Episode 14: Rethinking Hard Problems in Brain Science
When it comes to exploring the mindboggling complexity of living systems—ranging from the origins of human consciousness to treatments for neurodegenerative diseases such as Alzheimer’s—Susan Fitzpatrick has long been a critic of reductionist thinking. In this episode we talk with Fitzpatrick, who has spent three decades supporting brain research as president of the James S. McDonnell Foundation, about new ways to understand the human brain, the difficulty of developing an effective Alzheimer’s treatment, and how scientific research can more successfully confront complex problems.
- The James S. McDonnell Foundation website
- Susan Fitzpatrick’s review of Metazoa by Peter Godfrey-Smith and Life’s Edge by Carl Zimmer
- Her review of Brains Through Time by Georg F. Striedter and R. Glenn Northcutt
- Her review of Mind Fixers by Anne Harrington
- Her review of Chasing Men on Fire by Stephen G. Waxman and Understanding the Brain by John E. Dowling
- “Asking the Right Questions in Alzheimer’s Research,” Fitzpatrick’s Feature essay in the Fall 2018 Issues in Science and Technology
Jason Lloyd: Welcome to The Ongoing Transformation, a podcast from Issues in Science and Technology. Issues is a publication of the National Academies of Sciences, Engineering, and Medicine and Arizona State University. I’m Jason Lloyd, managing editor of Issues. On this episode, we’re in conversation with Susan Fitzpatrick, president of the James S. McDonald Foundation. Under Susan’s leadership, the foundation has supported research and scholarship in biological, behavioral, and complex system sciences. The philanthropy is particularly well-known for its work in funding emerging areas of research, such as cognitive neuroscience. Susan is a regular contributor to Issues. She’s written feature essays and book reviews on subjects, including Alzheimer’s disease, psychiatry and mental illness, and the evolution of consciousness. Susan, thank you for joining us on this episode of the ongoing transformation.
Susan Fitzpatrick: Hi Jay. Thank you. It’s a pleasure to be here.
Lloyd: One of the topics you’ve written about for Issues is research into Alzheimer’s disease. Could you tell us about the disease and how scientists’ understanding of it has evolved over time?
Fitzpatrick: I came of age in neuroscience when the—so I trained as a biochemist, biophysicist working in a neurology department in New York City. When I first started graduate school, the idea behind Alzheimer’s disease, it was that it had to be some more generalized sort of metabolic failure. Neurons in the brain are exquisitely sensitive to energy supplies, glucose, oxygen, and so somehow was it that the brain was no longer able to keep up with the energetic demands of computation, and so therefore began to evidence this pathology. It was difficult to really show that.
Then there was this idea that came along that it was actually acetylcholine, which is a neurotransmitter. It was a deficit in acetylcholine that was the problem. These poor individuals with Alzheimer’s disease were being fed massive amounts of lecithin, which is a way to deliver choline, which has a terrible, horrible smell.These poor souls were being loaded up with choline as a way to drive acetylcholine synthesis. This also turned out not to be helpful in terms of cognition.
And then, I think, there was a series of studies that really looked at molecular neuroscience. We were moving from this more metabolic physiological kind of neuroscience into this more molecular approach to neuroscience. The idea came that there had to be this target, right? If you wanted to have something druggable, you have to have a target. The discovery that these plaques and tangles contained this beta-amyloid protein and that there were individuals who got early-onset dementia that had an autosomal dominant gene that coded for this beta-amyloid protein, that really established the idea that the cause of Alzheimer’s disease was some problem with metabolizing or handling this protein. Beta-amyloid has a tendency to form aggregates, that these were toxic to neurons, and this was the cause of neuronal loss and hence cognitive degeneration in Alzheimer’s disease.
Lloyd: How did the scientific community respond to this new idea?
Fitzpatrick: Early on, there has always been resistance in the field to this idea. There was a competitive camp that was interested in another protein, which was called the tau protein. In fact, these two groups were nicknamed the Baptists, because they were interested in the Abeta protein, and the Taoists, who were interested in tau. You can find these funny stories about the Baptists and the Taoists arguing in different meetings and this stuff about who had the right protein. But in reality, their underlying conceptual approach was pretty similar. That there was this one thing that we should be targeting and it was this, whatever it was, whatever protein they were favoring, its aggregation in the brain.
