Human Herpesvirus-6 in Transplantation: An Emerging Pathogen

Methods

We identified studies on human herpesvirus-6 infection in the English-language literature published from 1986 (when human herpesvirus-6 was discovered) to the present through a MEDLINE search using the keyword “human herpesvirus-6.” Additional studies were identified by reviewing the bibliographies of original articles, review articles, and textbooks. Reports on human herpesvirus-6 infection in bone marrow transplant recipients and in solid organ transplant recipients were selected for review. Data from studies of human herpesvirus-6 infection in nontransplantation settings were also selected if they were pertinent to the epidemiology, pathogenesis, detection, or treatment of human herpesvirus-6 infection after transplantation.

Virology

Seroepidemiologic studies show that human herpesvirus-6 infection is endemic in humans and is commonly acquired in early childhood. Seroconversion occurs by 2 years of age, and seroprevalence in the healthy adult population exceeds 90% [24]. Like other herpesviruses, human herpesvirus-6 can persist in the host in a latent form after primary infection [3]. Although the precise site of latency after primary human herpesvirus-6 infection is unknown, detection of DNA sequences specific to human herpesvirus-6 have determined that the virus persists in a latent or chronic persistent state in the oropharynx, in the epithelia of the bronchial and salivary glands, and possibly in human monocytes and macrophages [25].

On the basis of genomic DNA sequences, cell tropism, and protein expression, two distinct variants of human herpesvirus-6 have been described; the variants are designated as variant A and variant B. Most human herpesvirus-6 virions isolated from bone marrow transplant recipients and immunologically intact children have been variant B [13, 26]. Variant A infections in transplant recipients have rarely been reported [27]. Although the natural history of variant A remains to be clarified, such infections are believed to be acquired later in life.

Variant A is more cytolytic and has a broader host cell range than does variant B [27]. Compared with equivalent doses of variant B, human herpesvirus-6 variant A has consistently shown greater virulence with respect to suppression of growth factor-induced maturation of marrow precursor cells and colony formation by marrow precursors [27]. A recent study showed that human herpesvirus-6 variant A has an intrinsically greater virulence than variant B. Variant B suppressed in vitro colony-forming units in 94% to 98% of the assays; variant A suppressed these units in 43% to 86% of assays [28].

The primary target cells of human herpesvirus-6 are CD4+ T lymphocytes; this characteristic is shared with HIV and separates human herpesvirus-6 from other herpesviruses (such as cytomegalovirus, herpes simplex virus, and Epstein-Barr virus). Human herpesvirus-6 preferentially replicates in CD4+ T cells, causing cytopathic effects and cell death [29]. Although human herpesvirus-6 most efficiently replicates in CD4+ T cells, the cellular host range of human herpesvirus-6 is wide and includes CD8+ T cells, macrophages, natural killer cells, megakaryocytes, and possibly epithelial cells [9]. In addition to directly infecting cells, human herpesvirus-6 is a powerful inducer of cytokines (Table 1). It has been proposed that immunomodulatory activities of human herpesvirus-6 may be secondary to the induction of interleukin-1β, tumor necrosis factor-α, and interferon-α [30-32].

Infection of peripheral blood mononuclear cells by human herpesvirus-6 results in suppression of T-lymphocyte function, which is shown by reduced interleukin-2 synthesis and cellular proliferation [30, 33]. Messenger RNA (mRNA) analysis by reverse transcriptase polymerase chain reaction (PCR) shows that interleukin 2-mRNA levels in mitogen that have been diminished by human herpesvirus-6 stimulate peripheral T cells [30]. Human herpesvirus-6 can compromise the host defenses in the same manner in which T cells are depleted in patients with AIDS. Because human herpesvirus-6 primarily affects the same cell line that HIV does—CD4+ T cells—it has been proposed that human herpesvirus-6 synergistically enhances immunodeficiency and accelerates the course of HIV infection [10, 11]. Profound destruction of the cellular immune system leading to fatal progressive immunodeficiency has also been documented with human herpesvirus-6 in non-HIV settings [34]. Thus, human herpesvirus-6 can be considered an emerging “immunotropic” herpesvirus that can directly infect or interfere with the function of several critical components of the immune system.

