Introduction
The vestibular system senses movement, and sends this information to the cerebellum and vestibular nuclei in the brainstem. Motion and other sensory information are processed and integrated: gaze is stabilised during head movement by means of the vestibulo-ocular reflex, and body and head stability is maintained via the vestibulospinal reflex and vestibulocollic reflex respectively.Reference Goebel and Goebel1, Reference Fife, Eggers and Zee2 A pathway that includes the saccule, inferior vestibular nerve and vestibulospinal tract mediates and actives the vestibulocollic reflex.Reference Ito, Ishimoto and Murofushi3, Reference Wilson, Boyle, Fukushima, Rose, Shinoda and Suqiuchi4 Cervical or collic vestibular-evoked myogenic potentials are a manifestation arising from the vestibulocollic reflex of the vestibulospinal tractReference Rosengren, Welgampola and Colebatch5 which is mediated ipsilaterally by a three neuron arc. Uchino and colleagues described the anatomical pathway of the vestibulocollic reflex, also known as the sacculocollic reflex.Reference Uchino, Sato, Sasaki, Imaqawa, Ikeqami and Isu6 Primary vestibular neurons that travel from the saccule via the inferior vestibular nerve, project into second order vestibular neurons in the lateral vestibular nucleus in the brainstem. From there, neurons descend in the medial vestibulospinal tracts and connect to the motor nuclei of the accessory nerve. Third order vestibular neurons descend to the flexor and extensor neck muscles via the medial vestibulospinal tract.Reference Uchino, Sato, Sasaki, Imaqawa, Ikeqami and Isu6
Cervical vestibular-evoked myogenic potentials are ipsilaterally evoked short latency responses, which can be measured with an active electrode over a contracted sternocleidomastoid muscle,Reference Colebatch, Halmagyi and Skuse7 capable of evoking the vestibulocollic reflex.Reference Colebatch and Rothwell8 This is perhaps the most direct way of testing vestibulocollic reflex functioning.Reference Welgampola and Colebatch9 Cervical vestibular-evoked myogenic potential abnormalities may indicate a lesion at any point along the vestibulocollic reflex pathway. Testing of the vestibulo-ocular reflex pathways, which include the horizontal semicircular canal, superior vestibular nerve and ascending neural path to the extra-ocular muscles,Reference Wester, Atcherson and Stoody10 is well characterised and the basis of many commonly used vestibular tests.Reference Fife, Eggers and Zee2 These include, but are not limited to the caloric testReference Baloh and Kerber11 and rotational tests.Reference Brandt and Strupp12 Testing of the vestibulospinal reflex pathways may include posturography.Reference Baloh and Kerber11
The vestibulo-ocular reflex and vestibulospinal reflex pathways have been examined and described in individuals with human immunodeficiency virus (HIV) and acquired immunodeficiency syndrome (AIDS). A recent systematic literature review summarised all vestibular tests and findings in subjects with HIV and/or AIDS, and demonstrated that tests of vestibular function concentrated on the vestibulo-ocular reflex and vestibulospinal reflex pathways only.Reference Heinze, Swanepoel and Hofmeyr13 To date, no studies have investigated the vestibulocollic reflex pathways, as examined using cervical vestibular-evoked myogenic potential testing, in subjects with HIV and/or AIDS. This may in part be attributable to that fact that cervical vestibular-evoked myogenic potential testing has only recently been included in clinical test batteries of vestibular function, having proved useful in identifying vestibular disorders.Reference Welgampola and Colebatch9
This study aimed to: (1) describe and compare the functioning of the vestibulocollic reflex and the well characterised vestibulo-ocular reflex in subjects with and without HIV; (2) describe the vestibulocollic reflex and vestibulo-ocular reflex throughout progression of the disease; and (3) compare the vestibulocollic reflex and vestibulo-ocular reflex in HIV-positive subjects receiving antiretroviral therapies with those who were not receiving antiretroviral therapies.
Materials and methods
The Research and Ethics Review Committee of the University of Pretoria and a tertiary referral hospital approved the current study. A cross-sectional comparative research design was employed and convenience sampling was used to recruit subjects. Each subject provided written informed consent to participate in the study.
