Eco-Friendly Alternatives to Toxic Household Cleaners

I. Introduction

Kawasaki disease (KD) is an acute, febrile systemic vasculitis of an unknown cause. KD was first described in Japan in 1967 by Dr. Tomasaku Kawasaki¹ and since then has been observed in children across the world of every racial group.² KD is now the leading cause of acquired heart disease in children.³ There have been great advances in the treatment of KD with the employment of intravenous immunoglobulins (IVIG) and high-dose aspirin (ASA);⁴ however, the lack of a diagnostic test can result in significant delays in treatment, often resulting in serious cardiac complications, such as coronary artery aneurysms, which occur in 20-25% of untreated children.⁵ Although there has been a great deal of research into the cause of KD, there is no known etiology. This review is intended to provide a background of the history of emergence, epidemiology, clinical manifestations, and pathogenesis of KD, and to concentrate on the past and present focuses of the search for a possible causative agent, including carpet cleaners, house dust mites, and potential infectious agents.

II. History of Kawasaki Disease

In 1967, Kawasaki published his first report in Japan describing the cluster of symptoms he observed in 50 patients seen since 1961, which he presented as an emergent disease called "mucocutaneous ocular syndrome" (MCOS).¹ Kawasaki's claim that MCOS was indeed a novel disease was met with a great deal of controversy.⁵ A more highly debated topic was the proposed association between the symptoms of MCOS and a series of cardiac complications, which was first recognized by Dr. Noboru Tanaka, a Japanese pathologist, following the autopsy of a child diagnosed with MCOS.⁵ The cardiac sequelae were first observed clinically and subsequently presented to the public in a report by Dr. Takajiro Yamamoto in 1968.⁶ A nationwide epidemiologic survey of MCOS in Japan released in 1970 confirmed the claims of Tanaka and Yamamoto and established the association between MCOS and cardiac complications caused by severe vasculitis.⁵

Kawasaki published an English translation of his original report of 50 patients in 1974.⁷ However, KD was independently recognized as a distinct disease in Hawaii in the early 1970's by Dr. Marian Melish, a specialist in pediatric infectious disease, and Dr. Raquel Hicks, a pediatric rheumatologist.⁵ The pair documented the disease in Hawaiian children, who were mostly Japanese Americans, and in 1973, after seeing photographs of children with KD from Japan and contacting Kawasaki, they concluded the syndrome they had encountered was the same as that recently documented in Japan.⁵ In addition, in 1971, Eunice Larson, a pediatric pathologist in Honolulu, observed the cardiac sequelae of a child (later recognized as having KD) and after further investigation made the connection with the Japanese KD.⁵ Following this independent discovery of the clinical and pathologic aspects of the disease, Melish, Hicks and Larson collaborated to publish a report in 1976 of their observations of KD in Asian/Pacific children from Hawaii.⁸

The history of the emergence of Kawasaki disease in Japan and the United States presents many questions that may have implications on the pathology and origin of the disease. It is matter of some interest whether there were any cases of KD prior to those observed by Kawasaki himself. An answer to this question may provide researchers with a reason for the independent recognition of KD by clinicians and pathologists around the world in the 1960's and 1970's⁵. Burns et al.⁵ presents three possible explanations for this phenomenon: (1) KD may have, in fact, been a novel disease that emerged in Japan and was in some way spread to Hawaii through Asian immigration; (2) KD and infantile periarteritis nodosa (IPN) may be on a spectrum of the same disease and, therefore, clinically mild KD may have been systematically misdiagnosed; (3) improvements in medical care, especially the use of antibiotics, may have assuaged the symptoms of KD, such as rash and fever, which allowed the clinical criteria to be independently established. It is possible that with thorough investigation, knowledge of the detailed history of the appearance and apparent dispersal of KD among children throughout the world may add direction to the search for a causative agent.

