Multichannel cochlear nucleus stimulation

Otolaryngology– Head and Neck Surgery SEPTEMBER 1999

VOLUME 121

NUMBER 3

ORIGINAL ARTICLES Multichannel cochlear nucleus stimulation HUSSAM K. EL-KASHLAN, MD, Ann Arbor, Michigan

Stimulation of the cochlear nucleus (CN) has been used on a limited basis for rehabilitation of a select group of patients with bilateral acoustic neuromas. These patients were implanted with an electrode placed on the surface of the CN after resection of their tumors. Animal studies have demonstrated greater efficiency of a penetrating CN electrode in activating the central auditory system than a surface electrode. The objective of this work was to study the electrically evoked middle latency response generated by stimulation through a penetrating multichannel CN electrode in an animal model. Six pigmented guinea pigs underwent implantation with a penetrating multichannel CN electrode. Threshold, latency, and input-output functions of electrically evoked middle latency responses with different stimulation pads were studied. There were systematic differences in the latency and amplitude of the input-output functions depending on the site of stimulation within the CN. The results support the hypothesis that discrete activation of neuronal subpopulations within the CN is possible with a penetrating multichannel microelectrode. (Otolaryngol Head Neck Surg 1999;121: 169-75.)

From the Department of Otolaryngology, University of Michigan. Supported by a Deafness Research Foundation grant. Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, New Orleans, LA, September 17-20, 1995. Reprint requests: Hussam K. El-Kashlan, MD, Department of Otolaryngology, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0312. Copyright © 1999 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/99/$8.00 + 0 23/1/93869

M

ost patients with profound hearing loss can be rehabilitated by cochlear implants placed in the scala tympani. However, this approach to auditory rehabilitation is not applicable to a significant number of patients, most of whom have bilateral acoustic neuromas requiring surgical removal. This intervention usually precludes the use of standard cochlear implants unless a functional auditory nerve can be preserved. Furthermore, other deafened individuals with auditory nerve degeneration caused by congenital, traumatic, ototoxic, and inflammatory events have been identified. These patients are likewise unable to benefit from the use of a standard cochlear implant. Efforts at the House Ear Institute led to the development of prostheses that bypass the peripheral auditory pathway and directly stimulate the cochlear nucleus (CN) through a surface electrode. Several reports have documented the usefulness of prostheses implanted on the surface of the CN in the auditory rehabilitation of this select group of patients after bilateral acoustic neuroma removal.1-8 Implants that penetrate the CN surface may offer potential benefits by achieving closer approximation to neurons subserving the entire frequency range. To date, a penetrating CN electrode has been studied only in animal models.9-15 The middle latency response (MLR) is an auditoryevoked response that occurs between 10 and 50 ms after stimulation. In the guinea pig, it is a triphasic waveform, with the components designated A, B, and C.16 It is thought to represent activation of higher auditory centers from the CN up to the auditory cortex.17 The MLR has been widely used in acoustic studies, as well as studies involving electrical stimulation of the cochlea.18,19 The electrically evoked MLR (EMLR) has the same shape as the acoustic MLR but with slightly shorter 169

170

EL-KASHLAN

Otolaryngology– Head and Neck Surgery September 1999

Fig 1. Comparison of acoustic MLR and EMLR in the guinea pig. Waveforms demonstrate the similar shape of both responses with shorter latency for EMLR.

