Viral Diseases of the Central Nervous System

Headache, lethargy and impaired concentration accompany many systemic viral infections, but invasion of the central nervous system itself is uncommon. When it does occur it matters out of all proportion to its frequency, because the central nervous system tolerates injury poorly and recovers slowly and often incompletely. Tumours, infections and autoimmune processes in the brain produce overlapping signs, and the patient’s history rarely identifies the cause on its own. The clinician’s real task is therefore one of triage: to separate the small number of treatable conditions, above all herpes simplex encephalitis and the non-viral mimics that respond to antibacterial, antifungal or antituberculous therapy, from the larger number of self-limited or currently untreatable viral infections, and to do so quickly enough to prevent avoidable damage.

Classifying central nervous system infection

The vocabulary of neurological viral disease follows the anatomical site of inflammation. Inflammation of the leptomeninges is meningitis, of the brain parenchyma is encephalitis, of the spinal cord is myelitis, and of the dorsal nerve roots and peripheral nerves is radiculitis and neuritis. The combinations are common: most patients with encephalitis have some meningeal involvement, and the term meningoencephalitis reflects this. The phrase aseptic meningitis is a long-standing misnomer for a benign, self-limited meningitis with a sterile routine bacterial culture; it is imprecise because the same picture is produced by fungal, tuberculous and other treatable infections, not only by viruses.

A second axis is the mechanism. Primary encephalitis follows direct viral entry into the brain, where viral replication is the initiating event and injury then comes from a combination of cytopathic effect and the host immune response; viral antigen and nucleic acid are demonstrable in the parenchyma. Post-infectious or para-infectious encephalomyelitis, by contrast, follows or accompanies a systemic infection without direct viral invasion of the brain. Its pathology is perivascular demyelination with immune-cell infiltration and no demonstrable virus, which is why it is regarded as an immune-mediated, probably autoimmune, process. A third category, the slow infections, runs over months to years and is considered below.

Two properties of a virus determine whether it causes neurological disease. Neuroinvasiveness is the capacity to enter the nervous system; neurovirulence is the capacity to cause damage once there. Their product is neurotropism. The two vary independently. Mumps virus is highly neuroinvasive, producing detectable cerebrospinal fluid changes in about half of all infections, but of low neurovirulence, so it rarely causes lasting harm. Herpes simplex virus is the opposite: rarely invasive, but devastating when it does invade. Not every neurotropic virus is neuronotropic, that is, able to infect neurons themselves: JC polyomavirus preferentially infects oligodendrocytes and causes demyelination rather than neuronal death.

The syndrome a patient presents with narrows the likely agents considerably, alongside the route by which each reaches the brain, the test that confirms it, and how it is treated. The acute syndromes and their agents are set out below; the slow infections are considered later.

Viruses by central nervous system syndrome

Virus	Route to CNS	Confirmatory test	Treatment
Meningitis
Enteroviruses (coxsackie, echo) and human parechoviruses	Blood	PCR of CSF	Supportive
Mumps virus	Blood	PCR or culture of CSF	Supportive
Herpes simplex virus type 2	Blood and neuronal	PCR of CSF	Aciclovir
Arboviruses (e.g. West Nile virus)	Blood	Serology, PCR	Supportive
Lymphocytic choriomeningitis virus	Blood	Serology, CSF culture	Supportive
Encephalitis
Herpes simplex virus type 1	Neuronal	PCR of CSF	Aciclovir (start empirically)
Arboviruses (Japanese encephalitis, West Nile, tick-borne, equine)	Blood	Serology (IgM), PCR	Supportive
Rabies virus	Neuronal	Antigen detection, PCR	Supportive; prevented by post-exposure prophylaxis
Measles virus	Blood	Serology, clinical	Supportive
Varicella-zoster, cytomegalovirus, Epstein-Barr, human herpesvirus 6	Blood and neuronal	PCR of CSF	Aciclovir for varicella-zoster; ganciclovir or foscarnet for cytomegalovirus
Enteroviruses, adenoviruses	Blood	PCR or culture	Supportive
Paralysis and acute myelitis
Polioviruses	Blood and neuronal	Culture, PCR	Supportive; vaccine-preventable
Enterovirus A71, coxsackievirus A7, enterovirus D68 (acute flaccid myelitis), enterovirus 70 (radiculomyelitis)	Blood	PCR	Supportive
Post-infectious encephalomyelitis
Measles, varicella-zoster, rubella, mumps, influenza	Immune-mediated; no direct CNS invasion	Clinical, MRI (virus absent from CNS)	Immunomodulation (unproven); supportive

