Congenital cytomegalovirus infection: Clinical features and diagnosis

INTRODUCTION — Congenital cytomegalovirus (CMV) infection is common worldwide. It is the leading cause of nonhereditary sensorineural hearing loss (SNHL) and can cause other long-term neurodevelopmental disabilities, including cerebral palsy, intellectual disability, vision impairment, and seizures.

The clinical features and diagnosis of congenital CMV infection will be reviewed below. The management and outcome of congenital CMV infection, other TORCH infections, CMV in pregnancy, and CMV infections in older infants and children are discussed separately:

●(See "Congenital cytomegalovirus infection: Management and outcome".)

●(See "Overview of TORCH infections".)

●(See "Cytomegalovirus infection in pregnancy".)

●(See "Overview of cytomegalovirus infections in children".)

TERMINOLOGY — Infants with congenital CMV infection are classified according to the presence or absence of apparent symptoms at the time of birth.

●The term "symptomatic" refers to infants with one or more symptoms at birth. (See 'Symptomatic neonate' below.)

●We use the term "primary neurophenotype" to refer to patients with only central nervous system manifestations. This is a newly described category of symptomatic infants, and the use of this term in the published literature is inconsistent. (See 'Primary neurophenotype' below and 'Neuroimaging' below.)

●The term "asymptomatic" refers to infants with no apparent symptoms at birth, although some of these infants may develop hearing loss or subtle symptoms later in life. (See 'Asymptomatic infection' below.)

●The term "asymptomatic with isolated hearing loss" refers to infants with isolated hearing loss at birth but no other symptoms. In studies of congenital CMV infection after the advent of universal newborn hearing screening, categorization of these infants as "symptomatic" or "asymptomatic" is inconsistent. Historically, they were often classified as "asymptomatic" because the hearing loss may not have been detected at birth. However, with universal newborn hearing screening, many such infants come to medical attention in the newborn period. We consider these infants to represent a distinct category because they are not truly asymptomatic, but their disease is generally milder than that of symptomatic infants.

VIROLOGY — CMV is a member of the herpesvirus family, along with Epstein-Barr virus, herpes simplex viruses-1 and -2, varicella-zoster virus, and human herpesviruses (HHV)-6, -7, and -8. These viruses share structural properties, including a genome of double-stranded linear DNA, a virus capsid of icosahedral symmetry, and a viral envelope [1].

Like other herpesviruses, CMV has the biologic properties of latency and reactivation. With latent CMV infection, a low level of virus replication is detectable and the viral genome may be present in the peripheral blood mononuclear cells and bone marrow. Recurrent infections with CMV occur through reactivation of the host's endogenous strain of CMV or reinfection with a new exogenous strain. CMV strains may compartmentalize such that the strain in the urine, for example, may be different from that in the blood, eye, or central nervous system [1].

CMV replicates slowly, often taking as long as 24 hours to produce progeny in infected cells and several days to weeks to produce a visible cytopathic effect in laboratory cell lines. Clinical samples with high titer of virus, such as urine or saliva samples from congenitally infected newborns, have multiple foci of cytopathic effect visible within a few days of incubation, whereas clinical samples with low titer of virus may require one to three weeks of incubation to show classic focal cytopathic effect (picture 1).

EPIDEMIOLOGY — Congenital CMV infection occurs worldwide, with a prevalence of 0.6 percent in developed countries [2-4]. Approximately 40,000 infants are born with congenital CMV infection annually in the United States [1].

The rate of congenital CMV infection, but not necessarily congenital CMV disease, is proportional to the seroprevalence of CMV in women of childbearing years. In areas of high CMV seroprevalence (80 to 100 percent), congenital CMV infection rates range from 1 to 5 percent, whereas in areas of relatively low CMV seroprevalence (40 to 70 percent), congenital CMV infection rates range from 0.4 to 2 percent [1,2,5-8].

Maternal CMV infection during pregnancy most often results from close contact with young children, particularly children attending daycare centers [9,10]. The risk of vertical transmission to the fetus is far higher with primary maternal infection than with recurrent infection (32 versus 1.4 percent) [2]. Other factors that can influence transmission of CMV infection during pregnancy include maternal age and parity (increased risk in younger primigravid women) [2]. (See "Cytomegalovirus infection in pregnancy", section on 'Maternal issues'.)

Infants infected as a result of a primary maternal infection are more likely to have symptoms at birth and suffer long-term sequelae than those infected as a result of maternal recurrent CMV infection. The risk of hearing loss, however, appears to be similar in both groups. [3,11,12].

Sequelae appear to be more severe when infection is acquired earlier in pregnancy, particularly in the first trimester [13]. However, symptomatic congenital CMV infection may result from maternal infection at any time during pregnancy.

