Keywords
HIV - hearing loss - audiometry - CD4 T-lymphocyte count
Introduction
The human immunodeficiency virus (HIV), being neurotropic and lymphotropic, may affect
the neuraxis at any level. The virus infects T-helper cells via CD4 T-cell receptors,
causing functional impairment of the T-cells and their depletion.
Hearing impairment has been documented as one of the manifestations in HIV-seropositive
individuals. While reducing morbidity and secondary manifestations are the main objectives
of antiretroviral treatment these days, hearing loss has been overlooked. Although
HIV-seropositive patients with otitis media usually visit or are referred to an ENT
clinic, asymptomatic or mild hearing loss due to other causes remains ignored or undetected.
Otitis media can lead to conductive hearing loss in these patients, whereas sensorineural
hearing loss has been attributed to ototoxicity of antiretroviral drugs or a direct
effect of HIV on the central nervous system, peripheral nerves, and cochleo-vestibular
apparatus.[1]
[2]
[3]
[4]
[5] Numerous studies have evaluated the auditory functions in HIV patients and found
hearing loss associated with progression of the disease.[6]
[7] However, these studies have two major limitations: i) they included subjects with
ear symptoms or previous systemic disease/treatment, which obviously required hearing
evaluation in general; ii) few of these studies performed evoked response audiometry,
which is not a routine clinical practice. Moreover, evoked response audiometry is
an extremely inefficient method of determining hearing thresholds and less precise
than behavioral pure tone audiometry (PTA). When possible (for example, with adults
and older children), behavioral audiometry is always the preferred method of hearing
assessment because it is faster, easier, cheaper, more precise, and widely available.
Therefore, we conducted a prospective, observational study in naïve HIV subjects (those
who have not yet started antiretroviral treatment and on whom audiometry of any kind
has not been performed before enrollment) with routine audiometry to find the effect
of HIV and antiretroviral treatment on auditory functions, so that any hearing impairment
in these patients can be detected even in a general ENT clinic and appropriate interventions
can be initiated timely to improve the quality of life.
Methods
After institutional ethical board approval, 25 naïve HIV-seropositive subjects (16
female and 9 male) with a mean age of 37.64 ± 6.55 years were enrolled after adjustment
of confounding variables and written informed consent. Subjects with abnormal (healed,
thin, retracted, bulging, perforated) tympanic membrane, tinnitus, vertigo, history
of systemic disease, opportunistic infection or treatment for it, diseases or any
other known causes of hearing loss and conductive hearing loss on PTA were excluded.
All subjects were placed in the same environmental conditions during the study period,
and any exposure to noise, otitis media, or ototoxic drugs needed to be disclosed
and would preclude the subject from participating in the study.
The subjects underwent detailed history, clinical, audiological, hematological and
biochemical evaluation at the time of enrollment and after 6 months of highly active
antiretroviral therapy (HAART: Efavirenz, Lamivudine, and Tenofovir Disoproxil Fumarate).
The subjects were evaluated for complete blood count, CD4 counts, lipid profile, liver,
and renal functions to exclude metabolic or endocrine disorders. The PTA was performed
in a certified sound-treated room with the clinical audiometer AC 40 (Interacoustics
A/S, Assens, Denmark) between 250–8,000 Hz for air conduction using circumaural headphones
and up to 4,000 Hz for bone conduction. This audiometer, calibrated annually, had
been calibrated 8 months before the first enrollment in the study. Subjects with conductive
hearing loss were excluded from the study as we believed that the disease, or the
medication used to treat it, could affect the outcome of our study. Subjects showing
sensorineural hearing loss on speech frequencies were then evaluated for tone decay
test (TDT) using the Olsen and Noffsinger method and short increment sensitivity index
(SISI) in similar test conditions. The SISI test was performed using the Jerger method,
which involves presentation of 20 dB suprathreshold tone in the range of 1,000–4,000 Hz
frequencies. The ear being tested was delivered a 5 dB increment in the tone every
5 seconds, and the test was repeated until the subject could identify the change in
loudness. After identification, the tone increment was reduced sequentially. Then
finally, the ear was delivered a 1 dB increment in the tone every 5 seconds for 20
times. The result is denoted as a percentage of the number of correct responses out
of 20. Observations were analyzed using non-parametric tests (Wilcoxon signed-rank
test and Spearman rank correlation coefficient).