Human herpesvirus-6 infection may predispose the host to superinfections by other viruses [35, 36]. Human herpesvirus-6 infection of human lymphoid cell lines that were genome positive for Epstein-Barr virus resulted in a reactivation of Epstein-Barr virus in these cells [37]. A high frequency of co-infections—such as those with cytomegalovirus, respiratory syncytial virus, and adenovirus—has been found in patients with human herpesvirus-6 pneumonitis [38]. These findings are particularly relevant in transplant recipients; human herpesvirus-6 may play a role in facilitating Epstein-Barr viral and cytomegaloviral infection after transplantation.

Detection of Human Herpesvirus-6

Laboratory tools currently available for detecting human herpesvirus-6 infection include viral isolation [13, 39], PCR assays [14, 40, 41], and serologic studies [42]. As with cytomegalovirus, tissue-cell cultures can be used to isolate human herpesvirus-6 on the basis of the appearance of cytopathic effects specific to human herpesvirus-6. Isolation in cell culture, however, is labor intensive and requires 5 to 21 days for detection.

Because latent human herpesvirus-6 infection is believed to occur commonly in the general population, using PCR to detect human herpesvirus-6 DNA in blood cells or tissues has limited value in diagnosing active or productive human herpesvirus-6 infections [43]. Latently infected peripheral blood mononuclear cells, however, contain fewer than 10 human herpesvirus-6 genomes per 10⁶ cells. Semiquantitative PCR techniques that cannot detect the latent virus but can detect actively infected cells are being explored. A sensitive and specific PCR assay for detecting cell-free viral DNA (a marker for active human herpesvirus-6 infection) has recently been described [41]. Human herpesvirus-6 serologic titers, although useful for the detection of latent infection or seroprevalence, may not be a reliable indicator of human herpesvirus-6 reactivation. Increases in human herpesvirus-6 antibody titers have been reported with other herpesvirus infections [42].

A rapid shell vial (early antigen) assay that can detect active human herpesvirus-6 infection within 72 hours was recently developed. In this assay, peripheral blood mononuclear cells are inoculated on human diploid fibroblasts for 48 hours and are then stained with antiserum specific for the major immediate early antigen of human herpesvirus-6. In bone marrow or liver transplant recipients, a sensitivity of 86% and a specificity of 100% were shown with this assay [44].

Immunohistochemical stains for detecting human herpesvirus-6 in formalin-fixed paraffin-embedded tissues (Figure 1) are also available [13, 22, 45]. Immunohistochemical staining of tissues with murine monoclonal antibody reactive against the structural protein p101 of variant B [20] and structural protein gp82 [46] of variant A detects cells productively infected with human herpesvirus-6. (“Productively infected cells” are cells that are actively infected with human herpesvirus-6 as opposed to cells that harbor latent human herpesvirus-6.)

View larger version:

• In this window
• In a new window

Figure 1. Staining shows cells infected with human herpesvirus-6 ( ) in the marrow of a bone marrow transplant recipient with chronic bone marrow suppression. (Vector red reaction substrate with hematoxylin counterstain. Original magnification, × 200.). Immunohistochemical staining with rabbit antiserum specific for human herpesvirus-6.arrows

Epidemiology

Human herpesvirus-6 infection has been reported in 38% to 60% of bone marrow transplant recipients [13, 21, 26, 40]. Only limited data on human herpesvirus-6 are available for solid organ transplant recipients; human herpesvirus-6 infection has been documented in 38% to 55% of renal transplant recipients [16-18] and 31% of liver transplant recipients [47].

Most human herpesvirus-6 infections occur 2 to 4 weeks after transplantation. This characteristic timing of onset contrasts with the timing of onset of cytomegaloviral infection, which usually occurs 6 to 12 weeks after transplantation. In immunocompetent hosts (such as children with exanthem subitum), human herpesvirus-6 viremia is usually of short duration, averaging 4 days. Prolonged viremia (lasting as long as 37 days) has been reported in transplant recipients [26].

Although the source of the virus has not been precisely defined, it has been proposed that most infections are caused by reactivations of the latent virus in the recipient. This assumption is based largely on the fact that nearly all donors and recipients in the reported studies were seropositive for human herpesvirus-6 before transplantation [16, 18, 21]. Yoshikawa and colleagues [21], however, have provided direct evidence of reactivation: Two human herpesvirus-6 strains were isolated from the blood of a child with acute lymphocytic leukemia before and after bone marrow transplantation. Genomic analyses of both strains indicated reactivation of the virus, which had been harbored latently in the patient.