CLINICAL MANIFESTATIONS

In utero — The fetus may be silently infected or manifest CMV disease in utero. The diagnosis and management of congenital infection often begins while the fetus is still in utero. Findings on prenatal ultrasound that may suggest CMV disease in the fetus include periventricular calcifications (image 1), ventriculomegaly (image 2), migrational abnormalities of the brain (polymicrogyria, pachygyria, and lissencephaly), microcephaly, hyperechogenic fetal bowel (image 3), fetal growth restriction, ascites and/or pleural effusion (image 4A-B), and hepatosplenomegaly. These and other findings are reviewed in a separate topic review. (See "Cytomegalovirus infection in pregnancy", section on 'Ultrasound markers suggestive of congenital CMV infection'.)

Symptomatic neonate — Approximately 10 percent of neonates with congenital CMV infection have symptoms at birth. Clinical findings in the symptomatic neonate can be similar to those in other congenital infections (table 1).

Clinical findings

Common findings — Clinical findings in the symptomatic neonate (picture 2) are similar to those in other congenital infections [7,8,14-17]:

●Petechiae (50 to 75 percent)

●Jaundice at birth (40 to 70 percent)

●Hepatosplenomegaly (40 to 60 percent)

●Small size for gestational age (SGA; 40 to 50 percent)

●Microcephaly (35 to 50 percent)

●Sensorineural hearing loss (SNHL; present at birth in approximately 35 percent, delayed SNHL can also occur)

●Lethargy and/or hypotonia (approximately 30 percent)

●Poor suck (approximately 20 percent)

●Chorioretinitis (10 to 15 percent)

●Seizures (5 to 10 percent)

●Hemolytic anemia (5 to 10 percent)

●Pneumonia (5 to 10 percent)

SNHL is a common sequela of congenital CMV and is detected in one-third to one-half of infants with symptomatic disease [1,3,16,18,19]. The hearing loss associated with symptomatic congenital CMV may be detectable at birth, but, in 18 to 30 percent of cases, it has delayed onset. In a systematic review, hearing loss was bilateral in 71 percent of children with symptomatic congenital CMV (compared with 43 percent of children with asymptomatic congenital CMV) [3]. The hearing loss associated with symptomatic congenital CMV is often progressive (18 to 63 percent of cases) [15,16,18,20,21] and eventually becomes severe to profound in the affected ear(s) of 78 percent of affected children [3].

Chorioretinitis is the most common ocular abnormality in symptomatic infants and correlates with poor long-term neurodevelopmental outcome. Other findings on eye examination may include retinal scars, optic atrophy, central vision loss, or strabismus [22,23]. Cataracts and microphthalmos are not typical in infants with congenital CMV and strongly suggest a disease other than CMV [24].

Ascites, myocarditis, cardiomyopathy, ventricular trabeculations, and enterocolitis are less common findings in symptomatic neonates.

Endocrinopathies, such as Graves' disease and diabetes insipidus, and renal disease, such as nephrotic syndrome, have been reported in newborns with symptomatic congenital CMV, but it is unclear whether these conditions are caused by CMV [25-27].

Life-threatening disease — Approximately 8 to 10 percent of newborns with symptomatic congenital CMV infection have severe, life-threatening disease. Life-threatening manifestations may include a sepsis-like illness, myocarditis, viral-induced hemophagocytic lymphohistiocytosis, and/or other severe end-organ involvement. Premature infants and infants with primary immune disorders of T cells or natural killer cells are at greater risk for mortality from congenital CMV infection. (See "Severe combined immunodeficiency (SCID): An overview", section on 'Clinical features' and "CD3/T cell receptor complex disorders causing immunodeficiency" and "NK cell deficiency syndromes: Clinical manifestations and diagnosis", section on 'Clinical features of NK disorders'.)

Many infants with fulminant disease at presentation die within days or weeks despite antiviral treatment and neonatal intensive care supportive measures. Mortality among such infants can be as high as 30 percent, whereas the overall mortality rate among infants with congenital CMV infection is approximately 4 to 8 percent within the first year of life [7,24,28].

Most infants die from viral-associated hemophagocytic syndrome or severe end-organ disease of the liver, lungs, bone marrow, or central nervous system [7,8,29]. In survivors, jaundice and hepatosplenomegaly may subside, but neurologic sequelae (eg, microcephaly, intellectual disability, cerebral palsy, and hearing disorders) persist. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Outcome'.)

Premature infants — Twenty-five to 35 percent of infants with symptomatic congenital CMV are born at <37 weeks gestation [17,18]. Premature neonates <32 weeks gestation with symptomatic congenital CMV are more likely to have pneumonitis, signs of viral sepsis, thrombocytopenia, and coinfections and less likely to have microcephaly or intracranial calcifications than term neonates [1,8,14,30].

Laboratory findings — Typical laboratory abnormalities associated with symptomatic congenital CMV include [1,7,8,15-18]:

●Elevated liver transaminases (50 to 83 percent)

●Thrombocytopenia (48 to 77 percent)

●Elevated direct and indirect serum bilirubin (36 to 69 percent)

Other less common findings include hemolytic anemia, neutropenia, lymphopenia, lymphocytosis, thrombocytosis, or leukemoid reaction. In newborns who undergo lumbar puncture, cerebrospinal fluid protein may be elevated (46 percent in one case series) [17].

