Coronaviruses

Introduction
This category of viruses is now known to include seven types that infect humans:
 * Common coronaviruses
 * hCoV-229E
 * hCoV-OC43
 * hCoV-NL63
 * hCoV-HKU1
 * Severe coronaviruses
 * SARS-CoV-1
 * SARS-CoV-2
 * MERS-CoV (HCoV-EMC/2012)

Severe acute respiratory syndrome due to SARS-CoV-1 first surfaced in the early 2000s, causing an outbreak of over 8,000 cases in 29 different countries from 2002-2004, with 774 patients dying (~10% mortality). There have been no known cases of SARS-CoV-1 infection since 2004.

Middle East respiratory syndrome (due to MERS-CoV) first surfaced in 2012, with most cases in the Arabian peninsula. This virus has persisted through 2020, with 2519 cases as of January 2020 and 866 deaths (~35% mortality).

In 2019, the SARS-CoV-2 virus emerged and caused a global pandemic, infecting over 3 million people worldwide as of May 2020.

Pathophysiology
In rats, a murine coronavirus strain (MHV-JHM) causes subacute demyelinating encephalomyelitis, which resembles multiple sclerosis. It begins with an acute infectious encephalitis causing inflammation of neurons and glial cells, followed by a subacute demyelination in which disease is restricted to the glial cells. In vitro studies from the late 1990s showed that CoV-OC43 has been found to infect astrocytes, oligodendrocytes, neurons, and microglia in cell culture, with persistent infections developing in some astrocytes, oligodendrocytes, and neurons. Similarly CoV-229E has been found to infect astrocytes, oligodendrocytes, and neurons in cell culture, with persistent infections in some oligodendrocyte and neuron cell lines. Studies have suggested that CoV-OC43 and 229E can be found in brain tissue on autopsy, suggesting neuroinvasion. In mice, human CoV-OC43 can cause an acute and chronic encephalitis, predominantly infecting neurons. In mouse models, CoV-OC43 has also been associated with flaccid paralysis and spinal cord demyelination with some mutations in the spike glycoprotein.

Epidemiology
In a series of 64 children with coronavirus infections of various species, seizures occurred in 3% of cases and meningoencephalitis in 5% of cases.

In a series of 183 patients in China with suspected encephalitis, 22 (12%) were infected with a coronavirus.

Neurological problems associated with hCoV-HKU1

 * Seizures in children (possibly febrile seizures)
 * Psychosis

Neurological problems associated with CoV-NL63

 * Psychosis
 * Mood disorders (possibly)

Neurological problems associated with hCoV-O43

 * Encephalitis: case of 11 month old boy with severe combined immunodeficiency.
 * Acute disseminated encephalomyelitis: case of a child
 * Acute flaccid paralysis: case of a child coinfected with hCoV-229E
 * Guillain-Barre syndrome: pediatric case
 * Parkinson's disease: hypothesized, but not clearly demonstrated
 * Multiple sclerosis: possibly

Neurological problems associated with hCoV-229E

 * Acute flaccid paralysis: case of a child coinfected with hCoV-O43
 * Multiple sclerosis: detected more frequently in patients with MS than those without in some studies  but not in others

Pathophysiology
Studies on rats given SARS-CoV-1 showed that the virus can infect the brain. SARS-CoV-1 appears to predominantly infect neurons in cell culture studies and did not infect cells of astrocytic lineage in another cell culture study. Mouse-adapted SARS-CoV-1 is found in the brain of infected mice. Several studies have shown ACE-2 expression in mRNA from brain tissue, but this may be due to the expression of ACE-2 in endothelial and smooth muscle cells. Although several studies did not find ACE-2 in nervous system cells themselves,  more recent ones have shown ACE-2 expression intracellularly in neurons, and possibly glia. The Human Protein Atlas shows that low levels of ACE-2 mRNA are expressed throughout the brain. It may be more expressed in tissues related to regulation of the autonomic system, such as the hypothalamus and brainstem nuclei.

Using a mouse model transgenic for human ACE-2, researchers have found that SARS-CoV-1 spread into the brain and infected cells in the cerebrum, thalamus, and brainstem (including respiratory centers). Interestingly there was not significant cellular inflammation. . SARS-CoV-2 also appears to enter cells using the ACE-2 receptor, and shares 79% of its RNA to SARS-CoV-1. Both also require the serine protease TMPRSS2 for spike protein priming. ACE-2 in the brain may be involved in regulating autonomic responses.

