Exp Brain Res DOI 10.1007/s00221-013-3782-z
Research Article
Ocular vestibular‑evoked myogenic potentials (oVEMP) to skull taps in normal and dehiscent ears: mechanisms and markers of superior canal dehiscence Rachael L. Taylor · Catherine Blaivie · Andreas P. Bom · Berit Holmeslet · Tony Pansell · Krister Brantberg · Miriam S. Welgampola
Received: 30 May 2013 / Accepted: 15 November 2013 © Springer-Verlag Berlin Heidelberg 2014
Abstract The site of stimulus delivery modulates the waveforms of cervical- and ocular vestibular-evoked myogenic potentials (cVEMP and oVEMP) to skull taps in healthy controls. We examine the influence of stimulus location on the oVEMP waveforms of 18 patients (24 ears) with superior canal dehiscence (SCD) and compare these with the results of 16 healthy control subjects (32 ears). oVEMPs were recorded in response to taps delivered with a triggered tendon-hammer and a hand-held minishaker at three midline locations; the hairline (Fz), vertex (Cz) and occiput (Oz). In controls, Fz stimulation evoked a consistent oVEMP waveform with a negative peak (n1) at 9.5 ± 0.5 ms. In SCD, stimulation at Fz produced large oVEMP waveforms with delayed n1 peaks (tendonhammer = 13.2 ± 1.0 ms and minitap = 11.5 ± 1.1 ms). Vertex taps produced diverse low-amplitude waveforms in controls with n1 peaks at 15.5 ± 1.2 and 13.2 ± 1.3 ms
for tendon-hammer taps and minitaps, respectively; in SCD, they produced large amplitude oVEMP waveforms with n1 peaks at 12.9 ± 0.8 ms (tendon-hammer) and 12.1 ± 0.5 ms (minitap). Occiput stimulation evoked oVEMPs with similar n1 latencies in both groups (tendon-hammer = 11.3 ± 1.3 and 10.7 ± 0.8; minitap = 10.3 ± 0.9 and 11.1 ± 0.4 for control and SCD ears, respectively). Compared to reflex amplitudes, n1 peak latencies to Fz taps provided clearer separation between SCD and control ears. The distinctly different effects of Fz and vertex taps on the oVEMP waveforms may represent an additional non-osseous mechanism of stimulus transmission in SCD. For skull taps at Fz, a prolonged n1 latency is an indicator of SCD.
R. L. Taylor · A. P. Bom · M. S. Welgampola (*) Institute of Clinical Neurosciences, Royal Prince Alfred Hospital, Central Clinical School, University of Sydney, Sydney, NSW, Australia e-mail: [email protected]
Introduction
C. Blaivie European Institute of Otorhinolaryngology, Antwerp, Belgium B. Holmeslet Department of Otolaryngology, St Olavs Hospital, Trondheim University Hospital, Trondheim, Norway T. Pansell Section for Ophthalmology and Vision, Department for Clinical Neurosciences, Karolinska Institutet, Stockholm, Sweden K. Brantberg Department of Audiology, Karolinska Hospital, Stockholm, Sweden
Keywords Superior canal dehiscence · Bone-conducted stimulation · Ocular vestibular-evoked myogenic potentials
Vertigo and imbalance evoked by loud sound or pressure may suggest an absence of bone overlying the superior semi-circular canal, known as superior canal dehiscence syndrome (SCD) (Minor et al. 1998). The defect introduces an additional low impedance pathway or ‘third mobile window’ through the inner ear, leading to an alteration in cochlear and vestibular responses to air- and bone-conducted stimuli. Typically in SCD, vestibular-evoked myogenic potentials (VEMP) to air-conducted (AC) sound have lowered reflex thresholds and large amplitudes. Ocular vestibular-evoked myogenic potentials (oVEMP) can be recorded from infraorbital surface electrodes in response to air- and bone-conducted (AC and BC) stimuli (Todd et al. 2007, 2008; Iwasaki et al. 2007). The typical
13
oVEMP waveform evoked by AC sound and forehead taps at Fz is characterised by an initial negative (excitatory) potential at around 10 ms, followed by a positive peak approximately 5 ms later. This surface waveform represents the synchronous modulation in activity of the inferior oblique muscles (Weber et al. 2012), mainly via a crossed pathway from the contralateral otolith organs (Iwasaki et al. 2008; Curthoys and Vulovic 2011). Stimulation by AC sound depends on transmission through the ossicular chain and the generation of pressure variations in the labyrinthine fluids. oVEMP responses to bone-conducted (BC) stimuli are presumed to originate through a combination of elastic wave propagation around the skull (dominant at high frequencies of stimulation) and whole-head acceleration (for low frequencies below ≈150–400 Hz) (Todd