The Promise of Gene Therapy in Neurological Disorders

Podcast Transcript

Intro: Neuro Pathways, a Cleveland Clinic podcast, exploring the latest research discoveries and clinical advances in the fields of neurology, neurosurgery, neuro rehab and psychiatry.

Glen Stevens, DO, PhD: Advances in gene discovery are paving the way for the promising use of targeted gene therapy for neurologic disorders in both pediatric and adult patients. In today's episode of Neuro Pathways, we're discussing the advanced techniques being used by today's genomic researchers and the profound application they may have in treating the most common and rare neurological disorders. I'm your host Glen Stevens, DO, PhD, neurologist, neuro-oncologist in Cleveland Clinic's Neurological Institute. Joining me for today's conversation is Dennis Lal, PhD, MS. Dr. Lal is Assistant Professor and Assistant Staff in Cleveland Clinic's Genomic Medicine Institute and Neurological Institute. He's also a visiting scientist at the Broad Institute at Harvard and MIT in Cambridge, Massachusetts, and group leader at the University of Cologne in Germany. Dennis, welcome to Neuro Pathways.

Dennis Lal, PhD, MS: Thank you very much.

Glen Stevens, DO, PhD: My first question is going to be a broad one. And that is that when we look at the genome itself, it certainly had a remarkable accelerated course. I'm a neuro-oncologist, and I think glioblastoma was one of the first tumors that was characterized for the genome. But can you start off the stage for today's conversation by taking us through the last 10 to 15 years, to what we know today about the genome itself?

Dennis Lal, PhD, MS: Yes. Thank you very much for inviting me today to speak. Very happy to talk a little bit about genetics and how it's affecting disease. So I think there are a couple of milestones, how genetics improved over the last 15 to 20 years. First of all, in the early 2000s, that was the time where we fully characterized the human genome, where the great hope started. Now around to 2000, we had the first complete human genome and everyone was very hopeful to that we basically can solve all human genetic disorders or disorders where we thought genetic plays a role, in the next 10 years, in the next decade. However, that didn't deliver actually. So, the next kind of revolution started more like ten or eight years ago when a new type of technology came to market, which was a new version of decoding the DNA through sequencing.

So before, this was a very, very slow process. But the technology improved significantly that you can decode one human's genome in round about a week of time, whereas in 2000 it would take you like five to seven years and round about 30 billion dollars to decode the single genome. But 10 years later, that can be done in a week. And this enabled then the clinical integration of, clinical sequencing, at least at the beginning for research and today for clinical screening. And with the massive integration of clinical sequencing, we learned a lot about genetic diseases. And today in many Western societies and many academic, and even now also non-academic centers, clinical genetic testing is now routine.

Glen Stevens, DO, PhD: So the goal is to identify, obviously, the right genes and the right patients that would fit the therapy correctly. Where are we in that? Do we have some that we can do that with? Or is this still in the future? What's the status?

Dennis Lal, PhD, MS: So this is quite interesting. So when it comes to moving towards genetic or targeted therapy, there's typically a game plan. So the first part is identifying the causal gene. And this is, at the end, not the gene, it's more or less the mutation in that gene, which leads to the abnormal functioning of the gene and ascitic protein. And then, typically, what people do is they try to identify the patient which really matches this defect. Because what happened is that people thought that patients have version one of the disease and that this could be explained by genetics. However, what genetics told us is that version one of the disease can be explained by 10 different genes. So there's more clinical heterogeneity, which is correlated with genetic heterogeneity.

So once it's figured out which gene does which subtype of the disease, then people have also clinical biomarkers like endpoints. And then people try to develop therapies where you, for example, replace the gene by a virus which brings in the healthy version of the disease into the body, or by improving the amount of remaining healthy gene ... also through some modifying nucleotides like therapies. And this is what people try now.

So there are remarkable advances which have been done for 10 to 20 examples for neurological disorders. These are typically rare disorders. But the beautiful thing is if you look at the last months to two years back, and the literature, and then at conferences, almost one in five rare diseases, there are already examples where people can try to cure the disease in mice. And let's say 1 in 20 or 30 diseases that can do it.

