Kawasaki Disease

Ralph D. Feigin

Frank Cecchin

Scott D. Wissman

Kawasaki disease is an acute, febrile, multisystem syndrome of unknown etiology that predominantly afflicts children younger than 5 years. The disease also is termed mucocutaneous lymph node syndrome. The diagnosis is based entirely on clinical features because no pathognomonic laboratory findings are extant.

The disease was recognized first by Tomisaku Kawasaki in 1967. Subsequently, numerous cases of the disease have been recognized throughout the world in all racial groups. Kawasaki disease is one of the most common causes of inflammatory arthritis and now is the leading cause of acquired heart disease in children in North America and Japan.

EPIDEMIOLOGY

Kawasaki disease affects children worldwide, primarily patients of Asian descent. Japan still has the highest incidence rate of Kawasaki disease of any country or population group at 108 to 111 cases per 100,000 children younger than 5 years of age, followed by persons of Japanese descent not living in Japan, who have a rate of 40 to 50 cases per 100,000 children younger than 5 years of age. The lastest published data from Japan show that the incidence rate in Japan has been increasing steadily during the last 11 years. The rate in 1998 was more than 1.5 times higher than in 1987.

Cases reported to the Centers for Disease Control and Prevention (CDC) indicate that the average yearly incidence among children 5 years old or younger in the United States is estimated to be approximately 12 to 15 cases per 100,000. Recent studies suggest that the incidence of Kawasaki disease has not changed markedly in the United States during the past decade. The incidence among Asian American children is three times higher than that in African American children and more than six times higher than that in white and Hispanic children. The incidence in the United States also varies by region, with the highest numbers of cases reported in the the Northeast and the West.

Kawasaki disease has been seen almost exclusively in children, and the male-to-female ratio is 1.5:1.0 in virtually all countries. Almost 80% of all cases occur in children younger than 5 years old, and 90% of all cases involve children younger than 8 years old. The median age of children diagnosed with Kawasaki disease in the United States is 2.3 years. Only 1.7% of cases occur in very young infants 90 days or younger, and fewer than 10 cases have been reported in the neonatal period, the youngest being less than 2 weeks old. This incidence pattern, which has been noted with other infectious diseases, suggests that transplacental antibody may offer some protection in young infants and that when maternal immunoglobulin G (IgG) concentrations begin to decline and infants no longer are immune, a cohort of susceptible children is produced. Because individuals have either clinical disease or acquired immunity to an as yet unknown agent, the incidence may decline toward zero. The foregoing incidence pattern, however, cannot be accepted as proof of an infectious cause for Kawasaki disease.

Fewer than 10% of cases are in children 8 years old and older, but this age group has a higher prevalence of abnormal echocardiograms. The diagnosis generally is made much later in the course of the illness in these patients, and, hence, treatment is delayed; this may account for the increased frequency of cardiac abnormalities. Rarely are cases reported in the adult population. In several adult cases, the reported illness more likely was caused by toxic shock syndrome than by Kawasaki disease.

Kawasaki disease is seen at all seasons of the year. However, in the United States and Japan, a slight increase in the number of cases occurs in the winter and spring months. An association also has been noted between Kawasaki disease and residence within 180 m (200 yards) of a body of water as well as history of humidifiers in the home. Some studies have found a greater proportion of patients from families with higher socioeconomic backgrounds, and one study suggested that a greater number of cases occurred in children whose parents work in the health profession.

Clustering of cases of Kawasaki disease has been observed. Often, families of children with the disease have had contact with other children who had Kawasaki disease. Outbreaks have been reported in several cities in the United States, Australia, and Japan. One outbreak in Japan started in Tokyo and spread northward and southward to involve the entire country within 6 months. An increased incidence of second cases occurs among siblings of children with Kawasaki disease. During epidemics of the disease in Japan, the rate of second cases among siblings was 10 to 30 times greater than the incidence in the general population.