et al. 2008; Jombik et al. 2011). Other possible contributors include the inertia of the middle ear ossicles, acoustic radiation, and pressure transmission through the cerebrospinal fluids (Stenfelt and Goode 2005; Stenfelt 2011). However, the relevance of these alternate BC mechanisms to the generation of tap oVEMPs in health and disease has not been determined. The recording of AC oVEMP amplitudes has emerged as a powerful clinical tool in the diagnosis of SCD. Whereas the distribution of AC cVEMP amplitudes in SCD overlaps with those of healthy controls, AC oVEMP amplitudes provide almost complete separation (100 % sensitivity and specificity) on comparison with age-matched normative data (Taylor et al. 2012; Janky et al. 2013). However, not all clinicians are in favour of using AC stimuli to elicit oVEMPs. Concerns regarding their use were recently aired in an article by Manzari (2013), who highlighted several limitations including low response rates in controls and sensitivity to middle ear pathology as reasons for adopting BC stimuli in clinical oVEMP testing. Unlike AC stimuli, BC taps applied at the hairline in the midline (Fz) are independent of the middle ear system and have been shown to generate robust oVEMP waveforms in healthy controls (Iwasaki et al. 2007, 2008; Rosengren et al. 2011). However, compared to AC oVEMP there are few reports of BC oVEMPs as a test for SCD. Manzari et al. (2012) reported good sensitivity and specificity for oVEMP amplitudes recorded in response to a 500 Hz BC stimulus. Others have shown overlap in the amplitude distributions of SCD patients and controls, implying BC oVEMP amplitudes to be a less-sensitive marker for dehiscence (Welgampola et al. 2008; Janky et al. 2013). The optimal stimulus location for eliciting BC oVEMPs in SCD is also unknown. In healthy controls, changes in oVEMP morphology, amplitudes and latencies are evident as the stimulus is shifted to different areas around the skull (Todd et al. 2008; Lin et al. 2010; Holmeslet et al. 2011; Jombik et al. 2011). Whether similar changes occur in the presence of an SCD has not been confirmed.
13
Exp Brain Res
The objectives of this study were twofold. First, we hypothesised that the presence of a dehiscence may alter the mechanism of BC transmission to the labyrinth. Differences in the pattern of end organ activation could result in identifiable changes in oVEMP morphology in response to stimuli delivered at different midline locations. Additionally, given the potential for widespread future use of BC stimuli in routine oVEMP testing, we felt that differences in amplitudes, latencies and waveform characteristics could be exploited to develop diagnostic markers that separate SCD from control ears.
Materials and methods Subjects oVEMPs were recorded from sixteen controls (aged 36.2 ± 9.8) and eighteen patients (24 ears) with SCD (aged 55.9 ± 12.8). The diagnosis of SCD was confirmed by 3D temporal bone CT scans with 0.3 mm slices and reconstructions in all six canal planes. Patients (23/24 ears) also demonstrated either augmented oVEMP amplitudes (n = 19 ears diagnosed after 2007) or reduced cVEMP thresholds (n = 4 SCD ears diagnosed prior to 2008) in response to AC click stimuli (based on laboratory normal data). One patient with a negative result on AC VEMP testing had previously undergone stapes surgery. This may have compromised middle ear transmission, as evidenced by a maximum conductive hearing loss across the entire frequency range. This patient experienced non-lateralising symptoms of autophony and conductive hyperacusis and had clear evidence of bilateral dehiscences on imaging. Participants gave written informed consent and were studied in accordance with the guidelines of the Helsinki Declaration. Stimuli The stimuli were skull taps, delivered either mechanically using an automated hand-held ‘minishaker’ (model 4810, Bruel and Kjaer) or manually via a triggered tendon-hammer. These stimuli have been shown to produce more reliable oVEMP responses than those produced by a conventional B71 bone-conductor and are widely used clinically (Rosengren et al. 2011). Taps were delivered at three midline locations (Fig. 1) in accordance with the international 10–20 system of electrode placement. The minishaker taps, referred to as ‘minitaps’, were generated in response to a 1 ms square-wave pulse (condensation polarity, 20 V amplitude). Calibration of the minishaker, using a Bruel and Kjaer sound level meter and artificial mastoid (Bruel and Kjaer 4930), confirmed an intensity of 147 dB Force level (24 N). For manually driven (tendon-hammer) taps,