There’s been two difficulties—actually, there’s a number of difficulties with this, but you can imagine that what happens when you get this kind of a theory, you start generating enormous amounts of tools to work with it. You develop transgenic animals, and you have ways of studying this and you have markers and you have all kinds of things that are focused on this idea. Then you’re developing interventions based around that idea.
One of the other criticisms has always been that the degree to which somebody has the hallmarks of Alzheimer’s disease is plaques and tangles in their brain, does not really perfectly correlate with their cognitive function. You can see people come to autopsy who have brains that don’t look so good who are doing just fine in their everyday life. It’s clear that yes, Abeta and Tau have something to do with damage to neurons and to brain tissue, but it’s not quite clear if it’s an effect rather than a cause. It’s secondary to something else. This has been further enhanced by all the recent clinical trials that have gone on over the last decade or so, where heroic attempts have been made to either limit the production of beta-amyloid in the brain, remove beta-amyloid in the brain. Some immune approaches that have been used. There was some studies for a while where people were draining CSF from people with Alzheimer’s disease as a way to try to lower their Abeta load.
Lloyd: That’s spinal fluid?
Fitzpatrick: Yep. And they don’t work. They don’t work in the terms of that they don’t stop what is most concerning to people with the disease: the degenerating cognitive impairment.
Lloyd: You can effectively get rid of these plaques and not have this build-up in the brain, but you don’t see an improvement in cognitive function.
Fitzpatrick: Right. And it’s probably, again, because if it’s a signal of damage that has occurred, the damage has occurred whether the beta-amyloid is in the brain or not at that point. Whatever caused these toxic proteins to build up and form these plaques and tangles in brain tissue, the process is probably still going on, and you haven’t solved what the actual problem is. Almost every iteration of this that’s tried really comes up with the same basic response: that you can clear the toxic protein, but you don’t seem to reverse or stop the degenerative decline.
Lloyd: We don’t have any treatments to reverse or stop cognitive decline, but the Food and Drug Administration recently approved a drug by a company called Biogen that’s called Aduhelm, which targets amyloid-beta. Could you tell us more about how this drug got approved and just why its approval was so controversial?
Fitzpatrick: Biogen went and had a clinical trial. They actually stopped the trial because they were not meeting their primary criteria of seeing any change in cognitive function. But there was this signal that it was decreasing the amount of beta-amyloid, based on marker scans of the brain. They went ahead and asked for FDA approval of this drug. When the data came before their 12-member panel, 11 of the members said no. They didn’t meet any of their primary endpoints. They’ve got this surrogate endpoint data, but the primary endpoint is not mattering. And it’s all based on this hypothesis, that if you believe that limiting beta-amyloid is going to help you then, OK—but actually that connection isn’t really made.
So 11 of the 12 members, 10 of the 11, I can’t remember the exact number, voted no, one person abstained, and the FDA approved the drug, even though it’s an immunotherapy, it’s extremely expensive, and it causes brain bleeds in patients. It’s not without serious side effects. This drug, which was supposed to be a blockbuster, and was the hope that everyone has been waiting for for generations, is an extremely expensive drug that seems to have no effect and can, in fact, be dangerous.
Several of the panel members that were reviewing this resigned over this. They were so appalled by the FDA’s decision—because giving this drug to people also means that you’re not going to be able to use those resources for something else.
Lloyd: I could imagine that that would have a pretty dramatic effect on, if you have this drug out there and you’re trying to help patients, but it’s effectiveness is limited, I imagine it’s quite frustrating to have patients wanting this to see if it would work and trying to focus on other things might be a bit of a challenge for both clinicians and researchers.
Fitzpatrick: It’s also been a real hit to the hypothesis. This has been a more visible failure than, I would say, some of the other drugs have been, because they just failed in clinical trial and then this is now out there in the public. Patients could be prescribed this, doctors know about it, families know about it. It’s something they could get access to because it’s been FDA approved. It’s not a clinical trial. But actually, there’s been very little usage of the drug.