Subjects
The experimental group comprised subjects with HIV, who were drafted from the infectious disease clinic of a tertiary referral hospital in South Africa. The subjects' HIV status was confirmed by blood serological testing. The researchers obtained written informed consent to access these individuals' medical records, which contained this data. The control group consisted of subjects confirmed to be HIV-negative. These were employees of the tertiary referral hospital and acquaintances of the researchers, all of whom agreed to undergo a blood serological test for HIV.
A total of 91 subjects, 53 adults with HIV and 38 without HIV, were evaluated for participation in the study. Table I summarises the characteristics of participating subjects. There were no statistically significant differences in mean ages between the groups (p = 0.26; t-test). Previous research indicates that age affects the vestibular system after 55 to 65 years;Reference Maes, Dhooge, D'haenens, Bockstael, Keppler and Philips14 in order to minimise the likelihood of age affecting the results, only subjects below the age of 50 years were allowed to participate in the study.
Table I Description of subjects
HIV = human immunodeficiency virus; SD = standard deviation; y = years; M:F = male to female ratio; CDC = Centers for Disease Control and Prevention Classification System for HIV Infection; n/a = not applicable; ARV = antiretroviral
The subjects with HIV were divided into categories according to their cluster of differentiation 4 cell count, as documented in their medical files (cell count was determined at the infectious disease clinic, at the start of the study). Subjects with counts higher than 500 cells/μl were assigned to category 1 of the Centers for Disease Control and Prevention Classification System for HIV Infection, while those with counts of 200–499 cells/μl and less than 200 cells/μl were assigned to categories 2 and 3 respectively.15 The HIV-positive subjects were evenly distributed between the three categories. Fifteen subjects were in category 1 (8 male, 7 female), 20 were in category 2 (8 male, 12 female) and 18 were in category 3 (8 male, 10 female).
The HIV subjects were further divided into two subgroups, one comprising those who were receiving antiretroviral therapies (n = 42) and another comprising those not receiving antiretroviral therapies (n = 11). Those undergoing antiretroviral therapy received a combination of at least three of the following drugs: tenofovir, lamivudine, efavirenz, emtricitabine, nevirapine, stavudine, zidovudine and lopinavir/ritonavir.
Otological and audiological examinations
An otoscopic examination was performed to inspect the external auditory canal for any debris or foreign objects that might cause occlusion of the ear canal, and to identify any possible perforation of the tympanic membrane. Subjects with obstructed ear canals were referred to a clinician for extraction prior to participation in the study.
Tympanometry was performed using a diagnostic 226 Hz probe tone (GSI TympStar™). The following criteria were used for normal adult acoustic admittance profiling: ear canal volume, 0.8 to 2.0 ml; compliance, 0.3 to 1.8 ml; and middle-ear pressure, −100 to +50 daPa.Reference Jerger16 Type A tympanograms were revealed for 49 subjects with HIV and for 38 subjects without HIV.
Pure tone audiometry (air and bone conduction) was performed to determine the presence of air–bone gaps (GSI 61™). Air–bone gaps larger than 10 dB were found in the four HIV-positive subjects with abnormal compliance and middle-ear pressure, suggesting a conductive component. No air–bone gaps were found for the 49 HIV-positive and 38 HIV-negative subjects with type A tympanograms.
Vestibulo-ocular reflex test
Caloric tests were used to describe the vestibulo-ocular reflex. Rotational testing was not available to the researchers. Tests of spontaneous nystagmus, with and without fixation, preceded caloric stimulation. Video nystagmography (performed with a Visual Eyes infrared video-based system from Micromedical Technologies, Chatham, Illinois, USA) was used to record any spontaneous nystagmus and caloric-induced nystagmus.
Bithermal (cool 24°C, warm 47°C) air caloric testing (AirFX; Micromedical Technologies) was used to irrigate the external auditory canal. Air caloric irrigation was chosen over water irrigation because water irrigation of the external ear canal may result in damage to the delicate skin lining of the outer ear, which in turn places it at risk of invasive external otitis due to bacterial invasion.Reference Zikk, Rapoport and Himelfarb17 This frequently occurs in those who are immunocompromised, such as persons with HIV. Subjects were placed in a supine position with the head tilted forward at an angle of 30° from the horizontal plane for correct positioning of the horizontal semicircular canals. Air was irrigated for 60 seconds, with 5-minute intervals between stimuli.