How viruses reach and injure the central nervous system

A virus reaches the brain by one of two routes. The great majority arrive through the bloodstream. After crossing an epithelial surface and replicating locally, the virus seeds the regional lymphoid tissue or enters the circulation directly, producing a primary viraemia that marks the onset of illness. Replication in highly vascular organs, particularly the liver and spleen, then generates a sustained, high-titre secondary viraemia, and it is this that delivers virus to the central nervous system. Entry across the vascular barrier occurs by infection of endothelial cells, passive leakage through damaged endothelium, or carriage inside migrating leucocytes, the last being the route HIV and measles virus exploit.

The alternative is neuronal spread along peripheral nerves, which shields the virus from circulating immunity. Herpes simplex virus and rabies virus are the exemplars. Rabies virus replicates at the inoculation site, binds acetylcholine receptors to enter peripheral nerve endings, and travels by axonal transport to the brainstem and limbic system. Herpes simplex virus can reach the brain along the trigeminal or olfactory nerves, which is consistent with its predilection for the inferomedial temporal lobe.

Several barriers oppose this journey. Skin and mucosal defences, innate immune responses that curtail viraemia, and the fixed macrophages of the liver and spleen all reduce the viral load reaching the brain. The blood-brain barrier, a layer of tight endothelial junctions sheathed by glia, is the last of these. It is double-edged: by excluding lymphocytes, antibody and complement as well as virus, it also impedes viral clearance once infection is established, which helps explain why persistent infections so often involve the nervous system.

Whether infection produces disease, and where, depends on viral and host factors. Neurovirulence can map to single genes: deletion of the gamma-34.5 gene attenuates herpes simplex virus in the brain, and among the coxsackie B viruses, types B1 to B5 readily cause neurological infection while type B6 rarely does. Host age is decisive in several infections. Neonates with systemic enteroviral infection may develop overwhelming disease, with around one in ten dying and up to three-quarters of survivors left with sequelae, whereas the same viruses cause trivial illness in older children.

Injury itself arises in two ways. Some viruses kill neurons directly, with neuronophagia and an inflammatory infiltrate; the hallmark is the haemorrhagic necrosis of the inferomedial temporal lobe in herpes simplex encephalitis. Others, in the post-infectious syndromes, cause damage indirectly through an immune attack on myelin, with little or no virus present in the brain by the time disease appears.

Viral meningitis

Viral meningitis is far commoner than bacterial meningitis and, in the normal host, far less severe; only meningeal and ependymal cells are involved and recovery is almost always complete. Enteroviruses cause the majority of cases, with a summer and autumn peak in temperate climates that mirrors the seasonality of enteroviral infection generally. Arboviruses are the next commonest cause in many settings, occurring in seasonal epidemics that reflect vector activity. Mumps virus remains an important cause wherever immunisation is not practised, and meningitis may be its only manifestation, since only about half of cases of mumps meningitis follow recognisable parotitis. Lymphocytic choriomeningitis virus, acquired from rodents, is a rarer cause.

The presentation is often non-specific. Enteroviral meningitis typically gives a few days of fever, malaise and headache, with nausea or vomiting in about half. Although neck stiffness and photophobia are the cardinal signs, about a third of patients with viral meningitis have no meningismus at all, and in infants under two years meningeal signs are usually absent, the child presenting instead with fever and irritability. A high index of suspicion is therefore needed at the extremes of age.

Herpes simplex virus type 2 causes meningitis during primary genital infection in a minority of patients, more often women than men, and is the usual agent of Mollaret meningitis, a benign recurrent lymphocytic meningitis. The immunocompromised host is a special case: a blunted cellular response means the cerebrospinal fluid may show little inflammation despite genuine infection, and herpesviruses are more likely to progress from the meninges into the parenchyma.

Viral encephalitis

Encephalitis is among the most serious of viral diseases. It often begins like meningitis, but altered consciousness signals that the parenchyma is involved: the patient becomes lethargic, then confused and stuporose, and may develop ataxia, seizures, focal deficits and coma. Survivors are frequently left with epilepsy, cognitive impairment, paralysis, deafness or blindness. Unlike meningitis, encephalitis carries a high mortality that varies sharply with the agent and the patient’s age.