So the beauty about these genetic approaches are that the framework is the same, but you only exchange the gene which you want to target. So basically the vehicle, the approach of how this medicine works is similar and you can exchange it. It's this same underlying premise, for example the mRNA vaccines, which we use now to immunize ourselves against the Coronavirus ... in theory, the coding part or the specific Corona part could be exchanged also for different virus, right? And the same underlying premise and/or idea's also behind the targeted therapies for the neurological diseases.

And that makes many, many, many people hopeful because if we look at the animal studies, which are way more ahead obviously than compared to human studies, you can see that you have a disease which we know for example since 30, 40 years, which is devastating, and is really disabling the person's lifelong, and they even die early. And if you generate then the mouse model, which has the same genetic disease, you inject twice, or sometimes even once, then this genetic therapy and the mice look from the outset at least, even after two years, that they look cured with one injection. So, that makes many, many, many, many people help really hopeful that this will translate to human success as well.

Glen Stevens, DO, PhD: So how much of a complication is heterogeneity?

Dennis Lal, PhD, MS: So are you asking clinically or genetically?

Glen Stevens, DO, PhD: Genetically.

Dennis Lal, PhD, MS: One of the lessons we learn, and this is across the board and from cancer to neurological diseases, pediatric and adults, is that what we can clinically capture through our methods, which we use in routine care or even advanced care, let's say looking at an MRI or an EEG or a PSG or whatever, we group people together which look more literally on a genetically level, not the same. So to us, they are the level of granularity, which we can access with the technologies which we have in hand. They are frequently not able to capture the full blown molecular heterogeneity. That means if you, for example, can you just give an example from epilepsy where we work, we can have a disorder which is called developmental epileptic encephalopathy, which is a severe pediatric brain disorder with a lot of daily seizures, which cannot be treated with any form of therapy which we have available, and the children have severe inter-disability and many other diseases.

We were able to group them clinically into 20 different groups, and some which don't really fit the bin. However, when we now look at the genes we have identified which can cause these disorders, we found more than 120. So there are many, many parts of an engine, if you think about the car, which can break, which leads to the situation that the car is slowing down, or is not driving at all. So, and if you want to precisely restore that, you need to identify what is wrong, to make the engine work again. It doesn't help if you repair one part which is actually working, if the underlying thing which is broken, is not exchanged.

Glen Stevens, DO, PhD: So you mentioned a little bit about the work you're doing in epilepsy. I see that you've come up with something called a polygenic risk score in epilepsy. Can you talk about that a little bit and how that might help you in treating patients?

Dennis Lal, PhD, MS: There's a general rule in genetics and in disease, the more common the disease is, the less likely a single mutation can play a role. Then, the other way around, the more rare and typically devastating and early onset the diseases, the more likely a single variant can explain it all. So this is something what we have which is called selection. So if you imagine you have a mutation, which is totally devastating, that will not be passed on, right? Because you will never have the chance to have children yourself because it's already as a child. It might be even lethal. However, if it's mild enough that you can have children, you might be disabled, but you can still have children, this can be passed on. So there's this general rule that common variant of mild effect leads to common disorders. However, these variants are considered as having really, really tiny effect.

So this is what we basically observe for all common disorders, from Alzheimer's to dementia, to Parkinson’s, to diabetes, to common forms of cancer or risk to develop tumors and overall to literally overweight, height. Everything what is even traits like risk, heart ticking, everything makes us up as a human, as these have all genetic components, but they play a modest role. And what you can do is you can do very large studies where you look at, say, like in a million people and look at their BMI and look and ask, "Hey, what is in the genome different between those people who have a high BMI, compared to those who have a low BMI?" Or people who have epilepsy versus who don't have epilepsy. And then what you will find is you will find between 5,000 to 200,000 mutations which you see in almost every 10th individual in the population, which you find a little bit more frequent in individuals which have a certain trait or disease ... just a tiny bit.

And these small variants, if you combine them, they add up and then you can use it as a score. And for example, this is what we did in epilepsy. We could identify 30,000 risk mutations, which individually have a super tiny effect. However, if you ask them in a cohort or like on group level, a thousand people with epilepsy, how many of the 30,000 risk mutations for epilepsy do you have? You will observe then on average, they will have let's say 17,000. And then if you take say 10,000 people who don't have epilepsy, and ask how many of these risk variants for epilepsy do you have? You might find only 13,000 on average. So this is then considered as a higher polygenic risk, which can first give you higher chance to develop epilepsy.

