Vestibulotoxicity: strategies for clinical diagnosis and rehabilitation

Abstract

Objective: The purpose of this article is to discuss the most commonly prescribed vestibulotoxic medications and their impact on the vestibular system, to describe the clinical features of vestibular ototoxicity including symptoms reported by patients, and to describe assessment tools that may be used in a monitoring programme, including the functional impact of vestibular loss. Recently published data from a cohort of patients exposed to systemic aminoglycosides (AGS) are summarised, which highlight the importance of monitoring. The role and importance of vestibular rehabilitation in treating affected individuals is discussed. Design: This is a descriptive article. Study sample: Recently published data from 71 patients with cystic fibrosis with AGS exposure are summarised. Results: Recently published data from a cohort of patients exposed to systemic AGS reveal a high prevalence of vestibular system involvement. Conclusions: Evidence suggests that including assessment of vestibular function in a programme to monitor for ototoxic damage is essential. While suggestions about possible components of a monitoring programme are made, the need for further study in order to determine an ideal protocol for assessing vestibular system function during and following exposure to toxic agents is stressed.

Key Words:

• Ototoxicity
• vestibulotoxicity
• aminoglycosides
• monitoring
• vestibular
• oscillopsia
• vestibular rehabilitation

Introduction

Ototoxicity is the focus of this special issue of International Journal of Audiology and other articles have addressed why it is a topic of concern, the role of audiology in clinical trials of new drugs that may be ototoxic, and ototoxic considerations for various specific patient populations. The concern about ototoxicity is not new, although audiologists continue to struggle with how best to understand the mechanisms of auditory and vestibular loss as well as how to evaluate patients undergoing exposure to these drugs, especially in terms of the evaluation of vestibular system function. While other articles in this issue address the impacts of ototoxic agents on the auditory system as well as protocols for monitoring auditory function, the purpose of this article is to specifically address the risks of vestibular ototoxicity and to discuss potential methods for monitoring vestibular system function in exposed patients. The toxic impact of these drugs in the inner ear is determined by a number of factors including the overall health of the patient and prior or concurrent exposures to other agents. For example, patients with cystic fibrosis (CF) are often routinely treated with inhaled aminoglycosides (AGS) including tobramycin and amikacin. In addition, pulmonary exacerbations may be treated with systemic AGS including gentamicin, tobramycin, and amikacin. The cumulative exposure to inhaled and systemic AGS may impact ototoxic damage (Garinis et al. 2017). Also, in CF patients with pancreatic insufficiency and diabetes, the functional impact of ototoxicity may be greater due to loss of other sensory system function associated with diabetes. This article will also include a discussion of recently published data that will support the need to independently monitor auditory and vestibular function in this patient population.

Clinical features of vestibular ototoxicity

Symptoms of vestibular ototoxicity, both unilateral and bilateral, are variable across patients and may include oscillopsia, which is the perception that viewed stationary objects or surroundings move coincident with head movement, dizziness, motion sickness, and unsteadiness when standing or walking, especially in the dark (Ahmed et al. 2012; Black et al. 2001, 2004; Ishiyama et al. 2006). With oscillopsia, the illusory movements occur on the same plane as head movement but in the opposite direction. For example, head movement resulting from walking may yield the perception that the environment is bobbing up and down, whereas turning the head may result in the sensation that the environment is whirling around. Severe oscillopsia can prevent an individual from having clear vision with even the slightest head movement, negatively impacting the ability to read and to participate in routine activities of daily living (Crawford 1952).

Unsteadiness when standing or walking can range from mild to extremely disabling, depending both on the severity of the vestibular loss as well as on the functional integrity of other sensory systems. Because patients with bilateral vestibular loss are more dependent on visual and somatosensory input for the maintenance of postural control, disorders that negatively impact the function of one or both of those systems will have a greater functional impact. For example, diabetic peripheral neuropathy and retinopathy would be expected to increase the functional impact of a bilateral vestibular loss for individuals who are otherwise healthy.

It is logical to assume that systemically administered medications would affect both ears in the same way. However, our data do not support that assumption. Specifically, while systemically administered medications may result in a bilateral vestibular loss, they may also result in non-lateralised peripheral vestibular impairment or unilateral vestibular loss (Handelsman et al. 2017). With unilateral vestibular paresis or non-lateralised vestibular loss, symptoms may include vertigo, which is the sensation of movement when there is none, nausea, and general disequilibrium. Furthermore, even when a bilateral weakness is present, asymmetries in vestibular function between the ears may occur, resulting in a symptom profile that is consistent with both a bilateral and a unilateral vestibular loss (Handelsman et al. 2017).

