Talking Trials

by Amy Madsen on Mon, 2013-04-01 09:19

Clinical trials expert Merit Cudkowicz discusses why it’s so hard to find therapies for ALS

Article Highlights:
  • Two of the biggest challenges in designing effective ALS clinical trials are late diagnosis and disease heterogeneity (variation in disease characteristics from person to person).
  • Developing an understanding of ALS disease heterogeneity must become a priority for researchers. This might be done by analyzing biofluids, imaging data and post-mortem samples from patients.
  • There are many more scientists and clinical investigators working in the ALS field today, generating more ideas about what might explain disease and progression. More therapeutic targets are being identified, but there is still much to do and learn.
  • A sidebar gives an example of ALS heterogeneity in a current stem cell trial.

Put together the words amyotrophic lateral sclerosis (ALS) and the phrase clinical trials expert, and one name that is sure to follow is Merit Cudkowicz.

Cudkowicz is chief of neurology and director of the Neurology Clinical Trials Unit and the ALS Center at Massachusetts General Hospital (MGH) in Boston. The longtime MDA grantee also directs the MDA/ALS Center at MGH (one of five sites in MDA’s ALS Clinical Research Network).

Merit Cudkowicz

In 1995, to help advance clinical testing in ALS and other motor neuron diseases, she co-founded the Northeast ALS Consortium (NEALS), a vast clinical trials network that brings together academic investigators and companies, research sites and scientific resources including an ALS biorepository and several trial databases.

In a February 2013 discussion, Cudkowicz described the current state of clinical trials in ALS, including the challenges they face, what investigators need to know to make them successful, and which ones to (maybe) get excited about.

Q: Why is it proving so difficult to find effective therapies for ALS?

A: The primary reason is that we don’t understand fully the underlying biology and disease heterogeneity enough to develop highly effective, targeted therapies.

We need biomarkers to help understand disease heterogeneity, and to provide the ability to accurately and reliably measure treatment effects in people.

More science is needed to understand the cause of disease, and targets at which to aim therapies. It may be that different treatments will be needed for different people with ALS — maybe even different treatments at different times in the same person.

Q: What would you say are the biggest challenges that investigators must keep in mind when designing and conducting ALS clinical trials today?

A: It isn’t necessarily the trial designs that are the challenge. Given the experience over the past several years in ALS trials and regulatory discussions, the field has a lot of knowledge about how to design ALS trials. There are established outcome measures, excellent trial centers and knowledge of the regulatory requirements.

Disease heterogeneity, late diagnosis, identification of the most relevant biological target and determining whether this target is the same in all patients — these are some of the greatest challenges.

Q: Can you talk about heterogeneity in ALS and the difficulties it presents with regard to trials?

A: Heterogeneity is a huge problem. Variability is large between patients. Survival can range in ALS from, unfortunately, less than one year to greater than 20 years. Ten to 15 percent of people with ALS may experience no progression over a yearlong period of time. Another 10 to 15 percent decline three to four times faster than average.

When we first see people, we cannot tell what progression will be like, as there are very few prognostic indicators. Because of this variability, currently phase 3 efficacy studies — any efficacy study! — needs to be very large.

Also, it is very likely that patients with a slow course versus those with a fast course have very different biologies, and it may be that they would respond to treatments differently. Treatments may need to be targeted differently to different forms of ALS, and for this to happen it must be a priority to develop understanding of the disease heterogeneity. This might be done by analyzing biofluids, imaging data and post-mortem samples from patients who have well-described phenotypes and genotypes.

I firmly believe we should learn from every patient by collecting and sharing clinical data, genetic data and biological samples. A national effort should be made to get clinical and genetic data, and biofluids, and to share widely with ALS scientists all over the world. NEALS has started this, but it needs to be done on a much larger scale.

Q: Why aren’t we having any success with drugs that make it to phase 3 clinical trials, such as the recent failure of dexpramipexole?

A: It is not unique to ALS that drugs fail in phase 3 trials. Unfortunately, this is more common than it should be for neurological disorders. Part of the reason is that the tools to determine what a therapy is doing in the central nervous system, or CNS, are limited — though there is a lot of new work going on now to develop better imaging tools for the CNS.