So to make a long story short, we can use now this score and look in an independent cohort, and what we see is that we can identify around about 1% of people, it's not many, who have a fourfold risk to develop epilepsy at birth. And the best example where this is almost clinically meaningful is Alzheimer's where you can score someone's risk to develop Alzheimer or dementia, depending what kind of score you use, with almost 90% accuracy at birth. So most of it is associated with the APOE e4 allele, but there's another 5% risk which can be attributed to variants outside of the APOE e4 allele. The biggest question is do you want to do this? Is it meaningful? There are examples for Alzheimer's that if you have lifestyle interventions such as learning a new language or being more mentally active, that this could protect you in certain situations. And there's good example for stroke. For example, if you have a better lifestyle in terms of nutrition, you lower your risk of having your blood clot, basically. Then you also are at lower risk to develop stroke, even if you have a high genetic score, and there's a lot of research in that direction.

Glen Stevens, DO, PhD: I guess one question that comes up would be the cost of doing these types of tests.

Dennis Lal, PhD, MS: If you want to do a polygenic score for any kind of disease, you would have to make an investment. If you would do it for research, you would have to pay $30. Then you have round about 500,000 regions in the genome characterized. And there's some magic which can happen through statistics where you can make out of 500,000 regions, 10 million. And with this, you can literally do a genetic risk score for all diseases. You can use bioinformatics, where you take the known risk variants, and ask to which degree an individual has them. This literally would cost $30, this is very cheap. But at the end, it has to be clear, certified. So for clinical purposes and someone needs to make sure that the quality is good and so it would be more expensive. But it's not crazy expensive compared to an MRI, which can cost five to seven thousand, right?

But in terms of sequencing, when you are interested in a specific mutation in which it's harder to assay than for research, it would cost you $600 to do a whole genome. But if you want to do it for clinical purposes and you need also an interpretation and these kind of things, that can range up to $5,000, and it goes up, depending on which level of quality you have, to significantly more.

Glen Stevens, DO, PhD: If you have someone that has epilepsy and they have a polygenic risk score that's at a certain level, how does that help you clinically with the patients, or doesn't it?

Dennis Lal, PhD, MS: Yeah, so the polygenic risk scores for epilepsy, they are not established. So this is just purely research. And for most diseases, I think it changes a little bit for atherosclerosis and Alzheimer's. But yet these polygenic scores for the common types of disorders, that's purely research. But people are trying to implement them. However, when it comes to sequencing, where you look at these monogenic disorders, the single mutation which makes it all, that can have, for epilepsy as a good example, have game-changing care paths. For example, there are people who have mutations in the voltage-gated sodium channel. If they have mutations which lead to increased function in that same gene, then you use sodium channel blockers, and the child has controlled seizures.

However, if you have a mutation in that same gene in channel at a different position, then you might have loss of the channel function. So here you want to avoid sodium channel blockers because it makes it worse. And this an example where the gene alone is not enough information. You need to have the interpretation of the individual variant. And I think there are also similar, very good examples, in cancer space, where the variant information is very, very important.

I have one wonderful example of how genetics can help clinical care. It's a rare disorder, or it's relatively rare disorder, where children have mild microcephaly seizures and sometimes also movement problems. So they're pretty much pretty severely disabled after the second year of life. And because they have mutations in the glucose channel, and because this mutation leads to loss of the glucose channel function, or glucose transporter... however, if you give them ketogenic diet, because then it's a different kind of metabolic pathway, they're basically cured. So they develop normally. However, if you look clinically at the child, you wouldn't be able to come up with a hypothesis that this child has this specific glucose transporter version of epilepsy, because it looks just like an epilepsy clinically. So you would need to do the genetic testing to come up with this hypothesis.

Glen Stevens, DO, PhD: So you mentioned some of the milestones that have been going on. What's the next frontier, what's the big thing coming next?

Dennis Lal, PhD, MS: So I think for the rare diseases it's really making sure that the right patients receive the correct therapy. I'm learning about the full magnitude of genes, which can lead to a disease. And while also learning the clinical correlates, which give you more confidence that the gene where you found the mutation is truly the reason for the disease, so that you have also clinical validation. And that's very, very important. So hopefully then enrolling the right patients and having the right dosages of the targeted therapies, that they will be successful.