III. Epidemiology

Although KD has been observed in children of all racial groups, the disease is over-expressed among Asian populations, especially the Japanese and those of Japanese descent living outside of the country.² The disease is most common among children, with 85% of those diagnosed aged less than 5 years, although the peak incidence in the U.S. occurs in infants aged 1-2 years and it is rarely observed in infants under the age of 1 and adolescents.⁹ The annual incidence rate in Japan, as measured in 1998, is 111/100,000 children under the age of 5.¹⁰ The annual incidence rate in the US, which is measured regionally, is between 4 and 15/100,000 children under the age of 5.⁵ The disease occurs in temporally limited regional epidemics, which are most common in the late winter and spring, and boys are more commonly affected than girls, with a male to female ratio of 1.5:1.⁹

The epidemiology of KD provides insight into the nature of the disease and aids in our understanding of possible causes or risk factors for KD by identifying populations that are more at risk of developing the disease. The fact that some ethnic groups are at higher risk for KD than others suggests that there may be ethnicity-related genetic differences or possibly shared environmental exposures that make some individuals more susceptible to KD. Further understanding of the role of genetics in KD and analysis of possible environmental causes may result in the determination of the causative agent and facilitate the investigation of other risk factors.¹¹

In addition, the epidemiologic patterns of KD provide evidence for the role of an infectious agent in its pathogenesis,⁹ such as: (1) the seasonal peaks of outbreaks in winter and spring, which are characteristic of infectious diseases; (2) the regional grouping of cases; (3) the rarity of illness in the infant and adolescent, suggesting maternal antibodies acquired transplacentally may protect infants and asymptomatic infection followed by immunity in older children; and (4) the clinical similarities between KD and other infectious diseases, such as scarlet fever. Although there has been no documentation of person-to-person spread of KD¹² and the etiology remains widely disputed, epidemiologic findings strongly support the hypothesis proposing a causative infectious agent of KD, and serve as the basis of etiological investigations.

IV. General Clinical Features

Currently there is no diagnostic test for KD; therefore, diagnosis is based on clinical criteria, which were established by the Centers for Disease Control in 1985¹³ and are listed in Table 1. KD is identified based on the presentation of a fever lasting at least 5 days and four out of five other criteria, or, if coronary artery aneurysms are present, three clinical criteria are sufficient for diagnosis of KD. Other helpful, although less common, signs for diagnosis of KD include meningeal irritation, hydrops of the gallbladder, diarrhea, vomiting, abdominal pain, and irritability, which can be related to aseptic meningitis.¹⁴

Almost all of the morbidity and mortality associated with Kawasaki disease are associated with the cardiovascular system and cardiac complications caused by severe vasculitis in the acute phase of the disease.¹⁵ KD is the leading cause of acquired heart disease in children,³ with coronary artery aneurysms occurring in 25% of untreated patients.¹⁶ The prognosis of disease in children with cardiac complications generally depends on the size of the aneurysm; small aneurysms can resolve completely.¹⁷

V. Etiology and Pathogenesis

It is likely that the immunostimulation of an inflammatory response plays a central role in the pathogenesis of KD. This is suggested by, for example, the demonstration of vascular inflammation of the coronary arteries in addition to other smaller blood vessels,¹⁸ the increased numbers of circulating activated T-cells and monocytes, and polyclonal B cell activation during the acute phase,^{19, 20} and clinical features associated with immune activation, such as fever and lymphadenopathy.²¹ Therefore, etiologic investigations are concentrated on determining a possible trigger of this hyper-reactive immune response.

Despite decades of investigation, the etiology of KD remains unknown. The focus of research since the emergence of the disease has shifted from a proposed association between exposure to freshly cleaned carpets and house dust mites to the search for a possible infectious agent responsible for KD pathogenesis. Although debate remains, it has been suggested that the causative agent is a toxin producing a superantigen that triggers the immune-mediated response leading to the pathology of KD. Despite the large number of hypotheses, the etiologic agent remains unknown. Continued research is integral to the evolution of the management of patients with KD because understanding the mechanism of pathogenesis would allow the development of a diagnostic test and, therefore, more rapid and comprehensive treatment.

Association between exposure to freshly cleaned carpets and KD

For a period of time in the 1980's and early 1990's, an association between exposure to carpet cleaners and KD was suspected and a number of studies were carried out investigating rug shampoo as a possible causative agent. Table 2 summarizes the studies and their results. In 1982, Patriarca et al.²² published a paper suggesting that an outbreak of KD in Colorado may have been related to the exposure of children to recently applied rug shampoo. The authors suggested that the association might be the result of hypersensitivity to the anionic detergents present in all of the shampoos used, which was unlikely due to the number of household cleaning products also containing such compounds, or due to the inhalation of an infectious agent that undergoes aerosolization during the cleaning process. Although the authors admitted that a causal association between rug shampoo and KD was speculative, this report incited an abundance of contradicting case-control studies and over a decade of controversy.

