40 BARROW QUARTERLY • Vol. 20, No. 4 • 2004

Auditory Brainstem Implantation

Since 1979 auditory brainstem im-plants (ABIs) have provided pros-

thetic hearing for patients deafened byneurofibromatosis-2 (NF-2). The firstABI was a comparatively primitive, sin-gle-channel device that aided users byincreasing awareness of environmentalsounds and lipreading. Contemporarymultichannel ABIs afford a level of per-formance comparable to that achievedwith single-channel cochlear implants,that is, the ability to discriminate envi-ronmental sounds, marked improve-ment in lipreading, and limited open-set speech comprehension. Some ABIusers can even communicate on thetelephone. This article reviews the de-velopment of ABIs, emphasizing rele-vant neuroanatomical and physiologicalprinciples, intraoperative technique, re-lated surgical anatomy, clinical out-comes, and future directions for re-search.

History of ABIs At the time of this writing, more

than 400 ABIs have been placed world-wide.17 The first ABI implant was a sim-ple ball electrode implanted into the pa-renchyma of the cochlear nucleus at theHouse Ear Institute in 1979. This elec-trode migrated from its implantationsite, and the design was soon modifiedinto a pair of platinum plates with a syn-thetic mesh backing. Subsequent iter-ations of ABI design culminated in thedevelopment of multichannel implantscomposed of 8 to 21 platinum discs em-bedded in plastic with a synthetic meshbacking, depending on the manufac-turer or model.

In 1994 a clinical trial was begun in

Auditory brainstem implants (ABIs) can restore meaningful hearing to patientsdeafened by neurofibromatosis type 2. These devices are similar to cochlearimplants on which their design is based. Current multichannel ABI technologyaffords a level of performance in aiding speech comprehension similar to that of-fered by single-channel cochlear implants. ABIs differ from cochlear implants inthat the brainstem implants stimulate the cochlear nuclei directly, bypassing thecochlear nerve. Multichannel ABIs can be implanted at the time of first or secondside surgery for tumors involving the auditory nerve. Most surgeons favor a trans-labyrinthine approach for ABI placement, because the stimulating leads need tobe implanted within the lateral recess of the fourth ventricle. At this location, theventral cochlear nucleus can be stimulated along its lateral to medial axis. Intra-operative monitoring of brainstem auditory evoked potentials is essential andhelps confirm adequate positioning of the device. After programming, about 85%of patients experience auditory percepts. Only a relatively rare, but significant, mi-nority of patients is able to attain sound-only open-set speech comprehension withtheir ABI. Most experience a significant improvement in open-set speech com-prehension when used as an adjunct to lipreading. Speech comprehension con-tinues to increase as long as 8 years after device implantation. More than 90%of recipients use their ABI daily. Most are satisfied with their decision to obtain anABI and would recommend the device to others.

Key Words: acoustic neuroma, auditory brainstem implantation, auditoryprosthesis, neurofibromatosis type 2, vestibular schwannoma

Division of Neurological Surgery and †Section of Neurotology, Barrow Neurological Institute, St.Joseph’s Hospital and Medical Center, Phoenix, Arizona

Gregory P. Lekovic, MD, PhD, JD

L. Fernando Gonzalez, MD

Mark J. Syms, MD†

C. Phillip Daspit, MD†

Randall W. Porter, MD

Copyright © 2004, Barrow Neurological Institute

the United States to evaluate the safetyand efficacy of multichannel ABIs. Thistrial was concluded in 2000 and culmi-nated in the Food and Drug Adminis-tration (FDA) approving the Nucleus24 Contour (Cochlear Corporation,Englewood, CO) and Nucleus 22 (Co-chlear Corporation, Englewood, CO)for implantation via the translabyrin-thine approach for the treatment ofdeafness related to NF-2. Currently, theNucleus 24 is the most widely implant-ed ABI in the world, including NorthAmerica, Australia, and the EuropeanUnion. Other available ABIs include theMXM Digisonic (Laboratories MXM,Valaurious Cedex, France), Med-EL(Med EL Corporation, Research Tri-angle Park, NC), and Clarion (Ad-vanced Bionics, Sylmar, CA). So far noevidence suggests that the efficacy of thevarious multichannel ABIs differs.

