Viral Infections of the Central Nervous System: An Update on Herpes Simplex Virus, Varicella-Zoster Virus, and Enterovirus D68

Viral infections of the central nervous system (CNS) vary in presentation and result in a number of clinical syndromes, including meningitis, encephalitis, acute flaccid myelitis (AFM), and chronic or progressive encephalopathies. An understanding of pathogen-specific epidemiology, clinical presentation, prognosis, and treatment options is important in improving outcomes. We present recent advances in 3 relatively common CNS viral infections: herpes simplex virus (HSV), varicella-zoster virus (VZV), and enterovirus D68 (EV-D68) (Table).

Updates in HSV

Herpes simplex viruses, including HSV-1 and HSV-2, are members of the Herpesviridae family, a group of double-stranded DNA viruses known to be pathogenic to humans. CNS infection by HSV-1 represents the most common sporadic infectious encephalitis and can occur in primary infection (30% of cases) or by reactivation following latency in sensory ganglia (70% of cases).¹ Well over half of adults worldwide are infected with latent HSV; a small percentage go on to develop HSV encephalitis, which has an incidence of 1 in 250,000 to 500,000. As discussed in the following, whereas acyclovir is the mainstay of treatment, corticosteroids have been evaluated as an adjunct. In addition, post-HSV autoimmune encephalitis (AE) can result in substantial morbidity, and recognition is essential.

Corticosteroids in Acute HSV Encephalitis

Promising results in early animal studies demonstrating decreased viral load and better outcomes when corticosteroids were used along with acyclovir led to human studies, which have not shown a clear benefit. The first multicenter, randomized, double-blind, placebo-controlled trial to test the effect of adjunct corticosteroids for HSV encephalitis suffered from methodologic issues and was only able to recruit 41 participants, with no difference in outcomes identified.² Moreover, although not limited to HSV encephalitis, a 2023 meta-analysis of 50 studies including 280 participants with viral encephalitis, 120 of whom were treated with steroids, did not show improved survival in the steroid-treated group.³ Data from the European DexEnceph study, a randomized, double-blinded study evaluating steroid use in HSV encephalitis that recruited >90 individuals, are eagerly awaited to further inform the debate regarding routine corticosteroid use in clinical practice. For the time being, there is no clear benefit to routinely adding corticosteroids in acute HSV encephalitis treatment, although it may be considered in the setting of substantial cerebral edema.

Post-HSV AE

Relapses of neurologic symptoms after HSV encephalitis have long been recognized, but only recently has it become clear that the large majority are driven by a postinfectious autoimmune process. Whereas the biologic mechanisms are incompletely understood, one attractive theory is that CNS damage incurred during the viral infection exposes cryptic antigens previously sequestered from the immune system, such as those associated with neuronal receptors, including the N-methyl-D-aspartate receptor (NMDAR). Antibodies develop against these newly exposed epitopes, resulting in an autoimmune-mediated attack on healthy CNS tissue.⁴

The presence of anti-NMDAR antibodies in individuals with HSV encephalitis was first observed in 2012,⁵ several years after the discovery of NMDAR antibodies as a mediator of AE. With time, further cohorts of individuals with post-HSV AE were studied, allowing for the epidemiology and clinical features of this phenomenon to be better characterized. Post-HSV AE occurs in up to 1 in 5 individuals. The median time after HSV infection to the onset of AE symptoms is ~1 month, with the start of autoimmune symptoms ranging from concomitantly with HSV treatment to about 2 to 3 months after the completion of treatment.⁶ There are differences between children and adults in the presentation of postinfectious AE. Young children are more likely to present with choreoathetosis; adolescents and adults are more likely to present with psychiatric manifestations.⁷

