Chronic Pain Chronicles with Dr Karmy
Join Dr. Grigory Karmy M.D., a distinguished chronic pain management physician with over 20 years of experience, on a captivating journey through the world of pain relief in his podcast series. Delving into the latest regenerative medical treatments like PRP, stem cell injections, and prolozone therapy, alongside educational discussions on pain transmission and the latest medical innovations, Dr. Karmy shares invaluable insights and real-life stories, empowering listeners to find relief and regain control over their chronic pain.
Chronic Pain Chronicles with Dr Karmy
Episode 21: The Role of AI in Chronic Pain Research
Artificial Intelligence (AI) is all over the news with headlines promising that AI will replace all human workers or possibly even take over the world. Hype aside, how will AI impact pain research?
Join me for a fascinating interview with Dr. Di Ieva, a neurosurgeon and computer scientist who is using AI to study chronic pain.
If you have any questions for Dr. Karmy, feel free to email us at karmychronicpain@gmail.com
Follow our social media:
Instagram
https://www.instagram.com/karmychronicpain?igsh=cHZycXdzeGhqN2Zn
Facebook
https://www.facebook.com/profile.php?id=61550237320641&mibextid=dGKdO6
Send us a text with your thoughts on this episode!
Learn more about pain management treatments offered at our clinic: https://karmyclinic.com/
when we select a patient who was responsive is not responsive anymore, doesn't mean that this patient will be failing to different kind of paradigms, because now they it's not like years ago there was one model, one brand, one paradigm of stimulation. We have so many different companies and brands and paradigm of programming with the tonic or bursa stimulation. Now they start coming even with the ultra low frequency that when one stimulator fails doesn't mean that that patient will not fail to another stimulation.
Dr Karmy:Hello, this is Dr. Karmy, and today with me I have Dr. Di Ieva. Hello.
Dr Di Ieva:Uh, hello. Very nice. Very nice to meet you. And thank you for the invitation grigory
Dr Karmy:Dr. Di Ieva has a very unusual background. He's actually a neurosurgeon. But who's trained in multiple countries from starting in Italy, moving to actually University of Toronto, St. Michael's Hospital, and then on onwards to Australia. I. So, uh, I guess that's in itself I find interesting because one is able to compare and contrast how medicine is practiced in different countries. But even more unusual is that he is actually interested in research. As much as he's interested in clinical practice, often academic physicians will do a little bit of research on the site, primarily clinical reviews of treatments that they already using. But Dr. Di Ieva went far beyond that. He's interested in mathematics, he's interested in computers. A lot of things he's doing is very much deeper inquiry than is typical for a clinician. So how did you end up getting interested in research?
Dr Di Ieva:Yeah, that's right. When I started medicine I already was interested in, uh, research because I was really thrilled by the clinical aspect of diseases and to discover their pathophysiology and to try to understand exactly what was going on in terms of diagnostics, therapeutics, and uh, to change really not just the patient's life, but literally also adding a little bit of a contribution to the big parameter science in which every one of us can contribute with a small or bigger research. So I think that I never found it like a different thing. It was a parallel thing. It was like a kind of a hobby becoming the job. So for me it was absolutely normal training to become a neurosurgeon during the day or night when I was operating. At the same time going to the wet lab, doing the research or to study in order to have some, a better understanding of what I was doing and to give some contribution. So again, I think that for me has been pretty normal to have a parallelism between research and the clinical applications.
Dr Karmy:Is just again I'm primarily community based but I understand that big academic institutions can be quite political.
Dr Di Ieva:Yes.
Dr Karmy:Is research encouraged in physicians once they finish their residency and fellowships, or is it a relatively more difficult thing to do?