Four studies, Lin et al.,²³ Rogers et al.,²⁴ and Glode et al.,²⁵ and Klein et al.²⁶ investigating the association between the exposure of children to freshly cleaned carpets and KD in cases from Maryland, New York, Michigan, Colorado and Wisconsin reported no significant associations. However, three of the four studies provided possible reasons for why their analyses could have incorrectly demonstrated negative results. Lin et al. stated that the findings were only conclusive in relation to the Maryland epidemic and did not feel confident that their results negated the possibility of some relationship between rug shampoo and KD. Rogers et al. suggested that their negative result could be due to the fact that the microorganisms that are aerosolized during carpet cleaning also reside in other areas of households, which would reduce the likelihood of recognizing such an association. In addition, Glode et al. suggested that failure to confirm the association between carpet cleaning and KD could be due to differences in endemic (which they studied) and epidemic cases or to the small number of cases analyzed (18 children with KD). Klein et al, on the other hand, assessed rug shampoo exposure, dust collection from homes, and serum specimen concentration, and concluded that their research was not in support of the hypothesis that there was an association between KD and carpet cleaning. Although drawing into question the relationship between carpet cleaning and KD, the four studies mentioned do not negate the possibility of some association among epidemic cases.

Following the initial proposal of the association between carpet cleaning and KD, there were three studies released confirming the positive results. Rauch et al. 1988 and 1991,^{27, 28} and Fatica et al.²⁹ demonstrated this association among children affected by epidemics in Wisconsin, Colorado, Montana, and New York. In Patriarca et al, Fatica et al. and Rauch et al. 1991, the intervals between the exposure to freshly cleaned carpets and the onset of illness were generally clustered around 13-30 days. This strengthens the argument for the relationship between carpet cleaning and KD because it suggests that this may be a standard period of incubation or induction for an infectious agent that may have been aerosolized during the process of cleaning.²⁸ In addition, the report by Fatica et al. included in its study sample a family with two siblings affected, both of whom became sick within 45 days of the application of rug shampoo and one of which had a recurrence 2 weeks following a second episode of rug cleaning. The authors indicated that such a strong association demonstrated by the recurrence was unlikely due to chance.

Despite the numerous case-control studies, controversy remains over the association between carpet cleaning and KD. There are a number of studies supporting and negating the relationship, yet no definitive consensus has been reached. Although the association has not been established, there has been a great deal of speculation about the possible mechanisms of this association.

Possible role of house dust mites in the pathogenesis of KD

Prior to the publication of Patriarca et al., two Japanese reports were released, Furusho et al.³⁰ and Hamashima et al.,³¹ suggesting a house dust mite antigen as a causative agent of KD. Furusho et al. reported increases in the serum levels of anti-mite specific IgG and also IgE during the acute phase of illness and Hamashima et al. found Rickettsia-like bodies, which have been found in biopsy specimens of KD,³² in the digestive canal of mites in the rooms of KD patients. The studies suggest that the house dust mites may act as an allergen stimulating an allergic reaction or, more likely, that the mites act as vectors carrying a microorganism, possibly rickettsiae, that may be the causative agent of KD. Although Furusho et al. suggested that the increased use of Tatami mats (which provide the moisture, darkness, human skin scales, and yeast that dust mites require) by Japanese people could account for the increasing incidence of KD in Japan, it was Patriarca et al. and a number of subsequent American studies that proposed the association between carpet cleaning, the aerosolization and inhalation of house dust mites, and KD. In 1983, Ohga et al.³³ found that within an outbreak in Japan, children with KD were more likely than control children to have rugs, which further supported this theory. Table three shows the results of house dust mite studies.

