Neurological, Radiological, Visual, and Auditory Alterations in Children with Intrauterine Exposure to the Zika Virus

1. Introduction

Congenital infection caused by the Zika virus was first described in Brazil after an outbreak of microcephaly initially identified in the northeast of the country and later named Congenital Zika Syndrome (CZS), a direct consequence of viral infection during fetal life. In addition to microcephaly, it is also responsible for brain, auditory, and visual abnormalities, with profound implications for the newborn and their family and a substantial impact on global health.

As time has passed since the onset of the Zika virus outbreak in Brazil, additional information is needed to fully describe the clinical spectrum of findings associated with congenital Zika virus infection. Publications suggest that microcephaly at birth is not an essential component of CZS [1,2,3]. Knowledge remains relatively limited regarding the growth and neurological outcomes of children with congenital syndrome in the early years of life, especially in those without congenital microcephaly.

Besides microcephaly, other neurological, visual, and auditory manifestations can substantially impact these patients’ lives, with significant individual and social consequences. There are few reports in the medical literature regarding the outpatient follow-up of children affected by congenital Zika virus infection, which is almost always associated with microcephaly, underscoring the importance of describing the effects on children with normal head circumference. The primary objective of this study was to describe neurological, visual, and auditory alterations in children whose mothers had confirmed Zika virus infection during pregnancy, with most of these children not presenting microcephaly at birth.

2. Materials and Methods

2.1. Study Design and Setting

An observational, longitudinal, and prospective study was conducted at the Maternity School Hospital, from the Federal University of Rio de Janeiro, Brazil (ME-UFRJ). The source population consisted of all pregnant women who agreed to participate in the study from 12/03/2015 to 01/31/2017 and who had confirmed infection by the Zika virus during pregnancy through (1) detection of viral RNA by reverse transcriptase reaction followed by polymerase chain reaction (RT-PCR), in any biological sample (blood, urine, amniotic fluid, placental fragment, etc.) and/or (2) positive IgM serology for the Zika virus, confirmed by an increase in IgG neutralizing antibodies ≥ 4 times concerning the titers of these antibodies against the dengue virus. The test was considered inconclusive if IgG against the Zika virus was < 4 times higher than IgG against the dengue virus. The study population consisted of children born to this source population, included consecutively, who were monitored at the ME-UFRJ follow-up clinic from birth to 30 months of age.

Pregnant women with suspected infection or those with microcephalic fetuses were selected for laboratory investigation. Blood collection for RT-PCR was ideally performed up to the 5th day from the onset of symptoms but could be extended up to the 14th day. A urine sample for RT-PCR was collected up to the 28th day. In addition, blood samples from pregnant women were collected between the 7th and 14th day to test for immunoglobulin M (IgM) and immunoglobulin G (IgG). Investigation for dengue and chikungunya, in addition to other diseases in the TORCH group, was also performed. All material was sent to the central laboratory of the Health Department of the State of Rio de Janeiro, and the pregnant women began to be monitored in the prenatal care of ME-UFRJ. At delivery, umbilical cord blood was collected for RT-PCR for ZIKV.

