The auditory midbrain implant (AMI) is a novel central auditory implant for hearing restoration in patients with neural deafness due to damage of the auditory nerve by bilateral vestibular schwannomas (mainly NF2). These patients cannot benefit from a cochlear implant nor from an auditory brainstem implant due to damage of the cochlear nucleus at the brainstem either by the tumor itself or its treatment.
The penetrating multichannel electrode arrays are inserted into the central nucleus perpendicular to the frequency layer. The tonotopic stimulation encodes different frequencies which can be used for speech discrimination. Five patients with a single shank and five patients with a double shank electrode have been implanted so far. The patients experience limited speech recognition scores comparable to those with auditory brainstem implants.
Overall the AMI provides an alternative treatment for hearing rehabilitation in NF2-patients.
18 Auditory Midbrain Implant
Patients with neural deafness cannot benefit from a cochlear implant (CI) due to damaged auditory nerves, such as those with neurofibromatosis type II (NF2) who have had removal of tumors involving the cochlear nerve. There are also deaf individuals with cochleas that cannot be implanted with a CI due to anatomical distortions/ossification or damage of the cochleas. In these types of patients, electrical stimulation of structures of the auditory pathway central to the auditory nerve can be used for hearing restoration. Two types of central auditory prostheses have been realized for clinical application: the auditory brainstem implant (ABI) and the auditory midbrain implant (AMI; Fig. 18.1). 1 , 2 , 3
The ABI has been in clinical use since 1979 in different versions including surface and penetrating electrodes. 5 Hearing performance with the ABI has remained substantially inferior to those achieved with cochlear implants and show a large variability in outcomes, spanning from no auditory sensations to some degree of open-set speech understanding. 6 , 7 , 8 , 9 Several possible reasons for the poor outcomes include the limited access or stimulation of the tonotopic organization of the cochlear nucleus, even with the penetrating electrode arrays used in a previous clinical study, and the preprocessing of auditory information from the cochlea to the brainstem that has been bypassed with the ABI. Current stimulation strategies for the ABI are derived from CIs and may not be able to sufficiently restore the auditory information at the brainstem level. Particularly in NF2 deaf patients, it has been proposed that poor ABI outcomes may be due to the damage caused at the brainstem level, associated with the tumor and/or tumor removal process, especially since hearing outcomes in nontumor patients have shown greater performance than those of tumor patients. 6 , 7 , 13
Considering these limitations, as well as to attempt to overcome some of these limitations, the AMI has been developed. 4 , 14 , 15 , 16 The main concept of the AMI is the use of penetrating single-shank or double-shank electrode array with ring contacts that can stimulate the different frequency layers of the inferior colliculus, beyond the damaged tumor region in the brainstem area (Fig. 18.2). This frequency layer stimulation could potentially improve speech discrimination by more precisely stimulating the tonotopic organization of the inferior colliculus in its central nucleus. In order to address the three-dimensional organization and the different neuronal presentation within one frequency layer, a double-shank electrode array was recently developed and brought into clinical studies with patients implanted in 2017 to 2019. This double-shank electrode array stems from research in animals and outcomes from a previous clinical study using the single-shank electrode array version in five deaf NF2 patients. The rationale, technology development and validation, and animal and human research supporting both human studies is described in detail in previous reviews. 3 , 17
The penetrating electrode arrays are attached to a CI receiver-stimulator developed by Cochlear Limited (Australia). The electrode arrays have a tapered tip (Fig. 18.2). For insertion purposes, each electrode shank is stabilized by a stylet with a sharp tip in a central canal of the shank. The stylet enables insertion of the shank into the brain and then can be pulled out after insertion, allowing the electrode array to become more flexible in order to adjust to the brain structure and creating less force on the brain tissue. The implant has a reference electrode with a separate wire which is ball-shaped placed into the temporalis muscle area in addition to the receiver-stimulator casing, similar to typical CIs developed by Cochlear Limited.
