On Innovation…

If you work in eyecare, it’s almost certain that you already know of John Marshall – his reputation precedes him. John is many things: educator, mentor, academic, entrepreneur and most definitely, a serial innovator. He’s accumulated a great deal of knowledge and experience of innovation over his career – so that’s precisely what we asked him about.

How important is experience to successful innovation?

It’s incredibly important. In my first “innovation exercise” with the excimer laser, I didn’t fully understand the value of what I had. I didn’t understand business practice, I didn’t understand financial input – and I certainly didn’t understand stock dilution! These are all things that you learn about as you innovate.

When I first had the idea of the excimer laser, despite having the original patents of the technique, I found it virtually impossible to raise money from conventional sources in the UK – people there didn’t believe in the idea of cutting away parts of the center of the cornea. But in the US, we managed to raise plenty of money by approaching their money markets, but this came at a cost: a big equity involvement from the investors…

Here’s the thing: most research workers are naïve and might not understand the value of what they have. They soon learn (like I did) that if you take company money, they will want a big share of it and may want to “gobble up” the intellectual property (IP).

But if you have a novel idea and good IP, it is relatively easy to raise money, and, importantly, if you can move forward quickly with the idea, then you shouldn’t encounter too many problems. Then there’s the “flash-to-bang” time.

What do you mean by “flash-to-bang”?

How quickly will your idea become essential to the community at large? This is “flash-to-bang”, and it matters in successful innovation. An example of this is vitreous fluorophotometry. The technique involved giving patients an intravenous bolus of fluorescein, then measuring the diffusion of the dye from the retinal architecture through the vitreous, into the anterior chamber – more dye would diffuse through in diseased eyes than in healthy eyes. Although it was a simple and elegant concept, a brilliant device, and was designed by two very bright people, it didn’t get anywhere. On the other hand, there’s the example of OCT – also a brilliant device designed by very bright people – which was introduced and was instantly successful. So in one case a brilliant idea didn’t really get anywhere, and in the other case a brilliant idea led to huge commercialization. It all relates back to this “flash-to-bang” time – the community embraced OCT as they immediately saw how useful it would be to their daily practice.

Are diagnostic devices easier to develop than therapeutics?

Both are hard. Treatment innovations are generally faster, as diagnostics can be very difficult. Not only is it difficult to secure funding to develop diagnostic innovations (because of concerns about obtaining reimbursement later), there’s also the problem of patients in the diagnostic process. If you need them to respond with an answer, you’re likely to have issues. Patients can be the biggest variable: they are trying to please you rather than doing the test, and frankly, if you can remove them from the process, diagnostic innovations may have more potential!

Here’s an example of a diagnostic technique with great potential that avoids the patients trying to please you with their answer: genetic screening. In the old days, you’d take a buccal swab, send it to a lab to perform PCR, and six weeks later you’d have your answer. Today, that’s still the same process, but it still takes 48 hours. Imagine something like red eye, where a simple test could discriminate if the cause is bacterial, viral or fungal, potentially even by using a coated strip that you push into a smartphone. That’s a wholly objective answer right there.

That doesn’t mean that we can stop treatment innovation – and we certainly need more of that. But even treatment innovation can be hard – and I think part of this is down to the procedures involved with innovation. Why don’t we have any new antibiotics? It isn’t because there aren’t any good laboratories capable of doing the work, but rather that the regulatory hurdles are so huge and expensive that a lot of companies don’t see the process of developing antibiotics viable, or in some cases, possible. And this is just one of the many ethical issues with innovation.

So what are the ethical issues attached to the innovation process?

When is a treatment really an experiment, and when does an experiment become a treatment? This is very difficult as an ethical problem, but it is also a scientific issue – and the problems only increase when commerce becomes involved. Let’s say a hypothetical new device has gone through fundraising and a number of limited trials, and receives endorsement from a regulatory authority. The inventor can then start charging money – charging patients for something that may only give them limited help yet is going to cost them a lot of money.

