Portal de Periódicos da CAPES

Human herpesvirus 6 infection in hematopoietic stem cell transplant patients

2004; Wiley; Volume: 124; Issue: 4 Linguagem: Inglês

10.1046/j.1365-2141.2003.04788.x

1365-2141

Tetsushi Yoshikawa,

Viral-associated cancers and disorders

Human herpesvirus 6 (HHV-6), a member of the β-herpesvirinae subfamily, shares many biological properties with other members of this family, human herpesvirus 7 and cytomegalovirus. Based on virological analysis, there are two distinct variants (HHV-6A and HHV-6B). While HHV-6B is ubiquitous in the human population, causing exanthem subitum, data regarding HHV-6A remains ambiguous. HHV-6 remains latent in the body after primary infection, and reactivates in immunocompromised patients. HHV-6 infection occurs in nearly half of all stem cell transplant (SCT) recipients within 2–4 weeks following transplantation. Viral infection results in clinical symptoms, including fever, skin rash, pneumonia, bone marrow suppression, encephalitis and rejection. Diagnosis of active HHV-6 infection in SCT recipients can be difficult, as this virus is ubiquitous and persistent in the human body. Therefore, accurate diagnostic measures are required to determine whether the virus is the true pathogen responsible for a clinical event. Although a double-blinded control study will be necessary to elucidate reliable treatment protocols, two anti-viral drugs (ganciclovir and foscarnet) are recommended for the treatment of HHV-6-associated diseases in these patients. Human herpesvirus 6 was discovered as a novel human herpesvirus in the peripheral blood of patients with lymphoproliferative disorders and acquired immunodeficiency syndrome. While it was originally designated ‘human B-lymphotropic virus’ (Salahuddin et al, 1986), subsequent studies established that the virus was primarily T-cell lymphotropic, causing it to be classified as HHV-6 (Lusso et al, 1988). Two genetically distinct variants of the virus exist, HHV-6A and HHV-6B (Schirmer et al, 1991). In most children, primary HHV-6B infection causes exanthem subitum (Yamanishi et al, 1988), a common febrile illness in infants that occurs between 6 months and 1 year of age (Yoshikawa et al, 1989). Several severe complications, including encephalitis/encephalopathy (Asano et al, 1992; Yoshikawa et al, 1992a; Suga et al, 1993), hepatitis (Asano et al, 1990; Tajiri et al, 1997; Ishikawa et al, 2002), and myocarditis (Yoshikawa et al, 2001a), can occur in rare cases following primary HHV-6 infection. This virus probably remains latent in the body after the primary infection, and then reactivates upon host immunosuppression in a manner similar to other human herpesviruses. All previously identified human herpesviruses, especially human cytomegalovirus (HCMV), are important contributors to morbidity and mortality in immunosuppressed organ transplant recipients. As HHV-6 shares many characteristics with HCMV, including DNA sequence homology, elements of genomic organization, antigenic cross-reactivity, and in vitro growth characteristics, HHV-6 may behave similarly as a complication of organ transplantation. Although the pathogenesis of HHV-6 is not well understood, recent evidence suggests that the virus is a serious, potentially life-threatening pathogen in the post-transplant period. This review summarizes the biology of HHV-6 and the clinical features of primary HHV-6 infection. With a focus on HHV-6 infection in haematopoietic SCT, we evaluate the clinical features, route of HHV-6 transmission, diagnostic methods, and important treatments for the management of patients with HHV-6. Human herpesvirus 6 and HHV-7 are recently discovered herpesviruses (Salahuddin et al, 1986; Frenkel et al, 1990), contained within the Roseolovirus genus of the β-herpesvirinae subfamily. HHV-6 and HHV-7 share many properties with ‘HCMV’, another β-herpesvirinae subfamily member, including virion structure, high prevalence in natural host, and the ability to establish latent infection in mononuclear cells and salivary glands. The lytic replication cycle in these viruses is generally longer than in other herpesvirinae subfamilies. Human herpesvirus 6 is an enveloped virion with an icosahedral nucleocapsid of 162 