Reemerging Infectious Diseases and Neuroimmunologic Complications

Authors: Avindra Nath, Dennis L. Kolson

Abstract

During the past decade (and beyond), neurologists have become aware of the emergence, persistence, and consequences of some familiar and new infections affecting the nervous system. Even among the familiar CNS infections, such as herpes virus, polyoma virus/JC, influenza, arbovirus, and hepatitis, challenges remain in developing effective antiviral treatments and treatments of postinfection sequelae. With the changing environment and increased global travel, arthropod vectors that mediate zoonotic disease transmission have spread unfamiliar viruses such as West Nile virus, dengue, chikungunya, equine encephalitis, and Zika, among others. Although the global health impact of these diseases has not risen to that of COVID-19 and HIV, it is likely to dramatically increase with continued spread of transmission vectors and the emergence of new zoonotic animal-to-human diseases mediated by those transmission vectors. Furthermore, specific virus-targeting treatments or effective vaccines for arboviral infections are not yet available, and this represents a major challenge in limiting the morbidity of these infections. By contrast, HIV-1, a disease that originated by direct transmission from nonhuman primates to humans (as early as the 1930s), after many years of intense study, is now targeted by highly specific and effective antiviral drugs that can limit the spread of infection and extend human life and health in all populations. Even with these dramatic therapeutic effects of suppressing HIV replication, neurologic dysfunction (primarily cognitive impairment) affects significant numbers of persons living with HIV. This emphasizes not only the importance of treating the underlying infection but also developing treatments for legacy effects of the initial infection even after antiviral therapy. Notably, the rapid emergence of SARS-CoV-2 infection was met with rapid implementation of highly effective and specific antiviral therapies. This resulted in early and dramatic lowering of the morbidity and mortality of SARS-CoV-2 infection. Nonetheless, the postinfectious complications of SARS-CoV-2 infection (long COVID) are now among the more costly consequences of emerging zoonotic infections worldwide. Developing new antiviral therapies that can penetrate the CNS, vaccines, and therapies that target host immune responses and metabolic dysfunction will be necessary for management of infectious and postinfectious complications of established and emerging infections.

Background

It is said that infections have killed more people than all the wars, famine, and climate change combined. By such killing and eliminating some civilizations, infections may have shaped human evolution. Plague, smallpox, measles, polio, and influenza viruses are examples. Certainly, the current SARS-CoV-2 pandemic is an example of the devastation caused by an infection even in today's society where tens of millions of people have died over a short period. Despite advances in technology, science, and medicine, we find ourselves extremely vulnerable to the onslaught of these infections. In today's jet-set society, infections that spread from humans to humans particularly through droplets can have a global impact at a very rapid rate. Vector-borne infections such as West Nile virus, dengue, Zika, and chikungunya have also spread recently in pandemic proportions. Most of these pandemic infections also cause neurologic complications. Long-term disability in the survivors is largely due to neurologic manifestations. Hence, it is important for neurologists to be well versed in their clinical manifestations, diagnosis, and management.

It is important to note that once a pathogen emerges, it continues to persist. Examples include HIV infection, which emerged in the 1980s and persists despite adequate antiretroviral therapy. SARS-CoV-2 is following a similar pattern. Other viral infections such as measles and polio that have been controlled with aggressive vaccination strategies can re-emerge, when vaccination efforts fail. Another interesting phenomenon is that once a virulent strain of a pathogen has been controlled, the less virulent strains take their place and over time evolve to become virulent. Examples include enterovirus D68, which causes a polio-like syndrome and which emerged approximately a decade ago. Zika virus, similarly first discovered in the 1950s, gradually mutated to cross the placental barrier and become neurovirulent (in the developing fetus) within the past decade. Only rarely has an infection of pandemic proportions been eradicated. This is true for smallpox. However, it seems that mpox (formerly known as monkeypox), which has cutaneous manifestations like smallpox and which can cause neurologic symptoms such as headaches, myalgias and fatigue (∼30% overall incidence), and seizures (2%) in immunocompromised hosts, may be emerging in its place.^1^

Within the past decade, much knowledge has been gained about many neurotropic viruses. This review addresses knowledge of established and emerging infectious diseases affecting the nervous system over the past 10 years. Similarities and distinctions between what are known about mechanisms of virus entry, replication and evolution in the CNS, neuropathogenesis, and treatment strategies are discussed (Figure 1). We highlight some of the articles that were published in Neurology: Neuroimmunology & Neuroinflammation within the past 10 years and place them in the context of the broader literature.