Two patterns coexist. Herpes simplex virus is the commonest cause of sporadic encephalitis, the isolated case that occurs without season or outbreak, and the one that most demands recognition because it is both treatable and, untreated, usually fatal.

The larger numbers worldwide are epidemic: the arboviruses, spread by mosquitoes and ticks, are the leading cause of encephalitis globally, with Japanese encephalitis virus foremost among them. Rabies, transmitted by an animal bite rather than by a vector, is the other major cause and is almost always fatal. In arboviral encephalitis asymptomatic infection greatly outnumbers symptomatic disease, and the case-fatality rate ranges from a few per cent to seventy per cent depending on the virus and the patient.

Herpes simplex encephalitis occurs at all ages and without seasonal variation. Its signature is a focal, haemorrhagic, necrotising encephalitis of the inferomedial temporal and frontal lobes, producing focal seizures, personality change, aphasia and a depressed level of consciousness; in the neonate the disease is instead diffuse. Untreated, herpes simplex encephalitis kills about seventy per cent of those affected, which is what makes its early recognition the central clinical imperative of this topic.

The clinical picture of other encephalitides reflects the regions each virus targets: rabies virus has a predilection for the limbic system, producing the alternating calm and agitation of furious rabies, while Japanese encephalitis virus targets the brainstem and deep grey matter, with tremor, dystonia and a parkinsonian syndrome.

Post-infectious encephalomyelitis, also called acute disseminated encephalomyelitis, is a demyelinating illness appearing one to two weeks after a systemic infection or, formerly, after vaccination with live vaccinia virus. It historically followed measles, mumps and the other exanthemata, and now follows influenza and varicella most often in immunised populations; measles remains the leading cause where vaccination coverage is low, complicating roughly one in a thousand infections. A related, immune-mediated process is now recognised after herpes simplex encephalitis: a proportion of patients develop antibodies to the N-methyl-D-aspartate receptor one to four weeks after the acute illness, producing a relapsing autoimmune encephalitis that complicates recovery.

Paralysis and acute myelitis

Several enteroviruses invade the anterior horn of the spinal cord and produce an acute flaccid paralysis. Poliovirus is the classic agent and remains a cause of both meningitis and paralytic poliomyelitis wherever it still circulates; the oral vaccine very rarely causes paralysis, chiefly in immunocompromised recipients. Enterovirus A71 and coxsackievirus A7 cause a sporadic paralytic disease essentially indistinguishable from poliomyelitis, and enterovirus 70 causes a usually reversible radiculomyelitis. Clusters of acute flaccid paralysis with anterior myelitis continue to occur in regions that have eliminated poliovirus, as in California in 2012 and 2013, with enterovirus D68 implicated; enterovirus D68 is notable for causing substantial morbidity where most enteroviruses do not. Because such cases may signal unsuspected poliovirus circulation, surveillance systems for acute flaccid paralysis exist to detect, investigate and exclude poliovirus and to characterise the other enteroviruses responsible.

Laboratory diagnosis

The lumbar puncture is the first investigation, and the cerebrospinal fluid picture orients the differential even though it cannot identify the cause on its own. Viral meningitis typically produces a clear fluid with a modest pleocytosis, a mildly raised protein and a preserved glucose, whereas pyogenic bacterial meningitis produces a turbid fluid with a high neutrophil count and a low glucose. The distinction is clinically vital but imperfect, because the two overlap: early in viral meningitis the cells may be predominantly neutrophils, shifting to lymphocytes over the first day or two, and a proportion of bacterial cases show atypical formulas. The fluid must therefore always be read alongside the clinical and epidemiological context.

