Brave New Visions
The promises and perils of imaging the brain
In separate interviews with Lens editor Bill Snyder, Richard Hargreaves, Ph.D., vice president of Imaging at Merck Research Laboratories, and Judy Illes, Ph.D., Canada Research Chair in Neuroethics at the University of British Columbia, describe recent progress in the use of imaging technologies, their potential and challenges to further development of the field. Can imaging lower the cost of developing new drugs? Are there places we shouldn't go?
Hargreaves earned his Ph.D. in physiology from King’s College London. At Merck since 1988, he led the biology core for the discovery of Maxalt, an anti-migraine medication, and Emend, which helps prevent chemotherapy-induced nausea and vomiting. Illes received her doctorate in Hearing and Speech Sciences from Stanford University, specializing in experimental neuropsychology. She co-founded the Neuroethics Society, and formerly directed the Program in Neuroethics at the Stanford Center for Biomedical Ethics.
Interview 1: Richard Hargreaves, Ph.D.
Radiotracer imaging such as positron emission tomography (PET) allows scientists to see whether a molecule “engages” its target sufficiently to test for therapeutic efficacy. If the molecule doesn’t bind to its target, or saturates the target and yet fails the efficacy test, “then we can move on to a new therapeutic concept,” Hargreaves says.
In the early stages of drug discovery and development, imaging “actually increases the attrition in the number of molecules you test,” he says. “You actually throw more molecules away to get the best one.”
Can imaging reduce the cost of developing drugs?
Built into the cost of any drug are all the failures of the molecules you test along the way. If earlier on you can select the best development candidates, you improve your chances of moving through the development process in a more informed and efficient way.
Remember, too, that making early “go/no-go” decisions also has ethical implications for the human subjects who take part in clinical trials, since fewer will be exposed to potentially ineffective therapies.
If you get it right, you’ll bring good medicines to the market faster.
Is imaging improving treatment?
In oncology, PET radiotracers can be used to assess different aspects of tumor physiology, such as glucose metabolism, cell proliferation, angiogenesis (growth of new blood vessels), apoptosis (cell death) and oxygenation.
PET tracers exist for some of these and are well embedded in clinical care. An example is 18F-fluorodeoxyglucose (FDG), which provides a measure of glucose metabolism. Others, such as 18F-fluorodeoxy-L-thymidine (FLT), which may reflect cell proliferation, are still being validated.
The beauty of being able to track aspects of tumor growth and viability means that you can see very early on after administering an experimental therapy whether it is impacting tumor physiology in any way at all. If it is, then you have a very good reason to carry on and look for tumor regression.
In our oncology projects, we’re also looking for molecular imaging agents that can tell you whether people are likely to respond to targeted therapies and how well our drugs engage them.
I think that imaging truly has the opportunity to revolutionize and personalize care in oncology. We’re going to see the fruits of that ripen in the near future.
How has imaging improved our understanding of the brain?
Functional imaging has fundamentally changed the way we can question brain systems because we can see them in action. It has given whole new dimensions to many areas of neuroscience such as pain, basic sensory systems, psychiatric disorders and their co-morbidity with other disorders of the central nervous system, cognition, language and neural mechanisms that underlie developmental plasticity and recovery of function.
Imaging studies that help us understand neurodegenerative brain disease can also help pave the way to better health care in the future.
At Merck, we have participated in the Alzheimer’s Disease Neuroimaging Initiative (ADNI), a collaborative consortium between industry, academia and the National Institute on Aging. The goal is to validate imaging tools such as PET and magnetic resonance imaging (MRI) and fluid biomarkers—alongside neuropsychiatric scores—for tracking the progression of Alzheimer’s disease in people currently on the best therapies available for this condition.
This is an important study because it will tell us what an aging Alzheimer’s disease population looks like in North America today. Once we have an understanding of that, then we have a reference library or baseline that we can use to evaluate whether our new medicines produce true improvements in clinical care by modifying disease progression.
It would be brilliant if you could prevent Alzheimer’s disease totally, but just think of a therapy that delays its onset by 10 years. That too would be a phenomenal achievement.
Can these collaborations overcome concerns about disclosure and conflicts of interest?
It’s in the interest of the patient, the National Institutes of Health, the FDA, healthcare providers, and all of us in the pharmaceutical industry to move safe and effective new medicines forward efficiently.
We will attempt internally to make imaging agents that validate target engagement for our new therapies and have the desired mechanistic effects. That’s proprietary to drug discovery programs. Eventually, when appropriate, we will try to make these agents available to all for research.
When we move into evaluating disease and finding imaging endpoints that chart progression, the story is somewhat different. Here the imaging endpoints are valuable to everybody trying to treat that disease—independent of the mechanism they are using to do so.