2.2. Newborn Monitoring in the Immediate Neonatal Period

For all confirmed cases of newborns with intrauterine exposure to the Zika virus, laboratory, imaging and other screening tests were performed: (1) RT-PCR for Zika virus, Chikungunya and dengue in the cord blood and cerebrospinal fluid of newborns with congenital microcephaly; (2) Serology (IgM and IgG) for Zika virus, Chikungunya in peripheral blood on the fifth day of life; (3) Collection of peripheral blood on the fifth day of life for TORCH screening; (4) Transfontanellar ultrasound with Doppler: performed in the first week of life, through the anterior and temporal fontanelle, performed with a Philips HD7 XE ultrasound device, with a C8-5 mini convex probe; (5) Neonatal Hearing Screening: Transient Evoked Otoacoustic Emissions (TEOAE), performed in the first 24 to 48 hours of life, following the guidelines of the Neonatal Hearing Screening Care Guidelines of the Ministry of Health [4]. For preterm infants, the Brainstem Auditory Evoked Potentials (BAEP) were performed at the corrected gestational age of 34 weeks, to avoid erroneous interpretations secondary to prematurity or immaturity of the auditory system. All exams were performed by the same examiner, using Otoport Lite OEA DP+TE (Otodynamics, Hatfield, Hertfordshire, England), which uses non-linear click stimulus (which covers frequencies from 1 to 5 KHz), with equivalent peak intensity of approximately 84 dB SPL (decibels sound pressure level). An acceptable stimulus stability of 70% or greater was considered for the exam to be valid, and the standard noise rejection value was 47 dBNPS. For the test to be considered positive (TEOAE present), otoacoustic emissions had to project above the noise level in most of the frequency bands covered, that is, the signal-to-noise ratio had to be greater than or equal to 3 dB and the reproducibility by frequency band or overall, greater than or equal to 50%; (6) Neonatal Eye Screening: performed by a trained pediatrician. Once any alteration was detected, the neonate was referred for evaluation by an ophthalmologist and, if necessary, directed to a specialized rehabilitation service (Specialized Center for Rehabilitation with visual modality or Visual Rehabilitation Unit) or even to specialized ophthalmology services, according to the “Guidelines for Eye Health Care in Childhood” of the Ministry of Health [5].

2.3. Outpatient Monitoring

General care for the newborn followed the standards recommended by the Ministry of Health [6], and the first consultation with the pediatrician was preferably during the first week of life. The remaining consultations occurred monthly in the first 6 months, bimonthly from 6 to 12 months, quarterly from 12 to 18 months, and quarterly from 18 months of life. Referral to specialized services was made whenever considered necessary. The intervals between consultations with the various specialists were determined individually in case functional impairment was confirmed.

Global neurodevelopment was assessed longitudinally by a pediatrician, pediatric neurologist, and physical therapist. The assessment by the neuropediatrician occurred with a 3-month interval between appointments, reducing the interval when the specialist deemed it necessary. A complete neurological examination was performed, and the Gesell neurodevelopment test was applied, corresponding to a direct assessment and observation of the quality and integration of behaviors. The analysis categories of this scale refer to the following areas: Adaptive Behavior, Gross Motor, Fine Motor, Language, and Personal-Social Behavior [7].

All patients were referred for computed tomography (CT) and/or magnetic resonance imaging (MRI) of the skull, even those without alterations initially revealed by transfontanellar ultrasound and with normal head circumference, undergoing a more accurate study of the brain anatomy. The images were acquired in the standard medical imaging format protocol 3.0 - Digital Imaging and Communications in Medicine (DICOM), distributed to a specific server for later analysis. Each participant underwent a battery of neuroimaging performed on a 16-channel multi-slice CT scanner (Siemens) and a 1.5T MRI magnet (General Electric). Anatomical MRI images were obtained with the following multiplanar pulse sequences: T1-weighted volumetric sequence, T2-weighted turbo spin-echo, FLAIR (Fluid attenuated inversion recovery), and diffusion.

The fundoscopy (FO) or retinal mapping was performed by the same ophthalmologist and occurred at three different times. The first evaluation was conducted in the first week of life, a second at 30 days, and a third at 6 months to identify possible changes compatible with congenital infection.

In most of the infants evaluated, a new evaluation with BAEP was repeated between 18 and 24 months of age (BAEP-a: Automated click BAEP; BAEP-neuro: Click BAEP for neurodiagnosis; BEAP-SF: Frequency-specific BAEP).

2.4. Ethical Aspects

The study was approved by the Research Ethics Committee of Maternity School Hospital, from the Federal University of Rio de Janeiro, Brazil (protocol code CAAE: 55465616.0.0000.5275, Opinion Number: 1,516,904, as of April 27, 2016). All participants signed a written informed consent form, agreeing to be included in the study along with their children.