Dr Di Ieva:That's an interesting question because I think it's a very political question. I've been training and working in different countries and I see that there is a different perception of this province according to where you are. There are countries in which I was doing mainly surgery and there was no, reason at all for doing research or time or there was no push at all. And the other countries in which basically there was not much expectation to do a lot of clinical surgical work. They were really expecting that you were publishing, getting grants and things like that. I think the best compromise I've had in my life was exactly Toronto when I was doing my skull based fellowship at the St. Michael. That was, uh, an extremely intensive, hardcore skull-based fellowship operating on complex, uh, brain tumors and skull-based tumors and the trauma, a lot of trauma there. Uh, but there was a very nice research environment in which our research was absolutely physiologically done without any constraints. And we had a lot of volunteers and medical students, undergrads and, um, and indeed in Canada, I, especially in Toronto, I discovered that, uh, for neurosurgeon is absolutely normal to, to publish. To have an academic curriculum and to, to be able to operate. Uh, when I came to Australia, uh, they support you they help you. They want that, you become a dual clinician researcher. Mm-hmm. Uh, but actually it's extremely rare. I think that we are really few neuro surgeons in Australia doing also neuro uh, research, basic research. Getting grants, investing time to get grants because at the end of the day, uh, when we do research, uh, at the professor level, you're not at the lab anymore. You are really trying to put the team together, though it's more a leadership position, so you have ideas. You, you wanna concretize this idea, you need money, you need grants. You need all the logistics before, during, and after the grant, and you wanna, uh, hire people or the good people, or you wanna train them and you want to create this kind of environment in which, we can work together. So again, many, many group of my colleagues, they don't do that. But we're trying to create this kind of academical surgeons of the future in which we have kind of a leadership research and, uh, of course we operate our priority are still patients.
Dr Karmy:Sounds like you've got a job of a CEO plus a job of a medical doctor.
Dr Di Ieva:I. Yeah, that's right. I would like also to tell you how I came to pain, because as a neurosurgeon's pain, there is more percentage except when we do spine or peripheral nerve surgery. What we call, uh pain functional neurosurgery is a very small percentage of our training and our practice. Mm-hmm. Indeed, I learned pain neuromodulation, and pain management. But for many years I haven't done much because I was extremely focused on brain tumors and skull-based tumors, which is still my first expertise. So I learned that. I did also some, uh, internship observership and the fellowship in pain, including the states and Europe, but for many years I haven't done, when I came to Australia, I was brought back to the battlefield by pain physician because I had so many pain patients and many of these were extremely complex. And, when they were, let's say, beyond the pain physician's expertise and he needed a neurosurgeon, he brought me back to the theater and we started just with extremely complex patients or complication done by someone else. And, uh, we established such a good routine that at the end we decided to create these pain and neuromodulation clinic. And nowadays we standardized that we see patient every time together. We decide our decision making together and the if they need a surgery, we do surgery together. By doing that, we improve our our results. The outcomes are better and the complications rate is much, much lower. And of course there is all the research aspect because I still have a bigger background research. I'm a computational neuroscientist. I have also a PhD in computational neuroscience and I have some expertise in, in wet lab. So I tried to put all these work with a huge expertise in brain tumors, translated to the pain. I wanna just tell you that basically I went to, into this domain from the back door, basically. Mm-hmm. I brought expertise in a translational fashion from different domains, brain tumors, wet lab, cadaver lab, computational to the pain domain, which is something, was something new for me.
Dr Karmy:Okay, so let's move into, uh, discussing a little bit about your research on pain. And from what I understand there is fundamentally two aspects that, uh, you are focused on. And maybe let me start with biomarkers in patients with chronic pain. Presumably there different physiologically, whether it's because there's different chemicals. Circulating in their blood or sometimes even different microbiota that lives in their stools. Uh, it's an active area of research. Can you tell me a little bit about this field?
Dr Di Ieva:We're looking for something which is a little bit more objective than the subjectivity of a pain. We're dealing not with the structural problems many times, uh, it's not like when I deal with the brain tumors is there, you can decide to take it out and doing also the other treatment afterwards. This is something that many, many times when you have a ruled out structural problem in the spine or in the nerve is a functional. And so you have trust of patients. What do the patient tell you? You start from a very simple uh, scores, the vision, neurological score, and basically they tell you zero to 10 how much pain they have. But of course, it's not enough. You have to keep going with more and more psychological test. And even when you spend hours to characterize the phenotype of patients by means of neuropsychological batteries of test. You still have to trust their subjectivity to describe phenomenon, how to sleep, how is, uh, their quality of life and so on. So what we are looking for is objective biomarkers. One of the biomarkers the most study ones are from the liquid biopsies. Mm-hmm. So. The best would be spinal fluid. So take the spinal fluid and look for your biomarkers. But that's an invasive procedure to put a, a needle to sample of CSF cerebral spinal fluid by menal lumbar puncture. So blood. Or other samples. But the blood is one of the easiest one because many times patients, they do blood sampling for other reasons. And when we have a sample of blood, we're looking for biomarkers, and that's the biomarkers that we're looking for. From the diagnostic point of view, is there a biomarker? So for example, a protein. Something in the blood, which is, uh, related to pain. And, uh, does this change after treatment if you have a successful treatment? Is this diagnostic biomarker also a prognostic one? Because basically is a changing, uh, the patient responsive to treatment. That biomarker, for example, goes down. Just an example. Mm-hmm. And, uh this is a prognostication, and if it goes up again and the patient is back, this is an objective quantification of a phenomenon. And it can be also therapeutic biomarker. Maybe that protein can be used as a target for a targeted treatment as many times we have done for the last decades. So looking for, uh proteins in the blood, for example, gives rise to the search of a three different kind of objective biomarkers. The diagnostic, also therapeutic and the prognostic bar biomarker related to the pain to make the pain, subjectivity more objective from the wet lab point of view.
