Optimizing Electrical Dose in Spinal Cord Stimulation


Physician Scientist, Coastal Research Institute

Dr. Krishnan Chakravarthy completed his undergraduate studies in math and finance at the University of Chicago. Later, he would enrol in the City of Buffalo’s lengthy MD/PhD programme, where he would spend three years researching on viral immunology at the CBC to create vaccinations for the 1918 flu. Before deciding to pursue a career in anesthesiology and realising his passion for patient care, he launched his first startup while still a graduate student. He joined the UCSD faculty and stayed there for the last five years before moving on to establish his own practise and launch a private research institute in southern California.

Where does SCS (Spinal Cord Stimulation) come from?

Jonathan: Does anesthesiology naturally lead you towards spinal cord stimulation? Perhaps you could give us a brief historical account of the origins of spinal cord stimulation.

Krishnan: I think it’s a fantastic question there’s this famous analogy “history kind of repeats itself” how we decide things in human history in terms of the applications of things and knowledge getting translated to new technology. 3000 years ago the Egyptians thought about using electricity as a way of pain control so the classic example was dipping patients in water with an electric eel that would send electrical stimulation at the point. Gait control didn’t even have words or acronyms to describe it, so around the late 60s Wall and Melzack came up with this basic concept that if you use gait control theory the basic principle is that ascending and descending inhibition through your peripheral nervous system all the way to your cortex. Everything in our body runs through electric current, nerves communicate through electricity there’s a level of chemical action potentials that are generated every time ions move across gradients so it isn’t surprising that it took a lot into the 60s to really coin that theory but to the credit of Norman Sheely, who was the first neurosurgeon who actually took that concept and applied it, where he placed an electrical lead in an oncology patient and essentially tried stimulating. He got about a day and a half worth of pain relief in palliative care patient so he had placed those leads in the subdural space so the point in that was the first instance somebody took a concept or theory that was largely at that point accepted that pain control was based on gait control and he applied it in a clinical sense and in the last 70 to 80 years we have seen an incredible cascade of innovation that’s surrounding Neuromodulations. It started with the movement of leads from the subject rule to the epidural space then you really came to the process study which was this idea that overlapping parasthetic stimulation to the area of pain and this tingling sensations patients feel when you stimulate was correlated to better pain control which really was the dogma till around the early 2000s when you had sub-parasthetic stimulation with high frequency therapy and with that you’ve seen a kind of a proliferation of different ways that people talk about stimulation whether it’s in the delivery model with closed-loop stimulation where you’re sensing something and delivering or you’re changing the actual parameters around stimulation.

At the end of the day, the practical part of it is that when you think about what are you actually stimulating, what are the variables really only four variables that you look at pulse-width, amplitude, frequency and impedance. Impedance is just a byproduct of where you’re actually putting the leads, if you’re putting in the epidural space versus a different space. It’s been amazing how the combination of clinical and industry has really shaped different patterns that are being put into the nervous system with the idea that you’re really trying to solve long-term therapy durability and or looking at the impact of how we keep these patients on these therapies for 3-5-7 years after implantation

What are the main impediments to the development of neuromodulation therapies?

Jonathan: I think you put your finger on a super important point, and the way it seems to me sometimes is that in pharmaceutical development, you have static therapeutic entities, it’s a drug and then it gets put onto the clinical treadmill and goes through phase 1 phase 2. With neuromodulation you know it’s a bit like one way of thinking about it is kind of it’s like a programmable drug, it’s morphable, it could change and so you have this changeable entity, and it does kind of throw a spanner in the works when considered with the traditional model of having a static entity that we can check this check that 

Krishnan: I think every speciality and field goes through phases of development and they kind of coined the word electroceutical. We’re in a renaissance time in the electroceutical industry but what we do need to do is we’re getting to the point, where everybody looks at how do I change one aspect of the care delivery model, am I changing the programming, am I changing the architecture of what I’m delivering and how I’m delivering it but to really get down to it, the big challenge for the neuromodulation space we still continue to ask this critical question is if I change the dosing paradigms around pulse-width, amplitude and frequency no matter whether you look at what was happening with traditional tonic stimulation, that’s been the most validated in terms of all the different therapy waveforms, the data supports about a 10 to 40 percent exponent rate and that’s a large number.When you think about the cost of these devices going into use today, how do we better design more standardization across development so that we’re addressing the critical issue on maintenance of these therapies because at the end of the day cost savings is what drives health care you can have a great therapy but it has to make cost sense for the entire different parties involved, so there’s definitely a component of that. Dosing is a part of it and how we approach dosing in terms of spinal cord stimulation but we also need to address the innovations that are happening in the space. Can they be more cost-effective for more durable therapy?

What is the importance of investing in big data collection?

Jonathan: If you had 10 million dollars to deploy would you put this behind tools for big data collection or tools for more accurate personalised models of patients?