Herpes Viruses (VZV, HSV, HHV-6)

Varicella zoster virus (VZV) is a member of the herpes virus family. It is well known to cause chickenpox in children after first exposure to the virus, and in adults, it causes herpes zoster rash due to reactivation of the virus in the dorsal root ganglia. An aggressive form of zoster has been described in patients with multiple sclerosis (MS) on dimethyl fumarate.^2^ It has also been implicated in giant cell arteritis although this remains controversial. In immunocompromised individuals, VZV can cause an encephalitis, CNS vasculitis, and necrotizing retinitis.^3^

Emerging evidence implicates certain host sequence variations in increasing risk of complications of VZV and other herpes virus infections. It was discovered that variations in the RNA polymerase III gene increases the susceptibility to developing VZV vasculitis in children.^4^ This adds to the list of other sequence variations described that specifically increase susceptibility to herpes virus infections.^5^ Of these, variations in toll-like receptor 3 have been most commonly described in children with herpes simplex encephalitis (HSE, associated with herpes simplex type 1 [HSV-1]) and has also been associated with an autoimmune encephalitis, postherpes simplex encephalitis.^6^ It is also being increasingly recognized that HSE can trigger autoimmune encephalitis in approximately 20%–24% of patients after several weeks of the HSE diagnosis.^7^ In these patients, neuronal antibodies (frequently NMDAR antibodies and less frequently others) appear several weeks after HSE resolves and new CNS autoimmunity symptoms appear, and the patient is no longer PCR-positive for HSV-1.

A less well-recognized herpes virus, human herpes virus 6 (HHV-6), has been associated with new-onset seizures in children, but in immunocompromised patients, it can cause a wide variety of CNS complications and hence requires a high degree of suspicion.^8^ Although VZV and HSV-1 infections are treatable with approved antiviral drugs, HHV-6 has no approved antiviral drug therapy.

Polyoma Viruses

Polyoma viruses are small DNA viruses that remain latent in healthy individuals but may become pathogenic in immunocompromised individuals. Two polyoma viruses, JC and BK, are human pathogens; BK occasionally causes disease in renal transplant patients while JC causes the CNS disease progressive multifocal leukoencephalopathy (PML) in immunocompromised individuals. PML is a fatal opportunistic disease caused by JC virus replication within the CNS. Most people are infected by the virus in early childhood. The infection is asymptomatic at the time, but in immunocompromised individuals, it can cause a subacute progressive yet fatal infection of oligodendrocytes and astrocytes. MRI findings can be diagnostic, generally appearing as white matter hyperintensities commonly in the frontal and parietotemporal lobes, but also in occipital and brainstem regions. Some patients with JC virus infection develop cerebellar atrophy, due to infection of the cerebellar granule cells. A novel mutation in the viral VP-1 protein has been identified in such patients, which may explain the altered viral tropism for cerebellar granule cells.^9^ The diagnosis of PML is confirmed by detection of JC virus in CSF. Notably, a case of CMV encephalitis in an immunocompromised patient that mimicked PML has been described.^10^ Therefore, demonstration of the virus in CSF or brain is important to establish the diagnosis. Formerly, PML was a rare complication in patients with lymphomas and other cancers, but it emerged as a major opportunistic infection in the era of AIDS, and now with proliferation of immunomodulatory drugs to treat cancer or autoimmune diseases such as MS, PML has re-emerged.^11^ Currently, there are no antiviral agents for treating the infection.