capsomers, enclosing a double-stranded DNA genome. Enveloped extracellular virions range in size from 160 to 200 nm in diameter. Under some fixation conditions, thin section electron microscopy reveals that HHV-6 teguments are smooth; these structures fill the space between the nucleocapsid and envelope, which is characteristic of this virus and is not present in other herpesviruses (Biberfeld et al, 1987). Genomic architecture is similar among the roseoloviruses; genomes contain a central unique segment (U) flanked by a pair of direct repeat structures, DRL and DRR (Lindquester & Pellett, 1991). In comparison with other herpesviruses, roseoloviruses have a low G + C content (35–40%). The HHV-6 genome also contains mammalian telomere-like sequences [(TAACCC)n] at the terminal and junctional DR segments (Gompels & Macaulay, 1995), the function of which is not known. These sequences do not appear to function in either nuclear retention of viral DNA, efficiency of DNA replication, or packaging-associated DNA cleavage (Deng & Dewhurst, 1989). Telomere-like sequences mediate the site-specific integration of the HHV-6 genome, observed proximal to the host chromosome telomere (Luppi et al, 1993; Daibata et al, 1998). Among the roseoloviruses, HHV-6 is closely related to HHV-7, with amino acid sequence identities throughout the genome ranging from 22 to 75% (Nicholas, 1996; Dominguez et al, 1999; Isegawa et al, 1999). HHV-6 has five characteristic genes (DR3, U6, U22, U83 and U94) that distinguish it from HHV-7. HHV-6A and HHV-6B are quite similar, with an overall nucleotide sequence identity of 90% (Dominguez et al, 1999; Isegawa et al, 1999). Variation of the DNA sequences between these strains is less than that between many other herpesvirus strains. Conservation can be even higher at the amino acid level; U94 amino acid sequence is identical among 13 independent HHV-6B isolates (Rapp et al, 2000). The U94 gene is a homologue of the parvovirus rep gene, suggested to play a role in site-specific integration. U94 gene has been suggested as a latency-associated transcript that may block the transcription of other viral genes (Rotola et al, 1998). As the integration of the HHV-6 genome into the human genome has been reported by several institutes (Luppi et al, 1993; Daibata et al, 1998), additional studies will be required to elucidate the function of the U94 gene in viral genome integration. The cellular receptor for HHV-6 entry is CD46 (Santoro et al, 1999), a member of the regulator of complement fixation family. This protein is expressed on the surface of all nucleated human cells. A population of T-cell lines expressing CD46 is unable to support replication of HHV-6A, suggesting the existence of a co-receptor. Several recent reports have demonstrated that binding to both CD46 receptor and the glycoprotein H and L complex are important for entry of the virus into host cells (Mori et al, 2003a,b; Santoro et al, 2003). The HHV-6 replicates within activated CD4 T lymphocytes in vivo (Takahashi et al, 1989; Lusso et al, 1991). HHV-6 can also infect CD8 T lymphocytes and natural killer cells, resulting in the induction of surface CD4 expression and subsequent susceptibility to human immunodeficiency virus (HIV) infection (Lusso et al, 1991, 1993, 1995). CD4 T lymphocytes can be infected concurrently with HHV-6 and HIV. As demonstrated by in vitro infection experiments, HHV-6 can also infect macrophages, dendritic cells, fibroblasts, epithelial cells and bone marrow progenitors (Robert et al, 1996; Kempf et al, 1997; Asada et al, 1999; Luppi et al, 1999; Kakimoto et al, 2002; Yoshikawa et al, 2003a). The results of several in vitro HHV-6 infection experiments have suggested that possible mechanisms for causing HHV-6-associated clinical events include bone marrow suppression (Knox & Carrigan, 1992; Isomura et al, 1997), a skin rash resembling acute graft versus host disease (GVHD) (Yoshikawa et al, 2002a), immunosuppression (Lusso et al, 1993; Gobbi et al, 1999) and encephalitis (He et al, 1996). Latency and reactivation are common biological characteristics of all herpesviruses. After cessation of the primary infection with HHV-6, the viral genome persists within peripheral blood mononuclear