Congenital Cytomegalovirus Infection

Cytomegalovirus (CMV) is a member of the herpesvirus family and is highly prevalent, infecting most people worldwide in childhood or early adulthood.1 The prevalence among women of reproductive age is estimated to be from 58% to 79% in North America and 86% globally.2,3

As a result, congenital CMV (cCMV) is the most common intrauterine infection seen,4 with an incidence of approximately 0.5% to 1.3% in the United States5,6 and higher in developing countries, resulting in in about 20,000 to 30,000 cCMV-infected babies in the United States each year.

Intrauterine transmission of CMV occurs during primary maternal infection or non-primary infection in HIV-positive pregnant women resulting from reactivation of latent virus or infection with a different strain; The type of maternal infection has a different impact on affected infants (Table 1). Primary maternal infection confers a 40% risk of transmission to the fetus, while non-primary infection carries a 0.5% to 2% risk. Given the high rate of maternal seropositivity, the majority of cCMV infections result from non-primary maternal infections.7

Studies suggest that severe manifestations in infants, including neurological deficits, are more likely to result from congenital infection following a primary maternal infection, but manifestations may also occur after a non-primary infection.8,9 Previous studies have shown that approximately 10% to 15% of babies with cCMV due to primary maternal infection are symptomatic at birth and 25% present sequelae at 2 years of age. Following non-primary maternal infection, less than 1% of infants with cCMV are expected to be symptomatic at birth and 8% will have sequelae by 2 years of age. However, a recent meta-analysis found no difference in the rate of symptomatic manifestations in primary and non-primary maternal infections (pooled odds ratio for symptomatic cCMV: 0.83, 95% confidence interval: 0.55–1.27 ), suggesting that more data collection is needed.10

Maternal infection in the first half of pregnancy is associated with a lower risk of transmission to the fetus, but the risk of serious sequelae increases if congenital infection occurs during this period. Despite available epidemiological data and given the limited sensitivity and positive predictive value of prenatal imaging, it is difficult to predict which infants will have cCMV sequelae. This highlights the need for further research in this area.

To prevent cCMV, preventive measures can be followed for pregnant women, including standard precautions (hand hygiene) and limiting contact with other people’s saliva. There is no universally recommended CMV screening test in pregnancy and, to date, there is no proven intervention to decrease transmission to the fetus. Prevention efforts currently under study include CMV-specific immunoglobulin11,12,13,14 (which has not shown reproducible beneficial prevention results), antivirals15,16 (in the case of recognized CMV infection during pregnancy), and most importantly, the development of vaccines against CMV.17

The clinical manifestations of cCMV range from the absence of any short- or long-term sequelae to multisystem involvement.

Symptoms of cCMV in the newborn include any of the following: central nervous system abnormalities (microcephaly, cortical malformations, ventriculomegaly, periventricular calcifications and/or germinal cysts), sensorineural hearing loss (SNHL), chorioretinitis, hepatosplenomegaly, transaminitis, direct hyperbilirubinemia , petechiae/thrombocytopenia and intrauterine growth restriction (IUGR). Patients with SNHL identified shortly after birth without any other clinical manifestations are classified as a subgroup of asymptomatic infection. An additional 5% to 15% of infants with cCMV who are asymptomatic at birth and without SNHL will develop late-onset sequelae, the most common being SNHL.

Given that SNHL is the most common sequela of cCMV, it is not surprising that cCMV is the most common nongenetic cause of hearing loss in children, accounting for about a quarter of cases. SNHL occurs in 20% to 65% of infants with symptomatic cCMV and 6% to 25% of infants with asymptomatic cCMV (this includes infants with isolated SNHL soon after birth and those patients who develop late-onset SNHL). ).18,19 Hearing loss varies from mild to profound, can be unilateral or bilateral, and can be stable, progressive or fluctuating. For infants affected with SNHL, the loss rarely improves over time and most children with SNHL (both symptomatic and asymptomatic) eventually have progression of the loss.20,21,22

The onset of hearing loss may be delayed and occur during the first years of life.19,20 Furthermore, affected patients who present with unilateral hearing loss are at high risk of developing SNHL in the contralateral ear. Among symptomatic cases of cCMV, IUGR, petechiae, microcephaly, and abnormal neuroimaging findings are associated with SNHL.23,24,25,26 Among asymptomatic infants with cCMV, prematurity and low birth weight are associated with SNHL. 18,27,28 There is no reliable method to predict which children with cCMV will develop SNHL.

