Comments on “CSF-Confirmed SARS-CoV-2 Acute Encephalitis”: SARS-CoV-2-Associated Encephalitis Is Autoimmune Rather Than Infectious

Dear Editor,

With interest we read the article by Tee et al.1 about a 69-year-old male with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-associated encephalitis classified as infectious upon confirmation of the virus in the cerebrospinal fluid (CSF) by PCR. The patient benefited significantly from ceftriaxone, meropenem, favipiravir, dexamethasone, methylprednisolone, clonazepam, valproic acid, and tuberculostatics.1 Complete recovery was achieved 10 days after the onset of the neurological compromise.1 The study results are appealing, but they raise several comments and concerns, as detailed below.

One serious limitation of the study report is that reference limits for blood and CSF parameters were not provided, which greatly impedes the ability to interpret the results. According to our own laboratory, a cycle threshold of 36.08 for the CSF PCR is normal and does not prove infectivity. Thus, based on the study results, the diagnosis of infectious encephalitis due to infection with SARS-CoV-2 remains unproven.

Another limitation is that “results of other CSF analyses” were not specified.1 It should be reported if investigations for viruses other than SARS-CoV-2 in the CSF were truly negative, if tuberculous meningitis and immune encephalitis were appropriately excluded, and if the CSF was tested for antibodies associated with immune encephalitis. Excluding immune encephalitis is crucial since SARS-CoV-2-associated encephalitis is due to an immune reaction against the virus rather than to a direct viral attack against the central nervous system.2 Moreover, recovery under steroids suggests immune encephalitis rather than infectious encephalitis. Virostatics are usually ineffective against COVID-19. A further argument for an immune mechanism is that most of the patients with SARS-CoV-2-associated encephalitis were pathophysiologically classified as autoimmune (Table 1). However, in several cases encephalitis was classified as infectious despite the virus not being reported in the CSF. Though the absence of the virus in the CSF does not generally exclude an infectious pathomechanism, it does make it rather unlikely.

Table 1
Selection of patients with SARS-CoV-2-associated infectious/autoimmune encephalitis

• Click for larger image
• Click for full table
• Download as Excel file

A further limitation is that cerebral magnetic resonance imaging (MRI) was not performed on admission. Encephalitis can usually be confirmed on T1-weighted images after administering gadolinium, by the presence of various degrees of focal or global enhancement. MRI performed 1 month after presentation will not reflect the acute disease. Additionally, fluorodeoxyglucose positron-emission tomography may show hypermetabolism within affected areas.3 Several other cases of SARS-CoV-2-associated infectious encephalitis reported in the literature4, 5 were not listed in the table of the case report.1

We do not agree that the case reported by Duong et al.6 can be classified as infectious encephalitis, since the presence of the virus could not be confirmed in that particular case. A comprehensive workup of the CSF, including a cytokine profile, was also missing. In particular, interleukin (IL)-8, IL-6, IL-1b, and tumor necrosis factor-alpha can be elevated in the CSF of patients with SARS-CoV-2-associated encephalitis or myelitis.7, 8

In summary, the study of Tee et al.1 has several limitations that challenge the reported results and their interpretation. The reported diagnosis of infectious encephalitis remains unproven, with the beneficial effect of steroids instead suggesting an autoimmune pathogenesis. Inflammatory cytokines should be measured in the CSF, which may reveal the immunological pathogenesis of SARS-CoV-2-associated encephalitis.