Exp Brain Res Cz
Cz
Fz Fz
Oz
-
+
-
+ Accelerometer
Ground
Fig. 1 Illustration of the two-channel electrode montage, mid-sagittal stimulation sites and accelerometry convention used in the present study. Tendon-hammer taps and minitaps were delivered normal to the skull surface at the locations indicated. Based on the international 10–20 system of electrode placement; Fz represents stimulation at the hairline, Cz corresponds to the vertex and Oz to the occiput. An example of the accelerometer positioning is shown for left mastoid
a voltage output greater than 2.5 V was required to activate the trigger pulse. This corresponded to a peak deceleration >8.0 g on delivery of the stimulus to the skull surface. 3D accelerometry The frequency and directional properties of the stimulusskull response were defined in 3D for six controls and two SCD patients using triaxial accelerometers (TMS international), which were secured over each mastoid process using adhesive tape and a tight elastic headband. The averaged results for the six controls are shown in Fig. 2. For all SCD and control ears, stimuli delivered at Fz and the vertex caused the mastoids to first accelerate outward (lateral, Y-axis), downward (Z-axis) and backward (X-axis). The direction reversed upon stimulation at the occiput. Acceleration responses for left and right mastoid recordings were averaged and are compared against SCD ears in “Appendix”. For Fz stimulation, the mean magnitude of acceleration was greatest in the X-axis for both stimuli. Stimulation at the occiput also produced the largest acceleration magnitude in the X-axis for the tendon-hammer, while for minitaps the Y component was largest. For vertex tendon-hammer taps, acceleration was greatest in the Z-axis; for minitaps, Z and Y components were of equal magnitude.
recordings. Three-dimensional acceleration for the left (aL) and right (aR) mastoid is defined inthe naso-occipital (X),inter-aural (Y) and axR (t) axL (t) R L R L vertical (Z) axes a (t) = ay (t) , a (t) = ay (t) azR (t) azL (t)
Fourier analysis, performed on a 10 ms Hanning window from stimulus onset, confirmed a consistent low frequency (peak
Research Article
Ocular vestibular‑evoked myogenic potentials (oVEMP) to skull taps in normal and dehiscent ears: mechanisms and markers of superior canal dehiscence Rachael L. Taylor · Catherine Blaivie · Andreas P. Bom · Berit Holmeslet · Tony Pansell · Krister Brantberg · Miriam S. Welgampola
Received: 30 May 2013 / Accepted: 15 November 2013 © Springer-Verlag Berlin Heidelberg 2014
Abstract The site of stimulus delivery modulates the waveforms of cervical- and ocular vestibular-evoked myogenic potentials (cVEMP and oVEMP) to skull taps in healthy controls. We examine the influence of stimulus location on the oVEMP waveforms of 18 patients (24 ears) with superior canal dehiscence (SCD) and compare these with the results of 16 healthy control subjects (32 ears). oVEMPs were recorded in response to taps delivered with a triggered tendon-hammer and a hand-held minishaker at three midline locations; the hairline (Fz), vertex (Cz) and occiput (Oz). In controls, Fz stimulation evoked a consistent oVEMP waveform with a negative peak (n1) at 9.5 ± 0.5 ms. In SCD, stimulation at Fz produced large oVEMP waveforms with delayed n1 peaks (tendonhammer = 13.2 ± 1.0 ms and minitap = 11.5 ± 1.1 ms). Vertex taps produced diverse low-amplitude waveforms in controls with n1 peaks at 15.5 ± 1.2 and 13.2 ± 1.3 ms
for tendon-hammer taps and minitaps, respectively; in SCD, they produced large amplitude oVEMP waveforms with n1 peaks at 12.9 ± 0.8 ms (tendon-hammer) and 12.1 ± 0.5 ms (minitap). Occiput stimulation evoked oVEMPs with similar n1 latencies in both groups (tendon-hammer = 11.3 ± 1.3 and 10.7 ± 0.8; minitap = 10.3 ± 0.9 and 11.1 ± 0.4 for control and SCD ears, respectively). Compared to reflex amplitudes, n1 peak latencies to Fz taps provided clearer separation between SCD and control ears. The distinctly different effects of Fz and vertex taps on the oVEMP waveforms may represent an additional non-osseous mechanism of stimulus transmission in SCD. For skull taps at Fz, a prolonged n1 latency is an indicator of SCD.