latencies because the peripheral auditory pathways are bypassed (Fig 1). The objective of this work was to characterize the EMLR generated by stimulation through a penetrating multichannel CN electrode and to determine the effect of using different combinations of electrode pairs on the characteristics of the electrophysiologic response. METHODS Six pigmented guinea pigs were used in this study. The study protocol was approved by the institutional committee on use and care of animals. Humane care principles were followed during handling of all animals, and NIH guidelines were adhered to for the use of laboratory animals. Animals were anesthetized with ketamine (40 mg/kg) and xylazine (8 mg/kg). A head screw was placed to restrain the animals. Epidural recording screws, including an active electrode, a ground electrode, and a reference electrode, were positioned, respectively, 1 cm lateral to the bregma over the temporal cortex contralateral to the CN prosthesis, 1 cm lateral to the bregma on the ipsilateral side, and in the midline, 2 cm anterior to the bregma. A posterior occipital craniotomy was then performed with a cutting burr. The dura was incised, and the flocculus of the cerebellum was exposed and partially removed to expose the CN on the dorsolateral brain stem. The electrode was mounted on a micromanipulator and lowered into the CN under direct microscopic visualization until all stimulation pads were within the CN. Figure 2 shows a schema of the experimental setup. These experiments used multichannel microelectrode arrays that were batch fabricated with silicon micromachining and thin-film technology.20,21 The electrodes consisted of a

Fig 2. Schema of experimental setup. Micromanipulator allows accurate placement of the electrode in the brain stem. Evoked responses are amplified, signal averaged, and stored on the computer.

silicon substrate that supported an array of thin film conductors. The silicon substrate was completely insulated except for openings in the upper layer that defined the active electrode sites on the distal aspect of the probe. The total size of each electrode was 2.5 mm long, 15 µm thick, and 80 µm wide, and each contained 5 stimulating pads. Each pad size was 2000 µm2, with a center-to-center spacing of 200 µm. The silicon substrate electrode was bonded to a printed circuit board that acted as both a carrier and an interconnection to the stimulating apparatus and allowed bipolar stimulation between any 2 of the 5 active sites. The circuit board was attached to a micromanipulator for placement into the CN. Figure 3 shows a schema of the electrode and a photograph of the electrode on a penny to demonstrate its relative size. The stimuli used in this study were biphasic charge-balanced constant-current pulses with a total duration of 400 µs and an amplitude varying from 5 to 200 µA. The repetition rate used for evoked potential recording was 5/second. The electrically evoked responses were amplified and filtered with a bandpass of 3 to 3000 Hz and signal averaged over 256 presentations for each run. Input-output functions were determined for all possible stimulating pads’ combination pairs. These were stored on a computer and analyzed for threshold, latency, and amplitude characteristics. The software used was developed at our institution.

Otolaryngology– Head and Neck Surgery Volume 121 Number 3

EL-KASHLAN

171

A

B Fig 3. A, Schema of multichannel electrode with 5 stimulation pads. Pad closest to the tip is designated pad 1, and on implantation, it is located in the ventral portion of the CN. B, Electrode on a penny.

Table 1. Mean EMLR thresholds with different combinations of stimulation pads Pad combination

Threshold (µA)

1-2

1-3

1-4

1-5

2-3

3-4

11 ± 3.94

9.5 ± 2.84

9.5 ± 2.84

0.5 ± 3.69

10 ± 4.38

11.25 ± 4.43

RESULTS

There was no significant difference in mean EMLR thresholds between different combinations of stimulation pads (Table 1). Figures 4 and 5 demonstrate the mean latency of MLR waves A and B, respectively,

4-5

10.5 ± 3.69

elicited by progressively wider separation between stimulating pads at different stimulus intensities. A systematic and consistent pattern was observed. The widest pad separation of 800 µm elicited the shortest latency for both waves A and B. The narrowest pad separation

172

Otolaryngology– Head and Neck Surgery September 1999

EL-KASHLAN

Fig 4. Mean latency for MLR wave A with different stimulation pad separation.

