Cochlear Implants: My Perspective

By William F. House, D.D.S., M.D. (Edited by David House)

Chapter I, A Brief History of Cochlear Implants

Every time history repeats itself, the price goes up.

~Anonymous

IT HAS BEEN MORE THAN A DECADE since the FDA approved for marketing the 3M/House cochlear implant, in November 1984, and amid national fanfare, as the first device ever approved to replace a human sense.

At the time I remember feeling that implants had finally arrived and that the future was now open for their rapid development and widespread use. I was premature: It is only in the last 2 or 3 years that implants seem to have become truly accepted, and progress, as compared with my expectations, has been quite slow.

Many factors prevented this rapid progress, including technological barriers which have only recently been overcome. But the space program, and particularly the moon shot, demonstrate clearly that determination, whenever it is not fettered by unreasonable hindrances or shattered by disunity, is easily able to overcome technologic barriers. With regard to these sorts of hindrances and the further development of cochlear implants, there remains an undercurrent of feeling that new approaches will only be detrimental to already established devices.

But these conditions will, I am confident, prove as temporary as the other barriers which previously slowed or stopped progress in this area.

The Sixties

The sixties saw a number of developments which had a significant impact on the development of implants. It was a time of great changes in otology generally.

The operating microscope, for example, opened up vast new possibilities. The tympanoplasty of Wullstein and Zolner, the stapes mobilization of Rosen, and the stapedectomy of Shea were new and exciting. I was busy pursuing new surgical approaches (intact canal wall and facial recess), exploring endolymphatic sac surgery and doing some early acoustic neuroma work. As well I remained interested in the possibility of restoration of hearing by electrical stimulation of the cochlea. During middle ear procedures under local anesthesia I was able to observe the percepts of patients when small electric currents were introduced to the promontory.

But technical barriers proved frustrating. During the early sixties I implanted several devices in totally deaf volunteer patients. Unfortunately these were rejected due to lack of biocompatibility of the insulating material. However, during the short time that these devices worked, it was obvious to me that this was an opening salvo to the conquest of sensorineural deafness.

During the same decade, Robin Michelson (at the University of California at San Francisco), Blair Simmons (at Stanford) and I formed a sort of implant group and we were actively pursing both animal and human research. There was much skepticism and even outright hostility that we, as clinicians, should be invading the cochlear domain of the neurophysiologists. The obvious feeling was "keep your hands out of our cochlea". Simmons later wrote:

While skepticism engendered by claimed miracles is healthy, outright denial that a genuine research problem exists is not. While my 1964-65 experiments were in progress I contacted a least six of the most prominent researchers in speech coding, and others in auditory psychophysics. None of these persons were willing or interested in suggesting experiments which might have helped define speech coding strategies for the future. I got the distinct impression, perhaps colored by a little personal paranoia after the first few rejections, that most everyone was either incapable of thinking about the many problems involved or would rather not risk tainting their scientific careers. I do not believe this problem has disappeared completely in the subsequent 20 years. [3]

Jack Urban

One of the most fortunate developments in cochlear implant development was that Jack Urban, a very innovative engineer, became interested in cochlear implants and teamed up with me to ultimately make cochlear implants a clinical reality.

We each brought different skills and differing outlooks to the process. My orientation was the selection of the patients and the surgical approach for implants. Jack applied his genius for electronics to the problems we faced.

At the time research money was completely unavailable, apparently because of the prevailing feeling that implant research "should not be done." I remember being bitterly disappointed when my request for funding was turned down by a prominent otologic research funding foundation. Fortunately for us all, Jack made his shop and electronic expertise available at no charge. I firmly believe that without Jack, cochlear implants would have taken many more years to develop. Many of us owe him an unpayable debt of gratitude: Jack died of cancer in 1985.

