Cytomegalovirus

Human cytomegalovirus, usually abbreviated to CMV or HCMV, is a betaherpesvirus carried for life by most of the world’s population, named for the massive cellular enlargement (cytomegaly) it induces.

In immunocompetent hosts, primary infection, lifelong latency, and periodic reactivations remain remarkably asymptomatic, despite consuming an unprecedented share of the host’s immune resources. However, HCMV exhibits significant pathogenicity when cellular immunity is immature or compromised. It is the leading infectious cause of congenital malformations and non-genetic childhood sensorineural hearing loss, and drives severe end-organ disease in transplant recipients and patients with advanced HIV.

It carries the largest genome of any human herpesvirus, a substantial part of which is devoted to manipulating and evading host immunity, and despite decades of effort no vaccine against it is licensed.

Discovery and historical significance

The enlarged cells bearing dense intranuclear inclusions were described in pathological specimens around 1900, defining the clinical entity cytomegalic inclusion disease long before the agent was known. The virus could not be isolated until human cell culture became available in the mid-1950s. It was isolated independently by three different investigators and named because of its cytopathic effect, the visible damage to cells in culture, which produced the large, swollen, refractile cells that give the cytomegaly for which it is named.

Later observations shaped current understanding. Serial studies in the 1970s showed that the quantity of virus excreted in the urine of congenitally infected infants tracks disease severity, an early sign of the viral-load relationship that now underpins diagnosis and treatment. Work on live-attenuated vaccines produced the attenuated Towne strain and the more virulent Toledo strain, a pairing that still informs vaccine design.

Classification, structure, and genome

Classification

Cytomegalovirus is the genus Cytomegalovirus in the subfamily Betaherpesvirinae, within the family Orthoherpesviridae, until recently named Herpesviridae. Its current International Committee on Taxonomy of Viruses (ICTV) species name is Cytomegalovirus humanbeta5, superseding the widely used Human betaherpesvirus 5 and the older human herpesvirus 5. Strains differ at their most variable genes but form neither distinct genotypes nor multiple serotypes: they are a single antigenic mosaic against which immunity is broadly cross-reactive. The closest human relatives are the other betaherpesviruses, human herpesvirus 6 and human herpesvirus 7.

Virion structure

The virion has the three layers common to all herpesviruses: an icosahedral capsid, a protein tegument, and a lipid envelope. The capsid is about 125 nanometres across, built on a triangulation number of 16 from the major capsid protein, with a single portal for genome insertion. The tegument delivers regulatory proteins into the cell at entry; its most abundant protein, the phosphoprotein pp65, makes up about a sixth of the virion, is the antigen detected in the antigenaemia test, and is a dominant T-cell target, while pp71 switches on the first viral genes. The envelope carries glycoprotein complexes built on glycoprotein B and on the glycoprotein H and L pair, the latter combining with either glycoprotein O or a five-protein assembly, the pentamer.

Genome organisation

The genome is a single linear molecule of double-stranded DNA of about 235 kilobases, the largest of any human herpesvirus, encoding roughly 165 canonical genes. A unique long and a unique short region, each flanked by inverted repeats, can invert to give four genome isomers of no clinical consequence. A 22-kilobase segment, ULb prime, is dispensable in cultured fibroblasts and lost rapidly on passage; its genes drive immune evasion and cell tropism, and their loss attenuates long-passaged laboratory strains.

Replication cycle

Attachment begins with binding to cell-surface proteoglycans before the fusion machinery engages its receptors, and entry follows either by direct fusion at the plasma membrane or by fusion after endocytosis, depending on the cell type. Glycoprotein B with the glycoprotein H and L pair drives fusion. The partner completing that pair sets the cellular range: the glycoprotein O form enters fibroblasts, while the pentamer (glycoprotein H and L with UL128, UL130 and UL131A) enters the epithelial and endothelial cells central to spread within and between hosts.

Once the genome reaches the nucleus, gene expression follows the ordered herpesvirus cascade. Immediate-early genes come first and encode the master regulators, principally immediate-early 1 and immediate-early 2 from a single strong promoter, which disable cellular defences and switch on the next phase; early genes supply the replication enzymes; late genes supply the structural proteins. The virus replicates its genome with its own machinery, which makes those enzymes, above all the DNA polymerase, the principal antiviral targets. Progeny genomes are cut to unit length and packaged through the capsid portal by the viral terminase. Capsids bud through the inner nuclear membrane, acquire tegument and envelope at a cytoplasmic compartment, and are released. The cycle is slow in cultured fibroblasts, long giving the virus a reputation for sluggishness, but brisk in the host, with a doubling time of about a day.