Lloyd: Are there lessons to be learned from the Aduhelm controversy?
Fitzpatrick: It’s interesting on some level. It’s interesting from an Alzheimer’s perspective of how do you rethink a disease process when you have decades of a dominant hypothesis?
Because that’s how people are trained, again, that’s what the tools are, that’s where the investments have been made. It’s very difficult to back up from that and go into another direction. But it also is, I think, a cautionary tale for neuroscience in general, because this situation is not unique to Alzheimer’s disease. We don’t have really effective treatments for almost any neurodegenerative disease, and we don’t have very effective population-level treatments for people suffering from neuropsychiatric diseases.
Is there something fundamental in neuroscience that we’ve kind of missed about what’s really happening in these diseases? How do they really occur? How do they play out? What kind of intervention should they really be looking at? Is there a way that we should be less dogmatic about the way that we pursue our research so that we don’t allow a hypothesis to become so dominant early on in the process—but we allow for these alternative hypotheses to continue and to be supported, so that when something doesn’t pan out, you’ve got other avenues that you can go down? And they’re also robust and healthy and they’ve got tools and people and training and literature and all those things that you can build on.
I think it’s an interesting time for the field to begin to think back and say, OK, where have we been placing our bets for the last 40 years? Where did we not place our bets over the last 40 years that we should begin to think about placing our bets? Do we really have to start from scratch? Or are there ways that we can bring other theories, other tools, other ways of approaching a problem into the field so that we don’t have to throw everything out that we’ve learned, but can we put it into a different framework? Can we put it into a different structure? Can we add other components in a way that may get us closer to where we actually want to be?
That’s one of the things that certainly in the last couple of years, we at the foundation have been thinking about a lot and really trying to push some of those ideas forward. One of the areas that we’ve been looking at, and we have a small working group that’s been working on this problem, is thinking about what a new framework would look like in the context of schizophrenia.
Lloyd: To turn some of those really interesting questions that you’ve raised back to you, do you see initiatives, programs that really do take a different approach that seem promising? What we’ve been talking to up to this point, it seemed gloomy, the prospects for someone with neurodegenerative disease or someone who may have that in the future. One of the things that you talk about in an Issues piece is looking at the individual in context, looking at the brain in context, and looking at not just even their physical aspects of how the disease presents, but also their social context. I think you call this the network context for an individual. I was wondering if there’s anybody doing that kind of work that you see maybe as more promising than what’s traditionally happened over the past 40 years.
Fitzpatrick: Yes. I definitely take your message to heart. One of the things that I don’t like is when we just say, “This has all been wrong.” That’s not very useful, and it’s not very hopeful. In many ways I do think the soul searching that’s going on right now in the field is actually quite hopeful and encouraging because it’s really hard to change direction if you think you’re going in the right direction. I think in some ways, the fact that people are willing now to take a step back and say, “Wait a minute, let’s really think about what might be going on here and let’s try some other things,” is actually hopeful, because it means that we might be able to make progress. It may even be that some of the things that we’re trying in combination with higher-level interventions.
We certainly see this in certain neuropsychiatric disorders that a combination of pharmaceutical interventions with cognitive behavioral therapy with social support actually is better than doing those things alone. But what we don’t have right now is a conceptual framework for thinking about that. People have been trying this on the clinical level, and a lot of progress does actually come from clinicians sort of running these essential clinical experiments. Like, let’s do this and let’s do that and let’s see what happens.
But we can’t learn from that if we don’t have the theoretical framework that we can stick it into and say, “OK, this makes sense if you would give this particular agent that’s going to have this perturbative effect on the current way the brain is working, which now allows us to introduce new strategies, new ways of functioning—and we’re going to build some environmental supports in there. We’re going to help family members actually see how they can support this different way that this brain is working.” I think that’s where we haven’t really been active enough, is this idea that we’re also going to go in and fix whatever this problem is.