cells (PBMCs), particularly adherent cells (monocytes/macrophages) (Kondo et al, 1991). The virus also persists in the salivary glands; viral DNA can frequently be detected in saliva through the use of polymerase chain reaction (PCR) (Cone et al, 1993a). As HHV-6 DNA can also be detected in cerebrospinal fluid (Yoshikawa et al, 1992a; Kondo et al, 1993; Suga et al, 1993) and brain tissue (Luppi et al, 1995; Cuomo et al, 2001; Cermelli et al, 2003), the central nervous system has been postulated as a site of HHV-6 latency. Unfortunately, no reliable animal model is available for the study of HHV-6 reactivation. HHV-6B reactivation from latently infected PBMCs can be induced by infection with HHV-7 in vitro (Katsafanas et al, 1996; Tanaka-Taya et al, 2000). These results support in vivo evidence that demonstrated HHV-6 reactivation at the time of HHV-7 infection (Tanaka-Taya et al, 1994). Latency-associated HHV-6 transcripts encoding open reading flames of the immediate early (IE) proteins IE1 and IE2 have recently been identified (Kondo et al, 2002, 2003). The molecular mechanisms governing HHV-6 latency and reactivation remain poorly understood. Infection with HHV-6B is distributed throughout the world with no seasonal predilection and generally occurs within the first year of life. In contrast to HCMV, most SCT recipients and donors, including children, are generally sero-positive for HHV-6B (Yoshikawa et al, 1991; Wang et al, 1996). The HCMV sero-status of the bone marrow donor and the recipient correlated with the risk of post-transplant HCMV infection. Such factors do not appear to be involved in the frequency of post-transplant HHV-6 infections. The epidemiology of HHV-6A, however, remains unclear. Most infections in either immunocompetent infants or immunocompromised patients are caused by HHV-6B (Drobyski et al, 1993a; Frenkel et al, 1994). HHV-6A has been isolated from immunocompromised patients and patients with neurological complications (Hall et al, 1998). Primary infection with HHV-6B causes exanthem subitum, a common febrile infant disease (Yamanishi et al, 1988). In Japan, 70–80% of infants with primary HHV-6 infection exhibit a typical clinical course of exanthem subitum (Asano et al, 1994). Other groups, however, have reported that only 17% of American children with primary HHV-6 infections develop exanthem subitum; the majority develop undefined febrile illness (Pruksananonda et al, 1992; Hall et al, 1994). The reason underlying this discrepancy remains unclear. Ethnic background may be associated with the typical clinical features in this disease. The clinical course of primary HHV-6B infection is generally benign and self-limiting (Asano et al, 1994). Several severe complications of primary viral infection have been reported, however, including encephalitis/encephalopathy (Asano et al, 1992; Yoshikawa et al, 1992a; Suga et al, 1993), hepatitis (Asano et al, 1990; Tajiri et al, 1997; Ishikawa et al, 2002), thrombocytopenia (Yoshikawa et al, 1993a), hemophagocytic syndrome (Huang et al, 1990), and myocarditis (Yoshikawa et al, 2001a). The clinical features of primary HHV-6A infection remain poorly defined. Most infections in transplant recipients are caused by HHV-6B (Drobyski et al, 1993a; Frenkel et al, 1994), but fatal cases exhibiting primary HHV-6A infection after liver transplantation have been reported (Rossi et al, 2001). There are numerous reports linking additional conditions with HHV-6 infection. As HHV-6 is a ubiquitous virus that persistently infects PBMCs throughout life, it is difficult to determine whether this virus is the primary cause of the clinical feature. The most common procedure for the diagnosis of viral infection is PCR. If clinical samples for PCR examination contain PBMCs harbouring HHV-6 DNA, a false positive PCR test result may occur, detecting only the quiescent viral DNA. Therefore, it is important to choose appropriate samples for PCR examination that do not contain PBMCs, such as serum or cerebrospinal fluid supernatant. The combination of PCR with additional examination methods, including viral isolation and serological assays, are also necessary for the accurate diagnosis of active HHV-6 infection. HHV-6 reactivation occurs in