Pathogenesis

Cytomegalovirus replicates in a wide range of tissues, recovered at autopsy of disseminated disease from lung, adrenal glands, brain, gut, liver and salivary glands, and it grows in epithelial and endothelial cells, smooth muscle and several classes of leucocyte. This breadth underlies the diversity of its syndromes.

Lifelong latency defines its natural history. Between episodes of replication the genome persists, chiefly in the CD34 bone-marrow progenitors and the monocytes derived from them, as a quiescent circular molecule expressing almost no genes. Reactivation is tied to cellular differentiation: as a latent monocyte matures into a macrophage, the chromatin over the major immediate-early promoter opens and productive infection resumes. This is why inflammation, immune activation and immunosuppression, all of which drive monocyte differentiation, are the settings in which the virus reawakens.

Much of the genome serves immune evasion rather than replication, centred on downregulating major histocompatibility complex (MHC) class I, the molecule that displays viral peptides to cytotoxic T cells. Several viral proteins act together: one holds class I in the endoplasmic reticulum, one blocks the transporter associated with antigen presentation (TAP) that loads it with peptide, and two return class I to the cytosol for destruction. Loss of class I would expose the cell to natural killer (NK) cells, which kill cells that have lost it, the missing-self response, so the virus deploys decoys: a class I mimic, and a protein supplying a peptide that stabilises the inhibitory ligand HLA-E, keeping the natural killer cell switched off. It also encodes chemokine-receptor mimics, an interleukin-10 lookalike that dampens inflammation, complement regulators, and proteins that strip the stress ligands by which natural killer cells recognise infected targets. The scale of this programme measures how completely the virus has co-evolved with human immunity.

Disease tracks viral load: the relationship between peak load and end-organ disease is steep, viral load is the strongest independent predictor, and serostatus and viraemia serve as markers of how much virus a person carries. The interval from acquisition to clinical illness is about four to eight weeks.

Epidemiology

Cytomegalovirus is found in every population. Seroprevalence ranges from about 40 to 99 per cent by region and socioeconomic conditions, rising with age and falling with affluence. Where it is high, as across much of Africa, Asia and Latin America, infection is acquired in early childhood and is near-universal by adolescence. Where acquisition comes later, a substantial minority of women reach childbearing age still seronegative and so susceptible to a first infection in pregnancy. Hence the paradox of congenital cytomegalovirus: the severe fetal disease that follows maternal primary infection is relatively more frequent in affluent settings, though the virus is most prevalent in the poorest.

Transmission needs close contact with infected secretions. Saliva is the principal route in childhood, and young children in group care readily infect carers and seronegative mothers. The virus also spreads sexually, in genital secretions and semen, with seroprevalence approaching saturation in sexually active groups. It passes to the child across the placenta, at delivery, and in breast milk, the last a common but usually harmless route by which about half of the breastfed infants of seropositive mothers are infected. It is transmitted by cellular blood products and transplanted organs, routes now largely controlled. Reinfection with a second strain occurs even in the immune, and there is no seasonal pattern.

Natural history

A single primary infection is followed by lifelong carriage. The virus is never cleared; it establishes latency and reactivates intermittently for life, shedding into saliva, urine, genital secretions and breast milk. At any time about a tenth of the population, and far more young children, are shedding. In the immunocompetent host the primary infection and these reactivations are almost always silent, yet they sustain transmission and, in pregnancy, allow spread to the fetus; reactivation becomes more frequent through pregnancy and peaks near term.

Whether infection ever causes disease turns on three things: the competence of the host’s immunity, which arm of infection is involved, and the viral load reached. Mature immunity contains the virus indefinitely. In the fetus, in advanced HIV infection, and under immunosuppressive therapy, control fails and the virus reaches the loads at which it injures tissue.

Clinical presentations and complications

Congenital cytomegalovirus infection

Cytomegalovirus infects about one in a hundred newborns and is the commonest infectious cause of damage to the developing child. Risk to the fetus depends on the type of maternal infection. A primary infection in pregnancy transmits to the fetus in roughly 30 to 40 per cent of cases, and about 10 to 15 per cent of those infants show clinical abnormalities. A non-primary infection, reactivation or reinfection in an already immune woman, transmits in around one per cent and seldom causes more than isolated hearing loss, because maternal immunity is substantially protective. Because immune women far outnumber those having a first infection, most congenitally infected babies are nonetheless born to immune mothers, each pregnancy carrying only a low individual risk.