Long-term morbidity in patients with cCMV results from neurological disability and includes cerebral palsy, motor/cognitive deficits, and seizure disorders. Additionally, visual impairment may occur, both due to ocular manifestations and cortical blindness. Mortality among infants with symptomatic cCMV in the United States is estimated to be less than 5%.29,30

Testing for cCMV should be performed within the first 3 weeks after birth to distinguish congenital infection from that acquired after birth (acquisition through saliva or breast milk). Postnatal CMV infection does not have the same constellation of symptoms or risk of hearing loss as cCMV,31 although postnatal CMV can cause clinical disease, particularly in premature infants.32 Thus, diagnostic and treatment dilemmas can be avoided with testing. early neonatal infections for CMV. Screening programs can be targeted (i.e., only screen newborns with signs or symptoms suspected of cCMV infection) or universal (i.e., screen all newborns).

> Targeted detection

At this time, targeted screening for cCMV is increasingly common in the United States. In this approach, neonates with abnormalities suspicious for cCMV are tested for CMV. Although some newborns diagnosed through screening will not meet criteria for antiviral treatment, routine targeted testing allows newborns who are diagnosed with cCMV to undergo a complete evaluation (brain imaging, laboratory tests, and ophthalmologic examination). ).

After a complete evaluation, the doctor can evaluate whether treatment is indicated and, if so, initiate therapy within the appropriate time frame. Additionally, infants with cCMV who do not qualify for antiviral treatment after birth may be appropriately monitored (e.g., repeat audiologic testing) and may have the opportunity to enroll in clinical trials. The ethical implications of this specific approach have been previously reviewed.33

Hearing-targeted screening for cCMV focuses on infants who have an abnormal hearing test during newborn hospitalization.34

Many states have legislation requiring hospitals to provide parents of babies who fail their hearing screening with information about cCMV and the opportunity to get it tested. Some health systems have policies for reflexive cCMV testing in newborns who fail hearing screening before hospital discharge. Some states have legislation requiring cCMV testing in newborns with suspected hearing impairment as a result of routine newborn hearing screening. In general, parents surveyed appear to support routine neonatal screening for CMV.35 Targeted screening for cCMV in infants with hearing impairment is particularly useful because the absence of other abnormalities found in patients with symptomatic cCMV does not rule out cCMV as a possible cause of the condition. hearing loss.

Some experts advocate that clinicians investigate additional causes of hearing loss among patients with a failed hearing test and a diagnosis of cCMV to identify overlapping causes of hearing loss. Genetic causes of hearing loss have been reported in a small percentage of this population.36,37 The discovery of an additional possible cause of hearing loss may affect the analysis of the possible risks and benefits of antiviral treatment for cCMV, particularly if the infant has no other manifestations of CMVc.

Expanded targeted screening is another approach by which testing for cCMV is performed in infants who have findings consistent with cCMV (other than SNHL) including thrombocytopenia/petechiae, conjugated hyperbilirubinemia, hepatosplenomegaly, hepatitis, IUGR, small for gestational age, microcephaly, rash consistent with cCMV, abnormal brain ultrasound with unexplained ventriculomegaly or periventricular calcifications.

> Universal detection

Congenital CMV is a leading cause of childhood disability and has a notably higher incidence than disorders currently included in neonatal screening programs.38

However, CMV is not currently included in any state newborn screening program.

Universal newborn hearing screening improved early detection of SNHL and increased early identification of infants with cCMV; However, a significant proportion of infants with cCMV are lost because hearing loss can occur beyond the neonatal period.39 This provides a rationale for universal CMV screening in newborns,38,40,41,42 which would lead to rapid diagnosis and complete evaluation of all infants with cCMV. This universal approach would identify children who need prospective audiological and developmental follow-up to detect late-onset manifestations.

In addition to health outcomes and quality of life benefits, a cost analysis study concluded that universal screening for cCMV would be cost-effective and result in net health care savings.43 Although the focus should remain on symptomatic infants who benefit from treatment, universal screening would also allow families with affected infants to make informed decisions about available treatment options. Options for a universal screening approach include screening all neonates during birth hospitalization or public health department-based neonatal screening programs; Specific testing modalities are discussed below.

The historical gold standard for CMV detection through culture has been replaced with DNA polymerase chain reaction (PCR) testing on urine, saliva, or blood samples.

