Epstein-Barr virus

Epstein-Barr virus, usually abbreviated to EBV, is a gammaherpesvirus carried for life by more than nine in ten adults worldwide, and the first virus shown to cause a human cancer.

In most people primary infection is silent and is followed by lifelong, asymptomatic latency in memory B cells. When primary infection is delayed to adolescence, as it increasingly is in wealthier settings, it presents as infectious mononucleosis. The virus’s lasting importance is oncogenic: it growth-transforms B lymphocytes and lies behind a family of malignancies, Burkitt lymphoma, nasopharyngeal carcinoma, Hodgkin lymphoma, several other lymphomas and gastric carcinoma, that together place it among the most important infectious causes of human cancer. It is now also regarded as a necessary cause of multiple sclerosis.

No vaccine is licensed, and antiviral drugs do not alter the common illness it causes.

Discovery and historical significance

The clinical syndrome came first. Infectious mononucleosis was named in 1920, its atypical blood lymphocytes were described soon after, and in 1932 Paul and Bunnell showed that patients’ sera contain heterophile antibodies, the basis of the diagnostic test still used. The link to cancer came by a separate route: in 1961 Denis Burkitt mapped a childhood jaw tumour across equatorial Africa and saw that its distribution followed temperature and rainfall, hinting at an insect-borne infection. In 1964 Anthony Epstein, Bert Achong and Yvonne Barr found herpesvirus particles by electron microscopy in cells cultured from one of those tumours, the first virus shown to cause a human cancer. Within a few years the virus had been tied to infectious mononucleosis and shown to transform B cells into immortal cell lines, and its genome was fully sequenced in 1984.

Classification, structure, and genome

Classification

Epstein-Barr virus is the prototype of the genus Lymphocryptovirus, in the subfamily Gammaherpesvirinae, within the family Orthoherpesviridae (until recently named Herpesviridae). Its current International Committee on Taxonomy of Viruses (ICTV) species name is Lymphocryptovirus humangamma4, superseding the earlier Human gammaherpesvirus 4 and human herpesvirus 4 (HHV-4). Its only close human relative is the other human gammaherpesvirus, Kaposi sarcoma-associated herpesvirus. Two types, 1 and 2, differ mainly in their nuclear-antigen genes; type 1 transforms B cells more efficiently and predominates outside equatorial Africa, but the two cause the same diseases.

Virion structure

The virion has the standard herpesvirus architecture: an icosahedral nucleocapsid, a protein tegument, and a lipid envelope. The envelope carries several glycoproteins, dominated by gp350/220, which attaches the virus to B cells, alongside the gp42 and gH/gL/gB machinery that completes entry.

Genome organisation

The genome is linear double-stranded DNA of about 172 kilobases, encoding roughly 80 to 100 proteins together with a set of abundant non-coding RNAs. On entering the cell the linear molecule circularises through its terminal repeats into a chromatin-coated episome, which the viral protein EBNA1 tethers to the host chromosomes and replicates with them; it is not integrated into the host genome. The number of terminal repeats is fixed when the episome forms, so it serves as a clonality marker that can prove a tumour arose from a single infected cell.

Replication cycle

Infection of a B cell begins with attachment, when gp350/220 binds complement receptor 2 (CD21, the receptor for the C3d fragment of complement) and tethers the virus to the surface. Penetration follows: gp42 engages an HLA class II molecule as coreceptor and the gH/gL heterodimer with gB drives membrane fusion, the virus entering through an endosome. Epithelial cells, which lack CD21, are attached and entered by other surface proteins, without gp42, and the resulting switch of tropism (virus made in a B cell carries little gp42 and infects epithelium well, while virus made in epithelium carries gp42 and infects B cells well) lets the two cell types amplify one another. Once in the cytoplasm the capsid is carried to the nucleus, where it uncoats and releases the genome.