Krishnan: I love that question because that’s exactly where I’m going. We are looking at a really interesting campaign, it’s called better together and its aim is to look at cost acts as a community around the neuromodulation space, the basic idea is we’re looking at a low-cost system today that’s around $200 that will be free to the patient but eventually a patient could buy off a website similar to an iPhone. The beauty of what we’re trying to do is we’re using large data gathering across using AI/ML technology so basically, today if Jonathan uses waveform x for a certain pain descriptor, does Krishnan in India use the same waveform or a different waveform and what can happen is as you collect all of that data and aggregate tens and thousands of other data points you start to see whether specific waveforms apply with certain dosing patterns to certain demographic patients for back pain and that’s really powerful, the only way you can do that is with a low-cost system you can’t do that when your implantable asp costs are $20,000 because the care delivery model doesn’t support big data either we are on to something here 

we have a solution for that and it’s now rolling out in the United States and what’s amazing about what we’re doing is the incredible amount of data that we’re collecting around patients. It’s their prescribing patterns, their compliance where they’re placing these units to stimulate and what’s ironic about that is that we’ve got our first sets of patients in India and Africa to study the demographic and cultural changes to see how people use these different devices in different parts of the world it’s really interesting I mean it just lends itself to a very interesting comment that a lot of pain space says we have great randomized controlled trial data but how come it never makes sense in my own practice why can’t I make it match the same set of outcomes? well in a controlled study setting you’re extremely particular about your inclusion criteria no two patients walk in the door that are identical so I completely agree with you, I think big data is the future for pain medicine and the way it drives therapy adoption and to some extent I would love to see waveforms that go off patents because then we can actually get to the real innovation which is cost access and hardware. it’s too complicated the delivery model today, all of the technologies take so much money to develop it doesn’t make sense for the broad masses so my point is if the patent office never patented 10 000 hertz we’d have a totally different conversation.

Is it harder to get modified hardware than software through the FDA?

Jonathan: Do you believe that the reason why there aren’t nearly as many modifications being made to hardware as there are to software is that it’s more difficult to upgrade hardware and have it approved by the FDA? Is it the main problem?

Krishnan: That’s the primary issue. I will give a simple analogy when the Apollo mission went to the moon, that was in the 1960s the computational capability was what they did with the ti-82 calculator, your iPhone today has way more capability, The beauty of it is that technology has advanced, it’s the direction that you’re pointing that technology to. So is it possible to change the entire infrastructure of the hardware? absolutely Is there an incentive to do that? maybe less so because you’re right it’s a huge pathway, seven years a lot of investments, and it’s a lot easier to do something that can change software where FDA looks at it as okay well from a safety perspective it’s not a big deal just a different waveform and if you have something that has a huge scope of uses you know you hear everything from now we’re doing ultra-low bert frequency to 10,000 to now different forms of ascending and descending charge accumulation to the multiplexed frequencies on the court, I mean wonderful and no doubt these are all important but are they fundamental shifts in the way neuromodulation is going? I don’t see it yet, but I’m certainly working hard to try to make some changes at least on the access, and the hardware costs 

Jonathan: Absolutely, because what we would like in neuromodulation is something that is Turin complete for neuromodulation, you can treat any condition with it. You have some hardware interface, it can receive an arbitrary program, it is in a neurological sense Turin complete and if I understand your point correctly, it is to get the possibility for stimulation rates up to 15,000 – 20,000 that’s a bit like just improving the sine function on your calculator. 

Krishnan: It almost seems to me like, you’re doing PKPD studies pharmacokinetic-pharmacodynamic studies for 25-30 years because we don’t understand the complete mechanisms of all the different areas. A classic example I put a 2d dipole in the spinal cord, we have been asking about the impact of neurons now we’ve shifted that conversation well what is in the spine? 90 percent of your glial cells? Most logical people would have said, well if I put the chord on a spinal cord stimulation I would be asking what all the impact of every different cell is and then I would go to what are the variables that I would be changing. It’s not that we haven’t come a long way, it’s incredible the amount of innovation, but I think that if you could standardise things, the speed of that will be way faster and much more logical

How much influence does the INS have in changing the field of nueromodulation?

Krishnan: They’re doing a great job, it’s just that I think it is how much can a society influence the changes in development of industry around these three therapies, you have a great job in new RCTs but I would love all of the clinical standardized the same development so that you see kind of that ability to say oh something is affecting, I have looked at the genetics, I’ve looked at the protein, I looked at FMRI in the clinical setting as a pilot before I run large RCTs, so cost has a different impact in that but at least we have those sets of conversations

Jonathan: Krishnan that’s very interesting I’m conscious of time but one thing that strikes me is that it’s so interesting we’ve been talking about the problems primarily in this space, whereas I initially came with the expectation that we’d be talking about technical issues but what’s clear from what we’ve been saying is that so much of this is organizational and cultural and to do with changing minds

Krishnan: I would say neuromodulation is phenomenal there are people who have incredibly benefited in my practice with it, I heard a fascinating story of somebody with phantom limb pain for 18 years and did fantastic, in any other circumstance you look at how impactful it has been for his life. What will be determining the next 15-20 years in our speciality is all of the other things around how innovation helps people get more of this therapy and I think that for that to happen we do have to make some broad-scale educational, policy and access changes or innovation has to support a much different approach of therapy and how it’s delivered.

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