Compared to viral culture, PCR provides faster results, is more sensitive, and requires only 1 sample. PCR for CMV in urine is the most sensitive and specific test. Saliva (buccal swab) may be a more convenient sample to collect in newborns and has high sensitivity but slightly lower specificity.44,45,46 False positive results are possible with saliva samples due to viral shedding in seropositive mothers with previous CMV infection; therefore, saliva samples should be collected 1 to 2 hours after breastfeeding to minimize this likelihood.47

Saliva PCR is commonly used as a screening test in daycares because samples are easy to obtain. Rapid detection platforms using pooled saliva samples are currently being validated and may help facilitate expeditious diagnosis.48,49 A positive saliva PCR result should be confirmed with repeat testing (preferably urine PCR).

In infants with cCMV, urine PCR will remain positive for months, but the window for definitive diagnosis of congenital infection is a positive urine PCR result within the first 3 weeks after birth; a positive urine PCR result beyond this period could also be compatible with a postnatally acquired CMV infection. There is no clear role for CMV IgG/IgM (antibody) testing in the infant because the presence of maternal IgG antibodies will confound the results, and CMV IgM testing has limited predictive value.

If screening for cCMV occurs beyond the first 3 weeks after birth, CMV testing using a dried blood spot obtained during the first 3 weeks of age and stored for neonatal screening programs may be useful. Previous studies have reported lower sensitivity of this test, but more recent studies show improved sensitivity, and it may be appropriate for retrospective diagnosis in infants and children with suspected cCMV infection.50,51 Dried blood spot cards are currently kept by health departments for varying periods of time before they are discarded and must be requested and retrieved with family permission.

The mainstay of antiviral treatment for cCMV disease includes intravenous ganciclovir and its oral prodrug valganciclovir.

The benefit of treatment has been demonstrated in clinical trials of symptomatic infants with cCMV (defined by the presence of at least 1 symptom of cCMV-related end-organ disease) with antiviral initiation in the first month after birth. In 1997, a phase II clinical trial reported an improvement in hearing outcomes after 6 weeks of ganciclovir in symptomatic infants with cCMV.52

Subsequently, a phase III study in symptomatic infants with CMV with neurological involvement reported improved hearing outcomes (assessed at 6 months to 1 year)53 and neurodevelopmental sequelae.54 Additionally, the pharmacokinetics of oral valganciclovir were found to be equivalent to intravenous ganciclovir. .55,56 In 2015, a phase III randomized controlled study compared symptomatic infants with cCMV who received 6 weeks versus 6 months of valganciclovir.57

Longer treatment resulted in better hearing outcomes and neurodevelopmental scores at 24 months of age.57 The most important adverse effect of ganciclovir and valganciclovir is neutropenia, which is dose-dependent and reversible. Additional side effects may include thrombocytopenia, anemia, renal failure, and transaminitis. Theoretically, this treatment could carry a risk of teratogenesis, carcinogenesis, and male infertility, which have been observed in animal studies.58

Based on clinical trial data, antiviral treatment is recommended in infants with symptomatic cCMV (at least 1 symptom of cCMV-related end-organ disease). It should be noted that these trials include a low number of infants with mildly symptomatic disease. In 2017, consensus recommendations were published, recommending treatment of infants with cCMV disease with moderate to severe symptoms, excluding patients with isolated SNHL or mild symptomatic disease.59 Moderate to severe cCMV is defined as infants presenting with multiple manifestations. of disease including thrombocytopenia/petechiae, IUGR, hepatitis, hepatosplenomegaly or central nervous system involvement (microcephaly, classic radiographic abnormalities, chorioretinitis).

SNHL is considered evidence of central nervous system involvement if there are other abnormalities suggestive of cCMV disease. An infant who has SNHL without other apparent manifestations of cCMV is classified as having “asymptomatic infection with isolated SNHL.”

Mildly symptomatic cCMV infection includes infants with manifestations such as an isolated low platelet count that resolves rapidly or mild transaminitis. Ideally, diagnosis of cCMV and eligibility for treatment should occur in the first month after birth. If indicated, treatment should begin at one month of age with a duration of antiviral therapy of 6 months. A summary of this treatment approach is provided in Table 2.

Some clinical scenarios require expert opinion and discussion with families to determine whether antiviral treatment should be initiated. Factors influencing treatment include the spectrum/severity of SNHL; findings associated with, but not pathognomonic for cCMV (such as periventricular cystic lesions); and confidence in the diagnosis of cCMV when the testing timeframe is beyond 3 weeks of age.

