Chapter 68. Cochlear Implants

An absence or disturbance of cochlear hair cells causes most cases of deafness. This defect in normal cochlear function, specifically, in the transduction of a mechanical acoustic signal into auditory nerve synaptic activity, represents a broken link in the delicate chain that constitutes the human sense of hearing. Cochlear implants afford an artificial means to bypass this disrupted link via direct electric stimulation of auditory nerve fibers.

Although current technological and scientific boundaries preclude the artificial transduction of sound by using the exact native cochlear patterns of synaptic activity at the level of each individual residual auditory nerve fiber, knowledge of these native patterns has aided the development of cochlear implants by allowing the processing of speech into novel synthetic electronic codes that contain the key features of spoken sound. By using these codes to systematically regulate the firing of intracochlear electrodes, it is possible to convey the timing, frequency, and intensity of sound. Cochlear implants have progressively evolved with increasing complexity and elegance from an experimental concept to a proven tool used in the management of patients with sensorineural hearing loss (SNHL). Worldwide, the number of implants is rapidly increasing. As with many other technology-driven medical treatment modalities, recent innovations in microcircuitry and computer science are continuing to drive the performance profiles of cochlear implants to new heights.

Currently, three separate corporations manufacture multichannel implant systems that are commercially available and approved by the FDA for use in both adults and children. Although expensive, multiple studies have demonstrated that the cost-utility of cochlear implantation is excellent and that it compares well with other common medical interventions.

All modern implant systems function by the use of the same basic components including a microphone, a speech processor, and an implanted receiver–stimulator (Figure 68–1).

Microphone & Receiver-Stimulator

Sound is first detected by a microphone (usually worn on the ear) and converted into an analog electrical signal. This signal is then sent to an external processor where, according to one of a number of different processing strategies, it is transformed into an electronic code. This code, a digital signal at this point, is transmitted via radiofrequency through the skin by a transmitting coil that is held externally over the receiver–stimulator by a magnet. Ultimately, this code is translated by the receiver–stimulator into rapid electrical impulses distributed to electrodes on an array implanted within the cochlea (Figures 68–2, 68-3, 68-4, and Figure 68–5).

Speech Processor

Current generations of speech processors are smaller in size and are continuously being redesigned to improve functionality, comfort, and cosmesis. Most adults and older children wear ear-level processors (behind-the-ear processors). Processors worn on the belt, clipped to clothing, or incorporated into small packs (body-worn processors) are still preferred for very young children as well as some adults (Figures 68–6, 68-7, and Figure 68–8). Entirely implantable devices are under development.

The literature uses the term speech processing, but this component may be more aptly termed sound processing, because the manipulations are not limited to speech only. In fact, a greater focus is now on enhancing the quality of all sound and specifically an effort to improve music appreciation. No matter what processing strategy is used, part of this process must include both amplification (ie, gain control) and compression. Since the deaf ear responds to electrical stimulation with a dynamic response in the range of 10–25 dB, processing must compress the signal to fit within this narrow range. How to best convert sound into an electrical signal is being actively investigated.

Electrical Stimulation Strategies

Multichannel Strategies

Some of the earliest multichannel strategies, known as analog filter bank strategies (continuous analog), channel speech through multiple frequency-dependent filters and deliver distinct outputs of sinusoidal analog signals directly to separate electrodes. Other so-called feature extraction strategies (F0, F1 and F0, F1, F2) work by rapidly drawing out frequency-based details that are considered to be the most essential in speech recognition; these include both fundamental frequency and vowel formants. Disbursement of this key information is accomplished through pulsatile signals having rates synchronous to the fundamental frequency and a tonotopic order that is derived from formants.

Pulsatile Stimulation

Modern adaptations of direct analog strategies have sought to overcome the problem of channel interaction or “spillover” that readily occurs when adjacent electrodes are simultaneously stimulated with continuous analog signals. The result of such efforts has led to the development of a strategy, continuous interleaved sampling, that delivers very rapid noncontinuous pulsatile stimulation over multiple filtered channels.

Spectral Analysis

In addition, newer high-rate spectral analysis strategies, SPEAK (spectral peak) and ACE (advanced combination encoders), for example, determine 6–10 spectral maxima for each input signal. Other newer approaches such as “n-of-m” strategies (n = filters, m = channels) are constantly undergoing innovation and refinement with the goal of combining the theoretical advantages of each type of system while incorporating ever-progressing new technologies.