Also in ALS, because the disease is very heterogeneous, small phase 2 studies are unable to predict effect size. While there has been great progress in understanding ALS, there is still much more to learn to better guide therapy development. For the majority of people with ALS, we don’t yet know the best targets.

There has been great progress, but there is still much more to discover and understand about ALS to help lead researchers to better treatments. Therefore, it is important to study ALS in people using biomarkers, imaging and genetic tools, and any other available tools and methods, to really understand what is going on in the body, as well as how it changes over the course of the disease. We need to know: Is it the same in all patients? At all times? This information will lead to much more targeted and ultimately effective treatments.

Q: Who designs clinical trials?

A: It is a team effort involving experienced clinical investigators, biostatisticians, nurses, physical therapists, project and data managers, pharmacologists, patient advocates and representatives of the study sponsor. For phase 2 and 3 studies, there is often input from the U.S. Food and Drug Administration (FDA) as well. It is a back-and-forth process, with people from industry and academia and the FDA identifying the question that needs to be answered, and then discussing trial design and the best approach to getting that answer.

Q: What can scientists do to make sure the results from a phase 2 trial will reliably signal whether a phase 3 trial is warranted?

A: Phase 2 studies are very hard to design in progressive disorders like ALS that have such high heterogeneity. There is no single answer here, but often more than one phase 2 trial is needed to address all phase 2-type questions before proceeding to phase 3.

There are many types of phase 2 trial questions. For example, what is the best dose? Does the drug get to its intended target and have the desired pharmacological effect? Determining the answers to these questions isn’t always possible in people, as the tools may not be available to do it, or the trial may lack in sample size or duration. We need good biomarkers — such as the use of imaging in multiple sclerosis, which was instrumental in the development of successful treatments for that disorder. We don’t yet have those in ALS.

Ultimately, a good phase 2 study provides the answers to important questions that you need to know before moving on to a phase 3, such as those concerning dose, target engagement, safety, outcomes feasibility and enrollment feasibility. But these trials are rarely, if ever, sufficiently large or robust enough to give confidence on efficacy.

Q: Are changes needed for phase 2 trials?

A: Yes and no. The phase 2 trials that have been done so far have not been bad studies. They may not have had sufficient data to proceed to phase 3 testing, though. The idea that you only need one phase 2 before going to phase 3 may be wrong for some drugs. Sometimes you need to do several phase 2 studies — enough to give you all the important data, such as proper dose, outcomes, duration, safety monitoring and pharmacokinetics — that you need for phase 3.

What’s problematic is trying to use phase 2 trials to say anything definitive about the efficacy of an experimental treatment. Phase 2 studies are rarely big enough or long enough to provide that type of confidence. I think the expectations from phase 2 trials have morphed in ways that lead people to make incorrect interpretations and think they can say something definitive about efficacy. They don’t. Today, with the tools and knowledge we have, it is not possible to determine efficacy with confidence from a small group of patients in a disease as heterogeneous as ALS.

Q: How do you address dropouts in a trial? What effect do they have on the reliability of any observed differences between participants?

A: It’s very important to make sure that if a trial participant stops taking the drug that’s being tested, that they agree — if possible — to allow complete follow-up. Missing data can really make interpretation of study results difficult. In addition, the study can lose the power to detect an effect if there are a large number of people who stop taking the treatment and from whom outcome data can’t be collected.

Participants in clinical trials are our partners in an effort to find treatments. It would not be possible to find a treatment for ALS without the people who have the disease and their families as a vital component of the research. We learn from every study, moving the field closer to meaningful treatments.

Q: Will genetic testing for known genetic mutations in ALS become important for future clinical trials?

A: Yes. It is going to be important both clinically and therapeutically. We may be close now to a point where genetic testing will be important not only in clinical trials, but also in clinical care at ALS centers.

Q: Where are we at with outcome measures in ALS clinical trials?