patients with drug-induced hypersensitivity syndrome that exhibit infectious mononucleosis-like symptoms (Suzuki et al, 1998; Tohyama et al, 1998). Correlation between this disease and HHV-6 infection appears to be conclusive, as the virus could be isolated from the patients. Despite the detection of HHV-6 antigen in tissues collected from patients with both multiple sclerosis (Carrigan et al, 1996; Goodman et al, 2003) and pityriasis rosea (Watanabe et al, 2002), further investigation is necessary to demonstrate a link between viral infection and these diseases. The clinical features of HHV-6 infection after SCT suggested by recent studies are summarized in Table I. Biological and molecular analyses indicate that HHV-6 is similar to HCMV. Although the exact frequency of reactivation is difficult to determine, approximately 40–50% of SCT recipients develop HHV-6 infection. The incidence appears to be dependent on the sensitivity of the diagnostic procedures for viral infection. Our recent prospective study on viral isolation analysis determined that about 40% of SCT recipients exhibited HHV-6 viraemia 2–4 weeks after SCT (Yoshikawa et al, 1991, 2002b), which was confirmed by additional studies (Maeda et al, 1999; Imbert-Marcille et al, 2000; Ljungman et al, 2000). Although the frequency of HHV-6 infection after SCT, when diagnosed by PCR, is similar to that after solid organ transplantation, the frequency of HHV-6 viraemia (Yoshikawa et al, 1991, 2002b) is likely to be higher after SCT than after solid organ transplantation (Yoshikawa et al, 1992b, 2000). The risk factors contributing to HHV-6 infection in SCT recipients are not fully understood (Table II). Our recent study demonstrated that the incidence of HHV-6 viraemia is significantly higher among allogeneic SCT recipients than in autologous SCT recipients (P = 0·011). A multivariate analysis of allogeneic SCT recipients revealed that underlying disease (leukaemia or lymphoma) is an independent risk factor (P = 0·02) (Yoshikawa et al, 2002b). Furthermore, the frequency of HHV-6 infection is higher in recipients from unrelated donors than in related donors (Ljungman et al, 2000). In addition, allogeneic bone marrow transplant recipients had a higher frequency of HHV-6 viraemia than allogeneic peripheral blood SCT recipients (Maeda et al, 1999). Sashihara et al (2002) reported that the incidence of HHV-6 infection after cord blood SCT was significantly higher than after either SCT or peripheral blood SCT (P < 0·05). Anti-CD3 monoclonal antibody treatment administered as prophylaxis for acute GVHD also increased the risk of both HHV-6 infection (odds ratio, 2·5; 95% confidence interval, 1·3–4·7) and encephalitis (Zerr et al, 2001). Therefore, SCT recipients with these risk factors should be carefully monitored following the procedure by reliable methods (e.g. real-time PCR). Elucidation of the risk factors for HHV-6 infection is important in the proper management of SCT patients. Several factors, including conditioning regimens, may result in the induction of HHV-6 infection. A multivariate analysis of a large number of cases will be necessary to accurately determine the risk factors for HHV-6 reactivation. Although numerous clinical syndromes have been postulated as HHV-6-associated diseases (Table I), many of these associations have been inconclusive. Therefore, if the physician observes patients with these complications, accurate diagnostic measures, described later in the ‘diagnosis’ section, should be performed to confirm active HHV-6 infection. Four major clinical events, including interstitial pneumonitis, a skin rash resembling acute GVHD, encephalitis and bone marrow suppression, are major prognostic indicators of potential HHV-6 infection and are reviewed in this section. Details of other clinical manifestations including thrombotic microangiopathy (Matsuda et al, 1999) and enteritis (Amo et al, 2003) should be studied from the original papers. Carrigan et al (1991) first reported the association of severe interstitial pneumonitis with HHV-6 infection in two marrow transplant recipients (one autologous SCT and one allogeneic SCT). The virus was repeatedly detected in respiratory specimens from one patient. HHV-6 infected