Observation and Results
At the time of enrollment of the subjects, their mean hearing thresholds for speech
frequencies (500 Hz, 1 KHZ and 2 KHZ) were 24.6 ± 16.8 dB and 24.06 ± 16.8 dB in the
right and the left ear, respectively. These thresholds remained almost at the same
levels (p > 0.05) on re-evaluation after 6 months (24.46 ± 16.8 dB and 24.06 ± 16.45 dB in
the right and the left ear, respectively). However, all subjects had increased hearing
thresholds at high frequencies (4 KHz and 8 KHz) in both ears at the time of enrollment
that persisted at the same level (p > 0.05) at the follow-up after 6 months ([Table 1]). None of the subjects displayed any other otological symptoms during the observation
period of 6 months.
Table 1
Hearing thresholds and Wilcoxon Signed-rank Test results
|
Frequency
(In Hz)
|
Hearing Thresholds (Mean ± SD in dB) and statistical significance
|
|
Right ear
|
Left ear
|
|
At 0 months
|
At 6 months
|
P-value
(2-tailed)
|
At 0 months
|
At 6 months
|
P-value
(2-tailed)
|
|
250
|
23.2 ± 15.19
|
23.2 ± 15.8
|
1.000
|
23.2 ± 17.31
|
23.6 ± 16.68
|
0.527
|
|
500
|
22.2 ± 16.39
|
23.8 ± 16.79
|
0.074
|
24.4 ± 17.99
|
24.4 ± 17.75
|
1.000
|
|
1,000
|
25.6 ± 18.56
|
25.2 ± 17.94
|
0.669
|
24.4 ± 18.1
|
23.2 ± 15.54
|
0.242
|
|
2,000
|
26 ± 16.39
|
24.4 ± 16.54
|
0.059
|
23.4 ± 15.39
|
24.6 ± 16.89
|
0.279
|
|
4,000
|
30.6 ± 15.23
|
30.2 ± 16.23
|
0.717
|
29.6 ± 15.47
|
29.6 ± 15.47
|
0.980
|
|
8,000
|
35 ± 18.37
|
34.8 ± 18.62
|
0.837
|
32 ± 16.39
|
32 ± 16.46
|
0.850
|
Seven subjects had sensorineural hearing loss in one or both ears with hearing thresholds
of 49.99 ± 24.92 dB (range 26.66–90 dB) and 50.33 ± 23.78 dB (range 30–88.33 dB) in
the right ear at 0 and 6 months, respectively, while the thresholds in the left ear
were 46.38 ± 23.00 dB (range 30–88.33 dB) and 45.83 ± 22.18 dB (range 30–85 dB) for
the same periods of time. These observations did not show any significant difference
on the Wilcoxon signed-rank test. These seven patients were further subjected to the
SISI and TDT tests to determine the type of sensorineural loss. They did not show
tone decay at 0 or 6 months after TDT, while the SISI (mean ± SD) results were 79.29 ± 5.35%
and 81.43 ± 2.44% at 0 and 6 months, respectively, indicating sensory (cochlear) hearing
loss.
The mean CD4 T-lymphocyte counts were 347.08 ± 245.9 cells/µL and 370.6 ± 203.13 cells/µL
at 0 and 6 months, respectively, that had no significance on the Wilcoxon signed-rank
test (p > 0.05). The Spearman correlation coefficient did not find a significant correlation
(p > 0.05) between CD4 T-lymphocyte counts and PTA during the study ([Tables 2] and [3]).
Table 2
Spearman correlation between hearing thresholds and CD4 at the time of enrollment
|
Frequency
(in Hz)
|
Right ear
|
Left Ear
|
|
Correlation Coefficient
|
P-value
(2-tailed)
|
Correlation Coefficient
|
P-value
(2-tailed)
|
|
250
|
–0.005
|
0.982
|
0.045
|
0.830
|
|
500
|
0.011
|
0.960
|
–0.012
|
0.954
|
|
1000
|
0.010
|
0.961
|
0.028
|
0.894
|
|
2000
|
–0.120
|
0.567
|
0.043
|
0.838
|
|
4000
|
–0.294
|
0.154
|
–0.142
|
0.499
|
|
8000
|
–0.389
|
0.055
|
–0.139
|
0.507
|
Table 3
Spearman correlation between hearing thresholds and CD4 at 6 months
|
Frequency
(in Hz)
|
Right ear
|
Left Ear
|
|
Correlation Coefficient
|
P-value
(2-tailed)
|
Correlation Coefficient
|
P-value
(2-tailed)
|
|
250
|
0.076
|
0.717
|
0.087
|
0.679
|
|
500
|
0.247
|
0.234
|
–0.062
|
0.767
|
|
1,000
|
0.001
|
0.996
|
0.040
|
0.849
|
|
2,000
|
0.008
|
0.969
|
–0.058
|
0.784
|
|
4,000
|
–0.194
|
0.354
|
–0.004
|
0.984
|
|
8,000
|
–0.297
|
0.149
|
–0.025
|
0.906
|
One subject had a marked decrease in her CD4 count (538 cells/µL to 82 cells/µL) at
follow-up but her hearing threshold remained normal for each frequency. For the subjects
with hearing loss at speech frequencies 500 HZ, 1 KHz and 2 KHz, the mean CD4 T-lymphocyte
count was 341.2 cells/µL in five patients with mild sensorineural hearing loss, 29
cells/µL in a patient with moderate–severe sensorineural hearing loss, and 26 cells/µL
in a patient with severe hearing loss that remained so despite improvement in the
CD4 count during the follow-up.
