Discussion

The mechanism underlying EBV-associated CNS infection is not well established. Current evidence suggests that EBV gains access to the CNS by utilizing infected lymphocytes to traverse the blood-brain barrier. Postmortem studies in specific cases have identified infiltrated lymphocytes containing EBV DNA in meninges and perivascular areas, with their absence in neurons indicating inflammatory brain damage resulting from an immune response rather than direct viral invasion, in contrast to herpes simplex virus (HSV) infection [5-9]. An in-vitro study provides an intriguing alternative perspective, demonstrating EBV’s capacity to infect various neural cells, generating progeny virus through lytic replication and causing host cell destruction, potentially infecting other neurons and mononuclear cells [10]. In a rat encephalitis model, anti-neuronal antibodies have been detected [11]. Despite these controlled laboratory findings, demonstrating this phenomenon in vivo remains inconclusive with current knowledge.
EBV-associated CNS infection can occur in 1-3 weeks following an acute infection [12], but can also occur in the absence of acute infection, potentially attributed to reactivation. There is currently no established diagnostic criteria of EBV-associated CNS infection. While a possible diagnosis has been suggested based on positive serologic findings alongside compatible neurological symptoms [13], the rarity of this condition dictates the exclusion of more common viruses in suspected cases among young immunocompetent patients [8]. Further support for diagnosis may be provided by a positive cerebrospinal fluid (CSF) polymerase chain reaction (PCR) for EBV DNA [6, 13], yet the specificity and sensitivity of this test remain undetermined. In a review of 23 cases diagnosed with EBV-associated CNS infection, 20 tested positive on the CSF PCR [11]. However, even when CSF PCR for EBV DNA is positive, co-infection is prevalent, found in 22% of patients with compatible symptoms. In such cases, it has been suggested that, instead of being the primary pathogen, EBV may be reactivated secondary to inflammatory responses from other pathogens, or even contamination from EBV-infected lymphocytes [9]. While direct isolation of the virus in brain biopsy or culture from the CSF could support the diagnosis [6], such approaches are less practical in clinical practice.
In our patient, the EBV DNA in CSF is measured at less than 200 copies/ml [3,4], which falls within the previously reported range of 51-216,000 copies/ml [9][14]. The absence of other pathogens on PCR and bacterial panels makes contamination unlikely. In cases involving potential co-infections with other pathogens, we concur that utilizing reverse transcription-polymerase chain reaction (RT-PCR) for messenger RNA (mRNA) of the lytic cycle gene BZLF in the CSF may be the optimal approach to establishing the pathogenic role of EBV [9]. The BZLF gene, responsible for encoding a transcription regulator protein, is expressed by EBV only during the transition from latent to lytic infection.
Our case also demonstrated the challenging nature of determining the chronicity of an EBV-related CNS infection. For an acute infection, a positive serum EBV-specific IgM and IgG to the Viral Capsid Antigen (VCA IgG, IgM), a positive IgG to the early antigen (EA IgG), and a negative IgG to nuclear antigen (EBNA IgG) is suggestive. If follow-up serology were obtained, significant changes in EBV-specific antibodies such as a rise in IgG level, the disappearance of VCA-IgM and EA IgG, and the appearance of EBNA IgG further support an acute infection [13]. The positive EBNA IgG in the context of positive VCA IgM and VCA IgG in our patient may initially appear perplexing. VCA IgM is a marker of acute infection. At the same time, EBNA IgG is usually expressed only during latent infection, beginning from 6 weeks after the first infection, as EBNA are proteins responsible for maintaining an episomal state of EBV DNA as well as immortalizing B-cells in which EBV persists during latent infection. However, this can be reconciled by the fact that VCA IgM can be presented up to 3 months following acute infection. Our patient’s reported symptoms of infectious mononucleosis 8 weeks prior and mild hepatosplenomegaly on the exam are consistent with that. Another possibility is a reactivation infection. In either case, the timing of the infection is not likely to be clinically significant, and we suggest serial follow-up serology in case chronicity needs to be elucidated.
Neuroimaging is normal in most of the patients with EBV-associated CNS infection, as in our patient. However, it is essential to perform an MRI to rule out other causes of CNS infection such as HSV encephalitis or acute disseminated encephalomyelitis (ADEM), which typically presents with lesions mostly in the deep and subcortical white matter [10]. The MRI findings in EBV-associated CNS infection can vary widely in hyperintensity in the basal ganglia, thalamus, cerebral cortex, brainstem, optic nerves, splenium, and corpus callosum [15, 16]. These manifestations may extend to brainstem hemorrhage [5], meningeal enhancement and multi-level spinal cord involvement [13]. While most MRI abnormalities are observed in the FLAIR and T2 sequences, restricted diffusion on the DWI sequence can also be presented. Additionally, heterogeneous signals on ADC sequences can be observed, ranging from hypointensity to hyperintensity [17]. However, none of these findings is specific to EBV infection.
There is no standard treatment guideline for EBV-associated CNS infection. There have been reports of ganciclovir or acyclovir treatment with or without corticosteroids. This is due to ganciclovir’s ability to suppress replication in DNA viruses and good in-vitro activity against EBV virus [16]. However, the efficacy remains unclear [9,18]. Suppose the main pathogenetic cause of EBV-associated CNS infection is the result of the inflammatory response by the body’s immune system rather than direct viral invasion. In that case, systemic corticosteroids are more likely to be efficacious. While the exact pathogenesis of EBV-associated CNS infection is not yet entirely clear, we believe it is reasonable, in the absence of contraindications, to initiate treatment with both antivirals and systemic corticosteroids. Our patient received both systemic corticosteroid and intravenous acyclovir empirically and they were discontinued once the patient showed continuous improvement, and we thus believe the duration of this treatment could be guided by the patient’s clinical response.

Conclusions

EBV infection involves a very heterogeneous constellation of symptoms. Isolated, EBV-associated CNS infection should be considered in young adults with altered mental status when other more common CNS viral infections have been ruled out. A positive CSF PCR for EBV DNA strongly supports the diagnosis in the absence of other pathogens in the PCR. MRI findings typical of other conditions, such as ADEM, also must be ruled out. Due to the unclear pathogenesis of EBV-associated CNS infection, we suggest the combination of antivirals and systemic corticosteroids if not contraindicated.