Kawasaki disease (KD), a vasculitis affecting small- and medium- sized blood vessels, is the second most common childhood vasculitis, and the most common cause of pediatric acquired heart disease in the developed world.¹ It was first described in 1967 and initially called mucocutaneous lymph node disease because of the predominant clinical features. Affected children present acutely with fever, bilateral nonexudative conjunctivitis, erythema of the lips and oral mucosa, cervical lymphadenopathy, rash, and extremity changes. These findings are usually self-limited, but 15% to 25% of untreated children develop coronary artery abnormalities, putting them at risk for developing ischemic heart disease or sudden death later in life.² Therapy with intravenous immunoglobulin in the acute phase is aimed at reducing inflammation, thereby minimizing the risk of coronary artery aneurysms and other late sequelae.

Children under 5 years of age are affected in 90% of cases; the median age is 2 years. There is a slight male preponderance. Children of all ethnicities may be affected, although the incidence is higher in the Asian population. The annual incidence of KD in children under 5 years of age in the United States and the United Kingdom is 20 and 8 per 100,000 children, respectively. The rates in Japan and Taiwan are approximately 5 times higher.³ Adults also may develop KD but far less often; fewer than 100 cases have been reported worldwide to date.

The pathogenesis of KD remains unknown, though most hypotheses center on an infectious trigger. Several epidemiologic features of the disease support this theory. For example, the incidence of KD peaks in winter and springtime, corresponding to peak seasons for infectious diseases. Furthermore, KD exhibits “epidemics,” as demonstrated by years with up to a tenfold spike in disease incidence in Japan in the 1980s and 1990s. Nonetheless, extensive testing using culture, serologic, and sequencing techniques has not implicated a specific transmissible agent as the cause. Additionally, anti-inflammatory medications such as intravenous immunoglobulins (IVIG) and TNF-inhibitors are clinically beneficial, while antibiotics are not. Taken together, these facts suggest that KD represents an exaggerated inflammatory response to one or more infectious agents in genetically susceptible individuals. Alternative hypotheses include immune activation by a superantigen, analogous to staphylococcal or streptococcal toxin-mediated illnesses,⁴ or a heretofore unrecognized organism as the causative agent. Ultrastructural, immunofluorescence, and RNA evidence support the hypothesis of a “new” virus associated with Kawasaki disease.

More recently, several genome-wide association studies have identified a variety of candidate genes and polymorphisms involved in inflammation and cardiovascular pathology showing preferential expression in children with KD.^5-7 Studies to determine the significance of these findings are ongoing.

The diagnosis of KD is a clinical one, with epidemiologic criteria as defined by Dr. Tomisaku Kawasaki in 1967 as follows:

Fever that persists for at least 5 days plus at least four of the following:

1. 
   

   Bilateral nonsuppurative conjunctival injection

   
   
2. 
   

   Changes of the lips and oral cavity

   
   
3. 
   

   Cervical lymphadenopathy

   
   
4. 
   

   Polymorphous exanthema

   
   
5. 
   

   Changes in the peripheral extremities

These features may not occur in all KD patients, and their temporal sequence is also very variable. KD tends to occur in three phases: the initial acute febrile phases lasts up to 14 days, followed by a subacute phase of 2 to 4 weeks, with a convalescent phase that may extend to months or years.

ACUTE FEBRILE PHASE

The typical patient demonstrates a persistent high fever which responds minimally to antipyretics. The child can be very irritable during this stage, often out of proportion to the height of the fever. Without treatment, the fever lasts an average of 12 days, with the remaining mucocutaneous features evolving over the next 1 to 2 weeks. The sequence in which they appear varies among patients, highlighting the importance of repeated evaluation in young children with persistent unexplained fevers.

Conjunctival injection is seen 85% of children and is bilateral, painless, nonsuppurative, and tends to spare the limbal areas. Other less common ocular manifestations include anterior uveitis, keratitis, papilledema, vitreous opacities, and subconjunctival hemorrhages, often with accompanying photophobia.

Changes to the lips and oral cavity are seen in 90% of children. These manifest as diffuse erythema of the oropharynx, dry, cracked, and sometimes bleeding lips, and a strawberry tongue. Mucosal ulcerations and pharyngeal exudates are not typical of KD.