Dr Karmy:I think there's been a lot of efforts in that direction. I'm not up to date on the research, but I think there has been quite a bit of, uh, efforts up to now. And a number of biomarkers found. But I don't know how reliable they are. Certainly there's nothing that's used clinically. So do you have a sense why even though there have been some biomarkers in research, nothing is used clinically at this point?
Dr Di Ieva:Yeah, you're absolutely right. Over the last decades though, there have been so many researchers and proposal of biomarkers, but none of them really hit the clinical world. We don't use a biomarker from the clinical point of view. There is a lot on the research domain. I could, uh, mention some biomarkers, but we are not there. So what we are trying to do now is, um, for example, we do different kind of analysis coming from the omics approach. So the omics, you wanna study all the genes, the genomics, you wanna study all the proteins, the proteomics. Or you wanna see the transcription of DNA in transcriptomics. So basically, uh, we can take the blood doing multiple omics approaches because every time they were looking for the Holy Grail protein, the single protein explaining everything, but actually that's the limitation of the, these reductionistic approaches over the last decades. It's not a holy grail. It is an interactions of proteins or there are epigenetics modification of the same protein, which might be or not, for example. So what we're looking for is multiple biomarkers, how they interact each other, and how these networks of biomarkers changes in longitudinal fashion over time according to response to treatment, to the treatment, or lack of response to treatment. Again, I would not go down the track to mention all the biomarkers found, but our approach that we're doing on this research is, uh, simply or not simply trying to understand the networks of a protein expressed in a patient before treatment to try to characterize the fingerprinting from the molecular point of view of pain and how these changes after all treatment. If the patient has a good response, and this has been shown on a neuropsychological assessment, we wanna see if these mirrors. Also in some biomolecular changes.
Dr Karmy:So first of all, a lot of these efforts seem to be primarily measuring proteins. Yes. The original approach, which has not been successful, is to try to find one protein that goes up when you have pain goes down as your pain is starting to go away. Or perhaps you hope that maybe even that protein is the cause of the pain and could be a target for treatment, uh, which you are doing is much more sophisticated. So, and that's partly because of progress in technology, because in the past we could only measure one protein at a time. Now with this omic approach, you can measure. Hundreds, maybe, I don't know, more proteins at any one time. And so what you're looking isn't so much a single protein, but a pattern.
Dr Di Ieva:Exactly interactions of a protein, clusters of biomarkers and patterns and how this pattern change. And my expertise actually doesn't come much from the wet lab point of view because we have a molecular biologist working on that. But it comes from the reduction and of diesel, big data set in order to make it more, uh, interpretable. So you wanna understand how these patterns can be quantified, stratified and compared in longitudinal fashion, and that's when the computational modeling comes up. We use machine learning in order to understand, uh, how this protein interact and especially to have some quantitators, which can be compared in longitudinal fashion over time. So that's why we need the domain from the wet lab point of view, the molecular biology point of view, but we need also the computational domain, and that's where artificial intelligence comes in because we use a lot of, machine learning techniques in order to understand things that are maybe they're not very easy to understand. And again, because we're not looking for a concentration of single protein going up and down, but we are looking for patterns. The pattern analysis is a very, very well known subfield of machine learning, well known and used in other domains. So we are just applying algorithms coming from other domains into our domain, which is the pain characterization.