Design of ABIsAll ABIs share certain features re-

gardless of specifications (Fig. 1). Thebasic electrode itself is placed within thelateral recess of the fourth ventricle andconnected via a cable implanted into agroove in the temporal bone to a re-ceiver-stimulator. The latter is seated inthe bone of the skull behind the helix of

the ear above the canthomeatal line. Asecond component consists of micro-phone, speech processor, and radio-transmitter coil that detects sound, con-verts it to a digital signal, and transmitsit to the receiver, respectively. Thetransmitter attaches to the receiver-stim-ulator by a magnet. Once an ABI is im-planted, monopolar cauterization isstrictly forbidden from use for fear ofdamaging the electrode, brain, or both.Recently, the magnet has been replacedwith a nonmagnetic plug at surgery, andthe transmitter coil is fixed to the scalpwith adhesive. These changes permitMR imaging from the outset.

ABIs are not activated until a patientrecovers from surgery. Because of con-cerns about the potential activation ofbrainstem structures, many centers pre-fer that the electrode be activated in amonitored setting. After successful ABIimplantation, medical and audiologicalfollow-up is performed at least every 3months for the first year and then an-nually thereafter.

Indications for ABIsThe FDA has approved the ABI for

patients with NF-2 who are 12 years orolder who have reasonable expectationsand motivation. The degree of hear-

ing loss is not a deciding factor (i.e.,there is no audiological cut-off for ABIplacement). Other factors also distin-guish optimal ABI candidates: goodoverall health, acceptable vision (becausethe ABI may aid most with lipreading),high motivation, excellent family sup-port, acceptable anatomical status, andan interest in spoken communication.Because the ABI creates an ‘unnatural’quality of sound, many users are disap-pointed with it at first. High motiva-tion and excellent family support areimportant to ensure that users do notdrop out of the ABI program.

These issues are exemplified in theHouse experience with placing ABIs inteenagers.17 Of 21 teens implanted, fourdropped out of the program. One laterreturned and has become an above av-erage ABI user as a result of strong fam-ily support and encouragement. Qual-ities most characteristic of programdropouts included low motivation andpoor family support. However, an im-portant lesson from the House experi-ence has been that the ideal candidate isself-motivated to hear again rather thanmotivated for the benefit of familymembers.

The only absolute contraindicationto the placement of an ABI is an activeinfection. Relative contraindications

41

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

B

Figure 1. (A) The auditory brainstem implant consists of a radio receiver-stimulator that is implanted in the temporal bone. A groundelectrode is inserted under the temporalis muscle, and the multichannel brainstem implant paddle is inserted into the lateral recess ofthe fourth ventricle. (B) Sound is picked up at the pinna by a microphone that sends the signal to a speech-processor/digitizer. Thelatter sends the signal to the brainstem implant receiver through the transmitter coil. Photographs courtesy of Cochlear Ltd.

A

Ground electrode

Paddle

Receiver-stimulator

Transmittercoil

Microphone Speechprocessor-digitizer

are rare and include patients who maymeet the formal criteria but who lackadequate support or motivation, as out-lined above. As discussed later, if the co-chlear nerve can be preserved intact afterresection of an acoustic neuroma, place-ment of a cochlear implant may be in-dicated if promontory testing is positive.

The FDA has approved the use ofABIs only in the setting of NF-2. How-ever, off-label use of the device has beenreported (predominantly in Europe) forother indications, including deafness re-lated to bilateral skull base trauma,3 spo-radic acoustic neuroma in patients withonly one hearing ear,4 cochlear nervehypoplasia or aplasia,2,5 cochlear ossifica-tion,11 bilateral auditory nerve deafness,22

or involvement with other neoplasticsyndromes such as Bourneville’s diseaseand von Hippel-Lindau disease.15

Since ABIs were approved by theFDA, they have been successfully im-planted at an increasing number of do-mestic and international centers. In Oc-tober 2004, the first ABI was placed atour institution (Fig. 2).

First Side vs. Second Side SurgeryInitially, ABIs were placed at the time

of second side surgery for tumor re-moval. Now more than a third of ABIsare placed during the first surgery (i.e.,while the patient still has serviceablehearing on the contralateral side).18

There are several advantages to plac-ing the device during tumor removal

on the first side. If the device is im-properly positioned within the lateralrecess of the fourth ventricle, does notelicit auditory sensations, or creates in-tolerable side effects that prevent its use,then the contralateral side can be im-planted when surgery is performed onthe second side. The patient therebyhas a second chance to obtain a func-tioning implant. Conversely, if the firstside is implanted and functions well, theuser can become accustomed to the de-vice while still able to hear with theother ear. This situation may help thepatient to acclimatize to and interpretthe ABI sound. However, most patientsimplanted during surgery on the firstside do not become daily users of ABIuntil their deafness is bilateral.