However, following the initial reports of Furusho and Hamashima, there have been a number of studies refuting the hypothesis that a mite antigen may be involved in the pathogenesis of KD.^{34, 35} In addition, there have been studies showing that there was not increased exposure of KD patients to house dust mites when compared to controls,^{25, 26} which may indicate that mites are not requisite vectors of the causative agent of KD. Glode et al.²⁵ analyzed the house dust mite content, serum anti-mite antibody levels, and exposure to freshly cleaned carpets in cases and controls of a Colorado epidemic and found no significant differences in the prevalence of mites (content was very low in all households), the exposure to freshly cleaned carpets, nor serum antibody concentrations. In addition, Klein et al.²⁶ found no variation in the density and species-specific prevalence of house dust mites in the home of KD patients. Although these results do not negate the possibility that house dust mites may act as a vector for the causative agent of KD, it does demonstrate that an epidemic can occur in the absence of significant mite exposure.

It is also possible that the association between the agitating of dust mites during carpet cleaning and KD is in some way masked by the properties of mite growth. It is clear that these mites do not reside only in carpets, but also in the dust that collects in heating vents, mattresses and pillows, stuffed toys and furniture.²⁴ This suggests that although the process of carpet cleaning may provoke the infusion of dust mites into the air, facilitating inhalation, it may not be the only source of mites, and that it is possible KD patients not exposed to rug shampoo came into contact with mites from another source.²⁴ In addition, mite prevalence is affected by atmospheric relative humidity and seasonal prevalence;²⁵ therefore, such factors may affect the validity of studies demonstrating no association between the presence of dust-mites in households of children with KD.

Current consensus on the role of carpet cleaning and mite exposure

Although there have been a number of case-control epidemiologic studies carried out to investigate the association between the exposure to freshly cleaned carpets and KD, it is clear that no such relationship has been concretely established. In addition, it has been demonstrated that a mite antigen is not involved in the pathogenesis of KD and that there is no significant association between KD and the presence or absence of house dust mites, although the possibility that house dust mites may act as vectors in the pathogenesis of KD remains undetermined.³⁶ As a result of such contradictory results and epidemiologic data, the focus of etiologic research has shifted from the proposed association with carpet cleaners and dust mites to investigation of potential infectious agents as the underlying cause.³⁷ Despite this shift of focus, the association has not been negated and, therefore, parents should take a conservative approach in preventing the exposure of children to freshly cleaned carpets and mite-containing furniture.

Current etiologic and pathogenic hypotheses

Although the etiology of KD remains unknown, the majority of evidence suggests an infectious causative agent, possibly leading to disease in genetically susceptible hosts.¹⁴ The epidemiologic factors suggesting an infectious cause, which were mentioned previously, are supported by the pathology of the disease, which resembles an immune-mediated inflammatory response resulting in endothelial damage. However, after years of searching, all possible known pathogens have been discarded.⁵

There have been a number of studies that suggest toxins acting as a superantigen may trigger the immune-mediated response responsible for KD, although this hypothesis has been highly debated. This theory proposes that an infectious agent, possibly a bacterial toxin such as streptococci or staphylococci, produces an antigen that is highly immunostimulant, leading to the activation of large portions (up to 20%) of beta chain T cells and an overreaction of the immune system.³⁷ KD is very similar in its clinical presentation to a number of diseases caused by exotoxins acting as superantigens, such as toxic shock syndrome (TSS) and scarlet fever.³⁶ In addition, Abe et al.³⁸ reported the selective expansion of Vbeta 2+ cells and Vbeta8.1 + T cells in KD patients during the acute phase but not the convalescent phase, which strongly supports the argument for a superantigen-mediated pathogenesis. However, subsequent studies have failed to show such Vbeta T cell expansion.³⁹

Additional hypotheses of the pathogenesis of KD include the role of pro-inflammatory cytokines⁴⁰ and anti-endothelial cell antibodies (AECA)⁴¹ in endothelial damage. In addition, the report of IgA plasma cells in the vascular walls of children with KD by Rowley et al.⁴² suggests that KD is the result of the entry of an etiologic agent through a respiratory or gastrointestinal portal leading to the stimulation of an antigen-driven immune response with prominent IgA plasma cell involvement. Although there have been many advances in our understanding of the immune-related response that leads to KD and possible triggers, there remains no definitive causative agent.