Dr Karmy:So basically you have a wet lab where you are doing all the measurements, the omics, and then if you have hundreds of thousands of proteins, there's, uh, exponential, billions of possible combinations of them all going up and down and trying from all of that noise is to see a pattern. Humans are not just not capable of doing that, but that's the strength of AI where it's able to pick up patterns that humans cannot because there's just too much information. and anything else you wanted to throw in about this aspect of your research?
Dr Di Ieva:So if you wanna do a serious epidemiological study, you would like to compare the typical healthy population where with the diseased population you have to stratify pain, as we said, is a big umbrella, so you have to stratify all the different phenotypes of a pain and you wanna see if each single phenotype has a fingerprinting and compare the healthy population. But this needs millions of dollars of grants and many, many big data set in order to have a statistical significance. So in order to make it more feasible and pragmatic we don't compare really to the general population, although we could because there are, there are shareware data set of health individuals proteomics, for example. But we do something a little bit more simpler we characterize the biomarkers before and after treatment, for example, before and after spinal cord stimulation. And, uh, so we do a binary yes or not. Is this patient responsive to spinal cord stimulator trial? Yes, we do the implant and if not, we don't do the implant. And if we go towards the implant, is anything changing in their blood, in their sample, or even in their brain? And this goes to the other research that eventually we're talking about today, but we're looking for these changes, not just from the subjective point of view of the patient, but also from the medical point of view, which can be from the lab point of view or from the computational point of view. Absolutely.
Dr Karmy:I guess, you know, pain is one of those fields that a lot of different specialties claim. On the one hand, it can be as simple as exercise, diet, and physiotherapy. At the very other extreme, it's what I typically think of surgeons doing, which is spine surgery. Yeah. And so, and then in between there's interventional pain management, which at least in Canada, is done sometimes by family physicians, sometimes by anesthesiologists, occasionally by physiatrists, uh, which generally involves a need of some description. So basically things like radiofrequency ablation where they burn the nerves or cortisone injections into the joints. There's a little bit of trend over regenerative medicine type procedures such as platelet rich plasma injections and nerve blocks. And of course spinal cord stimulators which I think are not as widely used in Canada as they are in US, but I think they're sort of starting to gain some traction initially, primarily for radiculopathy or leg pinch nerve. But now we are starting to use it more for back pain as well. When you are working together with a pain physician, is the decision surgery versus spinal cord stimulator versus a needle? What? What's that like?
Dr Di Ieva:Yeah, yeah, you're right. Pain is a huge, huge, very broad umbrella. And there are many specialists, dealing with that. So the way we deal with the pain, uh, patients is, uh, already the last tier of many, many things already done. And, uh, eventually failed in the past. So these are patients, for example, they under already underwent spine surgery or somewhere else and have a kind of a failed back surgery syndrome, although we don't call this anymore. But for example, they do spine surgery. They still have a chronic neuropathic pain, or they have had the complications or they have a trigeminal neuralgia, which basically has been treated all the tiers of starting with the medications, passing through a microvascular decompression and radiofrequency ablation, percutaneous treatment on no gas, sir or even gamma wave radio surgery sometimes. So these are complex patients that they've already tried many, many things and, by pain specialists and or neurosurgeons and all different kind of doctors. So at the beginning, when they came to our clinic, they were very selected by the fact that they already did everything else and they failed. And we tried to have a kind of a collegial decision making for the best management of complex patients. But when we became successful with these complex patients, we started to think maybe we should really start from scratch, any kind of a pain patients. Uh, of course if you have a disc herniation of causing radiculopathy, I don't need the pain physician. It's pretty easy, common problem. But if you have the patient with the years and years of a chronic neuropathic pain, or five neurosurgeons or spine surgeon they have proposed a fusion surgery. Some pain physician, they have tried radio frequency and there is really a melting pot of different information on that. Even the patient and doctors are so confused. So that's the case when we try to put together our expertise, sometimes even with the psychologist, to try to stratify their pain phenotype or a response to, potential treatment. And that's when we try to put our expertise together. Now, when we have selected someone towards a treatment, if it's a surgical, we are doing that together. Pain physician, neurosurgeon. If it's a purely interventional from the pain point of view, or radiofrequency ablation, for example this is done only by the pain physician, but if we select someone. I dunno, for fusion, of course, the pain physician steps down and this is just my operation, so that's why it's not a common thing that we do every day, every week. This is a clinic that we do every month to select, to filter this kind of patients. And we do two, 3 days a month in which, we try to put our expertise together, but many, many times they still go to the single expert rather than to the collegial team.
