Kawasaki disease (KD) is an acute, sometimes fatal vasculitis of childhood. 1 Although KD has replaced acute rheumatic fever as the most common cause of acquired heart disease in children in the United States and Japan, the pathogenesis of the illness remains uncertain. Many clinical features of KD are consistent with an infectious etiology, including fever, enanthem, exanthem, conjunctival injection, cervical adenopathy, and the self-limited nature of the illness. Also strongly supportive of an infectious cause are the epidemiologic features of KD, including well-defined epidemics with winter-spring predominance in temperate climates, and a geographic wave-like spread of epidemics. The peak incidence in early childhood and the virtual absence of the illness in adults suggest that the causative agent (s) is a ubiquitous microbe that causes an asymptomatic infection in most individuals with acquired immunity by adulthood. The rarity of the illness in infants less than 3 months of age may result from passive protection by antibody. The higher attack rate of KD in Asian populations suggests that genetic host factors may play a role in the ability of the microbe to cause symptomatic disease. Because most ubiquitous microbes enter the host via the respiratory or gastrointestinal tracts, one or both of these portals of entry would be likely for the putative KD agent (s).

Despite intensive investigation, the etiologic agent of KD remains unknown. Many viral and bacterial agents have been proposed as causative agents, but none has been convincingly related to the illness. The most recently proposed etiologic candidate has been a superantigen such as toxic shock syndrome toxin-1 (TSST-1) or another bacterial toxin. 2 The idea that a superantigen may be involved in KD was based on reports indicating that certain families of T cell receptor β genes (Vβ2 and Vβ8) were preferentially expressed in the peripheral blood lymphocytes of patients with acute KD. 3, 4 By binding to all members of a receptor family outside of the conventional antigen-binding site, superantigens can expand or delete one or more families of T cells with diverse antigen-binding sites. Individual superantigens have characteristic profiles of T cell family expansion or deletion. Thus, expansion of certain T cell receptor genes in KD suggests superantigen stimulation. However, other investigators have been unable to confirm preferential T cell receptor Vβ usage in acute KD. 5, 6 A relationship between KD and TSST-1 was proposed following a study performed in a single Boston hospital in which KD patients were more often colonized with TSST-1 producing strains of Staphylococcus aureus than were controls. 2 However, both earlier and subsequent studies have not shown a relationship between KD and staphylococci or streptococci. 7-10 Case reports indicating the isolation of streptococci or staphylococci from mucosal sites of children with KD 11 do not necessarily indicate a relationship between these bacteria and KD, because these organisms also are commonly recovered from healthy children and those hospitalized with other illnesses.