There the virus faces its central choice. In the lytic cycle the immediate-early proteins BZLF1 and BRLF1 switch on the early genes that replicate the genome and then the late genes that build the capsid, tegument and envelope; progeny virions assemble, the cell dies, and virus is released into the saliva. Far more often, though, the virus does not complete this cycle: the genome circularises into a chromatin-bound episome and enters latency, expressing only a restricted set of genes and producing no virus at all. The particular combination of genes expressed defines a small number of latency programmes, each tied to a stage of B-cell life and to a particular tumour.

Pathogenesis

EBV’s central capacity is to drive a resting B cell to proliferate. In its full latency programme the virus reproduces the signals a B cell would normally need from antigen and from helper T cells. The membrane protein LMP1 behaves as a constitutively active CD40 receptor, switching on the survival and proliferation pathway NF-kB without any ligand. LMP2A supplies the tonic signal a B cell normally receives through its antigen receptor, rescuing cells that would otherwise die. The nuclear antigens EBNA2 and EBNA-LP start the programme, further EBNA proteins drive the cell cycle, and abundant viral micro-RNAs blunt apoptosis; the virus even carries its own copy of the human anti-apoptotic gene bcl-2. A newly infected B cell is thereby pushed to proliferate, passes through the germinal centre, and settles as a memory B cell in which the virus rests for life.

The virus does not express this whole repertoire at once. The combination of genes a cell shows defines its latency programme, and the programme determines both how visible the cell is to the immune system and which tumour it can give rise to.

Programme	Genes expressed	Typical setting	Associated tumour
Latency 0	EBERs only	resting memory B cells (the reservoir)	none; immunologically silent
Latency I	EBNA1 and EBERs	dividing memory B cells	Burkitt lymphoma; gastric carcinoma
Latency II	EBNA1, LMP1, LMP2 and EBERs	germinal-centre B cells	nasopharyngeal carcinoma; Hodgkin lymphoma
Latency III	all six EBNAs, LMP1, LMP2 and EBERs (the growth programme)	newly infected B cells; immunosuppression	post-transplant and HIV-associated lymphomas

Two principles follow. The more genes a programme expresses, the harder it drives proliferation but the more visible it becomes to EBV-specific T cells, so the full growth programme survives only where T-cell control has failed. And the EBV-driven cancers cluster precisely where that control is weak: in human immunodeficiency virus (HIV) infection, under transplant immunosuppression, and with the immune exhaustion of chronic malaria. The dependence on T cells is underscored by the inherited immunodeficiencies that predispose to severe EBV disease, almost all of which are defects of the T-cell or natural killer (NK) cell response.

Epidemiology

More than 90 per cent of adults are infected in every population studied. The age of infection tracks living conditions: in low-income settings most children are infected in their first few years, often before their first birthday, whereas in wealthier populations infection is increasingly delayed into adolescence and early adulthood. That delay is what generates infectious mononucleosis, which is in essence a primary infection acquired late, and its incidence is rising as acquisition is pushed later. Symptomatic mononucleosis peaks between the ages of 15 and 24; a substantial share of young adults entering higher education are still seronegative, around half of them acquire EBV over the following years, and most of those infections are symptomatic. There is no seasonal pattern and no sex difference in incidence. Two of the EBV cancers have a striking geography: endemic Burkitt lymphoma follows the equatorial-African belt where Plasmodium falciparum malaria is intense, where it is the commonest childhood cancer, and nasopharyngeal carcinoma is concentrated in southern China and parts of South-East Asia.

Natural history

Primary infection takes hold in the oropharynx, in tonsillar B cells and the overlying epithelium, and spreads through the lymphoid tissue of the throat. After the primary episode the virus is never cleared; it withdraws into resting memory B cells, where roughly one in a hundred thousand to a million carries the silent genome for the rest of the person’s life. From time to time, when an infected memory B cell differentiates into a plasma cell, the virus switches into its lytic cycle, reactivates and is shed into the saliva, the main route of spread, so that most infections are caught from healthy, asymptomatically shedding contacts. At any time perhaps an eighth to a quarter of healthy seropositive people are shedding virus, and the immunocompromised shed far more often. The incubation period of symptomatic primary infection is about six weeks, and by the time mononucleosis appears a large fraction of the circulating memory B cells are infected.