Dr Karmy:You know, again, multidisciplinary teams I think is a little bit of a trend in chronic pain management.
Dr Di Ieva:Yeah. It's the workforce of a medicine, I think. Yes.
Dr Karmy:And, we all have patients that basically failed everything. Yes. Um, and there's no silver bullet unfortunately in pain management. And unless of course you develop one for us. But just out of curiosity, you've got somebody who tried everything, has a neuropathic pain, presumably from damaged nerves, right? Yeah. They would've probably gone through all the different medications, your gabapentin, your pregabalin, all the, cognitive behavioral therapy type stuff. Uh, you are the port of last call. Maybe they haven't, tried spinal cord stimulator, I suppose. I dunno how wide that the available is in Australia. So I guess my question is, what's the secret sauce? What is it that you guys are able to do?
Dr Di Ieva:Yeah. I should reemphasize that we're not able to fix every one. I'm just trying
Dr Karmy:What's percentage success rate?
Dr Di Ieva:We get complex patients and, you're right. So is the selection of criteria, because sometimes, for example, this patient have been in pain for a long, long time and they never tried the spinal cord stimulation. And if we think that might be successful, one we trial and if this is successful, we go to our implant. That's, the first sorcery using your word that we do. But sometimes we push even our boundaries further. In Australia spinal cord stimulation is pretty well spread, and, sometimes they already come with the spinal cord stimulator, which has worked at the beginning and doesn't work anymore. Mm-hmm. And, uh, for many, many colleagues, this would be the end of the story. There is not much to do. You have tried everything, including a neuromodulation device. Which is really on the spotlight nowadays for many, many reasons, economic reasons, because it's expensive. Sometimes they get complications and suddenly goes under the spot, which is something bad. But, uh, when we select a patient who was responsive is not responsive anymore, doesn't mean that this patient will be failing to different kind of paradigms, because now they it's not like years ago there was one model, one brand, one paradigm of stimulation. We have so many different companies and brands and paradigm of programming with the tonic or bursa stimulation. Now they start coming even with the ultra low frequency that when one stimulator fails doesn't mean that that patient will not fail to another stimulation.
Dr Karmy:Would you have to hook up a brand new stimulator or can you use the same probes? Would you have to pull it out or exchange it, or can you simply reprogram it or use the different machine?
Dr Di Ieva:Is a yes to all your question whoever had the times in which it was just enough to change the impulse generator. Same wires without touching the wires and changing generator. Other times the wires are gone, they're fractured. We try or we wanna change the position of the wires because it was a suboptimal. So in that case, we, we do everything from scratch. So we have had all the cases starting from scratch or readjusting, what was done in the past with the new models, new hardware, or a new paradigms of stimulation. Or combination of things. Sometimes I've had patients for example, they have dura problems and you wanna still put a spinal cord stimulator for a lower back pain. Uh, but there is also radiculopathy and you know that there is a mechanical reason there. So eventually we try to fix in the same time. Remove the mechanical compression and the remove the chronic neuropathic pain by means of stimulation. And by pushing the boundaries, we are getting more and more patients from different disciplines. I'm starting to get patients from endocrinologist nowadays because the the neuropathic pain or paresthesia, the BE neuropathy is a well responsive to high frequency stimulation as shown in multiple trials. So we are getting more patients of this kind nowadays. So again, we don't have a magic wand to fix, but, we try, to fix all the complex patients, but we try to put on the table all the data of, uh, how the patient is, what's the phenotype, what's the psychological status, and what has already tried. What has not been tried and what could be successful. And that's where, uh, we try also to predict and stratify what could be absolutely successful and what would be waste of time and money. And that's. When we don't do it.
Dr Karmy:So it sounds like they, they developed a lot more different protocols for spinal cord stimulation and that really changed the field quite a bit So, so let's maybe move on a little bit into the computational research. We have many ways of imaging the brain have improved drastically in the past 20 years. Um, and there's been also, there's some functional research, which we can just see the anatomy. We can also see which parts of the brain become activated with various tasks. All of that body of research has led to all kinds of suggestion that there's actually anatomical changes that are associated with chronic pain. I understand you have an interest in that aspect of it. How the brain and parts of the brain that process pain change when patients develop chronic pain. Can you talk a little about that?
