Auditory findings associated with Zika virus infection: an integrative review

Introduction Possible associations between Zika virus infection and hearing loss were observed during the epidemic in the Americas. Objective To describe the auditory alterations, pathogenesis and recommendations for follow-up in individuals with prenatal or acquired Zika virus infection. Methods Bibliographic research conducted in March/2018–April/2019 at the main available databases. Article selection, data extraction and quality evaluation were carried out by two independent reviewers. Studies containing auditory evaluation of patients with congenital or acquired Zika virus infection; and/or hypotheses or evidences on the pathophysiology of auditory impairment associated with Zika virus; and/or recommendations on screening and follow-up of patients with auditory impairment by Zika virus were included. Results A total of 27 articles were selected. Sensorineural and transient hearing loss were reported in six adults with acquired Zika virus infection. Of the 962 studied children, 482 had microcephaly and 145 had diagnostic confirmation of Zika virus; 515 of the 624 children with auditory evaluation performed only screening tests with otoacoustic emissions testing and/or automated click-stimuli auditory brainstem response testing. Studies in prenatally exposed children were very heterogeneous and great variations in the frequency of altered otoacoustic emissions and automated click-stimuli auditory brainstem response occurred across the studies. Altered otoacoustic emissions varied from 0% to 75%, while altered automated click-stimuli auditory brainstem response varied from 0% to 29.2%. Sensorineural, retrocochlear or central origin impairment could not be ruled out. One study with infected mice found no microscopic damage to cochlear hair cells. Studies on the pathogenesis of auditory changes in humans are limited to hypotheses and recommendations still include points of controversy. Conclusion The available data are still insufficient to understand the full spectrum of the involvement of the auditory organs by Zika virus, the pathogenesis of this involvement or even to confirm the causal association between auditory involvement and virus infection. The screening and follow-up recommendations still present points of controversy.


Introduction
The Zika virus (ZIKV) epidemic in the Americas which began in 2015, due to its significant proportions and deleterious consequences to humans, drew the attention of both the medical and lay community. Although the most emblematic fact of this epidemic was the association of ZIKV infection during pregnancy with an alarming increase in congenital microcephaly cases, other clinical manifestations were also attributed to infection by this virus, 1---4 such as possible associations with auditory alterations, both in acquired 5---7 and congenital 8---16 infections.
Evidence of causality between ZIKV infection and fetal abnormalities include evidence of infection during pregnancy, a specific and rare phenotype of central nervous system abnormalities in infected fetuses and newborns, and identification of the virus in the brain tissue of affected fetuses and neonates. 17 Pioneering studies have identified a central nervous system virus tropism in infected fetuses, with devastating consequences for development, including microcephaly and other severe central nervous system malformations. 18,19 Garcez et al., through immunohistochemistry and electron microscopy using neurospheres and brain organoids, demonstrated that the Zika virus affects human neural cells, reducing their growth and viability, strongly suggesting that the virus impairs neurogenesis during human fetal development. 20,21 Considering the indications of ZIKV neurotropism and its association with malformations in affected fetuses, it may be suggested that this virus is capable of affecting auditory neural pathways or causing malformations in auditory organs, leading to their developmental impairment and, consequently, increased risk of functional or morphological auditory impairment, especially in prenatal infections. Hearing loss has been described in infections by other agents that cause congenital syndromes, such as Toxoplasmosis, 22,23 Rubella, 24,25 Cytomegalovirus, 25---27 HIV 25,[28][29][30] and the Herpes Virus. 25,31 Direct auditory organ injury by the virus or local inflammatory changes induced by the infection comprise the involvement mechanisms described to date.
It is known that early congenital hearing loss diagnosis and intervention facilitate the possibility of improving language and communication development prognoses in affected individuals, 32---35 which becomes even more important in a context of association with other malformations, limiting neuropsychomotor development, as in the case of central nervous system malformations. Thus, it is also important to better elucidate the auditory system involvement pathogenesis in both congenital and acquired infections so that preventive, therapeutic, screening or follow-up strategies can be developed.
Hearing loss, both congenital and acquired, represents an important factor that reduces the quality of life of affected individuals. Changes in auditory pathways related to ZIKV infection have been described in case reports or in small case series concerning children presenting the congenital syndrome and adults with acquired infection, but little is known about the spectrum of these changes, or their pathogeneses and prognoses. Limitations concerning knowledge on hearing impairment in infected individuals and the need for specific guidelines for early diagnosis and auditory rehabilitation make a review of the subject both relevant and necessary.
In this context, this integrative review aims to: (1) describe the functional or morphological auditory changes related to prenatal ZIKV exposure and acquired ZIKV infection; (2) describe the pathogenesis of these auditory alterations and (3) review the recommendations for hearing screening and follow-up for these individuals.

