Systemic review analysis
Although the main target of organ damage by SARS-CoV-2 was the respiratory system, based on several previous studies, patients also experienced neurological problems ranging from the mild manifestations (such as, headaches and dizziness) to life-threatening complications (such as, cerebrovascular disorders and encephalitis).6 The study conducted by Helms et al. showed that agitation (69%) was the common complaint found in COVID-19 patients receiving treatment in the Intensive Care Unit (ICU).7
Based on this review, the common symptoms found in encephalitis patients were disorientation or confusion (72.72%), decreased consciousness (54.54%), and seizures (27.27%). These symptoms indicated damages in the central nervous system, especially the cerebral cortex, typically found in acute encephalitis. This result was similar to the clinical manifestations of encephalitis in which disorientation was the most common symptom reported in patients with positive HSV and PCR (92%).35 In COVID-19-associated encephalitis, the manifestations of seizures and headaches were found in 27.27% and 30.3% of the cases, while HSV encephalitis seizures were found in 56% of the cases and headaches in 83% of the cases.35 Abnormal behaviour that was characteristically found in HSV was rarely found in COVID-19 encephalitis. The clinical appearance of COVID-19-associated encephalitis was similar to MERS-CoV 2-associated encephalitis, with the symptoms of upper respiratory tract infection (fever, cough, fatigue) and a decrease in mental status and rapid respiratory failure.36 Meningeal irritation sign was only found in 4 cases (12.12%), which showed that the pathological process was more dominant in the cortex than in the meninges. This also indicated that the involvement of meninges in COVID-19-associated encephalitis was less common than HSV, as the meningeal irritation signs appearing in 29% of HSV cases.35 However, these results can also be produced due to the loss of meningeal irritation in severe states of consciousness.
The laboratory results showed an increase in inflammatory markers such as, CRP (48.48%), LDH (30.3%), and lymphophenia (27.27%). This result was found to be higher than the general systematic review of COVID-19 studies conducted by Rodriguez-Morales et al37, while the increase in CRP and LDH were found in 22.2% and 6.3% of the cases respectively. These differences were due to the frequent and severe inflammatory reaction in COVID-19-associated encephalitis. Wang's research showed that CRP levels in the group with severe symptoms were significantly higher than those with moderate or mild symptoms (54.15 mg / dL vs 16.76 mg / dL vs 1.52 mg / dL, p < 0.05). 38 The condition of lymphopenia indicated the involvement of T lymphocytes, which caused CD4 and CD837,39 cell depletion. Lymphopenia occurred due to lymphocyte sequestration at specific target organs such as, the lungs, gastrointestinal tract, and lymphoid tissue, via the activation of ACE2 receptors by SARS-CoV-2. Other hypotheses indicated that SARS-CoV-2 had a similar phenotype and mechanism as SARS-COV, which included the tendency for direct bone marrow suppression, the immune-mediated destruction of lymphocytes40,41,42, and lymphopenia manifestation40
The brain imaging studies in some of the cases did not show significant pathologic features (51.51%). The most common pathological common finding was the diffuse hyperintensity of T2 / fluid attenuated inversion recovery (FLAIR), with location most frequently in white matter (24.24%). These results were similar with the systematic review conducted by Katal et al, where most of the normal magnetic resonace imaging (MRI) images (41%) were found in COVID-19 patients. Therefore, the results of neuroimaging encephalitis associated with COVID-19 in this review resembled with those of encephalopathy / encephalitis in the previous coronavirus outbreak. Two severe acute respiratory syndrome coronavirus (SARS-CoV) cases accompanied by severe neurological symptoms in the form of decreased consciousness and seizures showed a normal picture of neuroimaging.44,45 The research conducted by Arabi et al in 2016 reported in the three cases of middle east respiratory syndrome coronavirus (MERS-CoV) patients and evaluation by MRI showed that the hyperintense lesions on T2 spread widely and bilaterally in white matter and subcortical areas, frontal lobes, temporal, parietal, and basal ganglia, as well as the corpus callosum.36 The neuroimaging findings suggested possible similarities in the pathomechanism of the central nervous system involvenent in SARS-CoV- 2, SARS-CoV, and MERS-CoV.
The electroencephalography (EEG) examination was carried out on 11 cases, with the result showing a non-specific general slowing EEG (63.30%). Meanwhile, the epileptiform foci were found in 3 cases (27.27%), 2 in the temporal and 1 in the frontal lobe. This was consistent with the study conducted by Canham et al, which stated that the predominant EEG features in severe COVID-19 patient experienced a general slowing.47 These findings indicated that the mechanism of central nervous system in COVID-19 was diffuse and widespread, in contrast to encephalitis caused by HSV, where a typical 2–3 Hz periodic lateralized epileptiform discharges, originating from the temporal lobes.48 The CSF analysis showed increase in protein level (42.42%), white blood cells (27.27%), and lymphocytes (24.24%) case. In addition, the cerebrospinal fluid (CSF) and polymerase chain reaction (PCR) examination showed a negative result for SARS-CoV-2, due to the direct invasion of the virus.