Convincing evidence suggests that donor transmission of human herpesvirus-6 also occurs. A report from Japan [18] documented simultaneous isolation of two human herpesvirus-6 strains from two renal transplant recipients who received grafts from the same cadaveric donor. Genomic analysis of both isolates showed the same DNA cleavage patterns suggesting virus transmission through the renal allograft. In vivo latency in the kidney has also been shown [18]. Human herpesvirus-6 was also isolated from the peripheral blood mononuclear cells of a liver transplant recipient 23 days after transplantation [48]. The patient was seronegative for human herpesvirus-6 before transplantation; donor transmission was believed to be the source of the patient's subsequent primary infection [48].

The role of rejection or augmented immunosuppression in facilitating human herpesvirus-6 infection is controversial. In a report of renal transplant recipients [16], rejection and increased immunosuppression were proposed to lead to a higher incidence of human herpesvirus-6 infection. Reactivation of human herpesvirus-6 infection after OKT3 monoclonal antibody treatment has been described in renal transplant recipients. Other reports have found no correlation between acute rejection or antirejection therapy and human herpesvirus-6 infection [18, 47].

Clinical Manifestations of Human Herpesvirus-6 Infection

Most studies delineating the pathogenicity of human herpesvirus-6 in patients receiving transplants have been done in bone marrow transplant recipients. Human herpesvirus-6 is an important cause of idiopathic marrow suppression after bone marrow transplantation [13, 39]. Human herpesvirus-6 was isolated in cultures taken from blood samples of 38% (6 of 16) of adult bone marrow transplant recipients [13]. Idiopathic marrow suppression occurred in 67% (4 of 6) of patients with concurrent human herpesvirus-6 viremia, compared with 10% (1 of 10) of patients without viremia (P < 0.05). An etiologic role for the virus is also supported by the isolation of human herpesvirus-6 from the bone marrow of all four patients with viremia at the time of marrow suppression and by in vitro colony-forming unit assays that showed the ability of human herpesvirus-6 to inhibit colony-forming unit-granulocyte macrophage and burst-forming unit-erythroid growth from human bone marrow [13]. In another study of bone marrow transplant recipients, bone marrow samples from patients with idiopathic marrow suppression were statistically significantly more likely to be positive for human herpesvirus-6 than were samples from patients with an identifiable cause for marrow suppression [39]. Febrile illness with profound thrombocytopenia after liver transplantation [19] and severe leukopenia after renal transplantation have also been reported to be associated with human herpesvirus-6 infection [17].

Leukocytes were the most common cell line suppressed (83%), followed by platelets (67%) and red blood cells (50%); depression of more than one bone marrow lineage may occur. The marrow-suppressive effect of human herpesvirus-6 ranges from a transient decrease in marrow function to chronic myelosuppression [49]. In bone marrow transplant recipients, the marrow suppression associated with human herpesvirus-6 variant B appeared to manifest most commonly as mild to moderately severe long-term suppression of marrow function [39]. In contrast, human herpesvirus-6 variant A infection was associated with an abrupt and dramatic appearance of aplastic anemia with an essentially empty bone marrow at biopsy [27]. These findings are consistent with greater in vitro virulence of variant A [28]. The marrow-suppressive effect of human herpesvirus-6 is believed to be at least partially mediated by cytokine- or virus-produced soluble factors [31, 32].