Cerebrospinal fluid findings: viral versus bacterial meningitis

Parameter	Viral meningitis	Pyogenic bacterial meningitis
Appearance	Clear	Turbid
White cell count (per mm³)	Usually 30 to 300, range up to 500	Hundreds to thousands
Predominant cell	Lymphocytes (neutrophils in the first day or two)	Neutrophils
Protein	Near normal to mildly raised, under 100 mg/dL	Raised
Glucose	Normal, above 40 per cent of the serum value	Low
Caveat	About 40 per cent show early neutrophil predominance	About 15 per cent show lymphocyte predominance

Detection of viral nucleic acid by polymerase chain reaction on cerebrospinal fluid has replaced viral culture and serology for acute diagnosis, and is the central test for most viral infections of the central nervous system, including those due to herpes simplex virus, enteroviruses, varicella-zoster virus, Epstein-Barr virus, cytomegalovirus and JC polyomavirus. Its impact is clearest in herpes simplex encephalitis, where only about four per cent of cerebrospinal fluid cultures are positive but polymerase chain reaction has a sensitivity above ninety-five per cent and a specificity approaching one hundred per cent, allowing diagnosis without the brain biopsy on which it once depended. Serology retains a limited role, chiefly for arboviral infections, where cerebrospinal fluid or serum immunoglobulin M may be informative but is rarely available in time to guide acute management.

Encephalitis, unlike meningitis, usually needs imaging and electroencephalography as well. Magnetic resonance imaging is the modality of choice, being more sensitive than computed tomography to the early changes of encephalitis and to demyelination, and diffusion-weighted sequences detect parenchymal change earliest; the demonstration of periventricular and deep white-matter demyelination also helps separate a post-infectious process from acute viral encephalitis. Before lumbar puncture, imaging and fundoscopy should exclude a mass lesion and raised intracranial pressure. The electroencephalogram shows diffuse or focal epileptiform discharges with background slowing; temporal-lobe changes strongly support herpes simplex encephalitis, although their absence does not exclude it.

Differential diagnosis

The priority in a patient with a cerebrospinal fluid pleocytosis and a negative routine bacterial culture is to identify the treatable conditions that mimic viral disease. Several bacteria evade routine culture: the spirochaetes of syphilis, Lyme disease and leptospirosis, the mycobacteria of tuberculous meningitis, and fastidious organisms such as Listeria, which can produce a mononuclear pleocytosis resembling a viral picture. Partially treated bacterial meningitis, after self-medication with leftover antibiotics, can do the same, and a parameningeal focus such as an infected sinus or a brain abscess presents with fever, focal signs and pleocytosis.

Fungi are important, particularly in the immunocompromised host: Cryptococcus and Coccidioides are the leading causes of fungal meningitis, while Candida, Aspergillus and the endemic mycoses tend to cause focal parenchymal disease. Free-living amoebae such as Naegleria and Balamuthia cause a fulminant meningoencephalitis.

Among the non-infectious causes, malignant infiltration of the meninges, drug-induced aseptic meningitis from non-steroidal anti-inflammatory drugs, immunoglobulin or trimethoprim, and the autoimmune and paraneoplastic encephalitides, including limbic encephalitis and anti-N-methyl-D-aspartate-receptor encephalitis, all enter the differential. The yield of this search is real: in the classic series of patients with suspected herpes simplex encephalitis taken to brain biopsy, nearly half had an alternative diagnosis, and many of those were treatable.

Chronic and slow viral infections of the central nervous system

A small group of viral infections of the brain runs not over days but over months to years. Subacute sclerosing panencephalitis is a rare, late and fatal sequel of measles, in which defective virus persists in the brain; progressive rubella panencephalitis is a rarer counterpart following congenital or childhood rubella. Both are demyelinating persistent infections.

Progressive multifocal leukoencephalopathy is different in mechanism: reactivation of JC polyomavirus in the setting of advanced HIV infection or of immunosuppression for transplantation or malignancy destroys oligodendrocytes and demyelinates the white matter.

HIV itself is highly neuroinvasive from early in infection, and in advanced disease causes HIV-associated neurocognitive disorder, with progressive dementia, myelopathy and neuropathy, alongside the many opportunistic infections and tumours that affect the brain in the acquired immunodeficiency syndrome.

Human T-lymphotropic virus type 1, after an incubation that can extend to decades, causes a slowly progressive myelopathy of the thoracic cord known as tropical spastic paraparesis, with leg weakness, spasticity and sphincter disturbance.