Fig 5. Mean latency for MLR wave B with different stimulation pad separation.

of 200 µm elicited the second shortest latency for both waves. Pad separation of 600 µm elicited the longest latency for both waves. This pattern was consistent among all animals included in the study. These differences in latency were statistically significant for both waves A and B at all stimulus intensities by use of 1way analysis of variance. Figures 6 and 7 demonstrate the mean latency of MLR waves A and B, respectively, with stimulation between different pairs of adjacent pads at different stimulus intensities. A systematic and consistent pattern was also observed. The distal-most pair of pads (1-2) located in the ventral part of the CN elicited the shortest latency for both waves A and B. The proximal-most pair (4-5) located in the dorsal part of the CN elicited the second shortest latency for both waves. Stimulation between pads (3-4) elicited the longest latency for both waves. This pattern was consistent among all animals in the study. Again, the differences were statistically significant for both waves at all stimulus intensities except for wave B at 50 µA. This may be partly because of

Fig 6. Mean latency for MLR wave A with different adjacent stimulation pads.

greater variability of data at lower stimulus intensities and partly because of the small number of animals. At higher stimulus intensities, recruitment of all available neurons may lead to decreased variability in the evoked response. Figure 8 demonstrates the mean MLR input-output functions elicited by wider separation between stimulating pads. The narrowest pad separation (200 µm) elicited the smallest response amplitude. The widest pad separation (800 µm) elicited an intermediate response amplitude, whereas intermediate pad separation (400 and 600 µm) elicited the steepest input-output function. This was consistent among all animals. Figure 9 demonstrates the mean MLR input-output functions elicited by stimulation between different pairs of adjacent pads. Similarly, a consistent pattern was observed. The distal-most pair of pads (1-2) elicited the smallest amplitude, the proximal-most pair (4-5) elicited an intermediate amplitude, and one of the middle pairs (2-3) elicited the largest response amplitude. DISCUSSION

Currently, rehabilitation of patients with profound hearing loss caused by surgical removal of bilateral acoustic neuromas involves implanting an electrode array on the surface of the CN at the time of tumor removal. This results in auditory perception that is comparable with that of single-channel cochlear implants.1-8 We may assume that with technical developments the clinical benefits of these brain stem auditory implants may increase, as they have with conventional cochlear implants. One potential technical improvement might be the development of penetrating multichannel electrodes. A penetrating electrode array offers many potential advantages over a surface electrode. By virtue of the proximity of a penetrating electrode to the target neuronal populations within the CN, as well as its accessi-

Otolaryngology– Head and Neck Surgery Volume 121 Number 3

Fig 7. Mean latency for MLR wave B with different adjacent stimulation pads.

bility to a larger neuronal population, the penetrating electrode has the potential to activate the central auditory system at a lower threshold than a surface electrode. Consequently, this may yield a lower operating current and a wider dynamic range. A penetrating CN electrode also has potential access to the tonotopic gradients and functionally distinct subunits within the CN. This, in turn, gives rise to the potential of using a multichannel electrode configuration that may, like cochlear implants, provide better pitch discrimination and speech perception. A previous study compared surface versus penetrating electrical stimulation of the CN in an animal model with electrophysiologic and autoradiographic techniques.14 The results indicated that at the level of the CN, a penetrating prosthesis was capable of activating the auditory system at a lower threshold and with a wider dynamic range than a surface prosthesis. Also, the penetrating prosthesis was more effective in activating the higher auditory centers. One issue that is fundamental to the development of next-generation implants is the ability of a penetrating CN electrode to selectively activate discrete portions of the central auditory pathway. Previous electrophysiologic22 and autoradiographic23-29 studies examined the tonotopic organization of the peripheral and central auditory pathways and the spatial projection patterns from peripheral to central auditory structures. Ryan et al29 examined the patterns of central auditory system activation produced by monopolar and bipolar cochlear prostheses under varying conditions of stimulus current and electrode position and compared the results to activity evoked by auditory stimulation. They demonstrated that pure-tone acoustic stimulation resulted in activation of spatially restricted regions within the central auditory pathways and that the location of the activated region within each

EL-KASHLAN

173

Fig 8. Mean MLR input-output functions with different stimulation pad separation.