Dr Di Ieva:Yes, correct. I have a, a strong interest in neuroimaging because that's what I use on daily basis for my brain tumor patients. And, in those patients, there is a structural problem. The brain tumor causes some structural abnormalities in that brain and functional problems because this patient can have, uh neurological or psychological or cognitive deficits caused by the disruption of the network of the brain cell? So first of all, at the imaging level, we use a lot of the MRI because the MRI gives information from the structural point of view. We can see things, how they are connected, what we call the myelo architecture, for example, how different part of the brains are connected, and the functional connectivity. We can check how different part of the brain connect each other in order to give rise to a function. And this is what now this is called the connectomics another omics, which is a multi-data analysis of the connections to the brain at the structural and the functional level. This can be done by different techniques when we talk about patients. MRI is the number one. There are other techniques and why are there other techniques? Because MRI is very good for a special resolution, tells you exactly where is where. So it tell you the location, but it doesn't give you the temporal resolution because when you are able to depict that there is a change in that part of the brain. Something in the neurons already happened many hundreds of milliseconds before. So the temporal resolution of the femoral, the functional MRI is relatively low. So we can add electroencephalography, EEG or magnetoencephalography, MEG. These techniques are, uh, extremely precise in the temporal resolution, but not in the special one. Uh, they can tell you exactly what when happens to something, but not exactly where. So by merging these two techniques, you have a both in high temporal and special resolution. So in the research we're doing nowadays, for example, at my lab, we use MRI. Uh, and the magnetoencephalography.. So we record the magnetic field of patients in order to have a very high temporal and, special resolution of what's happening in the brain. And, uh, we wanna see this how changes because the brain is affected by brain tumor and that's the easy part. But the difficult part is in pain.'cause many times pain, uh, does not give rise to any structural problem in the brain. Is something functional? That we need more tools. In the past, we were using only task-based functional MRI, for example, move your hand and you see the activation of the brain on the motor strip, on the contralateral hemisphere and the disease, the part of the brain giving activation, but in pain except in very uh, specific experiments. You cannot do that. I mean, you can trigger pain. And you can see in acute fashion where the pain is mapped on the brain. But the chronic neuropathic pain patient, you cannot do that because this is something chronic is always there. So the working hypothesis is that that brain is already different from the health indu population. So eventually there the structural and function, especially the functional connectome is different. So the research over the last decades is going towards what we call the pain connectome. So the connection of the brain related to the emergence of the pain as a disease. And if there is a pain connectome, which has been shown in many publications, we want also to understand how the treatment affects that pain Connectome. Now to put all these research under the same umbrella, you need advance the new imaging because you have to do and to understand the neuro imaging in terms of MRI and other techniques like EEG or MEG. And you need the computational tools to try to analyze these big data set because we're talking about really huge data set and information for each single patient. We're talking about many thousands of images, many uh, hundreds of thousands of times, series of spikes of activation of the neurons and this time series, they can be analyzed by different tools. One of the tool I use is related to the computational fractal based analysis, which quantifies the geometrical complexity of a pattern, how complex or smooth a time series is, for example. And the other one are machine learning techniques and the learning in which we use ai. To reduce the dimensionality of this huge data set and to try to make some sense of what they mean. When we have a descriptor from the computational point of view of the Connectome of the pain patient. We wanna track this over time according to the treatment we are doing. And one of the research I'm doing is et on the spinal cord stimulator because, uh, there are many decades of, uh, research why the spinal cord stimulator works. And, uh, there is the electrical effect. There is the chemical effect on the spinal cord. There are many, many theories. There is the gate theory, which was the basic theory decades ago which is still valid to explain to patients what do we do? But these are not really valid from the research point of view. But there is also the hypothesis that when we stimulate the spine, the nerves peripherally, we're changing something at the suprapsinal level. So the brain, at the end of the day, the brain is the perceptor, is the, elephant in the room really understanding and the perceiving pain, right? So we wanna try understand why this patient having a pain in the foot stimulated at the level of the spinal cord in the lower thoracic spine, uh, has a different perception of the pain. Or a different response to the same treatment. And that's why we go to the supraspinal level, studying the connectomics of that brain before and after treatment. And this can eventually give rise to some other biomarkers we were talking about before. So there are the liquid biopsy biomarker that are the computational biomarkers related to the image analysis.
