After 12 years in the lab, Robert Greenberg has brought a viable retinal prosthesis to market -- and returned at least some visual acuity to sightless patients. The bionic eye has finally arrived.
June, 2010. Deep within the airy, new building that's home to the Manchester Royal Eye Hospital, a well-built 52-year-old man with short, fair hair sits in a darkened room. He faces a computer screen, although the dark glasses he's wearing make it impossible to tell whether he's looking at it. A shape appears onscreen and, as it does so, John Rose (not his real name) begins to move his head almost mechanically from left to right and back again. After a few moments he calls out: "Triangle." Another shape appears. His head scans, left to right, left to right. "Circle," he says. Next a large letter -- 22 centimetres high -- appears before him. Rose's head sweeps back and forth across the screen before he reads it out. More letters come, then words. Then smaller letters and smaller words, until they are just a few centimetres high.
As eye tests go, this may not seem remarkable, even taking into account the hindrance of the shades. But what makes this exercise truly extraordinary is that this man has no visual perception -- his eyes do not function. Without the glasses and the movement of his head, he would be completely blind.
Rose, a former surgical technician who wishes to remain anonymous, began losing his vision in his twenties and, besides some limited light perception, has been clinically blind for the last seven years. But he is one of the first of a few dozen people to have received a transformational new eye implant that has restored a limited amount of his vision. Rose's retinal prosthesis receives images fed to it from a tiny camera mounted within the glasses. His head movements help to improve the quality of what he sees by allowing him to take in more of his surroundings by increasing his field of vision. "I can see letters and three- or four-letter words," he says. "I hadn't seen these things for years, so the chance to see them again is amazing."
This implant, the Argus II, is a fully functional commercial product. By the end of this year, the regulatory paperwork will have been processed and the device is expected to have received clinical approval in Europe, making it the first retinal prosthesis to come to market.
The age of the bionic eye, it seems, is finally upon us.
The level of vision the Argus II offers is still far short of that which sighted people experience, producing just 60 dots of light in the field of vision, compared to the 120 million or so produced in a healthy eye. Nevertheless, in trials Argus II patients have all identified some benefits. For some it allows them to make out basic shapes, so they can see where people are when talking to them, or navigate through doorways; others can even shoot basketball hoops, sort laundry and, in some cases, such as Rose's, read print.
The principle behind the device has long been known: gently zap the retina using an electrode with micro amps of current, and it's possible to replace the function of damaged light-sensing photoreceptor cells. Stimulate the remaining healthy nerve cells in the retina, and sensations of light can be created in the visual field. Zap enough cells, and images start to take shape.
The effort has been demonstrated numerous times over the past three decades in animals and on humans under local anaesthesia in surgical conditions. Doctors such as Eberhart Zrenner, director of the Institute for Ophthalmic Research at the University of Tübingen in Germany, have taken this further. In 2007, Zrenner carried out 11 so-called "acute implantations" whereby the device is implanted in the patient's eye for a relatively brief period of three months before being removed again. But, until recently, researchers hadn't designed a device that could be left permanently implanted in the body.
Robert Greenberg, 42, cofounder, president and CEO of Second Sight, the Sylmar, California-based company that developed the Argus II, knows how hard it is to build a prototype that can be placed in the eye, even temporarily. The harsh saline environment of the human body is as corrosive as the sea: if you leave an acute implant in for too long, the body will destroy it. "Creating a device that can be implanted permanently is like building a television set that you can throw into the ocean and will still work, and continue to work for another 20 years," Greenberg says. For Second Sight to restore vision to the blind it took years of innovative engineering, tens of millions of dollars of investment, a dash of rocket science and a deep sense of purpose -- not just the kind motivated by financial profit, but also that driven by the promise of profound personal reward for someone with nothing to lose and everything to gain.
Like many great ideas, the Argus II began with the coming together of two minds. Sitting in his holiday home near North Fork, in the southern part of California's Yosemite mountains, Greenberg, 42, recalls that first meeting in 1998. Greenberg -- a qualified biomedical engineer and medical doctor trained at Johns Hopkins Medical School in Baltimore -- was working for the Alfred Mann Foundation, a California-based non-profit medical research group.
There he had helped develop one of the first cochlear implants, devices that help hundreds of thousands of clinically deaf people to hear.
