Importance of Neuroimaging in the Evaluation of Zika Virus–Exposed Infants

A little more than 3 years ago, Zika virus (ZIKV) emerged as a newly recognized congenital infection associated with substantial risk to the developing fetal brain after infection during pregnancy. Although the virus had been circulating in French Polynesia and other regions before the epidemic in Central and South America in 2015 and 2016, its devastating effects on the developing brain were previously unrecognized. The study by Pool and colleagues¹ highlights the neuroimaging findings of a large number of ZIKV-exposed infants at a high-risk obstetric and pediatric referral center in Rio de Janeiro, Brazil, from March 2016 to June 2017. The medical team at the Fernandes Figueira Institute, Oswaldo Cruz Foundation, Rio de Janeiro, Brazil, rapidly established clinical care protocols for the evaluation of a large number of ZIKV-exposed infants. Infant evaluations included laboratory testing for ZIKV, neuroimaging with cranial ultrasonography (US), head computed tomography (CT), and brain magnetic resonance imaging (MRI). Infants were also evaluated by a multidisciplinary team including specialists in pediatric infectious disease and ophthalmology and had their hearing assessed by brainstem auditory evoked response. The phenotype of Zika syndrome was described at birth as severe, mild or moderate, or asymptomatic on the basis of clinical findings, head circumference, and neuroimaging findings. Infants with the severe phenotype had the findings described in association with congenital Zika syndrome.² The comprehensive infant evaluations were similar to the proposed guidelines set forth by the US Centers for Disease Control and Prevention for the diagnosis and evaluation of ZIKV-exposed infants.³

In the study by Pool et al,¹ 71 of 110 infants (65%) had abnormal head CT or brain MRI findings, with most having clinical classification at birth of the severe phenotype of congenital Zika syndrome. The high frequency of abnormal imaging findings would be expected in these infants with findings of microcephaly, fetal brain disruption sequence, arthrogryposis, and abnormalities on neurologic examination at birth. Conversely, among the 39 infants with a normal-to-mild clinical phenotype and normal neurologic examination findings, only 4 (10%) had abnormal neuroimaging findings.¹ Among the 4 infants who did not have findings of congenital Zika syndrome at birth, neuroimaging findings may be associated with other noninfectious causes. It is important to point out that the infants included in the study by Pool et al¹ underwent neuroimaging with either head CT or brain MRI after either abnormal findings on cranial US or because cranial US was not feasible owing to a small fontanelle, which was the result of either older infant age or microcephaly. The 110 infants included in the cohort thus represented a group of ZIKV-exposed infants who would be expected to have a high burden of neuroimaging abnormalities, which is a difference from other reported cohorts.⁴ Several brain abnormalities were seen in the infants, including calcifications and malformations in cortical development, which may be better imaged with CT and MRI, respectively.

Fortunately, many ZIKV-exposed infants do not have abnormal brain findings or a clinical phenotype associated with congenital Zika syndrome. In a prospective cohort of 82 pregnant ZIKV-exposed women,⁴ only 3 pregnancies (4%) were complicated by severe fetal brain abnormalities. Postnatal imaging with cranial US and brain MRI found additional abnormalities in the infants, but most were mild and not associated with a severe clinical phenotype.⁴ There appears to be a spectrum of abnormalities on brain imaging in ZIKV-exposed infants, including mild nonspecific changes seen at cranial US, such as lenticulostriate vasculopathy and germinolytic cysts, to more significant brain abnormalities, such as subcortical calcifications, ventriculomegaly, and, in its most severe form, thin cortical mantle and fetal brain disruption sequence.²

Congenital ZIKV infection has some unique neuroimaging features compared with other congenital infections.^2,5 The cerebral calcifications associated with ZIKV are typically subcortical and can be seen also in the basal ganglia,^1,2 but the location is distinctly different from the periventricular location classically seen in the setting of congenital cytomegalovirus infection.⁵ The cerebral mantle can be severely thin, and infection may even result in fetal brain disruption sequence, a finding that is unusual for other congenital infections that cause microcephaly.^2,5 The abnormal neuroimaging findings reported in the study by Pool et al¹ include the full range of brain abnormalities that are more uniquely seen in ZIKV, and this may be associated with the spectrum of clinical findings seen in affected infants.

As we continue to learn about ZIKV and the spectrum of sequelae to the developing brain, we are also now starting to understand the long-term neurodevelopmental consequences of ZIKV exposure and the neuroimaging changes in patients with microcephaly and ZIKV-associated brain damage.⁶ Infants born in Brazil and other areas during the epidemic are now of early childhood age, and those with the severe neurologic phenotype of congenital Zika syndrome have considerable neurologic impairments, including epilepsy, severe developmental delay, deafness, and visual impairment.⁷ Not as much is known about the less-affected children and their developmental outcome. Linking early neuroimaging abnormalities in ZIKV-exposed infants to neurodevelopmental trajectory is the next major important area for understanding the burden of ZIKV on the developing child. Public health and educational systems need to be prepared for the ongoing medical, physical, and educational needs of ZIKV-exposed children.

Centered on the findings by Pool et al¹ and others,^2,4 early neuroimaging remains one of the most valuable investigations of the ZIKV-exposed infant. However, access to neuroimaging is limited in some areas, and this may limit clinicians’ ability to recognize all affected infants. When available, cranial US should be performed as a first-line imaging modality, together with a neurologic examination, ophthalmologic examination, and brainstem auditory evoked potentials. If the findings of these studies are normal, no further imaging is warranted unless there is a developmental disturbance over time.³ If cranial US findings are abnormal or the infant cannot undergo cranial US, then neuroimaging with low-dose cranial CT or brain MRI should be performed to fully evaluate for ZIKV-associated brain injury. Neuroimaging thus remains an important component of the evaluation of any ZIKV-exposed infant, not only infants with congenital Zika syndrome.

Published: July 31, 2019. doi:10.1001/jamanetworkopen.2019.8137

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2019 Mulkey SB. JAMA Network Open.

Corresponding Author: Sarah B. Mulkey, MD, PhD, Division of Fetal and Transitional Medicine, Children’s National Health System, 111 Michigan Ave NW, Washington, DC 20010 (sbmulkey@childrensnational.org).

Conflict of Interest Disclosures: Dr Mulkey reported support for Zika research from the Thrasher Research Fund and has provided technical expertise to the US Centers for Disease Control and Prevention Zika virus studies.

Additional Information: This commentary was written on behalf of the Congenital Zika Program at Children’s National Health System, led by Roberta L. DeBiasi, MD, MS, and Adre J. du Plessis, MBChB, MPH.