Active middle ear implants (AMEIs) are sophisticated technologies designed to overcome many of the shortcomings of conventional hearing aids, including feedback, distortion, and occlusion effect. Three AMEIs are currently approved by the US Food and Drug Administration for implantation in patients with sensorineural hearing loss. In this article, the history of AMEI technologies is reviewed, individual component development is outlined, past and current implant systems are described, and design and implementation successes and dead ends are highlighted. Past and ongoing challenges facing AMEI development are reviewed.
Key points
- •
Over the last 20 years, there have been significant advances in active middle ear implant (AMEI) design.
- •
Many modern devices provide comparable objective audiometric performance with optimally fitted conventional hearing aids and afford a more natural, clear sound, with minimal feedback.
- •
Despite continued progress, the field of AMEIs is very young; there are still many theoretic benefits that have yet to be fully realized.
- •
With continued device innovation, expanding indications, and improvements in reimbursement, AMEIs will undoubtedly continue to develop.
Introduction
Hearing loss is one of the most common chronic disabilities and affects up to 48 million people in the United States. Approximately 1 in 4 adults older than the age of 60 years suffer from bilateral hearing loss, most of which is not surgically reversible. Despite a high prevalence of hearing impairment, only 15% of hearing aid candidates use conventional aids on a regular basis, making hearing loss the single largest chronic sensory impairment that remains untreated. This statistic seems surprising, because binaural hearing loss has been shown to significantly affect patient quality of life, leading to social isolation, anxiety, depression, and even cognitive decline. Common objections to hearing aid use frequently include cost of purchase and maintenance, social stigma associated with hearing aid use and concern over self-image, lack of audiometric benefit in those with severe loss or poor word recognition, and discomfort.
The primary impetus for AMEI development is the desire to overcome many of the shortcomings that are inherent in conventional hearing aid design. To receive sufficient usable gain, patients with advanced hearing loss require a tight-fitting ear mold, which can lead to occlusion effect, discomfort, and ear canal irritation. Despite sophisticated sound processing strategies, distortion and feedback still plague even the best hearing aid designs at high output levels. AMEIs receive acoustic signal and directly stimulate the cochlea through coupling of the long process of the incus, stapes suprastructure, footplate, or round window. Because many AMEIs bypass the external ear canal and do not use a speaker for signal amplification, they effectively circumvent symptoms of occlusion effect and offer the potential for improved sound clarity and enhanced functional gain.
Introduction
Hearing loss is one of the most common chronic disabilities and affects up to 48 million people in the United States. Approximately 1 in 4 adults older than the age of 60 years suffer from bilateral hearing loss, most of which is not surgically reversible. Despite a high prevalence of hearing impairment, only 15% of hearing aid candidates use conventional aids on a regular basis, making hearing loss the single largest chronic sensory impairment that remains untreated. This statistic seems surprising, because binaural hearing loss has been shown to significantly affect patient quality of life, leading to social isolation, anxiety, depression, and even cognitive decline. Common objections to hearing aid use frequently include cost of purchase and maintenance, social stigma associated with hearing aid use and concern over self-image, lack of audiometric benefit in those with severe loss or poor word recognition, and discomfort.
The primary impetus for AMEI development is the desire to overcome many of the shortcomings that are inherent in conventional hearing aid design. To receive sufficient usable gain, patients with advanced hearing loss require a tight-fitting ear mold, which can lead to occlusion effect, discomfort, and ear canal irritation. Despite sophisticated sound processing strategies, distortion and feedback still plague even the best hearing aid designs at high output levels. AMEIs receive acoustic signal and directly stimulate the cochlea through coupling of the long process of the incus, stapes suprastructure, footplate, or round window. Because many AMEIs bypass the external ear canal and do not use a speaker for signal amplification, they effectively circumvent symptoms of occlusion effect and offer the potential for improved sound clarity and enhanced functional gain.
Historical groundwork for active middle ear implant development
Although there are significant variations in modern AMEI designs, all devices have substantial component overlap with conventional hearing aids, the primary differences being that a coupled mechanical transducer replaces the receiver, and at least 1 component is surgically implanted. Therefore, AMEI development has benefited significantly from improvements in hearing aid design, including ear mold integration, device ergonomics, component miniaturization, microchip technologies, advancements in speech processing, microphone technology, and battery design. Although a comprehensive review of conventional hearing aid history is outside the scope of this article, the most important advance to affect AMEI development began with the application of integrated circuit technology developed by Jack Kilby of Texas Instruments in 1958. Subsequent improvements such as the microprocessor, borrowed from the computer industry, permitted implementation of developing complex sound processing strategies. The first hearing aids with digital speech processing became commercially available in the 1980s. Most hearing aid users own a fully digital aid that is capable of flexible gain processing, multichannel wide dynamic range compression, feedback suppression, and digital noise reduction.
