Techniques developed to study HIV infection revealed the existence of HHV-6. Initially found in the lymphocytes of patients with AIDS and lymphoproliferative diseases, this novel virus bore a familial resemblance to the herpesviruses and was adopted as their sixth member [1]. Composed of a large double-stranded DNA, DNA polymerase (but not thymidine kinase), and several important glycoproteins, HHV-6 can infect many types of cell but, like HIV, primarily affects CD4 cells. As with other herpesviruses, persistence or latency occurs; genomic material can then be detected in peripheral blood lymphocytes, secretions, and cerebrospinal fluid in healthy and immunosuppressed persons.

Human herpesvirus-6 actually comprises nearly identical twins, HHV-6A and HHV-6B. The genomic homogeneity of these variants is about 95%, but they differ in antigenicity; cell tropism; epidemiology; and, possibly, pathogenicity [2]. Human herpesvirus-6A was isolated first, and A strains have been recovered primarily from adults, especially immunosuppressed patients. Overall, however, HHV-6B isolates predominate. Human herpesvirus-6B is essentially the sole variant isolated from children with primary infection, and a direct causal relation with disease has been established for the B variant [2-4]. In adults, the quandary is whether the presence of HHV-6 is latent or causes disease. Adding to the mystery is the paucity of reports to date of primary infection with variant A.

The many miens and mimicries of HHV-6 are just beginning to be recognized. Seroepidemiologic studies quickly indicated universal acquisition during infancy: All newborns possess maternally derived antibody to HHV-6 that decreases to a nadir after several months [4]. This is followed by natural infection that proceeds over the next few months with an alacrity and universality that perhaps is exceeded by no other virus. This suggests that antibody is protective although not perfectly so) and that HHV-6 is readily available in the infant's environment, probably primarily in the secretions of caretakers. Detection of HHV-6 in maternal cervical secretions and in the peripheral blood lymphocytes of normal newborns also suggests that perinatal transmission is possible [3].

Since Yamanishi and colleagues [4] first isolated HHV-6 from 4 children with roseola, this disease has been almost the exclusive manifestation of primary HHV-6 infection reported from Japan [5]. In children in the United States, the initial expression of infection seems more varied. In a prospective study of more than 4000 children younger than 3 years of age who presented with acute illnesses in Rochester, New York, the most frequent HHV-6-associated diagnosis was nonspecific febrile illness, often with otitis and possible sepsis. However, gastrointestinal signs predominated in 25% of patients and respiratory signs predominated in another 25% [3]. Roseola accounted for only about 15% of illnesses. Human herpesvirus-6 caused 10% of visits for acute illness for children in the first 3 years of life and about one fourth of visits for infants 6 to 12 months of age. These visits, plus the evaluation often required to rule out bacterial and other serious infections, result in a potentially appreciable burden on our health care system.

Of concern and interest is the neurotropism of HHV-6, evidenced by the frequent central nervous system manifestations in infancy, including bulging fontanelle, meningoencephalitis, and febrile seizures. The HHV-6 genome persists in the cerebrospinal fluid of some children with acute or past infection [6] and in the brains of previously healthy persons who are dying of other causes [7]. Whether this is a bridge to adult disease is unknown.

In adults, HHV-6 is primarily associated with immunocompromising diseases, and its role is unclear. Since the original discovery of HHV-6 in patients with AIDS, the possible role of the virus as a cofactor in HIV infection has been questioned. Both viruses can infect CD4 cells concurrently in vitro, HHV-6 can upregulate the expression of CD4 cells, and the intermediate early gene can transactivate the long terminal repeat of HIV [2, 8]. Human herpesvirus-6 is detected frequently in HIV-infected patients, especially those with higher CD4 counts [9]. The chance of HHV-6 dissemination and the viral burden may be greater in HIV-infected patients. Reactivation of HHV-6 has been implicated as the cause of such complications as fatal pneumonitis and encephalitis in patients with AIDS and rash, fever, graft rejection, and bone marrow suppression in transplant recipients [2, 8, 10]).

In addition, a role for HHV-6 in the development of multiple sclerosis has been suggested by the differential location of the virus in brains of controls compared with patients with multiple sclerosis; in the latter, virion proteins were expressed only in neurons around multiple sclerosis plaques [11].

The bridge between infant infection and adult reactivation requires better delineation if we are to predict who is susceptible to the pathogenic potential of HHV-6. Each new piece of information adds to a complex collage with more conundrums rather than less. Strain variation within each variant group is so minimal that precise tracking of the transmission and course of HHV-6 remains problematic, and little is known about primary infection with HHV-6A. Therapeutic options generally have not been explored, despite the demonstration of in vitro inhibition by ganciclovir and foscarnet [2]. Future management of HHV-6 infection will require feasible and sensitive means of differentiating primary or active infection from persistent infection, latency, and reactivation.

