (peak in winter/spring); occasional outbreaks; 10-fold higher risk in siblings than in the general population; rarity in young infants (suggesting protection from maternally acquired antibody); low recurrence rate (4%, which suggests acquired immunity); as well as the resemblance of the clinical presentation to other self-limiting infectious diseases, such as measles, adenovirus infection and staphylococcal toxic shock syndrome. However, despite much investigation of a variety of viral and bacterial pathogens, there is as yet no good evidence to implicate any known organism in KD. The ethnic variation suggests a genetic predisposition.

There is debate as to whether the inflammatory response in KD is initiated by a conventional antigen or a superantigen, and there are some immunological features to support both hypotheses. Unreplicated reports of increased expression of specific T-cell receptor V β-regions suggests toxin activation, whereas infiltration of paratracheal and vascular tissue with reactive clonal IgA plasma cells suggests entry of a conventional antigen via the respiratory route.³

Pathophysiology

The pathophysiology of KD involves vasculitis of medium-sized vessels including coronary, renal, hepatic and splanchnic arteries, beginning in both adventitial and intimal surfaces and proceeding toward the media. Coronary changes occur in approximately 20% of untreated patients.⁴ Immune activation involving cytokines and growth factors leads to inflammation and aneurysm formation, with the risk of thrombosis. The process evolves for a long period after the acute illness. In the majority of patients with echocardiographically demonstrable coronary artery lesions, the vessels remodel and have a normal appearance within a year or so. However, there is evidence of subtle long-term changes in coronary artery function, the clinical significance of which remain unclear.⁵ The risk of early adult coronary artery disease in these patients is unknown.

A diffuse inflammatory process of a variety of tissues has been found in autopsy specimens including lymph nodes, liver and gall-bladder.² Endothelial changes are prominent, with hyperplasia, necrosis and thrombosis. Myocardial abnormalities include hypertrophy of myocytes and fibrosis.

Clinical features

KD should be considered in the differential diagnosis of all infants and young children with a fever, rash and red eyes, as well as those with a prolonged fever without an alternative explanation. The diagnostic criteria are outlined in Table 5.8.1.

The diagnosis can be made earlier than day 5 if other features are present. This is important, as there is evidence that earlier administration of intravenous immunoglobulin (IVIG) is associated with a shorter illness and reduced risk of coronary disease.⁶

The fever is often associated with irritability, which responds poorly to paracetamol. Over 90% of patients develop a rash, which can take any of a number of forms – scarlatiniform, morbilliform or maculopapular. It often has the appearance of a viral eruption or allergic reaction, but can appear urticarial. Petechiae are sometimes seen, as are micropustules. The red eyes demonstrate characteristic sparing of the avascular perilimbic region, giving the appearance of a halo around the cornea. There is not usually an exudate. The changes may resolve rapidly, such that this important feature may be present only on history rather than being present when the child is examined. The cervical nodes are commonly unilateral and in the anterior chain. Generalised lymphadenopathy and splenomegaly are absent. The subacute phase occurs from 2–4 weeks, with resolution of fever. Desquamation (periungual) of fingers and toes may occur during this time.

KD is a multisystem disease with many and varied clinical manifestations. In addition to those in the diagnostic criteria, common features include marked irritability, diarrhoea, cough, arthralgia/arthritis, urethritis with sterile pyuria, otitis media, mild hepatic dysfunction, hydrops of the gallbladder and aseptic meningitis. Diagnosis is more likely to be delayed in older children, in whom less classical manifestations such as gastrointestinal and joint symptoms often predominate.⁷

Incomplete KD

Cases of incomplete or ‘atypical KD’ are being increasingly recognised, in which full criteria are not met, but the patient has coronary artery abnormalities. This is more common in young infants.⁸ The incidence of coronary artery aneurysms is at least as high in incomplete KD as in classical cases. Given that infants under 6 months appear to be at increased risk of developing coronary abnormalities;⁸ a lower threshold for treatment is probably indicated in this group. KD should be considered in all children with unexplained fever for 5 days and at least two major features of KD, and any infant with unexplained fever for over a week.⁴

Differential diagnosis

Depending on which features are present, a broad differential diagnosis needs to be considered in cases of possible KD. These include viral infections (e.g. measles, adenovirus, enterovirus), streptococcal and staphylococcal toxic-mediated illness (scarlet fever, toxic shock syndrome), drug allergy (including Stevens–Johnson syndrome), inflammatory bowel disease, connective tissue disease, particularly Still’s disease, and malignancy.