The early history of active middle ear implant technology
All modern electronic devices designed for hearing rehabilitation include several basic components: a microphone, power source, signal processor, and a specialized end component designed to deliver conditioned signal to a portion of the auditory system. Examples of a specialized end component include a cochlear implant electrode array, an auditory brainstem implant electrode pad, the acoustic receiver of air conduction hearing aids, and an osseointegrated screw for implantable bone conduction hearing aids. The primary component that distinguishes an AMEI from other designs is the transducer, containing an input sensor and output actuator, converting electrical signal into mechanical vibration. Although many types of mechanical transducers exist, devices used in middle ear surgery must be lightweight, energy efficient, reliable and durable, biocompatible, and operate with high fidelity. Over the last century, piezoelectric and electromagnetic-based transducers have emerged as suitable options for use in AMEI devices. The development of these technologies is discussed later.
Development of Electromagnetic-Based Middle Ear Transducers (Also Known as Electrodynamic)
Alvar Wilska, a Finnish physiologist, is credited with the first attempt at mechanical stimulation of the auditory system through the use of an electromagnetic driver. In 1935, Wilska published the results of early experiments, in which he placed 10-mg iron pellets on the tympanic membrane and subjected them to a magnetic field generated by an electromagnetic coil inside an earphone. By adjusting the rate of oscillation, Wilska was able to reliably control the frequency of sound produced. He determined that vibratory amplitudes as small as ∼0.1 nM (the diameter of a hydrogen ion) could elicit sound thresholds at 1000 Hz. Wilska himself participated in many of these early experiments, despite the discomfort that was generated from the heat and physical contact of the metal substrates on the tympanic membrane.
Subsequently, in 1959, Rutschmann devised a method of fixing a tiny permanent magnet to the tympanic membrane at the umbo with water-soluble glue. By introducing an alternating current ranging between 0.1 and 3.0 A, pure tones between 2000 and 10,000 Hz could be generated. In his experiments, 2 individuals reported the ability to decipher an audio broadcast. Rutschmann realized early on that this technology carried a significant advantage over air conduction hearing aids for patients with advanced hearing loss by eliminating the problem of feedback at high output levels. In addition, he envisioned that with further refinement, this technology could bypass eroded ossicles or a diseased tympanic membrane and directly stimulate the inner ear through coupling of the oval or round window.
In the 1970s and 1980s, several researchers performed similar experiments with placement of a fixed magnet on the tympanic membrane outside the middle ear space. In 1973, Goode and Glattke performed a series of experiments on 5 individuals using an Alnico V magnet fixed at the umbo, which was driven by an electromagnetic coil located on the postauricular skin. One of the 5 individuals was evaluated during a middle ear operation for tympanic neurectomy, allowing the investigators to examine differences in audition with magnet placement at the umbo, long process of the incus, and the oval window. From their work, the investigators found that tympanic membrane loading did not seem to significantly affect air conduction thresholds, and word recognition score performance using electromagnetic induction was comparable with conventional audiometric testing.
In 1988, Heide and colleagues reported an important modification to previous studies by replacing a postauricular transducer with an in-the-canal electromagnetic induction coil located millimeters from a magnet fixed at the umbo. A mean functional gain of 17.5 dB was achieved, with no significant differences in word recognition compared with the users’ own hearing aids. Similar to previous experiments, patients cited benefits including a more natural sound without feedback and improved performance in the presence of background noise.
Two types of electromagnetic-based systems have been developed: the electromagnetic and the noncontact electromagnetic configuration. The noncontact electromagnetic design consists of a rare earth magnet, neodymium-iron-boron (NdFeB) or samarium-cobalt (SmCo), which is fixed to the tympanic membrane, ossicular chain, or round window membrane. A separate energizing coil, located millimeters away, creates a fluctuating magnetic field, which causes the magnet to vibrate in synchrony with incoming electrical signal, resulting in displacement of cochlear fluid and auditory stimulation. This configuration carries several theoretic advantages, including limited ossicular loading from the single small magnet, a unidirectional push of transmitted energy that replicates the natural action of middle ear energy transmission, and a noncontact coupling of the transducer. The primary disadvantage of this system is the short distance and coaxial alignment that must be kept between the transducer and the target magnet, because the strength of the magnetic field is lost inversely to the cube of the distance.