Promontory TestingPromontory testing is the direct stim-

ulation of the cochlear promontory viaan electrode placed surgically in the mid-dle ear. The procedure is performed un-der local anesthesia through a smallmyringotomy and can distinguish be-tween cochlear and retrocochlear causesof deafness. In a negative promontorytest, pitch is not perceived in response todirect electrical stimulation of the co-chlea. The finding implies a retroco-chlear cause of deafness. Conversely, ina positive test, pitch is perceived in re-sponse to stimulation, indicating that thecochlear nerve is capable of transmittingtonal information centrally.

Marangos et al.15 examined 19 earswith promontory testing in patientswith profound bilateral deafness. Thepromontory tests were positive in fiveof six ears that had not undergone pre-vious surgery and in three other surgi-cally treated ears with residual tumor.The patient who responded negativelyon promontory testing but who hadundergone no previous surgery had bi-lateral acoustic neuromas causing pro-found brainstem compression. A pa-tient presenting with deafness one yearafter undergoing a subtemporal ap-proach in which hearing was preservedhad a positive promontory test. Thispatient was fitted with a cochlear im-plant and had an excellent response.

Marangos et al.15 concluded that thegoal of first-side surgery should be to pre-serve hearing or at least the anatomicalintegrity of the auditory nerve to iden-tify patients who will respond to cochle-ar implantation. Preservation of the co-chlear nerve depends on the size of thetumor. Small tumors are usually found inpatients with a known family history ofNF-2. In sporadic cases, the tumors areoften diagnosed when too large for thecochlear nerve to be preserved. Theyfurther argued that ABI should be re-served for patients with a negative pre-operative promontory test, patients inwhom the cochlear nerve cannot be pre-served, or patients whose deafness is bothcochlear and retrocochlear (e.g., related totumor infiltration of the auditory nerve).

42

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

B

Figure 2. (A) Intraoperative views through the microscope as the auditory brainstem implant electrode paddle is positioned within thelateral recess and (B) as the receiver is secured to the temporal bone.

A

Neuroanatomy and Phys-iology of Audition

Sound is represented spatially in theauditory system. This spatial relation-ship, called tonotopy, is maintained withfidelity in ascending neuroanatomicalpathways from the peripheral auditoryend organ (the organ of Corti in hu-mans) to the primary auditory cortex.In the organ of Corti, hair cells line thebasal membrane of the cochlea in atonotopic fashion. Hair cells respond-ing best (i.e., with a characteristic fre-quency) to high pitches are located atthe basal turn of the cochlea, and haircells responding best to low frequenciesare located at the cochlear apex.

In his theoretical work on the natureof music, Tonemfindungen, Helmholtzproposed that sound delivers a standingwave to the basement membrane andthat regionalized movements of thisstructure responded to specific frequen-cies to determine tonotopy in the co-chlea. This model predicts that the haircells of the cochlea are passive transduc-ers of frequency. Although still widelytaught, this view is now understood tobe incomplete. In fact, hair cells are dy-namic, and various adaptations, includ-ing active motor processes and electri-cal tuning, determine their characteristicfrequency. These adaptations tune thehair cells, such that they preferentiallydepolarize in response to their charac-teristic frequency. Loud sounds, how-ever, flatten their tuning curve.

To preserve this tonotopic relation-ship, ascending primary afferents of thespiral ganglion must maintain fidelity tothe characteristic frequency of the innerhair cells that they supply. Thus the co-chlear division of the vestibulocochlearnerve is composed of afferent fibers,each of which represents a specificpitch-place within the cochlea. As thesefibers enter the brainstem, they termi-nate in regions of the cochlear nucleusthat again reflect spatial segregation offrequency encoding. Specifically, affer-ents with a high characteristic frequen-cy (in humans the upper limit of hear-ing is about 20 kHz) plunge deep intothe ventral cochlear nucleus before bi-furcating and sending a descending

branch to the dorsal cochlear nucleus.In contrast, low-frequency fibers bifur-cate superficially within the substanceof the ventral cochlear nucleus. All pri-mary afferent fibers bifurcate within theventral cochlear nucleus before termi-nating.