Methods
An integrative review was performed in a systematic way according to the recommendations of the Preferred Items for Systematic Reviews and Meta-Analysis ---PRISMA. 36 The PICO strategy was used, defining Participants (P) as individuals exposed to ZIKV in the prenatal period or individuals with acquired ZIKV infection; Intervention (I) as the functional or morphological evaluation of auditory pathways, and Outcomes (O) as functional or morphological auditory system alterations.

Literature search
The bibliographic research was carried out at the main available databases from March to April 2018 and updated until April 2019, namely PubMed/MEDLINE, LILACS, Scielo, Scopus and Web of Science, without restriction concerning language or publication date. The reference list of the selected articles, as well as Google and Google Scholar, were used to complement the search. The applied descriptors (MeSH terms) were: (Zika OR ZIKV) AND (Acoustic OR Audiometry OR Tympanometry OR Auditory OR ''Evoked Potentials'' OR Psychoacoustics OR ''Evoked Response'' OR P300 OR ABR OR BERA OR Hearing OR Hypoacusis OR Deafness OR Audition OR Dysacusis OR cochlear OR retrocochlear) (Appendix 1).

Study selection and data extraction
Two independent reviewers, an otolaryngologist and a pediatrician with training in epidemiology, selected articles in three stages: initially by title, then by reading the abstract, and finally by reading the full text, according to pre-established inclusion and exclusion criteria. The same reviewers independently extracted data from selected articles in digital form developed for the study (Smartsheet  ® , Inc., Bellevue, USA). Discrepancies concerning data selection and extraction were discussed among the reviewers at the end of each step, in order to reach a consensus.

Inclusion and exclusion criteria
All types of study design were included, as follows: (1) Description of patients with suspected or confirmed congenital ZIKV infection or patients with acquired ZIKV infection, undergoing any type of auditory evaluation; and/or (2) Hypotheses or evidence on the pathophysiology of auditory impairment associated with ZIKV infection; and/or (3) Recommendations on the screening and follow-up of patients presenting auditory impairment due to ZIKV infection.

Quality assessment of the included studies
The quality of each study was assessed by two independent reviewers with training in epidemiology. Case report studies were evaluated from a point of view of article  Figure 1 Flow diagram with the selection stages of the studies. The 27 studies included in the review were split according to the presence of each of the outcomes studied (some studies addressed more than one outcome).
writing quality, using CARE (Guidelines for Case Reports) recommendations. 37 Case series and experimental studies were evaluated for risk of bias with the instruments proposed by the NIH 38 Items of the STROBE (Strengthening of the reporting of observational studies in epidemiology) were added to the evaluation of writing quality of the case series 39 (Appendices 2 and 3). In addition, since many studies did not focus on auditory assessments, items were added to specifically assess the quality of the report and the methodology of this assessment. Both the overall quality of the studies and the quality of the applied auditory evaluations were categorized as low, moderate or high. Discrepancies were discussed among the reviewers in order to reach a consensus.
The GRADE system was applied to assess the quality of the scientific evidence. 40 The protocol of this review was recorded at the International Prospective Registry of Systematic Reviews (PROSPERO) under number CRD42018092819.

Data synthesis
The data were summarized in Figures and Tables.