Neuroimaging — Imaging of the brain with ultrasonography, unenhanced computed tomography (CT) (image 5A-B), or magnetic resonance imaging (MRI) (image 6) demonstrates abnormalities in 70 percent of infants with symptomatic congenital CMV infection [31-35]. Findings on neuroimaging include:

●Intracranial calcifications, usually periventricular (34 to 70 percent)

●Lenticulostriate vasculopathy (27 to 68 percent)

●White matter disease (22 to 57 percent)

●Ventriculomegaly (10 to 53 percent)

●Migrational abnormalities, including focal polymicrogyria, pachygyria, and lissencephaly (10 to 38 percent)

●Periventricular leukomalacia and cystic abnormalities (11 percent)

Additional findings that have been reported include ventricular septations and adhesions, cerebral atrophy, corpus callosum dysgenesis, and cerebellar hypoplasia [31,34,35].

Abnormalities on cranial imaging, particularly microcephaly, intracranial calcifications, and migrational abnormalities, correlate with poor long-term neurodevelopmental outcome [31,32,36]. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Outcome'.)

Primary neurophenotype — Some infants present with primarily neurologic findings (what we refer to as the "primary neurophenotype"). These infants lack the typical somatic manifestations, such as jaundice, petechiae, or hepatosplenomegaly. They may appear completely healthy at birth or may have mild microcephaly. They usually are not diagnosed with congenital CMV infection at birth unless they are cared for at a center where all newborns are screened for CMV. As they grow, they develop more significant microcephaly and neurologic manifestations (eg, global developmental delay, abnormal tone, seizures). A presentation similar to a genetic leukodystrophy has also been described [37]. If neuroimaging is performed, it typically shows polymicrogyria or other cortical dysplasia (see 'Neuroimaging' above). A high index of suspicion is necessary to identify congenital CMV infection as the etiology.

Asymptomatic infection — Approximately 90 percent of newborns who are congenitally infected with CMV are apparently asymptomatic at birth. Subtle differences, such as lower birth weight and slightly earlier gestational age, have been observed in newborns with asymptomatic congenital CMV infection [11].

Ten to 15 percent of apparently asymptomatic newborns experience SNHL [3,12,38]. (See 'Isolated hearing loss' below.)

Ocular abnormalities, including retinal lesions and strabismus, occur in 1 to 2 percent of infants with apparently asymptomatic congenital CMV but are rarely sight-threatening [22,23,39].

Abnormal brain imaging findings of periventricular leukomalacia, ventriculomegaly, and punctate calcifications have been observed in 5 to 20 percent of otherwise asymptomatic congenitally infected newborns (image 7).

Isolated hearing loss — Approximately 10 to 15 percent of apparently asymptomatic newborns with congenital CMV infection experience SNHL in infancy and early childhood; by age 18 years, up to 25 percent experience SNHL [3,12,19,38,40]. Some affected newborns have congenital hearing loss and may fail newborn hearing screening in one or both ears. In a systematic review of 37 observational studies, children with asymptomatic congenital CMV infection were less likely to have delayed-onset SNHL compared with symptomatic patients (9 versus 18 percent) and less frequently had bilateral severe to profound hearing loss (43 versus 65 percent) [3]. The two groups had similar rates of progressive hearing loss (approximately 20 percent) and fluctuating hearing loss (approximately 20 to 25 percent). In a study of 92 infants with asymptomatic congenital CMV infection, 20 children (22 percent) developed SNHL over a median follow-up of 17 years, including nine with congenital/early-onset SNHL and 11 with delayed-onset (at a median age of five years) [41]. Frequency-specific hearing thresholds were worse in ears with congenital/early-onset compared with delayed-onset SNHL. The severity of SNHL worsened over time in both groups. Approximately 2 percent of children with otherwise asymptomatic congenital CMV infection eventually develop SNHL severe enough to qualify for cochlear implantation [40]. With evolving considerations for unilateral SNHL and unilateral early cochlear implantation procedures, this percentage may be even higher. (See "Hearing loss in children: Treatment", section on 'Cochlear implants'.)

Universal newborn hearing screening programs may identify some of these otherwise asymptomatic infants. In a study of 572 newborns who failed newborn hearing screening, 6 percent were found to have congenital CMV infection; 75 percent of infants with congenital CMV infection were identified solely on the basis of the abnormal newborn hearing screen [42]. Studies in older infants and children indicate that 10 to 20 percent of hearing-impaired children have hearing loss as a result of congenital CMV [3]. (See 'Newborn screening for congenital cytomegalovirus' below and "Screening the newborn for hearing loss".)

