Congenital cytomegalovirus: the case for targeted infant screening in Australia

Australia lacks a comprehensive approach to CMV screening, and most infected infants go undiagnosed

For the developing fetus, congenital cytomegalovirus (CMV) represents one of the greatest threats to normal development posed by an infection. The virus is a leading cause of developmental disability and a common cause of permanent hearing loss in infants.1,2 Ninety percent of infected infants remain asymptomatic, making universal screening (testing all infants) difficult to justify.2 Despite the high asymptomatic rate, the significant prevalence and infectivity result in congenital CMV affecting more children than other disorders universally screened for in newborns in Australia. Congenital CMV infection is a significant if poorly recognised public health issue.

There is no CMV vaccine, despite 80 years of research and it being ranked the highest priority vaccine to develop relative to economic burden and years of life and disability that would be saved.3

Sensorineural hearing loss is the most common permanent sequela of congenital CMV infection and affects more than 50% of symptomatic cases.1 Peak national advisory bodies recommend congenital CMV testing for all infants with sensorineural hearing loss.4 Diagnosis and treatment of congenital CMV infection are time critical. In order to distinguish congenital and acquired CMV, testing must be completed by day 21 of life. This narrow window, limited congenital CMV awareness and the lack of a comprehensive approach across Australia means most cases go undiagnosed, and for symptomatic children this may represent a lifelong disability of unknown cause.1

Universal newborn screening programs have been established in Australia for around 25 conditions.5 Testing infant hearing via newborn hearing screening is an example of this. Universal screening for congenital CMV in Australia is possible but untried, and few overseas studies exist. In contrast, targeted testing identifies those at high risk for a condition and focuses testing on this group. Targeted congenital CMV testing typically uses a failed newborn hearing screening as a high risk indicator.

Targeted congenital CMV testing programs exist overseas; however, such programs have only occurred within a research setting in Australia.

We reviewed the evidence for congenital CMV screening to establish the case for a targeted approach in Australia, contextualised with reference to a current Western Australian trial (Australian New Zealand Clinical Trials Registry number ACTRN12621000484842; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=381390&isReview=true) assessing the feasibility of a state‐wide targeted congenital CMV screening program.

We searched online databases including PubMed, MEDLINE and the Cochrane Database of Systematic Reviews, limited to articles published in English since 2007. Search terms used included “congenital cytomegalovirus”, “congenital CMV”, “paediatric hearing loss”, “childhood hearing loss” and “congenital sensorineural hearing loss”.

Epidemiology, diagnosis and treatment

Despite the high rate of asymptomatic congenital CMV, the high prevalence rate (0.64%) results in congenital CMV being the leading cause of congenital malformations in developed countries, affecting more children than better known conditions such as Down syndrome or spina bifida.6,7 Hearing loss due to congenital CMV varies in age of onset (birth to 6 years of age), symmetry (bilateral or unilateral) and severity (mild to profound), and may fluctuate throughout childhood.8

Across Australia, the lack of an organised approach to screening has resulted in a very low diagnosis rate — less than 2% of cases according to a decade‐long study.9 The diagnostic process itself for congenital CMV infection is straightforward and aligns with the criteria for a suitable screening test.10 Polymerase chain reaction (PCR) testing of saliva or urine by day 21 is the gold standard and has high sensitivity and specificity (over 90%).11 PCR testing is rapid, widely available and cost‐effective (A$28).12,13 However, after day 21 of life, neither test can differentiate between acquired or congenital CMV, and dried blood spot specimens (as used in the Guthrie Card test) have low sensitivity for congenital CMV.14 A Queensland study found timely PCR testing for congenital CMV within a newborn hearing screening program to be achievable and at least cost‐neutral.15 The WA trial is using saliva PCR testing and confirming positive cases via urine samples to avoid false positive results. Retained breast milk in baby saliva may contain maternal CMV.

A diagnosis of congenital CMV‐related hearing loss prompts further clinical reviews, a broad approach to which is outlined in Box 1.

Antivirals such as ganciclovir and valganciclovir can treat certain sequelae of congenital CMV infection. Their use is relatively novel, with a limited evidence base. Antivirals have the potential to cause serious side effects so are usually reserved for infants with central nervous system complications, from which isolated sensorineural hearing loss is excluded. Day 30 of life is the usual upper limit for initiating antivirals due to the unclear efficacy when started after this point. Antivirals may reduce congenital CMV‐related hearing loss and can improve neurological outcomes in affected babies; however, longitudinal data and further trials are needed.17 Conclusive recommendations regarding antivirals in isolated congenital CMV‐related sensorineural hearing loss are not possible at present. In the setting of severe to profound sensorineural hearing loss, an approach among infectious disease experts is to broach the options with the affected family while noting the limited evidence and the associated risks.

