Episode 97 with Dr. David Baker on MS drug development and trial design
The process of taking a drug from bench to bedside is often challenging but always enlightening. In the first installment of a 2-part interview, Dr. David Baker shares what he's learned in his work developing new drugs for MS.
Professor of Neuroimmunology
Barts and The London School of Medicine and Dentistry
Queen Mary University of London
Full transcript below.
- References below.
Note: Each podcast includes an interview with a thought leader or newsmaker in MS and related demyelinating diseases. Listen to it here. Alternatively, you may subscribe to the podcast via iTunes or your favorite podcast app. In iTunes, for example, click File/Subscribe to Podcast and then enter this URL: http://msdiscovery.libsyn.com/rss
The 2015-16 series of MS podcasts is supported in part by a generous grant from Sanofi Genzyme. The content remains the sole responsibility of the Multiple Sclerosis Discovery Forum, an independent non-profit news organization.
Host – Dan Keller
Hello, and welcome to Episode Ninety-seven of Multiple Sclerosis Discovery, the podcast of the MS Discovery Forum. I’m Dan Keller.
Today's interview features Dr. David Baker, Professor of Neuroimmunology at Queen Mary University of London in the UK. We spoke at the ECTRIMS conference last fall, where I asked him about his work with cannabinoid compounds – work that has led to a better understanding of the cannabinoid system as well as to candidate drug compounds to treat spasticity.
Interviewer – Dan Keller
In terms of what you're doing now with cannabinoids, can you tell me what you are looking for, and what you've found?
Interviewee – David Baker
Many, many years ago, we actually were probably the first people to show that cannabis can actually alleviate muscle stiffness in animal models of multiple sclerosis, which then kind of underpinned the push to look for cannabis in MS. So people with MS were smoking cannabis and perceiving benefit. The question was, why? And what they didn't really understand that there was going to be an unfolding biology. And a few years later after our first discovery that actually cannabinoids can cause relaxation of the muscles, we understood that the function of the cannabinoid system is to regulate nerve signaling. And so because the cannabinoid system does regulate the strength of synaptic signaling, then it's obvious that it can inhibit signs and symptoms because of this excessive neurostimulation. So at the time of that, then we realized that the receptor is a CB1 receptor, and the compound within cannabis is THC, and they're the same molecules that cause all the side effects. So you could never really disassociate away the high from the medical benefit. So we started to think, well, how can we try and get the medical benefit from the cannabinoid system and at the same time try and limit the side effect potential.
So what we thought is, well, if we can stop the cannabinoid molecules getting in the brain, then they won't cause the side effects. But maybe we can target the aberrant signaling in the spinal cord and the peripheral system to try and get the benefits. And so that was our intention. So we tried to make a CNS-excluded drug. And that's, in fact, what we did. We made a drug that was very, very water soluble, so you know, you use the mechanism of the blood-brain barrier to actually exclude it from the brain. So we made the compound, and a few weeks later, we kind of started putting it into animal models, not really doing it the pharmaceutical way, which would be a methodical testing. So we showed that it didn't cause any of the unwanted side effects that are associated with cannabis in the animals. And then we put it in a system where we had a spasticity in a multiple sclerosis relevant system, and the drug worked.
Now what we did know is that the drug was blocked by the activity of the CB1 receptor antagonist, so it looked like we'd made what we set out to make. So we were really excited. And from that point, we started to try and see if we could develop it as a drug. Unfortunately what we realized very quickly actually is that it doesn't work by the known cannabinoid receptor system, and I think what we stumbled across is a whole new biology of the cannabinoid system.
And so we've been developing this drug bit by bit. We set up a university spinout company to try and develop it as a pharmaceutical drug. And over the years, bit by bit, we've been pushing it forward. So it's very safe in animals. It has a massive therapeutic window. And with grant funding agencies etc. we've managed to be able to take it into phase I study where it passed with flying colors. We tested it in 60 healthy humans. And a few weeks ago, we started our first testing in people with multiple sclerosis. So we'll have to see how it works. But we hope by early in 2016 we'll have the answer. So it could be a symptom modifying drug, but it doesn't have any of the side effects associated with drugs such as, you know, Sativex or baclofen as well. So it's not sedating as far as we know.
The way that the drug works is a new mechanism. And what we can probably say is it serves to block the excitation of nerves. So it dampens down excessive signaling, which are probably the consequences or the causes of spasms and spasticity and possibly the symptoms as well and maybe pain. We just have to do more work to see if it will work that way.
Is this a hyperexcitable system? Or is it a hypoinhibited system where you're getting this spasticity?
Well, I think spasticity is largely caused by loss of the inhibitory circuitry. So there's probably less GABAergic signaling. And so one can, you know, drive the inhibitory system, like you do with GABA, but likewise you can actually kind of block the excitation. And this mechanism actually probably only exists in pathology. So this is probably why there isn't the side effect potential that the real target that we're actually after really only occurs when the nervous system is going a bit haywire. So that's why we think we've got good safety margin.
And you had told me that this does not induce hunger, which I guess is another sign that it's not getting into the CNS?
Having said all that, it was made not to get in the central nervous system, but in reality, it doesn't matter if it does get in the central nervous system. So in fact, about 15% of the drug does get into the central nervous system, which would be as good as many drugs that are CNS penetrant. I guess when we were starting, we were hoping that, you know, it was going to be excluded because we thought it was a cannabinoid receptor agonist, but in reality, it doesn't matter. And in fact what we know is actually this targeting into the lesions. So there's actually more goes into the area. And what this kind of spins on to some other work that we've done with some of our sodium channel work.