cells were observed in lung tissue from both patients by immunohistochemical staining. Subsequently, Cone et al (1993b) performed a retrospective study of 15 cases of post-SCT pneumonia, 15 accidental deaths and six fetuses. The authors evaluated the quantity of viral DNA in lung tissues collected from each subject by semi-quantifiable PCR analysis, and showed that SCT recipients with idiopathic pneumonia had relatively high levels of HHV-6 DNA in their lung tissue. In contrast to two previous studies evaluating adult patients, our study examining paediatric patients who underwent allogeneic SCT could not identify a statistical correlation between HHV-6 infection and pneumonia (Yoshikawa et al, 2002b). In conclusion, although some evidence shows that HHV-6 infection plays a role in pneumonia following SCT, the association remains equivocal, particularly in paediatric patients. Human herpesvirus 6 has also been isolated from the blood of SCT recipients 15 d after transplant. Two of the three patients, subsequently treated for acute GVHD, had fever and macular rash at the time of virus isolation (Asano et al, 1991). To further explore an association between HHV-6 infection and acute GVHD, we examined peripheral blood specimens from 25 SCT recipients for the presence of virus by serological analysis (Yoshikawa et al, 1991). HHV-6 infection was confirmed in 12 (48%) of the 25 recipients. Four of the 12 developed skin rashes, with three of those four also undergoing a febrile episode during the period of viral isolation; none of the remaining 13 patients manifested these symptoms. These results suggest that HHV-6 infection may recur in almost half of all SCT recipients at approximately 2–3 weeks following the procedure. Viral reactivation may play an important role in acute GVHD or an acute GVHD-like illness. To clarify the association between HHV-6 and this skin rash, two institutes have attempted to detect the virus genome within skin tissues obtained from recipients. These studies suggest that HHV-6 plays an important role in the development of acute GVHD (Wilborn et al, 1994; Boutolleau et al, 2003) and the development of skin rash after allogeneic SCT (Appleton et al, 1995; Le Cleach et al, 1998; Cone et al, 1999). In a recent study, HHV-6 viraemia was found in nine of 15 (60%) cases with a skin rash that occurred within 1 month of SCT, in comparison with none of 10 (0%) cases exhibiting a skin rash occurring more than 1 month after transplantation (P = 0·008) (Yoshikawa et al, 2001d). HHV-6 may therefore be involved in the development of skin rashes that occur within the first month after allogeneic SCT. The question remains, however, whether HHV-6 causes acute GVHD or the virus causes an erythematous illness similar to acute GVHD. In addition, it is possible that acute GVHD induces HHV-6 reactivation, which is not associated with skin eruptions. HHV-6 infection induces the upregulation of several surface molecules, including human leucocyte antigen (HLA)-ABC, HLA-DR and intracellular adhesion molecule 1 (ICAM-1), on epidermal cells (A431 cells) in vitro (Yoshikawa et al, 2003a). The alterations in surface molecule expression may promote the infiltration of inflammatory cells into epidermal tissues, perhaps causing the skin manifestations seen in SCT recipients. Further histopathological analysis is needed to clarify this issue. In contrast to SCT recipients, immunocompromised patients, such as infants with acute lymphoblastic leukaemia and liver transplant recipients, do not appear to manifest skin rashes during primary HHV-6 infection after the fever subsides (Yoshikawa et al, 1993b, 2001b). Encephalitis caused by HHV-6 has been well documented in immunocompetent children with primary HHV-6 infections. The viral genome has been detected in the cerebrospinal fluid of a subpopulation of patients (Asano et al, 1992; Yoshikawa et al, 1992a; Suga et al, 1993), suggesting primary encephalitis. In contrast, the HHV-6 viral genome could not be detected in cerebrospinal fluid collected from other patients exhibiting neurological symptoms in the exanthematous period, suggesting secondary encephalitis. Recent in vitro analysis demonstrated that latent infection and reactivation of HHV-6 