Results
The search strategy initially identified 157 records, which, at the end of the selection process, resulted in the inclusion of 27 articles 5---7,8---16,41---55 ( Fig. 1 and Appendix 1). Three of these studies concerned acquired ZIKV infection in adults (n = 6), 5---7 16 studies concerned prenatal exposure to ZIKV infection (n = 962) 8---16,41---45,50,51 nine studies included considerations on the pathogenesis of auditory impairment 6---9,42,43,46,47,52 and 17 studies included recommendations on hearing screening and follow up for patients with congenital or acquired ZIKV infection 6---11,13,14,16,42,44,46,48,49,52,53,55 (Table 1). Most of the nineteen studies on acquired or prenatal exposure to ZIKV were considered of low or moderate quality, both in general and in relation to the performed auditory evaluations and the quality of scientific evidence for the prevalence of hearing impairment in children prenatally exposed to ZIKV is insufficient (Tables 1 and 2). These studies comprised 968 participants, composed of six adults with acquired ZIKV infection and 962 children exposed to the ZIKV during the prenatal period. Hearing screening in children exposed to Zika virus

Anals of II Brazilian Congress of Health Sciences
Case series Low Low  All adults presented clinical and epidemiological conditions compatible with ZIKV infection. Five were submitted for auditory evaluation exams, two of which confirmed ZIKV infection cases, while the others presented a diagnosis of probable or inconclusive infection due to possible cross-reaction with the Dengue Virus (DENV). One of the confirmed cases 5 reported only transient hearing impairment, but the patient did not undergo any type of objective auditory testing. The hearing losses supposedly related to acquired ZIKV infection displayed a predominantly sensorineural and transient nature ( Table 3).
The studies concerning children presented different inclusion criteria, which can be grouped as follows: (1) Children with positive ZIKV tests with or without microcephaly at term (n = 102) 9,10,51 ; (2) Children born to mothers with   (Table 4).
Among the 962 fetus or children studied, 482 presented microcephaly or neurologic changes; 245 were tested for ZIKV by PCR and/or IgM determinations in CSF and/or blood and/or urine, and a total of 145 presented laboratory confirmation of prenatal infection (Table 4).
Among the 624 children with informed auditory evaluations, 515 performed a hearing screening test with transient or distortion-product otoacoustic emissions testing (T-OAE or DP-OAE), and/or automated Auditory Brainstem Response testing (a-ABR) to determine auditory threshold values. These included 259 children in which the type of ABR stimulus was not specified, but were probably a-ABR. In one study including 102 children with microcephaly, the absolute number of children that performed ABR could not be determined. Nineteen children, in addition to a hearing screening, also underwent a diagnostic evaluation by a Frequency-Specific Auditory Brainstem Response testing (FS-ABR), one of which also performed a behavioral auditory evaluation. Forty-four children underwent tympanometry in addition to OAE, 68 children performed Hammersmith Infant Neurologic Exam (HINE) in addition to ABR, while 19 underwent only HINE ( Table 4).
The mean age of 296 children in the first hearing assessment was over 3 months old and not specified for the rest of the children. Most children underwent only one or two assessments at 1 month intervals. Only 43 children underwent three evaluations between 3, 6 and 24 months of age (Table 4).
Great variations in the frequency of altered OAE and a-ABR occurred across the studies: altered OAE varied from 0% to 75%, while altered a-ABR varied from 0% to 29.2% (Table 4).
Of the total number of patients who underwent OAE assessments (n = 244), 18.4% presented alterations, while 25% of microcephaly cases displayed alterations. Among the 448 patients who reportedly underwent the first a-ABR test (including those in which the type of ABR stimulus was not specified), 15.2% presented alterations, while in some studies data were not available to estimate the percentage of altered exams among microcephalic patients (Table 5).
The study carried out by Van der Linden et al. 51 examined children with late onset microcephaly and did not find any alterations in the performed hearing tests. The study performed by Marques Abramov et al. 43 was the only one to evaluate microcephalic children by neurodiagnostic click ABR, in order to evaluate auditory pathway integrity, and not auditory threshold values. The authors found only mild latency changes in three patients.
No histological studies in humans concerning auditory pathway involvement pathogenesis in ZIKV infection cases were found. The considerations of several authors on the pathogenesis are summarized in Table 6. In addition, no studies indicating morphological alterations in organs/auditory pathways related to ZIKV infection were reported. One experimental study was found addressing the auditory effects of the intrauterine exposure to ZIKV in no-human mammals. 54 The hearing evaluation was both functional, with electrophysiological tests ---ABR and Terminal Cochlear Action Potentials ---and histopathological, with the microscopy of the cochlear hair cells. Auditory alterations were found in 25---66% of the mice, in the FS-ABR, depending on the frequency of the tone-burst, with a greater number observed at higher frequencies. No deficits were observed when using click stimulus. It also reported the results of the microscopic analysis of the cochlea of these offsprings where no changes in the number of hair cells were observed.
Three official documents were found, formalizing recommendations for auditory screening and follow-up in patients exposed to the ZIKV during the prenatal period: two interim guidelines published in the Morbidity and Mortality Weekly Report by the Centers for Disease Control and Prevention      (Table 7). In the last CDC Interim Guideline update, 48 a change from the previous Interim Guideline 49 was noted, ruling out the need for a new ABR between 4 and 6 months or a 9 month behavioral audiometry for children exposed to ZIKV during gestation if they presented a normal ABR performed during the first month of life.
In 11 case series type articles, 6---11,13,14,16,42,44 one opinion piece 46 and two reviews, 52,53 the authors also suggest recommendations for screening and auditory monitoring concerning prenatal exposure to the ZIKV. Most studies recommend that this condition be considered as a risk indicator for hearing loss and, therefore, that children should be screened by OAE and a-ABR. They also highlight the need for regular follow-up regarding hearing development, due to the possibility of late hearing loss 9,10,42,46 (Table 7).
In two studies reporting cases of auditory impairment in ZIKV-acquired infections, the authors suggest auditory monitoring in these cases 6 and the need for further investigations to promote this recommendation during ZIKV outbreaks 7 (Table 7).