Vestibular reflexes

Before discussing the evaluation tools that are associated with the assessment of vestibular function, it is appropriate to provide an overview of the reflexes that are utilised in the assessment process. The vestibulo-ocular reflex (VOR) enables individuals to maintain clear vision when the head is in motion by keeping objects of interest on the fovea through inputs from the semicircular canals and the otolith organs. VOR function is evaluated by observing eye movements in the bedside setting and by recording eye movements in the laboratory setting using electrodes or infra-red video technology. The vestibulo-spinal reflex (VSR) stabilises the body and helps to maintain postural control using inputs from the visual and vestibular systems as well as from touch receptors on the skin coupled with proprioceptors on the soles of the feet, hands, and torso. These signals travel to the brain along the lateral, medial, and reticulospinal tracts and generate reflexes in the skeletal muscles of the legs and truck. Body position and VSR may be assessed using formalised or bedside tests of balance. Finally, the vestibulocollic reflex (VCR), which is a component of the VSR, plays a role in stabilising vision during body movement by keeping the head still and level during ambulation. VCR function is evaluated in the laboratory setting using vestibular-evoked myogenic potentials (VEMPs).

Evaluation tools

An important initial component of evaluating a patient is obtaining a thorough history of previous and current medical conditions as well as current symptoms. Relevant to patients with vestibular disorders is an assessment of how symptoms are impacting an individual’s quality of life. The sections that follow will discuss validated subjective measure of dizziness and balance handicap, bedside tests that can be used to assess vestibular reflex function, and formal laboratory tests that require specialised equipment. There is a third “hybrid” category of instrumented tests that may be utilised at the bedside. For example, vHIT, which is discussed later in the article, is an instrumented test of VOR function that can be used at the bedside. Because ototoxic medications are often administered to very ill patients with limited ability to be transported to the clinic, using a combination of strategies is ideal. Measures need to be stable within patients in order to facilitate monitoring changes in vestibular function over time. In addition, measures need to be sensitive enough to reliably detect subtle changes in function to identify early vestibular loss. A summary of the currently available bedside, laboratory, and hybrid tests follows the discussion of the assessment of handicap/disability.

Assessment of handicap/disability

The Dizziness Handicap Inventory (DHI) is a validated measure that is intended to assess the impact of dizziness and unsteadiness on a person’s quality of life (Jacobson and Newman 1990; Newman and Jacobson 1993). It is a 25-item scale that is intended to assess patients’ self-perceived handicap related to their dizziness and/or balance problems. Patients are asked to rate each item as “yes,” “sometimes” or “no.” “Yes” items receive four points, “sometimes” items receive two points, and “no” responses receive zero points. The maximum total points is 100 and the higher the score, the higher the self-perceived impact a patient’s symptoms are having his or her quality of life.

The DHI has been used broadly as a tool to assess patients’ perception of the impact of their disorders on everyday function. For example, when characterising the self-perceived handicap of patients with bilateral vestibular loss, Jacobson and Calder (2000) found significant differences for the DHI total and physical subscale scores between individuals with bilateral vestibular loss and those with unilateral vestibular loss. Individuals with bilateral vestibular system involvement experienced greater degrees of self -perceived handicap than did those with unilateral involvement (Jacobson and Calder 2000). Additionally, in a case report that discussed vestibular rehabilitation for treatment of acquired vestibular impairment secondary to ototoxicity, the DHI provided evidence of perceived functional improvement over time (Calder and Jacobson 2000).

There are also two validated measures for use with children. The Vanderbilt Pediatric Dizziness Handicap Inventory for Patient Caregivers (DHI-PC) is a reliable validated adaptation of the DHI for use with paediatric patients that is completed by their caregivers (McCaslin et al. 2015). It contains 21 items related to the impact the patient’s symptoms are having on daily life and is scored in the same way as the DHI with “yes” items receiving four points, “sometimes” items receiving two points, and “no” receiving zero points. As is the case with the DHI, the higher the score the more perceived handicap related to dizziness symptoms. The maximum possible score for the DHI-PC is 84% (McCaslin et al. 2015).

The Pediatric Vestibular System Questionnaire (PVSQ) aims to identify and measure the severity of common vestibular symptoms of dizziness and/or unsteadiness in children between the ages of 6 and 17 years of age. It was recently validated on a group of 168 healthy children and 56 children with post-concussion dizziness or a vestibular disorder (Pavlou et al. 2016). The PVSQ contains 10 items for which the child rates their symptoms over the past month on a four-point scale ranging from never to most of the time as well as one open-ended question. The results of the validation study demonstrated that the PVSQ was able to discriminate between healthy children and those with vestibular symptoms. As is the case with the DHI-PC, the higher the score, the greater the symptom severity (Pavlou et al. 2016).

Bedside tests

One of the most important parts of the clinical evaluation of patients being exposed to potentially ototoxic agents is asking targeted questions about subjective complaints. Unfortunately, patient reports of vague dizziness, fatigue, blurred vision, and unsteadiness when walking can be completely dismissed, attributed to the patient’s illness, or to deconditioning that results from inactivity that is associated with treatment (Ahmed et al. 2012). Because of the body’s ability to compensate over time for vestibular loss, recovery of function may be erroneously viewed as evidence that the symptoms reported were not real. For example, the reports of dizziness by a young adult with CF coincident with each course of AGS might be overlooked or misinterpreted because the symptoms subsided a few weeks following treatment. Also, patients may underreport symptoms, either because they attribute what they are feeling to their disease or because they do not want to risk changes in their medical management if problems associated with their treatment are detected. In conversations with patients, they affirmed both are true. For those reasons, asking targeted questions about changes in vision with head movement, motion-provoked dizziness, vertigo, and unsteadiness are important to any monitoring programme. Understanding the relevance of even subtle changes that are reported is important.

