Human Herpesvirus 6 Infection of the Central Nervous System: Is It Just a Case of Mistaken Association?

Human herpesvirus 6 (HHV-6) has been linked to a variety of CNS diseases, including meningitis, encephalitis, and, more recently, multiple sclerosis. Acute infection with HHV-6 results in roseola infantum, as first described in 1910 [1]. Since serological and virologic demonstration of the association between HHV-6 and roseola infantum by Yamanishi et al. [2], many investigators have attempted to associate HHV-6 infection with a variety of other clinical entities, including a mononucleosis syndrome [3, 4], hepatitis [5, 6], and febrile seizures [7], either alone or in combination with meningitis and encephalitis. The association between HHV-6 and CNS infection is, in large part, predicated on the recognition that seizures are a complication of primary infection [8]. Notably, seizures occur in as many as 13% of children who experience primary infection. This association between fever and seizures occurs in as many as one-third of children aged 12–15 months [9]. The literature is peppered with case reports of meningoencephalitis in children with roseola who have fever, depressed levels of consciousness, and seizures. Disease in many of these children progresses to CT-detectable lesions that are hypodense, multiple, and involve the cerebral cortex [10–12]. In each of these 3 studies, direct evidence of HHV-6 replication in the brain is lacking. In large part, the diagnosis was based on either serological assessment of these patients or the detection of viral DNA in the CSF by PCR, as is similarly reported in 2 studies by Isaacson et al. [13] and Birnbaum et al. [14] in this issue of Clinical Infectious Diseases. In these 2 articles, a total of 5 patients, one of whom was treated with ganciclovir, are described as having evidence of HHV-6 encephalitis. The evidence suggesting an association between HHV-6 and infection of the CNS is predicated on the detection of HHV-6 DNA in the CSF. The HHV-6 DNA load was defined for all 5 patients by means of real-time TaqMan-based PCR detection of viral DNA (Roche). In each circumstance, the viral copy numbers were significantly elevated. In the case reported by Birnbaum et al. [14], there was a significant decrease in viral load after the administration of ganciclovir. Several fundamental questions focus on the detection of HHV-6 DNA in the CSF of patients with altered mentation and findings of meningoencephalitis. First, and most importantly, it is well recognized that HHV-6 is latent in lymphocytes and, therefore, that if mononuclear cells are present in the CSF, false-positive qualitative PCR detection of HHV-6 DNA can occur. In the study by Birnbaum et al. [14], the relatively low lymphocyte count (although higher than normal) could not account for the high HHV-6 load detected in the CSF (>104 copies/mL for all patients). A more specific determination of the etiological relationship between HHV-6 and CNS disease could involve RT-PCR for detection of RNA of actively replicating virus. Such a laborious assay cannot be routinely employed in diagnostic virology laboratories. Second, what does the detection of HHV-6 in the CSF of a patient with encephalitis mean from a pathogenic standpoint? Historically, primary infection with HHV-6 has been associated with CNS complications, as noted above. In each of the 5 patients described in this issue [13, 14], only one was a young teenager; the other patients were adults. Two patients were seronegative for HHV-6 at the time of presentation. Two adult patients were seropositive for HHV-6; 1 of the 2 was seropositive shortly after admission to the hospital. A fifth patient was not tested. Thus, it appears as though at least 2 patients experienced primary infection as adults (age, 38 and 66 years). Notably, only an IgM assay was performed for the 66-year-old patient. Primary infections in adults are exceedingly uncommon; however, they have been reported. More importantly, the documentation of preexisting antibodies in 2 of the 5 patients implies that HHV-6 can be reactivated in the CNS of immunocompetent individuals. Part of the difficulty in studying the pathogenesis of HHV-6 infections of the CNS is associated with the observation that viral DNA can be detected in autopsy specimens of individuals who died of other causes. This has been most notably the case in the postmortem evaluation of individuals with multiple sclerosis, in whom plaques with evidence of HHV-6 DNA are common. The inability to directly isolate HHV-6 from brain tissue remains a confounding variable. Third, although there are antiviral drugs that inhibit the replication of HHV-6 in vitro, evaluation of such agents in clinical trials has not been rigorously tested. As a consequence, antiviral therapy for presumptive HHV-6 infection of the CNS is used on a case-by-case basis. Without more-definitive results from clinical trials, it is difficult to conclude that the administration of an antiviral drug (i.e., ganciclovir) results in clinical improvement. The probability that such trials will be performed is negligible, given the rarity of encephalitis. Our opinion is that HHV-6 infections of the CNS do exist. However, we believe that they are exceedingly uncommon. At the present time, the only available method for detecting these infections is by means of TaqMan quantitation of viral DNA in the CSF in identifying what would be presumably higher levels of viral DNA than would occur with atypical aseptic meningitis. The reports by Isaacson et al. [13] and Birnbaum et al. [14] define a need to further evaluate this syndrome and to identify diagnostic methods and treatment modalities that be of clinical value. Potential conflicts of interest. R.J.W. and F.D.L.: no conflicts.