Dr Karmy:So basically you can image the brain using functional MRI, which gives you which areas get activated. But because the changes are so slow, it doesn't really give you the time component that, the area get activated for one second. Or for 15 minutes. It's doesn't give you a very specific timeframe. So then you use another modality, which measures the electricity that nerve cells generate, which is either EMG or MEG. And that tells you that information about the time course. When you told describe pain connectome, and the way that I interpret it, and I could be wrong, is that you're basically saying that the way that different parts of the brain are connected, the normal people is different from the way brain parts are connected in patients with chronic pain. Is that what you're seeing?
Dr Di Ieva:That's exactly correct, and uh, I did not emphasize enough that, again, I spoke about the task based function. So you're moving the hand, you see the activation, but here is not task based. It's what we call resting state. So the patient is just going there in the machine, open eyes, just thinking about things, and we are just recording the baseline function of the brain. So that's why it's called resting state functional MRI or resting State magnetoencephalography. So we wanna see the baseline function of the brain. In a physiologic and pathological status and how these changes over time.
Dr Karmy:And what you're saying is that in people with chronic pain, the baseline function is different when you use these tools.
Dr Di Ieva:Not in all brain, but in some components of the brain or in some networks or there are some differences. And that's why some motors nowadays are talking about the pain connectome. So the, the functional network of the brain related to chronic pain, for example.
Dr Karmy:And what you're telling me perhaps maybe I'm reading into this too much, is people who respond to treatments, you also see normalization of some of these changes.
Dr Di Ieva:Yeah, yeah, of course. We try to understand what normalization means, and that's why we need more healthy individuals to try to depict what a normal brain looks like, and, uh, how this is different from a pain patient and how it's different from a pain patient undergoing successful treatment. So we try to create all these kind of fingerprinting in our longitudinal fashion between the different cohorts of patients and health individuals and, uh, amongst the same individuals. But over time.
Dr Karmy:There is a little bit of discussion, I think in neuroscience circles Certainly a lot, most people feel that a lot of patients with chronic pain have sensitization. Where essentially, you know, the tissue has healed, but the nervous system that transmits pain signals continues to fire off. Signals telling the brain that there's damage and there is some discussion. Does it happen primarily in the peripheral sensitization? Does it happen in the door, in the spinal cord, uh, substantia gelatinosa or does it happen in different structures in the brain? And does one lead to the other? Does peripheral sensitization drive central sensitization? Uh, so are you saying that these anatomical changes in pain, connectome in the brain is the substrate for the sensitization?
Dr Di Ieva:I. Not always. I, I would not like to generalize because we have so many other pains that clearly they come, they start from the periphery. I'm talking about, uh, neuroma I'm talking about, uh, complex original pain syndrome, starting as a peripheral problem and afterwards becoming a central problem. So I think there is always an overlapping between periphery spinal cord and the central system in the brain. So between interaction between, uh, the generator of the pain in the periphery, the carrier of the information, which is the nerve, the spinal cord. And the Supreme Court, which is in the brain. I don't wanna say that the brain is always involved, but we have, uh, cases of, think about the phantom limb pain. Mm-hmm. This patient, they don't have, uh, the arm anymore, for example, but the pain is perceived because there is an overactivation of the pain network, contralaterally to where the arm was. We have ablative procedural destroying some part of the spinal cord in order to interrupt that wrong circuitry. For example, though transectomy is a surgical technique we use, but there is also the neuromodulation domain in which putting a spinal cord stimulator on the spinal cord or directly on the brain helps to reduce or even cancel the phantom limb pain. So again, there is an overlapping between something which was in the periphery because eventually it's the damage of the nerve at the brachial plexus or uh, or post or pre ganglionic nerves before they go to the spinal cord. But after that it becomes a brain problems. There are so many different phenotypes and different problems that we try to, uh, to make my research a little bit, uh, feasible. We try to narrow the tunnel of the inclusion criteria and for example, an easy one is the lower back pain when there are no structural problems. Lower back pain if you've done an MRI. MRI CT scan is perfect. There is no structural problem in the bone. The bone scan with an knee doesn't show any hot spots. The MRI of the muscle does not show any fat infiltration of the multi fetus. There was no trauma. Is a profitable spine and why this patient have a pain. So maybe this is the easy model to use to see what's going on at the brain level perception. Of course, when you have done a good model for a very narrow question, you can start to extend to more difficult questions like A-C-R-P-S or phantom limb or uh, or diabetic neuropathy and uh, or other problems.