Almost no individuals with post-HSV AE demonstrate anti-NMDAR antibodies in the cerebrospinal fluid (CSF) at the time of HSV diagnosis, with time to detection of anti-NMDAR antibodies in the CSF ranging from 1 to 8 weeks after diagnosis. Up to one-third of individuals with post-HSV AE may have disease associated with other neuronal (non-NMDAR) autoantibodies.⁶ CSF PCR studies in these individuals are negative for HSV, further implicating autoimmunity rather than viral relapse as the mechanism for repeat neurologic symptoms. CSF studies in cases of post-HSV AE typically show pleocytosis and an elevation in CSF protein level, although the degree of pleocytosis may be decreased from initial HSV presentation.⁶ Treatment with immunomodulatory therapy, starting with high-dose corticosteroids and progressing to intravenous immunoglobulin (IVIg) and plasmapheresis, followed by rituximab if previous treatments are ineffective, is typically effective across age ranges.

Updates in VZV

VZV is another member of the Herpesviridae family that can infect the CNS, with diverse manifestations including vasculopathies, encephalitis, meningitis, and myelitis. Well over half of adults worldwide are latently infected with VZV; those who are immunocompromised because of comorbidities or increased age are at particular risk of experiencing VZV reactivation. CNS VZV infection is a challenging clinical diagnosis because of nonspecific clinical symptoms, particularly in vasculopathies and myelopathies, and the lack of a characteristic shingles rash in up to one-third of individuals.⁸ Here, we discuss recent advances in laboratory testing and vaccination.

CSF viral polymerase chain reaction (PCR) is the standard diagnostic method in many viral CNS infections. The sensitivity of viral PCR testing varies among CNS syndromes caused by VZV. VZV vasculopathy has a particularly low PCR sensitivity, with studies reporting sensitivities as low as 28%, although the test is highly specific.⁹ The presence of CSF VZV immunoglobulin G is reported to be highly sensitive, with a sensitivity of ~93%.¹⁰ In addition, although the reported sensitivity is low, at ~25%, CSF VZV immunoglobulin M may be considered in individuals with an acute onset of symptoms suspicious for CNS VZV who may not have yet developed immunoglobulin G antibodies.¹¹

Shingrix: The Changing Landscape of VZV

Since 1995, the inclusion of live attenuated varicella vaccines as part of the recommended childhood vaccine series in Western countries has prevented an estimated 91 million cases and >2000 deaths and has allowed adults previously infected with chickenpox to experience some degree of protection against reactivation.¹² The live attenuated vaccine, Zostavax, was introduced in 2006 in the United States and had ~65% efficacy for preventing shingles. Its recombinant successor, Shingrix, introduced in 2017, has a greatly improved efficacy of >80% at preventing shingles.¹³ However, despite >17 million people having received the Shingrix vaccine, few long-term data are available on the vaccine’s ability to prevent CNS complications, as clinical trials have focused on shingles and postherpetic neuralgia prevention. One survey-based study of 265,568 adults found a lower risk of stroke in the group that received the Shingrix vaccine, with unvaccinated adults having 1.71 odds of having a stroke when compared with age-adjusted vaccinated peers.¹⁴ Theoretically, some of this decreased risk may be attributable to decreases in VZV-associated vasculopathies, although data of this granularity were not available in the study. In addition, the very low incidence of clinically significant CNS VZV means that only a small percentage of stroke reduction could be possibly attributable to infection prevention. Overall, as more time passes since the vaccine’s approval, more data will become available on whether Shingrix is as effective at preventing CNS varicella reactivation as it is at preventing shingles.

Updates in EV-D68

Members of the Enterovirus genus are single-stranded RNA viruses that are widely diverse both in genotype and in tropism, with clinical manifestations ranging from respiratory illnesses caused by rhinoviruses to paralysis caused by poliovirus. EV-D68 was first detected in the 1960s, when it was thought to be a purely respiratory virus. Over time, evidence has emerged that EV-D68 infection is associated with an increased incidence of AFM. We review the evidence supporting a causal relationship between EV-D68 infection and AFM and discuss management of the condition.