Dynamic visual acuity testing

Dynamic visual acuity (DVA) testing is a simple way to look for evidence of vestibular loss, or oscillopsia. While oscillopsia is most typically associated with bilateral vestibular loss, it can also occur in the presence of unilateral loss, particularly in the acute phase. When the vestibular system is intact bilaterally, vision is as clear with head movement as it is with the head still. It is a reduction in the VOR that results in changes to visual acuity with head movement. At the bedside, DVA testing can be accomplished by having the patient read something from a book or magazine with the head still, followed by having the patient read the same passage while moving the head from side to side or up and down at 1–2 Hz. If the patient is able to read with and without head movement, evidence of oscillopsia is not present. On the other hand, changes in the patient’s ability to read with head movement provide evidence of oscillopsia. For young children who are not yet able to read, the same task could be performed while looking at shapes. A more formal assessment is possible using a Snellen chart and comparing the smallest line a patient is able to read with the head stationary with the smallest line he is able to read with head motion. In this case, a change in visual acuity of two lines or more is considered to be abnormal and suggestive of oscillopsia. The advantage of the latter strategy is that it provides a mechanism to determine when incremental changes in function occur over time. This more formal assessment can also be modified for young children using the Allen figures, Lea Symbols^®, or “Tumbling E” eye charts (American Academy of Pediatrics, American Association of Certified Orthoptists, American Association for Pediatric Ophthalmology and Strabismus, American Academy of Ophthalmology 2003).

Head thrust testing

The head thrust test, as originally described by Halmagyi and Curthoys (1988), is useful in identifying both unilateral and bilateral canal paresis. Because an intact VOR is necessary for the maintenance of stable gaze during rapid head movements, individuals with a canal paresis will demonstrate a catch-up saccade when the head is rotated quickly to the side of the weakness while the patient focuses on a stationary visual target. This test is also referred to as the head impulse test (HIT). In the case of a bilateral weakness, the examiner would expect to see corrective saccades in both directions of head rotation. There are a few limitations of this test for clinical use. One is that an eye blink can obscure visualisation of the corrective saccade. Another is that there is limited evidence regarding how much paresis is necessary to result in a positive head thrust test. Finally, it is not recommended for patients with neck mobility issues.

Head shaking nystagmus

The presence of nystagmus following head rotation from side to side is suggestive of an asymmetry in the vestibular system. Head rotation on the horizontal plane symmetrically stimulates the vestibular system in both ears. However, when an imbalance in the system is present, the stronger response on the healthy side builds up and is stored in the central velocity storage mechanism. When the head shaking stops, the stored energy is released in the form of slow eye movements that are typically away from the stronger ear, resulting in nystagmus that initially beats toward the stronger ear and away from the weaker ear (Hain et al. 1987). In the context of a bedside evaluation, the patient wears Frenzl lenses allowing the examiner to clearly view eye movements while simultaneously blurring the patient’s vision to prevent fixation and suppression of the VOR. A hybrid version of the test would involve recording the eye movements using infra-red video goggles. The patient’s eyes are observed for the presence of nystagmus for 10 s following cessation of movement. The presence of more than a few beats of nystagmus suggests an imbalance in the horizontal VOR (Hain et al. 1987).

Postural control

Bedside tests of postural control are intended to assess the VSR, which are mediated by the semicircular canals and the otolith organs. In patients with bilateral vestibular loss, postural control may be impaired due to a loss of those reflexes (Horak et al. 2002). The Fukuda stepping test (marching with eyes open and eyes closed for 30 s) and tandem (heel to toe) walking with eyes open and closed are measures of static imbalance of the horizontal canal reflexes. The Fukuda stepping test is considered to be abnormal if the patient rotates to the right or to the left (toward the side of the vestibular weakness) when marching with eyes closed. The Romberg tests are sensitive to an asymmetry affecting the vertical canal reflexes. The Romberg test involves having a patient stand with feet together both with eyes open and eyes closed. It is considered to be positive when the patient experiences increased postural sway or falls with eyes closed. The sensitivity of the test can be increased by having the patient stand with feet in the tandem position or by having him or her stand on soft foam, disrupting proprioceptive inputs (Lanska and Goetz 2000). Observation of postural stability during rapid body turns and in response to external perturbations can also be used to assess dynamic vestibulo-spinal function (Zee and Fletcher 1996). This could be accomplished by applying a gentle push to the shoulders of the patient.