The first autopsy study was published in July 2003, and two of three patients had edema around veins of the brain, with infiltration of the vascular walls by monocytes and lymphocytes, brain edema, demyelination, and focal neuronal degeneration. In June 2004, an autopsy study of 4 patients found that SARS-CoV-1 antigen and RNA were detectable in cerebral neurons in all four cases. These findings were confirmed in a 2005 autopsy study of 8 patients, with SARS genome sequences detected in the cytoplasm of numerous neurons in the hypothalamus and cortex in all eight cases, along with edema and scattered red degeneration of the neurons in 6 of 8 cases. Also in 2005, researchers did an autopsy of a patient who developed diffuse brain edema and multiple high-density lesions on CT, and confirmed presence of SARS-CoV-1 RNA and protein, as well as neuronal necrosis and striated encephalomalacia. Electron microscopy showed the presence of enveloped viral particles with coronavirus morphology in the brain itself.

How coronavirus get into the nervous system remains unclear. It may be hematogenous spread, or alternatively could enter in peripheral neurons and be transferred from there, as has been seen with the animal coronavirus, swine hemagglutinating encephalomyelitis virus , which shares significant DNA with SARS-CoV-1 and SARS-CoV-2. hCoV-OC43 may also be able to propagate from neuron to neuron. Accordingly, Toljan hypothesized that it may enter the CNS retrograde along nerves from the gastrointestinal tract. Baig et al. have proposed that SARS-CoV-2 may either travel across the cribiform plate to enter the CNS. However, evidence suggests that olfactory sensory neurons do not seem to express ACE2 or TMPRSS2, making it unlikely that these are infected themselves, but instead there may be infection in olfactory support cells, stem cells, and perivascular cells. In support of hematogenous spread, an autopsy study of a patient with encephalopathy with COVID-19 showed viral particles in the frontal lobe in small vesicles in endothelial cells and blebbing of viral particles in and out of the endothelial wall suggesting that they were the primary source for the nervous system. Neuronal cell bodies showed distended cytoplasmic vacuoles containing enveloped viral particles.

ACE-2 is expressed in skeletal muscles which may explain the occurrence of myalgias and rare cases of rhabdomyolysis reported with COVID-19. However, no SARS-CoV-1 was found in muscle in one study, although muscle inflammation with monocyte and lymphocyte infiltration was seen in SARS infection in another.

Inflammatory cytokines, including IL-1β, TNF-α, and IL-6 are upregulated in the brains of infected transgenic ACE-mice. It is hypothesized that cytokines and chemokines may open the blood-brain barrier, making the brain more susceptible to entry of SARS-CoV-1.

Others have hypothesized that some of the mortality associated with COVID-19 may be due to neurological effects on the cardiorespiratory centers in the brainstem.

On March 5th, 2020, non-peer-reviewed news reports first emerged of a patient who developed decreased consciousness and was found to have SARS-CoV-2 RNA in his CSF.

In cultures of rat neuronal and glial cells, nicotine decreased ACE2 levels. Similarly, direct nicotine inhalation in mice decreased ACE2 expression in the hypothalamus. However, studies in human lung tissue from lung cancer patients showed that the lungs had higher ACE2 expression in smokers compared with nonsmokers and upregulation of ACE2 in smokers has been hypothesized to potentially increase the risk of neurological disease in this population.

Epidemiology
The largest series of neurological involvement in COVID-19 included 214 patients from three hospitals in Wuhan, China, 78 (36%) of whom had neurological symptoms (included myalgias), and 46% of those with severe pulmonary infection. Patients with CNS symptoms were more likely to have lower platelets, lower lymphocytes, and higher BUN. Those with PNS symptoms had no clear predictors, and those with skeletal muscle injury were more likely to have higher neutrophil counts, lower lymphocyte counts, higher CRP, higher D-dimer, higher LDH, higher ALT, higher AST, higher BUN, and higher creatinine.

