Eva L. Feldman, M.D., Ph.D.
Professor of Neurology
University of Michigan
Q: In the last few years, we’ve witnessed significant advances in unraveling ALS and even testing potential therapies in clinical trials, but effective treatments still evade us. What’s been the problem?
A: The problem is two-fold. While we have made progress in understanding disease pathogenesis we’re still clearly not fully there, and that has made therapeutic development difficult—but not impossible. I think until we have a complete grasp on what is causing this devastating illness, we will not be able to completely cure it. However, I don’t at all think that this lack of knowledge needs to stop us from trying to develop new therapies.
The second part of the problem is that the therapeutics that have been tested have all failed, probably for multiple reasons. The obvious one is simply that the therapy was ineffective. Another reason has to do with the choice of endpoint measures—the clinical yardsticks by which we judge a drug’s effectiveness. This is still somewhat of a controversy in the field. Should we be looking at respiratory parameters such as how long the need for a ventilator is delayed? Should we look at arm strength, or the ability to walk? How do we define efficacy? These questions continue to be debated. Clearly no drug has been a home run, because if it were, you would see a positive effect regardless of the clinical parameter used.
A third reason for drug failure involves how it is delivered to the nervous system. Being able to get the therapy to the cells of interest is extremely important. It could be that a therapy might truly work if it could reach the motor neurons in the spinal cord and/or the brain area that’s affected. I think this has been a significant stumbling block for therapeutics. We are very interested in this path, as are many other laboratories, and we plan to begin a small pilot trial soon to determine if stem cells may be an effective way of delivering Insulin-like Growth Factor (IGF-1) directly to motor neurons (see below).
These are three key points that I think have hindered our ability to develop a therapy for ALS and bring it to fruition. We are left with no clear-cut therapy that is able to halt the progression of this disease to any significant degree. In terms of therapeutic treatment of this disorder, we’re not that much further along than we were in 1939 when Lou Gehrig was diagnosed.
Currently my hope for an ALS therapy is not necessarily a complete cure, though that of course remains the goal. But if we could discover a drug or intervention that would halt the progression of the disease—stabilize and control it—I would be very satisfied, realizing that a complete cure would still be pursued. You can liken it to diabetes: we may not be able to cure diabetes but we can use insulin to control it.
Q: Have the current mouse models of ALS failed us in terms of therapeutic development?
A: The ALS mouse model is very interesting. The most commonly used model is the G-93A mouse model, in which a mutated copy of the human superoxide dismutase (SOD1) gene, which is known to cause familial ALS in humans, has been placed in the mouse. These mice develop a disorder over time that clearly mimics ALS both clinically and pathologically.
One of the problems with the mouse model is that many therapies that have been tested in it and shown to be very effective have failed in human trials. That has certainly caused a high degree of frustration among not only clinical scientists but also patients. One explanation for this apparent disconnection between mouse and human results is that the SOD1 mutation that was used to develop the mouse model represents only a small portion of human ALS cases. Only 10 percent of ALS cases are inherited or genetic forms of the disease and of those only 20 percent are due to a mutation in SOD1.
The mouse model, therefore, clearly represents a genetic form of ALS and likely typifies some aspects of the 90 percent of ALS cases that are “idiopathic,” but it may not be a true, clear representative model of these idiopathic cases. The mouse is a good tool, but it is not a perfect tool.
Q: What needs to be done to overcome these obstacles and advance progress in therapeutic development for ALS?
A: Many things need to be done. The obvious need is to develop better animal models. We are clearly going to continue to need more and improved models as we try to understand what is causing the disease and test therapeutic hypotheses.
Beyond that, there is this idea out there of conducting what’s become known as a futility trial, an idea I’m very interested in but am just starting to learn about. Essentially, you set up a trial design where you need many fewer patients for a much shorter time to test whether or not a drug may have a clinical benefit. These kinds of novel trial designs may make it feasible—as long as the drug is shown to be safe—to replace animal-model testing of promising drugs with a human model. I’m completely serious about this. [Johns Hopkins neurologist] Jeffrey Rothstein just conducted a futility trial using the antibiotic ceftriaxone, and with only 60 patients they were able to tell that this drug may potentially offer a new therapy for ALS. Now they’re going to do a larger trial with 600 patients.
The idea is not only to be thinking about different animal models, but also to look at our human trial designs and how we can investigate the potential therapeutic of a drug using smaller numbers of patients. Let’s say I find a new drug in my lab tomorrow and I show that it works in the ALS mouse. Should the next step be a big, double-blind, placebo-controlled trial in some 300 patients? If I have a promising drug, and I know it’s safe, wouldn’t it be better if I could say whether or not it is effective by using a small number of patients?
Q: You have advocated for further study of insulin growth factor (IGF-1) as an ALS therapy, yet a recent IGF-1 clinical trial failed to show a benefit. What’s the best way forward in your view?