For children, standardised assessments include the Bruininks Oseretsky Test of Motor Proficiency II (BOT2) for children four years and older (Bruininks and Bruininks 2005) and the Peabody Developmental Motor Scales (Folio and Fewell 2000). Also, normal cut-offs have been established for children aged 4–15 years old for the following tests of static balance: single leg stance (SLS) eyes open (EO) on a firm surface; SLS eyes closed (EC) on a firm surface; and SLS EO on a foam surface; tandem stance (TS) EO on a firm surface; TS EC on a firm surface; and TS EO on a firm surface (Condon and Cremin 2014). Timed tests of single leg stance with eyes open and closed as well as tandem stance eyes closed have been effective in discriminating between children age ≥4 years with normal vestibular function and those with total bilateral vestibular loss (Oyewumi et al. 2016).

Laboratory tests

The purpose of laboratory tests of vestibular function is to use calibrated stimuli and objectively measured responses to determine whether an individual patient’s performance falls within the normal range, and, if not, to determine the likely source of the problem. Clearly, there are patient and examiner variables that can impact the validity of the results. Examples of some patient variables include medications, fatigue, attention to task, the ability to follow instructions, and mobility limitations. Most examiner variables are related to adequately controlling the patient variables as well as repeating tests as needed. The assumption for this discussion of laboratory tests is that patient and examiner variables have been adequately controlled. The components of a comprehensive vestibular evaluation are described in the following sections. Certainly, not all clinics will have access to this equipment and test protocols and may need to rely on bedside tests to monitor function.

Electronystagmography and videonystagmography

For both electronystagmography (ENG) and videonystagmography (VNG) testing, eye movements are recorded on the horizontal and vertical planes. Binocular ENG testing using two channel recordings involves placing electrodes adjacent to the inner and outer canthi of the eyes as well as above and below each eye whereas single channel recordings utilise electrodes only adjacent to the outer canthi and above and below one eye. When needed, eye movements can be recorded from only one eye. For VNG testing, eye movements are recorded using infra-red video cameras and digitally stored during VNG testing. While torsional eye movements can be observed and stored for later viewing, they are not typically quantified. The typical battery of tests includes recording and measurements of spontaneous eye movements, positional and positioning tests, and caloric testing. Because an intact central vestibulo-ocular pathway is essential to normal eye movement recordings for the purpose of determining vestibular function, initial testing requires an evaluation of oculomotor function.

Caloric testing is the only part of the vestibular test battery that allows the examiner to evaluate the performance of each ear independently. During caloric testing, a non-physiologic thermal stimulus (air or water) is delivered into the ear canals with the intention of inducing endolymph flow in the horizontal semicircular canals by creating a temperature gradient from one side of the canal to the other. Four irrigations are typically completed including one warm and one cool irrigation of each ear in alternating fashion, although monothermal calorics can be effectively utilised to predict unilateral vestibular loss in some circumstances (Bush et al. 2013; Shupak et al. 2010). The dependent measure is the peak slow-component eye velocity (SCEV) for each of the irrigations, which is then used to compare the total eye speed for each ear and the total eye speed for each direction (left-beating and right-beating) of nystagmus. The difference in total eye speed for each ear is expressed as a percentage relative to the overall total eye speed for all four irrigations. This difference is referred to as the unilateral weakness for the side with the lower percentage, while the difference in right-beating nystagmus versus left-beating nystagmus compared to the overall total eye speed is referred to as the directional preponderance. It is important that each clinic establish its own definitions of normal because stimulus variables such as type of stimulus (air or water), temperature, flow duration, and technique will impact the strength of the resultant nystagmus. It is noteworthy that paediatric patients may require protocol alterations such as decreasing flow time or adjusting the stimulus temperature to be closer to normal body temperature. For example, the flow time and/or warm stimulus temperature could be decreased if the patient is not able to tolerate the standard stimulus. These changes may be necessary to facilitate completion of the tests in children who may be intolerant of the stimuli or to the sensations of dizziness. While those changes are important to successful testing with children, they should be documented. When calorics are used to assess the impact of potentially toxic agents on vestibular function, it is essential that testing be conducted in the same manner each time for individual patients.

Rotational testing

As is the case with VNG testing, rotational testing typically involves the recording of eye movements using infra-red video technology. Electrode recording is possible for patients for whom the video goggles are not an option, and for young children eye movements may be observable using an infra-red camera attached to the chair. Rotational testing is accomplished by having a patient seated in an earth-vertical position in a chair that is attached to a motor capable of delivering calibrated sinusoidal, pseudo-random, or constant velocity stimuli in a darkened environment. Unlike caloric testing, both ears are stimulated simultaneously by a physiologic stimulus and audiologists are able to assess vestibular function throughout more of the frequency range to which it is responsive. Whereas caloric stimulation is equivalent to rotation at 0.002–0.004 Hz (Furman et al. 1988), the typical frequencies included during the sinusoidal test battery more closely align with head movements associated with standing balance and ambulation. The dependent measures during sinusoidal rotational testing are gain, phase, and symmetry. Gain is the amplitude of eye movement peak velocity relative to the chair, phase is the timing relationship between the head (chair) movement and the eyes, and symmetry is the relative amplitude of responses for rightward versus leftward chair (head) movements.