The main therapeutic approach to PML is to enhance the immune response against the virus. One approach is to use immune checkpoint inhibitors that can reverse T-cell exhaustion to regain their cytotoxic properties against the viral antigens.^12^ A similar approach has been successful in some cancers. However, this can only be successful if there are enough T cells to stimulate and they are exhausted as demonstrated by expression of the checkpoint molecules on their surface, in the absence of which such treatment may not be useful. Another alternative is to use cell-based therapy in which matched donor cells are stimulated in vitro with antigens that are homologous to BK virus and JC virus and then transfused into the patient intravenously.^13^

Although immune stimulation is important to eliminate the virus, excessive immune reaction can be detrimental and result in an immune reconstitution inflammatory syndrome (IRIS), which needs to be controlled with anti-inflammatory agents. On MRI, PML-associated IRIS typically presents as enhancing lesions with either circumferential or nodular enhancement. However, a miliary form of IRIS has also been described.^14^ Some have advocated for the use of maraviroc, which is a CCR5 receptor antagonist, for treatment of IRIS, because it decreases macrophage and T-cell activation.^15^

Arboviruses

Arboviruses (arthropod-borne viruses) continue to emerge as major human pathogens with a predilection for causing CNS injury. Dr. Anthony Fauci recently called the mosquito the “deadliest animal” after he recovered from a bout of West Nile encephalitis. West Nile virus was introduced into the United States in 1999. Since then, it has rapidly spread across the country and is a perfect example of how the presence of a vector, in this case the culex mosquito, can help cause an epidemic once a virus is introduced into that population even when no human-to-human transmission occurs (Figure 2). Here, birds serve as an intermediary host, which helps carry the virus long distances. The virus can cause a multitude of neurologic manifestations, most commonly encephalitis. Patients on anti–B-cell therapies can develop severe encephalitis if infected with this virus.^16^ Rare manifestations may include a myelitis, vasculitis, and intracerebral hemorrhage.^17^ Zika virus is another example of mosquito-borne viruses that emerged in Brazil, causing microcephaly in children through transplacental spread. The virus then rapidly spread across South America and beyond (Figure 3). In adults, it commonly causes GBS, which is associated with a spectrum of antibodies against the peripheral nerve.^18^ Eastern equine encephalitis virus is considered to be the most neurovirulent of all the arboviruses, with more than 10 human cases now reported in the northeastern United States in 2024 (cdc.gov). There is no specific antiviral therapy available. In isolated cases, IV immunoglobulin and high-dose steroids have been used.^19^ There is an urgent need for development of vaccines and antiviral agents that target these viruses.

Over the past few years, cases of dengue virus have increased exponentially and have spread to several countries including places where this infection was not previously seen.^20^ Previous attempts to control the infection by vaccination have failed. A major challenge to controlling this infection is that it has 4 serotypes and infection with one serotype does not protect against the other. In fact, infection with one serotype can sometimes produce enhancing antibodies that can then bind to the virus of a different serotype and drag the virus into the cell leading to increased complications by repeated infections. This is particularly important for causing neurologic manifestations. In rare instances, the virus may spread by cell-cell contact within the brain and this persistent virus can present as a neurodegenerative disease with mild neuroinflammation. In these cases, dengue virus may be detected within the brain but not in the CSF.^21^ It is also believed that infection with these viruses can lead to a postviral syndrome similar to that seen with long COVID; however, this remains underappreciated and understudied. Unfortunately, several attempts to make vaccines against dengue virus have been unsuccessful.

In chikungunya virus infection, fevers and arthralgias are common (approaching 100%), but neurologic complications with chikungunya virus are very rare. These include reported cases of GBS, acute tetraparesis, or paraparesis with MRI findings of T2 hyperintense lesions within the brainstem and deep cerebral white matter and along the surface of the spinal cord and anterior nerve roots of the cord.^22,23^