alters the synthesis of inflammatory cytokines in astrocytoma cells (Yoshikawa et al, 2002a). Thus, it is conceivable that reactivation of the virus in immunocompromised patients may result in the development of neurological disorders. Several cases of encephalitis, including fatalities linked to HHV-6 infection, have been reported in SCT recipients (Drobyski et al, 1994; Mookerjee & Vogelsang, 1997; Bosi et al, 1998; Cole et al, 1998). Moreover, allogeneic SCT recipients treated with CD3-specific monoclonal antibodies were more likely to develop encephalitis than patients not receiving the treatment (Zerr et al, 2001). As there are no consistent abnormal findings upon examination of patients with central nervous system complications caused by HHV-6, the careful follow-up of high-risk patients is critical. There is no specific finding in either radiological or cerebrospinal fluid examination in recipients with HHV-6 encephalitis. However, the numbers of organ transplant recipients with HHV-6 encephalitis who also exhibited amnesia and abnormal radiological findings within the temporal lobe have been increasing recently (Bollen et al, 2001; Wainwright et al, 2001; MacLean & Douen, 2002; Montejo et al, 2002). Histological analysis of a fatal HHV-6 encephalitis patient with amnesia detected HHV-6 antigen in the hippocampus tissues (Drobyski et al, 1994). Moreover, we recently examined two SCT recipients with similar clinical features and radiological findings (unpublished data). These features suggest that, in addition to the onset of neurological symptoms (2–4 weeks after SCT), the presence of amnesia and abnormal radiological findings within the temporal lobe may be a useful diagnostic indicator of HHV-6 as an aetiological agent for encephalitis in SCT recipients. Although the prognosis in such cases is generally poor, rapid diagnosis followed by immediate treatment with ganciclovir or foscarnet may improve the prognosis of some patients (Cole et al, 1998). Detection of viral DNA in cerebrospinal fluid by means of PCR is also a useful technique for the rapid diagnosis of patients (Wang et al, 1999). In contrast to other clinical features, finding an association between encephalitis/encephalopathy in SCT recipients and HHV-6 infection is likely to be conclusive. As in infants with primary HHV-6 infection (Yoshikawa et al, 1993a; Hashimoto et al, 2002), idiopathic thrombocytopenia has been documented in transplant recipients (Singh et al, 1995; Maeda et al, 1999; Ljungman et al, 2000). In addition to thrombocytopenia, myeloid cell suppression, such as delayed engraftment and secondary graft failure, also occurs in such patients (Drobyski et al, 1993b; Carrigan & Knox, 1994; Rosenfeld et al, 1995; Wang et al, 1996; Singh et al, 1997; Johnston et al, 1999; Boutolleau et al, 2003). However, several groups have suggested that there is no association between HHV-6 and bone marrow suppression in SCT recipients (Kadakia et al, 1996; Yoshikawa et al, 2002b). Two possible mechanisms have been postulated for bone marrow suppression caused by HHV-6 infection; an indirect effect, mediated by cytokines and HHV-6 soluble products, and direct damage by the infection of bone marrow progenitors (Isomura et al, 1997). As HHV-6 infection may be associated with severe clinical complications following SCT, it will be important to predict viral infection accurately and to clarify the route of viral transmission. Two likely sources of HHV-6 infection after SCT exist: reactivation in the recipient and infection acquired from the seropositive donor bone marrow. Therefore, the latently infected PBMCs of donors and recipients may be an important source of viral transmission. The presence of the HHV-6 genome in either the donor or recipient PBMCs prior to SCT is a valuable predictor of viral infection following the procedure (Yoshikawa et al, 1998). Moreover, restriction endonuclease analysis was used to analyse three HHV-6 strains isolated from a patient with leukaemia both before and after SCT. The analysis suggested that latently infected HHV-6 reactivates in the recipient. Possible mutation or superinfection of the virus is thus likely to occur in an immunocompromised patient (Yoshikawa et al, 1992c). In