The New Herpesviruses: Title and subTitle BreakEmerging Pathogens of Dermatological Interest FREE

RECENTLY, new herpesviruses have been discovered and named human herpesvirus (HHV) 6, 7, and 8. As the other members of the family Herpesviridae (Table 1), they may cause a primary infection, establish latent infection in a specific set of cells of their host, and then reactivate if conditions of altered immunity develop. They may have a profound impact on dermatological practice in the years to come. In this article, we review the current knowledge on them and try to provide a dermatological point of view.

Human herpesvirus 6 was first isolated in 1986 from peripheral blood lymphocytes of patients with the acquired immunodeficiency syndrome (AIDS) and a variety of lymphoproliferative disorders.¹ It is included in the β-HHV subfamily, mainly because of its genomic similarities with the human cytomegalovirus.²

Two distinct variants are recognized, named A and B, with different genetic, biologic, and immunologic features,³ with the variant A being more cytolytic and probably of a greater virulence. The etiologic role of the variant A has not been clearly identified, while HHV-6B is considered the major cause of exanthema subitum and other illnesses^{4 - 5} (Table 2).

Although HHV-6 is a lymphotropic virus with a tropism for CD4⁺ T lymphocytes,⁶ in which it replicates in vivo,⁷ it can also infect a variety of other cells.^{8 - 9} Human herpesvirus 6 induces a cytopathic effect in cultured lymphocytes, forming large ballooning cells and syncytia that correlate with the percentage of infected cells.^{1 ,10}

Serologic surveys have shown that most children become infected before the age of 3 years.^{11 - 12} Anti–HHV-6 IgG antibodies are present in 80% to 90% of adults and may disappear with aging.^{11 ,13} Anti–HHV-6 IgM antibodies are considered a marker of recent infection or reactivation, although approximately 5% of adults may be IgM seropositive at any time.¹⁴

The transmission is probably airborne, mainly from mother to child, most often via saliva.^{13 ,15} In addition, the presence of HHV-6 DNA in the cervixes of infected women in late pregnancy may account for perinatal infection.^{16 - 17} Human herpesvirus 6 antigens in abortive villous tissues from spontaneous abortions confirm such a possibility.¹⁸

Primary HHV-6 infection may present as exanthema subitum.¹⁹ In Japan, for example, 60% of HHV-6–infected children present with this disorder.²⁰ In contrast, in North America and Europe, approximately 70% of primary infections are inapparent or result in febrile illness without rash.^{21 - 23} Seizures, meningitis, and encephalitis are possible complications. In adulthood, the primary infection is very rare and its presentation includes hepatitis and a mononucleosislike illness.^{24 - 27}

Human herpesvirus 6 usually causes a lifelong latency after primary infection,²⁸ and probably is shed intermittently in the saliva without any symptoms or changes in the antibody titer.²⁹ The site of latency is still unknown, but the virus can be recovered from peripheral blood mononuclear cells (PBMC) in 17% to 90% of healthy adults.^{30 - 31} As it relates to Epstein-Barr virus (EBV), the lymphoreticular system harbors the latent HHV-6 while the salivary glands may be the sites of replication and virus reservoirs.^{32 - 33}

In immunocompromised patients undergoing kidney, liver, or bone marrow transplantation, latent HHV-6 may be reactivated, at times causing severe or fatal diseases,^{34 - 37} including pneumonia, bone marrow failure, and encephalopathy. In human immunodeficiency virus (HIV)-positive patients, HHV-6 may cause retinitis and disseminated infection that may expedite the progression of the disease.^{38 - 39} In culture, HHV-6 and HIV-1 have been shown to productively coinfect CD4⁺ T cells, resulting in accelerated HIV-1 expression and cellular death.¹⁰ Moreover, besides sharing the common primary tropism for CD4⁺ T cells, HHV-6 can infect and kill CD8⁺ T cells, natural killer cells, and mononuclear phagocytes.