Hepatitis E

The role of hepatitis E virus (HEV) in causing neurologic manifestations remains controversial. A report of 200 patients with hepatitis E infection revealed 16% with neurologic symptoms, including neuropathic pain and other sensory disturbances.^24^ In an observational case series of approximately 1,100 patients with MS receiving disease-modifying therapy, acute hepatitis E infection and transient liver injury were reported in 4 patients, one of whom was diagnosed with neuralgic amyotrophy.^25^ Another remarkable case report describes a patient with a 5-year history of fatigue, frequent painful muscular spasms in the extremities, and tingling sensations in the hands and feet.^26^ The patient had no abnormal neurologic findings, and EMG and MRI studies (brain and spinal cord) were also normal. Hepatitis E virus was detected in the patient's CSF, which also contained elevated IgG against HEV and no signs of inflammation. Sequencing of the virus showed that it had an eight-base pair deletion that was not detected in virus in the patient's serum, which suggests either mutation of the virus followed by invasion of the brain or independent virus genetic evolution within the CNS compartment. This mechanism of virus compartmentalization and evolution in the CNS is also typical for HIV-1 infection, which is associated with the establishment of persistent HIV-1 reservoirs in the CNS.

Human Immunodeficiency Virus Type 1

HIV-1 emerged in the late 1970s, causing massive devastation in the form of AIDS and now persisting in the population despite the use of a variety of antiretroviral drugs. As a human retrovirus, HIV is unique in its ability to reverse-transcribe its RNA genome into a complementary DNA that efficiently integrates into the human genome, thus establishing a permanent HIV reservoir within the body, including the brain. Despite highly efficient suppression of HIV replication by antiretroviral drugs and near-normal life expectancy in treated patients, HIV continues to cause neurocognitive dysfunction in approximately 20% of persons living with HIV. Despite extensive advancements in our understanding of the neuropathogenesis of HIV-1 infection, there are still major gaps in our knowledge. It has been shown that there is increased oxidative stress in the brain, but the reasons for this are not fully understood.^27^ Recent advances include the identification of polymorphisms in the promoter region of the heme oxygenase-1 gene (HMOX1), which encodes a ubiquitous antioxidant enzyme that protects host cells against oxidative injury. Common genetic variations in this promoter associate with reduced ability to suppress inflammation and oxidative injury associated with neurocognitive impairment in HIV-infected individuals.^28^

The persistence of HIV in the CNS can result in ongoing injury despite the ability of antiretroviral drugs to suppress HIV replication. Notably, current antiretroviral drugs do not include a viral transcription blocker. Hence, HIV transcripts and viral proteins are still formed and released extracellularly. HIV transcripts can be detected in extracellular vesicles in serum and CSF despite long-term suppression of viral replication. Sequencing of viral transcripts from the extracellular vesicles demonstrates compartmentalization of the virus in the CNS.^29^ A product of one of the HIV transcripts, the tat protein, is detectable in the CSF of these individuals.^30^ In these studies, the amounts of viral transcripts and viral protein correlated with neurocognitive deficits. This suggests that “lock and block strategies” that lock the HIV promoter in a deep latent state and block HIV transcription are going to be critical to prevent these complications of HIV infection. Multiple attempts to make vaccines against HIV have failed.

It is also being realized that there may be multiple biotypes of HIV-associated neurocognitive disorders; these biotypes are distinct in their pathophysiology and radiologic findings.^31^ This has important relevance for management of these patients because neuroprotective treatment strategies in addition to antiretroviral therapy may need to be tailored accordingly.

Influenza Virus

Influenza virus infection may have several neurologic complications, including postviral fatigue, encephalitis, myelitis, and GBS.^32^ A distinguishing clinical feature of GBS associated with influenza virus compared with other viruses may be cranial nerve involvement (Miller Fisher syndrome), which seems to be more common after influenza virus infection when compared with Campylobacter jejuni infection*.* Consistent with this, patients with influenza with Miller Fisher syndrome also more frequently had antibodies to glycolipid GQ1b.^33^ Similar to some cases of SARS-CoV-2 infection, acute transverse myelitis (ATM) and acute necrotizing encephalitis may rarely occur after influenza infection while postinfluenza fatigue may be a persisting symptom.^34^

