Cytomegalovirus in Pregnancy and the Newborn

Cytomegalovirus is the commonest congenital infection and the leading non-genetic cause of childhood sensorineural hearing loss (SNHL). Clinicians typically encounter it in three distinct settings, each demanding a specific diagnostic approach:

• The pregnant woman: the primary task is to recognise the infection and assess the risk of transmission to the fetus.
• The fetus: the focus shifts to determining whether the virus has crossed the placenta and, if so, whether it has caused developmental harm.
• The newborn: the goal is to confirm congenital infection, distinguish it from postnatal acquisition, and identify which infants require treatment.

Running through all three is the relationship between the timing of maternal infection and the outcome: transmission becomes more likely later in pregnancy, while the most severe damage follows infection in the first trimester. That gradient shapes who is screened, who is offered antiviral prophylaxis, and who needs the closest long-term follow-up.

Risk and transmission

Two variables set the risk of transplacental infection: the mother’s serostatus and the gestational age at which she is infected. A primary infection in a previously seronegative woman carries the highest per-event risk of transmission, but pre-existing immunity does not protect the fetus completely, because a seropositive woman can still transmit through reactivation of her latent virus or through reinfection with a different strain. Because the seroimmune population is so large, about two-thirds of congenital cytomegalovirus arises in already-immune (non-primary) mothers, even though each such pregnancy carries a much lower individual risk than a primary infection.

The relationship between timing and outcome contains a paradox that is central to management. The rate of transmission to the fetus rises steadily through gestation, from ~9% for a periconceptional infection to ~40% in the third trimester. The severity of damage moves in the opposite direction. Severe neurological injury follows almost only first-trimester infection, while after the first trimester the risk falls sharply and is largely limited to hearing loss. A virus acquired late in pregnancy is therefore more likely to reach the fetus but far less likely to harm it seriously.

Transmission and sequelae by timing of maternal primary infection

Timing of maternal primary infection	Transmission to the fetus	Risk of sequelae if the fetus is infected
Preconception	~9%	Low
Periconceptional	~19%	Low
First trimester	~31%	~32%
Second trimester	~34%	Severe neurological disease very rare; hearing loss can still occur
Third trimester	~40%	Severe neurological disease very rare; hearing loss can still occur

One corollary matters at the cot side. Once the fetus is infected, whether the maternal infection was primary or non-primary no longer predicts the outcome: the rates of symptomatic disease at birth and of long-term sequelae are similar either way. Serostatus governs the risk of transmission, not the severity of disease once transmission has occurred. Globally, cytomegalovirus infects 1 to 4% of pregnancies, the birth prevalence of congenital infection is around 0.6 to 0.7%, and 10 to 15% of congenitally infected newborns are symptomatic at birth while the remainder are silent.

Diagnosis in the three settings

Diagnosis has three distinct aims, each with its own specimen and its own pitfalls: the infected mother, the infected fetus, and the infected newborn. The newborn carries the extra task of separating congenital infection from infection acquired during or after birth, a distinction that turns entirely on timing.

The pregnant woman

The aim of maternal testing is to recognise infection early, ideally in the first trimester, while there is still time to intervene. Serology is the tool, and its interpretation rests on the kinetics of the antibody response: immunoglobulin M (IgM) appears within one to two weeks of infection, immunoglobulin G (IgG) follows a week or two later and then persists for life, and the avidity of that IgG matures slowly, reaching high avidity by about six months. Two traps recur. A positive IgM is weakly specific, since only 10 to 30% of pregnant women with a positive IgM actually have a primary infection, the rest reflecting persistent IgM, reactivation, or assay cross-reactivity. And current serology cannot risk-stratify a non-primary infection at all, because the mother is already IgG-positive.