Late complications — Approximately 70 to 80 percent of infants who are symptomatic at birth develop late complications that may include (see "Congenital cytomegalovirus infection: Management and outcome", section on 'Outcome') [1,24,43]:

●Hearing loss requiring hearing aids or cochlear implantation

●Vision impairment requiring glasses or corrective surgery

●Dental abnormalities

●Intellectual disability and delayed psychomotor development

●Behavioral problems such as inattention and hyperactivity

●Neuromuscular problems such as cerebral palsy and associated neuromuscular scoliosis and/or hip dislocation and dysplasia

Hearing loss is by far the most common late sequela seen in asymptomatic congenitally infected infants, occurring in up to 25 percent by age 18 years [3,12,38,40,41]. As discussed above, hearing loss often progresses over time [41,44] (see 'Isolated hearing loss' above). Thus, both ears should be closely monitored for progression of known hearing loss or emergence of new hearing loss in the normal hearing ear [44]. Vestibular and balance problems can also occur, with or without associated SNHL [45].

Some studies have identified neurocognitive and language delays in children with asymptomatic congenital CMV, but others found no lasting cognitive sequelae [46-49]. If children with asymptomatic congenital CMV differ cognitively from children without congenital CMV infection, the differences are likely to be subtle, involve one area or domain of learning, or require in-depth testing to detect. Children born with asymptomatic congenital CMV infection do not appear to have increased risk of behavioral differences (eg, attention and hyperactivity disorders) as those born with symptomatic infection do [50].

Lon-term outcomes and follow-up for children with congenital CMV infection are discussed in greater detail separately. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Outcome' and "Congenital cytomegalovirus infection: Management and outcome", section on 'Long-term follow-up'.)

DIAGNOSTIC APPROACH — A high index of suspicion is important for timely diagnosis. The manifestations of congenital CMV infection are variable and nonspecific, and infants may present with most, some, or just one clinical sign or symptom.

Clinical suspicion — Congenital CMV should be suspected in the following clinical scenarios [51]:

●Newborns with signs and symptoms consistent with congenital CMV disease – Findings that may warrant testing for CMV, especially if not explained by other causes, include microcephaly, small size for gestational age (SGA), thrombocytopenia, hepatosplenomegaly, and jaundice or direct hyperbilirubinemia at birth. Jaundice that fits other patterns does not necessarily warrant testing for CMV (eg, ABO incompatibility). (See 'Symptomatic neonate' above and "Unconjugated hyperbilirubinemia in neonates: Etiology and pathogenesis", section on 'Causes of significant unconjugated neonatal hyperbilirubinemia'.)

The yield of CMV testing is greatest when there are multiple suggestive clinical findings. The diagnostic yield of CMV testing in the setting of an isolated finding (eg, SGA) is relatively low [51-54]. The author of this topic review still favors testing for CMV in the setting of isolated SGA; however, other experts do not recommend CMV testing in this setting.

●Newborns with abnormal neuroimaging consistent with CMV, if the findings are not explained by other causes. Suggestive neuroimaging findings include periventricular calcifications, lenticulostriate vasculopathy, white matter disease, ventriculomegaly, migrational abnormalities (eg, polymicrogyria), or periventricular leukomalacia. (See 'Neuroimaging' above.)

●Newborns who have documented sensorineural hearing loss (SNHL), whether or not they have other symptoms of congenital CMV [51].

For infants who fail the newborn hearing screen but have not yet had a formal audiologic evaluation, the decision of whether or not to test for CMV is controversial and practice varies. In our practice, newborns who fail the initial hearing screen in one or both ears have a repeat screen performed before discharge from the nursery. If the newborn fails the repeat screen, he or she is referred for audiologic assessment and tested for congenital CMV before discharge or at least within the first three weeks of life. Other centers do not routinely screen for CMV in this setting. This issue is discussed in greater detail separately. (See "Screening the newborn for hearing loss", section on 'Screening for cytomegalovirus (CMV)'.)

●Newborns born to mothers with known or suspected CMV infection during pregnancy, including mothers with:

•Maternal seroconversion during pregnancy

•Presumptive maternal primary CMV infection with positive CMV immunoglobulin G (IgG) and CMV immunoglobulin M (IgM) antibody

•Mononucleosis-like illness during pregnancy

•Abnormal fetal imaging suggestive of in utero CMV infection (see 'In utero' above)

●Immune-compromised newborns, particularly infants with an abnormal T cell receptor excision circles (TRECs) result on newborn screening, which is indicative of severe combined immunodeficiency disorder (SCID), because congenital CMV infection can have devastating consequences in these severely immune-compromised hosts. (See "Newborn screening for inborn errors of immunity" and "Severe combined immunodeficiency (SCID): An overview".)

Approach to testing — Laboratory diagnosis of congenital CMV infection is accomplished by isolation or molecular detection of CMV from urine or saliva samples collected within the first three weeks of life.

We prefer urine samples over saliva samples for CMV testing. Although saliva is easier to collect, false-negatives and false-positives are more common with saliva samples compared with urine [55]. False-negatives may occur at a higher rate because saliva samples are more susceptible to sampling errors (ie, inadequate amount). False-positives may occur rarely in infants with CMV-infected mothers due to contamination of the saliva sample with retained breast milk in the mouth of the newborn [56]. When this occurs, the quantitative CMV DNA level is usually lower than in neonates with true infection.

