The human immunodeficiency virus (HIV) causes Acquired Immune Deficiency Syndrome (AIDS), the devastating pandemic that continues to affect millions of people worldwide.1,2

HIV infection and AIDS are distinct nosological entities. Many HIV-infected individuals present a normal number of immune cells, remaining asymptomatic for long periods of time, and could not be categorized as presenting the clinical definition of AIDS. To be clinically defined as AIDS, seropositive individuals older than 13 years of age, should present a CD4+ T lymphocyte count below 350 cells per mm3 (Ministério da Saúde, 1999)3 or develop at least one clinical condition that is consistent with AIDS.4

Since the advent of new antiretroviral drugs, there was a consistent shift in the treatment of HIV infection, providing to infected individuals a delay in the disease development and improving their clinical condition, although doubts regarding the toxic action of antiretroviral drugs on both peripheral and central auditory systems.

Until the 1990s, the most common treatment performed in antiretroviral therapy was the monotherapy, that is, the use of only one drug. Due to the evolution of the treatment, the combined therapy (better known as Highly Active Antiretroviral Therapy – HAART)5,6 was established, and showed better prognosis for patients due to the inhibition of HIV replication, thus, it increased survival indexes and, consequently, provided mortality reduction, and is currently used in the treatment of individuals with AIDS.7

AIDS and the therapies developed to combat it have many side effects. Specifically, auditory and vestibular disorders afflict 5–34% of adults with HIV/AIDS.8 Hearing loss affects approximately 20–50% of patients with HIV/AIDS, and 75% of adults with AIDS have some kind of hearing disorder.9

The HIV and infections related to it can affect multiple sites in the hearing system causing observable abnormalities in patients including altered tympanograms, as well as threshold audiograms and auditory brainstem responses. Sensorineural hearing loss associated with HIV/AIDS may result from Central Nervous System (CNS) neoplasms, from ototoxic drug administration,6,9–14 from the effects of HIV on the CNS or on the peripheral auditory nerve, or from opportunistic infections.15 Individuals with HIV/AIDS also often have external otitis and otitis media.16

Despite these factors, no consistent patterns of hearing impairment in adults with HIV/AIDS have been established.17 Thus, the problems that affect the hearing system of this population warrant further research. The purpose of the present study was to compare the audiological and electrophysiological data of HIV-positive patients with and without AIDS, who were receiving HAART, to healthy individuals.

Table 1 shows the distribution of gender and age in CG, RGI and RGII; hearing complaints, CD4+ T lymphocyte counts and time of HIV infection in RGI and RGII groups. There were no statistically significant differences between groups regarding gender (p=0.218), age (p=0.119), hearing complaints (p=0.064), CD4+ T lymphocyte counts (p=0.193) and time of HIV infection (p=0.168).

Fig. 1 shows the hearing complaints in HIV-positive individuals of RGI and RGII. The most frequent complaint was dizziness, followed by tinnitus in RGI. In RGII, the most frequent complaint was hearing loss, followed by tinnitus. No significant differences were observed among RGI and RGII groups (p=0.064); 61% of individuals in the RGI group and 89% of the RGII had at least one complaint of hearing.

Figure 1.

Hearing complaints (in %) in HIV-positive individuals of RGI and RGII (HL, hearing loss; T, tinnitus; D, dizziness; EF, ear fullness; E, earache; A, asymptomatic).

(0.04MB).

Table 2 shows the percentage of hearing loss type found in each group. Sensorineural represented the most common type of hearing loss in the three groups. The RGII had a higher percentage of hearing loss (44%), followed by the RGI with 25%. When RGI and RGII were compared to the CG regarding the risk of hearing loss, RGII had a risk rate and odds ratio significantly greater than RGI (Table 3).

The average hearing thresholds in conventional and high frequency audiometry were lower for CG than observed in the other two groups (Table 4). The differences in hearing thresholds between CG and RGII were statistically significant for all evaluated frequencies. Comparison between CG and RGI showed differences in threshold that were significant and starting at a frequency of 6kHz. In the comparison between RGI and RGII, significant differences were not found only for the frequencies of 8 and 9kHz.

Table 5 compares ABR components in the three groups. The CG displayed the shorter latencies, and significant differences among groups were detected for waves I, III and V and for interpeak interval I–V. Regarding medium-latency components, significant differences in latency and amplitude were found for Na and Pa waves only for the electrode positioned at C3 (Table 6). Finally, Table 6 indicates also that, regarding the P3 wave, CG had shorter latency then RGI and RGII, which were similar to each other.

The frequent occurrence of hearing abnormalities in individuals with HIV/AIDS has been long known,20,21 and many potential targets have been reported, from the middle ear to the central nervous auditory system. However, a discernible pattern has not been found, nor has a primary pathology that may result in hearing impairment.17 Thus, the present study focused on these issues by simultaneously assessing the peripheral and central auditory pathway in individuals with HIV/AIDS, subjected or not to HAART, and in healthy individuals.

Regarding hearing complaints, no significant differences were observed among RGI and RGII groups; 61% of individuals in the RGI group and 89% of the RGII had at least one complaint of hearing. Several studies reported an association between HIV infection and signs/symptoms neurotological. Individuals with AIDS often have hearing complaints, because the ENT manifestations are common at any stage of the disease, leading to specific symptoms such as hearing loss, tinnitus, dizziness and ear fullness,13,14,20–22 which also were observed in the present study.