When it comes to the more common disorders, we are still behind. Let's say, talk about Alzheimer's, Parkinson's or let's say neuro degeneration disorder, even common forms of epilepsy. Because here, the genetic architecture is a little bit more complicated. Many, many, many mutations play a role. And also the environment plays a role. And typically what we do is we only look at the patients once they have the full blown disease, and for later onset diseases that might be too late. So understanding the biology, which is really at the beginning of the disease, before it's clinically recognizable... understanding that one is probably the most important step that we can identify disease modifying therapies that really help. And it's not just act on a brain, for example, which is already half gone.

Glen Stevens, DO, PhD: Yeah, it would seem like you would have to intervene with a lot of the diseases very early on.

Dennis Lal, PhD, MS: Yeah, correct. So I think for the pediatric diseases, you naturally do this. It's a developing brain, and if you don't meet a specific milestone, that is very fast recognized. However, if someone has mild memory problems, or is not anymore on the top level, what they want where, I think it's not so easy to recognize because aging is also a process which happens once you're older. And typically, as I mentioned before, people only really do all the deep biomarkers, such an MRI, such as some movement of the test, once the disease is really obvious to everyone, and then they go to the clinic and that might be point in time which we're rather late to intervene effectively.

Glen Stevens, DO, PhD: Good. So other areas that we need to talk about? Anything that you want to really discuss that we've missed?

Dennis Lal, PhD, MS: I'm hopeful then very happy that we soon will have the Cleveland Clinic brain study, which is one of the first and probably the biggest prospective study to study human brain disorders, where we exactly try to identify these biological changes that occur in the aging population prior to disease onset.

Glen Stevens, DO, PhD: So I understand then that there's a large clinical trial that you're going to be starting, a brain study looking at neurodegenerative disorders. Would you like to discuss that for us a little bit?

Dennis Lal, PhD, MS: Yeah, thank you. So, one of the biggest challenges is that we see most patients coming to the clinic when they already have developed the disease. And there's a big hypothesis in the community that when a patient has already a clinically recognizable disease, then this is just the end of a long journey where the disease has developed. And at that point it's not reversible anymore. And the hypothesis is that for example, someone who develops Alzheimer's by the age of 60 may already have recognizable changes in their biology by the age of 40, or have some, for example, sleep problems or something else, which you may not intuitively would associate with a harm, as the first marker for the disease. But in theory, if you would be able to identify that someone is having the mildest version of the disease already 20 years ago, you may use the drugs which are out there today to not treat the disease, but really modify it to a degree that the person will never develop the severe disease.

And that's why we started here at the Cleveland Clinic, the so-called Cleveland Clinic brain study, where we look at people with multiple sclerosis and healthy individuals, just without any additional criteria, that don't have a neurological disease over the age of 50, and monitor them every year and very detailed for more than 30 assessments, most of them assessing the brain, from neuro psychological testing to MRI, to EEG, to PSG. But even something like echocardiogram. They get a genome and the transcriptome and so on and so forth.

We really monitor them, every year, in a standardized way with technologies from the world's experts. And then we follow them over time and then observe who develop for example, stroke and who will develop for example, Alzheimer's. And once we have done this, we can then look back in the data and ask for those who developed 10 years later Alzheimer's, how did their biomarker or clinical profile look like 10 years ago? Maybe that will give us some clues of the earliest risk factor or clinical recognizable markers for the disease. And you can even look in the molecular biomarkers and see if there maybe are some pathways which are dysregulated, which might be drug targets.

Glen Stevens, DO, PhD: And how many patients are you going to look at for the brain study?

Dennis Lal, PhD, MS: Yeah. So this is still something which is in development. So we are very fortunate. We got more excitement than we expected at the beginning. So we will start with a few thousand and the study is open ended. So we go for 10 years now, but it may go longer and it seems even at this point, even likely that we will enroll significant more people.

Glen Stevens, DO, PhD: Anything else you want to talk about that we haven't discussed?

Dennis Lal, PhD, MS: No, but maybe you have some thoughts on gene therapy from the glioblastoma field?

Glen Stevens, DO, PhD: I'm a neuro-oncologist in the most common primary malignant tumor of adults is a glioblastoma, that has a very high mutation rate. Historically, we've looked at blocking different pathways and of course, the tumor finds a different pathway to get down to grow and divide. There's been a number of genetic studies looking at putting retroviruses to try and treat tumor. Are you involved with any brain tumor treatment genetically?

