부산대학교 도서관

OBJECTIVE: To explore whether hospitalized patients with SARS-CoV-2 and neurologic symptoms have evidence of CNS infection, inflammation and injury using CSF biomarker measurements. METHODS: We assessed CSF SARS-CoV-2 RNA along with CSF biomarkers of intrathecal inflammation (CSF white blood cell count, neopterin, β2-microglobulin (β2M) and immunoglobulin G-index), blood-brain-barrier (BBB) integrity (albumin ratio), and axonal injury (CSF neurofilament light chain protein [NfL]) in 6 patients with moderate to severe COVID-19 and neurologic symptoms who had undergone a diagnostic lumbar puncture. Neurologic symptoms and signs included features of encephalopathies (4/6), suspected meningitis (1/6) and dysgeusia (1/6). SARS-CoV-2 infection was confirmed by rtPCR analysis of nasopharyngeal swabs. RESULTS: SARS-CoV-2 RNA was detected in the plasma of 2 patients (Cycle threshold [Ct] value 35.0–37.0) and in CSF at low levels (Ct 37.2, 38.0, 39.0) in 3 patients in one but not in a second rtPCR assay. CSF neopterin (median, 43.0 nmol/L) and β2-microglobulin (median, 3.1 mg/L) were increased in all. Median IgG-index (0.39), albumin ratio (5.35) and CSF white blood cell count (<3 cells/µL) were normal in all, while CSF NfL was elevated in 2 patients. CONCLUSION: Our results on patients with COVID-19 and neurologic symptoms suggest an unusual pattern of marked CSF inflammation in which soluble markers were increased but white cell response and other immunologic features typical of CNS viral infections were absent. While our initial hypothesis centered on CNS SARS-CoV-2 invasion, we could not convincingly detect SARS-CoV-2 as the underlying driver of CNS inflammation. These features distinguish COVID-19 CSF from other viral CNS infections, and raise fundamental questions about the CNS pathobiology of SARS-CoV-2 infection.