Neurological problems associated with SARS-1

 * Axonal neuropathy: described in several cases with normal CSF protein
 * Myopathy and rhabdomyolysis: described in two cases, with rhabdomyolysis in 3 patients.
 * Ischemic stroke: occurred in 2% of patients (5 of 206) in one study from Singapore, all large vessel
 * Seizure

Neurological problems associated with COVID-19

 * Hyperreflexia and extensor plantar responses: seen in 67% of cases in one series
 * Absence of dyspnea: dyspnea is absent in 62% of severe cases. This is hypothesized to possibly be due to damage to pulmonary C-fibers that usually cause dyspnea or alternatively to damage to the nucleus of the solitary tract.
 * Myalgia and skeletal muscle injury: 4-62% overall,         more common with severe pneumonia (19%) . CK elevations are described in 11-33% of cases.    but rhabdomyolysis is rare (0.2%)
 * Dizziness: 8-17%
 * Taste and smell impairment: 5-86%.
 * Unclear if this represents neural invasion into the olfactory nerve and bulb (which has been seen in animal models) or inflammation causing obstruction of olfactory clefts. Olfactory sensory neurons do not seem to express ACE2 or TMPRSS2, making it unlikely that these are infected themselves, but instead there may be infection in olfactory support cells, stem cells, and perivascular cells
 * Acute onset in 71% of those with smell or taste disorders, presenting symptom in 36% of those with smell or taste disorders
 * More than 70% of patients recovered olfaction within 8 days, and nearly all patients within two weeks.
 * Headache: 5-66%
 * PNS involvement: 9%
 * Miller Fisher syndrome and cranial neuropathies (four cases reported)
 * Guillain-Barre syndrome: preceded pneumonia onset in one case
 * Altered mental status: 8-9% overall in most series     , more common with severe pneumonia (15%) or in those who died (22%). In one series of 40 patients with ARDS, 65% had confusion on the CAM-ICU. Delirium is also likely to be common.  This may last longer, as 33% of patients (15 of 45) in one series had inattention, disorientation, or poorly organized movements after discharge.
 * Acute ischemic stroke: 1.6-5% overall, more common with severe disease (6-12%)
 * Includes small vessel, large vessel, and cardioembolic causes but perhaps more commonly large vessel occlusions.
 * Patient risk factors include: age ≥ 50, severe COVID-19 infection, history of hypertension, diabetes, or vascular disease.
 * Lab risk factors include: elevated WBCs, elevated neutrophils, lymphopenia, lower platelets, higher CRP, higher D-dimer, higher AST, higher BUN, and higher creatinine.
 * Occurred 8-24 days after COVID onset in one series of 6 patients
 * May be associated with lupus anticoagulant and antiphospholipid antibody, but not in all
 * May occur despite therapeutic anticoagulation
 * Not associated with presence of SARS-CoV-2 RT-PCR in the CSF
 * Nerve pain: 2%
 * Seizure: 1% but this is unclear as a study of 304 patients (108 severe) did not find any seizures. A probable seizure has been reported in a 6 week old infant coinfected with SARS-CoV-2 and rhinovirus.
 * Ataxia: 1%
 * Vision impairment: 1%
 * Intraparenchymal hemorrhage
 * Subarachnoid hemorrhage: one case report

Acute hemorrhagic leukoencephalitis or necrotizing encephalopathy
Although MRI was not performed, a case reported in 2005 of a patient infected with SARS-CoV-1 who had confirmed cerebral presence of virus is highly suggestive of acute hemorrhagic leukoencephalitis. A case of bilateral thalamic edema and hemorrhage has been reported with COVID-19 which was called acute necrotizing encephalopathy, and has been hypothesized to be due to the cytokine storm syndrome with elevated inflammatory markers.

SARS-CoV-1
In December 2003, Hung et al. reported a case of a 59 year old woman in Hong Kong who developed status epilepticus. NCHCT was unremarkable, and lumbar puncture showed normal protein and slightly elevated glucose. RT-PCR of CSF showed the presence of 6884 copies/mL in CSF and 6,750 copies in the serum, with no RBCs detected, making contamination from the patient's own blood very unlikely. Then, in February 2004, Lau et al. reported a case of a 32 year old woman with SARS-CoV-1 infection who had a seizure prompting lumbar puncture which showed no WBCs but RT-PCR showed the presence of SARS-CoV-1 in the CSF (there were only 20 RBCs, so less likely to be contaminated from the patient's own blood).