Neural Responses

In speech processing, consideration is given to the incoming acoustic signal; in addition, actual neural responses (neural response telemetry, neural response imaging, or auditory nerve response telemetry) to stimulation may be measured and accounted for in the formulation of a neural stimulation scheme. By measuring evoked action potentials from specific electrodes, it is possible to predict the needed amplitudes for each channel of the speech processor. Some audiologists find this information particularly helpful when programming very young children.

Audiological Assessment

Candidacy for cochlear implantation relies heavily on the audiological evaluation. Although the audiometric criteria continue to change, the goal remains the same—identify those patients in whom the implant is likely to provide better hearing. Because of device improvements, the ability to hear with an implant has dramatically improved over time. Therefore, the accepted audiometric criteria for implantation have expanded to include patients with more residual hearing. Some of these patients, having quite useful low-tone hearing in the setting of middle- and high-frequency deficits, may now be classified as having “partial deafness.” Hybrid or short electrode devices have been developed to allow the preservation of the native low-frequency hearing, thus the patient would combine electric (cochlear implant) and acoustic (hearing aid) hearing in the same ear.

For adults in the United States, candidacy is based on sentence recognition test scores (eg, Hearing-in-Noise Test or Arizona Biomedical Sentences) with properly fitted hearing aids. Scores of 60% or less are generally needed to establish candidacy.

In children undergoing an evaluation for cochlear implantation, it is first necessary to establish a hearing threshold. This may include otoacoustic emissions, auditory brainstem response testing, auditory steady-state responses, and behavioral testing. A hearing aid trial can then be initiated and speech and language development assessed. Input is elicited from audiologists, parents, teachers, and speech and language pathologists. The cochlear implant team then assimilates the information and a determination is made regarding the child's progress with amplification and suitability for implantation.

Otological Assessment

A thorough otological history and physical examination are obtained as part of the preimplantation assessment and include an investigation into the etiology of the hearing loss. (For a thorough discussion of the evaluation of SNHL in adults and children, see Chapter 52, Sensorineural Hearing Loss.)

Otitis Media and Eustachian Tube Dysfunction

In pediatric patients, it is important to ascertain whether there is a history of recurrent ear infections, pressure equalization (PE) tube placement, or other otological surgeries. Patients with acute otitis media should be both treated with appropriate conventional antibiotics and demonstrated to be clear of infection before proceeding with surgery. For patients with a chronic middle ear effusion or recurrent acute otitis media, myringotomy with PE tube placement may be considered. Cochlear implants can safely coexist with PE tubes, although, ideally, patients would have an intact tympanic membrane at the time of surgery.

Chronic Otitis Media

For adult implant recipients, an intact tympanic membrane is preferred. Accordingly, patients with a tympanic membrane perforation, a chronic draining ear, or cholesteatoma often require other surgical procedures before implantation. For patients with long-dormant chronic ear disease and a modified radical cavity, it is not uncommon to combine cochlear implantation, closure of the ear canal, and obliteration of the mastoid cavity and middle ear into a single surgical procedure. Patients with active chronic ear disease processes, however, are better served with initial conventional otological surgery with cochlear implantation delayed until the ear is stable.

Cochlear Patency

When deafness is a result of meningitis or cochlear otosclerosis, special attention is required preoperatively to account for the possibility of cochlear soft tissue obstruction or ossification. Cochlear patency is best evaluated with MRI. CT scanning should demonstrate cochlear ossification; however, obliteration due to fibrosis and the presence of soft tissue can be best assessed with a T2-weighted MRI. When the cochlea appears to be actively undergoing obliteration, the surgeon may wish to implant quickly, assuming that hearing is not expected to be recovered.

Cochlear and Vestibular Malformations

The specific type and severity of malformation presents a unique set of challenges to the implant surgeon. Preoperative imaging is invaluable in planning surgery, choosing the most appropriate electrode array and avoiding complications. Incomplete device insertion, cerebrospinal fluid leak, facial nerve injury, vestibulopathy, and poorer hearing outcomes are all more common in the setting of malformations.