In several outbreaks, children with Kawasaki disease had a higher incidence of antecedent illness—primarily of respiratory origin—than did control patients. A report of two cases of Kawasaki disease strongly suggested person-to-person transmission between first cousins, with a latent period of 16 to 18 days. Secondary or coprimary cases are rare, and nosocomial infection has not been reported. The recurrence rate of the disease is about 4%.

ETIOLOGY

Arguments that Favor an Infectious Etiology

The cause of Kawasaki disease is still unknown. Most investigators have favored the possibility of an infectious agent or
an immune response to an infectious agent or agents. This hypothesis has been supported by the epidemiologic patterns of the disease and the clinical appearance of oropharyngeal inflammation and cervical adenitis, which is consistent with the acquisition of a replicating agent by droplet transmission, of inflammation of the respiratory tract mucosa, of toxic appearance of children with fever, and of involvement of other organ systems. Laboratory features, which may include an elevated white blood cell count with a left shift, elevated levels of acute-phase reactants, and pyuria, also suggest an infectious origin.

Attempts to incriminate specific infectious agents have failed, and all attempts to culture bacteria or viruses have been unsuccessful. No culture or serologic evidence of infection with Lancefield group A streptococci or staphylococci exists with respect to the genesis of Kawasaki disease. Several investigators have proposed that a variant strain of Propionibacterium acnes may play a causative role in Kawasaki disease and that house dust mites may play a role as vectors. A causative link has not been established, however, despite multiple investigations. Single-case reports have documented recovery of various bacterial agents, among them Yersinia pseudotuberculosis, Salmonella, and Pseudomonas aeruginosa.

Several investigators have provided evidence of an agent similar to Rickettsia, but others have been unable to substantiate these findings. However, Kawasaki disease does not respond to treatment with antibiotics known to be effective against rickettsial agents, and most known rickettsial diseases are vector borne and seasonal. Mycoplasma pneumoniae also has been proposed as a causatice agent as has Chlamydia pneumoniae, but studies remain inconclusive.

The immunologic and clinical manifestations of Kawasaki disease bear remarkable similarity to diseases associated with superantigen production. The classic example is toxic shock syndrome, in which the staphylococcal enterotoxin functions as a superantigen that induces massive expansion of T cells expressing a specific Vbeta region on the T-cell receptor. This event, in turn, leads to excess cytokine production, causing clinical illness. Patients with acute Kawasaki disease have been shown to have selective expansion of T cells expressing T-cell–receptor variable regions Vbeta2 and Vbeta8, typically detected during the second week of illness. Further research to elucidate the source of the superantigen is ongoing.

Although many viruses have been implicated, an abnormal immune response to Epstein-Barr virus (EBV) is postulated. One study of patients with Kawasaki disease in Hawaii, however, showed no association between patients with EBV infection and control patients. In fact, none of the herpesviruses, including EBV, cytomegalovirus, human herpesvirus 6, varicella-zoster virus, and herpes simplex virus types 1 and 2, display a dominant role in the pathogenesis of Kawasaki disease in Hawaii. Patients studied during two outbreaks in Japan did have increased antibodies to adenovirus type 2, but no supportive data regarding a causative role exist. Some studies have suggested unusual immune responses to rubeola, rubella, and parainfluenza viruses.

Investigators have detected RNA-dependent DNA polymerase (reverse transcriptase) activity in cultured peripheral blood mononuclear cells from patients with Kawasaki disease. One study demonstrated that cultures taken between the third and ninth weeks after the onset of fever are the most likely to be associated with reverse transcriptase activity. In the early convalescent phase of Kawasaki disease, the cell can be detected most easily in older patients who mount a marked humoral immune response. However, all serologic tests for human immunodeficiency virus type 1 and human T-cell leukemia-lymphoma viruses types I and II have been negative. Other studies also rule out any retroviral cause.

