Congenital cytomegalovirus (cCMV) infection can result in permanent neurological problems and is a potentially preventable cause of sensorineural hearing loss in the UK. There is an urgency to diagnose and assess cCMV as antiviral treatment and has only been shown to be effective if started in the first 4 weeks of life. A recent randomised controlled trial of 6 months of treatment using oral valganciclovir has shown modest benefit in preventing hearing deterioration and in improving some neurodevelopmental outcomes. Parents and clinicians need to make a timely and informed choice regarding antiviral treatment and ensure that relevant non-pharmaceutical interventions are considered. This paper brings together the current evidence regarding the diagnosis and treatment of cCMV, consensus from two paediatric infectious diseases centres and outlines research priorities.
- Infectious Diseases
- Paediatric Practice
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Epidemiology and transmission
The prevalence of congenital cytomegalovirus (cCMV) in developed countries is 0.3%–0.8%.1 The risk of transmission is higher during later stages of pregnancy; however, transmission during early pregnancy is more likely to have severe consequences for the fetus.2 The rate of transmission during pregnancy is 30%–40% in primary maternal CMV infection but is only 1% when there is maternal CMV reactivation/reinfection. In the UK, up to 50% women of reproductive age are CMV seronegative so can have a primary CMV infection during pregnancy; however, in populations with high seroprevalence, more symptomatic babies will be born to mothers who have evidence of previous infection.3–5
Around 10% of infants with cCMV have evidence of CMV disease at birth (often called ‘symptomatic’ disease). This typically involves the reticuloendothelial, hepatobiliary or central nervous systems (CNS) and can be mild to severe, with up to 5% mortality. Those with clinical features suggestive of cCMV need testing (box 1). Although an over-representation of babies born prematurely has been reported in some cohorts, testing premature babies routinely for cCMV is controversial.6 Baseline screening to differentiate between congenital and postnatal CMV infection is, however, helpful in extremely premature infants (<28 weeks gestation), who are at risk of symptomatic postnatal infection later in their postnatal course.
Clinical features to trigger screening for congenital cytomegalovirus (CMV) infection
▸ Intracranial ventriculomegaly (without other explanation)
▸ Calcification on cranial ultrasound (often periventricular)
▸ Congenital cataracts
▸ Failed neonatal hearing screen
▸ Petechiae or purpura in the newborn
▸ Prolonged jaundice with transaminitis
▸ Unexplained thrombocytopaenia
▸ Evidence of maternal primary CMV infection in pregnancy*
▸ Intrauterine growth retardation
▸ Sensorineural hearing loss—new diagnosis
*Seek expert clinical virology advice for interpretation of virological investigations in pregnancy.
†Baseline screening to differentiate between congenital and postnatal CMV infection is helpful for extremely premature infants, who are at risk of symptomatic postnatal infection.
Of the 10% with symptomatic disease at birth, two-thirds will have long-term neurological impairment. Up to 15% of children who have no clinical signs of CMV disease at birth (often called ‘asymptomatic’) may also go on to develop sensorineural hearing loss (SNHL).7 A detailed review of the clinical epidemiology of cCMV with estimates of prevalence of disease in high and low seroprevalence settings has recently been published.8
Which tests are needed to confirm cCMV?
Babies can easily acquire CMV postnatally: during delivery, through breast milk or through close contact with body fluids. For prognosis and treatment, it is critical to distinguish between congenital and postnatally acquired CMV. The diagnosis of cCMV is established by detection of CMV in body fluids in the first 21 days of life, ideally with two independent samples. The sooner after birth the tests are performed the more confident the diagnosis of cCMV can be made. Urine and saliva samples are preferred due to greater detection sensitivity, but blood (including the newborn blood spot) can also be used.9 ,10 If an infant is breastfed, then the saliva swab should ideally be taken at least 1 hour after a feed. PCR assays are used to detect viral DNA; this is more rapid and cheaper than viral culture.
Once cCMV is confirmed, what are the next steps?
Determine whether there is evidence of cCMV disease.
It is important to determine whether there is evidence of symptomatic cCMV (disease) as opposed to asymptomatic cCMV (infection) as this will guide further management.