First steps

I will explain in more detail in the next chapter, but it is important to mention that we believed, at this point in time, and we did all of our early work starting with the assumption, that we would have to stimulate the snail-shell shaped cochlea discretely in order to mimic its natural function. That is, in a naturally functioning ear, the cochlea "sorts" sounds by their frequency along its length (with the highest frequencies nearest the round window) so that, as far as we know, the nerve fibers apparently responsible for a given frequency are all found in one place. It makes sense, then, that we should have to stimulate these groups of dendrites to cause the brain to interpret a given stimulation as having a certain frequency.[4]

This is known as the "tonotopic" (or sometimes as the "traveling wave") theory, and we believed it.

Because we wanted to stimulate only a portion of the cochlea with a signal which we could control completely and monitor continuously, and because we wanted to be able to inject a wide variety of inputs, Jack developed a through-the-skin pedestal for a direct-connection plug. This was connected internally to a five-wire electrode which had been inserted into the cochlea through a facial recess approach I had developed years earlier for chronic ear surgery. These systems had a plug-in button extending through the skin, and therefore offered complete control over separate stimulation of and direct connection of equipment to each electrode. Although there was a common ground which was placed external to the cochlea, the fact that we had direct access to any of the wires meant that we could use any of them as a ground or active electrode.

Because we had been repeatedly told that electrical currents would destroy the remaining neural tissue of the deafened cochlea, we moved with extraordinary caution, and stimulated only during brief sessions in the lab.

Three patients accepted this hard-wired device, and helped us in our trials of many different processing schemes. We worked most extensively with Charles (Chuck) Graser, a high school teacher who had been deafened by streptomycin, and who was an excellent observer of the hearing sensations that were generated through his system. It would often take months to develop a new processor, incorporating a new signal scheme, and sometimes within a couple of hours working together with Chuck, we might discover that the filter system was not right.

Many different systems of stimulation were tried, within the limits of the circuits we had available. Although we started with the idea that the sound signal would have to be separated into frequency bands, each of which would then have to be "presented" to different parts of the cochlea, in practice we could not make it work this way. For example, we tried a "vocoder" circuit, which shifted the higher frequencies lower before presenting them.

But among all the alternatives tried, the modulation scheme which Chuck felt produced the best sound, involved putting precisely the same signal into all electrodes, amplitude modulated on a 16 KHz carrier. Graser kept coming back to this as having the most natural sound, and our other patients agreed.

As time passed, evidence continued to accumulate which told us that, within the limits of the technology we had available, this would be the system of choice. I found that I had to let go of the tonotopic theory: it just did not work. As well, we saw absolutely no evidence of adverse neural effects as a result of the testing process.

Finally Jack and I decided to bite the bullet and give Graser a hard-wired, wearable device. He was delighted. In May of 1972, for the first time, he was able to walk out of the laboratory and perceive the sensation of sound. I called Barbara, his wife, daily for weeks. As each day of use continued to produce sounds that he found very useful, we became ever more convinced that cochlear implants had a great deal to offer.

But one swallow does not make a summer, and each success establishes conditions which enable further efforts. So Jack and I decided to make an entirely implantable device and to implant 8 or 10 patients that I had selected and who had talked to Graser. Because we had discovered that the best sound, as reported to us by our first few patients, was produced when the same signal was injected into all the electrodes, we decided to use only a single, short electrode. By all the evidence we had, nothing more was needed.

This was the next step, and we took it.

First frictions

I remember at the time of these studies requesting to present the very preliminary anecdotal findings on Graser a national meeting. I was turned down on the basis that reporters would be at the meeting and their reports of the implant would cause otologists to have to contend with a flood of patients with unrealistic expectations.

Finally, however, in 1973, The American Otological Society Saint Louis meeting held a session on cochlear implants.

Dr. Nelson Kiang, a very prominent neurophysiologist for Harvard, expressed the belief that a single-electrode device, such as all of our patients were then using, would only produce a kind of buzzing, Morse code-like sounds, and from a theoretical standpoint, this was a very reasonable position. Dr. Kiang felt strongly that, if an electric field was generated around the neural tissue in the inner ear, the nerve fibers would all fire, go into a refractory state, and then fire again, repeatedly, for as long as the stimulation lasted. [5] This would result in a buzz sound that turned on and off as the current was turned on and off.

