Encouraging postnatal cytomegalovirus (CMV) screening: the time is NOW for universal screening!

KEYWORDS:

• Cytomegalovirus
• congenital cytomegalovirus infection
• ganciclovir
• hearing
• newborn hearing screen
• screening
• valganciclovir

Congenital cytomegalovirus (CMV) infection is a huge public health problem that affects approximately 20,000–40,000 infants born each year in the United States and over a million newborns globally [1]. CMV acquired by the fetus is the most common nongenetic cause of sensorineural hearing loss (SNHL) as well as a major contributor to permanent neurologic disabilities and cognitive deficits in childhood [2]. With an estimated annual cost of up to $4 billion in the United States alone, the medical burden and societal costs of congenital CMV infection and the personal toll on affected families must be addressed – and we are more than capable of doing it now!

Currently and at best, only neonates with clinically apparent congenital CMV infection are identified, that is, those who are small for gestational age (SGA) and have such signs as petechiae, thrombocytopenia, jaundice, direct hyperbilirubinemia, hepatosplenomegaly, or microcephaly. But these so-called ‘symptomatic’ infants represent only 10–15% of all CMV-infected newborns! Certainly, these infants who have clinical, laboratory, or neuroimaging abnormalities are likely to have long-term sequelae. Approximately 50% of them will have or develop SNHL, mostly in the first 3 years of age although the risk is likely to be lifetime [2,3]. In addition, many will have intellectual disability with IQ < 70 as well as other neurodevelopmental impairments. Findings early in life that are predictive of adverse neurologic outcomes include microcephaly, chorioretinitis, and abnormal neuroimaging that demonstrate neuronal migration disorders such as cortical dysplasia and gyral abnormalities [4]. There should be no need for newborn CMV screening in these ‘symptomatic’ infants as they should be identified readily by health-care professionals – but in fact, many are not! Too often SGA, thrombocytopenia, and other signs of possible CMV infection are attributed to conditions such as ‘culture-negative sepsis’ or maternal exposures such as preeclampsia or illicit drug use, and testing for CMV sadly is never performed. Since awareness of congenital CMV infection is relatively low among health-care professionals, education programs focused on the varied presentations of congenital CMV infection are needed so that CMV can be part of the differential diagnosis and appropriate testing is performed in a timely fashion.

The vast majority of CMV-infected newborns, however, are never diagnosed as the infection usually is clinically inapparent at birth even though it may result in substantial sequelae [2]. Approximately 10% of so-called ‘asymptomatic’ neonates will develop SNHL, 5% neurodevelopmental impairments, and 2% chorioretinitis. When these later signs become apparent in childhood, it is virtually impossible to diagnose congenital CMV infection retrospectively. Congenital infection requires detection of the virus in the first 2–3 weeks of age. When CMV is identified by culture or polymerase chain reaction (PCR) in a child after 3 weeks of age, the precise mode of transmission is extremely difficult if not impossible to establish as it could represent persistent viral shedding from congenital infection, intrapartum acquisition, or postnatal infection by either horizontal transmission (e.g. daycare attendance, siblings) or breastfeeding. Hence the need for newborn CMV screening!

Since all newborns in most countries have dried blood spots (DBS) collected on filter paper for newborn screening programs, there has been considerable interest in utilizing these specimens for CMV DNA detection. In 2000, Barbi et al. [5] demonstrated their potential utility when DBSs that had been collected in the first week of age were retrieved and tested 15 days–4 years later for CMV DNA by a nested PCR method. Of 72 infants who were culture positive for CMV at <3 weeks of age, all had a positive CMV DNA PCR DBS. More recently, however, in the largest prospective study of DBS screening for CMV DNA detection among 20,448 newborns, Boppana et al. [6] demonstrated a sensitivity of only 28.3% but a specificity of 99.9% of the DBS for identification of newborns with congenital CMV infection. At the same time, Boppana et al. [7] showed that CMV screening using saliva swabs had excellent sensitivity of 97.4–100% and specificity of 99.9%. Other investigators have utilized urine and umbilical cords for congenital CMV screening with some success [8,9], but the ease of collection of saliva swabs makes this method the currently preferred one for routine newborn screening [10]. A positive CMV saliva screen should be confirmed with a urine CMV PCR test or culture, as collection of saliva after breastfeeding could detect contamination from CMV present in breast milk rather than actual infection [11]. In addition, if the diagnosis of congenital CMV infection is likely, promptly performing a urine CMV test avoids the delay in diagnosis using the two-step methodology.