If characteristic frequency is plottedagainst position in the brainstem, thecochlear nuclei are composed of isofre-quency laminae, or layers of neurons(like the layers of an onion) (Fig. 3). Inthis tonotopic organization, the super-ficial regions of the cochlear nuclei (thelateral and anterior laminae of the ante-rior and posterior divisions of the ven-tral cochlear nucleus) represent low fre-quency sounds, whereas the deeperregions of the nucleus represent higherfrequencies. Unfortunately, the aspectof the cochlear nuclei accessible to sur-face electrode stimulation is thereforerestricted to a very narrow range of fre-quencies. This limitation is the moti-vation behind the development of pen-etrating electrodes for implantation.

In addition to the tonotopic organi-zation of the cochlear nuclei, the richconstellation of neuronal cell types andinterconnections is only partly under-stood. Many ABI recipients are able toperceive pitch. Most speech is compre-hensible over a relatively narrow bandof frequencies. Although pitch rankingof multiple channels is thought to facil-itate speech comprehension, speechcomprehension may be improved withmultiple channels even if pitch rankingis impossible, and the extent of pitchranking does not necessarily correlatewith ABI performance outcomes.13 Thecochlear nuclei are composed of manydifferent neuronal populations, includ-ing fusiform cells, globular and spheri-cal bushy cells, and octopus cells, amongothers. Many neuronal cell types with-in the cochlear nuclei are specialized torespond to information other thanstrictly pitch. For example, some cellsrespond best to the onset of a soundstimulus, others to the cessation of sucha stimulus. This specialization may ex-plain why multiple channels may im-prove speech comprehension even if thepatient is unable to use the channels to

recognize different pitches. Alterna-tively, if improved speech comprehen-sion requires such ‘on’ and ‘off ’ signalsor other more complicated stimuli, thenstrategies such as penetrating ABIs thatfocus on the tonotopic organization ofthe cochlear nuclei may not improvespeech comprehension.

Surgical Considerationsand Techniques

Surgical AnatomyThe surgical goal of ABI implantation

is to position the electrode array on thesurface of the cochlear nuclei in the lat-eral recess of the fourth ventricle (Fig 4).The landmarks for proper identificationof the lateral recess include the origins ofthe facial, vestibulocochlear, glossopha-ryngeal, and vagus nerves; the flocculusof the cerebellum, the choroid plexusexiting the foramen of Luschka, and thetaenia. However, these landmarks canbe severely distorted by tumor.

The simplest way to identify the fo-ramen of Luschka is to follow the ves-tibulocochlear nerve proximally. Whenthe nerve is cut, its stump can be iden-tified and similarly followed. In a studyof the cerebellopontine angle in 100human specimens, the distance betweenthe facial and vestibulocochlear nerveswas 4.7 mm ± 0.9 mm.12 The distancebetween the vestibulocochlear and glos-sopharyngeal nerves was 5.5 mm ± 1.0mm. Thus, if the root of the vestibulo-cochlear nerve itself cannot be identi-fied with confidence, its root lies in atriangle approximately 5 x 6 mm with-in the origins of the facial and glosso-pharyngeal nerves.

In the same study12 the dimensionsof the foramen of Luschka were about3.4 mm x 2.0 mm. The foramen waswide open in only 24% of cases, openonly after incision of the arachnoid in53%, functionally closed (requiring ex-tensive dissection) in 18%, and anatom-ically occluded in 5%. The taenia of thechoroid was present in 92% of speci-mens, and it required incision to accessthe lateral recess in 51%.

Importantly, Matthies et al.16 analyzed

43

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

the outcomes of ABI placement as afunction of the amount of dissection re-quired to identify and open the foramenof Luschka. There was no relationshipbetween the density of arachnoid adhe-sions involving the foramen of Luschkaand outcomes with the ABI. Howev-er, these data were obtained from a smallsample (n = 8) and may lack the powerto demonstrate such a relationship.

The cochlear nucleus extends fromthe floor of the fourth ventricle medial-ly to the entry zone of the vestibuloco-chlear nerve at the ventrolateral brain-stem surface. In an anatomical study ofthe cochlear nuclei of humans, Questerand coworkers19,20 made thread measure-ments of the length of the cochlear nu-cleus complex from the auditory tuber-cle (which overlies the dorsal cochlearnucleus) to the entry zone of the ves-tibulocochlear nerve. The length of thecochlear nucleus was 12.8 mm ± 1.2mm. The length of the nucleus locat-ed within the lateral recess (i.e., primar-ily the ventral cochlear nucleus and siteof ABI placement) was 7.6 mm ± 1.5mm. Thus the dimensions of currentgenerations of ABI (i.e., about 3 x 8mm) are well suited to placement of thedevice completely within the lateral re-cess.