For those who had not actually seen any implant patients, this belief regarding the limitations of cochlear implants was widespread.

The Journal which resulted from the meeting printed Dr. Kiang's thoughts, as well as contrasting comments by noted researchers such as Dr. Merzenich of San Francisco. Dr. Merzenich had seen some of our patients and had begun work with Michelson. He said:

Dr. Kiang's remarks suggest little or no discriminative hearing can be generated from a single-electrode pair. However, it should be pointed out that these subjects do have some discriminative hearing in the sense that small differences in stimulus frequency can be detected. And subjects describe sounds which they hear as 'tones'. This must be faced up to and explained. These and other qualitative observations on hearing in these subjects have also been made by Dr. Simmons and others. They are bonafide, as you will appreciate the first time that you see a patient. They simply must be explained in terms of neural mechanisms. [6]

I remember hearing once that, according to standard aerodynamic theory, a bumblebee cannot fly. Apparently no one has successfully convinced the bumblebee of this, however, and these sound theoretical arguments regarding the response of the neural system likewise seemed not to be able to convince our patients that they were not hearing something useful.

As the saying goes, "there are none so blind as those who will not see." Despite repeated requests to observe the patients as Dr. Merzenich had done, there were those whose minds were made up and they continued to insist that no benefit was possible with a cochlear implant.

I remember one remark by a scientist at the 1973 meeting who said: "If I tell you that a lead balloon will not fly, and you go out and build a lead balloon and it does not fly, what have you learned?" I could not help but remark that I had flown to St. Louis in a lead balloon. (Two Wrights had not been proven wrong.)

Multiple electrodes

But eventually, of course, this sort of objection faded. It became clear that regardless of what we thought we knew about the neural system, in this instance we were wrong: for obviously much more than Morse code was being heard by these patients.

The conflict between the expectations of theory and the demonstrations of clinical practice has continued to this day, but its next manifestation during this period was found in the growing support for the thought that multiple electrodes were the wave of the future. This thinking was based in the tonotopic theory.

A few months after the American Otologic meeting, a conference entitled "The First International Conference on Electrical Stimulation of the Acoustic Nerve as a Treatment for Profound Sensorineural Deafness in Man" was held at the University of California in San Francisco. [7] In the foreword to the printed proceedings, it states that the purpose of the conference was:

To demonstrate to the scientific and otolaryngologic community the very marked limitations of the present devicesƅ

The implication was that then-existing implants would never provide anything more than limited contact with the environment through noise (becoming aware, for example, of a car horn) as contrasted with assistance in understanding speech. The foreword further said:

[Michelson and Merzenich's] work also suggested that the remaining obstacles to a practical implant, namely; multiple electrodes, multi-channel receivers, etc. could be solved in a relatively short time using standard neurophysiologic techniques and laboratory animals.

The thoughts contained in the foreword to the proceedings of the first major meeting on cochlear implants represent some of the attitudes that carried throughout the seventies.

Merzenich and Michelson were working on a multiple-electrode implant, and the common thought seemed to be that "multi-channel" devices [8] were just around the corner, and "single-channel" implants should not be pursued. In other words, solely on the basis of tonotopic theory, it was felt that since single electrode systems could not provide discrete stimulation of limited areas of the cochlea, they therefore could never provide frequency discrimination or access to speech; and since "better" devices would be widely available soon, we should stop using single-electrode devices. These thoughts were not merely implied: a vote was taken at the meeting, and the majority of those present felt I should no longer use single-channel implants.

Indeed, many apparently believed that it had already been demonstrated that multi-electrode implants, as represented at this conference by the as yet unimplanted Clarion device, were the only practical approach. However, it must be borne in mind that at that time there were only a few patients using single-channel implants and there were no patients using any form of multi-channel device: so this unfortunate belief had no possible basis in fact. Beyond this, it bears mention that it took almost 20 years before the Clarion, developed by the San Francisco program of Merzenick and Michelson, became available.