The peak of the slow phase eye velocity of caloric nystagmus post-irrigation was used as a parameter of superior vestibular nerve and horizontal canal function. The formula reported by Jongkees et al. was used to calculate unilateral weakness or asymmetry and directional preponderance.Reference Jongkees, Maas and Philipszoon18 A unilateral weakness or asymmetry of 20 per cent or more,Reference Baloh and Kerber11, Reference Jacobson, Newman, Peterson, Jacobson, Newman and Kartush19 directional preponderance of 30 per cent or more,Reference Baloh and Kerber11 bilateral weakness, or hyperreflexia was considered abnormal. Bilateral weakness or hypoactive responses were regarded as the total warm responses from both sides of less than 11 degrees per second, and the total cool responses from both sides of less than 6 degrees per second.Reference Barber and Stockwell20 Hyperreflexia or hyperactive responses were regarded as the total responses from both sides of more than 221 degrees per second.Reference Jacobson, Newman, Peterson, Jacobson, Newman and Kartush19
Vestibulocollic reflex test
Cervical vestibular-evoked myogenic potential testing was performed using an auditory-evoked potential system (Bio-logic Navigator Pro; Natus Medical, San Carlos, California, USA). Insert-type earphones (ER-3; Etymotic Research, Elk Grove Village, Illinois, USA) with disposable foam tips were used.
Subjects were comfortably seated with their head rotated approximately 45° to the opposite side of the ear being tested. A blood pressure manometer, with the rolled up inflatable cuff positioned between the subject's hand and jaw, was used to provide feedback of the contracted sternocleidomastoid muscle during the recording of the cervical vestibular-evoked myogenic potentials. The subjects pushed with their heads against the rolled up inflatable cuff and were asked to sustain a pressure of 40 mmHg. This allowed control of the sternocleidomastoid muscle contractions and ensured comparable muscle contractions between the left and right side.Reference Maes, Vinck, De Vel, D'Haenens, Bockstael and Keppler21 Both the subjects and the investigator monitored this sustained pressure.
Every measurement, even those indicating absent responses, was repeated twice to test for wave reproducibility and to eliminate potential artefacts. The average of the two recordings was used for analysis. The first peak on the waveforms was marked as P1, while the second (occurring within a period of 30 ms) was marked as N1.
The researchers recorded and measured P1 and N1 latencies (in milliseconds), inter-peak amplitude (in microvolts), and amplitude asymmetry (in percentages). The asymmetry ratio was determined by calculating the interaural amplitude difference according to the following formula: ((left ear amplitude – right ear amplitude)/(left ear amplitude + right ear amplitude)) × 100. Responses were interpreted as follows: (1) the absence of unilateral or bilateral waveforms was considered abnormal (absence of an identifiable P1 and N1 peak); (2) latencies above the upper limits for P1 and N1 latencies (17.0 ms and 26.3 ms respectively – calculated from two standard deviations above the mean of the HIV-negative group) were regarded as present yet delayed, and considered abnormal; and (3) the presence of an amplitude asymmetry ratio of 40 per cent or more was considered abnormal, as it indicated side-to-side differences in amplitude.Reference Akin, Murnane, Jacobson and Shepard22
Results
All analyses of data were performed using the statistical software package SPSS® for Windows, version 21. Means, standard deviations and percentages were used to describe the data. A one-way analysis of variance was used to compare the distribution of HIV-positive subjects between the three categories of the Classification System for HIV Infection.15 The one-sample Kolmogorov–Smirnov test was used to demonstrate normality of the data. The independent samples t-test was used to compare the mean values between the experimental and control groups. P-values of less than 0.05 were accepted as statistically significant. Odds ratios were calculated. The chi-square non-parametric test was used to compare the findings between the two study groups and the three Classification System categories.