Rotational step testing involves rapidly accelerating the patient and rotating him or her at a constant velocity long enough for the endolymph to catch up with the head, followed by rapid deceleration that provides a vestibular stimulus in the opposite direction. Step testing is typically completed sequentially beginning with both clockwise and counterclockwise rotations. For rotational step testing, the dependent measures of interest are the peak eye velocity and the time constant, which is defined as the time in seconds it takes for nystagmus speed to reduce to two-thirds of its peak. Abnormal rotational chair results can be associated with non-localised vestibular loss, unilateral vestibular loss, and bilateral vestibular loss (Table 1).

Table 1. Criteria for each category of vestibular disorders.

Category A: non-lateralised peripheral vestibular disorder	Category B: unilateral vestibular paresis	Category C: bilateral vestibular paresis
Spontaneous nystagmus	Caloric inter-ear asymmetry ≥25%	Each of the four caloric responses <10°/sec
Significant post-head shaking nystagmus	Any combination of Category A results may also be present	Gain reductions in rotational step testing
Significant positional nystagmus		Gain reductions in sinusoidal rotational testing
Short time constants for rotational steps		Any combination of Category A results may also be present
Increased phase leads at any sinusoidal frequency
Significant rotary chair asymmetries

Rotational testing is important for monitoring changes in function that may be the result of ototoxic damage, or even establishing whether unilateral or bilateral vestibular loss exists. Significant gain reductions are suggestive of bilateral canal paresis, although the extent and functional impact of the injury can be defined based on how many frequencies are affected. For example, an individual with gain reductions in the low frequencies and preserved high-frequency performance is likely to do better from a functional perspective than is someone with low gain throughout the frequency range tested. In terms of monitoring for changes in function over time related to toxic agents, having information about function across the frequency range is important. In a manner similar to ototoxic hearing loss, it may be that changes in vestibular function manifest earlier for some frequencies of rotation than others. Specifically, low-frequency vestibular function is likely to be impacted before high-frequency function. Similarly, for rotational step testing, time constants can move from the normal range to being short, suggesting new vestibular system involvement. These protocols can be adapted for young children with the use of a booster seat. Also, peak chair velocity may be decreased for sinusoidal and rotational step testing.

Static balance/posturography

Computerised dynamic posturography (CDP) serves as the gold standard test for evaluating an individual’s ability to use vision, proprioception, and vestibular system input for the maintenance of upright stance. Importantly, there are age-adjusted normal values that extend into early childhood for the sensory organisation test (SOT) (Foudriat et al. 1993; Ferber-Viart et al. 2007). During SOT, anterior and posterior sway and sheer are measured by sensors in the platform while the patient’s access to visual and support surface cues is systematically altered. An individual patient’s performance is described in terms of patterns of sway within and across the six testing conditions. Patients with uncompensated unilateral or bilateral peripheral vestibular loss typically exhibit increased postural sway or fall reactions in Conditions 5 (sway referenced support, absent vision) and 6 (sway referenced support and vision). Those with functionally compensated unilateral vestibular loss may perform normally across all conditions. Those with bilateral vestibular loss would not be able to perform normally. Specifically, they would exhibit fall reactions in Conditions 5 and 6. This test provides valuable information about how well a person is able to utilise and integrate sensory information, which may help provide insight about problems with gait and stance in real-word environments including on uneven surfaces and in darkness.

Video head impulse test

While head thrust testing at the bedside can be useful in detecting canal paresis, and quantitative HIT testing using scleral search coils to measure eye movements have been described for some time (Collewijn and Smeets 2000; Weber et al. 2008), video head impulse test (vHIT) provides some distinct advantages over both. With vHIT, eye movements are recorded during and following head perturbations in a non-invasive fashion using a small infra-red video camera that is attached to a pair of lightweight tightly fitting goggles. The patient is instructed to focus on a stationary visual target while the head is manipulated by the examiner along three planes that are aligned with the planes of the semicircular canals. VHIT is capable of assessing canal function for all three paired semicircular canals [i.e. horizontal, RALP (right anterior left posterior) and LARP (left anterior right posterior)] even in children as young as three years of age (Hamilton et al. 2015; Khater and Afifi 2016).