Complications

The major concern with KD is coronary artery disease. Younger children, especially under 12 months of age, are at highest risk. Changes seen include dilatation (ectasia) and discrete aneurysms. Aneurysms are classified according to the internal diameter into small (<3 mm), medium (3–6 mm), large (6–8 mm) and giant (>8 mm). Although aneurysms rarely form in the first 10 days of KD, echocardiographic signs of coronary arteritis may be seen, including perivascular brightness, ectasia and lack of tapering.⁴ Other findings may include decreased left ventricular contractility, mitral regurgitation, and pericardial effusion.

Coronary artery occlusion is most likely to occur in giant aneurysms, through a combination of sluggish blood flow and fibrotic stenosis at proximal and/or distal ends of the aneurysm. Thrombosis of coronary aneurysms can cause myocardial infarction and sudden death. Mortality is increased over the background rate within the first two months. Rarely, myocarditis can occur, causing congestive cardiac failure or arrhythmias. Thrombosis of peripheral arteries can lead to ischaemia and gangrene.

Investigations

Any child who presents with possible features of KD should be discussed with a paediatrician.

There is no definitive diagnostic test for KD. Investigation is directed toward excluding alternative diagnoses, as well as gathering supporting evidence for the diagnosis of KD. At least one set of blood cultures should be taken, and cultures of urine, cerebrospinal fluid and other sites may be indicated. Serological testing for Group A Streptococcus, and for specific viruses such as measles may be helpful. In KD, the peripheral white cell count and inflammatory markers are generally significantly raised. The erythrocyte sedimentation rate may continue to rise as the child improves clinically. A marked thrombocytosis is commonly seen in the second to third week of the illness. Pyuria, due to sterile urethritis, is often found on a voided sample (not on suprapubic aspirate (SPA) or catheter). Serum gamma glutamyl transferase is commonly elevated. Exclusion of markers for connective tissue disorders may be useful.

Echocardiography should be undertaken as soon as the diagnosis of KD is suspected. The American Heart Association guidelines suggest repeat echocardiograms at 2 weeks (time of maximal aneurysm formation), 6 weeks and 12 months (looking for late sequelae). If abnormalities are demonstrated on echocardiography, closer cardiological follow up may be required, depending on the lesion. A 12-lead ECG is most often normal but may show dysrhythmias, change in PR or QT intervals or non-specific ST changes.

Treatment

Treatment of KD is directed towards reducing the inflammation as rapidly as possible. The clinical markers of response to therapy are the temperature and patient’s general well-being, supported by the white cell count and CRP.

IVIG therapy has been demonstrated to induce resolution of fever as well as significantly reduce the risk of coronary artery abnormalities (from around 20% to around 3–5%) if given within the first 10 days of the illness.⁹ The precise mechanism of action of IVIG is unknown. Theories include saturation blockade of Fc receptors, direct antibody activity against bacterial superantigen, an unidentified causative pathogen or toxin, modulation of cytokine production or down-regulation of antibody synthesis. The optimal dose is 2 g kg^–1 day^–1 as a single infusion over 10–12 hours. There is some evidence that IVIG is also effective if given beyond 10 days;¹⁰ however, treatment as early as possible is optimal.⁶ IVIG should be given to patients with KD after the 10th day if they have persistent fever or other evidence of ongoing systemic inflammation.⁴ The effect of IVIG in patients who have already developed coronary artery aneurysms is unknown, though there may be some benefit.⁹ Different brands of IVIG may vary in their clinical effects, due to variation in sterilisation and other manufacturing procedures. As passive antibody acquisition may interfere with immunogenicity, live vaccine administration (e.g. measles, varicella) should be postponed by 3 months in children who have been given IVIG.

