Cerebral palsy and childhood mortality associated with congenital cytomegalovirus in Australia

Abstract

BACKGROUND Congenital cytomegalovirus (cCMV) is a known risk factor for neurodevelopmental disabilities including cerebral pasly (CP). Both primary and secondary prevention strategies exist to reduce the risk of cCMV disease and to diminish the severity of sequelae resulting from infection. In the absence of newborn screening for cCMV in Australia, the epidemiology of cCMV is poorly defined. METHOD This thesis explored: a) the epidemiology of children with CP and confirmed/probable cCMV registered on population-based CP registers b) national mortality data to determine the burden of cCMV to childhood mortality, c) the neuroimaging patterns of children with CP and cCMV and d) the frequency of CMV DNA detected on the newborn screening cards (NBSC) of children with CP who consented to PCR-based CMV testing of their NBSC. RESULTS The analyses revealed that 1.5% of children (2.91 cases per 100 000 live births, 95% CI 1.9–3.9) registered on population CP registers (1993–2003) had confirmed/probable cCMV reported as an attributable cause of their CP. The cCMV reported group had a significantly more severe disability profile including an over-representation of children with spastic quadriplegia. Females were also over-represented (64% vs 43%; p=0.014) suggesting a possible survivor bias favouring females. This hypothesis was investigated through the interrogation of national mortality datasets (1999–2011). Here 83 CMV-associated deaths in children <15 years were identified with 22 occurring in infants <1 month. Deaths associated with CMV were distributed relatively evenly across males (55%) and females (45%) which did not support the hypothesis of a possible survivor bias favouring females with severe cCMV disease. In an investigation of neuroimaging reports of children with CP and cCMV reported to the CP registers, distinctive patterns of cCMV disease were noted including brain malformations with accompanying white matter abnormalities, calcifications, cerebellar hypoplasia and/or ventricular dilation. These patterns are suggestive of cCMV infection in early gestation (12–24 weeks). The final chapter of this thesis described the testing of NBSCs of children with CP for CMV DNA. CMV viremia was highly prevalent (9.6%) compared with the general population (0.6%). Epilepsy was more common in children with CP and cCMV, but female sex, and spastic quadriplegia were not overrepresented in this group. CONCLUSION This thesis demonstrated the importance of cCMV as a contributor to both childhood mortality and neurodevelopmental disabilities. Neonatal cCMV viremia was found to be highly prevalent amongst children with CP, across all spastic subtypes, compared to the general population. However, in the absence of neonatal screening, cCMV remained largely unrecognised. Future research should investigate whether reducing the incidence of fetal CMV infection through primary prevention strategies will in turn prevent CP. The costs and benefits of routine neonatal screening should also be evaluated as early identification of cCMV disease provides important opportunities to reduce the severity of sequelae from cCMV disease including the use of antivirals and referral to appropriate early intervention services.