Dennis Lal, PhD, MS: Not at all in terms of treatment. So what we are doing is we have a study here at the clinic and also some brief collaborators, where we look at low-grade pediatric brain tumors, such as gangliogliomas and astrocytomas and DNETs, and where we just do the genome sequencing. And we see quite bit the BRAF/MAPK variants. But we are not involved in any kind of treatment. But one thing what is quite interesting is, in the gangliogliomas, it's expected that 50% of them will have the BRAF B600E variant. And the interesting thing what we saw in our data actually is that we see the 50%, but we see among those 50% that many of them have large structural variations. But not all. And we are looking now complete architecture of the genetics. Not only the individual point mutation, but also the other rearrangements that might have any clinical implications. But it's not therapeutic at this point.

Glen Stevens, DO, PhD: Interesting. We'll see on occasion BRAF mutations in glioblastoma, but it's not very common. But it does give us a targetable mutation to treat. So Dr. Lal, we have the benefit here at the Cleveland Clinic of having a Genomics Institute, but there may be a lot of physicians out there that don't have easy access. If I'm seeing patients with neurodegenerative risk and I'd like to have some involvement, how would I go about doing that if I don't have it at my hospital?

Dennis Lal, PhD, MS: This is not a question which is easy to answer. So overall, and one of the fastest growing and most demanded areas of medicine is genetics. And in particular, there's this relatively new type of position which is called the genetic counselor. This is an individual which is specialized typically on the disease type and the genetics of this disorder, and they are really trained in communicating genetic findings to patients. And they work together with a geneticist, molecular clinical geneticist, and that is probably the best way moving forward.

However, when I say this, I'm giving international meetings at the Cleveland Clinic, but also outside, a lot of training workshops for neurologists. One thing, what I really have to stress, and that's my personal opinion, but it's just experience what I observe, is that for many types of diseases, genetics is a really fast growing field because it can explain a significant amount of patient and it's growing. If you work as a treating physician in an area where you make the same observation that genetics seems to become more and more important, I really, really, really encourage you to not hope that the genetic counselor be able to guide you in terms of treatment, because that will not happen. The genetic counselor or the geneticist they will only be there to make sure that the variant is pathogenic or not. What kind of treatment would fit that patient, based on the genetic etiology, that will be the decision of the clinician. And also clinically, treatment in genetics is very close to research. So in many of the drugs or care pathways ... are a little bit more exploratory.

To have a good intuition and gut feeling, if a patient would fit a certain drug ... and where a certain skepticism is important. This is only possible if you familiarize yourself with genetics. So for those where you have to look at a genetic test results, and this happens from many, especially neurological domains, I would really encourage yourself to re-educate yourself, because what we learned, even four years ago, is old school compared to what we know today in the field of genetics, which is a really rapid evolving field. And I know it's really demanding to keep up the pace, even almost impossible that you could take care about patients, but there might be people in your department to talk to. And I would just try to stay open minded and try to keep up with the information and also reach out to people who know, or have more experience. And because you will be able to do, you will be able to better care for your patient with all this new knowledge which is available.

Glen Stevens, DO, PhD: I will just echo what you just stated and that is that within the Cancer Center, we do have genomic boards where patients will have molecular testing, and that will be discussed at a molecular board so that there are individuals that have better understanding than others to help guide treatment options for patients, which is very helpful.

So Dennis, I'm looking forward to the future of gene therapy. It sounds very exciting. And it sounds like if I go to sleep, there'll be something new tomorrow. Looking forward to the hope that we all believe that it will bring us for ourselves and for our patients, and I'd like to thank you for joining me today.

Dennis Lal, PhD, MS: Thank you very much.

Closing: This concludes this episode of Neuro Pathways. You can find additional podcast episodes on our website, clevelandclinic.org/neuropodcast, or subscribe to the podcast on iTunes, Google Play, Spotify, or wherever you get your podcasts. And don't forget, you can access real time updates from experts in Cleveland Clinic's Neurological Institute on our Consult QD website, that's ConsultQD.ClevelandClinic.org/neuro, or follow us on Twitter @CleClinicMD, all one word. And thank you for listening.