A: The outcome measures for phase 3 clinical trials in ALS are good. The ALS field is way ahead on this. We have several clinically relevant outcome measures — the ALS Functional Rating Scale-Revised, measurements of breathing function such as forced vital capacity, survival and muscle strength, for example. These are easy to obtain and accepted for use in trials by the FDA.

Q: How do you design a trial so that it will be "successful" — valuable — even if the drug fails in testing?

A: The first step is to identify clearly the primary study question. In phase 2 trials, for example, this may be selecting the right dosage that is safe and that engages the intended biological target. Only by doing this will you be able say definitively that the study has addressed whether the drug in testing works — or doesn’t — on the proper mechanism.

In the end, this is easier said than done because we don’t have great ways to tell in people with ALS or other neurological disorders whether a drug gets to its target in the brain and has the desired biological effects. If it is possible, then those studies should be done first — in the phase 2 testing of a compound.

Q: What issues do trial designers face when determining measures of success in a clinical trial?

A: In the beginning stages of designing a study, one of the first key questions you ask is what clinically meaningful effect do you expect to find? You figure out the sample size — number of people in the trial — so as to allow the study to be able to pick up that effect with statistical significance.

For example, if you treat 10,000 patients with a drug, you might find that a 10 percent change in the ALSFRS-R score is statistically significant. But this would not be considered clinically relevant. A study in 100 people, on the other hand, might only be able to pick up a very large effect, which may or may not be feasible depending on the treatment. It is important to pick an effect size that is both clinically meaningful and feasible to achieve. The effect size, along with considerations of disease heterogeneity, will determine how many people need to be enrolled in a study.

Q: What are the most promising drugs in testing right now?

A: There are several promising therapies under development, such as Cytokinetics’ CK-2017357 [Tirasemtiv], GlaxoSmithKline’s ozanezumab, mexilitine and fingolimid [Gilyena]. In addition, several groups, including Neuralstem and BrainStorm, are working on stem cell approaches, and other scientists are working on ways to modify genetic mutations.

Q: Why should people be excited about ALS research right now?

A: There are many more scientists and clinical investigators working in the ALS field today — more ideas about what might explain disease and progression, and more therapeutic targets being identified. There is still much to do and learn, but discoveries are rapid.

Q: What about clinical trials — what should people be excited about with regard to trials?

A: A lot is known about the pathway to getting drugs approved in ALS. This is a huge advantage for the field. The outcome measures and the proper way to approach trials are known. The network of clinical sites and patients exist, so trials can be done well throughout the U.S., Canada, Europe and Australia. And there is a growing pipeline of therapeutic targets and agents under development for people with ALS.

We have an extremely collaborative community sharing ideas and data, all with a shared passion to find treatments. The ALS field is leading the way in bringing gene-modifying approaches, stem cells and new small molecules to patients.

Q: Final thoughts?

A: Now is the time for a national effort to ensure that at all ALS centers’ clinical data, genetic data and biological samples are collected and shared, so that we can learn as much as possible about ALS from people with ALS, and so we can understand better the heterogeneity and underlying biology of the disease. From this, targets and therapies will flow.


Ted Harada

Neuralstem’s phase 1 trial of the safety of human spinal cord-derived stem cells demonstrates the variability of response that can be seen in a clinical trial. Ted Harada, 41, of McDonough, Ga., was one of several individuals with ALS who participated in the study that was designed to look at safety of the procedure, not efficacy. While the procedure was found to be safe and well-tolerated, Harada’s results were not typical of the other participants in the trial — surprisingly, he experienced an increase in strength and function. Some had slowing of disease progression, and others had no apparent benefit. Six trial participants have died (five of ALS progression). The variability in response is not at all understood at this time, but is an example of the biologic heterogeneity of the disease. Says Cudkowicz, “More science is needed to understand the cause of the disease and targets at which to aim therapies. It may be that different treatments will be needed for different people with ALS — maybe even different treatments at different times in the same person.”

Amy Madsen
Your rating: None Average: 4.7 (6 votes)
MDA cannot respond to questions asked in the comments field. For help with questions, contact your local MDA office or clinic or email See comment policy