Abnormalities
Abnormal cervical vestibular-evoked myogenic potential and caloric test results were found in 66 per cent (n = 35) of the subjects with HIV, compared with only 15.8 per cent (n = 6) of the subjects without HIV, indicating a significantly higher occurrence of pathology in subjects with HIV (p = 0.001; chi-square). The recordings indicated absent cervical vestibular-evoked myogenic potentials for the four HIV-positive subjects with abnormal tympanograms and air–bone gaps. This association between vestibular signs and HIV was further confirmed by the odds ratio, calculated as 10.2. This indicated a 10.2 times higher risk of abnormal cervical vestibular-evoked myogenic potential and caloric responses in persons who were HIV-positive.
Table II shows the distribution of abnormal cervical vestibular-evoked myogenic potential and caloric test results in the HIV-positive and HIV-negative groups. Abnormal cervical vestibular-evoked myogenic potential results were found in 43.4 per cent (n = 23) and abnormal caloric results in 35.8 per cent (n = 19) of subjects with HIV. Abnormal cervical vestibular-evoked myogenic potential results due to middle-ear pathology were found in 7.5 per cent (n = 4) of subjects with HIV. The cervical vestibular-evoked myogenic potential results were abnormal in 10.5 per cent (n = 4) and the caloric results were abnormal in 10.5 per cent (n = 4) of the subjects without HIV.
Table II Distribution of abnormal findings by group
*As demonstrated by abnormal middle-ear compliance and pressure, and air–bone gaps. †p = 0.001 (between-group difference). HIV = human immunodeficiency virus; VEMP = vestibular-evoked myogenic potential
Table III shows the occurrence of abnormal cervical vestibular-evoked myogenic potential results according to absent waveforms, delayed P1 and/or N1 latencies, and amplitude asymmetry of 40 per cent or more. In the HIV-positive group, the recordings for 20.8 per cent (n = 11) of subjects indicated absent cervical vestibular-evoked myogenic potentials, not including the four subjects with middle-ear pathology. Of the six subjects with unilaterally absent cervical vestibular-evoked myogenic potentials, the left side was affected in five subjects and the right side affected in one subject. Only one subject in the HIV-negative group had absent cervical vestibular-evoked myogenic potentials bilaterally. There was a significantly higher occurrence of absent cervical vestibular-evoked myogenic potentials among the subjects with HIV than for subjects without HIV (p = 0.003; chi-square). Table III further indicates that in the HIV-positive group 17 per cent (n = 9) of subjects presented with delayed latencies. Four subjects showed delayed latencies unilaterally and five bilaterally.
Table III Description of abnormal findings
*Data of the four subjects with middle-ear pathology excluded here. HIV = human immunodeficiency virus; VEMP = vestibular-evoked myogenic potential
There were no significant differences observed regarding the mean latencies of P1 and N1 to the left or right side in either of the study groups (p > 0.05; t-test). Table IV indicates the distribution of the mean P1 and N1 latencies, as well as the inter-peak amplitude differences in both the HIV-positive and HIV-negative groups as revealed by cervical vestibular-evoked myogenic potential testing. P1 latencies were statistically (but not clinically) significantly delayed in the HIV-positive group. N1 latencies showed no difference between the two groups.
Table IV Mean latency and inter-peak amplitude results*
* From cervical vestibular-evoked myogenic potential recordings. †Data for ears with absent cervical vestibular-evoked myogenic potentials not used. HIV = human immunodeficiency virus; SD = standard deviation
Disease progression
Figure 1 illustrates the cervical vestibular-evoked myogenic potential and caloric test results throughout the progression of the disease. The HIV-positive subjects were divided into the three categories of the Classification System for HIV Infection based upon their cluster of differentiation 4 cell counts at the time of participation.15 Normal cervical vestibular-evoked myogenic potential and caloric test results were recorded in 34 per cent of subjects with HIV. The occurrence of abnormal test results increased from 13.3 per cent for those in category 1, to 22.6 per cent in category 2 and 30.1 per cent in category 3 of disease progression (Figure 1).