Human herpesvirus-6 is associated with interstitial pneumonitis after bone marrow transplantation. The first report describing this association reported two bone marrow transplant recipients with human herpesvirus-6 pneumonitis [22]. Immunohistochemical studies showed human herpesvirus-6 at autopsy in the lung sample of a 19-year-old autologous marrow transplant recipient who had died of idiopathic pneumonitis [22]. The second case involved a 32-year-old allogeneic bone marrow transplant recipient who had developed interstitial pneumonitis 2 weeks after transplantation. Human herpesvirus-6 was isolated from the patient's bronchoalveolar lavage fluid, and immunohistochemistry showed human herpesvirus-6 in the lung biopsy specimen. Cells bearing nuclear or cytoplasmic inclusions of cytomegalovirus were not seen, and immunohistochemical staining for cytomegalovirus immediate early and structural antigens had negative results [22]. Increased levels of human herpesvirus-6 DNA in the lung tissue were significantly associated with idiopathic pneumonitis in a study of 15 bone marrow transplant recipients [23]. High human herpesvirus-6 DNA levels were associated with 1) more severe graft-versus-host disease, probably caused by a reactivation of human herpesvirus-6 exacerbating graft-versus-host disease or the reverse; and 2) a greater likelihood of idiopathic pneumonitis, as opposed to pneumonias with an identifiable cause [23].

Because of the ubiquitous nature of herpesviruses, detection of human herpesvirus-6 in respiratory tract secretions or even immunohistochemically in the lung tissue may not necessarily imply the existence of human herpesvirus-6 pneumonitis, a situation analogous to cytomegalovirus infection in transplant recipients. However, researchers have proposed that the density of productively infected cells correlates with human herpesvirus-6 pneumonitis; having fewer than 500 infected cells/cm² indicated that human herpesvirus-6 was an unlikely cause of pneumonitis [50]. Direct-cell lysis and immune-mediated destruction of infected cells is believed to be the pathogenesis of lung injury induced by human herpesvirus-6 [50].

A notable feature of human herpesvirus-6 is its propensity for neuroinvasion. Encephalitis caused by human herpesvirus-6 has been well documented in a nontransplantation setting [8], and herpesvirus-6 has been detected in the cerebrospinal fluid in association with exanthem subitum in children [7, 51]. A case of fatal encephalitis caused by human herpesvirus-6 has been reported after bone marrow transplantation [20]. Postmortem examination of the brain in a marrow transplant recipient with meningoencephalitis of undetermined cause showed infection of the astrocytes with human herpesvirus-6 [20]. The presence of p101 structural human herpesvirus-6 protein in the astrocytes indicated actively replicating as opposed to latent human herpesvirus-6 in these cells because this protein is not produced in latently infected cells.

Skin rash may also accompany febrile illness induced by human herpesvirus-6, although detection of human herpesvirus-6 in skin biopsy samples—and therefore causal association between human herpesvirus-6 and skin rash—have not been shown in the transplantation setting. In a study from Japan involving 44 bone marrow transplant recipients [26], human herpesvirus-6 was isolated from peripheral blood mononuclear cells of 41% (18 of 44) of the patients; skin rash (with fever) was present in 33% of the infected patients. In another study [21], human herpesvirus-6 was isolated from peripheral blood or bone marrow mononuclear cells in 40% (10 of 25) of the marrow transplant patients; 33% of the patients with human herpesvirus-6 infection developed skin rashes compared with 0% of the patients without human herpesvirus-6 infection. Skin rash secondary to leukocytoclastic vasculitis has been reported in a liver transplant recipient who had disseminated invasive human herpesvirus-6 infection [19].

A striking feature of human herpesvirus-6 is its potential to cause extremely high fever. Temperatures of 41.6 degreesC in a renal transplant recipient [15] and 41 degreesC in a liver transplant recipient have been reported in association with human herpesvirus-6 [19]. Association of human herpesvirus-6 with high fever has also been reported in nontransplantation settings [52]. In a study evaluating consecutive children with a febrile illness presenting to emergency departments [52], the fever was statistically significantly higher in children with than in children without human herpesvirus-6 viremia.

Management

The antiviral susceptibilities of human herpesvirus-6 generally resemble those of cytomegalovirus: Human herpesvirus-6 is less sensitive to acyclovir but is susceptible to ganciclovir and foscarnet [5, 53]. Acyclovir has an IC₅₀ (the concentration of antiviral drug that reduces the number of fluorescence-stained cells by 50%) for human herpesvirus-6 of 100 µmol/L [5]. Such a concentration is not readily achievable with the usual doses of acyclovir. Both ganciclovir and foscarnet are effective in vitro against human herpesvirus-6 [54-56]. The IC₅₀ of human herpesvirus-6 for ganciclovir has ranged from 1.0 to 2.5 µmol/L; for foscarnet, the IC₅₀ has ranged from 49 to 67 µmol/L [54-56]. For 14 clinical isolates of human herpesvirus-6 variant B, the mean IC₅₀ ± 1 SD was 5.9 ± 3.0 µmol/L (range, 2.2 to 13.5 µmol/L) for ganciclovir and 67 µmol/L ± 22 µmol/L (range, 21 to 117 µmol/L) for foscarnet (Carrigan DR. Unpublished data). Other drugs with in vitro activity against human herpesvirus-6 are guanosine analogue 9-[4-hydroxy-2-(hydroxymethyl)butyl]guanine, ampligen, and kutapressin [57, 58].