3. Results

Of the 2,882 pregnant women admitted during the study period, 116 (4%) had suspected ZIKV infection, and 33 had this infection confirmed. There was one twin pregnancy and three fetal deaths, so 31 newborns were included in the study. Of the 33 pregnant women included, 31 (93.9%) had diagnostic confirmation of ZIKV infection through a positive RT-PCR test and two (6.1%) through positive IgM serology for ZIKV. The two pregnant women who had diagnostic confirmation through IgM serology were simultaneously IgM negative for dengue virus, which corroborated the recent diagnosis of ZIKV in these patients (Table S1 in the Supplementary Material). The main clinical and epidemiological characteristics of the pregnant women included in the study are presented in Tables S2 and S3 (Supplementary Material).

Laboratory investigation for congenital infection in the TORCH group was performed during prenatal care in most patients eligible for the study (Table S4, Supplementary Material). Two patients had indeterminate IgM serology for herpes virus. Laboratory, imaging, and fundoscopy (FO) were performed in both newborns, and congenital infection was ruled out. Two pregnant women had syphilis during pregnancy; they were treated, and congenital syphilis was ruled out in their newborns. One pregnant woman was a carrier of the human immunodeficiency virus diagnosed before pregnancy. This patient received antiretroviral treatment during pregnancy and intrapartum, and her newborn received chemoprophylaxis for HIV exposure. None of the tests performed in newborns to investigate congenital infection in the TORCH group showed positive IgM for toxoplasmosis, rubella, CMV, or herpes (Table S5, Supplementary Material).

Of the 31 newborns included, twenty-seven (87.1%) were born with a gestational age between 37 and 41 weeks, three (9.7%) between 28 and 32 weeks, and one (3.2%) with a gestational age between 33 and 36 weeks, totaling 12.9% of premature births. There were no births with a gestational age of less than 28 weeks. Twenty-two (71%) were female. One newborn (3.2%) was classified as small for gestational age, and five (16.1%) were admitted to the neonatal ICU.

Only one newborn (3.2%) had microcephaly at birth (head circumference of 28.5 cm), classified as severe microcephaly. His exposure occurred in the first trimester of pregnancy. This patient underwent a lumbar puncture on the second day of life. The results of the cellularity and biochemistry of the cerebrospinal fluid were normal, and the RT-PCR for ZIKV was negative. The fetus identified as microcephalic in the obstetric ultrasound suffered a spontaneous abortion. No newborn developed microcephaly in the postnatal period, no other associated malformations were identified in the group of children studied, and there were no deaths in the neonatal period. The clinical and epidemiological characteristics of the newborns included in the study can be found in Tables S6 and S7 (Supplementary Material).

Of the 31 children in the study, four patients were lost to follow-up and 27 (87.1%) were monitored in the outpatient clinic throughout the study period.

3.1. Auditory Evaluation

All children underwent some hearing assessment during the study period. Thirty-one children (100%) underwent TEOAE. Only three children (9.7%) underwent all hearing assessment tests, and 21 (67.8%) underwent TEOAE + BAEP. Table 1 presents the results of the hearing tests performed for each child tested. Only one child (3.2%) presented alterations in TEOAE (failure in the left ear), but the subsequent hearing assessment with neuro BAEP was normal. No child presented alterations in BAEP a, one (8.3%) presented alterations in neuro BAEP, and no child who underwent BAEP FS presented alterations in the examination.

3.2. Ophthalmological Evaluation

Thirty children (96.8%) underwent retinal mapping exams in the first week of life. Table 2 presents the results of the retinal mappings performed. Three children (10%) presented some alteration in the first fundus exam, with 2 of them maintaining the alteration found in subsequent exams, and the other had the exam normalized in the third and final evaluation.