SARS-CoV-2

The presentation of SARS-CoV-2 infection, like infection by other neuroinvasive viruses, may range from an asymptomatic state to a critical illness, including high fever, dyspnea, palpitations, myalgias, weakness, fatigue, delirium, and stupor. Acute postinfectious SARS-CoV-2 syndromes are summarized in eTable 1.^35-44^ Among the more common presenting neurologic symptoms are headaches and meningismus, consistent with viral CNS invasion. In the first year of the pandemic, several ICU-hospitalized SARS-CoV-2–positive patients with encephalopathy and a prolonged comatose state had high CSF SARS-CoV-2 antibody titers with intrathecal antibody synthesis, disruption of the blood-brain barrier (BBB), and evidence for neurodegeneration based on 14-3-3 protein detection.^35^ In later large series, across the spectrum of mild-to-severe neurologic symptoms within the first 2–3 weeks of infection, the most consistent CSF findings were markers of BBB disruption (∼50% of patients), with pleocytosis (primarily lymphocytic) generally being more prevalent in patients with more severe neurologic symptoms.^36^ Severe acute SARS-CoV-2 infection is associated with robust CNS and systemic inflammation (‘cytokine storm’), glial activation (astrocytes, microglia), complement deposition, microthrombi, and hypoxia with some associated secondary neuronal loss.^37^ Other neuropathologic findings are leakage of fibrinogen into the perivascular region, macrophage infiltration, and deposition of antibodies and complement on endothelial cells. The brain endothelial cells also show expression of adhesion molecules and aggregates of activated platelets.^38^ The neuropathologic brain changes, if any, of mild or asymptomatic SARS-CoV-2 infection have not been directly observed, although BBB disruption is an associated phenomenon that often associates with the presence of cognitive symptoms.^39^ Some acute neurologic symptoms, such as encephalopathy, and rarer complications such as stroke and seizures may represent parainfectious sequelae associated with pathologic processes of microthrombi formation and hypoxia. Although clinical symptoms resolve within a month in approximately 90% of cases of SARS-CoV-2 infection, long-term sequelae (now known as long COVID), including cognitive impairment, have emerged as a major health problem worldwide.^40^

At the early peak of the COVID-19 pandemic, in 4 major hotspots, SARS-CoV-2–infected patients presented with GBS (n = 11), the prototypic viral-triggered acute autoimmune neurologic disease, highlighting a link between SARS-CoV-2 infection and autoimmune reactions against the peripheral nerves, muscle, cranial nerves, and brain even without neuroinvasion.^41,42^ Since then, several rarer autoimmune complications of SARS-CoV-2 infection have been acute disseminated encephalomyelitis (ADEM), acute hemorrhagic leukoencephalitis (AHLE), acute necrotizing encephalopathy (ANE), and ATM and acute necrotizing myelitis (ANM).^43-45^ Among these, ADEM and AHLE tend to occur in older patients with severe initial clinical presentations, and they typically present over a time range of hours to 30 days. In a series of 46 patients, typical MRI findings were T2/FLAIR (fluid attenuated inversion recovery) hyperintensities in the brain (95%) and spinal cord (54%), brain hemorrhage (42%), and contrast enhancement (36% brain, 22% spinal cord).^43^ Despite treatment with corticosteroids, IVIG, or plasmapheresis, high mortality rates (32%) were observed. Case reports of ANE describe a syndrome that also occurs (rarely) after other viral infections previously mentioned (influenza A, HSV-1, HHV-6) with the acute onset (1–3 days) of progressive obtundation and seizures and death. ATM is also a rare complication of COVID-19 infection and also COVID-19 vaccination.^45,46^ Longitudinally extensive cord lesions are commonly observed (55%–83%) in SARS-CoV-2 infection–associated ATM, and a generally good response to steroids is expected.^45^ Vaccines against SARS-CoV-2 are affective at preventing the major complications but are not fully effective in preventing infection.