Avidity testing resolves much of the ambiguity, with the important caveat that it dates an infection rather than excluding one. Low-avidity IgG indicates a recent infection, within the last three to four months, and therefore the possibility of fetal transmission. High avidity indicates an infection more than six months old, but a high-avidity result obtained after 20 weeks of gestation cannot exclude a first-trimester primary infection, because avidity will already have matured by the time the sample is taken. The European consensus algorithm formalises this into four serological starting points.

Interpretation of maternal cytomegalovirus serology

Result	Interpretation and next step
IgG negative, IgM negative	Susceptible: retest every four weeks until 14 to 16 weeks to catch seroconversion
IgG positive, IgM negative	Past infection: a first-trimester or periconceptional primary infection is excluded with high probability
IgG positive, IgM positive	Test IgG avidity: high avidity cannot exclude a first-trimester infection, low avidity indicates a probable recent primary infection
IgG negative, IgM positive	Blood polymerase chain reaction now plus repeat serology in 10 to 15 days: seroconversion confirms primary infection, its absence excludes it

The fetus

Once a maternal primary infection is established, the question becomes whether the fetus is infected. Imaging is insensitive: ultrasound detects only ~15% of infected fetuses, and magnetic resonance imaging (MRI) adds an anomaly in ~6% of those with a normal ultrasound. The findings, when present, include periventricular calcification, ventriculomegaly, microcephaly, hyperechogenic bowel, growth restriction, ascites and hepatosplenomegaly.

The definitive prenatal test is polymerase chain reaction (PCR) on amniotic fluid, which works because the infected fetus excretes virus in its urine into the amniotic fluid. Two intervals must both be satisfied for the test to be reliable: it should be performed at least 6 to 8 weeks after the maternal infection, and from around 17 weeks of gestation, by which time placental transfer and fetal renal excretion have had time to occur. Under those conditions its sensitivity is 90 to 95%. The residual false-negative rate is ~8%, and reassuringly those missed cases tend to be the ones without fetal injury.

The newborn

In the newborn the diagnosis must be both made and dated. Virus must be detected within the first three weeks of life to prove congenital infection, because after three weeks a positive urine or saliva result can no longer be distinguished from infection acquired through the birth canal or breast milk. Urine is the specimen of choice, ahead of saliva, which is easier to collect but prone to a false positive from cytomegalovirus in breast milk retained in the mouth; a positive saliva result is therefore confirmed on urine, and saliva is collected at least one to two hours after a feed. Blood is unreliable, since only 10 to 20% of infected newborns are viraemic at birth, and neonatal IgM is too insensitive to use. A useful rule works in the other direction: a negative CMV IgG in the mother or the newborn effectively excludes congenital infection.

Beyond three weeks of age the diagnosis becomes retrospective. PCR on the newborn-screening dried blood spot is the standard tool here: a positive result confirms congenital infection, but a negative one does not exclude it, since many mildly infected infants were not viraemic when the spot was taken. This same logic separates congenital from postnatal infection: a newborn who tests negative at birth and positive only later, after the onset of symptoms, has acquired the virus postnatally rather than in utero.

A confirmed infant is then evaluated for end-organ disease: a full examination including head circumference, a full blood count, liver and renal function, quantitative CMV viral load, an auditory brainstem response, an ophthalmology examination, and neuroimaging, beginning with cranial ultrasound in every infant and adding MRI where the examination, imaging or hearing is abnormal. Lumbar puncture is not routine, and a negative result does not reassure, since CMV is detectable in the cerebrospinal fluid in only 13 to 15% of infants even when the central nervous system is involved.

The infected newborn: clinical features and classification

About 90% of congenitally infected newborns are asymptomatic at birth. The symptomatic minority present with some combination of growth restriction, microcephaly, a petechial or purpuric rash, the blue-red nodules of a blueberry muffin rash, jaundice, hepatosplenomegaly and an abnormal neurological examination, against a laboratory background of thrombocytopenia, anaemia, neutropenia, raised transaminases and a conjugated hyperbilirubinaemia. Chorioretinitis is the characteristic ocular lesion, while cataracts and microphthalmos are not typical of cytomegalovirus and should prompt a search for another cause.

