Cervical vestibular evoked myogenic potential with chirp sounds

Abstract

BACKGROUND:

There are only a few reports concerning cervical vestibular evoked myogenic potential (cVEMP) using chirp sound, and clinical indications/advantages of it are still unclear.

OBJECTIVE:

To compare cVEMP using CE-chirp LS^® with cVEMP using 500 Hz and 1000 Hz tone bursts (TB) and to investigate clinical indications/advantages of CE-chirp LS^® for recording cVEMP.

METHODS:

Sixteen patients with vestibular disorders (2men and 14 women) (18∼62, mean 42.9 years of age) were enrolled in this study. Participants underwent cVEMP testing using 500 Hz and 1000 Hz tone bursts (TB) and CE-chirp LS^®. Response rate of P1-N1, corrected/normalized amplitude of P1-N1, latencies of P1 and N1, asymmetry ratio, and correlation of P1 latency to SLOPE in tuning property test (an index of endolymphatic hydrops) were compared.

RESULTS:

Corrected/normalized amplitude of P1-N1 to CE-chirp LS^® was smaller than corrected/normalized amplitude of P1-N1 to 500 Hz TB. Peak latencies to CE-chirp LS^® were the shortest among the 3 types of stimulation. EH-positive ears according to the tuning property test had tendency of prolonged P1 latencies to CE-chirp LS^®.

CONCLUSION:

CE-chirp LS^® is applicable for recording cVEMP with a similar diagnostic accuracy to TB. Prolongation of P1 latency in CE-chirp LS^® might be an indicator of endolymphatic hydrops in the saccule.

Keywords

Cervical vestibular evoked myogenic potential chirp tuning property endolymphatic hydrops latency Meniere’s disease

1 Introduction

For recording cervical vestibular evoked myogenic potentials (cVEMP), 500 Hz and 1000 Hz air-conducted tone bursts (TB) or clicks have been mainly used [8 , 15]. Among them, 500 Hz TB has been used most frequently because the otolith organs, especially the saccular sensory cells, best respond to sounds around 500–700 Hz [7, 16]. Furthermore, 1000 Hz TB has been used supplementally to assess tuning property change in cVEMP in order to detect endolymphatic hydrops (EH) [10 , 13]. Chirp sound, which can be described as a sound in which frequency varies with time, has been applied in auditory electrophysiological tests such as auditory brainstem response (ABR) and auditory steady state response (ASSR) [1, 3]. While chirp sound has been also applied for recording VEMPs [14, 20], clinical indications/advantages of chirp sound for VEMPs are not clear yet.

As chirp sound has a feature in which frequency varies with time, we hypothesized that tuning property of cVEMP might be reflected in the peak latency of chirp-cVEMP. Herein, we studied chirp-cVEMP with TB-cVEMP in patients with vestibular disorders to clarify its features and advantages with a special interest to association of chirp-cVEMP latencies with EH.

2 Subjects and methods

2.1 Subjects

Sixteen patients with vestibular disorders (2men and 14 women) (18∼62, mean 42.9 years of age) were enrolled in this study. Seven patients had Meniere’s disease (MD) or related disorders (one man and 6 women). Three of the 7 patients were diagnosed with having definite MD and 4 were diagnosed with having delayed endolymphatic hydrops (ipsilateral type) [6, 17]. Seven patients had vestibular migraine (one man and 6 women) [5]. Two had recurrent peripheral vestibulopathy (2 women) [13].

2.2 Methods

Each subject underwent cVEMP testing. Recording was performed using Eclipse system (Interacoustics, Denmark). For recording, surface electrodes were placed on the belly of the sternocleidomastoid muscle (SCM) (active), on the upper end of the sternum (reference), and on the nasion (ground). During the recording, subjects were asked to rotate their necks to the opposite side to the stimulated ear in order to keep contracting the target muscle. During the recording, muscle activity was kept between 20–200μV RMS. CE-chirp LS^® (duration 10 msec, 100dBnHL equivalent to 131.5 dBpeSPL) [4], 500 Hz tone burst (TB) (rise/fall time = 4 msec, plateau time = 2 msec, 100dBnHL equivalent to 123.5 dBpeSPL) and 1000 Hz TB (rise/fall time = 2msec, plateau time = 1msec, 100 dBnHL equivalent to 121.5 dBpeSPL) were presented through the insert-earphones at the random order to obtain cVEMP responses at the rate of 5.1 Hz. CE-Chirp LS^® is the revised version of the standard CE-Chirp^® that minimizes the upward spread of excitation at high intensity levels [2].