The pathogenesis of neurotropic CNS virus infections is highly variable, ranging from cellular destruction by virus replications, production of proinflammatory and pro-oxidant mediators, and pathogenic immune responses. Neurotropic viruses (some arboviruses, HSV-1, VZV) cause neuronal injury through direct infection and lysis of neurons while others (HIV-1) can induce neuronal injury through indirect mechanisms such as glial cell infection and activation and associated release of neurotoxic mediators, cytokines, reactive oxygen species, and toxic viral gene products.^47,48^ One distinguishing feature of most cases of SARS-CoV-2 infection of the CNS from pathogenic arbovirus infections, such as Zika, WNV, and dengue, is that there is little, if any, evidence for direct infection of neurons by SARS-CoV-2.^47,49^ SARS-CoV-2 has been shown to infect endothelial cells in the lung, but not brain endothelial cells. Disruption of the BBB may be indirectly mediated by complement-dependent antibody toxicity.^38^

Multiple attempts have been made to identify SARS-CoV-2 genome or proteins in autopsied tissue from patients with COVID-19. Early autopsy studies did not demonstrate viral invasion of neurons despite robust infection of the nasal mucosa.^50^ A detailed rapid autopsy study of en bloc dissections of the nasopharynx, olfactory nerve, and bulb in 85 patients dying within one day of SARS-CoV-2 infection revealed virus only in sustentacular and ciliated epithelial cells within olfactory mucosa, with no virus in olfactory nerve terminals or the olfactory bulb.^49^ An autopsy study of 44 patients with COVID-19 demonstrated genomic and subgenomic SARS-CoV-2 fragments in several CNS regions (cervical cord, thalamus) at day 230 (one patient) but no replication-competent virus beyond day 13 in any patient's brain, indicating that replicating virus is likely cleared from the brain compartment with weeks of infection.^51^ These studies suggest rare persistence of SARS-CoV-2 RNA, but not infectious virus, in the brain after early clearance of the replicating virus. The significance of rare residual SARS-CoV-2 genomic components in the CNS remains unanswered. As with HIV-1 CNS infection, the possibility of production of SARS-CoV-2 viral proteins and associated pathologic responses (toxicity, immune responses) in the absence of replicating virus requires investigation but evidence for this is currently lacking.

Like many neurotropic virus infections, SARS-CoV-2 infection causes not only acute manifestations but also long-term neurologic sequelae in a subset of patients. Postacute sequelae of SARS-CoV-2 infection is now termed “long COVID”; a new definition was recently established by the National Academies of Sciences, Engineering and Medicine (NASEM) Committee on Examining the Working Definition for Long COVID.^52^ This working definition has been adopted by the CDC. Accordingly, long COVID (LC) is defined as an infection-associated chronic condition that occurs after SARS-CoV-2 infection and is present for at least 3 months as a continuous, relapsing and remitting, or progressive disease state that affects one or more organ systems.

The emerging health burden of LC has provoked a new awareness of the potential for persistent, long-term effects of CNS virus infection, even after viral clearance from the brain. Postinfectious fatigue, postexertional malaise, sleep disturbances, mood disorders, pain syndromes, and cognitive impairment have been described after other viral infections, including EBV, influenza, chikungunya, West Nile virus, dengue, and SARS-CoV-1, among others, but not with the consistency of SARS-CoV-2.^34^

The mechanisms that drive LC have not been identified with certainty. Some investigators consider it to be a neurodegenerative disorder, considering a large MRI study in the UK Biobank initiative (n = 401 patients with COVID-19, n = 385 controls) that compared pre–COVID-19 and post–COVID-19 brain volumes.^53^ These investigators found significant volume reductions in the orbitofrontal cortex, parahippocampal gyrus, and total brain size over a mean time between scans of approximately 3 years and mean duration of infection of 141 days. Many studies have attempted to identify biomarkers (serum, CSF, neuroimaging) to associate with or predict neurologic dysfunction in LC, but none has been validated. In vitro studies have shown that the SARS-CoV-2 spike protein and protease can cause aggregation of amyloid, tau, or synuclein proteins.^54^ Consistent with this, it is possible that SARS-CoV-2 infection can accelerate or unmask neurodegenerative diseases. In fact, elevated levels of tau, p-tau, and neurofilament and decreased levels of amyloid beta peptides in the CSF of patients with post–COVID-19 have been reported.^55^ As a chronic condition initiated by SARS-CoV-2 infection and without persistence of infectious virus within the CNS, what drives neurologic symptoms and complications? Evidence exists for a role of both innate and adaptive immunity.