The success of the implant had prompted one of its investors, Sam Williams, to approach Greenberg's boss, Alfred Mann, to ask if a similar device could be developed for the eye. Williams, then 77, was a talented and highly successful engineer and philanthropist who had made his fortune designing miniature engines for private jets and Tomahawk cruise missiles. It was a Williams jet that was to help SpaceshipOne win the $10 million X Prize in 2004 when it became the first private spacecraft to go beyond the stratosphere. Three years before his meeting with Greenberg, Sam Williams had been awarded the US's National Medal of Technology and Innovation -- other recipients of which have included Steve Jobs and Bill Gates.
Knowing that Greenberg had spent six years carrying out research in retinal prostheses as part of his PhD at Johns Hopkins between 1991 and 1997, Mann had called a meeting to bring the two together in Palm Springs. Greenberg arrived at Williams's home and was stunned by the view of the valley below. He made his pitch, outlining his earlier research on retinal cell-stimulation. "I talked about the work I had done at Johns Hopkins, I talked about some of the ideas I had at that point and what I'd like to do to take the research to the next step," Greenberg recalls. "Sam was an impressive man with broad technical knowledge. I remember being surprised by how quickly he grasped the issues we would have to tackle to build a successful retinal prosthesis." Greenberg finished talking. Now it was Williams's turn.
His proposal was nothing short of outrageous: Williams, a multimillionaire, wanted to bankroll Greenberg to develop a commercially available cure for blindness. (Greenberg won't reveal the scale of Williams's investment.) And he needed it done yesterday. "He wanted this to happen in his lifetime," Greenberg says.
Williams's demands were audacious, but Greenberg didn't doubt his sincerity for a moment. For all Williams's fortune and success, Greenberg knew that the picturesque view no longer held an allure for the engineer: Williams, a victim of the degenerative eye disease retinitis pigmentosa, was blind.
Keen to find a way to restore his own sight, Williams had previously funded university research into retinal prostheses but had grown frustrated by the slow progress. He had come to believe that the only way to speed up development was by making the search for a cure a commercial endeavour. "Sam was adamant about setting up a separate company," Greenberg says. "As an entrepreneur he thought that it was important to engage the employees with stock options and give them a focused incentive to produce a device. He felt that this would be a more productive way of approaching the project, and I think he was right."
And, in Greenberg, Williams had found the right man for the job.
Not just because of Greenberg's entrepreneurial streak (he had developed and sold software programs and medical devices even when still at high school). Nor was it Greenberg's biomedical-engineering expertise. What Greenberg brought to the table was a passion that matched Williams's drive. Long before their meeting, Greenberg had already made the decision to devote his professional life to making retinal prostheses a reality.
In 1991, while at Johns Hopkins Medical School, Greenberg had been given the rare opportunity to join pioneering retinal surgeon Eugene de Juan in carrying out some of the first experiments in stimulating the retinas of blind patients. "This patient first had one electrode put into his eye, and when the technician turned on the electronics he saw a spot of light. Then Eugene put in two electrodes, alternately flipping on one electrode or two, and the patient would either see one spot of light or two," Greenberg says. "At that point I was hooked. It was pretty obvious at that moment, at least in my mind, that it was going to be possible to create a prosthesis for the blind."
With funding from Williams, Mann and two other investors, Second Sight was set up in a matter of months. Greenberg was its first employee, and his former mentor, de Juan, and biomedical engineer Mark Humayun, then also at Johns Hopkins but now associate director of the Doheny Eye Institute at the University of Southern California, were brought in as consultants. "I spent the spent the first six months travelling around the world and hiring the best research team I could find," Greenberg says.
From the outset, Second Sight's approach was to focus on how to make a device suitable for chronic implantation. Greenberg was aware of the rigorous testing and scrutiny a device would have to undergo: in 1997 he had spent nine months working for the US Food And Drugs Administration, the formidable agency responsible for giving clinical approval to medical devices. So Greenberg knew only too well that in addition to the challenges he faced in building a chronic implant, if it was ever to make it to market, what lay ahead was bench-testing, animal trials and clinical trials to prove the device's longevity, safety and functionality.
A retinal prosthesis is made up of three main components. An external unit captures and processes images from a camera and feeds them to a second component within the body which translates these signals into electrical stimulations. These are then fed to the third component, an array of electrodes that sit either behind or in front of the retina. As a proof of principle the decision was made early on to use existing technology wherever possible, and in particular to borrow the tech used in cochlear implants. "With our first system, the Argus I, we essentially took a cochlear implant developed by our sister company, Advanced Bionics, and put a different electrode array on it," Greenberg says.