Most infected newborns appear healthy. About an eighth are born with overt cytomegalic inclusion disease: growth restriction, jaundice, hepatosplenomegaly, thrombocytopenia with a petechial or purpuric rash, microcephaly, intracranial calcification, and chorioretinitis. A few of these severely affected infants die, and about half of survivors are left with permanent neurological impairment such as developmental delay, spastic weakness, seizures or visual loss. The signature injury is sensorineural hearing loss, and it is not confined to symptomatic infants: a meaningful fraction of those normal at birth develop hearing loss that is often progressive and may not appear until two to four years of age. In all, close to a fifth of congenitally infected children sustain lasting damage.

Perinatal infection

A child may also acquire cytomegalovirus around birth, from genital secretions at delivery or, more often, from breast milk, with urinary excretion appearing after about three weeks of life, which distinguishes it from congenital infection. It is usually harmless in the healthy term infant. The exceptions are the premature, very-low-birthweight infant and the infant of a seronegative mother, who lacks transplacental antibody; in them perinatal infection, including infection acquired from transfused blood, can cause a sepsis-like illness or pneumonitis.

Infection in the immunocompetent host

A first infection in an older child or adult is nearly always silent. When it causes illness, the usual picture is a mononucleosis like that of Epstein–Barr virus but negative for heterophile antibodies: prolonged fever, malaise, a biochemical hepatitis, and atypical lymphocytes on the blood film. Sore throat and prominent lymphadenopathy, typical of Epstein–Barr virus, are uncommon, and the illness is self-limiting, tending to occur about a decade later in life than Epstein–Barr mononucleosis. Serious disease in the immunocompetent host is rare but recognised, including hepatitis, gastrointestinal ulceration and bleeding, and an association with Guillain–Barré syndrome.

End-organ disease in advanced HIV infection

Before effective antiretroviral therapy, cytomegalovirus was among the most feared opportunistic infections of the acquired immunodeficiency syndrome (AIDS), affecting a quarter or more of patients, and it remains a threat wherever HIV goes untreated and the CD4 T-cell count falls below about 50 cells per microlitre. The commonest manifestation is retinitis, rare in other settings, presenting with floaters or visual loss and showing white retinal necrosis with haemorrhage that threatens sight through retinal detachment; restoring immunity with antiretroviral therapy can itself inflame the affected eye. Next most often the gut is involved, with ulcerating oesophagitis causing painful swallowing or colitis causing pain, diarrhoea and bleeding, at times mimicking inflammatory bowel disease. Less often the virus causes a diffuse encephalitis resembling HIV-associated dementia or a rapidly progressive ventriculoencephalitis, or infects the nerve roots to produce a subacute ascending weakness with a neutrophil predominance in the spinal fluid. The adrenal glands are frequently involved. Pneumonitis, by contrast, is uncommon in AIDS, because the lung injury depends partly on an immune response these patients cannot mount.

Disease in the transplant recipient

Cytomegalovirus is the single most important viral pathogen of organ and stem-cell transplantation. In solid-organ recipients it causes fever with cytopenias, hepatitis, gastrointestinal disease and, especially after lung transplantation, pneumonitis, and it harms the graft indirectly by promoting rejection and predisposing to other infections. In stem-cell recipients the dominant threat is reactivation of the recipient’s own virus, and pneumonitis is the most dangerous outcome, often aggravated by graft-versus-host disease.

Cytomegalovirus and the ageing immune system

Lifelong carriage leaves its mark. In older people, decades of containing cytomegalovirus generate a large population of differentiated virus-specific T cells that crowds the immune repertoire, contributing to the decline of immune function with age and associated with increased mortality in the elderly.

Diagnosis

Quantitative real-time polymerase chain reaction (PCR) for cytomegalovirus DNA is the mainstay for active infection, viral load and the response to treatment. The older antigenaemia assay detects the tegument protein pp65 in blood leucocytes and is rapid, but it is labour-intensive and unreliable when the white-cell count is low, and it is largely replaced by PCR. Culture in human fibroblasts is slow, taking up to three weeks, though it can be accelerated by detecting early viral antigens and is rapid for the high-titre urine of a congenitally infected infant. In tissue, the diagnosis rests on the enlarged cells bearing owl’s-eye intranuclear inclusions or, more sensitively, on immunohistochemistry.

Serology has defined uses rather than a role in active disease. Immunoglobulin G (IgG) marks past infection and carriage. Immunoglobulin M (IgM) indicates recent or current infection; it is too non-specific for universal screening in pregnancy, but has a targeted role, in the pregnant woman with a prolonged unexplained mononucleosis-like illness and in selected high-risk women with a known exposure, such as those caring for young children. A positive IgM is also the prerequisite for the decisive test in pregnancy, IgG avidity, whose binding strength matures over the sixteen to twenty weeks after a primary infection, so that low-avidity antibody marks a recent first infection, the situation of greatest fetal risk, while high avidity shows long-standing infection.