Viral culture, modified culture (also called rapid culture or shell vial assay), and polymerase chain reaction (PCR) are the preferred diagnostic tests for newborns with suspected congenital CMV infection (table 2). PCR is more sensitive compared with rapid culture and may be more accurate, especially if the sample must be transported to a reference lab over a long distance. In addition, PCR provides quantitative results [57]. Most newborns with congenital CMV have high levels of CMV DNA in their urine and saliva. If a low level is detected on CMV PCR, the results should be confirmed by culture or repeat PCR testing.

The choice of test may depend upon availability at different laboratories. Testing blood samples for CMV is not recommended as a first-line test because not all infected infants are viremic. However, detection of CMV by PCR in blood or plasma samples can be diagnostic. Serologic testing for CMV IgM antibody is not recommended for neonates and young infants, because CMV IgG antibody in the blood at this age reflects maternal IgG level and is not diagnostic of a congenital CMV infection.

Prenatal — Diagnosis of in utero CMV infection can be made by viral culture or CMV DNA detection in amniotic fluid or also by CMV IgM antibody measurement in fetal blood of the symptomatic fetus. (See "Cytomegalovirus infection in pregnancy", section on 'Fetal issues'.)

Birth to three weeks — The diagnosis of congenital CMV infection can be established within the first three weeks of life by detection of CMV in the urine or saliva (table 2). Testing should be carried out as soon as the diagnosis is suspected so that evaluations for end-organ disease can be performed and, if indicated, antiviral therapy can be started promptly. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Antiviral treatment'.)

Three weeks to one year — After three weeks, the detection of CMV in urine or saliva may indicate either congenital or postnatal CMV infection. Postnatal CMV infection usually is clinically benign or self-limited. (See "Overview of cytomegalovirus infections in children", section on 'Early postnatal infection'.)

Congenital CMV infection may be retrospectively diagnosed on the basis of PCR analysis of dried blood samples (Guthrie cards) obtained for newborn screening [58-62]. Detection of CMV DNA in the newborn dried blood spot sample indicates that CMV viremia was present and confirms congenital CMV infection.

In the United States, with written parental permission, newborn dried blood spot cards may be retrieved from state newborn screening laboratories and tested for CMV DNA by selected research or public health laboratories experienced in this specialized testing. Positive CMV PCR on a dried blood spot confirms the diagnosis of congenital CMV infection. However, a negative result does not exclude CMV infection; false-negative results may occur, particularly in newborns with few or mild symptoms. (See "Overview of newborn screening".)

If testing of the dried blood sample is negative or cannot be performed, alternative methods of testing in this setting include:

●Testing the urine or saliva for CMV (via viral culture, shell vial assay, or PCR)

●Measurement of CMV IgG antibody in the blood

Detection of CMV (by any of these methods) in a symptomatic infant at age three weeks to one year suggests, but does not confirm, congenital CMV infection because of the possibility of postnatal infection.

Older than one year — Establishing a diagnosis of congenital CMV infection beyond the first year of life is generally not feasible. Retrospective diagnosis of congenital CMV via PCR analysis of dried blood spots as described above may be possible in some regions; however, this is not possible in the United States because most state health laboratories that perform newborn screening discard the dried blood spot samples after one year. In addition, many young children, particularly those attending daycare, become infected with CMV during the first few years of life. Hence, detection of CMV in the urine or saliva or CMV antibodies in the blood of a child older than one year is far more likely to represent postnatal infection than congenital infection. However, a negative CMV IgG antibody and a negative CMV culture of urine or saliva provides evidence against a congenital CMV infection.

A small number of studies have reported late diagnosis of congenital CMV infection using PCR methods on preserved umbilical cords [63-65]. However, this method is not recommended, because it has not been adequately studied and the reliability of results of umbilical cord testing is unclear. In addition, contamination of the sample with maternal blood or other sources may cause false-positive results.

Interpretation/diagnosis — The presence or absence of congenital CMV infection can be determined based on the timing and results of diagnostic tests, as follows:

●Virologically proven congenital CMV infection can be diagnosed on the basis of any of the following:

•Detection of CMV via viral culture in urine or saliva samples obtained within the first three weeks of life

•Detection of CMV via shell vial assay in urine or saliva samples obtained within the first three weeks of life, with a positive confirmatory test (either viral culture or PCR)

•Detection of CMV via PCR in urine, saliva, or blood samples obtained within the first three weeks of life, confirmed on repeat testing

•Detection of CMV via PCR in the newborn screening dried blood spot

●Possible congenital CMV infection – It may not be possible to confirm the diagnosis of congenital CMV infection if testing is performed after the first three weeks of life (because of the possibility of postnatal acquisition). Newborn dried blood spot testing can be helpful if available, but negative results do not exclude congenital CMV infection. A diagnosis of "possible" congenital CMV infection may be made if all of the following criteria are met:

•One or more signs or symptoms of congenital CMV

•Other conditions that cause these abnormalities have been excluded (see 'Differential diagnosis' below)

•CMV is detected in urine or saliva samples (via viral culture, shell vial assay, or PCR) or CMV IgG antibody is detected in the blood after the first three weeks of life, up to one year of age

●Not infected – Infants in whom CMV is not detected in urine or saliva (via viral culture, shell vial assay, or PCR) during the newborn period do not have congenital CMV. Because of the high specificity of these tests (table 2), a negative result excludes the diagnosis of congenital CMV infection. In our practice, we usually perform at least two tests (urine or saliva or both) to exclude the diagnosis with confidence. Congenital CMV infection can also be excluded beyond the newborn period if CMV IgG antibody testing is negative (provided the infant has a normal immune system), indicating that CMV infection never occurred in the child.

POST-DIAGNOSIS EVALUATION — Infants with virologically confirmed congenital CMV infection should be evaluated for evidence of organ involvement. Comprehensive evaluation should be performed even in infants who appear asymptomatic to detect subtle or subclinical subtle findings. The evaluation includes:

●Thorough physical, neurologic, and neurodevelopmental examination, including measurements of weight, length, and head circumference (to detect microcephaly).

●Laboratory testing, including:

•Complete blood count with differential and platelet count

•Liver function tests

•Kidney function tests

•Coagulation studies in selected cases with liver disease or viral sepsis

●Hearing evaluation by auditory brainstem response.

●Ophthalmology evaluation.

●Neuroimaging – Ultrasonography should be performed in all infected infants. Ultrasonography is better able to detect lenticulostriate vasculopathy than advanced imaging and will detect most major, obvious abnormalities. Infants with abnormal ultrasonography, abnormal neurologic examination (eg, focal abnormalities, globally increased or decreased tone, abnormal reflexes), seizures, or abnormal head circumference should undergo additional advanced neuroimaging with computed tomography (CT) and/or magnetic resonance imaging (MRI) of the brain. CT is quickly performed and is useful in detecting ventriculomegaly or calcifications but entails radiation exposure, whereas MRI requires a longer procedure time, often requires sedation, but does not entail initial radiation exposure and is more sensitive in detecting vasculitis, polymicrogyria, white matter abnormalities, and other neuronal migrational abnormalities.

Cranial imaging with ultrasound, CT, or MRI helps to assess the degree of central nervous system involvement. Ultrasound is the preferred initial study for most infants, but MRI should be obtained if there are abnormal neurologic examination findings (eg, microcephaly, focal deficits, seizures or other abnormal movements, delayed milestones, increased or decreased tone), if there is evidence of clinically significant CMV disease, or if the ultrasound is abnormal. The choice between MRI versus CT should be individualized based on the relative advantages and disadvantages. The advantage of MRI is that it is far more sensitive than CT for identifying structural abnormalities, though it is less sensitive for identifying calcification. Newer MRI imaging techniques, such as susceptibility-weighted imaging, may visualize calcium better and can be performed in infants and newborns.

The main disadvantage of MRI is that some infants may require sedation/anesthesia and it is a longer procedure. In some institutions, a "neonatal MRI protocol" may be available, where the infant is swaddled and the MRI is performed under natural sleep. The main disadvantage of CT is that it exposes the infant to ionizing radiation; however, because it is quick and generally does not require sedation, it is sometimes the preferred choice for imaging unstable infants. If the infant also has documented significant sensorineural hearing loss (SNHL), coordination with the otolaryngologist may be important because imaging of the temporal bones and the internal auditory canals also may be needed, especially if a cochlear implant procedure is under consideration, and can be performed at the same time as brain imaging.

●We suggest measuring CMV DNAemia by quantitative polymerase chain reaction (PCR) of whole blood or plasma for any infant receiving antiviral therapy. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Who to treat'.)

DIFFERENTIAL DIAGNOSIS

●Other congenital and neonatal infections – The classic findings of symptomatic congenital CMV infection include petechiae, jaundice, hepatosplenomegaly, small size for gestational age (SGA), and microcephaly. These nonspecific findings may also occur in:

•Other TORCH infections (including Zika virus infection) (see "Overview of TORCH infections" and "Congenital Zika virus infection: Clinical features, evaluation, and management of the neonate")

•Other less common pathogens such as lymphocytic choriomeningitis virus (see "Viral meningitis in children: Epidemiology, pathogenesis, and etiology", section on 'Lymphocytic choriomeningitis virus')

•Neonatal sepsis (see "Clinical features, evaluation, and diagnosis of sepsis in term and late preterm neonates")

Although other congenital infections may have characteristic clinical findings (table 1), appropriate virologic and microbiologic studies generally are necessary to make a specific diagnosis. (See "Overview of TORCH infections", section on 'Initial evaluation' and "Clinical features, evaluation, and diagnosis of sepsis in term and late preterm neonates", section on 'Evaluation and initial management'.)