In the 1980s, high-dose aspirin was shown to decrease the incidence of coronary artery involvement in KD.¹¹ The current role of aspirin in KD is difficult to determine as it has been used in combination with IVIG in the major trials. Many centres have used high-dose aspirin (80–100 mg kg^–1 day^–1 in 3–4 divided doses, for anti-inflammatory effect) initially, and then switched to low-dose (3–5 mg kg^–1 day^–1, antiplatelet effect) after the patient’s fever resolves. However, in patients treated with IVIG, concomitant use of high-dose aspirin initially does not appear to result in shorter duration of fever or hospitalisation than low-dose.¹² Furthermore, the incidence of coronary artery aneurysm appears to be independent of aspirin dose.¹³ Therefore low-dose aspirin seems to be sufficient for initial treatment. Low-dose aspirin is continued for 6–8 weeks, and then stopped if there is no coronary involvement.

The role of corticosteroids in KD is a subject of continuing research. Studies of the addition of a single pulsed dose of steroids to IVIG as initial therapy have yielded conflicting results.^14,15 It is possible that a subset of patients at highest risk of developing coronary artery aneurysms may benefit; however, risk scores for stratification of patients have not been validated.

Refractory KD

Up to 15% of patients with KD treated with IVIG and aspirin have persistence or early recrudescence of fever, indicative of an ongoing vasculitic process.¹⁶ This is a strong risk factor for the development of coronary artery aneurysms.¹⁷ Children with ongoing or recurring fever beyond 36 hours after treatment with IVIG should be given a second dose of IVIG.⁴

Pulses of methylprednisolone (30 mg kg^–1 daily for 1-3 days) have been used in cases unresponsive to two doses of IVIG.¹⁸ Corticosteroids have been shown to reduce fever in patients with resistant KD; however, their effect on coronary artery abnormalities is uncertain.⁴ A number of other treatments including plasma exchange, cyclophosphamide, and tumour necrosis factor-α antagonists have been reported in children with refractory KD. The role of these therapies remains unclear.⁴

Prognosis

Mortality is less than 1%, being highest in those less than 12 months old. Recurrence is most likely to occur in children aged less than 3 years who had cardiac involvement initially, and usually within 12 months of the initial episode.¹⁹ Patients with recurrent KD appear to be at increased risk for cardiac sequelae.

Children without demonstrable cardiac disease appear to have an excellent prognosis, with long-term follow-up studies demonstrating absence of clinical sequelae for up to 21 years.²⁰ It is possible, however, that individuals who have had KD are at risk of early atherosclerotic heart disease.

The prognosis for children who have had coronary artery aneurysms is less clear. Most small to medium-sized aneurysms resolve echocardiographically;²¹ but healing involves fibrosis and calcification, with associated loss of vascular distensibility and reactivity. A proportion of coronary artery aneurysms progress to stenosis over time. Therefore children with KD who have had coronary artery aneurysms should have indefinite cardiology follow up. These patients will generally be treated with long-term antithrombotic therapy to prevent myocardial ischaemia. Children receiving long-term aspirin therapy should receive the influenza vaccine annually, to prevent Reye syndrome. New antiplatelet agents are under investigation.²² There have been reports of the use of thrombolytic agents in patients with KD, with thrombus seen within coronary aneurysms.²³ Anticoagulation with warfarin is required for patients with ‘giant’ (≥8 mm) or multiple aneurysms, and surgical intervention (e.g. angioplasty, bypass grafting) is occasionally necessary.

Controversies

The immunological mechanism of the inflammatory response in KD is the subject of continuing research – there is some evidence to support a conventional antigen, but other evidence suggests a superantigen.

The optimal dose of aspirin in the acute febrile phase is uncertain.

The role of steroids in KD is debated.

The clinical significance of subtle long-term changes in coronary artery function in some patients who have had KD remains unknown.