Consultation with an infectious disease specialist can be important and useful in interpreting the data, determining a treatment decision, and advising families. During shared decision-making discussions with families about initiating valganciclovir, clinicians should be transparent about the level of evidence, applicability in the patient’s individual scenario, and potential adverse effects of treatment.

The scientific community must also commit to generating high-quality data to guide evidence-based treatment of infants with cCMV. Delay in starting therapy, treatment of more than 6 months duration for very severe disease, treatment of isolated or late-onset SNHL, and treatment of mild or asymptomatic disease are being actively studied and could be of great benefit.60,61,62

Before initiating treatment, complete evaluation of infants with cCMV should include physical examination, complete blood count, bilirubin and transaminase levels, evaluation of kidney function, brain imaging (ultrasound, CT scan, or MRI, with ultrasound considered first). line for babies without neurological symptoms or microcephaly), ophthalmological examination and a complete diagnostic audiological evaluation.

Infants receiving antiviral medications should be closely monitored with frequent blood counts with differential counts including absolute neutrophil counts (typically weekly or biweekly for the first month(s) of treatment, then monthly for the duration of treatment), as well as monitoring routine monitoring of transaminases and kidney function.

The incidence of late-onset, progressive and fluctuating hearing loss requires continuous audiologic monitoring of all patients with cCMV.

Despite attempts to identify risk factors for hearing loss, it is not possible to predict which patients with cCMV will develop late-onset hearing loss or which patients with hearing loss are at risk for further progression. Audiologic evaluation should be completed every 6 to 12 months with consideration of more frequent testing during the first year.22,63 Routine audiologic monitoring is recommended until age 4 to 6 years in patients with cCMV, after which resume regular hearing screening (typically performed at school) in those without hearing loss.19 Children with SNHL should undergo auditory rehabilitation, with amplification and early intervention in speech development, to optimize hearing outcomes and prevent delays in speech development. speech and language.

Cochlear implantation is an effective treatment for patients with severe to profound SNHL and deafness. The benefits of cochlear implantation are well established and include improvement in hearing thresholds, speech perception, and speech expression.64,65,66,67 In patients with cCMV and unilateral SNHL, cochlear implantation is recommended early in the ear with SNHL due to the high risk of SNHL progression in the contralateral ear. This also helps prevent a prolonged period of hearing deprivation and further developmental delay. Children with additional disabilities, including neurological impairment, may also benefit from a cochlear implant.68

CMV infection is a common congenital infection with a spectrum of significant manifestations and morbidities in a subset of affected infants. Late-onset sequelae, especially SNHL, may occur in infants with cCMV who are asymptomatic or symptomatic at birth. A broad and rapid screening approach is essential to identify neonates affected by cCMV to ensure that a complete evaluation is performed and, where appropriate, treatment with antiviral medications is initiated.

The critical window for timely and definitive diagnosis is within the first 3 weeks after birth using CMV PCR in saliva or urine, with a confirmatory urine PCR recommended in the case of a positive saliva PCR. A complete evaluation of infants with cCMV includes a complete physical examination, complete blood count, liver and kidney function tests, neuroimaging, ophthalmologic evaluation, and audiologic testing.

Antiviral treatment may improve eventual hearing and neurodevelopmental outcomes in symptomatic infants with moderate to severe manifestations of cCMV, and active studies are currently underway to understand whether other infants with cCMV may benefit from valganciclovir. Infants not treated with antivirals require audiologic monitoring for late-onset hearing loss and/or progression of SNHL. For patients with hearing loss, rapid initiation of interventional therapies, hearing augmentation, and, in the case of severe to profound loss, cochlear implantation, may improve outcomes.

Table 1. Impact on infants with cCMV as a result of primary versus non-primary maternal infection5,6,7,8,9

Table 2. Recommendations for antiviral treatment (valganciclovir)59

Congenital CMV is the most common intrauterine infection, and can occur during a primary maternal infection or a non-primary infection in seropositive pregnant women by reactivation of the latent virus or infection with a different strain, having a different impact on affected infants.

Maternal infection in the first half of pregnancy is associated with a lower risk of transmission to the fetus, but with a higher risk of serious sequelae if infection occurs during this period.

It is important to implement broad and rapid screening to identify affected neonates, perform a complete evaluation and, when appropriate, initiate treatment with antivirals. To prevent cCMV, preventive measures can be followed, such as hand hygiene and avoiding contact with other people’s saliva.

There is no universally recommended CMV screening test in pregnancy and, to date, there is no proven intervention to decrease fetal transmission. Work is currently underway on new treatment modalities, and on the development of possible vaccines against CMV.