Dr Karmy:So here's, I guess a somewhat related question and I'm probably gonna end with that. So if you are able to find, uh, abnormal pain, connectome does that impact the treatments. For example, when you are, try doing the brain stimulation, would you be able to say in this patient the abnormalities in that part of the brain based on pain, connectome, and that's where we should put electrodes and another patient abnormality is a different part of the connectome and we should put electrode somewhere else or maybe even uh, suggest what type of stimulation protocol to use based on some of these.
Dr Di Ieva:Yeah. Yeah. That's the final$1 million question because, we are going towards diagnostic and prognostic, uh, signature. Now the therapeutic one is much more difficult because when we have confirmed that there is something different, how we can change that, connectome towards the normality in order to reduce the pain, and this is purely speculative and theoretical. We have some tools, but this is a black space in which black domain, which we are trying to add more information. In theory, if we have a connectome, we wanna buzz that network in order to go back to the normal function. This can be done by epidural stimulation stimulator, director in the brain. Deep brain stimulation stimulating a deep part of the brain. In the past, there was the periaqueductal gray stimulation. For example, nowadays we can do thalamic or single lung stimulation. And, uh, last but not least, even non-invasive procedures. We are talking about opening the, the skull here, but. There are also the non-invasive one, like transcranial magnetic stimulation, and, uh, there are a lot of anecdotal cases nowadays in which TMS works very well for chronic robotic pain. But it, it's like for depression, it works pretty well in pharmaco resistant depression. We start just out to understand why. The same will be with the pain. Some people have a very good response, not just uh, from TMS for depression, but also for chronic neuropathic pain with the TMS. But we have to try to understand what's the program, what's the location of the stimulation? What is the frequency? How many times you have to do, and when you have had the successful patient, is this for good or for short period of time? So these are the next question. The treatment, of course, will be the final goal that we are looking for, but we are not completely there yet. Unfortunately.
Dr Karmy:Thank you, Dr. Di Ieva, for taking the time to talk to me. Good night.
Dr Di Ieva:Thanks you, Dr. Karmy. I really appreciate that. I enjoyed it. Cheers.
Dr Karmy:So what are my final thoughts about Dr. Di Ieva's research? Well, first of all, we have very rudimentary understanding of chronic pain. I. And very rudimentary ways of diagnosing it. At present in order to find out whether somebody has chronic pain or not, we just ask'em. We don't have a blood test and the treatments are mostly trial and error because we don't have a full understanding what causes the pain? The reason for this limitation is that the tools that we had for yes, to investigate the human body, were very limited. For instance, you could only measure one chemical out of millions in human body at a time, which means that if in fact there are some chemicals that are involved in chronic pain, it would be kind of like looking for a needle in a haystack. Now we have advances in how we measure things. We have omics, proteomics, genomics, which enable us to measure hundreds or even thousands of proteins at once, which makes it easier to find the proteins involved in chronic pain. Equally importantly, the other big advance in our tools is advance and analyzing all these data, and the reason for this advance is the development of artificial intelligence. In the past, it would've been very difficult to look at thousands of proteins especially if they work in combinations with some proteins going up and other proteins going down to find the combination that causes chronic pain. There are millions of possible combination and human brain is simply unable to analyze that much information. But AI is actually very good at sifting through very large amounts of information to find patterns human brain is just unable to see. Likewise, we can measure multiple nerves in the brain when they produce electrical signals, which is how the nerves communicate. But the problem is that there is something like a billion nerve cells in our body. Only a small fraction of the, those cells are responsible for transmitting signals involved with chronic pain. It's a little bit like being in a noisy room with a billion other people and trying to find a hundred people who are discussing the topic you're interested in. In the past, that task would've been impossible. But again, with artificial intelligence, this task could very well be manageable. The hope, of course, is that if we understand better what causes chronic pain, then it'll lead to development of newer and better treatments and will take a lot of the guesswork from both diagnosing chronic pain, but also selecting the best treatment for each patient. Thank you.
Raveena:Disclaimer when it comes to your health, always consult with your own physician or healthcare provider for personalized advice and guidance. The information provided in this podcast is for educational and informational purposes only, and should not be considered medical advice or a substitute for professional medical care. Be sure to follow our Instagram at@karmychronicpain for updates on new episodes and more educational content today.