In immunocompetent adults, the full spectrum of diseases is still unknown. Several disorders have been attributed to HHV-6, including infectious mononucleosislike illnesses, sinus histiocytosis,⁴⁰ and "gloves and socks" syndrome.⁴¹ Lymphoid and hematopoietic diseases are associated with elevated titers of HHV-6 antibody and yield HHV-6 DNA sequences. They include Hodgkin and non-Hodgkin lymphoma, atypical polyclonal lymphoproliferation, and angioimmunoblastic lymphadenopathy.^{42 - 43}

In addition, a transactivation effect on EBV replicative cycle has been observed⁴⁴ and specimens from angioimmunoblastic lymphadenopathy have shown simultaneously HHV-6 and EBV DNA sequences.⁴³ In vitro, HHV-6 is also capable to transform human epidermal keratinocytes. The transformed keratinocytes produce tumors when injected into nude mice.⁴⁵ So far, however, the direct involvement of HHV-6 in human malignant diseases has not been proved.

Although its role is unclear, HHV-6 DNA has also been demonstrated in lesions of histiocytosis X⁴⁶ and in the papular exanthema of bone marrow transplant recipients.⁴⁷ At least some of the rashes mentioned in literature are probably manifestations of a graft-vs-host disease.⁴⁸ In fact, a preceding or concurrent HHV infection had already been identified as a risk factor for the development of acute graft-vs-host disease.⁴⁹ Other inflammatory diseases, such as chronic fatigue syndrome and systemic lupus erythematosus, have been associated with HHV-6 infection, but the evidence is still equivocal.

Human herpesvirus 7 was isolated from activated CD4⁺ peripheral blood T cells of a healthy individual in 1990.⁵⁰ The second independent isolation from a patient with chronic fatigue syndrome was reported in 1992.⁵¹

DNA analyses have shown that HHV-7 is closely related to HHV-6 and cytomegalovirus.⁵¹ Human herpesvirus 7 has been included, therefore, in the β-HHV subfamily. Like HHV-6, HHV-7 is a prevalent lymphotropic virus with CD4⁺ T-cell tropism.^{50 - 51} Human herpesvirus 7 utilizes CD4 as a receptor, but cannot establish a productive infection in T cells. Nonetheless, it may inhibit HIV-1 replication by destroying CD4 receptors.⁵²

Human herpesvirus 7 can provide a transactivating function for HHV-6. Latent HHV-6 can be recovered after the cells are infected with HHV-7 and, once reactivated, the HHV-6 genome becomes prominent and HHV-7 disappears.⁵³

Human herpesvirus 7 is ubiquitous and infects more than 80% of children at infancy,⁵⁴ but the primary infection may appear later in life than that of HHV-6.⁵⁵

According to serologic studies, the HHV-7 prevalence rate in the United States exceeds 85%. In Japan, a lower rate has been reported. So far, the portal of entry of HHV-7, the site of primary infection, and the mechanism of latency have not been determined. The salivary glands are a likely site of HHV-7 replication as infectious viruses are present in the cell-free filtered saliva from which the virus can be more readily isolated than HHV-6.⁵⁶ On the contrary, PBMC are the site where the latent infection state is established. No disease has been reported that is clearly linked to HHV-7. The virus has been suggested to be a cause of exanthema subitum,⁵⁷ childhood recurrent fever,^{58 - 60} chronic fatigue syndrome,⁵¹ and, with more evidence, pityriasis rosea^{61 - 62} (Table 2). In pityriasis rosea, viral "footprints" were found, such as interferon alpha in the serum, and a cytopathic effect in cocultured mononuclear cells. This effect was not seen in controls or in the patients who were cured. It reappeared in the patient who relapsed. Various stages of virus morphogenesis were observed in the culture supernatant. Etiologic evidence of HHV-7 was supported also by HHV-7 DNA sequences isolated from PBMC, skin, and plasma of patients with acute pityriasis rosea and by their absence in the plasma of cured patients and in controls' skin.^{61 - 62} Particularly important is the detection of cell-free viral DNA in body fluids, which reflects viral replication and virulence and strongly supports a causative role.^{63 - 64} There have been no reports of HHV-7 DNA being detected in plasma of healthy people or of patients with other illnesses. Pityriasis rosea may be a manifestation of a late primary infection, but the large diffusion of HHV-7 through the general population, the low rate of human-to-human transmission, the rare relapses, and the occurrence in states of immunodeficiency all support the hypothesis of viral activation. Additional studies are needed to confirm the pathogenetic role of HHV-7 in pityriasis rosea and its possible correlation with other viruses.⁶⁵