VI. Summary and Conclusions

Since its first description in Japan in 1967 by Kawasaki, KD has affected children of all racial groups, especially those under the age of 5. Due to the absence of a diagnostic test for KD, the disease often masquerades as other exanthems and is misdiagnosed, leading to an unnecessary increase in the incidence of cardiac complications, especially coronary artery aneurysms. Although during the 1980's scientists explored an association between KD and carpet cleaners, the numerous contradictory case-control studies published during this period have resulted in no consensus on the matter. It has been demonstrated that a dust mite antigen is not likely to be the causative agent of KD, but it remains to be determined if mites may act as vectors of the causative agent. Although no concrete link has been established, parents should be advised to be conservative in allowing their children to be exposed to freshly cleaned carpets. Future areas of etiologic and pathogenic research include bacterial toxins producing superantigens, anti-endothelial antibodies, and normal antigen-driven processes, all of which may produce new insight into the nature of the disease. Despite the broad scope of research into this disease, the causative agent of KD remains a mystery.

Editor's Note: This paper is a project of the Community-Based Learning Initiative (CBLI) at Princeton University. Students and faculty collaborate on research projects with community organizations, matching student interests with research needs. CBLI and CHEC continue to enjoy a productive collaboration with the goal of facilitating people's access to knowledge that affects their lives.

References:

1. Kawasaki, T., [Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children]. Arerugi, 1967. 16(3): p. 178-222.
2. Leung, D.Y., P.M. Schlievert, and H.C. Meissner, The immunopathogenesis and management of Kawasaki syndrome. Arthritis Rheum, 1998. 41(9): p. 1538-47.
3. Curtis, N., Kawasaki disease. BMJ, 1997. 315(7104): p. 322-323.
4. Lang, B. and C.M. Duffy, Controversies in the management of Kawasaki disease. Best Pract Res Clin Rheumatol, 2002. 16(3): p. 427-42.
5. Burns, J.C., et al., Kawasaki Disease: A Brief History. Pediatrics, 2000. 106(2): p. 27-35.
6. Yamamoto, T. and J. Kimura, Acute febrile mucocutaneous lymph node syndrome (Kawasaki): subtype of mucocutaneous ocular syndrome of erythema multiforme complicated with carditis [in Japanese]. Jpn J Pediatr, 1968. 21: p. 336-339.
7. Kawasaki, T., et al., A new infantile acute febrile mucocutaneous lymph node syndrome (MLNS) prevailing in Japan. Pediatrics, 1974. 54(3): p. 271-6.
8. Melish, M.E., R.M. Hicks, and E.J. Larson, Mucocutaneous lymph node syndrome in the United States. Am J Dis Child, 1976. 130(6): p. 599-607.
9. Newburger, J.W. and J.C. Burns, Kawasaki Disease. Vascular Medicine, 1999. 4: p. 187-202.
10. Yanagawa, H., et al., Incidence Survey of Kawasaki Disease in 1997 and 1998 in Japan. Pediatrics, 2001. 107(3): p. e33-.
11. Gardner-Medwin, J.M., et al., Incidence of Henoch-Schonlein purpura, Kawasaki disease, and rare vasculitides in children of different ethnic origins. Lancet, 2002. 360(9341): p. 1197-202.
12. Yanagawa, H., et al., Nationwide surveillance of Kawasaki disease in Japan, 1984 to 1993. Pediatr Infect Dis J, 1995. 14(1): p. 69-71.
13. Rauch, A.M. and E.S. Hurwitz, Centers for Disease Control (CDC) case definition for Kawasaki syndrome. Pediatr Infect Dis, 1985. 4(6): p. 702-3.
14. Cimaz, R. and F. Falcini, An update on Kawasaki disease. Autoimmunity Reviews, 2003. Article In Press.
15. Laupland, K.B. and H. Dele Davies, Epidemiology, etiology, and management of Kawasaki disease: state of the art. Pediatr Cardiol, 1999. 20(3): p. 177-83.
16. Durongpisitkul, K., et al., The prevention of coronary artery aneurysm in Kawasaki disease: a meta-analysis on the efficacy of aspirin and immunoglobulin treatment. Pediatrics, 1995. 96(6): p. 1057-61.
17. Kato, H., E. Ichinose, and T. Kawasaki, Myocardial infarction in Kawasaki disease: clinical analyses in 195 cases. J Pediatr, 1986. 108(6): p. 923-7.
18. Leung, D.Y., Kawasaki Disease. Curr Opin Rheumatol, 1993. 5: p. 41-50.
19. Leung, D.Y., et al., Immunoregulatory T cell abnormalities in mucocutaneous lymph node syndrome. J Immunol, 1983. 130: p. 2002-2004.
20. Leung, D.Y., et al., Endothelial cell activation and high interleukin-1 secretion in the pathogenesis of acute Kawasaki disease. Lancet, 1989. 2: p. 1298-1302.
21. Leung, D.Y., et al., Superantigens and Kawasaki Syndrome. Clinical Immunology and Ummunopathology, 1995. 77(2): p. 119-126.
22. Patriarca, P.A., et al., Kawasaki syndrome: association with the application of rug shampoo. Lancet, 1982. 2(8298): p. 578-80.
23. Lin, F.Y., et al., Kawasaki syndrome. A case-control study during an outbreak in Maryland. Am J Dis Child, 1985. 139(3): p. 277-9.
24. Rogers, M.F., et al., Kawasaki syndrome. Is exposure to rug shampoo important? Am J Dis Child, 1985. 139(8): p. 777-9.
25. Glode, M.P., et al., Kawasaki syndrome and house dust mite exposure. Pediatr Infect Dis, 1986. 5(6): p. 644-8.
26. Klein, B.S., et al., Kawasaki syndrome: a controlled study of an outbreak in Wisconsin. Am J Epidemiol, 1986. 124(2): p. 306-16.
27. Rauch, A.M., Kawasaki syndrome: issues in etiology and treatment. Adv Pediatr Infect Dis, 1989. 4: p. 163-82.
28. Rauch, A.M., et al., Outbreak of Kawasaki syndrome in Denver, Colorado: association with rug and carpet cleaning. Pediatrics, 1991. 87(5): p. 663-9.
29. Fatica, N.S., et al., Rug shampoo and Kawasaki disease. Pediatrics, 1989. 84(2): p. 231-4.
30. Furusho, K., et al., Possible role for mite antigen In Kawasaki disease. Lancet, 1981. 2(8239): p. 194-5.
31. Hamashima, Y., et al., Mite-associated particles in Kawasaki disease. Lancet, 1982. 2(8292): p. 266.
32. Hamashima, Y., K. Kishi, and K. Tasaka, Rickettsia-like bodies in infantile acute febrile mucocutaneous lymph-node syndrome. Lancet, 1973. 2(7819): p. 42.
33. Ohga, K., et al., Kawasaki disease and rug shampoo. Lancet, 1983. 1(8330): p. 930.
34. Jordan, S., et al., Lack of evidence for mite-antigen-mediated pathogenesis in Kawasaki disease. Lancet, 1983. 1(8330): p. 931.
35. Tang, R., et al., The significance of mite antigens in Kawasaki disease. Chinese Journal of Microbiology and Immunology, 1987. 20(1): p. 29-36.
36. Lloyd, A.J., C. Walker, and M. Wilkinso, Kawasaki disease: is it caused by an infectious agent? Br J Biomed Sci, 2001. 58(2): p. 122-8.
37. Fischer, P., et al., Kawasaki disease: update on diagnosis, treatment, and a still controversial etiology. Pediatr Hematol Oncol, 1996. 13(6): p. 487-501.
38. Abe, J., B. Kotzin, and C. Meissner, Characterization of T-cell repertoire changes in acute Kawasaki disease. J Exp Med, 1993. 177: p. 791-6.
39. Pietra, B., et al., TCR Vbeta family repertoire and T cell activation markers in Kawasaki disease. J Immunol, 1994. 153: p. 1881-1888.
40. Matsubara, T., S. Furukawa, and K. Yabuta, Serum levels of tumor necrosis factor, interleukin 2 receptor, and interferon-gamma in Kawasaki disease involved coronary-artery lesions. Clin Immunol Immunopathol, 1990. 56(1): p. 29-36.
41. Grunebaum, E., et al., The role of anti-endothelial cell antibodies in Kawasaki disease - in vitro and in vivo studies. Clin Exp Immunol, 2002. 130(2): p. 233-40.
42. Rowley, A., et al., IgA plasma cells in vascular tissue of patients with Kawasaki syndrome. J Immunol, 1997. 159(12): p. 5946-5955.