We need to define success, and we must set the end-points at a level where true benefit to the patient is taken into account. It seems to me that there is a situation where any limited improvement in a patient’s vision is deemed to be successful, and the media adds to this hype through describing treatments as “sight restoring.” But is being able to see a bit of light coming through a window worth $100,000?

Another issue is at what point do you release Mark I devices knowing that you have also developed a Mark II version? Commercially, you need to release the Mark I device onto the market to demonstrate that your innovation works, but your research and development team know that there is something better coming along. So what do you do? Do you let people know that there are these further developed technologies in the pipeline? Or do you let people know that this is where we are now, but there will be an improvement?

The ethics of innovation can be complex, and I think it is why we have seen a number of so-called treatments which are really experiments coming out.

Does regulation help?

Regulatory frameworks exist to look at efficacy and safety, but the elements of these frameworks are not necessarily policed well. For instance, there is a huge regulatory framework for breast implants, but a company in France chose to ignore it and it led to a significant number of problems. It’s no good having tables of performance and efficacy if at the end of the day you’re not going to enforce them – enforcing them after the problems have arose is just too late.

In my opinion, all these factors have led to the situation we have now, where companies with a pre-production product will have key opinion leaders speak at meetings. By doing this, companies are getting information out there without going through the necessary peer review process – other than standing in front of an audience.

How can you avoid innovation pitfalls?

If you are an innovator today, there are a lot of pitfalls. One lesson I have learned is that surgeons and scientists make the world’s worst CEOs – you shouldn’t try to do everything yourself outside your area of confidence. If you have an idea, there is research technology available to help you determine if others have similar ones. There are then pitfalls with IP. Filing for a provisional patent gives you a year’s grace but also makes others aware of what you are doing. A good pre-patent tip is to get your lab books notarized by a lawyer – whilst it isn’t a patent, it is a legal document confirming that you had the idea at that time. After filing a patent, there are several routes to take. You can keep the product to yourself and start paying for IP protection, or you can look for a commercial partner in the field to license the technology.

There are also commercial aspects to this. The real issues start when you get to a trial stage – clinical trials these days just cost a fortune. We have seen them migrate from the US to Europe, and now to China, simply because they progress faster and at a lower cost than the “creeping death” we now see in countries with more regulatory restrictions. With trials, you want to look at safety and efficacy, but instead of trying to look at progressing your device, you end up focusing on getting through all the steps in the process.

But it can be fun too, right?

Yes. This was a brilliant idea that went to the military – retinal biometrics as barcodes. An infrared camera that hunts for the highest infrared reflection in your eye – which is your optic disc. The camera goes into scan mode, and it scans the transition of blood vessels as they go from myelinated fibers of the nervous background to the tissue. Barcodes these days are 13 digits in length. Everyone has between roughly 14 and 17 vessels coming out of their optic nerve. If you measure the relative proportions of vessel to non-vessel – essentially a black bar then a white bar – around the circumference, you got a perfect barcode, which was unique to each eye. So this is a very, very high security biometric.

One cracking invention was an idea that came to me in a pub. I walked in and said to the barman: “I’m gasping for a cup of tea, any chance?” The answer I got was “Nope.” I asked “Why?” He said “It’s no, unless you want to come out to the kitchen.” I went “Fine, I don’t mind going to the kitchen… but why?” He said, “You have no idea! People come in at the end of lunch time, ask for a cup of tea, I come back, and they’ve nicked all the money out of the fruit machine!” And then I thought, well, it’s so simple to design a device to give you a lot of warning. So I put a little device together that had a light trap, so if anyone handled the system in an improper way, a huge klaxon went off. The barman said “It’s fantastic!” – and and it actually made money – it was adopted by several elements of the gambling industry!

John Marshall is the Frost Professor of Ophthalmology at the Institute of Ophthalmology in association with Moorfield’s Eye Hospital, University College London, UK. He invented and patented the excimer laser – an innovation which has been used in over 50 million procedures world-wide.