3.3. Radiological Evaluation

Of the 31 children included, 28 (90.3%) underwent some neuroimaging examination: transfontanellar ultrasonography, cranial tomography, and/or cranial magnetic resonance imaging. Twenty-seven (87%) children underwent TFUS during their hospitalization. Twenty-six (96.3%) had normal results, and one (3.7%) showed abnormalities. Twenty-five children (80.6%) underwent brain MRI, and some also underwent computed tomography when the radiologist deemed it necessary to complement the examination. Seven children (28%) showed some alteration in the image. Table 3 shows the results of the radiological examinations of these patients.

3.4. Neurological Evaluation

In the evaluation performed by the neuropediatrician, twenty-two (81.5%) had the Gesell Neurodevelopmental Test applied between the ages of 12 and 27 months. Table 4 separately shows the neurological examination findings in the 22 patients who were evaluated between the ages of 18 and 27 months. Ten children (45.4%) presented some alteration in the examination. The most frequent alteration was language delay, which occurred in 31.8% of the cases, a suspected intellectual deficit in 9%, and hyperreflexia in 9%. All other findings, gait with hemiparesis, increased tone on one side, asymmetric posture, behavioral change with suspicion of autism spectrum disorder, microcephaly, global increase in tone, internal eversion of the feet during gait, clonus, opisthotonus, epilepsy, included thumbs, sleep disorder and irritability in 4.5% of cases each. The most serious findings were found in the microcephalic patient.

4. Discussion

This study demonstrated that children with intrauterine exposure to the Zika virus who did not have microcephaly at birth did not develop significant ophthalmological or auditory impairment. However, during outpatient follow-up of these children, it was observed that 50% of them presented some alteration in brain imaging, neurological examination, and/or neurodevelopment. The global delay in neurodevelopment of the only microcephalic patient was considered severe, and the findings in the other patients were considered mild in their absolute majority.

Microcephaly was uncommon in our sample, being present in only one of the 31 live births (3.2%). It was classified as severe microcephaly, and the head circumference of this child remained below three standard deviations (SD) throughout the first two years of life. Microcephaly was also identified in one of the fetuses on maternal ultrasound examination, but this evolved to death in the third trimester of pregnancy. Despite the importance and impact of microcephaly, studies have shown that the absence of microcephaly does not rule out the possibility of neurological and other system impairments. França et al., in a case series study of 1,501 newborns, classified suspected cases of congenital Zika into five categories based on neuroimaging and laboratory results for Zika virus. They identified many cases classified as definitive and probable, with the head circumference within normal limits [8]. A systematic review and meta-analysis of a sample of 2,941 infected women estimated the prevalence of microcephaly at 2.3% among all pregnancies and 2.7% when considering only live births [9], a value very close to that observed in our study. The same can be observed in the work of Brasil et al. (2016) [10], who identified microcephaly at birth in 3.4% of infected pregnant women; however, half of them in newborns small for gestational age and with proportional microcephaly, that is, the size of the head small, but proportional to the weight and length of the newborn, which may suggest that in these cases the microcephaly may not have been caused by congenital infection by ZIKV. In a cohort study of 182 children born to pregnant women with infection confirmed by RT-PCR testing for ZIKV, no microcephaly was found in the sample studied [11].

None of the children in the present study developed microcephaly after birth. Different findings were observed by Linden et al. (2016) [12] in 2016, in the state of Ceará, northeast Brazil, a region heavily affected by the ZIKV epidemic. In a retrospective study of 13 children with normal head circumference at birth, they observed deceleration of head growth at 5 months of age and 11 of them were classified as having microcephaly. In this same study, however, all children presented abnormalities in radiological findings consistent with congenital Zika syndrome, including decreased brain volume, ventriculomegaly, subcortical calcifications, and cortical malformations, showing the severity of neurological impairment in the sample studied.