Two settings demand discipline. Congenital infection must be confirmed by PCR on urine or saliva within the first three weeks of life; later, a positive result cannot separate congenital infection from the common, benign infection acquired around birth. Where a maternal primary infection threatens the fetus, prenatal assessment combines ultrasound with PCR on amniotic fluid, a high fluid viral load indicating established fetal infection.

Management

Treatment targets the virus’s replication enzymes. First line are ganciclovir, given intravenously, and its oral prodrug valganciclovir. Cytomegalovirus lacks the thymidine kinase that activates aciclovir against other herpesviruses, so ganciclovir is activated instead by the viral UL97 protein kinase before it blocks the viral DNA polymerase; its main toxicity is bone-marrow suppression, particularly of the white cells and platelets. Foscarnet and cidofovir inhibit the same polymerase by other means and are limited mainly by nephrotoxicity. Two newer agents act elsewhere: letermovir blocks the terminase that packages the genome and spares the marrow and kidneys, and maribavir inhibits the UL97 kinase and is used for resistant or refractory infection. Resistance under prolonged treatment arises most often by UL97 mutation and less often by DNA-polymerase mutation, some forms of which cross-resist the older agents.

Two non-transplant settings warrant note. In the infant with symptomatic congenital infection, especially with central nervous system involvement, several months of oral valganciclovir improve hearing and neurodevelopmental outcomes and are the standard of care, a short course being less effective. In advanced HIV infection the essential step is restoring immunity with antiretroviral therapy; specific antivirals treat active end-organ disease such as retinitis, and maintenance can stop once the CD4 count is durably restored.

Prevention and public health

Vaccination

No cytomegalovirus vaccine is licensed, though one has long been a recognised priority. The obstacles are intrinsic: natural immunity is only partial, since reinfection and reactivation occur in the immune, so a vaccine must outperform natural infection; latency lies beyond the reach of the immune response; primary infection is usually silent and hard to target; and there is no good animal model. The candidate with most clinical experience, a subunit vaccine of glycoprotein B with an oil-in-water adjuvant, halved the rate of primary infection in seronegative women and transplant candidates without yet yielding a licensed product. Live-attenuated strains and their recombinants, vectored vaccines and nucleic-acid vaccines have all been explored.

Infection prevention and control

Without a vaccine, prevention interrupts transmission to those at risk. Seronegative pregnant women can be advised on hygiene, chiefly hand-washing and care with the saliva and urine of young children, since a toddler in group care is the commonest source of a maternal primary infection; trained healthcare staff are at no special occupational risk and need only standard hygiene. For transfusion-dependent risk groups, the seronegative pregnant woman, the premature infant and the immunocompromised patient, cytomegalovirus-seronegative blood or the removal of white cells from cellular products by filtration essentially eliminates transfusion-transmitted infection. In transplantation, matching a seronegative recipient to a seronegative donor avoids the highest-risk pairing where allocation allows, with antiviral strategies covering the rest.

Population screening has a narrower place. Routine antenatal serological screening is not generally recommended, because a positive antibody result does not reliably identify which pregnancies will produce an infected infant, and detecting primary infection must be weighed against the anxiety and intervention it generates. Newborn screening, often linked to hearing-loss detection, is increasingly discussed as a way to find the infants who would benefit from early treatment and audiological follow-up.

South African context

No South African data appear in the standard references, but the general epidemiology points clearly to the local picture. In high-density, lower-income populations cytomegalovirus seroprevalence approaches saturation and infection is acquired in early childhood, so most South African women enter pregnancy already immune. The congenital burden is therefore driven less by maternal primary infection than by non-primary infection in immune women, a lower per-pregnancy risk applied across very many pregnancies. High maternal HIV prevalence adds a further dimension, both because cytomegalovirus end-organ disease accompanies advanced untreated HIV and because the two viruses interact, high cytomegalovirus loads being linked to poorer HIV outcomes.

References and recommended reading

• Griffiths P, Reeves M. Cytomegalovirus. In: Richman DD, Whitley RJ, Hayden FG (eds.), Clinical Virology, 4th edition. Washington: ASM Press; 2016. The principal source for the virology, pathogenesis, congenital outcomes, diagnosis and treatment 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 transmission, congenital epidemiology and the rationale for immunoglobulin G avidity testing.
• Goodrum F, Britt WJ, Mocarski ES. Cytomegaloviruses. In: Fields Virology, 7th edition. Wolters Kluwer; 2023. The current reference for global seroprevalence and the contemporary antiviral landscape, including letermovir and maribavir.