general, as most recipients and donors are seropositive for HHV-6, reactivation from recipients and re-infection via donor marrow may both occur. In a manner similar to HCMV, the HHV-6 viral genome persists in PBMCs throughout life after primary infection. Detection of viral genome by means of PCR in clinical samples containing PBMCs may detect both latent and active viral infection. The laboratory tools currently available for the detection of HHV-6 infection are summarized in Table II. Although isolation of the infectious virus remains the best procedure for detecting active viral infection, such a process is labour intensive and requires 1–2 weeks for completion. As the rapid diagnosis of viral infection is important to establish a treatment regime, the PCR assay is a valuable tool. As latent HHV-6 infection probably occurs commonly throughout the general population, the use of PCR to detect HHV-6 DNA in blood cells or tissues has limited value in diagnosing active HHV-6 infection. Two strategies have been introduced to solve the problem, plasma PCR (Secchiero et al, 1995; Allen et al, 2001; Ihira et al, 2002) and quantifiable PCR. Recipients with HHV-6 viraemia had higher HHV-6 DNA levels than those without HHV-6 viraemia by quantifiable PCR analysis (Ljungman et al, 2000). Quantifiable PCR utilizing a Taq-Man system has been demonstrated to be useful for monitoring active HHV-6 infection in organ transplant recipients (Locatelli et al, 2000; Tanaka et al, 2000; Gautheret-Dejean et al, 2002). It is difficult, however, to determine threshold levels that distinguish between active and latent viral infection. Monitoring of viral load by means of real-time PCR, which can indirectly demonstrate the kinetics of viral replication, will provide important information for the management of these patients. As real-time PCR is generally available in many laboratories, this is the best method for monitoring active viral infection. In addition to DNA PCR, reverse transcription PCR is another candidate for a diagnostic method for detecting active HHV-6 infection (Norton et al, 1999; Van den Bosch et al, 2001; Yoshikawa et al, 2003b). Further clinical studies are needed to determine whether reverse transcription PCR is a reliable method for monitoring active viral infection in SCT recipients. As SCT recipients have an impaired immunological response, the reliability of a serological assay is considered to be insufficient in these patients. However, all the paediatric SCT recipients with HHV-6 viraemia showed a significant increase in HHV-6 antibody titres (indirect immunofluorescent assay) in our recent analysis (unpublished observations). Therefore serological assays may also be useful for the diagnosis of active viral infection in limited subjects. However, cross-reactivity between HHV-6 and HHV-7 antibodies have proved to be problematic (Wyatt et al, 1991; Yoshikawa et al, 2001c). In addition, an interaction between these viruses has been postulated. Although indirect immunofluorescence is commonly used to determine antibody titres against these viruses, the inability of this assay to distinguish between HHV-6 and HHV-7 antibodies makes this assay unreliable. Recently, we demonstrated that a neutralization test, in conjunction with an immunoblot assay that detects antibodies against an HHV-6B-specific major immunogenic protein, is a sufficient specific serological assay for the diagnosis of active HHV-6 infection (Yoshikawa et al, 2001c). Several additional assays for the detection of variant-specific antibodies have been reported, but a convenient laboratory method is not yet available. Immunohistochemical stains detecting HHV-6 in formalin-fixed, paraffin-embedded tissues are also available. Such stains, performed using a murine monoclonal antibody reactive against the HHV-6B p101 structural protein, detect cells productively infected with HHV-6. Such an antigenaemia assay is a reliable tool for monitoring active HCMV infection in SCT recipients. As seen for HCMV, one institute has successfully used the HHV-6 antigenaemia assay to monitor active HHV-6 infection in liver transplant recipients (Lautenschlager et al, 2000). As the number of subjects was too small to accurately determine the