General Medical Considerations

It is important to be aware of individuals at particular risk of infection or surgical site complications during the implant workup process. Pertinent nonotological medical conditions to consider are previously irradiated surgical field, immunodeficiency, poorly controlled diabetes mellitus, tobacco use, malnutrition, a prominent history of allergic hypersensitivity reactions, or widespread dermatological disease.

Vestibular Evaluation

A vestibular evaluation, including at least electronystagmography, while not required preoperatively can be helpful in selecting the ear for implantation and for assessing the risk of postoperative balance problems. Although the risk of losing balance function in the ear being implanted is low with current minimally traumatic surgical technique, if function was lost and the contralateral ear already lacked function, the resulting balance problems can be devastating, particularly in elderly adults. Patients with an existing balance problem or suspicion for unilateral or bilateral vestibular hypofunction should undergo preoperative testing.

Radiological Assessment

Radiological assessment for cochlear implantation typically consists of fine-cut CT scanning and/or MRI of the temporal bone. CT is preferred to delineate the detailed bony anatomy of the labyrinth, which may include evidence of cochlear ossification, abnormal facial nerve course, or congenital abnormalities. In patients in whom soft tissue detail is required, as in a patient with an elevated suspicion of central pathology or when cochlear patency is in question, MRI with and without gadolinium plus high-resolution T2-weighted images may prove to be valuable. The modification of MRI pulse sequences, such as fast-spin echo, and the development of new coils are leading to continuously improving resolution of the inner ear and internal auditory canal. Specifically, MRI can now allow reliable imaging of the internal auditory canal contents and afford confirmation of the presence of an auditory nerve. The Michel deformity (ie, congenital cochlear agenesis) and absence of the auditory nerve, which may be present with the narrow internal auditory canal malformation or in the setting of auditory neuropathy, are the two absolute contraindications to cochlear implantation that may be found on radiological assessment.

General Considerations

In addition to meeting audiometric and medical criteria, the basic evaluation of cochlear implant candidates involves analyzing multiple other factors (Table 68–1). The goals of the evaluation are (1) to determine whether the patient is likely to benefit from the implant, (2) to establish realistic expectations concerning the outcome, and (3) to assess the needs and expectations of the patient, the patient's family, or both. Not all patients who meet audiometric and medical criteria are appropriate implant candidates. There must be grounded expectations as well as a firm commitment to follow through with the necessary postimplantation rehabilitation and programming.

Patients with Other Cognitive or Developmental Disorders

A unique group of individuals requiring careful consideration are with hearing loss and other developmental and cognitive deficits. Historically, children with cerebral palsy or children with other conditions in addition to hearing loss were denied implantation. It is now clear, however, that many of these patients are very good candidates. In fact, if a hearing disability can be ameliorated with a cochlear implant, other disabilities (eg, a learning disability) may become less pronounced or more manageable. In contrast, in a child with very severe developmental issues and a poor prognosis for cognitive development, a cochlear implant may simply be another burden and not result in an improved quality of life.

Timing of Implantation

The timing of implantation is very important. Earlier implantation in children generally yields more favorable results, and many centers routinely implant children under 12 months of age. In engaging this pursuit, complex questions arise with regard to how early children should undergo cochlear implantation in order to optimize eventual developmental outcomes. Although the answers to all these questions have yet to be definitively answered, success in progressively younger patient populations is driving the age of implant recipients lower. Early implantation essentially limits how far behind the child is in language development. Likewise, adults do better with a shorter duration of deafness. It has been shown that outcomes in adults over age 80 rival those in young adults. Excellent results can be obtained in older children and adults with long-term deafness, but expectations for the outcome should be modified.

Implant Selection

All three currently available devices are excellent, and rarely there are strong reasons to prefer one to another. The hearing outcomes seem to be similar regardless of the device, thereby indicating that patient factors are more important than the device variations. There are a few clinical situations that may influence the physician to favor one device over another. For example, a device with MRI compatibility or a removable magnet may be the best type to implant in a patient who will need future MRI scans. Some companies have multiple electrode configurations that are useful in an obliterated or malformed cochlea. To avoid unwanted facial nerve stimulation in the setting of cochlear otosclerosis, a perimodiolar electrode array may be preferable. Finally, multiple family members or friends may have implants, and being able to share experiences and tips is facilitated if the devices are the same.