Adapting electronics designed for audio processing to handle video signals may sound an unconventional approach, but there were sound scientific principles behind the decision. The Alfred Mann foundation had alreadyinvested$50milliontodevelopthesechronically implantable devices and they were already approved for clinical use. With the clock ticking, it would save time as well as money.
These devices had a wireless means of getting the signals from the external unit to the processor inside the head by using an induction loop. This also provided a convenient way of powering the internal device without having wires leave the body.
But that still left the delicate task of developing the electrode array. "I expected that to be a six-month project -- but it took a couple of years," Greenberg says. The retina resembles a wet and very delicate piece of one-ply tissue paper. So the challenge was to make an interface that could be placed close enough for the electrodes to stimulate the underlying nerve cells, but without damaging the retina.
To do this the interface has to be almost as pliable as the tissue, so that it will conform to its shape. This ensures that each electrode sits as close to its target nerves as possible.
Adding to this complexity, the device also has to be completely sealed to protect it from bodily fluids and yet allow the electrodes to protrude without causing leaks.
With the help of a $25 million grant from the US National Institutes of Health, Greenberg says, in 2000 Second Sight eventually managed to solve these issues, although he won't reveal how. But he says it was problems like this that consumed a large part of company development time and resulted in Second Sight notching up around 100 patents for everything from techniques for biocompatibility bonding and encasing the electrodes, to motion compensation techniques for the video capture and signal processing for the video feed. During this period Williams was ever present, impatiently urging on the researchers. "He wasn't the only one, we had other investors breathing down our necks," Greenberg says. But Williams was different: his engineering expertise was invaluable.
On paper he may have been just one of the company's board members, but, according to Greenberg, he was highly involved in the research. "In the early days, he and I spoke often about technological issues that we were facing. Sam was a great sounding board and many times helped me think through the design challenges," Greenberg says. In developing the electrodes, some of the toughest challenges demanded answers from material science, such as finding ways to miniaturise metal and ceramic components.
Given that Williams had built his career on taking large objects, such as jet engines, and then making them smaller, his insight was invaluable."He had a lot of input about how we might solve these problems," he says.
After extensive bench testing and animal trials, the team was ready to begin its human implantations in 2002. The first Argus I patient was a 74-yearold retired concession-stand vendor from Maryland. "He was actually the first to volunteer for the intraoperative studies at Johns Hopkins many years earlier,"
Greenberg says. "At some point, he was told that if we ever had a long-term implant, he would be the first to receive it. It was very gratifying to have been able to keep this pledge."
The procedure was a success and, between 2002 and 2005, six further patients received Argus I implants. Although a proof of principle with only 16 electrodes (one for each dot of light or pixel the patient would see), the implants worked better than expected, allowing patients to make out basic shapes and detect edges and lines. In part this was thanks to a behaviour the patients had developed as a result of their disease.
With retinitis pigmentosa -- one of the biggest causes of blindness in the developed world, and the disease at which this technology was targeted -- loss of vision is gradual. Patients increasingly experience tunnel vision until, eventually, they see nothing. During this process they often resort to "scanning" in order to see their environment. "If you look through a pinhole and move your head around, you can build up quite a high-res image of the world around you," Greenberg explains. "And that's what these patients were doing with the Argus I." With the camera mounted on the glasses, moving their heads enabled them to get a better resolution than the device alone offered.
In theory Second Sight could have applied for approval and gone to market with the Argus I. But, despite requests from patient groups to do so, the company still didn't feel the technology was ready. For one thing, the cochlear-implant processor they used had been discontinued. Secondly, the implantation procedure was complex, taking eight hours and requiring four of the world's top surgeons -- Dennis Maceri, an LA-based ENT surgeon attached the cochlear implant; Michael Burnstine, then at the Doheny Eye Institute, California, ran the cable from the cochlear implant to the eye; and de Juan and Humayun attached the array inside the eye.
If it was to become a widely available treatment, a simpler approach was needed.
As far as Williams was concerned, such an approach was just around the corner. From the beginning Second Sight had been working on a second-generation device. Rather than using a cochlear implant, this device (the Argus II) was to have a dedicated processor which could handle more channels, or electrodes (to boost the resolution) and which would simplify the surgery by being small enough to fit within the eye cavity.
But, much to Williams's frustration, the Argus II took several years longer to realise than anyone had anticipated. This, Greenberg says, was largely because it involved such an enormous challenge: the equivalent of taking a device as complex as a pacemaker and shrinking it down to the size of an aspirin. "That was a truly monumental, world-changing leap in the neural-prosthesis field," he says. It involved miniaturising all the electronics and circuitry within a biocompatible watertight casing that would sit within the eye cavity on the outside of the eyeball. Simultaneously this had to allow each of the electrodes in the retinal array to be connected to it by a separate cable, again without any risk of seepage. "It went through several iterations before we arrived at the present design," Greenberg says. But what they ended up with made it possible to implant the device in just a few hours with only one surgeon.