Discussion
Despite the involvement of several research groups in the study of ZIKV infection, this review found only 27 records that specifically contained one or more of the three proposed outcomes related to congenital or acquired ZIKV infection: auditory, functional and/or morphological manifestations, hypotheses concerning the pathogenesis of auditory system alterations and recommendations ''Hearing compatible with normal cochlear function presenting with neural synchrony changes can be classified as a neuropathy. As such, it is verified that this definition is compatible with the characteristics observed in both cases in this study''. ''In view of all the information quoted and analyzed, it is suggested that ZIKV can damage auditory nerve pathways and, in that way, impair communication in adult patients''. ''In this study, earlier absolute latencies were observed in children with microcephaly, with high-intensity stimuli for the investigation of the neurophysiological integrity of the brainstem. Considering the affinity of ZIKV for nerve tissue, it is possible that the neuroconduction of the acoustic stimulus in children exposed to the virus and with microcephaly is different from that of other children, even those exposed to other congenital infections. It is plausible to assume that the norms available in studies validated and used for the purpose of audiological diagnosis in term and preterm neonates do not apply to this population''. ''Our results about brainstem functional normality are challenging, considering the substantial disruption of brain development as well as evidence suggesting the action of ZV on progenitor cells, from the cell proliferation phase. Although the brainstem develops in parallel with the telencephalon in the early stages of embryogenesis, the development of the brainstem does not exhibit the same neuronal migration processes observed in the telencephalon, suggesting that the ZV primarily acts on specific mechanisms of cerebral cortex formation extending from the first to the third trimester.'' ''The functional organization of the brainstem, as observed in this study, indicates an adequate centripetal development process in children with microcephaly, with a neuronal and synaptic organization comparable to typical development, restricting the disorder produced by the ZV to more specific processes of CNS development, probably limited to the telencephalon.'' ''The increase of the deviation of the normality of latencies of wave I with the age at the time of examination suggests that the ZV infection leads to a progressive process in peripheral auditory nerve or sensorineural structures.' ' Mittal et al. (2017); USA 46 ''Many infants with microcephaly exhibit SNHL without clear injury to the inner ear structures, possibly representing auditory impairment at the brainstem or cortical level. Another possibility could be inflammatory changes within the cochlea because the virus may have direct access to these structures via the cochlear aqueduct. But because of the association with microcephaly, some might question whether the HL in infants with ZIKV infection is central in origin due to brain malformation, rather than abnormality at the cochlear level''. ''In summary, ZIKV exposure is associated with HL in infants and adults. Hearing loss can occur as a result of the damage to the inner ear or auditory nerve, by a direct viral effect or mediated by an autoimmune process as demonstrated in the case of other viral infections.''  (Diaph 1 and 3) can cause microcephaly and hearing loss in humans and mice.'' ''Neurotropic viruses can significantly affect NPC (neuroprogenitor cell) functions, including cell fate decisions, proliferation, migration and survival.'' ''Most viruses express gene products that can functionally 'hijack' the Rho-Diaph system, sequestering these proteins for the purpose of modifying the actin cytoskeleton to promote viral entry, assembly and spread.'' ''We hypothesize that viral 'repurposing' of Diaphs disrupts normal host cellular functions, resulting in premature NPC differentiation or apoptosis, as mentioned previously. If viruses sequester Diaphs in NPCs, preventing them from executing their normal cellular functions, it follows that neurological phenotypes associated with these infections would be similar to those caused by genetic mutations of ''In all studies, it seems clear that the auditory impairment is closely related to the presence of microcephaly and its severity.'' ''An issue not yet clarified is the topography of the lesion responsible for hearing loss produced by