Although frequently hypothesized, Kawasaki disease has not been associated consistently with exposure to environmental pesticides, chemicals, heavy metals, toxins, or pollutants. Usually, poisoning with environmental agents does not simulate an acute infectious disease, although similarities between acrodynia (mercury poisoning) and Kawasaki disease have been noted. Children with Kawasaki disease have had normal mercury levels, with the exception of six patients from the Great Lakes area whose urinary excretion of mercury was increased.

An outbreak of Kawasaki disease in Denver, Colorado was considered to be associated with the use of rug shampoo. Eleven of 23 patients with the syndrome had been exposed to rug shampoo in the 30 days before the onset of illness. Six case-control studies have been completed in an attempt to delineate the association between Kawasaki disease and rug shampooing. It was hypothesized that anionic detergents could trigger a hypersensitive response, or a causitive agent, either infectious or allergenic, was aerosolized during the cleaning process. Three of the studies demonstrated a significant association, whereas the other three did not.

Another suggestion is that Kawasaki disease may be an allergic phenomenon. Several studies suggest that the incidence of allergies in children with Kawasaki disease or in members of their families is higher than in control patients. The prevalence of atopic dermatitis in children with Kawasaki disease is nine times greater than in age-matched control children. In addition, numerous children with Kawasaki disease show a twofold to fourfold elevation in total serum IgE levels during the acute phase of the illness, followed by a decline to the normal range in the ensuing 1 to 2 months. Peak IgE levels do not correlate with the severity of disease or the incidence of arthritis and carditis. The relationship of IgE to the pathogenesis of Kawasaki disease remains unclear.

Relationship of Kawasaki Disease to Infantile Periarteritis Nodosa

A pathologic similarity between infantile periarteritis nodosa and fatal infantile Kawasaki disease appears to exist. Discussions of the similarities between these disorders have been published by numerous investigators. The two diseases cannot be shown to be identical because their causes are entirely unknown, and experience with gross and histologic investigation is relatively embryonic. Distinguishing infantile periarteritis nodosa with coronary artery involvement from fatal infantile Kawasaki disease pathologically is impossible. Clinically, most patients with infantile periarteritis nodosa associated with coronary artery involvement do not meet the other criteria established by the CDC for Kawasaki disease. However, when pathologic and clinical criteria are combined, the two diseases appear to be indistinguishable, thus raising a question about the novelty of Kawasaki disease. This issue is true particularly in the United States, where infantile periarteritis nodosa has been documented since the 1940s, whereas Kawasaki disease was not recognized as a clinical entity until 1974. A male-to-female ratio of 3:1 exists in patients with periarteritis nodosa. Whether any of the cases of periarteritis nodosa with coronary artery aneurysms were examples of Kawasaki disease is speculative because histories in the past often were scant, and deaths often were attributed, perhaps erroneously, to other disorders, such as scarlet fever.

PATHOLOGY

Grossly, cardiac hypertrophy is common. Multiple single beadlike or fusiform aneurysms of the coronary arteries and their branches usually are found in fatal cases. During the various
clinical stages, specific pathologic findings are noted, and during days 0 to 9 of the illness, the coronary arteries have perivasculitis and endarteritis but medial sparing. Pericarditis, myocarditis, endocarditis, valvulitis, and conduction system inflammation are observed, with polymorphonuclear infiltrates. During days 12 to 25, coronary artery panvasculitis and aneurysm formation occur, with inflammation and necrosis of the media resulting in “true” aneurysms. By the second week, the inflammatory infiltrate has evolved into lymphocytic and plasma cell dominance.

Resolution of the coronary inflammation occurs near day 30, with subsequent granulation formation. Coronary artery scarring, stenosis, and endocardial fibroelastosis are described after day 40. Aneurysms of other arteries, such as the renal, iliac, and brachial arteries, may be found. Phlebitis is common, with vascular inflammation that most often and most severely affects larger musculoelastic arteries in their extraparenchymal portions. Sites of arteritis include the lung, pancreas, spleen, kidney, testis, mesentery, adrenal gland, and gastrointestinal tract.

Only gold members can continue reading. Log In or Register to continue