Surgery

Implant Location

Surgery is performed under a general anesthetic without muscle relaxation to allow for facial nerve monitoring. The device location is marked out with the use of templates. It is important to place the internal device far enough posteriorly so that the processor that is placed behind the ear does not lie against it and render the underlying skin at risk. In an effort to avoid the complications associated with a skin flap breakdown, it is imperative to plan the skin incision in order to provide adequate exposure while both preserving tissue viability and avoiding the placement of a suture line directly over implanted hardware. The location of prior incisions should be taken into account. Most surgeons use a postauricular incision that may be extended slightly more superiorly than what might be typically used in a routine ear surgery. The current devices can be placed through a 3–4 cm cosmetically acceptable incision. In children, the scar is less likely to widen with head growth if it is located only in the postauricular area and does not extend up into the scalp. Due diligence should be rendered in manipulating the tissues of the skin flap to assure minimal trauma.

Mastoidectomy & Cochleostomy

After the initial incision, the periosteum is elevated from the mastoid and a mastoidectomy is performed. The facial recess is opened to gain access to the middle ear—specifically to the promontory, the round window niche, and the stapes (Figure 68–7). The chorda tympani nerve is preserved and the incus buttress can be left in place. According to the shape and size of the particular device chosen, a well may be drilled posterior to the mastoid cavity in the cortex to harbor the receiver–stimulator package. In children, this dissection is often carried down to the dura so that the device can be recessed. This better protects the device from trauma and is more cosmetically appealing. In adults, because of thicker bone, the device can be adequately recessed by removing bone to the inner table of the skull. The device can be placed in a tight subperiosteal pocket or secured by various means such as suture.

After the device is seated, a cochleostomy is then made inferior to the round window membrane with a goal of affording access to the scala tympani (Figure 68–9). Principles of minimally traumatic surgery should be employed in all cases in an effort to preserve structure and function. These atraumatic techniques include elements such as a more inferiorly based cochleostomy site to assure wide clearance of the osseous spiral lamina, minimizing access of bone dust or blood into the cochlea, avoiding suctioning of perilymph, slowing methodical electrode array insertion, and utilizing adjuvants such as corticosteroids and/or lubricants. Following successful insertion, small pieces of fascia or periosteum are used to carefully seal around the cochleostomy.

Hearing Preservation

Preservation of residual hearing is possible with a full insertion of a standard electrode array when minimally traumatic surgical technique is utilized, but with current techniques it cannot be reliably achieved. Hearing preservation, and thus electroacoustic stimulation are best obtained by implanting a shortened electrode that does not extend into and disturb the apical cochlear neural elements. Because these devices are quite small and flexible, residual hearing can be preserved in more than 90% of patients at the time of surgery. Unfortunately, there is the potential for delayed hearing loss, typically in the first 3–6 months, the etiology of this remains unknown and is actively being investigated. A direct round window membrane incision may be favored over cochleostomy by some surgeons for its potential to be less traumatic.

Intraoperative Electrical Tests

Depending on the device and the availability of audiology support, intraoperative electrical tests can be performed to confirm the proper functioning of the device. Evoked potentials and stapedial reflexes can be measured, which may be particularly helpful in programming with young children. Finally, if cochlear nerve action potentials can be recorded and a stapedial reflex is elicited, the surgeon can be quite confident that the device is indeed in the cochlea. After wound closure, a Stenver's view x-ray can be obtained for further confirmation of the device location and for reference in the case of future trauma or device migration (Figure 68–10). Patients may be discharged from the hospital on the same day, or they may spend one night in the hospital.

Cochlear Ossification

When cochlear ossification is encountered, various approaches may be used to ultimately attain satisfactory electrode placement. Often the obliteration involves only the first few millimeters of the basal turn, which can be removed with a drill or other small instruments. In these cases, the device can then be fully inserted. A preoperative MRI can usually predict this situation. For more extensive disease, other options exist such as partial electrode insertion, insertion into the scala vestibuli, or other more elaborate drill-out approaches. Split electrodes (ie, two distinct electrodes) have been designed so that one electrode can be partially inserted in the scala tympani and the second can be inserted in the scala vestibuli or further along the scala tympani.