EV-D68: AFM

AFM is a condition characterized by flaccid limb weakness accompanied by gray matter lesions spanning 1 or more spinal cord segments. The first case of AFM potentially associated with EV-D68 infection was first described in 2005, but the first large-scale outbreak resulting in public health concern did not occur in the United States until 2014.¹⁵ In that year, >1000 children were hospitalized for EV-D68 infections in the United States, most of whom had respiratory illnesses, ranging from mild to causing respiratory failure. An increase in AFM, particularly in children, was also observed during this time, with ~10% of children hospitalized in the United States for EV-D68 receiving an AFM diagnosis.

AFM is much more common in young children, although cases in adults have been reported, particularly in immunocompromised individuals. Individuals typically present with an infectious prodrome of either respiratory or gastrointestinal symptoms of varying severity, followed by paralysis that most commonly affects the proximal upper limbs. CSF analysis is usually notable for a pleocytosis with lymphocytic predominance without identification of a causative organism by PCR or other means. Nasopharyngeal swabs may be positive for EV-D68 or other enteroviruses.¹⁶ MRI findings include gray matter hyperintensity with absence of edema in the spinal cord, although these findings may take several days to develop after symptom onset. Recovery from weakness is poor.¹⁷

EV-D68 rarely has been isolated in CSF analysis of people with AFM attributable to EV-D68 infection; other viruses with known neurotropism, such as poliovirus, are also not routinely found in CSF. This may be due to viral replication in the CSF preceding symptom onset or to virus presence in the brain parenchyma but not in the CSF. Despite the lack of isolation of EV-D68 RNA in CSF, there is strong evidence that EV-D68 is implicated in the development of AFM. The incidence of AFM consistently rises sharply in geographic areas experiencing an outbreak of EV-D68–associated respiratory infection, with individuals with AFM having 4 to 10 increased odds of testing positive for EV-D69 in respiratory samples.¹⁸ In addition, individuals with AFM often have a respiratory or gastrointestinal illness that precedes the development of neurologic symptoms, with testing for prodromal symptoms revealing EV-D68 in about half of the cases. Few autopsy reports exist, but 1 case report identified cytotoxic T-cell infiltration into gray matter in a man who died of EV-D68 pneumonia who was also found to have EV-D68 in the CSF by PCR.¹⁹ Recent animal models have also supported the hypothesis that the virus can infect neurons in the CNS. In neonatal mice, intracerebral injection of modern EV-D68 strains results in flaccid paralysis, with enterovirus virions seen in motor neurons.²⁰ Overall, geographic and epidemiologic data, along with animal studies, are greatly supportive, although not yet conclusive, of a causal relationship between EV-D68 and AFM.

Specific, effective treatments for EV-D68 are lacking. IVIg is frequently used clinically, although no large studies on its efficacy exist. Preclinical studies in mice show a reduction in viral load and paralysis in animals being treated with IVIg.²¹ Commercially available IVIg in the United States contains EV-D68 neutralizing antibodies.²² The antiviral agent telaprevir has been shown to inhibit EV-D68 replication in vitro and has recently been shown to improve paralysis in a mouse model of AFM, although it has not yet been studied in humans.²³ Corticosteroids are effective in treatment of weakness mediated by spinal cord edema, but there is little edema associated with AFM, and research in mice has shown that corticosteroid administration can increase viral load and worsen paralysis.²⁴

As is the case with many virally mediated illnesses, the most effective treatment is prevention. No vaccine for EV-D68 exists, but recent work in primates has shown that an EV-D68 viral-like particle with a carbomer-based adjuvant elicits CD4+ T-helper response and neutralizing antibody generation.²⁵

Conclusion

The landscape of viral infections of the CNS is ever-changing, with new knowledge of infectious sequelae, novel prevention and treatment options, and new disease entities emerging continuously. It is essential that clinicians keep abreast of this evolving field in order to be well-versed in advances in epidemiology, diagnosis, and management to improve outcomes.