Closed-Loop Spinal Cord Stimulation

Dr Timothy Ray Deer

President and chief executive officer of The Spine & Nerve Centers in Charleston, West Virginia

Dr Tim Deer is President Emeritus of the West Virginia Society of Interventional Pain Physicians (WVSIPP) and past President of the International Neuromodulation Society (INS). Dr Deer is a former member of the board of directors of American Society of Interventional Pain Physicians (ASIPP) and has served as faculty, moderator and presenter on many important subjects at many national and international meetings. Dr Deer has published extensively on a wide range of topics including injection techniques, minimally-invasive disc procedures, intrathecal drug delivery, and spinal cord and peripheral nerve stimulation, as well as served tirelessly on many editorial boards. He has recently served as author and series editor on a number of textbooks relating to interventional techniques.

When Did Neuromodulation Begin and How Did It Evolve?

People talk about Mesopotamia, regarding electrical stimulation with fish, and my friends from Egypt often mentioned that as the origins of pain. Then it came through for many centuries where people, like Benjamin Franklin, was doing really early experiments on neurostimulation in America. We had the Leiden jar, we had galvanic stimulation with Galvani making muscles move. A lot of things were going on, leading to Mel Zuckerman’s 1965 gate control theory, which may be one of the mechanisms of stimulation. Now we’re learning that probably isn’t the main mechanism of stimulation for many areas we work in, but that was the main thought around what we call tonic stimulation, where you stimulate a nerve and make it change its way and function. So that began a great deal of research. In 1967 Norman Shealy publishes the first case in a cancer patient who was terminal with intrathecal stimulation and that grew into where we are today. It’s been an evolution since early times, but it wasn’t a sudden jump from one to one. It was a progression to a lot of developments, and I’ve certainly written about that in a couple of books in the past.

What Considerations Drive the Choice Between Spinal Cord, Dorsal Root Ganglion (DRG) and Peripheral Nerves as Targets in Neurostimulation?

If you had to pick between the spinal cord, the dorsal root ganglion and the peripheral nerves as targets for neuromodulation, the dorsal root ganglion is probably the most likely target to give a good outcome in most patients, because that’s where the peripheral nerve fibres come in. There’s a volume knob there at the DRG so you can turn down the pain -that’s proven to multiple animal models-. Now we think it may even work in things like nociceptive or joint pain, which before we thought was all nerve pain. The advantage of the peripheral nerve is that it’s minimally invasive, so I can do a peripheral nerve implant without much risk to the patient.

So the advantage of the DRG is that it’s probably the best target; the advantage of the peripheral nerve is it’s certainly a very low risk and we can go out to the nerve that’s injured and try to change things there. Then the spinal cord’s advantage is that some people aren’t appropriate for the dorsal root ganglion or the peripheral nerve because of the severity of the injury. So they each have their place. We try to think algorithmically and then follow the data. There are certain patients where you know that DRG is their best chance and you’re going to go straight to that target. If the peripheral nerve is their best chance, you’re going straight to that target like the shoulder for example.

Where Are We with Data-Driven Decisions in Neuromodulation Today?

I think a lot of people don’t use data-driven decision making. What we are trying to do is exactly that, we try to look at studies that have been done, and again we always need more studies to further define this. But let’s say, for example, god forbid Jonathan, you came to see someone like me, and you’d had a crush injury of your foot, and now you have some minor nerve damage. I could tell you without any hesitation, based on the data, that dorsal root ganglion is the best thing for you. If you’ve had a big back surgery and now you have severe back and leg pain, then it’s debatable. Is it the spinal cord, or is it the DRG at different locations? In both of those situations though, the peripheral nerve is probably not appropriate, because you have a severe nerve injury.

I think it comes down to looking at the literature and giving the best evidence-based discussion to the patient. Then it’s down to tailoring the therapy to that patient. For example, they may not match the study’s criteria that was done to show that great outcome. Reproducibility of studies is one of the most important things in our field. Can you reproduce the great study that was done in a normal patient population?

Why Do We Need Objective Correlates of Pain?

In the past, studies have always been about VAS, Visual Analog Scale, or NRS, Numeric Rating Scale. And so a patient comes into your office and you ask how’s your pain, well they may rate their pain too low because they want to please you, they may rate their pain too high because their knee hurts, even though you’re studying their arm. It’s really hard for the patient. With things like the promise 29, for example, we can look at the quality of life function, well-being and all those factors.