Although activation of innate immune responses and neuroinflammation are features of acute COVID-19, they are not consistently observed in LC.^56^ Severe acute COVID-19 is associated with robust inflammation (‘cytokine storm’), BBB disruption, lymphocytic and monocytic infiltration, astrocytic and microglial activation, complement deposition, microthrombi, and hypoxic injury with neuronal loss. Neuropathologic studies of patients with LC have not been published. Although some patients with LC-associated cognitive dysfunction have elevated blood markers of inflammation and immune activation (IL-6, TNF-α, IFN-γ, and IL-1β and others), up to 40% of patients with LC do not.^57,58^ Analysis of CSF of patients with LC and of autopsied brains of COVID-19 decedents provide inconsistent evidence for neuroinflammation.^51,59^ This suggests that not all patients with LC demonstrate inflammation in different body compartments, including the CNS.^58^ These studies, along with functional neuroimaging (PET) studies,^e1^ suggest that CNS neuroinflammation and innate immune responses may play a role in neurocognitive symptomatology in some patients with LC, but they also suggest that pathologic processes other than neuroinflammation drive symptoms in many.

Enhanced autoimmunity and induction of specific antiviral immunity are also common consequences of SARS-CoV-2 infection.^e2,e3^ Autoimmune antibodies directed against antigens in CNS neurons, glia, endothelial cells, and others have been detected in infected patients, although whether they drive neuropathologic responses is unknown. Some antibodies may modulate neuronal and/or glial cell functions in vitro. An early case study reported MOG antibodies in an acutely ill patient with COVID-19 with rapid clinical deterioration and MRI evidence for inflammatory vasculopathy (perivascular enhancement surrounding T2 hyperintense lesions).^e4^ Rapid and sustained improvement was observed after IV methylprednisolone and plasma exchange. One study of patients with post–COVID-19 detected functional serum antibodies against the muscarinic M2 acetylcholine receptor (30/31 patients) with neurologic symptoms.^e5^ Notably, these antibodies produced negative chronotropic effects in cultured cardiac myocytes. This effect was blocked by atropine, thus providing evidence for the antibody's specific effects on the acetylcholine receptor. In the same study, functionally active serum antibodies against alpha-adrenergic and beta-adrenergic receptors and nociceptive receptors were also detected in most patients. These results suggest an antibody linkage to LC symptoms, such as postural orthostatic tachycardia syndrome, other autonomic symptoms, and pain syndromes through modulating effects on neurotransmitter receptors.

Additional studies have detected autoantibodies directed against proteins that may modulate signaling pathways in patients with post–COVID-19. A study of 194 patients with COVID-19 (compared with uninfected individuals) found a high prevalence of serum antibodies against immunomodulatory proteins (cytokines, chemokines, complement components) and cell surface proteins, some of which may perturb CNS receptor signaling.^e2^ Reactivity against CNS antigens was demonstrated, including the orexin receptor (HCRTR2), which regulates wakefulness and appetite, suggesting a possible link to sleep disorders and fatigue in LC.

In addition to detection of autoantibodies in serum, autoantibodies have been observed in CSF from patients with COVID-19. A CSF study of hospitalized patients with COVID-19 (n = 7) with various neurologic symptoms detected anti-neuronal antibodies in 5 patients.^56^ Subcloning of CSF B lymphocytes from several patients identified clonal production of an antibody (not found in peripheral blood) directed against the SARS-CoV-2 spike protein and additional antibodies that labeled (in mouse brain slices) neurons in the olfactory bulb, cortex, hippocampus, brainstem, cerebellum, and cerebral vasculature. Immunoprecipitation analyses of CSF from 2 patients identified 2 cellular antigens bound by autoantibodies, transport protein 88 homolog and THAP domain-containing 3, which are of unclear significance to the genesis of neurologic dysfunction. Notably, none of these 7 patients had neuroinflammation by CSF analysis or neuroimaging, thus emphasizing the potential for multiple pathogenic mechanisms underlying LC neurologic dysfunction. The challenge of demonstrating a cause-effect relationship between autoantibody production and neurologic sequelae in patients with post–COVID-19 remains.