The large variety of pharmacological agents was as a result of the dynamic changes of COVID-19 guideline and protocols in each center. The use of acyclovir and ceftriaxone were generally given as the empirical therapy, and was stopped when there was no evidence of a bacterial infection or HSV. There was an increase in proinflammatory cytokines in CSF18,24,27,33,34, indicating the possible role of intravenous steroids and immunotherapy (IVIg and plasmapheresis) in the management of encephalitis associated with COVID-19. The methylprednisolone and IVIg therapy given in the 4 cases of COVID-19-associated encephalitis18,26,29,32 showed a positive response, with 3 patients discharged in stable conditions18,29,32, while 1 was still in care with significant improvement.26 A study conducted by Dogan et al stated that plasmapheresis therapy showed dramatic improvements in both clinical and laboratory findings.34 This positive result supported the theory of a cytokine-mediated hyperinflammatory response as the basis for pathomechanism of COVID-19-associated encephalitis.18
Cytokine-immune-mediated inflammation as the underlying patomechanism in SARS-CoV-2-associated inflammation
The mechanism of encephalitis in COVID-19 was found to be unclear, however, previous studies had indicated that there was a neurotropism of the SARS-CoV-2 virus which allowed the invasion of the virus towards the central nervous system. There were two pathways that allowed this invasion, namely through the systemic circulation and the cribriform plate of ethmoid bone.49 The SARS-CoV-2 virus bound with the angiotensin-converting enzyme 2 (ACE2) receptor via spike protein S1, allowing the attachment of virions to cell membranes.49,50 The systemic dissemination resulted from the attachment of SARS-CoV-2 to the ACE2 receptor in the capillary endothelium.50 ACE2 expression in glia cells and neurons was the pathway mechanism for cerebral damage.49,50 The occurrence of hyposmia or anosmia due to the spread of the virus in the olfafactory bulb via the cribroform plate was the alternative pathway for invading the central nervous system.49
Furthermore, this study opposed that the pathomechanism of central nervous system damage in COVID-19-associated encephalitis was caused by direct invasion of the virus. The MRI images that were predominantly normal in this study were consistent with the features of autoimmune encephalitis, while based on previous research, 60% of the N-methyl-D-Aspartate receptor (NMDA-R) encephalitis wwere generally normal.51 EEG findings indicated diffuse cerebral abnormalities, possibly the results of severe and extensive inflammation. CSF analysis showed that the inflammatory process (denote an proteinorrachia) mediated by cytokines, was supported by the results of increased proinflammatory in the CSF. The predominantly negative CSF PCR (89%) against SARS-CoV-2 opposed the hypothesis regarding direct viral invasion of the brain as the cause of encephalitis. The possibility of an autoimmune mechanism was considered based on several previous studies linking COVID-19 with GBS.52 There were two reported cases with autoimmune features namely, the NMDA-R33 and limbic encephalitis (T2 hyperintense signal abnormalities in limbic lobes, bilateral medial thalamus, and frontal white matter).32 The anti-neuronal autoantibodies were only positive in 1 case (NMDA-R)33 of the 8 cases examined (12.5%), indicating the emergence of new antigens, mediating immune reactions to the central nervous system. The positive response showed by COVID-19 patients towards intravenous steroid therapy and immunotherapy (IVIg and plasmapheresis) commonly used in in the autoimmune conditions such as, myasthenia gravis and GBS proved that there was an immune process in the occurrence of encephalitis.53
Based on the theories from the previous studies54,55,56 the pathomechanism of immune-mediated cerebral damage in COVID-19-associated encephalitis were illustrated (Fig. 2). The binding of SARS-CoV-2 to ACE2 receptor via spike protein, caused ACE2 downregulation50,54, followed by an increase in ACE / Angiotensin II / AT1R axis, and a decrease in ACE2 / Mas receptor (MaSR) axis.54 Consequently, these resulted to the activation of the nuclear factor kappa B (NF- kappa B) pathway, Mitogen-Activated Protein Kinase (MAPK) pathway, upregulation of tumor growth factor beta (TGF-beta) accompanied by downregulation of anti-inflammatory cytokine IL-10.54.55 Therefore, the increased proinflammatory pathways rised the levels of monocyte chemotactic protein 1 (MCP-1) ), vascular cell adhesion molecule 1 (VCAM-1), selectin-E, and interlekukin 6 (IL-6).55 The lL-6 was found to be the core part of the cytokine storm. 54,55,56 The lL-6 activated the CD4 + T cells into Th17, which aggravated the pro-inflammatory cytokines namely, IL-17, IL-21, and IL-22.57 The cytokine storms and the endothelial dysfunction caused blood-brain-barrier damage, making it easier for pro-inflammatory cytokines to enter the brain parenchyma and also caused further neural damage.56
It was previously confirmed that the pathomechanism underlying COVID-19 encephalitis was a cytokine-immune-inflammatory process.18,34,56 However, further research was needed to explain and confirm this hypothesis.
The limitation encountered was due to the large variety of case reports and some investigation such as, EEG and PCR which were not reported in all studies. The negative CSF and PCR pointed out that the cause of COVID-19 encephalitis was not a result of viral direct invasion. The inflammatory markers and the anti-neuronal autoantibodies assessment had only been done in few studies, therefore, it was inconclusive that the cause of COVID-19 encephalitis was cytokine-immune mediated inflammatory process.