Human herpesvirus–like DNA sequences have been identified in tissue samples from a patient with AIDS-associated Kaposi sarcoma (KS) using representational difference analysis,⁶⁶ and, subsequently, in more than 90% of tissue samples from such patients by using polymerase chain reaction (PCR).⁶⁶ Those sequences also have been found in most HIV-negative persons with KS, namely, African endemic KS, and in posttransplantation or iatrogenic and Mediterranean KS.^{67 - 74} Having a high degree of homology with EBV and Herpesvirus saimiri of the Gammaherpesvirinae subfamily,⁶⁶ they have been considered a new member of the subfamily and named HHV-8 or KS-associated HHV.

The infectious nature of HHV-8 has been confirmed by the isolation and propagation of a filterable agent from skin lesions of patients with AIDS-associated KS. This agent produced a cytopathic effect in cultured cells and revealed ultrastructural HHV-like features in their nuclei and cytoplasm.⁷⁵

Human herpesvirus 8, like other HHVs, is probably under immunological control. The virus has been found in PBMC of a healthy blood donor who did not belong to any risk group⁷⁶ and in PBMC of about 50% of patients with KS, but only occasionally is found in their plasma.^{70 ,77} In the PBMC of patients with KS, the HHV-8 detection rate seems to correlate with the reduced number of CD4⁺ T cells and with the stage of KS.^{77 - 78}

The virus seems to have a tropism for the flat endothelial cells lining the vascular spaces and the spindle cells of KS lesions,⁷⁹ in which it produces HHV cytopathic effects, and can be found as HHV nucleocapsids and virions.⁸⁰ Those cells, therefore, may be the site of latency or may harbor the virus as a mere passenger, although they also may be sites of active viral replication.⁷¹ That the urogenital tract, prostate, and sperm are sites of latent infection is highly controversial. Besides positive findings of viral DNA in prostate tissue and semen of healthy immunocompetent persons,⁸¹ stands a recent retraction⁸² of previous positive findings in most HIV-positive homosexual men and healthy immunocompetent persons.⁸³ The route of transmission is controversial as well.

Although seropositivity in children and the detection of the virus in nasal secretions and saliva⁸⁴ may suggest other mechanisms of transmission, there is enough evidence in favor of a predominant sexual transmission.⁸⁵ This evidence consists in 3 different observations. First, HIV-positive homosexuals have a higher seroprevalence than HIV-positive hemophiliacs and drug abusers.^{86 - 90} Second, patients attending sexually transmitted disease center clinics have a higher seroprevalence than general population.⁸⁹ Third, non–KS HIV-positive patients may have a normal rectal mucosa infected.⁹¹ In Africa, however, this may be not the case (Thomas Schultz, MD, oral communication, 1997).

The primary infection, if any, has not been identified. As for the other HHVs, the diseases so far related to it are more probably viral reactivations than primary infections (Table 2).