Other attitudes

Another position represented at this conference was that animal work was all that was needed to solve the major problems regarding implants. This perspective fueled the controversy over human implantation.

It is difficult to see, however, how work with animals could assist with solving many of the pivotal research problems for a system which is intended to provide access to speech.

Beyond the fact that the alternative was, essentially, non-existent, the criticism seemed to ignore the fact that we exercised utmost caution, worked only with volunteer adults who had demonstrated that they had no measurable auditory function, and who were well-aware and fully advised of the risks.

There were others who felt that all implant work was useless. A prominent ENT department head stated: [9]

I have the utmost admiration for the courage of those surgeons who have implanted humans, and I will admit that we need a new operation in otology, but I am afraid this is not itƅ

This reluctance to break new ground was not an isolated thought. I remember in 1974 at the American Otologic meeting joining Walter Work for breakfast. (At that time, he was the head of the ENT department at the University of Michigan.) After some preliminary discussion, he said to me, "Bill, you know that the cochlear implant is no better than vibro-tactile devices."

I protested that I had read the vibro-tactile literature starting with Gault [10] in 1926. At that time there were no wearable vibro-tactile devices and I pointed out that they were not practical because the speech signal vibrations were overwhelmed by the background noise. The three-year study by Geers and Moog [11] of 13 matched groups of children in the same educational setting, one group continuing on with hearing aids, one with a vibro-tactile device and one with a 22-channel implant, has finally put the matter to rest: vibro-tactile devices cannot provide sufficient speech information, and implants do.

These sorts of attitudes did not disappear overnight, even after considerable success had been clearly demonstrated. The 3M/House single-channel cochlear implants in children had been on-going for several years, and many of these children were demonstrating excellent speech reception. Yet in a 1984 news magazine article, a well-known department head and pediatric otolaryngologist was interviewed on the topic of cochlear implants in children:

"There is no moral justification for an invasive electrode for children." Speaking for himself, he says he finds the cochlear implant a costly and 'cruel incentive', designed to appeal to conscientious parents who may seek any means that will enable their children to hear. "It's a toboggan ride for those parents, and at the end of the ride is only a deep depression — and you may hurt the kid."

(Medical World News, June 11, 1984, p. 34)

Meningitis

This doctor may have held these views in part because of a fear of meningitis. At a major meeting the same otologist quoted above stated:

If one child develops meningitis [from the cochlear implant] it will not have been worth it.

In other words, the concern, in children particularly, was that infections such as otitis media would spread along the electrode to the cochlea, and from there to the spinal fluid, causing meningitis. [12]

I reasoned that the risk was not high because we had not seen otitis media progress to meningitis in the many thousands of stapedectomy patients, who had a wire running from the middle ear into the cochlea's oval window. By this time we also had considerable experience with implants in adults, including the early cases where a plug penetrated the skin barrier directly, and the wires to which it was attached then penetrated the cochlea. While adults are less susceptible to middle-ear infections than children are, an increased risk would surely have demonstrated itself in some manner. There had been no such effect.

Still, this was a concern. In our first 5 or 6 years of implants in children we constantly queried pediatricians and parents about otitis media in the implanted children. The children who developed otitis media were successfully given the same course of treatment as before the implant. There were no reports of meningitis, and there was no increase in the child's susceptibility to otitis media. And so this concern subsided as well.

Multiple electrodes, revisited

But the criticism which was then (and to a degree it remains today), most pervasive and apparently persuasive, is the thought that single-electrode implants can never provide patients with the auditory information that multiple-electrode implants can.

As I will demonstrate below, there is no evidence on which to base this conclusion, but it is one of those ideas which seems so right, so logical, that it has gained great strength in spite of its meager diet, which is devoid of facts.

It is difficult to trace the precise genesis of such ideas, but we can point to a few places where it was fostered.