The dependent variables of interest in vHIT testing are comparison of head and eye movement gain and phase as well as the presence of overt or covert saccades. Because sensors attached to the goggles quantify head direction and velocity for each perturbation, and eye movement recordings quantify eye direction and velocity for each perturbation, a direct comparison of the two is represented graphically for each direction of head movement. In addition, both overt and covert saccades are apparent in the eye tracings. Overt saccades are those that occur following the head movement and may represent the corrective saccades that are apparent during bedside head thrust testing as described earlier. Covert saccades, on the other hand, are saccades that occur during the head thrusts and may be missed during bedside testing. Covert saccades are important since they may also be indicative of vestibular paresis but would have been missed with clinical bedside testing. In a study describing gentamicin ototoxicity, Ahmed and colleagues reported that the VOR for all six semicircular canals was significantly reduced and both overt and covert saccades were present (Ahmed et al. 2012). Additionally, in a recent study evaluating whether vHIT could be used to define the severity of bilateral vestibular hypofunction, the results revealed a significant positive linear relationship between lateral canal vHIT gain and rotary chair gains. In addition, vHIT VOR gain increased as the severity of the bilateral hypofunction decreased (Judge, Janky, and Barin 2017). These findings are encouraging for including vHIT testing in a monitoring programme for vestibular ototoxicity although additional research is needed.

Vestibular-evoked myogenic potential

Most of the clinical tests of vestibular function focus on assessing the function of the semicircular canals, which until recently has been limited to the horizontal canals. In the same way that vHIT testing allows us to evaluate all of the semicircular canals, vestibular-evoked myogenic potential (VEMP) testing can provide insight into the health of the otolith organs of the inner ear. There is a paucity of evidence about the application of VEMP testing in assessing vestibular function relative to ototoxicity due to systemic exposure. However, VEMP testing has been used to demonstrate loss of saccular function following intratympanic gentamicin treatment in patients with Meniere’s disease (Helling et al. 2007). Specifically, while most patients exhibited VEMP responses prior to treatment with intratympanic gentamicin treatment, VEMPs were absent in all treated patients.

Cervical-Evoked Myogenic Potentials (cVEMP). cVEMP is thought to provide information about the saccule and inferior vestibular nerve, although animal research suggests that other variables, such as threat-induced anxiety, fear, and arousal can contribute to the response (Naranjo et al. 2016). It relies on a relaxation response of the sternocleidomastoid (SCM) muscle in response to a loud sound delivered to the saccule through the ear canal and middle ear, which is turn activates the inferior vestibular nerve, lateral vestibular nucleus, the spinal accessory nerve nucleus, and then the ipsilateral SCM muscle (Hain 2016). cVEMP responses can be elicited by clicks or tone bursts presented at a high intensity level (e.g. ≥90–95 dB SPL), although research supports the use of 500 Hz tone bursts as the preferred stimulus (Akin et al. 2003; Janky and Shepard 2009). Recording parameters may need to be adjusted based on the age of the patient. For example, creative methods may be required with children to ensure adequate contraction of the SCM throughout testing.

Testing typically involves placement of electrodes on the body of the SCM around the middle third of the muscle with the reference electrode on the clavicular joint. It is important to have symmetric placements from side to side. The ground electrode can be placed on the forehead. Because the cVEMP is a relaxation response, it is important that the patient is able to generate and maintain contraction of the SCM during testing. Various techniques have been recommended including elevating and turning the head to one side. Recording of the baseline EMG activity prior to and during cVEMP testing has also been recommended, especially since the evaluation of function is not only dependent on the absolute amplitude of the cVEMP responses obtained, but on being able to compare the responses from one ear against the other. For that reason, ensuring that testing for both ears occurred under identical conditions is essential, as is the ability to account for SCM contraction via EMG.

Ocular-Evoked Myogenic Potential (oVEMP). oVEMP testing is intended to provide information about utricular and superior vestibular nerve function. Unlike cVEMP, this is an activation response from the inferior oblique muscles inferior to the eye. Patients are instructed to look up, resulting in muscle contraction. As is the case with cVEMP, acoustic stimuli are delivered through insert earphones placed into the ear canal or via bone conduction. No monitoring of baseline EMG is required because the activity of looking up is likely to result in symmetric engagement of the inferior oblique muscles. While the cVEMP is an ipsilateral response, the oVEMP is a contralateral response that should be observable following very few stimulus presentations (e.g. ≤200 sweeps). Normative data for adult patients have been published (Janky and Shepard 2009; Piker et al. 2011; Murnane et al. 2011). While oVEMP responses can be recorded in very young children and child friendly paradigms are available, paediatric normal data are emerging.

Proposed ototoxicity monitoring programme

There are no generally accepted protocols for monitoring vestibular function during or following exposure to potentially ototoxic agents, in part due to the expense associated with laboratory equipment and in part because patients receiving ototoxic medications may be too ill to fully cooperate with formal testing. In addition, data continues to emerge about test modifications for use with children as well as the reliability and sensitivity of various measures. The previous sections have provided information about various tools that are available for assessing vestibular function and a summary of those tools, and what each is intended to measure is provided in Table 2. Additional study is needed to determine the relative value of each measure in a monitoring programme. For professionals who are fortunate enough to have access to objective measures of vestibular integrity, a comprehensive approach that evaluates semicircular canal and otolith function as well as measures of functional abilities would be ideal. Additionally, because the vestibular system is sensitive across a wide range of frequencies (accelerations), sampling performance across the range is important. In settings where laboratory testing is available, caloric and rotational testing would be recommended to evaluate semicircular canal function because they are stable measures within individual patients. The combination of those tests also provides individual ear data as well as information about the VOR across a wide frequency range. VHIT testing may prove to be a valuable tool for screening, although head accelerations are fairly high. If laboratory testing is not possible, DVA is an informal test that is also incredibly valuable as are measures of standing balance as described earlier. Enlisting the help of a vestibular physical therapist to explore functional measures is useful.