Sensorineural hearing loss represented the most common type in the three groups, which is in agreement with previous studies.9 The RGII displayed the highest percentage loss (44%), followed by RGI (25%). These findings agree with previous work showing losses between 21% and 49%.23 It is worth mentioning that these losses may vary depending on the criteria adopted for hearing loss and the age range included in each study.9,15,17

We also observed an increased risk for hearing loss in RGII. We found a risk ratio of 6.6 for RGII, almost double the one found for RGI (3.75). The odds ratio for the presence of hearing loss was 11 for RGII and 4.66 for RGI in comparison with the CG. These finding agree with a previous work showing that HIV-positive individuals subjected to HAART had higher odds ratio for hearing loss than healthy individuals. However, these authors did not make comparisons with HIV-positive individuals who did not receive HAART. Our results contrast with those reported previously16 that found a weak positive association between HIV status and poor cochlear function, with an odds ratio slightly higher than one. However, this previous work relied on a different technique (otoacoustic emissions) and did not separate individuals regarding their HAART status.

The CG had lower hearing thresholds than the other two groups, and RGI had lower thresholds than RGII with significant differences in a majority of the frequencies tested. These findings suggest the existence of mechanisms that underlie the worsening of hearing thresholds in the two groups of patients with HIV/AIDS: viral presence in the cochlea previously24; a combination of HIV with opportunistic infections; and/or the effects ototoxic therapeutic agents.2,9,15,17 Specifically in RGII (individuals with AIDS), these mechanisms may act in synergy, whereas the non-drug administration of HAART in the RGI suggests the absence of the effects of ototoxic drugs specific to AIDS. This hypothesis is supported by work using HEI-OC1 cells that evaluated the ototoxic potential of 14 anti-HIV agents, a majority of which was used by RGII patients.6 In this previous study, the authors suggest that many of these drugs, used as antiretroviral agents, may have deleterious effects for the auditory system of patients and further research is under way to validate this idea.

It is worth noting that sensorineural hearing loss more often affects patients with severe HIV infection.22 This fact may also explain why RGI individuals, who were not on HAART and likely had less severe infections than RGII, had less intense hearing problems.

Regarding ABR, the CG had shorter latencies than RGI and RGII, which were similar between them. Other previous studies have compared the ABR of HIV-positive patients and controls.25–31 Results from these reports included latency delays of one or more waves, and increased latencies of one or more interval interpeaks, all of which agree with the present study.

HIV can affect the subcortical and cortical areas of the CNS, and the ABR-generating system depends on the temporal synchronization of neuronal activity. Thus, the central conduction of auditory information in patients with various degrees of HIV infection may be affected, which can generate changes to ABR waves, often causing increased latencies of waves III and V, whereas waves I and II maintain normal absolute latencies.32

Regarding ABR in individuals with HIV/AIDS, these results indicate the usefulness of this procedure for the detection of early signs of neurodegeneration in this population, and in monitoring how fast lesions evolve.25,28 These findings also emphasize the role of immunosuppression in the development of neural abnormalities involving the brainstem auditory pathway during the course of disease.26

Regarding MLR, we observed significant differences in latency and amplitude of Na and Pa waves with the electrode positioned at C3. In two other studies using MLR in HIV/AIDS patients, one found a tendency toward the increase in latency and reduction in amplitude of the PA wave,33 whereas the other found ear and electrode effects happening concomitantly.31 The two studies suggested that auditory information was impaired in cortical and subcortical regions, reinforcing the need for the detailed investigation of auditory function in individuals with HIV/AIDS.

This result may be explained by the fact that the first CNS alteration in this population includes subcortical demyelination, even before clinical neurological manifestations are present.34 Along the same line, other work suggested that HIV affects subcortical and cortical CNS areas, which are crucial for the integrity of MLR wave generation, thus explaining abnormalities.32 Nevertheless, studies that assess the MLR of HIV-positive individuals remain scant, as most research utilizes ABR and P300.

Concerning P300, we found significant differences among the three groups for the P3 wave, with shorter latencies the CG and no differences between RGI and RGII. Previous studies have used long latency auditory evoked potentials to assess the central auditory pathway of individuals with HIV/AIDS.31,33,35–39 All of these studies reported alterations, mainly an increase in latency or reduction of amplitude of the waves N1, P2 and/or P3, in the comparison with healthy subjects. These findings may be associated with impaired cognition, a problem that may be present in AIDS patients.35,37 Other authors added that P300 provides an early indicator of cognitive deficits in seropositive patients, for whom reduced amplitude suggests lower attention, and longer latency a slower processing of information.38,39

The novel approach of the present study was to simultaneously evaluate the peripheral and central auditory pathways of HIV-positive individuals subjected or not to HAART, and to compare them to healthy individuals. Our findings suggest that HIV-positive individuals may develop alterations of the peripheral and central auditory systems that may result from direct viral action, from the presence of opportunistic infections, and/or from the use of ototoxic drugs.

Regarding the number of individuals, it is worth mentioning that some challenges were found to enlarge the sample, especially in relation to the referral of the RGI individuals, once the HIV positive population who do not underwent antiretroviral treatment is scarce. For this reason, it was not possible to achieve the same number of individuals in the RGII.

Specifically concerning RGII, since each patient received at least three medications among those available for HAART, it was not possible to obtain a homogeneous sample as to the type of medication used.

HIV-positive patients with and without AIDS, who were receiving HAART, when compared to healthy individuals, presented:

• -

  Elevated hearing thresholds in behavioral hearing evaluations (Conventional Tonal Audiometry and High Frequency Audiometry), where the group subjected to HAART had the highest thresholds.

• -

  Longer latencies regarding waves I, III and V as well as interpeak I–V for ABR, suggesting lower speeds of neuroelectric impulse transmission throughout the brainstem auditory pathway.

• -

  Longer P300 latency, suggesting reduced processing speed of auditory information in cortical regions.

• -

  A reduction in the amplitude of the Pa wave in MLR in the comparison of RGII with RGI, suggesting alterations to the auditory pathway in cortical and subcortical regions.

The authors declare no conflicts of interest.