The Bilger Report

In 1975 the National Institutes of Health (NIH) asked Dr. Michelson, Jack and I if they could sponsor an independent evaluation of our patients. We readily agreed and our patients, who were by that time enthusiastic about their implants, also agreed to this. A research contract was awarded to Pittsburgh Eye and Ear Hospital for this study.

This extensive report, published in 1977, became known as the Bilger report. [13] This was a study of what were then the world's total population of cochlear implant patients: 13 adults. Eleven of these used an early version of what was later to become known as the 3M/House implant, and 2 used the Robin Michelson's U.C. San Francisco implant, which never became a production unit. [14] Thus, all 13 patients had implants which used a single electrode.

This report for the first time firmly legitimized cochlear implants, but the authors went considerably beyond clinical observation in order to speculate about how and how well implants work, and what should be done in the future.

For example, they assumed that single-electrode implant patients would never be able to understand speech. On page 4 of the report they state:

The psychoacoustic protocol was designed primarily to specify the nature of auditory discriminations possible with present-day auditory prostheses and did not stress tasks that would require the subjects to provide an absolute identification of the stimulus (e.g., repeat the word), since it is well-accepted that subjects using [these] auditory prostheses cannot understand speech with them.

It is worthy of note that what was then "well-accepted" had no scientific basis for being widely assumed to be true, if indeed it was. Further consider that the implication of the statement seems to be that certain experiments did not need to be done: because certain limits of implants were well-accepted, these limits did not need to be clinically tested. What is truly surprising is that some of these first few patients were reporting and demonstrating speech reception, albeit limited. One may take refuge in the fact that reality has never been subject to popular vote, or the earth would today be flat.

On page 9 of that same report, under Conclusions the authors state:

To the extent that the effectiveness of single-channel auditory prostheses has been demonstrated here, the next step lies in the exploration of a multi-channel prosthesis. A multi-electrode implant is technologically feasible, but that electrode assembly must be situated in the impaired auditory system in a location that permits stimulation of independent groups of auditory neurons. (Whether such a multi-channel prosthesis will allow the deafened person to understand speech, as such, remains to be seen.) In the development of multi-channel prosthesis, it is hoped that more attention will be paid to stimulus coding and to the external electronic package than characterized the development of single-channel prostheses. [15] Until such multi-channel prosthesis become a reality, one must consider the question of whether or not it is reasonable to continue implanting single-channel prostheses.

It is difficult to understand how such sweeping conclusions can be reliably reached on the basis of the information then — or even now-- available, and in fact the statements made offer their own proof that these conclusions were based on the authors' assumptions, theories and biases, and not on clinical observation.

To it's credit, the report had a very positive effect on acceptance of the general idea of implants. The conclusion that cochlear implants work (offering more than 'Morse code' input) could no longer be denied, even if their ultimate utility remained to be discovered. At the same time, however, my opinion is that the report tended to suppress exploration in the field by introducing terms and providing conclusions, based solely in theory, which had no underpinning of solid, data and clinical observation. That is, whether or not it can be traced to this report, it at least pre-figured the very odd and pervasive tendency in this field to develop beliefs which are unsupported by any facts whatsoever.

These beliefs became associated with the use of certain terms, and primary among the set of terms which were promulgated to poor effect were "multi-channel" and "single-channel". The idea behind these phrases was appealing: again, that it is somehow necessary to supply the damaged auditory system with information separated by frequency band, and delivered to the "appropriate spot" within the cochlea.

These terms were, of course, conjured out of the air, and although this would have been a useful act of creation if they had been placed in a context of exploration, they were instead introduced as absolutes: "single-channel" implants were inferior or primitive, [16] and "multi-channel" implants were the necessary future.

In sum, in spite of the fact that these conclusions had no basis whatever, they were repeated by others as if they were factually based, and whether from this or other sources, these or similar ideas have become entrenched as a set of popular truths, which are now "well-accepted".

Progress in spite of it all

Even with the pejorative image of single-electrode implants, interest in implants grew, and progress was made. By the mid-seventies numerous teams were established to study and develop implants.