Cochlear Malformation

Cochlear malformations provide another surgical challenge. Everything from finding the cavity to inserting and stabilizing the electrode may be problematic. Fortunately, specifically designed electrodes are available to facilitate implantation. A CSF leak should be anticipated and may require plugging the eustachian tube and packing the middle ear. Additionally, electrode insertion under fluoroscopic guidance may be of benefit.

Bilateral Implantation

The benefits of bilateral cochlear implantation have been studied in adults and children. Specifically, improvements in sound localization and speech understanding in noise have been noted. Bilateral implantation can be performed sequentially, and staged months apart or simultaneously in one operation. For children who are implant candidates, it is routine in many large centers to proceed with bilateral simultaneous implantation, even in children under 12 mo of age.

After the patient has healed from surgery, usually in 1–4 weeks, the device hardware is fully engaged and programmed. The initial programming is often done over 2–3 days. There are a myriad of variables that can be adjusted to improve the sound quality. After the first day, most adults report that speech sounds like static or voices sound either like “Donald Duck” or sound metallic in character. Amazingly, without any changes to the device, over the next 24 hours the sound quality improves. The brain somehow manages to adapt to the signal. This learning by the brain occurs mostly within the first 3–6 months, after which the rate of improvement in sound quality slows. Most adults have programming sessions two to four times in the first year, then annually or as needed.

Children (particularly infants) are more difficult to program because of the lack of a consistent feedback regarding volume and clarity. Objective intraoperative measurements are helpful in estimating hearing thresholds and comfort levels. It is obviously very important to not provide too much gain. Children are seen more frequently for programming. Programming is critical to the success of the device, and experienced audiologists are able to achieve better outcomes than less-experienced audiologists.

Cochlear implantation requires a surgical procedure under general anesthesia and therefore carries some risk. In particular, risks such as those encountered when removing a cholesteatoma or performing any surgery for chronic ear pathology do exist, including wound infection, facial nerve injury, taste disturbance, tinnitus, and balance problems. Overall, the complication rate of cochlear implantation has been reported as being 5–10%.

Wound Infection

The frequency of cochlear implant site wound complications and cochlear implant infections have been reported to range from 4.5% to 11.2% and 1.7% to 4.1%, respectively. Collectively, these comprise the largest subset of cochlear implant-related complications. It should be noted that implant site wound complications such as flap necrosis or dehiscence often lead to the development of infection and vice versa; thus, from the standpoint of prevention and treatment, these need to be considered together.

In general, cochlear implant infections come in two broad categories. By far, the most common type involves the postauricular soft tissues including the suture line, flap, and/or hardware itself consisting of cellulitis and/or abscess. Often these originate from a focus of scalp pressure necrosis caused by dynamic contact with an external object (including an excessively strong magnet) or by spread from a nearby infected cutaneous lesion. The second type of cochlear implant infection involves a bout of acute otitis media.

In the rare event that an implant recipient would present with a severe systemic or life-threatening infectious complication such as a brain abscess or unstable meningitis, immediate device explantation and surgical debridement are required. Fortunately, extreme complications are rare and most cases can be safely managed with intravenous antibiotics and local wound care.

Most uncomplicated cases of acute otitis media can be treated promptly on an outpatient basis with oral antibiotics and close follow-up to confirm a complete response to therapy, with more aggressive inpatient treatment reserved for those unresponsive to conservative treatment and those with negative risk factors such as a major congenital cochlear anomaly. The pathogens causing acute otitis media in the implanted patient population are the same usual pathogens encountered in the nonimplanted population.

In cases of infection unresponsive to prolonged culture-directed intravenous antibiotic therapy, hardware explantation may be required. It has been hypothesized that bacterial biofilms may play a role in some cases of implant infections and that the presence of a biofilm will render infection essentially impossible to clear with antibiotic medications and local wound care in many of these cases. During explantation, most surgeons advocate leaving the electrode array within the cochlea to prevent the development of obliterative soft tissue and/or new bone formation, thereby facilitating subsequent reimplantation.

Facial Nerve Injury

Facial nerve injury has been reported which may not be surprising because of the wide array of aberrant anatomy potentially encountered in this unique patient population. The anticipation of an abnormal nerve location and the use of intraoperative facial nerve monitoring should result in very few cases of temporary or permanent nerve injury.