Table 2. This table provides a summary of the various tests that are available for assessing vestibular system function including the test type, dependent variables of interest, and what system in being evaluated.

Test name	Test type	Outcome measure	Generator(s)
DVA	Informal/formal	Decreased visual acuity with head movement	Horizontal canal
Head thrust	Bedside	Catch-up saccades	Horizontal canal
Post head shake	Bedside/hybrid	Nystagmus present	Horizontal canal
Postural control	Bedside	Postural sway/fall reaction or rotation while marching EC	Vertical canals and horizontal canal
VNG (calorics)	Laboratory	VOR gain	Horizontal canal
Sinusoidal rotational tests	Laboratory	VOR phase, gain, and symmetry	Horizontal canal
Rotational step tests	Laboratory	VOR gain and time constant	Horizontal canal
vHIT	Laboratory	VOR gain/presence of covert and overt saccades	Anterior, posterior, and horizontal canals
cVEMP	Laboratory	Response amplitude	Saccule/inferior vestibular nerve
oVEMP	Laboratory	Response amplitude	Utricle/superior vestibular nerve
Posturography SOT	Laboratory	Postural sway/fall reactions	Vision, vestibular, and proprioception

Vestibular rehabilitation

An essential component of a programme intended to address vestibular ototoxicity is vestibular rehabilitation, so having access to a skilled interdisciplinary team is important. Specifically, in addition to the audiologists and physicians who are a part of the care team, physical therapists with training in vestibular disorders are essential to the team. While the results of bedside and instrumented tests provide valuable information about the presence and extent of vestibular loss, an evaluation of the functional impact of the loss is necessary in order to develop and implement a therapy plan. Because vestibular loss can create functional limitations that negatively impact clear vision and mobility, there are safety issues that need to be addressed. The evaluation should address factors that are necessary for static and dynamic balance, including strength, muscle tone, sensation, and coordination as well as impairment measures, basic functional tasks, specific gross motor skills, and high demand functional tasks. Specific activities can be adapted based upon the age and ability of the patient.

The role of vestibular rehabilitation is to promote central nervous system compensation and substitution when a vestibular loss is present, to promote adaptation, and to facilitate habituation of symptoms that are associated with a vestibular loss. More specifically, compensation activities stimulate the remaining sensory, motor, cognitive, and neurologic systems so they can take over while adaptation activities provide movement to generate new neural activity. Habituation and cerebellar suppression activities use repetition to help limit the motion-provoked symptoms that results from a vestibular imbalance. The stimulus for all of these activities is movement and systematic progression is based on symptoms. Specific goals and therapy plans are based on the outcome of a comprehensive therapeutic assessment (Cronin 2013).

Recent evidence regarding vestibular ototoxicity

While there is ample evidence of vestibular ototoxicity in the literature based on animal models, most of what has been published about vestibular ototoxicity in humans is based on case reports or retrospective case series of patients with known vestibular losses (Black et al. 2004; Dhanireddy et al. 2005; Ahmed et al. 2012). However, preliminary data related to the prevalence of auditory and vestibular loss in patients with CF who were treated with IV AGS were summarised elsewhere (Handelsman 2011). Additionally, the overall prevalence data for 71 patients with CF who were exposed to IV AGS were recently published (Handelsman et al. 2017) and are briefly summarised here.

Hearing test outcomes were based on individual ear pure tone air and bone conduction threshold at octave frequencies from 250 to 8000 Hz and were considered to be abnormal if thresholds at any frequency for either ear exceeded 25 dB HL. Vestibular test outcomes were based on results from VNG, sinusoidal rotational testing, and rotational step testing. Abnormal results were sorted into three categories: non-lateralised peripheral vestibular system involvement, unilateral vestibular paresis, or bilateral vestibular paresis (Handelsman et al. 2017).

Of the 71 patients in the sample, 27 (38%) demonstrated evidence of non-lateralised vestibular system involvement, eight (11%) had significant unilateral loss, and 21 (30%) showed evidence of bilateral vestibular paresis. Overall, the prevalence of vestibular loss at 79% (N = 56) is high in this group of patients, and it is much higher than the prevalence of hearing loss at 23% (N = 16). Only 12 (10%) patients had no evidence of auditory or vestibular sensory loss and 13 (18%) demonstrated evidence of both hearing loss and vestibular loss. Of the 55 (78%) patients with normal hearing, 43 (61%) have abnormal vestibular function. These data show no clear relationship between hearing loss and vestibular system involvement. Some patients with severe bilateral vestibular loss had normal hearing, while some patients with significant sensorineural hearing loss had normal vestibular system function. These data support the need to include both hearing and vestibular testing in any ototoxic monitoring protocol.