Clinical presentations and complications

Infectious mononucleosis

In young children EBV infection is asymptomatic or very mild. When it is delayed to adolescence or early adulthood, as it increasingly is in wealthier settings, primary infection often presents as infectious mononucleosis. After a long incubation of four to seven weeks the illness begins insidiously with headache, malaise and fatigue, then settles into its three regular features: a high, fluctuating fever, pharyngitis, and generalised lymphadenopathy. The pharyngitis is often striking, with a grey-white exudate over the tonsils that is occasionally severe enough to threaten the airway; the spleen is frequently enlarged and the liver enzymes mildly abnormal. The acute illness lasts two to three weeks, though convalescence and fatigue can be protracted. The symptoms come not from the virus directly but from the vigorous T-cell response to it, and the atypical lymphocytes on the blood film are reactive T cells, not infected B cells.

Two rashes should be distinguished. About a tenth of patients develop a faint, widespread, non-itchy maculopapular rash that lasts roughly a week as part of the illness itself. Quite separate is the intense, itchy rash over the extensor surfaces and pressure points that very often follows treatment with amoxicillin or ampicillin (but not penicillin), a hypersensitivity reaction rather than a true penicillin allergy. The enlarged spleen can rupture, spontaneously or after minor trauma, in roughly 0.1 to 0.5 per cent of cases; rare but potentially fatal, this is the basis for advising against contact sport. A wide range of further complications is occasionally seen: neurological forms including Guillain-Barré syndrome, Bell’s palsy, meningoencephalitis and transverse myelitis; haematological ones including haemolytic anaemia and thrombocytopenia; and carditis, nephritis and pneumonitis. A small minority of patients go on to a prolonged, sometimes disabling, chronic fatigue.

Infection in the immunocompromised host

Where the T-cell control of EBV fails, infected B cells proliferate unchecked. The inherited prototype is X-linked lymphoproliferative syndrome, in which affected boys meeting EBV for the first time mount a catastrophic response: an overwhelming proliferation of infected B cells with marrow failure, and in those who survive, hypogammaglobulinaemia or B-cell lymphoma. Acquired failure is far more common, in advanced HIV, in transplant recipients and in immunodeficient children. The infection may look like mononucleosis or present atypically as pneumonitis or hepatitis; infants with AIDS develop a lymphocytic interstitial pneumonitis, and a primary central nervous system lymphoma is characteristic of advanced HIV. In transplant recipients the same process produces post-transplant lymphoproliferative disorder, a spectrum from a reactive, mononucleosis-like overgrowth of infected B cells to frank lymphoma.

EBV-associated malignancies

EBV is implicated in a family of cancers. Burkitt lymphoma presents in three distinct forms. The endemic African form is almost universally EBV-positive and strongly linked to heavy malaria exposure, which drives B-cell proliferation while weakening EBV-specific immunity. The sporadic form, found worldwide, is EBV-positive in only a minority of cases. The immunodeficiency-associated form mainly complicates HIV infection. All three share a translocation that places the c-myc oncogene under the control of an immunoglobulin gene, classically t(8;14), an accident of the antibody-gene machinery during the germinal-centre reaction.