Surgical ApproachThe original clinical trial protocol

called for ABI placement via a translab-yrinthine approach, which remains themost popular approach. The translaby-rinthine approach offers the most directangle of attack to the lateral recess of thefourth ventricle. However, some sur-geons, particularly in Europe, have ad-vocated for alternative approaches toABI placement. Chief among these al-ternatives is the retrosigmoid or lateralsuboccipital approach.6,7

The main argument for use of theretrosigmoid approach is the potentialto preserve the cochlea and cochlearnerve if promontory testing indicatesthat the patient may benefit from a co-chlear implant rather than an ABI.14

Other advantages include shorter oper-ative times and less risk of cerebrospinalfluid (CSF) leakage. Its disadvantages

include difficulty in placing the lead ap-propriately. In one series, two of threeelectrodes placed via the retrosigmoidapproach required revision for malpo-sitioning.23

Other innovations to ABI placementinclude subtonsillar or telovelar ap-proaches via a midline suboccipital cra-niotomy21 and ABI placement throughone of the standard approaches but withendoscope-assisted technique.9 The ad-vantages of a subtonsillar approach in-

clude superior visualization of the floorof the fourth ventricle and the absenceof scarring of the surgical corridor.Some authors have asserted that the dor-sal cochlear nucleus, which can be ob-served directly via a subtonsillar ap-proach where it is found under theauditory tubercle, is a superior target forABI placement, because the ventral co-chlear nuclei are partially obscured bythe cerebellar peduncle. However, thedorsal cochlear nucleus responds best to

44

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

Figure 3. Schematic representation of the cochlear nucleus complex (comprising the ven-tral and dorsal cochlear nuclei) and surrounding structures. Alternating shaded bandsrepresent isofrequency laminae of afferent fibers and second-order neurons tuned tolow-frequency (short dashed line) and high-frequency (long dashed line) sounds. Be-cause of the topography of the cochlear nucleus (tonotopy), a paddle electrode placedwithin the lateral recess tends to contact low isofrequency laminae. High-frequency neu-rons, of which the dorsal cochlear nucleus is primarily composed, are consequently onlypartially accessible to paddle electrode stimulation.

Vestibulocochlearnerve

Ventralcochlearnucleus

Dorsalcochlearnucleus

Low-frequency fiber

High-frequency fiber

higher frequencies, including those out-side the normal range used in speech.

Regardless of the suitability of thedorsal cochlear nucleus as a target forABI, the telovelar approach, in whichthe inferior medullary velum is split andthe taenia are divided from their attach-ment to the floor of the fourth ventri-cle, requires a midline suboccipital cra-niotomy or craniectomy. However, anacoustic neuroma cannot be resected viasuch an approach. An ABI can only beplaced if the surgeon is willing to ex-tend a retrosigmoid craniotomy inferi-orly, possibly including a C1 laminec-tomy. Nevertheless, the telovelar orsubtonsillar approach may be consideredwhen the surgeon anticipates too muchscarring or anatomical distortion relat-ed to the tumor to allow accurate place-ment of the electrode through either atranspetrosal or lateral suboccipital ap-proach (i.e., when an ABI is to be placedafter the operation for tumor resection).

In a cadaveric study of endoscopical-ly assisted translabyrinthine, retrosigmoid,and even middle fossa approaches, Fried-land and Wackym9 demonstrated thatvisualization of the lateral recess is facil-itated with the use of an endoscope. An

endoscope can be wielded to excellenteffect in such situations because a 30-degree angled endoscope gives the sur-geon the ability to look ‘around the cor-ner.’ Although theoretically appealing,use of an endoscope is hindered by thenecessity of holding the camera and ad-vancing the ABI paddle parallel to thescope before the device can be placed.Advances in endoscopy may increase theuse of endoscopically assisted proce-dures.

Intraoperative MonitoringFor several reasons intraoperative

monitoring of evoked auditory poten-tials and monitoring of the trigeminal,facial, and glossopharyngeal nerves iscrucial during ABI placement.1,10 Mon-itoring auditory brainstem responses cancorroborate accurate electrode place-ment, whereas monitoring neighbor-ing cranial nerves may detect inadver-tent stimulation of regional structures.When the labyrinth has not been drilledand the cochlear nerve has been pre-served, promontory testing also may beuseful to predict the possible benefit ofcochlear implantation.