ZikV infection. None of these children had any anatomical abnormalities of the inner ear on imaging examinations, but it is well known that other congenital infections can cause deafness without any cochlear malformation; therefore, it is not possible to rule out a sensory impairment in these cases.'' ''. . .severe changes in the central nervous system of many of these newborns are a possible origin of the auditory impairment, as already demonstrated in other conditions that involve the auditory pathways.'' ''We are far from understanding if the lesions are produced directly by the virus itself or a result of an inflammatory tissue reaction. To date, no study has been addressed regarding a histologic analysis of the auditory sensory or neural structures.'' regarding the screening or follow-up of individuals with hearing disorders related to ZIKV infection.
Only three case reports were published describing auditory changes related to acquired ZIKV infection in adults, totaling six participants. 5---7 The predominantly sensorineural and transient nature of auditory alterations raises the hypothesis of auditory organ cochlear or neural involvement. An association between flavivirus infection and hearing loss was suggested in two other reports concerning DENV infection cases. However, vascular impairment could not be ruled out in one of the reports, 56 describing an adult with hemorrhagic dengue, who presented deep and bilateral sensorineural hearing loss after 5 days of symptoms. In another report 57 of a child presenting vertical infection by DENV and a clinical condition of hemorrhagic shock who presented alteration during hearing screening, other risk indicators for hearing impairment were also present, such as low birth weight, use of ototoxic medications and permanence in the Neonatal ICU. These findings do not allow us to state that acquired ZIKV or DENV infections cause auditory impairment. However, unlike DENV cases, which comprised other risk factors for hearing loss, in the three ZIKV infection cases, the absence of other events that justify auditory pathway involvement, alongside the strong tropism of the virus by the nervous system, reinforce the hypothesis that damage to these pathways was due to ZIKV infection.
In studies that assessed auditory impairment in prenatal exposure to the ZIKV, it is noteworthy that only about twothirds of all children underwent hearing assessment, with the majority of them being at an age far above one month as recommended by national and international protocols. 48,55 Hearing assessments were carried out predominantly by means of OAE and/or ABR to obtain auditory thresholds. The large variation in the frequency of alterations found during auditory evaluations may be partially explained by the great heterogeneity of the study population, methods of hearing assessment and study designs.
The findings favor the hypothesis of an auditory risk of probable peripheral origin, since the predominant alterations were OAE failure and auditory threshold changes assessed by the ABR. On the other hand, conductive alterations, suggested by children who failed the first OAE screening and presented a normal electrophysiological response in the ABR click assessment, 42 sensorineural alterations, suggested by children who failed both tests, or retrocochlear alterations, in the cases which passed the T-OAE and failed the ABR, 11 have also been described in various studies. Some children presented other risk factors for hearing loss, and very few studies performed a broad auditory evaluation, including FS-ABR, behavioral auditory assessment 8 or tympanometry. 13,44 In addition, losses of a central origin could not be ruled out.
In a small sample of children who presented with late microcephaly 51 and, therefore, less severe neurological damage than microcephaly at birth, the absence of alterations in the auditory evaluations performed by a-ABR and FS-ABR was observed. Similarly, in a transversal study of hearing screening including children congenitally exposed to ZIKV with or without microcephaly, no alterations suggesting sensorineural hearing loss were observed. 13 These results raise suggest a possible relation between the degree of neurological damage and auditory impairment, also emphasized by Leal et al. 52 in a review piece. On the other hand, Marques Abramov et al. 43 concluded, in an evaluation of the   Hearing screening in children exposed to Zika virus during pregnancy. Case series The health services that provide care to this population (children exposed to Zika virus during pregnancy) should make parents or caregivers aware of the need to continue monitoring of hearing development up to 24  These findings allow for anticipation of medical and social service needs of affected children and their families, including early intervention services, and planning for resources to support these families in health care and community settings in Brazil, the United States, and other countries.