The wider family of EBV-associated cancers, with the latency programme each expresses:

Malignancy	EBV-positive	Latency	Notes
Burkitt lymphoma (endemic)	almost all	I	African jaw tumour of children; malaria cofactor; t(8;14)
Burkitt lymphoma (sporadic)	minority	I	worldwide
Burkitt lymphoma (HIV-associated)	about a third	I
Nasopharyngeal carcinoma	essentially all	II	southern China; middle-aged men
Hodgkin lymphoma	variable	II	EBV in Reed-Sternberg cells; higher in children, low-income settings and HIV
Diffuse large B-cell lymphoma	a subset	II or III
Gastric carcinoma	about one in ten	I or II	a very common cancer, so large in absolute number
Extranodal NK/T-cell lymphoma, nasal type	almost all	I or II	aggressive; common in Asia
Immunosuppression-related lymphomas	high	III	post-transplant and HIV-associated, including primary CNS and plasmablastic

Because gastric cancer is itself so common, EBV-positive gastric carcinoma may be the single most frequent EBV-associated malignancy by number of cases. A previous attack of infectious mononucleosis raises the later risk of EBV-positive Hodgkin lymphoma severalfold. In nasopharyngeal carcinoma a rising serum immunoglobulin A (IgA) antibody to the viral capsid antigen can precede the tumour and is used to screen for it in high-incidence regions.

Other associations

Several non-malignant conditions also belong to EBV. Oral hairy leukoplakia, a corrugated white lesion on the side of the tongue, is productive EBV replication in the immunosuppressed, classically in HIV. Chronic active EBV disease, marked by months of fever, lymphadenopathy and a very high viral load, and EBV-driven haemophagocytic lymphohistiocytosis, an overwhelming activation of macrophages with fever, falling blood counts and organ dysfunction, are rare but severe disorders in which the virus infects T or NK cells rather than B cells; both carry a poor prognosis and may require stem-cell transplantation. EBV is now also regarded as a necessary cause of multiple sclerosis: a large prospective cohort found that infection raised the risk of multiple sclerosis around thirtyfold, with markers of nerve injury rising only after infection, and the leading explanation is that the immune response to the viral nuclear antigen EBNA1 cross-reacts with a protein in the central nervous system.

Diagnosis

In a typical case the blood film shows a lymphocytosis with many atypical (reactive) lymphocytes, and the diagnosis is confirmed by a heterophile antibody test, the Paul-Bunnell or Monospot test, which detects antibodies that agglutinate animal red cells. The heterophile test has limits: it is often negative in young children, can be negative in the first week or two of illness, and gives occasional false positives in liver disease, lymphoma and autoimmune conditions. Where it is unhelpful, EBV-specific serology both confirms infection and dates it. Antibody of the immunoglobulin M (IgM) class against the viral capsid antigen appears at the onset of symptoms and fades within weeks to months, marking an acute primary infection; antibody of the immunoglobulin G (IgG) class against the same antigen rises early and persists for life as a marker of past infection; antibody to the early antigen accompanies active replication; and antibody to the Epstein-Barr nuclear antigen (EBNA) appears only weeks to months into convalescence and then persists, so its presence at first testing shows the illness is not a primary infection. Read together, these antibodies stage the infection.

Antibody pattern	VCA IgM	VCA IgG	EA IgG	EBNA IgG	Interpretation
Susceptible	absent	absent	absent	absent	never infected
Acute primary infection	present	present	present	absent	current or recent first infection
Past infection	absent	present	absent	present	infected long ago, immune
Reactivation	absent or present	present, often high	present	present	reactivation or chronic active infection

(VCA is the viral capsid antigen, EA the early antigen; EA antibody can persist at low levels in a minority of healthy people, so it supports rather than proves active infection.) Detecting EBV DNA by polymerase chain reaction (PCR) is unhelpful in the healthy, who carry the virus anyway, but valuable for surveillance in the immunocompromised and as a quantitative marker for the EBV-driven tumours; in tissue, the abundant EBER RNAs are detected by in-situ hybridisation, the standard way to prove a tumour is EBV-related.

