Lack of Association Between Kawasaki Syndrome and Infection With Parvovirus B19, Human Herpesvirus 8, Tt Virus, Gb Virus C/Hepatitis G Virus or Chlamydia Pneumoniae

The etiology of Kawasaki syndrome (KS), an acute, self-limited multisystem vasculitis with clinical and epidemiologic features consistent with an infectious process, remains undiscovered, despite many investigations using classic as well as newly developed molecular biologic techniques. The marked restriction to young children suggests that a causative microbe may be easily contracted because it appears to result in disease or immunity early in life. Certain febrile childhood exanthematous diseases with etiologies that had been obscure have recently been found to be caused by ubiquitous rather than rare microbial pathogens. A recent example is roseola infantum, a relatively benign exanthematous disease of infancy, long thought to be an infectious disease. Roseola or exanthem subitum is now known to be caused by human herpesvirus 6 (HHV 6),1 which, like human herpesvirus 5 (or human cytomegalovirus), usually results in asymptomatic infection or poorly differentiated illnesses during childhood. Following this logic KS may be caused by a ubiquitous infectious agent commonly acquired during infancy or early childhood, which has not been detected by classical culture or molecular techniques. Many groups of viruses, including retroviruses, have been excluded as a cause of KS,2 and no differences have been found in infection rates with cytomegalovirus, Epstein-Barr virus and HHV 6 between patients with KS and age-matched controls.3 Human parvovirus causes erythema infectiosum but also asymptomatic infections and undifferentiated febrile illnesses. Prompted by the controversy over the etiologic link between parvovirus B19 infection and KS4-7 and by the recent discovery of two viruses with no definitive disease associations, specifically GB virus C/hepatitis G virus (GBV-C/HGV) and TT virus (TTV), along with studies implicating Chlamydia pneumoniae in the pathogenesis of atherosclerotic heart disease through molecular mimicry,8 as well as the possible link between C. pneumoniae infection and KS,9 we utilized PCR to determine whether KS could be the result of infection by these microbes. In addition we included human herpesvirus 8 (HHV 8), a gamma herpesvirus causing Kaposi's sarcoma, in our search. Materials and methods. A total of 123 patients (55 female and 68 male; age range, 1 month to 8 years; mean and median ages, 2.4 and 1.5 years, respectively), who presented with KS at the Kapi'olani Medical Center for Women and Children between 1980 and 1999, were studied. Whole blood, plasma/serum, peripheral blood mononuclear cells (PBMC) and urine were collected during the acute phase (≤7 days) of KS, after obtaining informed consent from parents or guardians. This study was approved by the research and ethics committee of Kapi'olani Medical Center for Women and Children. RNA and DNA were extracted from plasma or serum using the QIAGEN Viral RNA Extraction kit (QIAGEN, Chatsworth, CA), and DNA was extracted from whole blood and PBMC, using the QIAGEN Blood Extraction kit (QIAGEN).10 DNA from urine was extracted by a conventional extraction procedure, consisting of proteinase K digestion and phenolchloroform purification.11 Oligonucleotide primers and thermocycling conditions for the amplification of parvovirus B19,12 HHV 8,13 TTV,10 GBV-C/HGV14 and C. pneumoniae15 are shown in Table 1. For GBV-C/HGV amplification RNA was initially reversed transcribed into complementary DNA in a 20-μl reaction mixture, containing 8 μl RNA, 500 ng of random hexamers (Promega, Inc., Madison, WI), 20 units of RNase inhibitors, 10 mM dithiothreitol, 1 mM each deoxynucleotide triphosphate, 75 mM KCl, 3 mM MgCl2 and 150 units of Superscript reverse transcriptase H− (Life Technologies, Gaithersburg, MD) and incubated at 42°C for 50 min and then at 72°C for 15 min, before storage at −30°C. PCR was performed in 50-μl reaction mixtures, consisting of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.8 mM MgCl2, 0.1 μM each primer, 0.2 μM each deoxynucleotide triphosphate, 1.25 units of Thermus aquaticus DNA polymerase (Perkin Elmer, Foster City, CA) and 5 μl of complementary DNA or DNA. For nested PCR 5 μl of the original amplicon were subjected to 45 cycles (HHV 8, TTV and GBV-C/HGV) or 35 cycles (parvovirus B19). PCR amplicons were size-fractionated by electrophoresis on 2% agarose gels and visualized by ethidium bromide staining.TABLE 1: Primer sequences and cycling conditions for amplification of microbial sequences in Kawasaki syndrome In addition acute phase sera from 46 of the 123 patients were tested, in duplicate, for IgM and IgG antibodies against parvovirus B19 by enzyme immunoassay (MarDx Diagnostics, Inc., Carlsbad, CA). Results. As shown in Table 2 IgG or IgM antibodies against parvovirus B19 were detected in the acute phase of 26 of 46 (57%) KS patients, whereas only 5 of 46 (11%) had both IgG and IgM antibodies. Patients in the 0- to 2-year age group had the highest age-specific seroprevalence (15 of 20 or 75%) of IgG or IgM antibodies against parvovirus B19, whereas there was a decrease in seroprevalence with increasing age. On the other hand only 2 of the 46 patients had parvovirus B19 DNA in their blood samples, as detected by PCR.TABLE 2: Seroprevalence of parvovirus B19 infection among 46 patients with Kawasaki syndrome from Hawaii C. pneumoniae gene sequences were not detected in PBMC and urine of 26 KS patients, and HHV 8 sequences were not detected in plasma or serum from 77 KS patients. Of these 77 KS patients we amplified TTV DNA in only 2 female patients (ages, 4 and 5.4 years), who subsequently cleared the infection within 2 months. Similarly, GBV-C/HGV was detected in 2 other patients, both female, 1.5 and 2 years old. No convalescent phase samples were available to determine the persistence of GBV-C/HGV in these 2 patients. Discussion. The presence of IgA-secreting plasma cells within vascular lesions of KS patients16 suggests entry of the etiologic agent by way of the gastrointestinal or respiratory tracts. In this regard the pathogens we examined are transmitted by either the fecal-oral or respiratory aerosol routes. Moreover parvovirus B19 and HHV 8 appear to target vascular endothelial cells. TTV has been found in young children and infants17 as well as shown to be transmissible by fecal-oral routes,18 indicating possible acquisition of the virus during childhood. Our data do not support an etiologic association between KS and infection with parvovirus B19 or HHV 8 or with TTV, GBV-C/HGV and C. pneumoniae. Although we failed to demonstrate an etiologic link between KS and infection with the above named microbes, these data do not exclude the possibility that KS is caused by an infectious agent. A plausible explanation for the largely negative results may be that KS is caused by a previously unidentified microbial agent. Alternatively the agent may have already been cleared from the sampled compartment at the time of specimen collection. A molecular approach, with the use of representational difference analysis, to identify cryptic, nonhost, microbial genomic sequences in KS patients, currently under way, may offer further clues about an infectious etiology. Acknowledgments. This work was supported by US Public Health Service Grant G12RR/AI-03061 from the Research Centers in Minority Institutions Program of the National Center for Research Resources, National Institutes of Health. Pong K. Chua, B.S. Vivek R. Nerurkar, Ph.D. Qigui Yu, M.D., Ph.D. Cora L. Woodward, B.S. Marian E. Melish, M.D. Richard Yanagihara, M.D. Retrovirology Research Laboratory; Pacific Biomedical Research Center (PKC, VRN, QY, CLW, RY) Department of Pediatrics; John A.