Signals were bandpass-filtered (10–750 Hz) and averaged. One hundred responses were averaged for each run. Reproducibility was confirmed by 2 runs. Responses were normalized according to muscle activity.

The initial biphasic positive-negative response was studied as P1-N1. Normalized peak-to-peak amplitude of P1-N1 and peak latencies of P1 and N1 were measured. In addition, asymmetry ratio (AR) and 500–1000 Hz cVEMP slope (SLOPE) were calculated. AR is an index of unilateral vestibular weakness, while SLOPE is an index of the tuning property [12].

AR = 100×(Al-As)/(Al+As) SLOPE = 100×(A500-A1000)/(A500 + A1000)

Al: normalized P1-N1 amplitude (larger side), As: normalized P1-N1 amplitude (smaller side)

A500: normalized P1-N1 amplitude to 500 Hz TB, A1000: normalized P1-N1 amplitude to 1000 Hz TB.

According to the previous studies [12, 13], SLOPE < –19.9 was regarded as EH-positive.

For statistical analyses, chi-square test, Kruskal-Wallis test, u-test, and t-test were used. For multiple comparison, Scheffe’s test was used. P < 0.05 was regarded as significant.

Informed consent was obtained from each subject. This study was approved by the Ethics Committee of Teikyo University (TR17-159).

3 Results

3.1 Response rate of cVEMP responses

P1-N1 s were present in 26 ears (81.2%) to 500 Hz TB, 27 ears (81.2%) to 1000 Hz TB, and 26 ears (81.2%) to CE-chirp LS^® (Fig. 1). There was no significant difference among the 3 types of stimulation concerning the response rate of cVEMP responses (chi-square test, p > 0.05).

Fig. 1

An example of recording (a 33-year-old woman, vestibular migraine, right ear).

3.2 P1-N1 amplitude

Normalized P1-N1 amplitude (mean + SD, median) to each type of stimulation was shown in Table 1 and Fig. 2. Statistically significant differences of the median P1-N1 amplitudes among the 3 types of stimulation were observed (Kruskal-Wallis test, p < 0.001). Multiple comparison revealed that significant difference existed between 500 Hz TB and CE-chirp LS^® (Scheffe’s test, p < 0.001).

Table 1
Summary of amplitude and latencies

Normalized P1-N1 amplitude P1 latency (msec) N1 latency (msec)

Stimulation N (ears) mean + SD, median mean + SD, median mean + SD, median

500 Hz TB 26 0.91 + 0.48, 0.88 18.1 + 1.83, 17.8 28.4 + 2.72, 28.1

1000 Hz TB 26 0.60 + 0.28, 0.50 16.9 + 3.02, 17.3 26.1 + 3.96, 25.7

CE-chirp LS^® 26 0.44 + 0.21, 0.43 14.52 + 2.50, 13.6 23.6 + 2.81, 23.1

Kruskal-Wallis test p < 0.001 p < 0.001 p < 0.001

* *

		Normalized P1-N1 amplitude	P1 latency (msec)	N1 latency (msec)
500 Hz TB	26	0.91 + 0.48, 0.88	18.1 + 1.83, 17.8	28.4 + 2.72, 28.1
1000 Hz TB	26	0.60 + 0.28, 0.50	16.9 + 3.02, 17.3	26.1 + 3.96, 25.7
CE-chirp LS^®	26	0.44 + 0.21, 0.43	14.52 + 2.50, 13.6	23.6 + 2.81, 23.1
Kruskal-Wallis test		p < 0.001	p < 0.001	p < 0.001
		*	**	***

*: Multiple comparison (Scheffe’s test) showed significant difference between 500 Hz TB and CE-chirp LS^® (p < 0.001). **: Multiple comparison (Scheffe’s test) showed significant difference between 500 Hz TB and CE-chirp LS^® (p < 0.001) and between 1000 Hz TB and CE-chirp LS^® (p = 0.017). ***: Multiple comparison (Scheffe’s test) showed significant difference between 500 H Hz TB and 1000 Hz TB (p = 0.019) and between 500 Hz TB and CE-chip LS^® (p < 0.001).

Fig. 2

Box plots of P1-N1 amplitude.

Correlation coefficients of P1-N1 amplitude to one type of stimulation to others were summarized in Table 2. They showed weak or moderate correlations.

