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 acquired5–7 and congenital8–16 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 HIV25,28–30 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.

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.

The search strategy initially identified 157 records, which, at the end of the selection process, resulted in the inclusion of 27 articles5–7,8–16,41–55 (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 impairment6–9,42,43,46,47,52 and 17 studies included recommendations on hearing screening and follow up for patients with congenital or acquired ZIKV infection6–11,13,14,16,42,44,46,48,49,52,53,55 (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.

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).

(0.31MB).

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 cases5 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 confirmed ZIKV infection during pregnancy (n=416)12,15,16; (3) Fetus or newborns or children with suspected CZS based on epidemiologic and clinical grounds (n=444)11,13,14,41–45,50 (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).

Among the three studies that included only children with laboratory confirmation of congenital ZIKV infection (n=102),9,10,51 18 had hearing alterations (17,6%): five in the ABR and 13 in the HINE (Table 4).

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 (CDC, Atlanta, GA, USA), with updates for the evaluation and management of children presenting possible congenital ZIKV infection48,49 and a Manual by the Brazilian Ministry of Health (Protocol on Health Care and Response to the Occurrence of Microcephaly Related to Zika Virus Infection)55 (Table 7). In the last CDC Interim Guideline update,48 a change from the previous Interim Guideline49 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 piece46 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 loss9,10,42,46 (Table 7).

In two studies reporting cases of auditory impairment in ZIKV-acquired infections, the authors suggest auditory monitoring in these cases6 and the need for further investigations to promote this recommendation during ZIKV outbreaks7 (Table 7).

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 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 report57 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 two-thirds 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 assessment8 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 microcephaly51 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 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 malformations9,46 or if the neuroconduction of the acoustic stimulus is involved, both in prenatal42 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 Guideline48 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.

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 long-term 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.

The authors declare no conflicts of interest.