This study demonstrated that the group of children whose mothers were proven to have ZIKV infection during pregnancy had predominantly normal hearing assessment exams concerning the conduction of the auditory neural pathways and the hearing threshold. It was possible to perform a long-term hearing assessment, a pioneering aspect of this study, with children evaluated both in the neonatal period and between 18 and 30 months later. A single child who presented alterations in the hearing assessment in the neurodiagnostic BAEP, demonstrating a delay in absolute wave latencies, was referred for medical evaluation by a specialist, and the presence of chronic effusion in the middle ear was diagnosed bilaterally. The results in our sample indicated that antenatal exposure to ZIKV did not significantly affect the hearing of the children studied. A similar result was found in the study by Fandiño-Cárdenas et al. (2018) [13], who did not find hearing loss in the first two years of life of children whose mothers had Zika during pregnancy. In other study carried out in the state of Ceará, northeast Brazil (as already mentioned, a region with a high prevalence of microcephaly during the Zika epidemic), involving children with laboratory-confirmed SZC or diagnosed based on clinical-epidemiological criteria, Leite et al. (2018) performed hearing screening with transient evoked otoacoustic emissions testing, immittance testing, and cochlear-palpebral reflex research. They observed that most of the children evaluated had preserved cochlear function, which is associated with cases suggestive of middle ear pathology [14]

Of the three children who presented alterations in the first retinal mapping examination (10%), one had their examination normalized in the third evaluation, with a frequency of 6.4% of children with altered ophthalmological examination at the end of the first year of life. Children who persisted with the alteration in the examination showed alterations in the optic nerve. This finding was also highlighted in the work of Zin et al. (2017) [15], who, in a cohort of 112 children with ZIKV infection confirmed by RT-PCR in maternal samples, abnormalities of the optic nerve and retina were the most common findings. In the study mentioned above, 21.4% of the children examined presented ocular abnormalities, of which 41.7% did not present microcephaly, and 33% did not present any abnormality of the CNS. The authors draw attention in this study to the fact that ocular abnormalities may be the only initial finding in children with congenital ZIKV infection, proving evaluation by a specialist in all exposed children, regardless of the stage of exposure during pregnancy and the presence of CNS alterations. In our study, the microcephalic patient did not present alterations in retinal mapping, which does not rule out the possibility of cortical visual impairment, a visual impairment related to brain damage that usually affects the visual processing centers and visual pathways of the brain. Visual impairment is still a topic little addressed in the follow-up of these patients. A study using tools to test visual acuity, function, and visual development milestones was developed by Ventura and Lawrence (2017) [16], who concluded that regardless of structural ocular manifestations, all children with SZC have some visual impairment.

Since the beginning of the microcephaly epidemic, studies have been published showing a characteristic radiological profile in patients with ZIKV. A study conducted in the state of Paraíba, northeast Brazil, an area heavily affected by the ZIKV epidemic, described the results of MRI and/or CT scans of the skull of fetuses and newborns with confirmed or strongly suspected ZIKV infection. The most notable change in the brain parenchyma described in all neonatal images was reduced brain volume. Abnormalities in cortical development associated with volume changes were observed in almost all cases analyzed, as well as abnormalities in the corpus callosum, ventriculomegaly, changes in cerebral gyrification, calcifications, changes in the brain stem and corpus callosum [17]. Sanz Cortes et al. also described the same findings in a study conducted in Colombia [2]. Such changes were also highlighted in the present study of microcephalic patients. The Microcephaly Epidemic Research Group studied the clinical and radiological characteristics of 104 infants with microcephaly in the state of Pernambuco, northeast Brazil, in 2015 and showed similar results [18]. Levine et al. (2017) [19] believe that two radiological findings are strongly suggestive of Zika: 1) severe microcephaly with microencephaly; 2) gross calcifications at the junction of the white and gray matter, findings that are pretty uncommon in other congenital infections that are part of the differential diagnosis of ZIKV infection.

It is worth noting, however, the presence of abnormalities in the brain imaging of children in the present study, which do not fit into the classification described here as part of Congenital Zika Syndrome, drawing attention to the greater scope of radiological signs that may be associated with congenital ZIKV infection, especially in children without microcephaly and with less severity of clinical signs and symptoms. Other studies of the follow-up of children with intrauterine exposure to ZIKV point in the same direction [11,20], although few studies focus on this approach.