The auditory brainstem response(ABR) is thought to reflect the sum-mated activity of ascending portions ofthe auditory system. Classically, fiveABR waves (i.e., waves I-V) are thoughtto reflect the activity of the cochlea, co-chlear nerve, cochlear nuclei, olivarynuclei, and nuclei of the lateral lemnis-cus, respectively. In the setting of ABIplacement, the cochlear or cochlearnerve waves of the ABR are not ex-pected because the cochlear nuclearcomplex is stimulated directly.24

Intraoperatively, the first wave of theABR seen would be wave III, that ofthe cochlear nuclei. Ideally, three waveswould be expected: those of the co-chlear nuclei (III), olivary nuclei (IV),and nuclei of the lateral lemniscus (V).In practice, anywhere between one andthree responses, termed P1-P3, areseen.24

The presence of one or more re-sponses helps corroborate proper elec-trode placement. However, there arepitfalls in the interpretation of intraop-erative responses, including the presenceof motor contamination (i.e., motor re-sponses transmitted from inadvertentstimulations of the trigeminal, facial,

45

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

Figure 4. (A) Schematic representation of the right cerebellopontine angle. The ventral and dorsal cochlear nuclei border the lateral re-cess and extend from the anterolateral aspect of the brainstem to the floor of the fourth ventricle. Surgical landmarks for identifying thelateral recess include the flocculus of the cerebellum and the choroid plexus within the foramen of Luschka. (B) The auditory brainstemimplant is placed within the lateral recess of the fourth ventricle. The leads of the paddle electrode are apposed to the dorsolateral sur-face of the ventral and dorsal cochlear nuclei. The electrode is held in place with a Teflon plug (not depicted).

BA

Facialnerve

Vestibulocochlearnerve Ventral & dorsal

cochlear nuclei

Choroid plexus in foramen ofLuschka

Flocculus

Facialnerve Stump of

vestibulocochlearnerve

Electrodepaddlein lateral recess

Ventralcochlearnucleus Dorsal

cochlearnucleus

Inferiorcerebellarpeduncle

glossopharyngeal, or vagus nerves).Motor responses can be differentiatedfrom auditory responses by both laten-cy and amplitude. Higher amplitudewaves with longer latencies are indica-tive of a motor response.

ABR monitoring is best understoodas an adjunct to anatomical means of lo-calizing the cochlear nuclei. Its role issimply to confirm that the ABI stimu-lates the auditory brainstem, with mini-mal or no stimulation of other structures.No correlations between the number orquality of waveform responses (i.e., P1,P2, P3) and the efficacy of ABI havebeen found. Moreover, as long as thesurgeon is confident of adequate elec-trode placement based on reliable ana-tomical landmarks, even the completeabsence of ABR waves does not neces-sarily predict that the ABI will fail toimpart useful hearing.

Surgical Results

American ExperienceThe United States clinical trial of the

multichannel ABI, composed of data col-lected from 10 domestic centers, consti-tutes the largest published series. It in-volved 92 patients receiving the Nucleus22 (Cochlear Corporation, Englewood,CO), eight-electrode ABI, 85% of whomreported receiving auditory sensations.8

At their six-month follow-up examina-tion, 93% of patients receiving auditorysensations demonstrated improved lip-reading, reflected in a mean improvementof 24% on the CUNY sentence test (anopen-set speech comprehension test)over lipreading alone. Importantly, 85%of patients reported satisfaction with theirABI, and 74% would recommend it toothers. Of patients who used the devicedaily (97% of those patients implanted atthe second surgery (i.e., deaf patients),65% reported using the device more than8 hours a day.

In 13 patients the failure of the deviceto elicit auditory stimulation was attrib-uted to improper electrode placementrelated to distorted anatomy. Therewere two deaths, although neither weredirectly attributable to the ABI. Non-

auditory side effects, most commonlyparesthesias involving the ipsilateral body,were common. They are attributed tostimulation of the inferior cerebellar pe-duncle, which lies close to the cochlearnuclei (Fig. 4).

In the House Ear Institute experi-ence with the long-term follow-up of71 patients receiving an eight-electrodeABI, 10 patients did not participate infollow-up.18 Of the remaining 61 pa-tients, 6 detected no auditory sensationsafter implantation. Two CSF leaks weretreated by pressure bandage and lumbardrainage, and one patient developedmeningitis. Eventually, 24% of the elec-trodes were inactivated because ofnonauditory perception.