SARS-Cov-2
On March 4th, 2020, non-peer-reviewed news reports first emerged of a patient who developed decreased consciousness and was found to have SARS-CoV-2 RNA in his CSF. Imaging was reportedly normal. Then on April 3rd, 2020, Japanese researchers published a report of a 24 year old man with headache who had encephalopathy and seizure associated with COVID-19, and was found to have SARS-CoV-2 RNA in CSF (no RBCs) and evidence of DWI hyperintensity in the hippocampus. Additional cases of possible encephalitis have been reported in which CSF was negative for SARS-CoV-2 RNA, so they were suspected but not confirmed. However, one series of of 304 patients with COVID-19 (108 severe) did not find any seizures in the patients, suggesting a low seizure risk.

A 41 year old woman in Los Angeles was reported to have viral meningitis with fever and a new seizure, had inflammatory CSF, and tested positive for COVID-19 in the absence of respiratory failure. No CSF RT-PCR was able to be sent. A probable seizure has been reported in a 6 week old infant coinfected with SARS-CoV-2 and rhinovirus, with negative CSF RT-PCR and a paroxysmal episode of hypertonia has been reported in a 4 week old infant with COVID-19. A 30-year-old woman in Iran also was reported to have multiple seizures in the setting of COVID-19. A 60 year-old man in Italy presented with akinetic mutism and had mild lymphocytic pleocytosis and elevated protein in the CSF, with a positive serum but negative CSF RT-PCR. MRI was unremarkable. The patient seemed to respond to steroid treatment.

An autopsy study of a patient with encephalopathy showed viral particles in the frontal lobe in small vesicles in endothelial cells and blebbing of viral particles in and out of the endothelial wall. Neuronal cell bodies showed distended cytoplasmic vacuoles containing enveloped viral particles.

For children, in a series of 27 patients with COVID-19 pediatric multisystem inflammatory syndrome, a total of 4 patients (15%, ages 8-15) had neurological involvement. All had encephalopathy, with some having headache, brainstem signs with dysarthria or dysphagia, meningismus, or cerebellar ataxia.

Diagnostic testing in encephalitis/encephalopathy

 * Imaging
 * Hypoperfusion of bilateral frontotemporal regions on perfusion imaging (100% of patients assessed, n=11)
 * Leptomeningeal enhancement (62% of patients assessed, n=13)
 * Ischemic strokes (23% of patients assessed, n=13)
 * Splenium of the corpus callosum hyperintensity: in pediatric patients, seen in all four patients described with encephalopathy in the setting of COVID-19 pediatric multisystem inflammatory syndrome.
 * Normal in some cases.
 * EEG: nonspecific changes (n=8)
 * CSF
 * No cells in most patients with COVID (n=7) but 18 WBCs were reported in one case of encephalopathy
 * Elevated protein
 * Oligoclonal bands (29% of patients assessed, n=7)
 * Negative SARS-CoV-2 RT-PCR in 100% of patients in one study (n=7), although there are reports of CSF positivity.

Hypoxic encephalopathy
Not unexpectedly, this has been described to occur in 9% of patients overall and 20% of those who died. Cardiac arrest in patients with COVID-19 is associated with extremely poor outcome. Of 136 patients with cardiac arrest in one series, 90% had initial rhythm of asystole, 4% PEA, and 6% shockable rhythm. ROSC was achieved in 18 patients (13%), only 4 patients (3%) survived 30 days, and only 1 achieved favorable outcome (CPC 1-2) at 30 days.

Of patients with COVID-19 associated ARDS who survive, one series of 41 patients suggested that 39% achieved mRS 0-1, 39% achieved mRS 2, and 22% were no longer independent at 4 months.

Psychosis
After SARS-CoV-1 infection, patients have been reported to have a variety of neurological and psychiatric problems, many of which may be psychological in nature. Some, however, such as psychosis, may be neurological, although this is not at all proven.

Peripheral neuropathy and/or Guillain-Barre syndrome
Several patients have been described with neuropathy after SARS-CoV-1 infection, which appeared to be axonal and likely critical illness neuropathy and myopathy. In COVID-19, two cases have been reported of inflammatory cranial neuropathies (Miller-Fisher syndrome associated with anti-GD1b antibody, and polyarteritis cranialis). Both had cytoalbuminologic dissociation and negative CSF RT-PCR for SARS-CoV-2, and both improved significantly with IVIG. A 61 year old woman also presented with Guillain-Barre syndrome and was found to have COVID-19. CSF showed cytoalbuminologic dissociation and nerve conduction studies showed absent F waves and prolonged distal latencies consistent with demyelination. She was treated with IVIG and symptoms resolved by day 30. Interestingly, she developed COVID pneumonia on post-admission day 8.

