Response to: Before SARS-CoV-2-related Encephalitis Can Be Attributed to Anti-GFAP Antibodies, Alternative Etiologies Must Be Ruled out

The Authors Reply We were pleased to receive a letter from Finsterer concerning our paper (1) and have responded to his concerns. As Finsterer pointed out, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antigen test does not have high sensitivity in asymptomatic cases but has high specificity. Since our patient had symptoms consistent with SARS-CoV-2 and was antigen-positive, we believe that the diagnosis of coronavirus disease 2019 (COVID-19) was appropriate. In our patient, negativity in polymerase chain reaction (PCR) analysis of nasal swabs for SARS-CoV-2 was confirmed on day 26 (26 days after the onset of COVID-19 and 22 days after the disappearance of COVID-19 symptoms). It has been reported that the PCR positive rate of nasal swabs is approximately 10% when conducted between 14 and 21 days after the disappearance of SARS-CoV-2 symptoms (2). Therefore, a negative PCR test at our institution is not a basis for denying SARS-CoV-2 infection (a PCR analysis of nasal swabs for SARS-CoV-2 was not performed during the acute phase of infection). In addition, positivity of SARS-CoV-2 anti-nucleocapsid antibody supports the presence of SARS-CoV-2 infection. The fact that the SARS-CoV-2 antigen in the cerebrospinal fluid (CSF) is negative indicates that encephalopathy is not caused by the SARS-CoV-2 infection itself but by autoimmune encephalitis secondary to the activation of autoimmunity caused by the virus infection. This is compatible with the clinical course in which encephalopathy occurred after the SARS-CoV-2 infection had resolved rather than during the peak stage of the infection.

As Finsterer pointed out, we did not perform a virology panel test for encephalitis/meningitis. However, it is obvious that our patient had autoimmune encephalitis rather than viral encephalitis, as mentioned above. As described in a case report (1), a serum sample was found to be negative for both anti-aquaporin-4 antibody and anti-myelin oligodendrocyte glycoprotein antibody. However, his CSF was positive for anti-glial fibrillary acidic protein (GFAP) antibodies in cell- and tissue-based immunofluorescence assays. No other antineuronal antibodies were detected in tissue-based assays using frozen rat brain sections. Inflammatory cytokines in the CSF are interesting, but they increase nonspecifically in inflammatory encephalitis. Measuring inflammatory cytokines may be useful for elucidating the pathological mechanism, but it was considered to have little clinical utility in influencing the direction of treatment, so such analyses were not conducted in our patient.

In the follow-up CSF study, a CSF analysis on day 43 showed persistent lymphocytic pleocytosis (90 /μL; 100% mononuclear cells). CSF cell counts reportedly continue to increase even after the disease has ameliorated in autoimmune GFAP astrocytopathy (3). Therefore, we consider the discrepancy between clinical improvement and follow-up CSF findings in our patient to have been compatible with autoimmune GFAP astrocytopathy.

Finally, we apologize that we could not provide clear magnetic resonance imaging (MRI) scans due to motion artifacts. However, the MRI findings were confirmed by neuroradiologists.

The authors state that they have no Conflict of Interest (COI).