ABI patients benefited from significantqualitative and quantitative improvementsin hearing, including awareness of envi-ronmental sounds and improved closedand open set speech recognition. Fewpatients, however, demonstrated signif-icant open-set speech recognition in thesound-only mode. Environmental sounddiscrimination was at or above 50% onclosed-set tests. On a closed-set test ofword recognition (the MTS test), 87%of patients scored significantly higherthan chance. Similar to the results ofthe clinical trial in the United States, themean improvement in CUNY sentencerecognition scores over lipreading alonewas 26%.

Perhaps the most significant findingfrom the House experience, the largestsingle center experience in the world,was that a significant proportion of pa-tients continued to show improvementin audiological indices as long as 8 yearsafter implantation. This finding testi-fies to the use-dependence of improvingspeech recognition.

European ExperienceThe European experience with ABI

differs from the experience in the Unit-ed States in several significant ways. Thelatter clinical trial primarily was per-formed with the Nucleus 22 eight-elec-trode ABI while in Europe 12-, 16- and21-electrode multichannel ABIs havebeen placed preferentially over eight-electrode models (including the clinical

trials of Digisonix, Med-El, and theNucleus 24 ABIs). European centershave implanted the devices for more in-dications than have been used in theUnited States. European also have usedthe retrosigmoid approach more oftenthan the translabyrinthine approach.

Nevertheless, the European experi-ence broadly mirrors the experience inthe United States in terms of rates ofsurgical morbidity and functional out-comes. Unfortunately, heterogeneity inaudiological tests performed makes di-rect comparison of the audiological testsimpracticable. The combined results ofthe European Nucleus (i.e., 21-channelABI) experience (n=58), the Digisonixclinical trial results (n = 14), and theMED-EL clinical trial results (n=16)yielded 88 patients. Three patients wereimplanted for indications other thanNF-2, including bilateral auditory neu-ropathy and miscellaneous tumor syn-dromes affecting the vestibulocochlearnerve bilaterally. Three patients diedand one wound infection required ex-plantation of the device. Of 24 patientswho underwent a retrosigmoid crani-otomy for electrode placement, threerequired revision surgery. One patienthad transient but significant symptom-atic cerebellar swelling. Between 86 and100% of patients experienced auditorysensations, 63% to 100% of whom weredaily users. All patients who receivedauditory percepts demonstrated im-proved speech recognition in conjunc-tion with lipreading. A limited numberof patients achieved some open-setsound-only speech recognition.

Future DirectionsIn the last decade ABI research has

focused on increasing the number ofavailable channels for stimulation. Theideal number of channels has not yetbeen established and whether morechannels necessarily correlate with bet-ter chances of comprehending speech isunclear.13 The biggest innovation inABI design has been the recent devel-opment of penetrating electrodes forstimulation of regions of the ventral co-chlear nucleus complex that have been

46

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004

inaccessible to surface electrodes. Suchelectrodes offer the potential of stimu-lating higher frequency cell populationsand may therefore increase the numberof patients receiving ABI who are ableto perform pitch scaling. Even thoughfewer electrodes are used, such an in-creased range of pitch sensation mightfacilitate speech recognition.

ConclusionsABIs restore meaningful hearing to

patients deafened by NF-2. Almost allpatients can expect to receive some au-ditory perceptions and awareness of en-vironmental sounds. The ability to com-municate with lipreading improvessignificantly in most patients, but fewwill achieve sufficient open-set sound-only speech comprehension to enableuse of the telephone.

Appropriate expectations and moti-vation on the part of the patient ismandatory. Many ABI patients are ini-tially discouraged by the quality ofsound heard with an ABI. They mustunderstand that participation in theirauditory rehabilitation program is cru-cial to continue to improve their profi-ciency with the device. Such improve-ment can be expected to continue overmany years.