Fig. 1 Cervical vestibular-evoked myogenic potential (cVEMP) and caloric test results throughout disease progression. CDC = Centers for Disease Control and Prevention Classification System for HIV Infection; HIV = human immunodeficiency virus
Antiretroviral therapy
Table V shows the distribution of abnormal cervical vestibular-evoked myogenic potential and caloric results of the HIV-positive subjects receiving antiretroviral therapies compared with those not receiving antiretroviral therapies. Of the 42 antiretroviral therapy users, 66.7 per cent (n = 28) showed abnormal cervical vestibular-evoked myogenic potential and caloric test results. Of the 11 non-antiretroviral therapy users, 63.6 per cent (n = 7) showed abnormal cervical vestibular-evoked myogenic potential and caloric test results. Although the occurrence of abnormal test results was higher among the antiretroviral therapy users, this difference was not statistically significant (p = 0.91; chi-square).
Table V Distribution of abnormal findings for HIV-positive antiretroviral therapy users versus non-users
*n = 42; †n = 11. ‡p = 0.91 (between-group difference). HIV = human immunodeficiency virus; ARV = antiretroviral; VEMP = vestibular-evoked myogenic potential
Discussion
In this study, subjects with HIV presented with significantly more abnormal cervical vestibular-evoked myogenic potential and caloric test findings than those without HIV. Two-thirds of the subjects with HIV (66 per cent) presented with at least one abnormality on the cervical vestibular-evoked myogenic potential and caloric test parameters, compared with only 15.8 per cent of subjects without HIV. The calculated odds ratio suggested that subjects with HIV had a 10.2 higher risk of having abnormal cervical vestibular-evoked myogenic potential and caloric test results than subjects without HIV.
Regarding the reliable recording of cervical vestibular-evoked myogenic potential responses, the use of a correction algorithm remains the favoured method, as well as monitoring the electromyographic (EMG) activity of the sternocleidomastoid muscle.Reference Maes, Vinck, De Vel, D'Haenens, Bockstael and Keppler21 The current study did not have access to these algorithms or EMG systems, which is a limitation of the study. However, the results of a recent study concluded that the use of the blood pressure manometer may be a useful alternative for reliable recordings of these responses.Reference Maes, Vinck, De Vel, D'Haenens, Bockstael and Keppler21
The caloric findings in the current study can be compared with those of four previous research reports that employed a group study design of adult subjects with HIV.Reference Castello, Baroni and Pallestrini23–Reference Teggi, Ceserani, Luce, Lazzarin and Bussi26 Abnormal caloric test findings demonstrate the presence of abnormal functioning of the vestibulo-ocular reflex pathway. Three of the four studies showed abnormal caloric test findings in subjects with HIV that were similar to those in the current study (in which 35.8 per cent of HIV subjects (n = 19) were affected). The two most recent studies found abnormal caloric test results in 43.3 and 50 per cent respectively in their sample of HIV subjects.Reference Castello, Baroni and Pallestrini23, Reference Hausler, Vibert, Koralnik and Hirschel24 Their subjects were all symptomatic as they suffered from chronic dizziness. An earlier study found slightly fewer cases of abnormal caloric test results compared with the current study; specifically, 11.6 per cent of their sample of HIV subjects were affected.Reference Teggi, Ceserani, Luce, Lazzarin and Bussi26 In contrast, only one study reported no abnormal caloric test findings in HIV subjects.Reference Teggi, Giordano, Pistorio and Bussi25 No noticeable differences were found between the caloric test protocols employed in these reports, but the subjects in the study by Castello and colleagues were all asymptomatic without any vestibular symptoms (e.g. vertigo, dizziness or disequilibrium),Reference Teggi, Giordano, Pistorio and Bussi25 which could explain the absence of abnormal caloric responses. It is interesting to note that 66.4 per cent (n = 29) of HIV-positive subjects in the study by Hausler and colleagues were asymptomatic, yet the authors reported abnormal caloric test findings among those subjects too.Reference Teggi, Ceserani, Luce, Lazzarin and Bussi26 However, these authors did not indicate whether their subjects used antiretroviral therapies or any medication for secondary infections that could contribute to the abnormal caloric test findings.