Several patients have been described with neuropathy after SARS-CoV-1 infection, which appeared to be axonal and likely critical illness neuropathy and myopathy.

In a series of 4 pediatric patients with encephalopathy in the setting of COVID-19 pediatric multisystem inflammatory syndrome, all four had proximal muscle weakness, and 2 also had reduced reflexes.

Ischemic stroke
Ischemic stroke occurred in 2% of patients (5 of 206) with SARS-CoV-1 infection in one study from Singapore, all large vessel occlusions. This was hypothesized to be due to a hypercoagulable state. A cohort analysis of 1916 COVID-19 patients in New York City showed that 1.6% had an acute ischemic stroke. In a similar cohort of influenza patients from 2016-2018, of 1486 patients, 0.2% had ischemic stroke. After adjusting for age, sex, and race, the likelihood of stroke was higher with COVID-19 infection than influenza (OR 7.6, 95% CI 2.3-25.2).

Pathophysiology
For MERS-CoV, the virus enters cells using human dipeptidyl peptidase 4 (DPP4; CD26). In transgenic mice for human DPP4, MERS-CoV infection occurred predominantly in the brainstem, thalamus, and cerebrum. Infected brains had perivascular cuffing, cellular degeneration, and debris, with degenerating and dying neurons that stained for viral antigen. The virus was able to infect human glioblastoma, neuroblastoma, astrocytoma, and neuron-committed teratocarcinoma cell lines. DPP4 does appear to be expressed in human brain tissue, although at low levels, and this may be due to it being expressed on vasculature. It also may be upregulated in neurological disorders, such as Alzheimer's disease. In transgenic mouse models, intranasal MERS-CoV spread to the lungs and the brain, led to lethargy and weight loss, and was found to have infected cerebral neurons and glia. Interestingly, neurological disease occurred earlier and overall disease was more severe in mice infected intranasally vs. via aerosolization. Neurological disease does seem to generally appear after lung disease. The degree of inflammation is variable, with one model showing only mild perivascular cuffing, and another showing congestion of cerebral vessels and several areas of cellular necrosis.

Symptoms:

 * Myalgia: 14-63% of cases
 * Headache: 7-35% of cases
 * Altered mental status: 11-26% of cases
 * Convulsions: 10% of cases
 * Focal neurological deficit: 4% of cases
 * Ataxia
 * Intracerebral hemorrhage (possibly, case report)
 * Guillain-Barre syndrome / Bickerstaff's encephalitis / acute sensory neuropathies
 * Ophthalmoplegia
 * Painful peripheral neuropathy with leg weakness
 * Distal sensory paresthesias

In 2015, Arabi et al. published three case of patients with MERS and severe neurological disorders, with varied presentations and radiographic patterns. Importantly, all patients had multiorgan failure and lymphopenia. Symptoms included altered mental status, coma, ataxia, and focal motor deficits. Algahtani et al. published two cases, one of which was intracerebral hemorrhage in a patient with coagulopathy and thrombocytopenia, and the other being critical illness neuromyopathy, so neither clearly due to MERS neuronal infection. Kim et al. published four cases of patients with acute neuropathic complaints

Possible clinical syndromes

 * Encephalitis
 * Vasculitis (large, medium, or small-vessel)

Imaging
Varied radiographic patterns of injury have been seen, including:
 * Scattered subcortical patchy and globular diffusion restriction with T2/FLAIR hyperintensity (possibly similar to ADEM or medium to small vessel vasculitis)
 * Confluent periventricular diffusion restriction with T2/FLAIR hyperintensity and edema, associated with multiple large vessel near-total occlusions This appearance also looks possibly consistent with fungal infection (e.g. hematogenous Mucor), and LP was not obtained so hard to know if this is truly from MERS.
 * Faint confluent subcortical white matter T2/FLAIR hyperintensity without diffusion restriction.

Labs
In two cases in which CSF was analyzed, cell counts were negative, MERS-CoV RT-PCR was negative, and only protein was elevated.