In contrast, the (contact) electromagnetic configuration incorporates an energizing coil and magnet together in a single housing, which is physically coupled to the ossicular chain or round window membrane. The fluctuating magnetic field results in magnet oscillation, creating vibrations. Although this design is usually more complex than the contactless electromagnetic system, it is generally more efficient and is not limited by the requirements of coaxial alignment and close proximity.
Development of Piezoelectric-Based Middle Ear Transducers
Piezoelectricity was first discovered by Jacques and Pierre Curie in 1880, after observing that certain solid substrates developed an electrical charge proportional to an applied mechanical stress. The piezoelectric effect is a reversible event, whereby an applied electrical current can also result in a reproducible temporary deformity. In 1954, the metallic oxide-based piezoelectric material, lead zirconate titanate (PZT), was developed by Jaffe and remains one of the most commonly used piezoelectric materials. When a voltage is applied, PZT crystals are deformed by approximately 0.1% of their original dimension. Thus, small PZT ceramics can be used as middle ear transducers to create predictable microvibrations to drive the ossicular chain.
Most piezoelectric transducers use a ceramic monomorph or a bimorph system. The monomorph design uses a single ceramic platform, which results in expansion and contraction after voltage application. In contrast, the bimorph configuration uses 2 stacked crystals oriented in reverse polarities. When an electrical charge is applied, the platform bends, or oscillates, back and forth. In 1984, the RION Device was the first commercially approved piezoelectric-based AMEI to be implanted. Since this time, several additional AMEIs have used this technology, including the Implex TICA and the Envoy Esteem.
Electromagnetic-based active middle ear implant systems
Experimental Noncommercially Approved Device Designs
Semi-implantable middle ear electromagnetic hearing device
In 1986, Maniglia and colleagues at Case Western Reserve University began investigating one of the first contactless electromagnetic-based AMEI systems. Through industry collaboration with Wilson Greatbatch (Clarence, NY), several successive models were developed and tested in animals. An early electromagnetic prototype incorporated placement of a target magnet on the stapes but required incudostapedial joint disarticulation. Realizing the drawbacks of permanently altering the ossicular chain in an otherwise healthy middle ear, the primary objective for subsequent designs was to develop an implant that did not require contact coupling, but rather, worked through a contactless system. Theoretically, such a design would allow the patient unimpeded use of a conventional hearing aid should the device fail, even without device explantation. Furthermore, because the actuator would not be rigidly connected to the ossicular chain, there would be less risk of ossicular erosion.
Early on, Maniglia and his coinvestigators determined that an electromagnetic system using an air-core actuator would provide the best functional characteristics to achieve their goal. Because of the significant technical obstacles associated with fully implantable designs, a semi-implantable device was created, using an external microphone, radiofrequency amplifier, transmitting antenna, and battery. The external unit was designed to sit in a postauricular soft-tissue sling that was created using local rotational skin flaps. The internal component was to be implanted via mastoidectomy with atticotomy. A supporting frame was then secured to the mastoid cortex, and the transducer was placed within 0.5 to 1.0 mm of an NdFeB magnet, which was secured to the body of the incus with titanium-bone cement ( Fig. 1 ).
Using a cat model, Maniglia and colleagues found that the semi-implantable middle ear electromagnetic hearing device system did not interfere with normal acoustic hearing and afforded a mean 22 dB gain. Aside from antenna breakage, which was believed to be a problem unique to implantation within a cat, the device proved durable (9.6 months) and biocompatible. The results of these experiments led to US Food and Drug Administration (FDA) investigational device exemption approval for a limited clinical trial in May, 1996 for patients with moderate to severe sensorineural hearing loss (SNHL); however, to our knowledge, no further data were published regarding this device.