Since HHV-7, the closest relative of HHV-6, was first isolated in 1990, much has been revealed about its inner structural core, but little is known about its outer clinical personality and importance [12]. Human herpesvirus-7 has morphologic similarity to and appreciable homology with HHV-6, resulting in serologic cross-reactivity and confusion in diagnosis [13]. Like HHV-6, HHV-7 is ubiquitous, and infection with HHV-7 generally occurs after infection with HHV-6. The parasitic prowess of HHV-7 is shown by its isolation from the saliva of 75% of healthy adults [12, 13]. The HHV-7 genome can also be detected in peripheral blood lymphocytes and, occasionally, in cervical secretions.

The clinical manifestations and implications of HHV-7 infection have yet to be clarified. A few cases of primary infection, with documented viremia and seroconversion, have been reported in Japanese children with roseola and in more varied febrile illnesses in children in the United States [14, 15]. Thus, HHV-7 may account for some cases of roseola (especially recurrent ones) and may sometimes cause febrile seizures. Of eight children with HHV-7 viremia in the prospective Rochester study [3], 75% had complicating seizures [15].

A piquant part of HHV-7's personality is its interactions with other viruses. Human herpesvirus-7 may reactivate HHV-6 and vice versa, thus confounding disease etiology [12, 14, 15]. Both HHV-7 and HHV-8 primarily infect CD4 cells, and HHV-7 downregulates CD4 expression, an essential part of HHV-7's membrane receptor [16]. The downregulation of CD4 expression and HHV-7's marked reciprocal interference with HIV replication in vitro suggest an investigative approach to controlling HIV infection. Whether HHV-7 is a cofactor that can influence the natural course of HIV infection awaits study.

One hundred twenty-two years after Kaposi first described an idiopathic pigmented sarcoma of the skin, Chang and colleagues [17] discovered Kaposi sarcoma-associated herpesvirus, or HHV-8. Two novel DNA fragments isolated from Kaposi lesions showed homology to herpesviruses. Subsequent characterization of the large (160- to 165-kb) genome of this virus identified at least 97 genes, some closely homologous to the human genes cyclin D, bcl-2, and interleukin-6. These genes have a potential role in oncogenesis and the inhibition of apoptosis. Gene expression in Kaposi sarcoma seems highly restricted, which suggests that HHV-8 infects most Kaposi sarcoma cells latently rather than lytically.

Human herpesvirus-8 has been detected primarily in patients with Kaposi sarcoma, body cavity-based lymphomas, and multicentric Castleman disease and in the skin lesions of patients who received transplants. Of 224 AIDS-related and non-AIDS-related Kaposi sarcoma lesions, 97% contained HHV-8 compared with only 2% of 449 control tissue specimens [18]. The causal role of HHV-8 in these lesions is supported by studies showing that the presence of HHV-8 in peripheral blood lymphocytes precedes and predicts the risk for Kaposi sarcoma [18].

The sexual transmission of HHV-8 seems likely from the presence of HHV-8 in semen; this is supported by epidemiologic data showing that HIV-infected homosexual men have a much greater risk for Kaposi sarcoma than do HIV-infected patients with hemophilia [18, 19]. Additional routes of infection, such as saliva, are likely, given the different epidemiology of Kaposi sarcoma in such areas as the Mediterranean.

Information on the prevalence of HHV-8 in healthy populations is controversial and conflicting. An Italian study [19] detected HHV-8 in about 10% of patients with and patients without an underlying pathologic condition; another study found that HHV-8 is highly associated with Kaposi sarcoma or related conditions [20]. These conflicting data may arise partly from the currently limited and variably sensitive diagnostic tests and from divergences in viral load in healthy and high-risk populations; this, in turn, depends on differing cofactors, such as genetics, geography, and concurrent infections.

The extraordinary growth in the importance of HHV-8 in the past 3 years portends rapid progress in assembling the jigsaw puzzle of the clinical associations, covert cofactors, and epidemiology of this virus, including routes of transmission and risk to sexual partners and infants of infected mothers. Integral to solving the puzzles that surround these new herpesviruses is the development of improved diagnostic assays and therapies. The latter would provide not only clinical benefit but also options for intervention studies to determine whether the role of these herpesviruses in disease is that of actor or audience. The herpesvirus family saga continues.