We've been developing new sodium channel blockers as potential neuroprotectants. And what we've done is certain molecules actually get excluded by CNS drug pumps, and what we'd noticed in MS is that some of these drug pumps disappear. So we made a drug that was actually targeted specifically to one of those drug pumps, which would normally mean it would be excluded from the brain, but what we showed is that with these new sodium channel blockers, that actually they physically target into the lesion where the pump disappears. And so again, you increase this therapeutic window between effect versus side effect, because again with the sodium channels, you need them for health. You block them, and you have side effects. But what we've found with the sodium channel blockers is that in the animal models, sodium channel blockers were neuroprotective, and we then took that idea forward actually into the clinical trial.
So we first of all thought the trials with sodium channel blockers had failed. Why had they failed? Well, the reason they failed was the trial outcomes weren't right, and suddenly actually because of this unpleasant side effect, 50% of the people didn't take the drug. So the trial was doomed before it ever started. And then what we had was we had the bloods of the people in the trial. So we looked two years after the trial had finished and was seen to be a failure, and we found that 50% of people weren't taking the drugs. But if you look to the people who were taking the drugs, we could see that there was less neurofilament in their blood indicative that there is less nerve damage. And so actually in reality, the trial actually was positive, but it was seen to be negative because of this failure to take the drug.
So the question was, how could we then develop that forward? So the clinical guy said, well, let's think how we could best do a quick trial. And they came up with the idea of the optic nerve being the ideal target. And so what they said to us was, can you, you know, model this in the animal model? So we developed a new animal model. So we took Vijay Kuchroo's 2D2 mice, which are preprogrammed to get optic neuritis, and then we just made their eyes florescent so we could just look in their eyes and see nerves in real time and in life. And as a consequence of using the transgenic, which targeted myelin oligodendrocyte glycoprotein, the cells would go in, cause optic neuritis, that would cause nerve death, and then we could monitor the nerve death just by looking into the eye, because each nerve was labeled with a fluorescent protein. We'd see one single nerve die.
And so we started to use that as a way of testing different drugs for neuroprotection. And we put a whole stack of different compounds, minocycline, sodium channel blockers, glutamate receptor antagonist, we did a few. And we got some hits with the sodium channel blockers, and we tried a few of the different ones, some of them better than others. And unfortunately the one that they chose for the trial is probably the worst one in the animals, but they decided that you had to load drug quickly, so they selected phenytoin. So we showed that the sodium channel could work in the optic neuritis, and then the idea was then we translate that and then do a trial with optic neuritis in the human.
So this was a trial that Raj Kapoor did. And so the idea was that people go blind, and then you go to the doctors. And then they were randomized to either get steroids, which is the standard treatment, or they'd get steroids plus a sodium channel blocker, which was phenytoin at the time. And that was done because you can dose very quickly. So the idea was to get people on drug very quickly. So within seven days of their first symptom, people were on active drug. And people were treated for about six months. And then they looked at the retinal fiber thickness. So as a consequence of the ganglion in the retina dying, the retina thins, and then you can measure that with a machine called OCT, optical coherence tomography. And that was slowed. So they saved 30% of the nerves from dying, even though there were people getting a steroid. So it tells us that really certain channel blockers are neuroprotective.
And then the question is, is how then can we show that in reality? So what we've done from there is we've actually gone on with another sodium channel blocker, which was called oxcarbazepine, which was much more effective in the animal models. And we've been trying to initiate a new trial design whereby we're looking for people who are on current DMTs by showing evidence of neurofilament release, which is indicative that their nerves are being destroyed, because as the nerves are destroyed, they liberate their contents, and then we can pick that up in the biological fluids. So the idea is that if they've got neurofilaments in their cerebrospinal fluid, they get the option of having a sodium channel blocker in addition to their DMT. And then we'll monitor them by serial lumbar punctures to see if the neurofilament levels decrease as a way of a trench push on the trial design for phase II.
Because if you're thinking about the standard phase III, phase II trial for neuroprotection, you're talking about a two- or three-year trial, which will take you two years to recruit the 600 people and another year to do the analysis. So you're really talking about a seven-year trial with 600 people. This trial design will kind of push it down probably to 12 months to 18 months with 60 people. So we can do 10 times more people and a lot quicker this way. So that's started where we've been recruiting, and we're still recruiting, but fingers crossed that would be another way forward in terms of developing neuroprotection. I think it shows how we've been trying to use our animal models to translate things into the human. Because at the end of the day, there has been really, really poor translation between the animal models and humans. And I guess the question is, is why?
We’ll pick up on that question in part two of our interview with Dr. Baker next time, when he’ll describe some of the deficiencies he sees in the design and interpretation of animal experiments and how they could be improved to better relate to clinical trials and the clinical situation.
Thank you for listening to Episode Ninety-seven of Multiple Sclerosis Discovery. This podcast was produced by the MS Discovery Forum, MSDF, the premier source of independent news and information on MS research. Msdiscovery.org is part of the nonprofit Accelerated Cure Project for Multiple Sclerosis. Robert McBurney is our President and CEO, and Hollie Schmidt is Vice President of Scientific Operations.
Msdiscovery.org aims to focus attention on what is known and not yet known about the causes of MS and related conditions, their pathological mechanisms, and potential ways to intervene. By communicating this information in a way that builds bridges among different disciplines, we hope to open new routes toward significant clinical advances.
We’re interested in your opinions. Please join the discussion on one of our online forums or send comments, criticisms, and suggestions to firstname.lastname@example.org.
For Multiple Sclerosis Discovery, I'm Dan Keller.