Table 2

Correlation coefficients (r) and p-values of P1-N1 amplitude to one type of stimulation to others

	500 Hz TB	1000 Hz TB	CE-chirp LS^®
500 Hz TB		p = 0.023	p = 0.441
1000 Hz TB	r = 0.46		p = 0.010
CE-chirp LS^®	r = 0.16	r = 0.51

Correlations of AR to one type of stimulation to others were summarized in Table 3. They showed moderate or strong correlations. Correlation between ARs in CE-chirp LS^® and 500 Hz TB was the strongest (r = 0.73).

Table 3

Correlation coefficients (r) and p-values of AR in one type of stimulation to others

	500 Hz TB	1000 Hz TB	CE-chirp LS^®
500 Hz TB		p = 0.023	p = 0.0012
1000 Hz TB	r = 0.56		p = 0.0038
CE-chirp LS^®	r = 0.73	r = 0.66

3.3 P1 & N1 latencies

P1 and N1 latencies (mean + SD, median) to each type of stimulation were shown in Table 1 and Fig. 3. Statistically significant differences of the medians among the 3 types of stimulation were observed in both of P1 and N1 (p < 0.001, Kruskal-Wallis test). Multiple comparison revealed that significant differences in P1 were observed between 500 Hz TB and CE-chirp LS^® (p < 0.001, Scheffe’s test) and between 1000 Hz TB and CE-chirp LS^® (p = 0.017, Scheffe’s test). Significant differences in N1 were observed between 500 H Hz TB and 1000 Hz TB (p = 0.019, Scheffe’s test) and between 500 Hz TB and CE-chip LS^® (p < 0.001, Scheffe’s test). In both of P1 and N1, CE-chirp LS^® had the shortest latencies.

Fig. 3

Box plots of P1 and N1 latencies.

Correlation coefficients of P1 and N1 latencies to one type of stimulation to others were summarized in Table 4. P1 latencies showed stronger correlations than N1 latencies

Table 4

Correlation coefficients (r) and p-values of P1 and N1 latencies to one type of stimulation to others

P1
	500 Hz TB	1000 Hz TB	CE-chirp LS^®
500 Hz TB		p < 0.001	p < 0.001
1000 Hz TB	r = 0.77		p < 0.001
CE-chirp LS^®	r = 0.63	r = 0.70
N1
	500 Hz TB	1000 Hz TB	CE-chirp LS^®
500 Hz TB		p = 0.002	p = 0.001
1000 Hz TB	r = 0.59		p = 0.012
CE-chirp LS^®	r = 0.60	r = 0.50

3.4 Association of tuning property (SLOPE) with P1 latency

SLOPE < –19.9, which is suggestive of EH (+) [10, 12], was observed in 4 ears (4 subjects). Twenty-five ears (13 subjects) were EH (–), while 3 ears (3 subjects) had no responses. Diagnoses of the 4 subjects were MD in 3 subjects and ipsilateral type of DEH in one subject.

Correlation coefficients of P1 latency to SLOPE were summarized in Table 5. Significant correlations were observed in CE-chirp LS^® (r = –0.567) and 1000 Hz TB (r = –0.495) (Table 5).

Table 5
Correlation coefficients between SLOPE and P1 latency in each type of stimulation

500 Hz TB 1000 Hz TB CE-chirp LS^®

Correlation coefficient –0.294 –0.495 –0.567

p-value 0.143 0.010 0.0031

	500 Hz TB	1000 Hz TB	CE-chirp LS^®
Correlation coefficient	–0.294	–0.495	–0.567
p-value	0.143	0.010	0.0031

EH (+) ears had significantly longer P1 latency to CE-chirp LS^® than EH (–) ears (p = 0.016, u-test) (Fig. 4). Concerning 500 Hz TB and 1000 Hz TB, there was no significant differences between the 2 groups (p > 0.05, u-test).

Fig. 4

Box plots of P1 latency in EH(+) ears and EH(–) ears.

4 Discussion

Results of cVEMP using CE-chirp LS^® in this study are summarized as follow. (a) Response rate of cVEMP responses to CE-chirp LS^® is almost the same as those to 500 Hz TB and 1000 Hz TB. (b) Amplitude of P1-N1 to CE-chirp LS^® was smaller than amplitude of P1-N1 to 500 Hz TB. Amplitude of P1-N1 to CE-chirp LS^® had moderate correlation to amplitude to 1000 Hz TB while only weak correlation to amplitude to 500 Hz TB. (c) AR in CE-chirp LS^® had a moderate correlation to AR in 500 Hz TB. (d) Peak latencies to CE-chirp LS^® were the shortest among the 3 types of stimulation. Concerning peak latencies, moderate correlations were observed to both of 500 Hz TB and 1000 Hz TB. (e) Subjects who were considered to be EH(+) according to the tuning property test using 500 Hz TB and 1000 Hz TB cVEMP amplitudes had tendency of prolonged P1 latencies to CE-chirp LS^®. This tendency was not observed in 500 Hz TB and 1000 Hz TB.