●Abnormal neurologic findings – A variety of genetic and metabolic disorders, as well as in utero exposure to drugs and toxins (eg, alcohol, cocaine, isotretinoin), may mimic the neurologic findings of congenital CMV infection. The presence and pattern of calcifications on computed tomography (CT) imaging of the brain can be helpful in distinguishing between CMV infection and other causes of neurologic disease. Intracranial calcifications, particularly in periventricular distribution, are a common finding in CMV-infected neonates. Additional testing such as chromosomal analysis, metabolic studies, and drug screening may be used to exclude other causes of neurologic disease in newborns, including [1]:

•Tuberous sclerosis complex (see "Tuberous sclerosis complex: Clinical features")

•Sturge-Weber syndrome (see "Sturge-Weber syndrome")

•Aicardi syndrome

•Galactosemia (see "Galactosemia: Clinical features and diagnosis")

•Urea cycle deficiencies (see "Urea cycle disorders: Clinical features and diagnosis")

•Organic acidemias (see "Organic acidemias: An overview and specific defects")

•Lysosomal storage disorders (see "Inborn errors of metabolism: Classification", section on 'Lysosomal storage disorders')

•Peroxisomal disorders (eg, Zellweger syndrome, infantile Refsum disease, adrenoleukodystrophy) (see "Peroxisomal disorders" and "Clinical features, evaluation, and diagnosis of X-linked adrenoleukodystrophy")

•Other inherited leukodystrophies (see "Metachromatic leukodystrophy" and "Alexander disease")

•In utero exposure to drugs and toxins (eg, alcohol, cocaine, isotretinoin) (see "Fetal alcohol spectrum disorder: Clinical features and diagnosis" and "Neonatal abstinence syndrome (NAS): Clinical features and diagnosis", section on 'Cocaine')

●Hepatitis and hyperbilirubinemia – The differential diagnosis for the hepatic findings associated with congenital CMV infection includes:

•Other causes of viral hepatitis (hepatitis A, hepatitis B, hepatitis C, Epstein-Barr virus, herpes simplex virus, enterovirus, adenovirus) (see "Hepatitis viruses and the newborn: Clinical manifestations and treatment")

•Ischemic injury

•Thrombosis

•Hemolytic disease (see "Unconjugated hyperbilirubinemia in neonates: Etiology and pathogenesis", section on 'Causes of significant unconjugated neonatal hyperbilirubinemia')

•Biliary atresia (if conjugated hyperbilirubinemia is present and persistent) (see "Biliary atresia")

•Metabolic and genetic disorders that involve the liver (see "Inborn errors of metabolism: Identifying the specific disorder", section on 'Hepatosplenomegaly')

The clinical course and appropriate virologic tests distinguish CMV from these other causes. Infants with congenital CMV and persistent liver dysfunction or persistent conjugated hyperbilirubinemia should be evaluated for biliary atresia or alpha-1 antitrypsin deficiency since CMV may coexist with other disorders of liver function, presenting a diagnostic challenge. (See "Unconjugated hyperbilirubinemia in neonates: Etiology and pathogenesis" and "Unconjugated hyperbilirubinemia in term and late preterm newborns: Screening".)

NEWBORN SCREENING FOR CONGENITAL CYTOMEGALOVIRUS — Newborns are not routinely screened for CMV. However, given the substantial public health impact of congenital CMV, most CMV experts support targeted and/or universal newborn screening for congenital CMV infection. The goals of newborn screening include early identification of infected infants with subtle symptoms who may benefit from antiviral therapy, and identification of asymptomatic infants who are at risk for delayed hearing loss and warrant more frequent audiologic evaluation [66-71]. (See "Congenital cytomegalovirus infection: Management and outcome", section on 'Antiviral treatment' and "Congenital cytomegalovirus infection: Management and outcome", section on 'Long-term follow-up'.)

Targeted newborn screening — Targeted newborn screening involves testing for CMV in all newborns who fail their newborn hearing screen in either one or both ears. Many hospitals have adopted targeted CMV screening programs that aim to identify infected infants in the newborn period, so that appropriate evaluations for end-organ involvement can be performed and antiviral treatment, if indicated, may be provided. As discussed above, our practice is to screen for CMV in newborns who fail both the initial and repeat follow-up hearing screen in one or both ears (see 'Clinical suspicion' above). In such neonates, we perform testing for congenital CMV before discharge or at least within the first three weeks of life. However, other centers do not routinely screen for CMV in this setting.

Targeted screening for CMV in the setting of a failed newborn hearing screen is discussed in greater detail separately. (See "Screening the newborn for hearing loss", section on 'Screening for cytomegalovirus (CMV)'.)

Universal newborn screening — Universal newborn screening for congenital CMV has been proposed by many CMV experts, audiologists, otolaryngologists, and public health officials [68,72,73]. The main goal of this approach is to identify asymptomatic infected infants in order to provide careful monitoring for delayed-onset hearing loss. As discussed above, asymptomatic newborns with congenital CMV infection have a risk of delayed-onset hearing loss that is not detected by newborn hearing screen protocols (see 'Isolated hearing loss' above). If delayed-onset hearing loss is identified, educational accommodations, speech/language therapies, and other appropriate interventions can be initiated at an early stage in order to optimize the child's language development and learning. (See "Hearing loss in children: Treatment".)