In the present study, it was observed that only the microcephalic child presented neurological sequelae considered severe. However, 50% of the children studied presented some alteration in the brain imaging, neurological examination, or the Gesell Neurodevelopment Test. Premature children were evaluated for neurodevelopment according to the corrected gestational age. However, it is impossible to rule out that some delay in the child development milestone in this specific group may be associated with prematurity. A similar result could be observed in the study by Faiçal et al. (2019) [3], who studied 29 normocephalic children with intrauterine exposure to the Zika virus and described neurodevelopmental delay in 35%, with language delay in 31%, cognitive delay in 4% and motor delay in 3% of the children evaluated using the Bayley scale of infant neurodevelopment. The same can be observed by Moreira, Nielsen-Saines, and Brasil (2018) [11], who, in a study to evaluate neurodevelopment in children exposed to ZIKV, described changes in at least one of the assessment areas of the Bayley scale in 37.2% of the children submitted to evaluation (25.5% of them between 1SD and 2SD of normality and 11.7% below 2SD of normality). Moreira, Nielsen-Saines, and Brasil (2018) [11] also describe in the same study cognitive changes in 11.7% of the children evaluated, language in 26.6%, and motor in 19.1% of the cases. Language delay was the most commonly affected area of ​​child development in the present study and in the aforementioned scientific publications. These results are also compatible with the work published by Nielsen-Saines and collaborators (2019) [21], who evaluated neurodevelopment in 223 children with intrauterine exposure to the Zika virus and found neurodevelopmental delay in 28.7% of the assessed children and 3.7% of microcephalic children, once again showing that not only patients who presented microcephaly at birth are at risk of complications and sequelae.

Neurological findings such as global hypertonia, hyperreflexia, clonus, opisthotonus, double hemiparesis, sleep disturbance, and epilepsy were present only in the microcephalic patient in the present study. Similar findings in patients with microcephaly are described by several authors [22,23]. In this case, the severity of the neurological findings is naturally justified by the severe involvement of the CNS caused by the tropism of ZIKV for the brain, targeting neural progenitor cells, which magnetic resonance imaging findings can confirm. On the other hand, we also draw attention in the present study to a normocephalic child, whose radiological investigation performed through CT and MRI of the brain showed subcortical calcifications in the parietal lobes and predominantly in the frontal lobes and a slight reduction in the volume of the cerebral gyri. This patient developed asymmetric gait due to left hemiparesis, again drawing attention to possible sequelae in patients with a head circumference within the normal range at birth. Other findings, such as suspected intellectual deficit and behavioral changes with suspected autism spectrum disorder, were also observed in the present study. The same was reported in the study by Nielsen-Saines et al. (2019) [21], who observed a frequency of 2.1% of autism spectrum disorder in 2-year-old children exposed to ZIKV during pregnancy.

One of the limitations of our study is the small number of pregnant women included. This is because we adopted laboratory confirmation of Zika virus infection as the inclusion criterion in the study. Since the viremia period is short, most pregnant women were outside the ideal period for collecting biological material for the conclusive diagnostic test, RT-PCR. However, although many cohorts describe a larger sample, few are as rigorous in selecting cases, with many lacking proper laboratory confirmation. Another limitation is that we cannot confirm that the exposed fetuses were actually infected by the Zika virus, which is a limitation in the vast majority of studies, given the difficulty in diagnosing the disease in the neonatal period. Finally, it is impossible to rule out that the alterations found in the four premature infants in our study may be related solely to prematurity.

5. Conclusions

This study demonstrated that newborns of mothers proven to be infected with the Zika virus during pregnancy may present varying degrees of visual, auditory, and neurological impairment associated or not with microcephaly. However, the most severe neurological findings appear to be associated with microcephaly at birth and early maternal infection. Congenital ZIKV infection is undoubtedly a challenging disease for researchers and health professionals, mainly due to the number of uncertainties involved. Many factors related to its pathogenesis and the potential for impairment, especially in non-microcephalic newborns, have not yet been clarified, requiring more studies to clarify these points further.