Infant screening

Universal congenital CMV screening would enable diagnosis of all affected infants, including asymptomatic cases. Guidelines recommend monitoring asymptomatic cases; however, without universal testing, none are likely to be recognised.4 Identification of an infant with asymptomatic congenital CMV infection should prompt reviews similar to those outlined in Box 1.

Modelling data from the United States suggest that congenital CMV testing may reduce severe to profound hearing loss by 4.2–13.0% and is cost‐saving.13 While screening increases caseload, it also reduces unnecessary tests and focuses on early intervention, which are highly cost‐effective.1 The strongest argument in support of universal congenital CMV screening lies in the cohort of infected infants who are initially asymptomatic but develop delayed sequelae (eg, hearing loss or intellectual disability) — this group represents the highest disease burden and comprises those who may otherwise be fated with a lifelong disability of unknown cause.1

Universal congenital CMV screening in Australia would diagnose almost 2000 infants annually and warrants consideration.9 Longitudinal data from health services could confirm predictive studies and guide decision making. Targeted congenital CMV screening is an emerging field which would benefit from randomised controlled trials and further longitudinal data. The WA trial will contribute valuable data, including those of a longitudinal nature.

Some studies of targeted screening for congenital CMV have resulted in targeted screening programs being established and others have not. The United Kingdom National Screening Committee decided against the approach in 2017.18 Whether targeted screening for congenital CMV is worthwhile may be subject to geographic variability and screening protocol differences. This was demonstrated by reviews of two programs in the US, where a Utah program reported a 6% incidence and a Connecticut study reported a much lower incidence of 0.6%.8,19 Further research would inform the best approach to targeted congenital CMV screening.

Across Australia, newborn hearing screening programs provide testing to all newborns. Given the success of these programs and the frequency of sensorineural hearing loss in congenital CMV, newborn hearing screening provides a strong basis for targeted congenital CMV screening.20 This is the basis on which the WA trial is operating — all infants across the state who meet the inclusion criteria and who do not pass newborn hearing screening are offered testing.

Targeted congenital CMV testing is feasible, cost‐effective and enables early intervention, all of which align with the World Health Organization criteria for a screening program (Box 2).10,15,21,22 Diagnosis can provide peace of mind and clarity and potentially curtails the exhaustive “diagnostic odyssey” investigative pathway. Targeted congenital CMV screening can be acceptable to families and need not provoke parental anxiety.21 Children with congenital CMV shed high viral loads for years, which for close contacts, especially women who are pregnant or planning pregnancy, warrants counselling to reduce transmission risk.16

Certain reported benefits of targeted congenital CMV screening (eg, costs or quality of life measures) are based on modelling or are drawn from congenital hearing loss data more generally.23 Whether targeted congenital CMV screening ultimately results in better clinical outcomes for the affected child is uncertain. Unlike sensorineural hearing loss, some congenital CMV sequelae are not readily evident in infancy, and targeted congenital CMV screening can provide benefit through the enhanced monitoring afforded to congenital CMV‐positive infants (Box 1). This can aid early intervention. A weakness of targeted screening is that it does not identify asymptomatic newborns, some of whom will develop delayed sequelae.14

The limitations of these analyses and the caution required when extrapolating international studies to a local setting must be acknowledged. A targeted approach to congenital CMV has limitations and risks; however, it also appears to be feasible and an improvement on the current approach in Australia.

Conclusion

The current approach to the diagnosis of congenital CMV infection across Australia is inadequate. Consequently, affected infants regularly fail to receive best practice care. Despite comprehensive newborn hearing screening programs, many children with sensorineural hearing loss are not routinely screened for congenital CMV and this is arguably an important gap within the program.

Targeted congenital CMV testing is a feasible approach to identifying affected infants and a move towards improved outcomes. Major obstetric and paediatric hospitals around Australia are increasingly recognising the importance of targeted congenital CMV screening, and further local studies focusing on longitudinal outcomes would offer insight into disease burden and guide prevention and treatment efforts.

The present approach across Australia warrants scrutiny and a shift is required in order to provide best practice care to affected children. Challenges to improvement relate in part to acquiring further scientific understanding; however, the limited level of awareness of the virus among parents, clinicians and policymakers may be the biggest obstacle to effectively addressing congenital CMV in Australia.