Children with disabilities related to congenital Zika virus infection will need multidisciplinary care from various pediatric subspecialists (10). Long-term follow-up and measurement of developmental progression of children affected by Zika virus can inform intervention services and sub-specialties needed to provide optimal care for these children. Otolaryngologists should monitor ZIKV exposed infants without hearing impairment at birth because they may develop HL at later stages of life. The early diagnosis and detection of HL in ZIKV-exposed infants will improve auditory rehabilitation, leading to improved long-term developmental outcomes. In view of the rapid spread of ZIKV in Brazil, it is suggested that patients should monitor their auditory health following ZIKV infection, since even though those patients may not report any alterations in their hearing accuracy; it is possible that the central auditory system could be affected. Zika virus exposure in pregnancy and its association with newborn visual anomalies and hearing loss. Narrative review Hearing examination of infants with suspected CZS infection, even in the absence of microcephaly, is essential, because the associated impairments might be underestimated if microcephaly continues to be the only inclusion criterion during the screening of this group of infants. Children with microcephaly related to CZS need regular follow-ups, even the ones with normal initial screening tests, because hearing loss, like in other congenital viral infections, can be delayed and progressive. conduction of the auditory pathways up to the brainstem in patients with prenatal exposure to ZIKV, that the physiology of auditory pathways of the brainstem is not affected by ZIKV congenital infection, even when it occurs during the first trimester, and that there is no direct correlation between the degree of microcephaly and auditory pathway function in the brainstem. Finally, the need for long-term follow-up remains controversial. The presence of auditory alterations in late evaluations, at 19---24 months of life, in 68.4% of microcephalic children at birth, with laboratory evidence of congenital infection by ZIKV, reinforces the need for late follow-up. These evaluations, however, were restricted to the Hammersmith Infant Neurologic Exam (HINE) ---a response to sound stimulation with a rattle or bell in a small sample of children, where it was not possible to identify the type of auditory alteration, central or peripheral, sensorineural or conductive. 10 On the other hand, in a recent review Leal et al. 52 mentioned a not yet published study that found an incidence of 4.3% of hearing loss in newborns with congenital ZIKV infection and a normal auditory follow-up of those children after 18 months, contradicting the possibility of a progressive or late onset hearing loss. In the only experimental study in no-human mammals, 54 it is noteworthy that in several animals the deficits improved in subsequent measurements, suggesting a transitory hearing loss, as observed in the acquired human infection, 5---7 but also contradicting the hypothesis of a progressive hearing loss. 9,10,43,52 Some of the analyzed studies make suggestions concerning the pathogenesis of auditory involvement by the ZIKV, generally based on what is observed in other viral infections (Table 5). In the animal study previously mentioned, the microscopic analysis detected viral antigens in the cochlea, but no changes in the number of hair cells were observed, leading to the postulation that the hearing loss associated to ZIKV infection does not seem to involve damaged hair cells.
It is not known if the tissue lesion is due to the direct effect of the virus or the host's immune reaction, 7,8 where it may involve only the cochlea or may originate from the central nervous system, especially in cases of CNS malformations 9,46 or if the neuroconduction of the acoustic stimulus is involved, both in prenatal 42 and postnatal exposure, 6 which are issues that require clarification. There are indications of preservation of functional organization in the brainstem of microcephalic children exposed to the ZIKV during pregnancy, suggesting that the disorders caused by the ZIKV are restricted to more central regions. 43 More importantly, there is evidence of a possible progressive damage to the peripheral regions of the auditory nerve or to sensorineural structures, as suggested by Marques Abramov et al. 43 from the finding of increased I wave latency in the ABR with age. This progressive damage could be responsible for late hearing loss.