Discussion
The HIV causes profound immunological impairment by invading the cells related to
the immune system, which leads to reduction in lymphocytes and to opportunistic infections.
The T-helper lymphocyte via CD4 receptor is primarily infected, which leads to its
functional impairment and depletion.[1]
The prevalence of hearing impairment in HIV-seropositive patients is 14–49%.[6]
[8]
[9] This hearing loss can be due to secondary opportunistic infections and ototoxic
drugs or can be a direct effect of the virus on the central nervous system, peripheral
nerves, vestibulocochlear nerve, or cochlea. Secondary infection leading to otitis
media is the most common presentation in HIV patients and can lead to conductive,
sensorineural, or mixed hearing loss. However, systemic infections in the form of
cytomegalovirus, mumps, measles, Cryptococcus, tuberculosis, or ototoxicity of drugs
that are used to treat these infections also contribute prominently in sensorineural
hearing loss.[5]
[10]
The direct effect of HIV on the cochlea has been established in a study by Pappas
et al,[2] who identified extracellular viral-like particles in the tectorial membrane of three
HIV positive patients with sensorineural hearing loss. They also identified pathologic
changes in the labyrinthine wall, epithelial lining of macula and crista, and inclusion
bodies in the supporting cells. Torre et al[10] performed low- and high-frequency PTA assessment of HIV-positive subjects on HAART
and age-matched HIV-negative controls. They found significantly higher hearing thresholds
at both low (250–2,000 Hz) and high frequencies (3–8 KHz) among the HIV-seropositive
individuals in comparison to the HIV-negative controls. There was no correlation between
CD4 T-lymphocyte count and hearing thresholds in HIV seropositive individuals. They
concluded that low-frequency hearing loss can impair communication in HIV patients.
We also found increased hearing thresholds at 4–8 KHz in all subjects with no significant
changes after 6 months. Seven subjects with sensorineural hearing loss also did not
show any significant difference after 6 months. The TDT and SISI tests suggested cochlear
pathology in these seven subjects. We also did not find any statistically significant
correlation between the overall CD4 counts and sensorineural hearing loss. However,
individually, one subject had moderate–severe hearing loss (CD4 count of 29 cells/µL)
and another had severe sensorineural hearing loss (CD4 count of 26 cells/µL). These
finding may suggest a relationship between the severity of the sensorineural hearing
loss and CD4 T-lymphocyte counts. However, contrary to this assumption, another subject
had a marked decrease in CD4 count (from 538 cells/µL on enrollment to 82 cells/µL
after 6 months) on the follow-up; yet, her hearing thresholds remained normal at each
frequency. Palacios et al[11] have found abnormal audiologic observations in subjects with prolonged HIV infection,
viral load, and low CD4 counts.
Assuiti et al[12] have conducted a review and did not find any direct association between hearing
loss and antiretroviral drugs. They concluded that hearing loss in HIV-positive/seropositive
patients is multifactorial and can be attributed to the direct effect of HIV, opportunistic
infection, and ototoxicity due to HAART or drugs used to treat opportunistic infections.
We found that hearing impairment had no relation with HAART, duration of treatment,
or severity of the disease, indicating that cochlear functions are impaired by HIV
either directly or indirectly through premature aging.
This study indicated cochlear damage, which could have been caused due to some unknown
etiology (secondary infections or drugs used to treat them) that subjects failed to
recall. Evoked response audiometry may be taken as a limitation of this study, but
we deliberately excluded the use of evoked response audiometry for evaluation as we
wanted to observe changes through PTA, which is quick, inexpensive, and widely available
in comparison with evoked response audiometry and can enable even a general ENT clinic
to find subtle changes in the hearing thresholds, so that timely and appropriate interventions
can be initiated.
Conclusion
Auditory functions are impaired in HIV infection, but its assessment remains within
the research area only. We conducted a study to assess auditory functions in HIV-infected
patients with routine PTA and we found hearing loss at high frequencies in all HIV-positive/seropositive
subjects. We advocate the use of PTA in all HIV-positive/seropositive patients so
that adequate intervention can be initiated as early as possible to avoid morbidity.