Similar to CNS complications potentially linked to innate and adaptive immunity/autoantibodies, peripheral nervous system complications of SARS-CoV-2 infection have also been linked to immune responses. Among these are small fiber neuropathy (SFN) and GBS.^e6-e8^ Antiganglioside antibodies have been observed (one patient), which suggests cross-reactivity between peripheral nerve glycolipids and epitopes within the COVID-19 spike-bearing gangliosides that mediate binding to target cell gangliosides.^41^ Similar molecular mimicry has been shown between peripheral nerve glycolipids and Campylobacter jejuni or Zika virus, which also triggers GBS.^41^ Some COVID-19–triggered GBS cases have improved with IVIG, similar to other viral-triggered GBS cases. In post–COVID-19 SFN cases, however, conclusions about treatment responses are not possible, because of the small number of cases and lack of disease controls. In a retrospective chart review study of 16 patients with SFN symptoms (10 confirmed by skin biopsy), improvement in symptoms with IVIG treatment was reported in most.^e7^ No link to markers of autoimmunity or inflammatory markers was established. Another study of 17 patients referred for post-COVID-19 neuropathy evaluation identified that at least 10 patients had SFN confirmed by skin biopsy.^e8^ Five of these patients with SFN improved with IVIG, and 2 reported improvement with corticosteroids alone. Controlled studies with larger sample sizes and systematic evaluation of inflammatory and innate immunity markers are clearly needed.^e6^ The presence of a wide variety of autoantibodies after COVID-19 suggests that the infection causes a polyclonal B-cell activation, which is seen in a variety of infections, which are often self-limiting. However, in others, it may unmask an immune-mediated disorder.

Do Cross-Reactive Anti–SARS-CoV-2 Antibodies Play a Role in COVID-19 Neuropathogenesis?

Antiviral antibodies elicited after infection play an important role in neutralization of virus infectivity, typically by binding to glycoproteins on the surface of the virion (receptor-binding domain antibodies) that interact with target cell receptors. SARS-CoV-2 infection elicits a variety of virus-reactive antibodies, including neutralizing antibody responses against the receptor-binding domain. These antibodies, along with CD8^+^ T-lymphocyte responses, aid in clearing the virus from tissue compartments, including the CNS. One untoward effect of some antiviral antibodies is reactivity against epitopes expressed in CNS cells. Analysis of CSF and blood from patients with COVID-19 with neurologic symptoms showed CSF antibodies against SARS-CoV-2 epitopes that were distinct from paired serum-derived anti–SARS-CoV-2 antibodies from the same patients.^56^ One CSF-produced monoclonal antibody that reacted to the virus spike protein cross-reacted to neural tissue. This is the first report of naturally produced cross-reactive SARS-CoV-2 antibody within the CNS, and it suggests that, like other neurotropic viruses, the presence of replicating SARS-CoV-2 virus and also viral antigens (spike protein) may induce pathogenic responses. Additional work is needed to determine the prevalence and pathogenic potential of such cross-reactive SARS-CoV-2 antibodies.

Summary

The past decade has seen the emergence of highly pathogenic infections that have challenged neurologists and the entire health care system. The long-term consequences of these pandemics are largely neurologic, and hence, they require long-term care. We are ill-equipped to handle this crisis.^e9^ We need targeted treatments based on our understanding of the current pathophysiology of the disease. While we have made substantial progress in the clinical characterization and in understanding the pathophysiology of these illnesses, we lack disease-modifying therapies and the medical personnel to provide long-term care. We are even less prepared to handle the next pandemic. It is thus critically important for us to develop an organized plan and invest resources toward managing these crises.