It’s unlikely that we’re going to be able to fix this problem. I think what we might be able to do is slow down processes, but I think you’re mainly going to have to say, “OK, how do we support a brain that just functions in a different way than neurotypically?” I think this is going to be true for neurodevelopmental disorders. So you have seen work going on in these areas. You see work in autism, you see work going on in depression, but it has not yet reached the point where we have a framework for thinking about it.
Let’s say, if you are someone who does suffer from serious depression, you may get put on a drug. It works, it doesn’t work, so they try another drug, but there’s no principled way for deciding what order we should be doing that. They may start with like, “Let’s start with the oldest drug,” or, “Let’s start with most popular drug”—but there’s nothing theoretical that’s sort of saying, “Well, we should start with something that affects the serotonin system, and then we should move to the dopamine system, and then at that point, we should begin to introduce cognitive behavioral therapy—or should we start with cognitive behavioral therapy and then introduce drugs?” We don’t really have that framework. I think it’s the kind of rethinking that’s going to be important, and again, I don’t want it to overnight become the dominant hypothesis. I don’t want any of these things to become dominant. I want us to be kicking them around more and really paying attention to what’s happening clinically.
Lloyd: Could you tell us a little bit more about what the research community could learn from the clinicians?
Fitzpatrick: We’re acquiring an enormous amount of information by people who are on the front line of trying to treat individuals who are suffering. I feel sometimes that the research end of the spectrum doesn’t often pay as much attention to what we’re seeing clinically. I think anything new that people are proposing, I want to see it constrained by what we actually know about patients. What do we actually see? Before you go running off with your one finding, your one gene, or your one molecule, your one whatever, and develop an animal model over it, and then you’re going to study that thing, this one thing—there has to be a bigger rationale.
In the many ways, I feel actually more optimistic than I’ve ever felt because I do feel that there is this change.I think the younger neurologists and psychiatrists who are coming into the field are more open to looking at some of these different approaches and are not so wedded to these molecular approaches. The difference is that—and this is going to sound really strange—but the other weird thing is that we don’t have the business model for this. We have a business model for what happens when you identify a druggable target and you develop a drug. We have a distribution model, we have a production model, we have all those kinds of things. We don’t have a model for thinking about how do you actually manage a complex neurodegenerative or neuropsychiatric disease in the community? We keep hearing about community health. We really don’t have a system. We don’t even have the model for what that system would really have to look like.
I was at a small meeting one time where a psychiatrist who does a lot of clinical trials was there. We were talking about the high placebo rate that there is in most neurological disorder clinical trials, and it’s quite high. Even deep sham deep brain stimulation has a pretty high placebo approach. There’s no current being turned on and off, and people get better. We were talking about this and he said, and this has so struck me and stuck with me. He said, “Maybe if we treated everyone the way we treat patients who are enrolled in clinical trials, a lot more people would be getting better.” What he meant was, when you’re enrolled in a clinical trial, you’re getting a lot of support. You’re getting your medical needs taken care of. You’re being seen by a lot of eyes, and you’re getting positive reinforcement, and you feel like you’re taking some agency over what’s happening to you, and your family thinks that you’re trying, and the whole ecosystem changes.
I said, “Well, why don’t we?” He said, “Well, we can’t afford it.” I was like, “I mean, we can’t afford not to do it to some extent.”
In the meantime, we’re willing to approve a drug for $50,000 that has minimal to no effect—possibly even negative effects. That’s what I mean. What if we were putting that into saying, “Okay, what does it mean to really support somebody who is suffering?”
So I’m excited about the fact that these kinds of conversations are happening and that there is work going on that is beginning to look at this. There are people who are looking at higher-level interventions. There are people thinking about the brain from dynamical systems perspective, that there are people who are looking at interactions with the environment. There are people who are beginning to think about, what are the different drug interactions that occur over time?
I think one of the weird things about biomedicine, in general, is that we’re always treating the disease you had. We totally ignore the dynamic, developmental nature of—I mean, we call these diseases “degenerative” or “neurodevelopmental” for a reason. They’re changing over time. We’re often looking at the disease that you had and not the disease that you have because every treatment is introducing some kind of change. We really don’t always have a good handle on that.