During the late seventies and eighties a number of problems had been solved. One significant problem which had come up involved keeping the external processor centered over the implanted hardware.

All internal devices worked on the principle of induction. That is, a current had to be induced in a coil of wire implanted under the skin. Besides delivering a signal, the external processor, in essence, had to deliver the power for the internal electrodes. If the external power/signal coil was not centered over the internal coil, induction was poor, and there was a significant loss of signal.

Many approaches to keeping the two coils matched up were tried, such as attaching the external devices to an ear mold or to glasses, but all patients noted significant fluctuation of the signal even with slight movements of the jaw or when laughing. Dr. Dorman in Oklahoma and Steve Waldron, an engineer I was working with in California, came up with the idea of including a magnet in the internal coil that could serve as a kind of anchor for a matching magnet in the external coil. Such magnets have now become standard in all implant systems.

With the development of solutions to such problems, and with the increase in knowledge which previous studies provided, we realized that further possibilities had opened: specifically, they made possible the implantation of children.

And so in the early eighties we began implanting children.

At the time this was seen by some as an unwarranted risk, or even an unethical act. I have always thought that concern about unknown dangers was clearly justified as long as questions existed, but I could not understand why the well-known and profound liabilities of deafness itself were never factored in.

Deaf children face a number of wrenching difficulties and extraordinary challenges. The human community is characterized by speech, and that speech is summarized by a mother saying "I love you" to her child, and the child hearing and responding in kind. If we allow opportunities to offer this birthright to pass away, unremarked, if we fail to act to correct a problem when we have the means at hand, we have violated our oath as healers. Implanting children had nothing to do with glory; it had to do with responsibility.

Major concerns evaporate

In any case, this step became possible because by the late eighties, virtually all of the major concerns about the long-term success and safety of cochlear implants were largely resolved.

The first concern was that long-term, 18-hour-a-day electrical stimulation of the cochlea would eventually destroy the neural tissue. This concern faded entirely as animal studies were done, [17] [18] [19] and as the person-years of stimulation accumulated and there were no reports of implant failure due to loss of the ability of the VIII nerve to be stimulated. Even though no one has used a cochlear implant for more than 25 years, if we are going to see VIII nerve damage it would be appearing by now. Beyond this, a number of temporal bones have been donated by long-term implant wearers to be studied after their death. No hint of damage traceable to electrical stimulation has appeared in these. [20] Thus, by now it should be clear that cochlear implants are a lifetime solution, even for young children.

Fears concerning meningitis have also faded, for the reasons outlined a few pages previously.

A further concern was ossification: rabbits implanted with cochlear electrodes developed ossification of their cochleas and this seemed to wipe out the VIII nerve. Again, our experience with adults in post-operative polytome X-ray studies had not demonstrated any hint of this problem, but we needed to be wary.

Animal models, however, do not always provide us with a clear picture what will happen in humans. For example, I remember that in the early days of my practice, after stapedectomy had been practiced for several years, experimental stapedectomy in cats showed widespread destruction of their cochleas. At the 1959 meeting of the American Otological Society when the paper about this was presented, many otologists got up and said (in effect): "I have 10, 20, or 50 stapedectomy patients and I have not seen the problems your cats have developed."

In any case, concerns about ossification also subsided.

The Nucleus Implant

During this same period, work outside the USA was progressing, most notably in Australia where Clark and colleagues were developing a multi-channel cochlear implant that, in the last half of the eighties, was to become the single-most used implant in the world under the name "Nucleus Multi-channel Cochlear Implant". [21]

The commercial success of the Nucleus device signaled the final acceptance of implants as assistive devices. As implant patients became more numerous, and many clinicians and teachers of the deaf had first-hand observation of these patients, more and more accepted that implants are here to stay, have very few risks, and are very beneficial. For these reasons increasing numbers of cochlear implants are now being recommended.

It's an interesting saga, but I never dreamed it would take so long.