Fig 9. Mean MLR input-output functions with different adjacent stimulation pads.

nucleus varied systematically with the frequency of stimulus. They also demonstrated that a similar restricted pattern of activation can be produced by electrical stimulation of different frequency regions within the cochlea, from base to apex, with a bipolar electrode configuration. Several neuroanatomic30-34 and functional35 studies examined the ascending pathways and projections from the CN and found evidence of functional specialization. This may permit selective stimulation of functionally distinct neuronal subpopulations in the CN in a fashion similar to stimulation of different frequency regions in the cochlea with multichannel cochlear implants. A previous study examined the patterns of 2deoxyglucose uptake as a measure of the metabolic response in auditory pathways evoked by stimulation of discrete regions within the CN.12 A single-channel penetrating bipolar electrode was implanted in either the anterior or the posterior part of the CN. Qualitative differences were observed in the pattern of the evoked response between the 2 electrode-placement sites.

174

Otolaryngology– Head and Neck Surgery September 1999

EL-KASHLAN

These findings suggested that spatially distinct regions of the central auditory tract may be differentially activated depending on the site of stimulation by a penetrating CN prosthesis.12 The demonstration of consistent and orderly differences in the characteristics of the EMLR with different stimulation pads on a penetrating multichannel CN implant in this study further supports the hypothesis of feasibility of discrete activation of neuronal subpopulations with the CN. A single report in the literature studied the current levels required to obtain behavioral thresholds in human beings with a penetrating CN electrode.1 Several other reports provide information on thresholds of patients receiving surface CN stimulation.3,4,7,8 The data from these studies indicate that the current levels required to obtain a behavioral threshold by a penetrating electrode are much lower than those required by a surface electrode. This difference was consistently present across all frequencies tested. Given the wide variation observed in individual patient performance with auditory brain stem prostheses, the retrospective nature of this comparison, the uncontrolled patient groups, and the fact that only 1 patient with the penetrating electrode configuration was tested, conclusions based on these data should be reached cautiously. However, these are promising and correspond well with data generated in animal models. A more recent study by Shannon36 compared psychophysical data on temporal resolution from listeners with normal hearing, listeners with cochlear implants, and patients electrically stimulated on the CN. The study demonstrated a relatively similar temporal processing across the 3 groups once the obvious differences in the dynamic range were taken into consideration. The results imply that central auditory systems of patients with cochlear implants are able to fully use the nonnatural patterns of temporal neural information produced by electrical stimulation. Given the potential morbidity associated with these devices (eg, insertion trauma, potential for migration, unknown long-term effects of brain stem stimulation), additional animal studies are needed to provide an adequate basis to help guide clinical advances. CONCLUSION

This study supports the hypothesis that discrete activation of neuronal subpopulations within the CN is possible with a multichannel penetrating microelectrode. Studies are currently under way with double-labeled autoradiographic techniques to further characterize the patterns of central auditory system activation with penetrating multichannel CN electrodes. This would enhance the development of prostheses and coding strategies that might provide multiple channels of infor-