Taste Disturbance

Drilling a posterior tympanostomy via the facial recess requires surgical negotiation of the chorda tympani nerve. At times, especially when the facial recess is poorly pneumatized or when round window visualization is otherwise poor, trauma to the chorda tympani nerve and associated taste disturbance may occur. This typically resolves by 6 mo but may be permanent in 2–3% of cases.

Tinnitus

Patients need to understand that the residual hearing in the ear with the implant may be lost and that a hearing aid may be of no benefit. Cochlear trauma from device insertion not only results in a loss of hearing but also may lead to or exacerbate tinnitus. When encountered in this setting, tinnitus typically lessens in time and often markedly improves after device programming.

Vestibular Dysfunction

Breaching the confines of the inner ear may also result in vestibular dysfunction with temporary, or more rarely permanent, balance difficulty. A patient who suffers an acute vestibulopathy should be provided vestibular rehabilitation therapy to maximize recovery. Benign paroxysmal positioning vertigo may occur in the postoperative setting and can be treated in standard fashion.

Device Failure

Although the implanted device has no moving parts to wear out, there are still instances of electronic malfunction or failure due to trauma. In fact, device failure is one of the most common postoperative complications. Most devices can be explanted and a new device is reimplanted without compromising performance.

Risk of Meningitis

The risk of meningitis in implant recipients has been scrutinized. Patients with inner ear malformations have a higher risk of meningitis pre- and postoperatively unrelated to the cochlear implantation. The role of the electrode design and its impact on the risk of meningitis has also been investigated and it is accepted that a particular historical design led to an increased risk. The Centers for Disease Control and Prevention has specific recommendations for vaccination of adult and pediatric implant candidates and recipients.

Subjective Measurements

Cochlear implantation, not long ago viewed as experimental, is now a proven treatment for SNHL in properly selected patients. Adult implant recipients with positive outcomes have seen benefits as far-reaching as a restored capability to communicate on the telephone (attained by roughly 60% of adult recipients) and the ability to converse without the necessity of lip-reading. More modest outcomes have included the improved reception of environmental sounds and augmented lip-reading capabilities. Other selected benefits that have been described include the treatment of tinnitus, the improvement of preimplantation depression, and a perceived overall improvement in the quality of life (reported to be as high as 96% of recipients in one report).

Rarely, in medicine, there is a procedure that has such a profoundly positive impact on the quality of life. Successful cochlear implantation is extremely rewarding for implant team members and patients alike. Yet, it is essential to stress that the outcomes seen with cochlear implantation vary widely both within given patient populations and among differing groups. Multiple factors have been shown to have a bearing on the degree of benefit obtained from implantation (Table 68–2). Although these factors are helpful in anticipating performance levels, additional unaccounted-for dynamics, which are difficult to gauge and recognize, do exist and account for about 50% of the variance in performance.

Objective Measurements

Open-Set Sentence and Word Recognition Scores

More specific objective measures in postlingual deafened adults following implantation include an evaluation of both open-set sentence and word recognition scores. Various reports have documented open-set sentence recognition scores of 60–70% and word recognition scores of 30–50%. Note that patient variability, evolving inclusion criteria, and ever-changing technological innovations render the objective analysis between various implant systems and processing strategies quite difficult. Overall, average performance continues to improve.

Measurements for Pediatric Patients

In children, results seem to have greater variability and are more difficult to measure.

Mainstream Schooling

A common goal (and one that has been frequently attained) for implant recipients in the very young pediatric population is to achieve communication abilities sufficient to allow enrollment in mainstream schooling by the second grade. In fact, an especially high number of children who have received an implant before age 3 have been known to eventually achieve age-appropriate speech recognition and production, with the most frequent success coming in the subset of patients who are younger than 18 months when they receive the implant.

Word Understanding

The objective markers of pediatric outcomes in postlingual deafened children (the minority of deaf children) include word understanding test scores 3 years after implantation that are documented to reach as high as 100%. The prelingual deafened child, who represents the majority of pediatric deaf patients, has shown to make slower and more variable improvements over a longer time span, including reported progress at up to 8 years after implantation. As previously alluded to, evidence seems to indicate that children do better if implantation is undertaken at the youngest possible age with the best outcomes usually obtained in children less than 2 years old.