reliability of the methods, a large number of cases should be analysed in future studies. In addition to detecting active HHV-6 infection, another critical point for the accurate diagnosis of HHV-6-associated disease is to determine whether HHV-6 is responsible for the symptoms, as it is necessary to rule out other possible causative agents. General considerations for the prevention and treatment of HHV-6 infection are summarized in Table III. Ganciclovir, phosphonoformate (foscarnet), and cidofovir are effective against HHV-6 and HHV-7, as seen in in vitro analyses. Acyclovir and other thymidine kinase-dependent drugs, however, are only marginally effective (Agut et al, 1989; Burns & Sandford, 1990; Yoshida et al, 1998). The HHV-6 U69 gene product, the presumed functional homologue of HCMV UL97, phosphorylates ganciclovir (Ansari & Emery, 1999; De Bolle et al, 2002). A HHV-6 mutant exhibiting decreased sensitivity to ganciclovir in vitro was isolated after serial passages in the medium containing the drug; this mutant was later found to encode mutations in the U69 gene (Manichanh et al, 2001). Although ganciclovir-resistant HHV-6 has not been isolated to date in vivo, surveillance of emerging ganciclovir-resistant virus is also important for managing SCT recipients. Although several case reports suggesting antiviral effects of ganciclovir and foscarnet in transplant recipients with severe HHV-6 related diseases have been published (Mookerjee & Vogelsang, 1997; Rieux et al, 1998; Kawano et al, 2000; Bollen et al, 2001; Carvajal et al, 2001; Chik et al, 2002), no controlled trials assessing the efficacy of antiviral treatment against HHV-6 infection has been performed. Singh and Paterson (2000) reviewed cases of HHV-6 encephalitis following bone marrow or solid organ transplants, and found that seven of eight patients who were cured received ganciclovir or foscarnet, in comparison with none of four patients who did not receive the treatments. Zerr et al (2002) also carried out a retrospective analysis to determine whether ganciclovir and/or foscarnet treatment is effective for the treatment of HHV-6 encephalitis following SCT. They found eight SCT recipients with HHV-6-associated central nervous system manifestations; viral load in their cerebrospinal fluid decreased concurrent with antiviral treatment. Ganciclovir and foscarnet, however, may cause bone marrow suppression and renal toxicity respectively. Therefore, physicians must consider the patient's general condition prior to choosing an antiviral drug. Although in vitro studies suggest that HHV-6 is less sensitive to acyclovir, Wang et al (1996) have suggested the efficacy of high dose acyclovir prophylaxis for the prevention of HHV-6 infection. As acyclovir is less toxic than either ganciclovir or foscarnet, the efficacy of the drug as a treatment for HHV-6 infection should be evaluated in a double-blinded control study. It has recently been reported that prophylactic ganciclovir administration may reduce HHV-6 load in SCT recipients (Rapaport et al, 2002; Tokimasa et al, 2002). In contrast to acyclovir, ganciclovir can have severe side-effects; thus, prophylactic administration of the drug is not recommended for all SCT recipients. If patients at high risk for HHV-6-associated manifestations, such as encephalitis or bone marrow suppression, could be identified, prophylactic treatment with ganciclovir may be a good strategy for the management of disease in recipients. Furthermore, an appropriate treatment protocol (dose and timing of ganciclovir administration) should be carefully tested to establish a reliable prophylactic treatment with minimal side-effects. Human herpesvirus 6 can be a serious pathogen in SCT recipients. Although an association between some post-transplant clinical complications and HHV-6 infection appears to be conclusive, definitive proof of this association is still under debate. To clarify the pathophysiology behind this phenomenon, however, prompt recognition of the disease spectrum associated with the viral infection is critical. Future clinical virological studies, in combination with basic research, will provide valuable knowledge that can be used to improve the prognosis of transplant recipients.