Chronic and slow viral infections of the central nervous system

Disease	Virus	Onset	Mechanism
Subacute sclerosing panencephalitis	Measles virus	Years after measles	Persistent defective virus; demyelinating panencephalitis
Progressive rubella panencephalitis	Rubella virus	Years after congenital or childhood rubella	Rare persistent demyelinating infection
Progressive multifocal leukoencephalopathy	JC polyomavirus	During advanced HIV or immunosuppression	Lytic infection of oligodendrocytes; white-matter demyelination
HIV-associated neurocognitive disorder	Human immunodeficiency virus	Advanced HIV infection	Direct neuroinvasion; progressive dementia and myelopathy
Tropical spastic paraparesis	Human T-lymphotropic virus type 1	Up to decades after infection	Immune-mediated myelopathy of the thoracic cord

A separate group of slow, fatal brain diseases, the transmissible spongiform encephalopathies, are caused not by viruses but by misfolded prion protein; they are addressed in the companion prion diseases article.

Treatment and the approach to the patient

The approach to a patient with a suspected viral infection of the central nervous system is governed by one principle: find and treat the treatable.

After stabilising the airway, breathing and circulation and gauging the extent of neurological involvement, the clinician must vigorously exclude the treatable mimics, herpes simplex encephalitis chief among the viral causes and tuberculous, fungal, partially treated bacterial and parameningeal infection among the rest, before attributing the illness to an untreatable agent.

In practice this means starting intravenous aciclovir empirically whenever herpes simplex encephalitis is plausible and continuing it until the diagnosis is excluded, because the drug is most effective when given early and the cost of waiting is high: aciclovir has reduced the mortality of neonatal herpes simplex disease from about seventy to about forty per cent and lessened the neurological damage in survivors.

Aciclovir is also used for varicella-zoster virus infection of the nervous system, and ganciclovir and foscarnet for cytomegalovirus infection in the immunocompromised; specific dosing is a matter for current national guidelines and is not given here.

Most viral meningitis needs only supportive care, since no specific therapy has proven benefit; for post-infectious encephalomyelitis, immunomodulatory treatment is widely used but rests on case series rather than controlled trials.

Throughout, the complications of encephalitis demand active management: seizures, cerebral oedema and raised intracranial pressure, the syndrome of inappropriate antidiuretic hormone secretion, and the cardiac and respiratory instability that can accompany brainstem involvement.

Prevention remains the most effective intervention. Measles, mumps and rubella vaccination has greatly reduced the incidence of encephalitis where it is practised, poliovirus vaccination has eliminated paralytic poliomyelitis from most of the world, and vaccines exist for Japanese encephalitis and for rabies. Where no vaccine is available, vector control and avoidance reduce arboviral disease, and prompt post-exposure prophylaxis prevents rabies.

South African context

In the South African setting the differential diagnosis of a lymphocytic meningitis or a meningoencephalitis is dominated by the high prevalence of HIV. Tuberculous meningitis and cryptococcal meningitis are common and treatable causes that must be excluded early in any patient with a lymphocytic cerebrospinal fluid, and HIV-associated neurocognitive disorder and the opportunistic infections of advanced immunodeficiency, including progressive multifocal leukoencephalopathy and cytomegalovirus disease, are frequent. Cerebrospinal fluid polymerase chain reaction for the herpesviruses and enteroviruses is available through the National Health Laboratory Service, and a positive herpes simplex virus result confirms a diagnosis that should already have prompted empirical aciclovir. Rabies remains endemic and is invariably fatal once symptomatic, which places the emphasis on wound care and post-exposure prophylaxis. Acute flaccid paralysis is a notifiable event under the surveillance system that protects the country’s polio-free status, and a case demands urgent investigation to exclude poliovirus.

References and recommended reading

• Cassady KA, Whitley RJ. Viral Infections of the Central Nervous System. In: Richman DD, Whitley RJ, Hayden FG, editors. Clinical Virology, 4th edition, Chapter 3. ASM Press; 2017. The principal source for the definitions and classification of neurological viral disease, the pathogenesis of meningitis and encephalitis, the cerebrospinal fluid and polymerase chain reaction findings, the differential diagnosis, and the clinical approach to the patient.
• Burrell CJ, Howard CR, Murphy FA. Viral Syndromes. In: Fenner and White’s Medical Virology, 5th edition, Chapter 39. Academic Press / Elsevier; 2017. The source for the syndrome-based classification of viral diseases of the central nervous system (Table 39.5), the distinction between neuroinvasiveness and neurovirulence, and the chronic and slow infections including subacute sclerosing panencephalitis, progressive multifocal leukoencephalopathy and tropical spastic paraparesis.