To our knowledge, no other published research has utilised cervical vestibular-evoked myogenic potential testing to describe the functioning of the vestibulocollic reflex pathways in subjects with HIV. The current study demonstrated a significantly higher number of abnormal cervical vestibular-evoked myogenic potential results in HIV subjects than in non-HIV subjects. The mean P1 latencies were statistically significantly delayed in the HIV-positive group. The abnormal cervical vestibular-evoked myogenic potential test results suggest a high occurrence of abnormal functioning of the vestibulocollic reflex pathways in subjects with HIV. The abnormal caloric test results also suggest a higher occurrence of abnormal functioning of the vestibulo-ocular reflex pathways in subjects with HIV compared with those without the disease. The caloric test is a very low frequency test, namely 0.003 Hz;Reference White, Hughes and Pensak27 the mid-range and higher frequencies were therefore not assessed, meaning that the occurrence of vestibulo-ocular reflex abnormality as assessed with the caloric test may be even higher.
Possible mechanisms for the increased vestibulocollic reflex and vestibulo-ocular reflex abnormalities include opportunistic infections, ototoxic treatments and direct effects of HIV.Reference Heinze, Swanepoel and Hofmeyr13 Opportunistic infections may affect the functioning and integrity of both branches of the vestibular nerve (and the structures that they innervate). Young reported five patients with the herpes zoster virus, a common HIV-related opportunistic infection, suffering from vertigo.Reference Young28 Unfortunately, the author did not indicate in the report whether these subjects were HIV-positive. Nonetheless, recordings revealed absent cervical vestibular-evoked myogenic potentials for all five subjects (100 per cent). In addition, recordings revealed absent caloric responses for four subjects (80 per cent). In another study, 10 subjects with Ramsay Hunt syndrome, another common HIV-related opportunistic infection, underwent cervical vestibular-evoked myogenic potential and caloric testing.Reference Ozeki, Iwasaki, Ushio, Takeuchi and Murofushi29 Once again, the authors did not indicate the subjects' HIV status. Similar to Young,Reference Young28 the study revealed abnormal cervical vestibular-evoked myogenic potentials in 7 subjects (70 per cent) and abnormal caloric results in all 10 subjects (100 per cent). Another study of subjects infected with cytomegalovirus demonstrated abnormal cervical vestibular-evoked myogenic potential and caloric test results (23.1 per cent and 30.8 per cent respectively).Reference Zagólski30 Cytomegalovirus is another common HIV-related opportunistic infection that has been reported to cause sensorineural hearing loss, and peripheral and central neurological manifestations, in subjects infected with HIV (the HIV status of subjects in the aforementioned study was not indicated).Reference Meynard, El Amrani, Meyohas, Fligny, Gozlan and Rozenbaum31, Reference Vancíková and Dvorák32 These findings suggest that opportunistic infections like herpes zoster virus and Ramsay Hunt syndrome may result in involvement of both the vestibulocollic reflex and vestibulo-ocular reflex pathways.
The use of antiretroviral therapies could also contribute to the higher occurrence of abnormal cervical vestibular-evoked myogenic potential and caloric test findings in subjects with HIV. Those exposed to antiretroviral therapies presented with slightly more cervical vestibular-evoked myogenic potential and caloric testing abnormalities (66.7 per cent) than those not receiving antiretroviral therapies (63.6 per cent); these differences, however, were not statistically significant. Recent studies found similar results with auditory brainstem response (ABR) testing, demonstrating a higher occurrence of abnormal ABR findings in subjects receiving antiretroviral therapies (62.5 per cent) compared with those not receiving antiretroviral therapies (50 per cent), although the difference was not statistically significant.Reference Matas, Silva, de Almeida Marcon and Gonçalves33 Such findings suggest that the auditory and vestibular nerves, and the structures that they innervate, are at risk as a result of the possible ototoxic effects of some antiretroviral therapies.
Antiretroviral regimes may consist of three or more classes of drugs, and one or more of these are nucleoside or nucleotide analogue reverse transcriptase inhibitors.Reference Moyle34 One adverse effect of these inhibitors is mitochondrial toxicity, which is responsible for, among other things, myopathy and neuropathy.Reference Warnke, Barreto and Temesgen35 Neuropathy is a dysfunction of the nervous system, and may therefore involve the vestibular branches of the VIIIth cranial nerve. Additionally, there are case reports of ototoxic sensorineural hearing loss associated with the use of nucleoside or nucleotide analogue reverse transcriptase inhibitors which may have been induced by a reduction in the mitochondrial DNA content, although ageing and the virus itself could have contributed to mitochondrial DNA mutations.Reference Simdon, Watters, Bartlett and Connick36 If these drugs can cause sensorineural hearing loss and affect the auditory brainstem pathways, it is also likely that they can affect the vestibular nerves and/or end organs in subjects receiving antiretroviral therapies.