mation to improve the potential benefits from a CN implant. The author thanks the University of Michigan Center for Neural Communication Technology (sponsored by NIH/ NCRR grant P41-RR09754-04) for providing the electrodes used in this study. REFERENCES 1. Edgerton B, House W, Hitselberger W. Hearing by cochlear nucleus stimulation in humans. Ann Otol Rhinol Laryngol 1982;(Suppl)91:117-24. 2. Hitselberger WE, House WF, Edgerton BJ, et al. Cochlear nucleus implant. Otolaryngol Head Neck Surg 1984;92:52-4. 3. McElveen JT, Hitselberger WE, House WF, et al. Electrical stimulation of cochlear nucleus in man. Am J Otol 1985;(Suppl)6:8891. 4. Eisenberg LS, Maltan AA, Portillo F, et al. Electrical stimulation of the auditory brainstem structure in deafened adults. J Rehabil Res Dev 1987;24:9-22. 5. Otto SR, House WF, Brackmann DE, et al. Auditory brainstem implant: effect of tumor size and preoperative hearing level on function. Ann Otol Rhinol Laryngol 1990;99:789-90. 6. Brackmann DE, Hitselberger WE, Nelson RA, et al. Auditory brainstem implant: issues in surgical implantation. Otolaryngol Head Neck Surg 1993;108:624-33. 7. Shannon RV, Fayad J, Moore J, et al. Auditory brainstem implant: II. Postsurgical issues and performance. Otolaryngol Head Neck Surg 1993;108:634-42. 8. Shannon RV. Threshold functions for electrical stimulation of the human cochlear nucleus. Hear Res 1989;40:173-8. 9. Niparko J, Altschuler R, Xue X, et al. Surgical implantation and biocompatibility of CNS auditory prosthesis. Ann Otol Rhinol Laryngol 1989;98:965-70. 10. Niparko J, Altschuler R, Evans D, et al. Auditory brainstem prosthesis: biocompatibility of stimulation. Otolaryngol Head Neck Surg 1989;101:344-52. 11. Evans DA, Niparko JK, Altschuler RA, et al. Demonstration of prosthetic activation of central auditory pathways using [14C]-2deoxyglucose. Laryngoscope 1990;100:128-37. 12. El-Kashlan HK, Niparko JK, Kileny PR, et al. Direct electrical stimulation of the cochlear nucleus: autoradiographic patterns of central auditory pathway activation [abstract]. Otolaryngol Head Neck Surg 1990;103:189. 13. El-Kashlan HK, Niparko JK, Altschuler RA, et al. Prosthetic stimulation of the cochlear nucleus: dynamic range of action. ARO Abstracts 1991;14:158. 14. El-Kashlan HK, Niparko JK, Altschuler RA, et al. Direct electrical stimulation of the cochlear nucleus: surface versus penetrating stimulation. Otolaryngol Head Neck Surg 1991;105:533-43. 15. El-Kashlan HK, Noorily AD, Niparko JK, et al. Metabolic activity of central auditory structures following prolonged deafferentation. Laryngoscope 1993;103:399-405. 16. Kraus N, Smith I, McGee T. Rate and filter effects on the developing middle latency response. Audiology 1987;26:257-68. 17. Kaga K, Hink RF, Shinoda Y, et al. Evidence for a primary cortical origin of a middle latency auditory evoked potential in cats. Electroencephalogr Clin Neurophysiol 1980;50:254-66. 18. Kileny PR, Kemink JL. Electrically evoked middle-latency auditory potentials in cochlear implant candidates. Arch Otolaryngol Head Neck Surg 1987;113:1072-7. 19. Burton MJ, Miller JM, Kileny PR. Middle latency responses: I. Electrical and acoustic excitation. Arch Otolaryngol Head Neck Surg 1989;115:59-62. 20. Stimulating electrodes based on thin-film technology. Quarterly report #9. Contract NIH-NINDS-N01-NS-9-2359. Michigan Center for Neural Communication Technology; 1992. 21. Hetke JF, Lund JL, Najafi K, et al. Silicon ribbon cables for

Otolaryngology– Head and Neck Surgery Volume 121 Number 3

22.

23. 24. 25. 26. 27.

28.

chronically implantable microelectrode arrays. IEEE Trans Biomed Eng 1994;41:314-21. Rose JE, Greenwood DD, Goldberg JM, et al. Some discharge characteristics of single neurons in the inferior colliculus of the cat. I. Tonotopical organization, relation to spike-counts to tone intensity, and firing patterns of single elements. J Neurophysiol 1963;26:294-320. Ryan AF, Woolf NK, Sharp FR. Tonotopic organization in the central auditory pathway of the Mongolian gerbil: a 2-deoxyglucose study. J Comp Neurol 1982;207:369-80. Ryan AF, Woolf NK. Development of tonotopic representation in the central auditory system of the Mongolian gerbil: a 2deoxyglucose study. Dev Brain Res 1988;469:61-70. Ryan AF, Braverman S, Woolf NK, et al. Auditory neural activity evoked by pure-tone stimulation as a function of intensity. Brain Res 1988;483:293-302. Huang C, Fex J. Tonotopic organization in the inferior colliculus of the rat demonstrated with the 2-deoxyglucose method. Exp Brain Res 1986;61:506-12. Webster WR, Serviere J, Batini C, et al. Autoradiographic demonstration with 2-(14C)-deoxyglucose of frequency selectivity in the auditory system of cats under conditions of functional activity. Neurosci Lett 1978;10:43-8. Serviere J, Webster WR, Calford MB. Isofrequency labeling