R. L. Taylor · A. P. Bom · M. S. Welgampola (*) Institute of Clinical Neurosciences, Royal Prince Alfred Hospital, Central Clinical School, University of Sydney, Sydney, NSW, Australia e-mail: [email protected]
Introduction
C. Blaivie European Institute of Otorhinolaryngology, Antwerp, Belgium B. Holmeslet Department of Otolaryngology, St Olavs Hospital, Trondheim University Hospital, Trondheim, Norway T. Pansell Section for Ophthalmology and Vision, Department for Clinical Neurosciences, Karolinska Institutet, Stockholm, Sweden K. Brantberg Department of Audiology, Karolinska Hospital, Stockholm, Sweden
Keywords Superior canal dehiscence · Bone-conducted stimulation · Ocular vestibular-evoked myogenic potentials
Vertigo and imbalance evoked by loud sound or pressure may suggest an absence of bone overlying the superior semi-circular canal, known as superior canal dehiscence syndrome (SCD) (Minor et al. 1998). The defect introduces an additional low impedance pathway or ‘third mobile window’ through the inner ear, leading to an alteration in cochlear and vestibular responses to air- and bone-conducted stimuli. Typically in SCD, vestibular-evoked myogenic potentials (VEMP) to air-conducted (AC) sound have lowered reflex thresholds and large amplitudes. Ocular vestibular-evoked myogenic potentials (oVEMP) can be recorded from infraorbital surface electrodes in response to air- and bone-conducted (AC and BC) stimuli (Todd et al. 2007, 2008; Iwasaki et al. 2007). The typical
13
oVEMP waveform evoked by AC sound and forehead taps at Fz is characterised by an initial negative (excitatory) potential at around 10 ms, followed by a positive peak approximately 5 ms later. This surface waveform represents the synchronous modulation in activity of the inferior oblique muscles (Weber et al. 2012), mainly via a crossed pathway from the contralateral otolith organs (Iwasaki et al. 2008; Curthoys and Vulovic 2011). Stimulation by AC sound depends on transmission through the ossicular chain and the generation of pressure variations in the labyrinthine fluids. oVEMP responses to bone-conducted (BC) stimuli are presumed to originate through a combination of elastic wave propagation around the skull (dominant at high frequencies of stimulation) and whole-head acceleration (for low frequencies below ≈150–400 Hz) (Todd et al. 2008; Jombik et al. 2011). Other possible contributors include the inertia of the middle ear ossicles, acoustic radiation, and pressure transmission through the cerebrospinal fluids (Stenfelt and Goode 2005; Stenfelt 2011). However, the relevance of these alternate BC mechanisms to the generation of tap oVEMPs in health and disease has not been determined. The recording of AC oVEMP amplitudes has emerged as a powerful clinical tool in the diagnosis of SCD. Whereas the distribution of AC cVEMP amplitudes in SCD overlaps with those of healthy controls, AC oVEMP amplitudes provide almost complete separation (100 % sensitivity and specificity) on comparison with age-matched normative data (Taylor et al. 2012; Janky et al. 2013). However, not all clinicians are in favour of using AC stimuli to elicit oVEMPs. Concerns regarding their use were recently aired in an article by Manzari (2013), who highlighted several limitations including low response rates in controls and sensitivity to middle ear pathology as reasons for adopting BC stimuli in clinical oVEMP testing. Unlike AC stimuli, BC taps applied at the hairline in the midline (Fz) are independent of the middle ear system and have been shown to generate robust oVEMP waveforms in healthy controls (Iwasaki et al. 2007, 2008; Rosengren et al. 2011). However, compared to AC oVEMP there are few reports of BC oVEMPs as a test for SCD. Manzari et al. (2012) reported good sensitivity and specificity for oVEMP amplitudes recorded in response to a 500 Hz BC stimulus. Others have shown overlap in the amplitude distributions of SCD patients and controls, implying BC oVEMP amplitudes to be a less-sensitive marker for dehiscence (Welgampola et al. 2008; Janky et al. 2013). The optimal stimulus location for eliciting BC oVEMPs in SCD is also unknown. In healthy controls, changes in oVEMP morphology, amplitudes and latencies are evident as the stimulus is shifted to different areas around the skull (Todd et al. 2008; Lin et al. 2010; Holmeslet et al. 2011; Jombik et al. 2011). Whether similar changes occur in the presence of an SCD has not been confirmed.