Clinical findings in symptomatic congenital cytomegalovirus

Finding	Frequency at birth
Petechiae or purpura	50 to 75%
Jaundice	40 to 70%
Hepatosplenomegaly	40 to 60%
Small for gestational age	40 to 50%
Microcephaly	35 to 50%
Sensorineural hearing loss at birth	~35%
Chorioretinitis	10 to 15%
Raised transaminases	50 to 83%
Thrombocytopenia	48 to 77%

Neuroimaging is abnormal in ~70% of symptomatic infants, and the abnormalities fall into two groups: inflammatory and destructive lesions from the infection itself (lenticulostriate vasculopathy, germinolytic pseudocysts, ventriculomegaly, periventricular calcification and white-matter change), and disorders of brain development (cortical malformation such as polymicrogyria, and cerebellar hypoplasia). Around 8 to 10% of symptomatic newborns have life-threatening disease, a sepsis-like syndrome with hepatitis, pneumonitis and haemophagocytic features, and ~30% of these most severely infected infants die.

Three clinical categories drive management and follow-up: symptomatic disease (any sign on examination, laboratory testing or imaging), asymptomatic infection (confirmed virus, no signs), and isolated hearing loss (SNHL with no other sign, treated as a distinct group because it is neither truly silent nor part of overt disease). Hearing loss is the defining sequela across all of them. It affects around a third to a half of symptomatic infants and 10 to 15% of those asymptomatic at birth, and because it is frequently delayed in onset and progressive, up to a quarter of asymptomatic infected children have hearing loss by the age of 18. Every infected infant therefore needs structured audiological follow-up, with hearing assessed every three to six months in early childhood and then at least annually.

Management

Management follows the same host-by-host logic, with antiviral drugs used in two quite different ways: to reduce transmission in the pregnant woman, and to treat established disease in the newborn.

In pregnancy

When a maternal primary infection is confirmed but fetal infection has not yet been established, the goal is to prevent transmission. Oral valaciclovir, 8 grams a day given as 2 grams four times daily, reduces transmission to the fetus after a first-trimester primary infection (an odds ratio of ~0.37 overall, and ~0.34 for first-trimester infection), and shifts the balance of outcomes towards milder, asymptomatic infection. It is started as soon as possible after diagnosis and before 16 weeks, in women with a primary infection in the first trimester or periconceptional period, and continued until a negative amniotic-fluid result; renal function and a full blood count are checked at baseline and every two weeks, the main toxicity being a reversible renal impairment in ~2%. The benefit is clearest for first-trimester infection and is not established for periconceptional or later infection.

Cytomegalovirus hyperimmune globulin, by contrast, does not work: two randomised trials found no reduction in transmission or disease, and it has been abandoned for this purpose. When fetal infection is confirmed, management turns to prognostication rather than prevention, using serial fetal ultrasound and a third-trimester MRI. Normal fetal imaging has a negative predictive value approaching 100% for moderate-to-severe sequelae, but it leaves a residual risk of ~17% of unilateral hearing loss, so normal imaging is reassuring about severe disease without excluding hearing impairment.

The newborn

Treatment of the infected newborn is decided by phenotype. Antiviral therapy is given for six months when there is any central nervous system involvement, defined broadly to include abnormal neuroimaging, microcephaly, chorioretinitis or hearing loss; for symptomatic disease without central nervous system involvement, such as an isolated hepatitis or thrombocytopenia, six weeks is sufficient. Isolated hearing loss is treated, with valganciclovir started before four to six weeks of age, beyond which there is no evidence of benefit. Isolated growth restriction is not treated, and an asymptomatic infant with normal hearing is not treated either.