Availability of methods for the rapid diagnosis of human herpesvirus-6 would allow for surveillance of patients and timely therapeutic interventions for human herpesvirus-6 infection. Isolation in cell culture takes 5 to 21 days (median, 11 days) and may not be routinely available because of its difficulty and expense. The rapid shell vial assay, serum PCR assay, and immunohistochemical stains for viral structural proteins p101 and gp82 can be done with a turnaround time of about 1 to 3 days. The two monoclonal antibodies capable of detecting variant A and variant B of human herpesvirus-6 in formalin-fixed paraffin-embedded tissues using standard immunohistochemical staining techniques are now commercially available. Any routine histology laboratory could implement immunohistochemical staining with these antibodies. Finally, immediate early monoclonal antibodies for rapid shell vial assay are also available.

We recommend the weekly assessment of blood samples using either the shell vial or serum PCR assay for the first 2 or 3 months after transplantation. When clinically indicated, samples of bone marrow, cerebrospinal fluid, bronchoalveolar lavage fluid, bronchoalveolar cytology, and tissue biopsy specimens should also be analyzed by appropriate diagnostic methods. We propose that the criteria in Table 2 be used to determine which patients should be treated for human herpesvirus-6 infection after transplantation. If the listed criteria are met, the coexistence of another infectious agent should not preclude treatment of human herpesvirus-6 infection, because of the possible interaction between the two pathogens and the ability of human herpesvirus-6 to enhance the virulence of other pathogens. Either ganciclovir or foscarnet may be used to treat human herpesvirus-6 infection depending on clinical circumstances, such as renal failure or bone marrow function. For example, ganciclovir may be preferable in patients with renal dysfunction because foscarnet is a potentially nephrotoxic drug. On the other hand, foscarnet may be preferable in patients with marrow suppression because it does not possess the myelosuppressive effect of ganciclovir. Moreover, ganciclovir and human herpesvirus-6 may synergistically worsen the cytokine-mediated marrow suppression associated with human herpesvirus-6 (Carrigan DR. Unpublished data).

Although the optimal approach toward, timing of initiation of, duration of, and efficacy of prophylaxis for human herpesvirus-6 infection after transplantation remain to be defined, the possibility of prophylaxis using intravenous ganciclovir, intravenous foscarnet, or oral ganciclovir exists. Intravenous ganciclovir could clearly achieve serum concentrations high enough to provide effective prophylaxis against human herpesvirus-6 reactivation, but the role of oral ganciclovir remains to be determined. Administration of 1000 mg of oral ganciclovir three times per day resulted in a maximum serum concentration of 1.2 µg/mL (4.7 µmol/L) (Product monograph, CYTOVENE, Roche Laboratories, Palo Alto, California). This level may not be adequate for all strains of human herpesvirus-6.

In summary, human herpesvirus-6 can be an opportunistic pathogen in transplant recipients. Between 40% and 50% of transplant recipients have human herpesvirus-6 infection; nearly half of these cases are associated with symptomatic disease, predominantly bone marrow suppression. Future research should focus on the delineation of the complete clinical spectrum of human herpesvirus-6, the sources of human herpesvirus-6 after transplantation (including the role of donor allograft), the relative prevalence and virulence of the two variants of human herpesvirus-6, and the immunologic sequelae of human herpesvirus-6 after transplantation, including its potential to facilitate superinfections by other pathogens. Such research would lead to rational strategies for prophylaxis of human herpesvirus-6 infection in transplant recipients.

Dr. Carrigan: Medical College of Wisconsin, Department of Pathology, 8700 West Wisconsin Avenue, Milwaukee, WI 53226.