13
Exp Brain Res
The objectives of this study were twofold. First, we hypothesised that the presence of a dehiscence may alter the mechanism of BC transmission to the labyrinth. Differences in the pattern of end organ activation could result in identifiable changes in oVEMP morphology in response to stimuli delivered at different midline locations. Additionally, given the potential for widespread future use of BC stimuli in routine oVEMP testing, we felt that differences in amplitudes, latencies and waveform characteristics could be exploited to develop diagnostic markers that separate SCD from control ears.
Materials and methods Subjects oVEMPs were recorded from sixteen controls (aged 36.2 ± 9.8) and eighteen patients (24 ears) with SCD (aged 55.9 ± 12.8). The diagnosis of SCD was confirmed by 3D temporal bone CT scans with 0.3 mm slices and reconstructions in all six canal planes. Patients (23/24 ears) also demonstrated either augmented oVEMP amplitudes (n = 19 ears diagnosed after 2007) or reduced cVEMP thresholds (n = 4 SCD ears diagnosed prior to 2008) in response to AC click stimuli (based on laboratory normal data). One patient with a negative result on AC VEMP testing had previously undergone stapes surgery. This may have compromised middle ear transmission, as evidenced by a maximum conductive hearing loss across the entire frequency range. This patient experienced non-lateralising symptoms of autophony and conductive hyperacusis and had clear evidence of bilateral dehiscences on imaging. Participants gave written informed consent and were studied in accordance with the guidelines of the Helsinki Declaration. Stimuli The stimuli were skull taps, delivered either mechanically using an automated hand-held ‘minishaker’ (model 4810, Bruel and Kjaer) or manually via a triggered tendon-hammer. These stimuli have been shown to produce more reliable oVEMP responses than those produced by a conventional B71 bone-conductor and are widely used clinically (Rosengren et al. 2011). Taps were delivered at three midline locations (Fig. 1) in accordance with the international 10–20 system of electrode placement. The minishaker taps, referred to as ‘minitaps’, were generated in response to a 1 ms square-wave pulse (condensation polarity, 20 V amplitude). Calibration of the minishaker, using a Bruel and Kjaer sound level meter and artificial mastoid (Bruel and Kjaer 4930), confirmed an intensity of 147 dB Force level (24 N). For manually driven (tendon-hammer) taps,
Exp Brain Res Cz
Cz
Fz Fz
Oz
-
+
-
+ Accelerometer
Ground
Fig. 1 Illustration of the two-channel electrode montage, mid-sagittal stimulation sites and accelerometry convention used in the present study. Tendon-hammer taps and minitaps were delivered normal to the skull surface at the locations indicated. Based on the international 10–20 system of electrode placement; Fz represents stimulation at the hairline, Cz corresponds to the vertex and Oz to the occiput. An example of the accelerometer positioning is shown for left mastoid
a voltage output greater than 2.5 V was required to activate the trigger pulse. This corresponded to a peak deceleration >8.0 g on delivery of the stimulus to the skull surface. 3D accelerometry The frequency and directional properties of the stimulusskull response were defined in 3D for six controls and two SCD patients using triaxial accelerometers (TMS international), which were secured over each mastoid process using adhesive tape and a tight elastic headband. The averaged results for the six controls are shown in Fig. 2. For all SCD and control ears, stimuli delivered at Fz and the vertex caused the mastoids to first accelerate outward (lateral, Y-axis), downward (Z-axis) and backward (X-axis). The direction reversed upon stimulation at the occiput. Acceleration responses for left and right mastoid recordings were averaged and are compared against SCD ears in “Appendix”. For Fz stimulation, the mean magnitude of acceleration was greatest in the X-axis for both stimuli. Stimulation at the occiput also produced the largest acceleration magnitude in the X-axis for the tendon-hammer, while for minitaps the Y component was largest. For vertex tendon-hammer taps, acceleration was greatest in the Z-axis; for minitaps, Z and Y components were of equal magnitude.
recordings. Three-dimensional acceleration for the left (aL) and right (aR) mastoid is defined inthe naso-occipital (X),inter-aural (Y) and axR (t) axL (t) R L R L vertical (Z) axes a (t) = ay (t) , a (t) = ay (t) azR (t) azL (t)
Fourier analysis, performed on a 10 ms Hanning window from stimulus onset, confirmed a consistent low frequency (peak