In addition to the prevalence data described above, Handelsman (2011) previously summarised data for a limited group of patients (N = 22) for whom serial data were available. We were interested in determining whether systematic changes in function as measured by VNG and rotational tests can be detected over time with repeated AGS exposure. Data from the patients for whom serial testing was completed suggest that it is possible. It is noteworthy that while two of five patients remained normal over time and 2 of 14 patients from the non-lateralised vestibular function category became normal, test results revealed significant declines in the VOR gain with repeated exposures for the remaining patients. For example, the patient who initially fell within in the unilateral loss category developed a bilateral vestibular loss and all of the patients who were initially in the bilateral loss category remained there. Further, even within categories there was evidence of decreased function. Specifically, individuals within the bilateral vestibular loss category became more severe, as evidenced by further reduction in the VOR with repeated exposure. These data from a limited number of patients suggest that it may be possible to see systematic changes in function over time. Additionally, because for 20 of 22 patients there was no evidence of improvement in vestibular status over time, the ototoxic damage typically appears to be cumulative and irreversible.

In a recently published systematic review that aimed to investigate the prevalence and characteristics of AGS vestibular ototoxicity in humans, the authors included 27 studies, most of which demonstrated vestibulotoxic effects ranging from 0 to 60% (Van Hecke et al. 2017). In 21 studies, the prevalence of AGS vestibulotoxicity was identified, whereas in the other six studies, the aim was to define the vestibular outcome in patients who were selected based on their known AGS-induced vestibulotoxicity. For patients included in the prevalence studies (N = 117), most with vestibulotoxicity demonstrated bilateral vestibular hypofunction (N = 46) while unilateral paresis was noted (N = 11) across six studies. As was the case in the data from Handelsman et al. (2017), coexisting hearing loss did not consistently occur with vestibulotoxicity (Van Hecke et al. 2017).

Summary and conclusions

Damage to the inner ear from toxic agents remains a concern for patients being treated for potentially life-threatening conditions with medications that put them at risk for hearing loss and/or vestibular loss. It is clear that the damage is unpredictable, so monitoring auditory and vestibular function during and following treatment with AGS is important. Existing evidence is insufficient to accurately estimate the prevalence of AGS vestibuloxicity in adults and children, to identify the most vestibulotoxic AGS type, or to know what specific risk factors to cochleotoxicity also apply to the vestibular system (Van Hecke et al. 2017). Other articles in this issue have thoroughly addressed the need and strategies for monitoring hearing. This article has provided a rationale for evaluating vestibular function as well, and has discussed tools that are available to do so. It is important to consider some pragmatic issues, however, when developing a monitoring programme. For example, while conducting testing with each treatment course is likely ideal, patients are often too ill during treatment to fully participate in a comprehensive evaluation. In addition, their ability to be transported to a clinical laboratory during treatment might be limited. For those reasons, prioritising tools based on ease of use as well as relative sensitivity and specificity of identifying vestibular loss is recommended. Unfortunately, while the data to date support the use of VNG and rotational testing in evaluating at risk patients for vestibular system involvement, it may be difficult to utilise only those tests in a monitoring programme. Recent evidence suggests that VHIT and VEMP testing may prove to be sensitive measures of changes in vestibular function secondary to ototoxicity although additional study is needed. Also, there is little evidence about the vestibular impact of chemotherapeutic agents such as cisplatin on vestibular system function. As a result, additional study is needed to assess the impact of chemotherapeutic agents on the vestibular system and to determine the role of each test in a comprehensive vestibular ototoxicity monitoring programme.

Abbreviations
cVEMP	=	cervical vestibular-evoked myogenic potentials
CP	=	cystic fibrosis
DVA	=	dynamic visual acuity
EMG	=	electromyography
HIT	=	head impulse test
Hz	=	Hertz
IJA	=	International Journal of Audiology
LARP	=	left anterior right posterior
oVEMP	=	ocular-evoked myogenic potentials
RALP	=	right anterior left posterior
SCEV	=	slow-component eye velocity
SOT	=	sensory organisation test
SCM	=	sternocleidomastoid muscle
vHIT	=	video head impulse test
VNG	=	videonystagmography
VOR	=	vestibulo-ocular reflex

Declaration of interest

No potential conflict of interest was reported by the author. This project was partially funded by the Cystic Fibrosis Foundation Therapeutics, Inc., grant number [N008681-385013].

Acknowledgements

I would like to thank my collaborators on the University of Michigan quality improvement clinical protocol for referring patients with CF who are treated with systemic aminoglycosides, W. Michael King, Ph.D., Samya Z. Nasr, M.D., and Crystal Pitts, Au.D.

Additional information

Funding

This project was partially funded by the Cystic Fibrosis Foundation Therapeutics, Inc., grant number [N008681-385013].