Lloyd: Your response to that last question raises some huge questions about how we, as a society, as a scientific enterprise, conduct research and think about some of these problems and approaches to solving them. This really shines through, I think, in your book reviews, where you have a real appreciation of authors who can navigate the complexity of the issue that they’re talking about, who explain the nuances, who don’t take this reductionist approach. I think one of the through-lines, in the several book reviews that you’ve written for Issues, is that you really appreciate and promote this idea that things are far more complicated than we understand. The complexity is often irreducible when you’re talking about consciousness (that was the most recent subject of a book review). But embracing that complexity goes against the reductionist tendency of scientific research as it’s currently practiced, scientific funding, the training of scientists.
Where the focus is, you find the causal mechanism at the lowest level, at the most basic level. The molecule, the protein that’s causing this, and then you develop a drug or pharmaceutical to treat it, or maybe a more complex device or something like that, but that’s what you focus on. That’s where all the money goes. That’s what you’re trained to do. And that’s how the thinking of the research enterprise is structured. I’m wondering, this is sort of a big question, but if there were a way to intervene in some aspect of that whole enterprise to better embrace the complexities that you think are really, really important in addressing and confronting these diseases, what would that be?
Fitzpatrick: It’s very difficult. One of our colleagues at ASU, Jason Robert, has this great phrase that I absolutely love where he talks about a “hedgeless hedging.” He’s a philosopher of science and has done a lot in neuroscience. It’s not that anyone in neuroscience doesn’t know that the brain is complex. They all do. In fact, what you’re seeing now sometimes is a lot of, “The brain is complex. It’s an ecosystem. It’s organized, blah, blah, blah, dynamical, adaptive, changing over time—now let me tell you about my gene.” So it’s like by acknowledging the complexity, you can brush it aside to some extent. I love this idea that it’s like a hedgeless hedge. You’re not really hedging your bets. You’ve got no hedge, really. This is what you do. And that’s because this is what you know how to do.
I was sitting next to a young neurologist one time at a talk that was really trying to embrace the complexity of some of the problems that we’re trying to solve. He looked at me and he said, “If you took this seriously, you’d have to stop what you were doing.” I said, “Yeah.” And like, I can’t! I’ve got to get my grant out. I’ve got to get my paper published. I’ve got to get my tenure. I’ve got to get promoted. This is the way you do it. Even though it was hitting him that there was something really wrong with what we’re doing, what’s the way out of it?
I think we really do have to take this seriously and change the reward structure: that you don’t get rewarded for tiny bits of incremental work on highly artificial systems that don’t reflect anything in nature. One of the things I love about some of the books that I’ve been able to review for Issues is that the writers are also very grounded in the natural world. They’re either looking at organisms that are living in the natural world, they’re looking at evolution across time, but they’re using real artifacts. They’re not trying to create this in the laboratory. I think we have to find a way of getting biology back into nature—and then getting biomedical research back into biology.
And then, I think, a lot of this falls into place because now you cannot do real organism biology without thinking of whatever it is that you’re interested in the context of the organism, in the context of its life, in the context of its environment, and the context of the problems it’s trying to solve as it goes through both its own developmental ontogeny and across evolutionary time scales. You can see, again, that there are attempts to do this. I know NSF has these new biological integration institutes that they’re trying to work on. I think it’s the right idea. I think the idea is, can you be patient enough to realize that this kind of cultural change doesn’t happen overnight?
Lloyd: Even if those changes don’t happen overnight, they do sound really promising. I think there’s a lot to learn from our progress and missteps with how the scientific and policy communities have handled Alzheimer’s research. Thank you, Susan, for taking the time to talk to us.
Fitzpatrick: Oh, thank you. This is actually a delight for me and a pleasure.
Lloyd: And thank you for joining us for this episode of The Ongoing Transformation. For more of Susan’s work, and I really encourage you to check it out, please visit the show notes for this episode, where we’ll link to her recent essays in Issues. Please subscribe to The Ongoing Transformation wherever you get your podcasts. Email us at [email protected] with any comments or suggestions. If you enjoy conversations like this one, visit us at issues.org and subscribe to the magazine. I’m Jason Lloyd, managing editor of Issues in Science and Technology. Thanks for listening.