Antiviral treatment is generally not recommended for infected infants who pass the newborn hearing screen and are otherwise asymptomatic, although an ongoing prospective study is evaluating whether antiviral therapy reduces the risk of late-onset hearing loss in asymptomatic infected newborns [74] (see "Congenital cytomegalovirus infection: Management and outcome", section on 'Who to treat'). A National Institutes of Health-sponsored multicenter clinical trial evaluating the potential benefit of treatment of asymptomatically infected newborns to reduce the risk of later-onset hearing loss is now being conducted. However, another potential benefit of universal newborn screening for CMV is that it may eliminate the diagnostic odyssey experienced by many newborns with atypical signs and symptoms of congenital CMV or those with failed newborn hearing screens.

Universal newborn screening appears to be cost effective [75,76]. However, the most reliable and cost-effective method for universal newborn screening for congenital CMV is not established. It may involve detection of CMV in saliva or urine collected at birth and tested by central laboratory high-throughput technology or by simple point-of-care detection methods that are under development; or testing of dried blood spots (DBS) collected for routine newborn screening panels [62,72,77]. Improvements in PCR testing methodologies has led to improved sensitivity of DBS testing [62]. In a report from two centers in Europe, performing testing on pooled saliva samples reduced the cost of screening [78]. (See "Overview of newborn screening".)

In the absence of universal screening, the diagnosis of congenital CMV is often delayed because many infected newborns do not exhibit classic signs and symptoms. Among newborns who fail newborn hearing screening, testing for CMV often is not performed until completion of confirmatory auditory testing and evaluation by otolaryngology specialists, which often is around two to three months of age.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: TORCH infections".)

SUMMARY AND RECOMMENDATIONS

●Epidemiology – Congenital cytomegalovirus (CMV) infection affects approximately 30,000 to 40,000 infants annually in the United States and is associated with long-term neurodevelopmental disabilities, including sensorineural hearing loss (SNHL), cerebral palsy, intellectual disability, vision impairment, and seizures. It is the leading cause of nonhereditary SNHL. (See 'Introduction' above and 'Epidemiology' above.)

●Clinical manifestations

•Clinical findings in symptomatic newborns – Approximately 90 percent of newborns who are congenitally infected with CMV are apparently asymptomatic at birth. Clinical findings in symptomatic neonates are nonspecific and include petechiae, jaundice, hepatosplenomegaly, small size for gestational age (SGA), and microcephaly. (See 'Symptomatic neonate' above.)

•Hearing loss – SNHL is the most common sequela of congenital CMV infection, occurring in 33 to 50 percent of symptomatic infants and approximately 25 percent of asymptomatic infants. The newborn hearing screen is a valuable tool for detecting otherwise asymptomatic infants with CMV infection. (See 'Clinical findings' above and 'Isolated hearing loss' above and "Screening the newborn for hearing loss".)

•Laboratory and imaging findings – Common laboratory findings in the symptomatic neonate include thrombocytopenia, elevated transaminases, and elevated direct and indirect serum bilirubin. Imaging of the brain often shows periventricular intracranial calcifications, migrational abnormalities, or other abnormalities (image 5A-B and image 6). (See 'Laboratory findings' above and 'Neuroimaging' above.)

●Clinical suspicion – Congenital CMV infection should be suspected in infants with (see 'Clinical suspicion' above):

•One or more of the typical signs of symptomatic disease (eg, microcephaly, SGA, thrombocytopenia, petechiae or purpura, hepatosplenomegaly, or jaundice at birth)

•Abnormal neuroimaging with otherwise unexplained periventricular calcifications, lenticulostriate vasculopathy, white matter disease, ventriculomegaly, migrational abnormalities (eg, polymicrogyria), or periventricular leukomalacia

•Documented SNHL in one or both ears

•Mother who had known or suspected maternal CMV infection during pregnancy or abnormal fetal imaging suggestive of in utero CMV infection

●Diagnosis – The diagnosis of congenital CMV infection is confirmed by isolation or molecular detection of CMV from urine or saliva samples collected within the first three weeks of life (table 2). Both viral culture and polymerase chain reaction (PCR) tests have high sensitivity and specificity for detection of CMV in infected neonates. Serology should not be used for the routine diagnosis of congenital CMV infection. (See 'Diagnostic approach' above.)

●Differential diagnosis – The differential diagnosis of congenital CMV infection includes other congenital infections, neonatal sepsis, genetic and metabolic disorders, and in utero exposure to drugs and toxins. Appropriate virologic and microbiologic studies, chromosomal analysis, metabolic studies, drug screening, and presence and pattern of calcifications on computed tomography (CT) imaging of the brain will aid in differentiating among these possibilities. (See 'Differential diagnosis' above.)