Antiviral treatment of the infected newborn

Clinical picture	Treatment	Duration
Any central nervous system involvement (abnormal imaging, microcephaly, chorioretinitis or hearing loss)	Oral valganciclovir, or intravenous ganciclovir if severe or the enteral route is unavailable	Six months
Symptomatic without central nervous system involvement	Oral valganciclovir	Six weeks
Isolated sensorineural hearing loss	Oral valganciclovir, started before four to six weeks of age	Six weeks
Isolated growth restriction, or asymptomatic with normal hearing	None	None

The drug of choice is oral valganciclovir at 16 mg/kg per dose twice daily, which achieves levels equivalent to intravenous ganciclovir; intravenous ganciclovir at 6 mg/kg per dose twice daily is reserved for severe disease or an infant who cannot take the oral solution, and is switched to oral therapy as soon as possible and avoided beyond six weeks. Treatment is reserved for infants above 1.8 kg or beyond 32 weeks of gestation at delivery, with specialist advice and drug-level monitoring for smaller or more premature babies, and is started as soon as possible after birth, before four weeks of age where the diagnosis is known, and up to 12 weeks where it is delayed. A randomised trial established that six months of valganciclovir preserves hearing and improves neurodevelopmental scores better than six weeks.

The dominant toxicity is neutropenia from bone-marrow suppression, severe in ~18% of treated infants, and most likely in those given intravenous ganciclovir for more than two weeks, born prematurely, or starting with a low neutrophil count. The neutrophil count is checked at baseline, at two and four weeks, and again after treatment, and is acted on by threshold.

Neutrophil-count thresholds during antiviral treatment

Absolute neutrophil count	Action
Above 1.0 ×10⁹ per litre	Continue with routine monitoring
0.8 to 1.0 ×10⁹ per litre	Continue, recheck in two weeks
0.5 to 0.8 ×10⁹ per litre	Continue, recheck in one week
Below 0.5 ×10⁹ per litre	Stop, recheck in one to two weeks, restart once above 0.8 to 1.0

Treatment response is tracked by the blood viral load, which usually falls over the first two to four weeks. Failure of the viral load to fall, or a sustained rise alongside progressive end-organ disease, defines refractory infection and should prompt genotypic testing for ganciclovir resistance and a search for an underlying immunodeficiency. Most ganciclovir-resistant strains remain susceptible to foscarnet, the preferred second-line agent, given as an induction course of 60 mg/kg per dose every eight hours before maintenance, with cidofovir a further option; both demand close attention to renal function, and sight-threatening chorioretinitis may additionally need intravitreal therapy.

Children at risk, namely those infected in the first trimester or at unknown timing, those symptomatic at birth, and those with hearing loss, chorioretinitis or abnormal imaging, have a formal neurodevelopmental assessment at 24 to 36 months and audiological follow-up to at least five to six years of age. A child with a documented second or third-trimester maternal infection is at much lower risk of serious sequelae and the intensity of surveillance is tailored accordingly, though hearing is still monitored because hearing loss can follow infection at any stage.

Postnatal cytomegalovirus

Cytomegalovirus acquired after birth is a different disease. It is acquired most often through breast milk, and less often through cervicovaginal secretions at delivery or through blood transfusion. In the healthy term infant it is benign, usually silent and at most a transient illness with no demonstrable neurodevelopmental consequence. The exception is the very preterm or very-low-birth-weight infant, in whom postnatally-acquired infection can cause a sepsis-like syndrome with hepatitis, pneumonitis and cytopenias in ~15% of those infected, occasionally fatal and associated with necrotising enterocolitis and severe retinopathy of prematurity. Such disease is treated like congenital disease but for a shorter course, typically three to six weeks.

Prevention in the vulnerable preterm infant works on the routes of acquisition. Freezing and thawing expressed breast milk reduces transmission without eliminating it, and pasteurisation reduces it further at the cost of some beneficial components; transfusion-transmitted infection is prevented by using cytomegalovirus-safe blood, meaning leucoreduced or seronegative products, for all neonates. For a healthy term infant the benefits of breastfeeding outweigh the small risk, and the milk need not be treated.