A recent study compared the vestibular function of HIV-positive subjects receiving highly active antiretroviral therapy with that of age- and gender-matched HIV-negative subjects.Reference Cohen, Cox, Springer, Hoffman, Young and Margolick37 They performed vestibular screening tests, which consisted of head thrust tests, Dix–Hallpike manoeuvres and Romberg balance tests, and found no significant difference between the two groups. The subjects were excluded if they had vestibular complaints, as this could suggest that they were centrally compensated. Therefore, the vestibular screening tests employed may have been underpowered for the detection of subclinical vestibular involvement in subjects who were centrally compensated.
The occurrence of abnormal cervical vestibular-evoked myogenic potential and caloric test findings increased with progression of the disease, rising from 13.3 per cent in the early stages (Classification System for HIV Infection15 category 1) to 30.1 per cent in the advanced stages (category 3). Three previous studies that also used a cross-sectional research design also demonstrated an increase in vestibular involvement from early to advanced disease stages.Reference Castello, Baroni and Pallestrini23, Reference Hausler, Vibert, Koralnik and Hirschel24, Reference Teggi, Ceserani, Luce, Lazzarin and Bussi26 A detailed summary of those findings has been reviewed elsewhere.Reference Heinze, Swanepoel and Hofmeyr13 There is a higher occurrence of abnormal cervical vestibular-evoked myogenic potential and caloric test findings in the advanced stages of HIV due to the reduction in cluster of differentiation 4 cell counts, which places the infected individual at risk of various opportunistic infections. This necessitates the use of antiretroviral therapies to strengthen immunity in order to combat opportunistic infections; however, the vestibular nerves and structures of the vestibular end organs may be susceptible to ototoxicity and to the infections themselves, resulting in abnormal cervical vestibular-evoked myogenic potential and caloric test results.
• The vestibulocollic reflex stabilises the head during active movements
• Testing of the vestibulo-ocular reflex is the basis of many vestibular tests
• The occurrence of abnormal vestibular function is high in adults infected with human immunodeficiency virus (HIV)
• Previous reports focused on vestibulo-ocular reflex tests; it was not known whether HIV affected vestibulocollic reflex pathways
• Cervical or collic vestibular-evoked myogenic potentials allow direct testing of vestibulocollic reflex pathways
• The study results indicated a high occurrence of vestibulocollic reflex pathway abnormalities in adults with HIV
The VIIIth cranial nerve and brainstem pathways may undergo neuropathological changes such as subcortical demyelination because of the HIV infection itself.Reference Bankaitis and Keith38, Reference Reyes-Contreras, Silva-Rojas, Ysunza-Rivera, Jiménez-Ruíz, Berruecos-Villalobos and Romo-Gutiérrez39 This may explain the ABR abnormalities observed in HIV-positive individuals with and without clinical features of the disease, irrespective of normal hearing thresholds.Reference Bankaitis and Keith38 As ABRs may indicate subclinical pathological changes in the peripheral auditory nervous system, cervical vestibular-evoked myogenic potential and caloric testing may indicate pathological changes in the vestibulo-ocular reflex and vestibulocollic reflex pathways respectively in adults infected with HIV. Posturography may be useful too in detecting pathological changes in the vestibulospinal reflex pathways;Reference Dellepiane, Medicina, Mora and Salami40 however, this test procedure was unavailable to the researchers.
Conclusion
There was a significantly higher occurrence of abnormal cervical vestibular-evoked myogenic potential responses and caloric test results in the adults with HIV than in those without HIV. The abnormalities shown by the cervical vestibular-evoked myogenic potential and caloric tests were probably due to pathology of the vestibulocollic reflex and vestibulo-ocular reflex pathways respectively. A combination of cervical vestibular-evoked myogenic potential and caloric tests in the vestibular test battery for adults infected with HIV may offer a tool for detecting early neurological involvement, irrespective of disease progression and clinical manifestations.