So how should screening be performed? Two approaches have been proposed and recommended for the timely identification of neonates with congenital CMV infection, namely targeted and universal screening. Targeted CMV screening has been in the forefront recently as an approach for identification of CMV infection among newborns who do not pass the hearing screen. In 2008, Stehel et al. [12] showed that it is both feasible and important, as 6% of newborns who referred on the hearing screen had congenital infection and 75% of them were identified solely on the basis of the failed hearing screen. More recently, however, Fowler et al. [13] showed that although such a targeted approach identified the majority of infants with CMV-related SNHL at birth, it failed to identify as many as 43% of infants who develop CMV-related hearing loss in later infancy and childhood. Other problems exist with this targeted approach. Hearing screening may not be able to be performed before 3 weeks of age on premature infants less than 34 weeks’ gestation and older gestational age ill infants. Subsequent interpretation of CMV positivity when these infants refer on the hearing screen at discharge is problematic. Detection of CMV DNA from stored DBSs has been advocated [14], but the sensitivity is suboptimal and the testing usually is performed only in research laboratories. In addition, many states discard the DBS after a certain period of time or the parents did not provide the state with permission to keep the DBS after routine newborn screening was performed. Currently, Utah and Connecticut mandate CMV testing of neonates who refer on the hearing screen, while Illinois requires that parents be informed of CMV as a possible cause of the failed hearing screen and be provided the opportunity for CMV testing of the infant before discharge home.

A targeted approach also should include CMV testing of any newborn who has any clinical, laboratory, or radiographic sign of congenital CMV infection, e.g. SGA, hepatosplenomegaly, thrombocytopenia, periventricular calcifications, or lenticulostriate vasculopathy. In addition, 2–7% of infants born to mothers infected with the human immunodeficiency virus have congenital CMV infection and thus should be part of a targeted screening program [15].

It is clear that in 2017, targeted CMV screening should be an oxymoron! Clinicians should want to know the etiology of clinical and laboratory abnormalities in their patients, and if a newborn refers on a hearing screen or has any abnormality where CMV is in the differential, then appropriate testing for CMV should be performed. The problem remains that targeted screening fails to identify a substantial number of CMV-infected neonates who are at risk of developing late-onset hearing loss and neurodevelopmental problems.

The alternative approach is universal CMV screening of all newborns [10,16]. Such a strategy would circumvent all of the previously mentioned problems with targeted screening. But are we there? Certainly, congenital CMV screening conforms to the basic definition of a screening strategy, that is, one that is used in a population to detect a disease in individuals without signs or symptoms of that particular disease. As many as 90% of CMV-infected newborns have clinically inapparent infection and are not identified in the neonatal period, yet, months to years later, about 15–20% will develop late-onset SNHL [2].

In 1968, the World Health Organization established principles to base implementation of newborn screening programs, namely the prevalence of the disease (yes for CMV), availability of a suitable screening test (yes for CMV), and the accessibility of safe and effective therapies (yes for CMV). It certainly appears that congenital CMV screening has achieved such a status.

Among CMV-infected infants with clinically apparent disease, treatment with intravenous ganciclovir or oral valganciclovir improves hearing outcomes at 24 months of age and possibly developmental outcomes as well [17–19]. Nonetheless, debate remains among pediatric infectious diseases specialists as to which ‘symptomatic’ infant qualifies for treatment, especially since hearing deterioration may still occur later in life despite treatment in the first months of age. However, the importance of maintaining some hearing in the first 12–24 month of age cannot be overemphasized, as it is critical for establishment of basic speech and learning patterns.

It is possible that universal CMV screening could result in psychological discomfort and anxiety among new parents. However, it seems that such screening is acceptable to most mothers of CMV-infected newborns since they often are frustrated by their lack of knowledge of CMV and the effects that it may have on their child [20].

A continuing and unresolved dilemma is how best to evaluate the neonate who has a normal physical examination and screens positive for CMV infection. Beyond hearing and ophthalmologic evaluations, there is no consensus, even among pediatric infectious diseases specialists, as to what tests should be performed. Should all of these neonates have complete blood cell and platelet counts, liver enzyme concentrations, and/or neuroimaging (e.g. cranial ultrasound)? In short, they should, as some infants with congenital CMV infection and a normal physical examination have abnormalities detected on expanded testing that impact the decision to institute antiviral therapy. On the other hand, the completely normal neonate with congenital CMV infection and a normal evaluation should not receive antiviral treatment, but instead have hearing screening performed every 6 months up to 4 years of age and then yearly for early detection of late-onset hearing loss. Ultimately, predicting which neonate with clinically inapparent CMV infection will develop sequelae is the black box of congenital CMV, and biomarkers to identify such infants are needed urgently to help guide optimal and appropriate management.

In conclusion, the time for universal CMV screening is NOW! Although both targeted and universal CMV screening has been shown to be cost-effective [21,22], universal screening provides larger net savings and the greatest opportunity for directed care.

The prevalence of congenital CMV infection, its associated sequelae, the availability of a simple saliva screening tool, available antiviral treatment, and directed therapies for hearing impairment mandate that we act now to make universal screening a reality!

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper was not funded.