Cervical lymphadenopathy affects about 50% of children with KD. The enlarged lymph node is typically unilateral, larger than 1.5 cm in diameter, limited to the anterior chain, firm and nontender, and without significant overlying erythema. Patients with KD have been misdiagnosed with having infectious lymphadenopathy because of the prominence of the enlarged lymph node. Imaging with ultrasound or CT scan will reveal multiple enlarged nodes without suppuration, giving the appearance of a “bunch of grapes” similar to the lymphadenopathy seen in Epstein-Barr virus infections. Biopsies, though generally not necessary, reveal focal necrosis with an inflammatory infiltrate consisting of neutrophils and macrophages and fibrin thrombi within the microvasculature.

A rash is evident in 70% to 90% of affected patients and is polymorphous and nonspecific. A macular, targetoid, or morbilliform exanthem that is usually nonpruritic is most common. The rash may occasionally be urticarial, erythrodermic, or pustular. In most cases, the rash is widespread, involving the trunk and limbs. In up to half of cases, it begins in the perineal region, typically desquamating after 2 to 3 days. Children with prior bacille Calmette-Guérin (BCG) vaccination tend to have prominent erythema and induration over the BCG scar.

Changes in the extremities in the first week of KD appear as redness over the palms and soles, and/or swelling of the dorsum of the hands and feet. Later, periungal desquamation is seen, usually by the third or fourth week.

Children with KD may have other clinical features, including anorexia, arthralgias/arthritis (initially involving large joints, then later, small joints), diarrhea, vomiting, abdominal pain, tachycardia out of proportion to the fever, S3 gallop, and muffled heart sounds. These may assist in distinguishing KD from other childhood febrile exanthema, although they are not included among the diagnostic criteria. Table 147-1 highlights the range of clinical findings in KD.

SUBACUTE PHASE

This phase begins as the fever resolves, typically by the end of the second week in patients not treated with IVIG. The affected child demonstrates periungual desquamation, beginning just under the nails and progressing to sheets of skin sloughing from the digits and at times from the palms and soles. Joint inflammation, typically a self-limited polyarthritis involving the small joints, may also develop during this period. Apart from these, the child treated with immunoglobulin tends to be asymptomatic. An untreated child may still have the mucocutaneous features, although these usually resolve spontaneously as well. At this stage, most coronary artery abnormalities that will develop are evident on echocardiography, though they may continue to expand for weeks due to both hemodynamic factors and ongoing inflammation, especially in untreated cases. Laboratory tests begin to return toward normal, particularly the white blood cell count and C-reactive protein (CRP). Other aberrations, however, including lymphocytosis, thrombocytosis and elevated erythrocyte sedimentation rate (ESR) may persist for up to 2 months.

CONVALESCENT PHASE

Children tend to be asymptomatic during the convalescent phase, with normalization of all inflammatory markers and stabilization of coronary artery dilatation by about 2 months after the onset of the disease. Beau’s lines—deep transverse grooves across the nails reflecting a non-specific response to significant inflammation—may appear at this time.

INCOMPLETE OR ATYPICAL KAWASAKI DISEASE

The terms incomplete or atypical Kawasaki disease refer to children who only have some of the clinical features of KD and do not meet diagnostic criteria for the disease. Nonetheless, these children may develop coronary artery abnormalities, particularly infants under 6 months of age who are less likely to demonstrate the full spectrum of mucocutaneous manifestations. Because the diagnosis is not necessarily apparent, physicians must remain vigilant and should consider KD in children with unexplained fever for more than 5 days. An expert panel has developed guidelines for evaluating such children (Figure 147-1).⁸ Children under 6 months with persistent fever for 7 days and evidence of inflammation should have an echocardiogram to screen for atypical KD even in the absence of clinical criteria. Older children with three of the principal clinical features and supportive laboratory findings should be similarly evaluated. Findings such as perivascular brightness and ectasia or lack of tapering of the coronary arteries may represent early signs of vasculitis, though they are also reported in viral syndromes and systemic onset juvenile idiopathic arthritis. Decreased left ventricular contractility, mild valvular regurgitation (most commonly mitral regurgitation), and pericardial effusion also support a diagnosis of incomplete KD. Such patients should be treated no differently than those who fulfill traditional diagnostic criteria.