The hypothesis raised by Racicot et al., 47 suggesting that the ZIKV causes sequestration and redirection of proteins (Diaphanous-related formins ---Diaphs) in progenitor neural cells, determining microcephaly or hearing loss in humans, similar to what is noted in individuals presenting mutations in the genes responsible for the expression of these proteins, opens a promising field to be explored in future research.
Regarding the guidelines for screening and auditory follow-up in individuals exposed to ZIKV, controversies regarding the need to repeat subsequent auditory evaluations in the presence of a normal ABR in the neonatal period were observed herein. Although the authors of the last CDC Interim Guideline 48 justify the lack of need for reevaluation based on the lack of data suggesting late hearing loss in congenital ZIKV infection, the opinion of the authors of this review is that the evidence on the absence of late impairment are still insufficient to support these recommendations. Long-term follow-up studies of children exposed to ZIKV during gestation are necessary for the establishment of evidence-based recommendations.
This review presents some limitations. Most of the studies are retrospective reports or case series, assessing small sample sizes and carrying out cross-sectional auditory evaluations, which makes it difficult to establish causal links and prognostic projections to generate enough evidence to propose well-founded protocols. Moreover, the included studies were very heterogeneous in many aspects, which makes it difficult to compare and synthesize the data collected. Additionally, not all studies focused on hearing assessments, restricted to only neonatal screening and providing minimal details on the performed tests or the obtained results. Many studies limited the hearing evaluation to small sub-samples of patients presenting the more severe viral infection spectrum. Therefore, it was not possible to estimate the overall frequency of altered hearing exams in the subgroup of children with confirmed congenital ZIKV infection, since laboratory confirmation was not specified in the children who performed auditory evaluations. However, three studies that included only children with confirmed ZIKV infection allowed the frequency estimation of auditory alterations in this subset of children. We opted to analyze the frequency of auditory alterations in the microcephaly group, to determine if a worst spectrum of clinical manifestations could be related with an increased frequency of hearing alterations. Unfortunately, the relative frequency of altered a-ABR, which was the most frequently performed test and more specific than de OAE, could not be estimated in children with microcephaly, because of missing data. Finally, the absence of studies evaluating auditory organ histopathology in human ZIKV infection restricts knowledge of the pathogenesis of auditory impairment to hypotheses and theories, limiting the effectiveness of prevention, monitoring and therapeutic management actions. However, in the case of a recent and large epidemic, whose tragic legacy has only recently been recognized, these early studies become a valuable source of data.

Conclusions
Evidence for the involvement of the auditory pathways in congenital or acquired infection by ZIKV is still scarce. The data available to date do not allow the knowledge of the entire spectrum of auditory organ involvement by ZIKV, nor do they confirm the causal association between this involvement and the virus infection, nor they rule out progressive hearing impairment. Especially with regard to individuals with central nervous system malformations, data are still missing to confirm hearing loss of central origin. Multidisciplinary monitoring of all children exposed to ZIKV during pregnancy should improve, with earlier hearing assessment and follow-up. Future research resulting from the longterm follow-up of children presenting the full spectrum of ZIKV involvement, as well as necropsies of stillborn hearing organs and more studies with animal models may provide answers to these yet unanswered questions.

Conflicts of interest
The authors declare no conflicts of interest.