EL-KASHLAN

29. 30. 31. 32. 33. 34. 35. 36.

175

revealed by a combined [14C]-2-deoxyglucose, electrophysiological, and horseradish peroxidase study of the inferior colliculus in the cat. J Comp Neurol 1984;228:463-77. Ryan AF, Miller JM, Wang Z, et al. Spatial distribution of neural activity evoked by electrical stimulation of the cochlea. Hear Res 1990;50:57-70. Warr WB. Fiber degeneration following lesions in the multipolar and globular cell areas in the ventral cochlear nucleus of the cat. Brain Res 1972;40:247-70. Warr WB. Fiber degeneration following lesions in the posteroventral cochlear nucleus of the cat. Exp Neurol 1969;23:14055. Warr WB. Fiber degeneration following lesions in the anterior ventral cochlear nucleus of the cat. Exp Neurol 1966;14:453-74. Osen KK. Projection of the cochlear nuclei on the inferior colliculus in the cat. J Comp Neurol 1972;144:355-72. Bourk TR, Mielcarz JP, Norris BE. Tonotopic organization of the anteroventral cochlear nucleus of the cat. Hear Res 1981;4:215-41. Frederickson CJ, Gerken GM. Functional characteristics of cochlear nucleus in behaving cat examined by acoustic masking of electrical stimuli. J Neurophysiol 1978;41:1535-45. Shannon RV. Quantitative comparison of electrically and acoustically evoked auditory perception: implications for the location of perceptual mechanisms. Prog Brain Res 1993;97:261-9.

Your comprehensive reference to valuable research information...

O

tolaryngology–Head and Neck Surgery provides its readers with timely information on recent treatments and advances in otolaryngology.

YES!

Please send me the Five-Year Cumulative Index to Otolaryngology–Head and Neck Surgery (1991-1995). I’ve indicated my method of payment below. All prices include shipping. ❏ $45.00 U.S

❏ $52.96 Canada*

❏ $49.50 International

* Includes Canadian GST.

Now you can have easy access to that information with the five-year cumulative author and subject index to Volumes 104-113 of Otolaryngology–Head and Neck Surgery.

Payment must accompany your order. ❏ Check enclosed (U.S. funds, payable to Otolaryngology– Head and Neck Surgery) ❏ MasterCard ❏ VISA ❏ Discover

The Cumulative Index is your comprehensive reference guide, including a list of every article in the journal from 1991 through 1995. All articles are indexed both by subject and by author.

Credit Card # _____________________ Exp. Date _______

The Cumulative Index features:

Name _______________________________________________ Institution ____________________________________________ Address _____________________________________________

•Offset and bold Subject Headings •Subject Entries containing complete article title, author(s), year of publication, volume, and page •Author Entries listing author(s), article title, author-toauthor referral, year of publication, volume, and page Order the Otolaryngology–Head and Neck Surgery Cumulative Index today!

Signature ____________________________________________

City _______________________________ State____________ ZIP ____________________ Country_____________________

☎ To order, just call 1-800-453-4351 Outside the U.S.: 314-453-4351 Fax: 314-432-1158

Or return coupon to: Mosby, Inc. Journal Subscription Services 11830 Westline Industrial Dr. St. Louis, MO 63146 U.S.A.