A precise estimate of HHV-8 seroprevalence is lacking.⁸⁵ Antibodies to lytic and latent HHV-8 antigens have been found in about 25% of adults (including blood donors).⁸⁶ In African people, a prevalence of 50% has been detected.^{86 - 88} A lower seroprevalence, however, ie, less than 5% of general population, has been found in Britain, North America, and Italy.^{87 - 88} In children, the 2% to 8% North American prevalence has been confirmed by semiquantitative PCR in a group of HIV-seronegative febrile infants in Zambia.⁹² Technical considerations may account for such contradictory findings. Nonspecific reactivity from lytic immunofluorescence assays⁸⁶ may lead to higher prevalences, while less sensitive reactivities to latent proteins may account for lower rates.^{87 - 89}

Other apparent contradictions emerge. Seroprevalences in different geographic areas do not correlate with KS development. Eighty-four percent to 100% of serum samples from the general population in non–KS-endemic African areas proved to be positive. Such a rate is higher than the one of endemic regions, and is the same as among patients with KS.⁹³

Vice versa, not all PCR-positive tissue samples are also seropositive, although both enzyme-linked immunosorbent assay and immunofluorescence assay are more sensitive indicators than PCR.

A more precise estimate of HHV-8 seroprevalence would require larger population-based surveys using more standardized assays and possibly including additional antigens.

Another controversial point is the significance to be given to antibodies to HHV-8 antigens. It is well known that antibodies directed to latent viral antigens are present in nearly all EBV-infected individuals. However, in several cohorts of patients with KS, the antibodies to antilatent HHV-8 antigens had a lower prevalence than antilytic antibodies.⁸⁶

Furthermore, despite the accumulating literature and some personal convinctions, whether HHV-8 plays a causal role in KS, is a coagent, or is a mere bystander is also still unclear.^{79 ,85} Besides the aforementioned controversies on serologic findings, there have been no reports of linear forms of the viral genome, characteristic of viral replication, being detected in KS tissues. They have been detected in PBMC,^{94 - 95} however. In addition, a low copy number of HHV-8 DNA sequences has been found even in the clinically normal skin adjacent to the tumor⁶⁹ and in a normal skin sample from an organ transplant recipient.⁹⁶ This evidence has been later denied, however.⁹⁷

Also controversial is the hypothetical relationship of HHV-8 with other HHVs. Human herpesvirus 8 and EBV have been found in the same KS lesions⁹⁸ and, incidentally, in the same AIDS-related body cavity–based non-Hodgkin lymphomas,⁹⁹ which are an unusual subset of AIDS-related lymphomas growing as lymphomatous effusion mainly in the body cavities without an identifiable contiguous tumor mass.

All these observations are difficult to reconcile with a simple model of KS pathogenesis.^{100 - 101} Other factors could be considered, such as the possibility of multiple viral interference as cofactors and transactivators as in the case of other known HHVs.

In addition, HHV-8 DNA has been found in other proliferative skin lesions of both immunocompromised and immunocompetent patients.^{96 ,102 - 104} This was not confirmed by others, however.^{97 ,105} Noteworthy, on the other hand, is the HHV-8 DNA detection in benign proliferation of endothelial cells.^{102 - 104} This finding may mean that the presence of the virus is not enough to produce specific lesions or that the virus is a mere passenger even in KS lesions. In the latter, HHV-8 detection may be easier because its replication is favored by the overexpression of cytokines in the KS tissues.⁷⁹

In addition, HHV-8 has been identified in pemphigus lesions,¹⁰⁶ in sarcoid tissues,¹⁰⁷ in some blood cell lines,⁷⁷ in bone marrow dendritic cells of patients with multiple myeloma,¹⁰⁸ and in a variety of lymphoid disorders. The latter include multicentric Castleman disease¹⁰⁹ and primary effusion lymphomas.⁹⁹ Some authors go so far to state that only KS, primary effusion lymphomas, and multicentric Castleman disease are the human diseases truly associated with the virus.¹¹⁰

We would like to emphasize here that several techniques should be concomitantly used to clarify the role of HHV-8. They are virus isolation, immunohistochemical staining, serum PCR, and quantitative PCR to determine intratissue virus load and distinguish between latent and productive infection.¹⁰¹

Accepted for publication July 20, 1998.

Reprints: Francesco Drago, MD, Clinica Dermatologica dell'Università, Viale Benedetto XV, 7. 16132 Genova, Italy.