Round window electromagnetic system
Similar to several other contemporary AMEI investigators, Spindel and colleagues recognized early on the importance of creating a coupling system that would bypass a healthy ossicular chain. Spindel’s research team at the University of Virginia sought to develop a system that would provide direct cochlear activation through vibromechanical stimulation of the round window membrane. To investigate their prototype design, an animal study was devised, including 19 guinea pigs. Using a skin surface electromagnetic coil and an NdFeB target magnet on the round window membrane, the investigators reported a high degree of correlation between acoustic and electromagnetic stimulation using auditory brainstem response latency and amplitude data. Furthermore, the magnet did not seem to have any deleterious effects on normal acoustic hearing, and, because the design was not coupled to the ossicular chain, there was less risk of reverse energy transfer to the tympanic membrane, minimizing the possibility of reverse feedback. The investigators acknowledged difficulty with determining the best method for permanent attachment of the magnet to the round window membrane. Although this prototype design has yet to be adapted to clinical applications, the investigators showed the feasibility of round window membrane coupling.
Electromagnetic ossicular augmentation device
In 1986, Kartush and Tos, at the Michigan Ear Institute, collaborated with Smith & Nephew Richards, (Memphis, TN) to create an electromagnetic-based partially implantable AMEI that used an in-the-canal electromagnetic coil with a custom ear mold housing. These investigators participated in an FDA clinical trial using 3 different target magnetic configurations. In the first phase, 6 patients with bilateral sloping, high-frequency SNHL had a 30-mg to 45-mg SmCo magnet secured to the tympanic membrane with glue, approximately 2.5 to 3 mm from the ear canal transducer. The results of this trial showed audiologic outcomes on par with optimally fitted hearing aids, and most patients reported no feedback, a more natural sound, and a subjective performance improvement when in environments with competing background noise.
In the second phase, an electromagnetic ossicular replacement device was created, using a total ossicular reconstruction prosthesis (TORP) design fitted with a target magnet. Patients were required to have mixed hearing loss associated with inactive chronic ear disease. In total, 7 national and 3 international coinvestigators participated. In their 1995 report, a preliminary summary of the first 50 implanted patients was published. At the time of analysis, 26 patients were still using their device, 8 ears had device extrusion, 4 ossicular replacement devices were displaced, and 1 patient noted intermittent electromagnetic interference. Once again, audiologic outcomes were not significantly different from optimally fitted hearing aids; however, patients cited no problems with feedback, which permitted enhanced usable gain in addition to improved performance in noise. In a 2002 report, the long-term results of a 9-patient Danish cohort were published. Although initial audiologic outcomes were promising, at last evaluation, none of the patients was using their device. In addition to device extrusion, the primary reason for nonuse was difficulty with properly fitting the in-the-canal electromagnetic driver.
The phase 3 trial allowed for the implantation of 10 patients with SNHL using a target magnet, which was fitted to the undersurface of the tympanic membrane. At the time of the last report, 3 patients had undergone implantation, and no audiologic outcomes were provided. Although more than 50 patients were implanted using these technologies, device fitting remained problematic, and with time, funding was lost.
EarLens tympanic contact transducer
In 1993, Perkins and Shennib patented the EarLens platform (Redwood City, CA). This design used an SmCo magnet, which was embedded in a soft silicone lens held to the tympanic membrane by surface tension after placement of a thin layer of mineral oil. The primary advantage of this technology includes a less invasive means of magnetic fixation, which can be easily removed if needed. The initial design using an induction loop worn around the neck proved to be inefficient, prompting the development of a small in-the-canal induction coil placed several millimeters from the EarLens magnet. In 1996, a feasibility study was performed, including 7 patients evaluated over a 3-month period. Tympanic membrane loading resulted in a ∼5 dB loss; however, the maximum mean functional gain was 25 dB. There was no evidence of tympanic membrane irritation over the course of follow-up. A 16-patient follow-up study was published in 2010, using a refined in-the-canal electromagnetic coil and a 9-mg gold-coated disk magnet. Again, the EarLens system was well tolerated, and the ear canal transducer provided patients with enough output to reach thresholds of 60 dB HL. A recent modification includes use of a behind-the-ear processor containing a low-power laser diode that emits coded infrared light, which, in turn, activates a small tympanic membrane electrokinetic actuator embedded in the soft lens platform. At the time of writing, the EarLens system has not sought FDA clearance.