In September 2006, a woman (Second Sight will not reveal her identity) in Mexico was the first to receive the device. Today there are 29 people around the world, including John Rose, who have had some of their vision restored thanks to this new implant. The 60 electrodes offer considerably better resolution than the 16 of its predecessor but, according to Rose, are still a long way from restoring full vision. Rose, who started to lose his sight back in his 20s and has been completely blind (with only limited light perception) for the last seven years, received his Argus II in June 2009. After a two-month period to let the implant bed in, it was switched on. At first, Rose was disappointed. "I was probably expecting to see more," he says, citing overly high expectations.
Initially the team just tested the electrodes, turning them on sequentially. "It was very obvious when they came on," he says. He would see yellowish-white balls of cotton shimmering in his field of vision. Following this, his surgeon Paulo Stanga, associate professor of ophthalmology at the University of Manchester, started to create shapes using the electrodes and eventually turned on the camera. "There has to be a lot of light about," Rose says, but he can make out objects and has since shown that in high-contrast conditions he can make out shapes appearing on a computer screen.
And with training, to help his brain adapt, he has since learned to recognise letters and words.
Rose still uses his walking cane for guidance, but says the implant helps him in daily life. "I use it getting around my flat, going through doors," he says. "I can see where chairs are. You can't always make out what something is, but you can see that there's something there, the TV and a picture on the wall. I can't see the picture, but I see its outline." Rose suffers confusion outside home because there is too much visual information for him to make sense of images, and it's hard to tell if something is near or far. "But when I look out the window I can see the cars on the road below."
Stanga says that it's a significant improvement on Rose's former vision: in clinical terms it's the equivalent of a two-category improvement, giving Rose the kind of vision he would have had about 15 years ago. "You have to bear in mind that retinitis pigmentosa has no treatment, so such an improvement is dramatic," Stanga says. "Until now, his vision would only get worse."
And, with Second Sight planning on developing devices with larger numbers of electrodes, the technology is likely to improve.
During the development of Argus II, the company experimented with thousand plus electrode arrays. But making such devices suitable for lifelong chronic implantation is incredibly difficult, Greenberg says. In the end it comes down to a trade-off: adding electrodes means increasing the number of individual wires connecting the internal processor to the array; keep adding them and pretty soon the device becomes bulky, unwieldy and increases the chances of seepage or something else going wrong. The way Greenberg sees it, Second Sight could have spent years trying to improve the resolution, or it could bring the benefits of partial sight to market now.
It's a roadmap modelled on the first cochlear implants. These had just one channel, enabling implantees to hear either sound, or no sound. With the Argus II, the aim is to provide patients with limited vision to assist them with navigation and orientation.
Techniques like multiplexing -- whereby multiple signals are sent down a single cable -- will contribute to improvements, Greenberg says, although even cochlear implants are only just beginning to do this. There are also ways to produce sub-pixel stimulation of retinal nerve cells, by using pairs of neighbouring electrodes to address cells that lie between them.
Other groups are pursuing Second Sight's market with very different approaches. Zrenner's company, Retina Implant AG, which was founded in 2003 near Tübingen, has a device which needs no camera, but instead uses the eye's working components to focus light on an array of electrodes behind the retina, each with its own light sensor. The device still requires a power supply similar to those for cochlear implants to boost the signal, but Zrenner says it removes the need for a patient to have to scan his or her head. And, because it doesn't require the electrode array to be connected to a separate processor via lots of tiny cables, Zrenner's chip has 1,500 electrodes -- which potentially offer greater resolution.
Although Zrenner is upbeat about the trials in the 11 patients who have undergone the procedure, saying that one can even read print, Stanga is not so convinced. "At this point the results of devices with a larger number of electrodes are disappointing," he says. According to Stanga, only one of the patients with a 1,500-electrode device experienced an improvement in vision on a par with the seven best Argus II recipients. And, since they were implanted temporarily, the long-term clinical benefits remain unclear. InStanga's opinion, the best improvements will come from improved signal processing and software, not necessarily an increase in the number of electrodes. Zrenner counters: "Argus II requires continuous and rapid head shaking to refresh the percept mediated by 60 electrodes, but the last three patients in our study perceived images continuously in a natural way. This is clearly a unique, huge qualitative difference."