At the Pittsburgh sleep study, a recent closed-loop study we did, we show how we could improve sleep dramatically. If the FDA is hopefully moving in the same direction that we are, then we’re going to start seeing pain as the only outcome going away. We’re going to look holistically at the patient. Are we are changing their lives? Which really changes their families’ lives, work environment, and everyone around them. I think that’s a much better measure, and I think it will economically be better for society as I believe we’ll get better healthcare utilisation in that way.

Does Peripheral Neuromodulation Cause

Central Plastic Changes?

We published a paper recently that showed that if you assimilate a nerve in the periphery for several weeks, maybe even a couple of months, you can change the brain. You can change the actual brain function, so you may make permanent changes. Regarding the DRG, we published a paper showing that we’re changing the chemical milieu around the spine when we stimulate the DRG, and that might be why it helps nociceptive pain in the joints. It’s hopefully something I’m going to study very soon.

And then in the brain, we know from De Ridder’s work, that we can change the medial thalamus to change the way people suffer. So now there are well-defined central changes. I’ve had patients saying “we’re just covering up the pain”. We are trying to cover up your pain if you will so you don’t hurt as much, but we may be changing your brain and spinal cord by these targets. I think that will be the long-term benefit of these therapies as they get more and more evolved. We are early in the field; it feels like we’re a long way down the road, but I think we’re still early as we bring in things like virtual reality and other things to our field.

Is Software the Future of Neuromodulation?

The gate control theory would suggest that you’re just changing one impulse. And that’s why I said I don’t think that’s what we’re doing in the new waveforms. We’re actually changing the nerve itself, not just that one signal when you actually block that nerve. So I think that’s where we’re going, things like closed-loop, for example. We will see changes in artificial intelligence where these devices learn what Tim Deer or Jonathan does, so it changes with the person, and I think that’s going to be in the next generation we see going forward. Plus, working on some great new waveforms I can’t talk about publicly at the moment that may work differently in the brain. So I think we’re going to see another evolution of waveform software technology as well.

Dr Dirk’s been a leader in AI and data integration, so I’m working with Dirk on some of the ideas I have now. I think we’re going to see a collaboration like that. Dirk, a neurosurgeon, moving from Belgium and over to the Auckland area of New Zealand. So well, you have him, me in West Virginia, and we have some great basic scientists up in Madison, Wisconsin, we have the Europeans – all of us working together is going to be how we evolve this. I think the software will probably be more important than hardware, and this is how we develop new waveforms that change the brain and the cord that have never been seen before. We’re going to find a personalised therapy that really works in a person, and maybe other things don’t work. I think that’s exciting for patients who may not be suffering from pain right now but may later be injured and have something coming along that’s new for them all the time; it’s going to be an evolving progression.

How Can Urine Biomarkers Be Used to Measure Pain?

I believe every single target in the body where we’re using neurostimulation, over the next decade,  will have feedback. Whether it be peripheral nerve, DRG, spine or brain, with the exception of waveforms that work in areas other than where you put them. And that probably wouldn’t be a feedback loop mechanism we could look at at the current time.

Regarding multi-factorial outcome measures; I believe accelerometers haven’t been much help in the field so far. They’ve been pretty much what I would call a rock, it just lays there and changes with the body a little bit. Accelerometers don’t impress me. I think that would have been generation 1, now we are generation 50 already. But I do believe as we measure some of those things we talked about earlier, we’re going to have the help from the actual measurement of the real nerve, which is a feedback loop.

The other thing I think we’re going to get into, and I’ve published a few papers on, is urine biomarkers. And the importance of urine biomarkers for predicting someone’s outcome. In our paper, we showed that the biomarkers in chronic pain patients are abnormal in the urine, and before, we thought this had to be a certain test which is very expensive. But now we’re learning we can get urine biomarkers that may be predictive and then if you can change the behaviour, whether it be diet, vitamins, smoking or other factors with the exception of genetics, we can make actually changes based on those biomarkers and trail how they do. And I think that may help us also predict how people do with the outcome of these devices.

Why Is Psychological Assessment So Important

for Patient Selection in Neuromodulation?

It is importnant that we keep modifying and changing our selection process for patients. Now we’re seeing an 80% outcome in five years in a lot of therapies we’re offering; that’s a pretty good outcome. It used to be 50% of outcome in five years. Now remember, that they were all failing when they came in, so they were 0% good outcomes in the beginning. But if we get 80% of people doing well in five years after implant, that’s a pretty good selection. What I’d like to see us do is get to 95%. I think that the biomarkers and all those things come in will help, but the one area that’s lagged in this whole process has been the psychological assessment of patients. Because looking backwards you could have predicted failure in many patients because of other social issues like family, or drug issues etc.

This is the highlight of the interview. If you like to explore more, please visit our YouTube Channel.