Clinical features should be documented and blood, neuroimaging, audiology and ophthalmology assessments should be completed as soon as possible (table 1).
Consider early referral to a paediatric infectious diseases consultant for consideration of treatment, even while further investigations are being planned.
All families should be offered the advice and support of the national CMV support group (http://cmvaction.org.uk/).
Treatment and follow-up
Any baby with a diagnosis of cCMV and evidence of disease should be considered for 6 months of oral antiviral treatment. Treatment is with ganciclovir intravenously or its oral prodrug valganciclovir which inhibits CMV replication. Neonatal pharmacokinetic data support 16 mg/kg/dose of valganciclovir oral solution administered twice daily which provides ganciclovir exposure comparable with that of 6 mg/kg/dose of intravenous ganciclovir. This dose of ganciclovir has been shown to be effective in clinical trials in infants born at 32 weeks gestation or more (see below).11 ,12 A randomised controlled trial is currently evaluating benefits of treating older children with confirmed SNHL and cCMV (ClinicalTrials.gov NCT01649869).
Table 2 provides guidance on which infants should be offered treatment following a risk versus benefit discussion with the family.
There have been two randomised controlled clinical trials informing cCMV antiviral treatment decisions. In both trials, medication was started in the first month of life in infants over 32 weeks gestation. The first showed that 6 weeks of intravenous ganciclovir reduced the risk of progression or development of hearing loss at 12 months of age in infants with CNS involvement and had a positive effect on neurodevelopmental outcome.12 ,13 The second trial compared 6 weeks with 6 months oral valganciclovir in symptomatic cCMV disease, with and without CNS involvement, and demonstrated that the 6-month course modestly improved audiological and neurodevelopmental outcomes to at least 2 years of age.14 The benefit of 6 months versus 6 weeks treatment on hearing was more marked when there was baseline CNS disease, compared with infants with no CNS involvement. The studies were relatively small and the numbers of patients with a specific clinical feature were insufficient to show benefit of treatment in less symptomatic subgroups.
Treatment does have risks; short-term toxicity, including neutropenia, can be anticipated in around half of the patients treated with ganciclovir and in a fifth on valganciclovir.12 ,14 ,15 This may require treatment interruption and, rarely, administration of granulocyte colony-stimulating factor. Thrombocytopaenia and hepatotoxicity can occur in up to 30% of patients. For treatment with intravenous ganciclovir, risks of prolonged intravenous access need to be considered. Long-term toxicity, including potential effects on malignancy risk and fertility, is theoretical but parents should be counselled accordingly.16
Monitoring of toxicity and response to medication requires a pragmatic approach. Myelosuppression is most likely to develop in the first month, hepatotoxicity usually happens after the fourth month and renal toxicity is negligible, but any renal impairment may increase toxicity, as these drugs are renally excreted. If the viral load increases during treatment, it may indicate poor adherence or, more rarely, resistance.17 Table 3 suggests guidelines for monitoring treatment and follow-up with audiology, ophthalmology and neurodevelopmental services.
Our understanding of cCMV is still limited and individual centres will only care for small numbers of patients necessitating multicentre, national and international collaboration for research. The following are important research priorities:
What are the benefits of treatment in babies with milder symptoms or no symptoms at birth?
Is there a benefit of starting treatment after the neonatal period (study is currently recruiting in UK ClinicalTrials.gov NCT01649869)?
What is the safety profile of medication before 32 weeks gestation?
What are the long-term outcomes for babies treated with antiviral medication?
When is it cost-effective to screen for cCMV?
What is the added benefit of conducting MRI above cranial ultrasound performed by a skilled operator on prognosis and long-term outcomes?
cCMV is an important cause of morbidity. The diagnosis and workup for cCMV need to be timely, so that early treatment can be considered when all relevant investigations are available. The family and infant will commence a long-term relationship with health services and prompt care and counselling is crucial.
Test your knowledge in the multiple choice questions (see online supplementary appendix 1).
Contributors TS wrote the first draft. SL, SK, MS, PH and HL reviewed the content and suggested amendments that TS incorporated. All authors approved the final version.
Competing interests None declared.
Provenance and peer review Commissioned; internally peer reviewed.