Management

Infectious mononucleosis is treated supportively, with rest, fluids and analgesia, and with advice to avoid contact sport while the spleen is enlarged. Antiviral drugs do not help: aciclovir and its relatives act only on the virus’s lytic replication, whereas the illness is immune-mediated and the viral load has already peaked by the time it presents, so they shorten neither the illness nor its course, even though they reduce shedding. Corticosteroids are reserved for specific complications such as threatened airway obstruction, not given routinely. EBV-driven lymphoproliferative disease is approached first by restoring immune control, reducing immunosuppression where that is the cause, and by depleting the infected B cells with the anti-CD20 antibody rituximab; transferring EBV-specific cytotoxic T cells can restore control directly, with the best results in stem-cell recipients whose disease runs the full growth programme, and is increasingly available off the shelf from banks of donor cells. Cancers that express fewer viral genes, such as nasopharyngeal carcinoma and Hodgkin lymphoma, give the transferred T cells fewer targets and respond less well; these are treated by oncology with radiotherapy and chemotherapy, as their non-viral counterparts are. No antiviral clears latency, and strategies to force the latent virus into its drug-sensitive lytic phase remain investigational.

Prevention and public health

Vaccination

No EBV vaccine is licensed. Part of the difficulty is that the protective target is unclear: natural infection prevents neither reinfection nor the cancers, so a vaccine must do something natural immunity does not. The most studied candidate, a subunit vaccine based on the attachment glycoprotein gp350, reduced the incidence of infectious mononucleosis by about three-quarters in seronegative young adults but did not prevent infection itself. Newer candidates present several entry glycoproteins together, on nanoparticles or by messenger RNA (mRNA), to raise stronger neutralising responses, and are in clinical trials. The recognition that EBV is a necessary cause of multiple sclerosis has sharpened interest in a prophylactic vaccine, since preventing infection might prevent the disease.

Infection prevention and control

Little can be done about transmission in the community, since the virus is shed in the saliva of healthy people for life and most infection is acquired silently in childhood, so no routine screening or isolation is warranted. In the transplant setting the recipient’s EBV serostatus is checked before grafting, and a seronegative recipient of an organ from a seropositive donor, the highest-risk pairing, is monitored for rising EBV DNA so that immunosuppression can be adjusted early.

South African context

The standard references carry no South African data, but the general epidemiology indicates the local picture. In lower-income, higher-density populations EBV is acquired in early childhood and seroprevalence is near-universal, so primary infection is largely a silent childhood event rather than adolescent mononucleosis. The classic endemic Burkitt lymphoma of equatorial Africa is tied to intense Plasmodium falciparum malaria, and most of South Africa lies south of that holoendemic belt, with malaria confined to the low-lying north-east, so the endemic form is not the typical South African pattern. The South African burden of EBV malignancy instead follows the large HIV epidemic: HIV-associated lymphomas, including primary central nervous system lymphoma and plasmablastic lymphoma, and the higher EBV-positivity of Hodgkin lymphoma in this setting.

References and recommended reading

• Fingeroth JD. Epstein-Barr Virus. In: Richman DD, Whitley RJ, Hayden FG (eds.), Clinical Virology, 4th edition. Washington: ASM Press; 2016. The principal source for the virology, infectious mononucleosis, the malignancies, diagnosis and management set out here.
• Burrell CJ, Howard CR, Murphy FA. Fenner and White’s Medical Virology, 5th edition. Academic Press / Elsevier; 2017. The concise account of the herpesvirus group, the natural history and the clinical features.
• Gewurz BE, Longnecker R, Cohen JI. Epstein-Barr Virus. In: Fields Virology, 7th edition. Wolters Kluwer; 2022. The current reference for the epidemiology of the EBV cancers and the vaccine landscape.
• Damania B, Kenney SC, Raab-Traub N. Epstein-Barr Virus: Biology and Clinical Disease. Cell 2022;185(20):3652-3670. DOI 10.1016/j.cell.2022.08.026. The current review underpinning the latency-to-tumour map, the oncogenic mechanisms, and the multiple sclerosis association.