References1. Brackmann DE, Hitselberger WE, Nelson RA, et

al: Auditory brainstem implant: I. Issues in surgi-cal implantation. Otolaryngol Head Neck Surg108:624-633, 1993

2. Colletti V, Carner M, Fiorino F, et al: Hearingrestoration with auditory brainstem implant inthree children with cochlear nerve aplasia. OtolNeurotol 23:682-693, 2002

3. Colletti V, Carner M, Miorelli V, et al: Auditory brain-stem implant in posttraumatic cochlear nerveavulsion. Audiol Neurootol 9:247-255, 2004

4. Colletti V, Fiorino F, Carner M, et al: Auditory brain-stem implantation: The University of Verona ex-perience. Otolaryngol Head Neck Surg 127:84-96, 2002

5. Colletti V, Fiorino F, Sacchetto L, et al: Hearinghabilitation with auditory brainstem implantationin two children with cochlear nerve aplasia. Int JPediatr Otorhinolaryngol 60:99-111, 2001

6. Colletti V, Fiorino FG, Carner M, et al: The retro-sigmoid approach for auditory brainstem im-plantation. Am J Otol 21:826-836, 2000

7. Colletti V, Sacchetto L, Giarbini N, et al: Retro-sigmoid approach for auditory brainstem implant.J Laryngol Otol 114 (Suppl 27):37-40, 2000

8. Ebinger K, Otto S, Arcaroli J, et al: Multichannelauditory brainstem implant: US clinical trial results.J Laryngol Otol 114 (Suppl 27):50-53, 2000

9. Friedland DR, Wackym PA: Evaluation of surgicalapproaches to endoscopic auditory brainstemimplantation. Laryngoscope 109:175-180, 1999

10. Frohne C, Matthies C, Lesinski-Schiedat A, et al:Extensive monitoring during auditory brainstemimplant surgery. J Laryngol Otol 114 (Suppl27):11-14, 2000

11. Grayeli AB, Bouccara D, Kalamarides M, et al: Au-ditory brainstem implant in bilateral and completelyossified cochleae. Otol Neurotol 24:79-82, 2003

12. Klose AK, Sollmann WP: Anatomical variations oflandmarks for implantation at the cochlear nucle-us. J Laryngol Otol 114 (Suppl 27):8-10, 2000

13. Kuchta J, Otto SR, Shannon RV, et al: The multi-channel auditory brainstem implant: How manyelectrodes make sense? J Neurosurg 100:16-23, 2004

14. Marangos N, Stecker M, Laszig R: Topodiagno-sis of deafness: Strategy for treatment of neu-rofibromatosis type 2. J Laryngol Otol 114(Suppl 27):3-7, 2000

15. Marangos N, Stecker M, Sollmann WP, et al: Stim-ulation of the cochlear nucleus with multichan-nel auditory brainstem implants and long-termresults: Freiburg patients. J Laryngol Otol 114(Suppl 27):27-31, 2000

16. Matthies C, Thomas S, Moshrefi M, et al: Audito-ry brainstem implants: Current neurosurgical ex-periences and perspective. J Laryngol Otol 114(Suppl 27):32-36, 2000

17. Otto SR, Brackmann DE, Hitselberger W: Audi-tory brainstem implantation in 12- to 18-year-olds.Arch Otolaryngol Head Neck Surg 130:656-659, 2004

18. Otto SR, Brackmann DE, Hitselberger WE, et al:Multichannel auditory brainstem implant: Updateon performance in 61 patients. J Neurosurg96:1063-1071, 2002

19. Quester R, Schroder R: Topographic anatomy ofthe cochlear nuclear region at the floor of thefourth ventricle in humans. J Neurosurg 91(3):466-476, 1999

20. Quester R, Schroder R, Klug N: Optimization ofmicrosurgical operation technique to insert au-ditory brainstem implants, taking into accountthe results of a morphometric study [in German].HNO 52:706-713, 2004

21. Seki Y, Samejima N, Kumakawa K, et al: Subton-sillar placement of auditory brainstem implant.Acta Neurochir Suppl 87:85-87, 2003

22. Sollmann WP, Laszig R, Marangos N: Surgicalexperiences in 58 cases using the Nucleus 22multichannel auditory brainstem implant. JLaryngol Otol 114 (Suppl 27):23-26, 2000

23. Vincent C, Zini C, Gandolfi A, et al: Results of theMXM Digisonic auditory brainstem implant clini-cal trials in Europe. Otol Neurotol 23:56-60,2002

24. Waring MD: Auditory brain-stem responsesevoked by electrical stimulation of the cochle-ar nucleus in human subjects. Electroen-cephalogr Clin Neurophysiol 96:338-347,1995

47

Lekovic et al: Auditory Brainstem Implantation

BARROW QUARTERLY • Vol. 20, No. 4 • 2004