Prevention

Three levers reduce the burden of congenital cytomegalovirus. The first is behavioural: since the main route of maternal infection is contact with the saliva and urine of young children, hygiene measures, chiefly handwashing and not sharing food or utensils, substantially reduce the risk of a primary infection, though they do nothing for non-primary infection. The second is vaccination, which remains unrealised: no cytomegalovirus vaccine is licensed, the glycoprotein B and pentameric-complex surface targets are understood, but the most advanced candidate, the replication-defective V160, fell short of its efficacy target. The third is screening. The case for routine first-trimester maternal serology rests on the timing of the disease: because the harm that can be prevented is concentrated in first-trimester infection, where valaciclovir works, early screening is the precondition for acting at all. Newborn screening, whether targeted at infants who fail the hearing screen or applied universally, mainly serves to identify children at risk of delayed-onset hearing loss who would otherwise be missed.

South African context

South Africa sits at the high-seroprevalence, high-HIV-prevalence end of this disease. Most women of childbearing age are already immune, so the congenital burden is driven less by maternal primary infection than by non-primary infection across a very large immune population, and the local birth prevalence is correspondingly high. A Gauteng cohort of 2,685 neonates found a congenital cytomegalovirus prevalence of ~2.5%, several times the global figure, and a rural Eastern Cape cohort of 302 neonates found ~6%; the two disagreed on whether HIV exposure raised the risk, which the Gauteng study found and the Eastern Cape study did not.

The interaction with HIV is the locally distinctive feature. In the Gauteng cohort congenital cytomegalovirus was commoner in HIV-exposed than HIV-unexposed neonates, at ~5.2% against ~1.4%, and a congenitally infected HIV-exposed neonate carried roughly a twentyfold higher risk of acquiring HIV in utero, independent of maternal CD4 count or treatment duration, so a positive cytomegalovirus result in this setting flags a pregnancy at high risk of vertical HIV transmission. The high-seroprevalence, near-universal breastfeeding context also shapes the virology, with reinfection from breast milk and compartmentalised viral strains between the cervix and breast milk described in South African women. Against this burden, routine maternal cytomegalovirus screening and newborn cytomegalovirus screening are not established in the public programme, and access to valganciclovir for treatment is limited, so practice centres on recognising and confirming disease in the symptomatic newborn.

References and recommended reading

• Leruez-Ville M, Chatzakis C, Lilleri D, et al. Consensus recommendation for prenatal, neonatal and postnatal management of congenital cytomegalovirus infection from the European congenital infection initiative (ECCI). The Lancet Regional Health – Europe 2024;40:100892. The principal source for the maternal serology algorithm, the prenatal-diagnosis intervals, the valaciclovir and neonatal valganciclovir regimens, and the timing-based prognostic framework.
• Demmler-Harrison GJ. Congenital cytomegalovirus infection: clinical features and diagnosis; and Management and outcome. UpToDate; 2026. The source for the clinical-feature frequencies, the neonatal diagnostic logic, and the treatment thresholds and monitoring schedules.
• Boppana SB, Hui SL. Cytomegalovirus infection in pregnancy. UpToDate; 2025. The source for maternal diagnosis and the prevention of transmission.
• Pathirana J, Groome M, Dorfman J, et al. Prevalence of congenital cytomegalovirus infection and associated risk of in utero HIV acquisition in a high-HIV prevalence setting, South Africa. Clinical Infectious Diseases 2019;69(10):1789-1796. The source for the South African birth prevalence and the cytomegalovirus and in-utero HIV association.
• Tshabalala D, Newman H, Businge C, et al. Prevalence and determinants of congenital cytomegalovirus infection at a rural South African central hospital in the Eastern Cape. Southern African Journal of Infectious Diseases 2018;33(4):89-92. The source for the rural Eastern Cape birth prevalence.