If Zrenner is able to establish long-term clinical benefits, his approach may prove to be the way forward. That is because, unlike Second Sight's approach, Zrenner's allows the number of electrodes to be increased without increasing the number of wires.
Regardless of which technology prevails, the arrival of the Argus II on the market will be an important milestone, Zrenner says. The clinical benefits may now seem limited, but the technology has arrived. "The Wright brothers showed flying was possible," Zrenner says. "They first flew just 37 metres but showed the principle even though it would take another quarter of a century before Lindbergh crossed the Atlantic." Today, technology moves faster: within three to five years Zrenner hopes to have a 1,500-electrode device on the market.
As well as improved resolution there is the distant possibility of colour. Argus II patients tend to see in monochrome, or at least a yellow version of it. The retina contains more red and green light receptors than blue ones; stimulating them indiscriminately creates a yellowish-white effect. Greenberg says that in the long term the potential is there to target these nerves more precisely and produce colour images.
John Wyatt, an electrical engineer at MIT who is cofounder of the Boston Retinal Implant Project, a collaboration between the Massachusetts Eye and Ear Infirmary, Harvard Medical School and MIT to create engineering solutions to treat blindness, says better vision is likely to be achieved, but he is doubtful that full-colour vision will ever be possible. Currently, electrodes are relatively large compared to their target cells, so with each zap hundreds, if not thousands, of cells are stimulated simultaneously. To allow full vision the electrodes would have to stimulate cells individually, which is a tall order. "It's like learning to play classical piano while wearing boxing gloves," Wyatt says.
But, as technology improves, genuinely useful vision should be achievable, he believes. What remains to be seen is how well the visual cortex learns to adapt to these electrode stimulations. The signals are not biological in nature, so when groups of neurons are stimulated simultaneously the brain needs to switch on some cells that should be off. Cochlear implants have shown that our brains can adapt to these artificial signals, Wyatt says, but with vision it's a lot more complicated.
Wyatt maintains that what Second Sight has done in developing the Argus II and making it fit for chronic implantation is a remarkable achievement. "It's engineering that no one else has been able to do," he says. And forming a private company helped to achieve this: Wyatt has received a similar amount to Second Sight in public funding, around $30 million, but the additional private funding that Second Sight has attracted -- a figure Greenberg will not disclose -- may have tipped the balance. And it's not just the money that has an impact on the research: investors will expect steady progress, Wyatt says. With around 200,000 people suffering from retinitis pigmentosa in the US and Europe, there should be a big enough market to support a private-sector approach.
Sadly for Williams, all this comes too late. He died last year, just 12 months short of having his dream fulfilled. Williams had remained active within the company until 2008, despite battling illness. He was aware of and excited by the progress the company was making and of how close its product was to market. And even though it became clear that the technology would arrive too late for him, the legacy he was leaving gave Williams a great sense of pride. "It was important to him even if he wasn't going to personally benefit from it," Greenberg says. "He was very proud that he had contributed to something so important."
Sitting on his sofa, staring out at the Yosemite mountains, Greenberg doesn't come across as a pioneer who has helped to cure blindness. He looks worn out from more than a decade devoted to a single end. It is as if he has been working so intently on his goal that he hardly recognises that he's achieved it.
But, then again, Greenberg doesn't see this as a cure. "A cure would be to restore their normal vision. We're not on the threshold of a cure," he says. "But we are on the threshold of a new class of product that will allow these folks to get useful vision back."
Greenberg has every reason to be wary of boosterism. When the Argus II hits the market, initially in Europe, Second Sight expects to be inundated with patients desperate for the technology. Yet with a price tag of $100,000, it is likely only the very rich will benefit.
But it won't always be like that. When cochlear implants first came to market, they cost about the same as a retinal implant (when adjusted for inflation) -- today, they are about $50,000. As government regulators begin to endorse the technology, Second Sight will gradually be able to increase its production of devices and training of surgeons, making it accessible to patients who may not be able to afford them privately.
And that was the ultimate aim of the work, Greenberg says: to get this technology out of the lab and make it available to patients. It may have taken them longer than planned, and it may be too late for Williams, but to turn it round in just 12 years is a remarkable achievement." Sam would be very proud," Greenberg says. "He always had confidence that we were going to do it. And today I think he would be ecstatic to see that we have.
Duncan Graham-Rowels a technology journalist based in Brighton and a Wired contributor
This article was originally published by WIRED UK