MED-EL Vibrant Soundbridge (formerly the Symphonix device)
The Vibrant Soundbridge (VSB) was the first FDA approved AMEI system for implantation of patients with SNHL, receiving approval in August, 2000. Initial device development began in 1996 and was pioneered by Geoffrey Ball, a cofounder of Symphonix Devices (San Jose, CA). The relatively slow adoption of the VSB after FDA approval led to the dissolution of Symphonix, and in March, 2003, the VSB technology was purchased by MED-EL (Innsbruck, Austria). Several years later, the VSB received the European Union CE Marking for treatment of conductive and mixed hearing loss in adults (2008) and children (2009).
The VSB consists of 2 primary components: an external audio processor and a surgically implanted vibrating ossicular prosthesis (VORP). The external audio processor contains a microphone, signal processor, telemetry coil, and a replaceable battery, all housed within in a single unit. The external unit is designed to sit against the postauricular skin, in the hairline, through magnetic attraction with the implanted receiver. The original external device, the Vibrant P, used analogue processing. Since that time, several generations of digital processors have been developed; the current fifth-generation design, the Amadé Audio Processor, offers a directional microphone, 3 programs to choose from, and several new sound processing features designed to minimize wind noise and attenuate background noise. The device uses a nonrechargeable zinc air battery, which requires replacement approximately once per week.
The implanted VORP comprises 3 functional units, including a receiver, demodulator unit, and an electromechanical floating mass transducer (FMT); the footprint of the implanted system is similar to modern cochlear implants ( Fig. 2 ). The implanted receiver, containing a magnet and coil, collects signal data transcutaneously, which are processed in a demodulator unit and are subsequently sent through a flexible conductor link to the transducer. The 2.3 × 1.6 mm FMT consists of 2 polymide-coated gold coil wires wrapped around a hermetically sealed titanium casing. Within the transducer housing sits an SmCo magnet, which is suspended by a set of silicone elastomer springs. The application of an alternating current induces a magnetic field that stimulates movement of the encased magnet, resulting in vibration of the entire FMT unit. The VORP can be implanted via a mastoidectomy with facial recess approach or through transmeatal access. With either approach, a postauricular incision is performed to create a well and trough for the implanted receiver and demodulator. Most commonly, the FMT is attached to the incus with a prosthesis clip; however, several other FMT coupling options exist, including round window, oval window, and TORP or partial ossicular reconstruction prosthesis vibroplasty.
Since its clinical debut, there have been more than 50 publications on the VSB, making it the most heavily clinically studied AMEI. Most data show good short-term and long-term audiometric performance comparable with optimally fit hearing aids; however, subjective patient-centered outcome data consistently favor the VSB over conventional hearing aids.
Ototronix Maxum (formerly the SOUNDTEC direct system)
In the early 1980s, the Hough Ear Institute, located in Oklahoma City, began a series of experiments that led to the development of the SOUNDTEC Direct Drive Hearing System, becoming only the second AMEI to receive FDA approval, in September, 2001. Initial research focused on the development of an optimal target magnet design. Various NdFeB magnets were tested in several locations, including a donut-shaped design over the stapes suprastructure, a magnet placed between the malleus handle and promontory, and one positioned between the stapes capitulum and the malleus. In 1988, 5 patients with moderate SNHL were implanted with a target magnet placed at the incudostapedial joint. Oxidation of the magnet occurred as a result of moisture exposure during preimplant preparation. A redesigned implant using an SmCo rare earth magnet was reimplanted in 4 of the 5 original patients. However, despite audiologic benefit, 6-month testing showed a significant decline in unaided high-frequency hearing thresholds; therefore, device explantation was required.
Over the following 6 years, improvements in device design led to an FDA-sanctioned phase 1 clinical trial in 1998, including 5 patients with moderate to moderately severe SNHL. Upgrades from the earlier prototypes included a stronger and lighter magnet, a hermetically sound titanium laser welded canister, and a wire clip for attachment at the incudostapedial joint, designed to provide optimal coaxial alignment with the in-the-canal electromagnetic coil. Implantation involves a ∼45-minute procedure, using a transcanal approach. The incudostapedial joint is carefully divided, and gentle back-traction on the incus is performed while the ring wire is slipped over the capitulum. The incus is then repositioned over the stapes capitulum to allow for fibrous union. Patients are usually fitted with an ear mold coil assembly with a behind-the-ear processor 10 weeks after surgery ( Fig. 3 ). The results of the first clinical trial showed a 50% (∼15 dB) improvement in functional gain, and ∼20% improvement in speech recognition scores over the patients’ previously worn hearing aids. Furthermore, subjective measures, including sound quality and patient satisfaction, were superior to conventional amplification.