A: I think that targeted drug delivery is the way to move forward with IGF-1. I am almost certain that IGF-1 could provide a high degree of neuroprotection to motor neurons under attack in ALS if IGF-1 could be delivered to the site of injury. From my point of view, one of the reasons that the recent 330-patient, double-blind, placebo-controlled IGF-1 trial directed by the Mayo Clinic failed is that IGF-1 was given twice a day by subcutaneous injection. We had no clear evidence that IGF was going to reach the nervous system using this delivery route. In a mouse model of ALS, if you give IGF-1 via subcutaneous injection, it is not effective. However, if you deliver IGF-1 via gene-therapy, with the idea that it can actually reach the spinal cord and the motor neurons that are diseased, it is effective in blocking the progression of ALS. The human trial was not a gene therapy trial but rather a trial in which IGF-1 was delivered subcutaneously by injection, so it is not surprising that it failed.
I think gene therapy is the easiest way to do targeted drug delivery with IGF-1. There have been several approaches to doing this so far. One has used an adeno-associated viral vector, which is one step up from the common cold virus. The infectious part of the virus is removed and the gene for IGF-1 is inserted, producing a vector that is genetically engineered to produce IGF-1. You can then inject these vectors into muscle where they will be taken up by nerve terminals and transported back to nerve cells. It’s a very good approach; it clearly works in animals and one can envision it working in humans. The problem is that it takes a lot of injections, and the injections have to be near motor nerve terminals. You might need to make several hundred injections to get enough IGF-1 transported back to the spinal cord for it to have an effect. While the concept is good, I think it may not be completely practical.
A more practical approach to gene therapy may lie with the rabies virus. The rabies virus expresses a “coat” protein that has a high affinity for motor neurons. When someone is infected with rabies from an animal bite, the virus is selectively picked up by motor neurons via the coat protein and carried back to the spinal cord. The idea is to isolate the rabies coat protein and incorporate it into a viral vector producing IGF-1. In this scenario, you create a viral vector that is both making IGF-1 and that motor neurons selectively latch onto. Some people are even asking if you could deliver the IGF-1-producing vector by some sort of local or topical application that is absorbed into the skin rather than injecting it, to eliminate the need for multiple injections. The approach may or may not work, but at least it’s a way to be thinking toward new therapies for the future.
Q: You are now collaborating with a biotech company to use stem cells to deliver IGF-1. Where are you in that program?
A: We are pursuing an approach to induce neural stem cells (stem cells that are committed to becoming nerve or glial cells) to make more IGF-1, then to transplant these cells into the spinal cord or brain near the diseased area. The idea is to have these stem cells produce and deliver IGF-1. It would be analogous to delivering chicken soup to a sick neighbor. This is something that I definitely think holds therapeutic promise.
Toward that end, we are engaged in two research tracks. We have a rat model of ALS that carries the same genetic defect that the mouse model does, but in a larger animal. This allows us to do intraspinal transplantation of stem cells that produce IGF-1 to see if we can slow progression of the disease and prolong meaningful life in the rat.
We have also applied to the FDA for approval to transplant neural progenitor cells that endogenously make IGF-1 into the spinal cords of our patients with ALS to see if there is therapeutic benefit. It’s a fairly aggressive approach requiring a one-time surgery that will entail multiple injections of the stem cells.
We have designed the study as a “ramped” Phase 1 safety trial, which means we’ll do it in steps: enroll three patients, watch these patients carefully and if they do well, enroll three more and so on. If we can show safety then we of course want to conduct a larger trial. We have injected these cells safely into the spinal cords of pigs, which are very similar to humans’, and the pigs did extraordinarily well. Based on that work, we believe we can safely transplant stem cells into human spinal cords.
I am the principal investigator on the trial, and we are doing it in collaboration with Jonathon Glass and neurosurgeon Nicholas Boulis at Emory University, where the trial will be physically conducted. The trial is funded by a company called Neural Stem.
Dr. Clyde Svensen’s group at the University of Wisconsin is also looking at stem cell therapy in ALS, and has worked with Dr. Boulis to produce very promising results.
Q: If approved, this will be one of the first trials in the country to investigate the use of stem cells for a neurological disease, a big step forward that some argue we’re not ready for. What has been the reaction from the scientific and patient communities?
A: The ALS community is amazingly collegial and supportive but there are some physicians who feel that this is too aggressive or too bold of a move right now, based on the scientific data that is available on stem cells. I wouldn’t say that is the prevailing wisdom among the ALS scientific community, but rather that the jury is still out. I have presented the human trial only a few times publicly and it has been embraced, but there have been investigators who are concerned about doing surgery and transplanting these stem cells into a patient with ALS, who is already debilitated.
We understand that transplanting stem cells directly into the spinal cord is a very bold step, but in this aggressive disease we think that bold steps need to be taken. Of course I can’t be sure that it is going to work, but I think that it holds sufficient therapeutic promise to go forward. The patients are very enthusiastic, and as long as they fully understand the potential risks, I think it is something that we should pursue.
You know, right now, when I diagnose a patient with ALS, I have nothing I can offer them. I have patients who are going to China or Mexico to receive stem cell transplants. I would rather have them stay here and know that we are giving them the best possible care, which someday, we hope, will include stem cell therapy.