Nowadays, 500 Hz TB is mainly used for cVEMP recording, because it evokes larger responses than 1000 Hz TB and clicks [8 , 16]. Furthermore, in combination with 1000 Hz TB, tuning property, an index of EH, can be assessed [8 , 13]. As the present study revealed that the correlation of AR in CE-chirp LS^® to AR in 500 Hz TB was strong and the response rate was the same as 500 Hz TB, CE-chirp LS^® can be used as stimulation for cVEMP although P1-N1 amplitude in CE-chirp LS^® was smaller than in 500 Hz TB. Then, what is an advantage of using CE-chirp LS^® cVEMP instead of TB cVEMP or in addition to TB cVEMP? Chirp sound has a feature in which frequency varies with time [1 , 20]. It was developed to get good responses in the field of auditory electrophysiology [1 , 19]. The temporal feature concerning frequency of chirp, especially up-chirp, matches to tonotopicity in the cochlea. However, the saccule, an origin of cVEMP, does not have such tonotopicity. Therefore, delay of presentation of higher frequency components of stimulation could be reflected as delay of cVEMP responses. We speculated that prolongation of P1 latency in EH(+) ears might imply shift of frequency tuning to higher frequency in EH(+) ears which has been demonstrated by cVEMP study using TB [12 , 18]. If so, possibility of assessment of EH using delay of P1 latency could be an advantage of application of chirp sound for cVEMP recording. The best cut-off line of P1 latency of chirp cVEMP for EH detection should be determined in a future study with a larger number of participants.

Several studies concerning chirp cVEMP have been reported. Walther and Cebulla reported that “CW-VEMP-chirp (250–1000 Hz)” (up chirp) showed significantly larger amplitude than click and 500 Hz TB and that it had equal diagnostic accuracy to those [20]. Ozgur et al. reported that healthy subjects had shorter cVEMP latency and smaller cVEMP amplitude to narrow-band-chirp VEMP than 500 Hz TB and click [14]. While shorter cVEMP latencies to chirp sound than to TB have been constantly reported, larger cVEMP amplitudes to chirp sound are still controversial. Different findings might be due to differences of the type of chirp sound. Various types of chirp sound have been reported [3, 4]. Relationship between chirp-cVEMP latency and tuning property has not been studied yet. This paper first reported that prolongation of P1 latency in CE-chirp LS^® cVEMP might be an indicator of EH. This finding should be confirmed in a larger-sized study.

In conclusion, CE-chirp LS^® is applicable for recording cVEMP with an equal diagnostic accuracy to TB concerning response rate and AR. Prolongation of P1 latency in CE-chirp LS^® might be an indicator of endolymphatic hydrops in the saccule. It might be additional clinical indications/advantages of CE-chirp LS^® cVEMP.

Funding

This study was supported by AMED under Grant Number JP19dk0310092.

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author contributions

All authors contributed extensively to the work presented in this paper. All authors collected data. TM wrote the manuscript. MT, YT, and EY reviewed and edited the manuscript.

References

Dau

, Wegner

, Mellert

and Kallmeir

, Auditory brainstem responses with optimized chirp signals compensating basilar-membrane distortion, J Acoust Soc Am 107 (2000), 1530–1540.

Dzulkarnain

A.A.A.

, Ibrahim

S.H.M.N.

, Farah

, Anuar

, Abdullah

, Zam

T.Z.H.T. Z.

, Rahmat

and Ruzai

M.A.M.

, Influence of two-electrode montages on the level-specific (LS) CE-Chirp auditory brainstem response (ABR) at multiple intensity levels, Int J Audiol 56 (2017), 723–732.

Eder

, Schuster

M.E.

, Polterauer

, Neuling

, Hoster

, Hempel

J.M.

and Semmelbauer

, Comparison of ABR and ASSR using NB-chirp stimuli in children with severe and profound hearing loss, Int J Ped Otorhinolaryngol 131 (2020), 109864. doi: 10.1016/j.ijporl.2020.109864

Elberling

, Callo

and Don

, Evaluating auditory brainstem responses to different chirp stimuli at three levels of stimulation, J Acoust Soc Am 128 (2010), 215–223.