18.2 Preoperative Diagnostics and Selection Criteria for Patients
The AMI is indicated in patients with bilateral neural deafness due to damage of the auditory nerve, mainly in NF2 patients with bilateral vestibular schwannomas (Fig. 18.3). In those NF2 patients, the tumor itself or subsequent treatment such as microsurgical tumor removal and/or radiotherapy can create damage at the brainstem level, including the nearby area of the cochlear nucleus and/or the lateral recess of the fourth ventricle. The AMI can bypass the damaged brainstem areas by targeting the central nucleus of the inferior colliculus. Targeting of the inferior colliculus is possible through preoperative imaging of the midbrain structures and fusion of magnetic resonance imaging (MRI) and computed tomography (CT) images (Fig. 18.4). These images can help identify key anatomical landmarks for the inferior colliculus, which includes the division line between the inferior colliculus and the superior colliculus, the midbrain midline, the tentorium, and the third ventricle. Audiological tests must document the neural deafness on both sides. The audiological profile is characterized by a severe to profound sensorineural hearing loss with pure speech discrimination in comparison to the pure tone threshold, missing auditory brainstem responses, and a negative (transtympanic) promontory test. These patients would not benefit from acoustic amplification.
AMI should be especially considered in patients with distorted or damaged brainstem due to previous treatment or by the tumor itself. The anatomic situation shall be well documented using high-resolution imaging including MRI of the temporal bone and the brain as well as a CT scan of the temporal bone and the skull.
18.3 Surgical Technique
AMI implantation can be performed either with or without tumor removal depending on the individual case. Surgery is possible either in the semi-sitting or supine position. The preferred position is semi-sitting for retrosigmoid (suboccipital) approach with medial extension up to the midline (Fig. 18.5). Navigation is advised in order to safely identify target structures, and determine the angles of electrode insertion. Bone anchored fiducials are placed in the skull around the craniotomy and a CT scan is taken. Surgery is done with the head fixed in the Mayfield clamp and using monitoring for facial nerve and long tracts and, depending on tumor extension, also other cranial nerves. After the exposure of the skull, a retrosigmoid craniotomy with extension to the midline is performed. After incision of the dura and opening the cistern for cerebellar relaxation a brain spatula is placed to hold the cerebellum, and tumor removal (e.g., vestibular schwannoma in the cerebellopontine angle) is initiated.
The next step is the exposure of the auditory midbrain through an infratentorial supracerebellar approach. In the semi-sitting position after removing the brain spatula from the cerebellopontine angle, the cerebellum shows a significant shift to the caudal direction due to gravity. This provides a sufficient surgical corridor to access the midbrain without any further retraction. The bridging veins from the tentorium can be typically preserved (Fig. 18.6). The surface of the inferior colliculus is exposed (Fig. 18.7). The inferior and superior colliculi, the brachium, and the trochlear nerve are exposed with proper dissection of vascular structures that should be preserved (Fig. 18.8).
The penetrating electrode arrays shall be placed into the central nucleus of the inferior colliculus with perpendicular penetration across the frequency layers shown to exist in previous anatomical and imaging studies in humans. 18 , 19 , 20 From the previous electrophysiological experiments and cadaver studies the optimum entry point on the surface of the inferior colliculus can be determined using anatomical landmarks (Fig. 18.9). The landmarks include the midline between the two inferior colliculi, the horizontal division line between the superior and inferior colliculi, the brachium to the lateral side, and the exit point of the trochlear nerve as the inferior point of reference. The electrode shanks of the double-shank array are then inserted into the inferior colliculus in a rostral-lateral toward inferior-medial direction with an insertion angle of 40 degrees toward the midline. Special navigation tools have been developed to determine this angle (Fig. 18.10). Placement of millimeter paper is useful to identify the two entry points, the rostral one being more lateral and the caudal one being more medial. The pia mater is first perforated with a special surgical tool (Fig. 18.11). After placement of the receiver-stimulator in the drilled-out bony bed superior to the craniotomy, the electrode arrays are pre-positioned with their leads bifurcated intracranially. First, the rostral shank is inserted (Fig. 18.12 and Fig. 18.13), followed by the caudal shank. The stylet is removed. The Dacron mesh defines the insertion depth (Fig. 18.14). Electrode cables are protected in the bone canal between the bony bed and the craniotomy. The reference electrode is placed under the temporalis muscle. Intraoperative electrophysiology is performed to determine the electrode impedances and measure the so-called neural responses with the telemetry system of the implant. In addition, the electrically evoked middle latency responses (E-MLRs) are recorded (Fig. 18.15). Watertight closure of the dura is (Fig. 18.16) performed and the craniotomy is closed by cranioplastic approach with artificial bone cement wound closure.
18.3.1 Location of the Electrode Arrays
Postoperative CT scans are taken and the position of the electrodes within the inferior colliculus are reconstructed with superposition into the preoperative MRI scans (Fig. 18.4). These data are still being analyzed to create three-dimensional anatomical reconstructions of the midbrain and shank positions, and will be published in a future publication.