Lempert

, Olesen

, Furman

, Waterston

, Seemungal

, Carey

, Bisdorff

, Versino

, Evers

and Newman-Toker

, Vestibular migraine: diagnostic criteria, J Vestib Res 22 (2012), 167–172.

Lopez-Escamez

J.A.

, Carey

, Chung

W.H.

, Goebel

J.A.

, Magnusson

, Mandalà

, Newman-Toker

D.E.

, Strupp

, Suzuki

, Trabalzini

and Bisdorff

, Diagnostic criteria for Meniere’s disease, J Vestib Res 25 (2015), 1–7.

McCue

M.P.

and Guinan

J.J.

Jr, , Spontaneous activity and frequency selectivity of acoustically responsive vestibular afferents in the cat, J Neurophysiol 74 (1995), 1563–1572.

Murofushi

, Clinical application of vestibular evoked myogenic potential (VEMP), Auris Nasus Larynx 43 (2016), 367–376.

Murofushi

and Kaga

, VEMP: its basics and clinical applications, Springer, Tokyo, 2009.

10.

Murofushi

, Komiyama

and Suizu

, Detection of saccular endolymphatic hydrops in Ménière’s disease using a modified glycerol cVEMP test in combination with the tuning property test, Otol Neurotol 37 (2016), 1131–1136.

11.

Murofushi

, Matsuzaki

and Wu

C.H.

, Short tone burst-evoked myogenic potentials on the sternocleidomastoid muscle, Arch Otolaryngol Head Neck Surg 125 (1999), 660–664.

12.

Murofushi

, Tsubota

, Suizu

and Yoshimura

, cVEMP tuning property of patients with recurrent peripheral vestibulopathy: Is it Meniere’s disease without hearing loss?, Clinical Neurophysiol 128 (2017), 2491–2492.

13.

Murofushi

, Tsubota

, Suizu

and Yoshimura

, Is alteration of tuning property in cVEMP specific for Meniere’s disease?, Front Neurol 8 (2017), 193. doi: 10.3389/fneur.2017.00193

14.

Ozgur

, Erdivanlt

O.C.

, Coskun

Z.O.

, Terzi

, Yigit

, Demirci

and Dursun

, Comparison of tone burst, click and chirp stimulation in vestibular evoked myogenic potential testing in healthy people, Int Adv Otol 11 (2015), 33–35.

15.

Papathanasiou

E.S.

, Murofushi

, Akin

F.W.

and Colebatch

J.G.

, International guidelines for the clinical application of cervical vestibular evoked myogenic potentials: An expert consensus report, Clin Neurophysiol 125 (2014), 658–666.

16.

Rauch

S.D.

, Zhou

, Kujawa

S.G.

, Guinan

J.J.

and Herrmann

B.S.

, Vestibular evoked myogenic potentials show altered tuning in patients with Meniere’s disease, Otol Neurotol 25 (2004), 333–338.

17.

Schuknecht

H.F.

, Delayed endolymphatic hydrops, Ann Otol Rhinol Laryngol 87 (1978), 743–748.

18.

T. Seo

, Node

, Yukimasa

and Sakagami

, Furosemide loading vestibular evoked myogenic potential for unilateral Meniere’s disease, Otol Neurotol 24 (2003), 283–288.

19.

Shore

S.E.

and Nuttall

A.L.

, “High-synchrony cochlear compound action potentials evoked by rising-frequency-swept tone bursts,”, J Acoust Soc Am 78 (1985), 1286–1295.

20.

Walther

L.E.

and Cebulla

, Band limited chirp stimulation in vestibular evoked myogenic potentials, Eur Arch Otorhinolaryngol 273 (2016), 2983–2991.

Cervical vestibular evoked myogenic potential with chirp sounds

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1 Introduction

2 Subjects and methods

2.1 Subjects

2.2 Methods

3 Results

3.1 Response rate of cVEMP responses

Table 5 Correlation coefficients between SLOPE and P1 latency in each type of stimulation 500 Hz TB 1000 Hz TB CE-chirp LS® Correlation coefficient –0.294 –0.495 –0.567 p-value 0.143 0.010 0.0031

Funding

Conflict of interest statement

Author contributions

References

Table 5
Correlation coefficients between SLOPE and P1 latency in each type of stimulation

500 Hz TB 1000 Hz TB CE-chirp LS^®

Correlation coefficient –0.294 –0.495 –0.567

p-value 0.143 0.010 0.0031