In this case there is a very good case for a consortia approach, which combines the efforts of interested pharmaceutical companies with academia and perhaps also diagnostic companies. The goal: defining new surrogates that speed drug assessment and approval, particularly in areas where medical need is poorly met.
I think the ADNI has the potential to be a wonderful example of this approach. We’re going to set the “goal posts” for imaging biomarkers and the evaluation of therapies in Alzheimer’s disease.
Will imaging increase the emphasis on using drugs to treat disease, or will it help validate behavioral/cognitive approaches as well?
I think the two go together.
The industry spends its time making highly selective pharmacological agents. But if you combine those together with the amazing imaging of brain function, then you can understand the functional/chemical neuroanatomy of health and disease as well.
Functional brain imaging should help you design better therapies—be they pharmacological or non-pharmacological. Indeed, it’s already begun to help us understand what the placebo component of responses really is, and to segregate responders from non-responders in the field of pain research.
If brain scans become a practical, reproducible method of predicting risk for disease, what kind of ethical issues will it raise?
Clinically, brain structural imaging is used to assess disease and to make treatment decisions based on prediction of long-term sequelae. I think functional brain imaging has a way to go before it reaches this point.
One concern that has been raised about functional brain imaging in the past is that we may find things that are unexpected, or image things that may be predictive of events or behaviors for which we have no solution or control. This is clearly an ethical issue, and we need to think how to respond in advance.
Do you want to bury your head in the sand and not know, or do you want to develop a strategy or set of rules that can help you deal ethically with the issues that could be raised as you strive, through imaging, to understand brain disease and provide a platform for discovery of better therapies?
The bottom line for me is that many new prognostic and diagnostic medical technologies raise these types of issues. You need to think about them and plan how to respond in advance, or reconsider what you are doing. It’s not just science for science’s sake.
Is there concern that the field is over-hyped?
It’s very visual and so it’s very powerful. It tends to get jazzed up. Scientists need to be central in communicating what the functional MRI technique is.
What are its limitations? What are we actually measuring, and how certain can we be about what we see? How has it been validated? If we do the same thing twice, do we get the same answer?
Does the field need standards? Yes, undoubtedly. Let’s be sure to use this tool to ask the unique questions that can actually be answered with it, rather than as a more complicated and expensive route to answers that could be obtained more simply and reliably another way. Otherwise, it can be neo-phrenology, can’t it?
At the end of the day, we need to use functional brain imaging carefully, and manage our excitement to make sure that we give clear context and boundaries to what we do and what we see.
Interview 2: Judy Illes, Ph.D.
What are some of the ethical concerns about imaging?
What if we could use those data to predict behavior in the future? What if we could in an adolescent predict propensity to aggression or sociopathy or suicide? How are we going to handle people in whom we are able to predict potentially devastating behaviors?
What will it mean to be able to predict the onset of a disease may occur 30 years down the line, especially when there’s no treatment? Alzheimer’s a perfect case of that. The issues of prediction are immense.
Some hardcore MR (magnetic resonance) physicists who developed this technology would say, ‘Nah! Nah! Never!’ But I don’t know. I think that the evidence is that we haven’t been stopped yet in our innovations. It’s just a matter of time.
Are there places we shouldn’t go?
Is everything allowable as long as it’s done ethically? Should there be boundaries imposed on our science because of their new potential real-world applications?
I will argue that limitations on ethically conducted science are not appropriate. It’s just part of the human condition to be curious and innovate and push the envelope. But now we have every good reason to couple our ethical thinking with our neuroscience.
I think part of the ethical construction of research is not only good protocols and protection of human subjects, but actually thinking about the downstream implications of the research.
What if you’re doing a study of people with schizophrenia and you find a mass? How do you handle that with that kind of population? What about children? What if something comes out and they end up with a result that may be stigmatizing? That could have a lifelong implication.
What if somebody develops a drug that can check addictive behaviors? What kind of interventions are developed and become available? All the more reason to start thinking about the ethics of it. If in fact we can use an imaging technology to predict addiction, then we definitely want to have a response ready.
And to the extent that we continue to do these studies that probe personhood, it’s not enough just to say “I’m not going to hurt somebody in my experiments.” But it is becoming, I believe, a requirement to think about, “If I find out that there is a locus or loci or network for making race judgments in the human brain, this is how I’m going to handle the information in terms of its dissemination.”
Without being too alarmist and certainly without being negative, I do feel that unless we start to introduce a reasonable—not a heavy, but a reasonable—ethical component to the kind of work that we’re doing, there are risks of adverse effects on people that then have the adverse effect of potentially slowing down the progress of research.
And so by being proactive, and by trying to address these issues jointly from within the neuroscience and bioethics community, we can get a very good handle on what the issues are and what are the ways that we can empower our science ethically so that we can either prevent the adverse events down the road or at least be able to manage them very efficiently.
What are the limitations of these technologies?
They start at the beginning with the design of an experiment. Any experiment, especially one that probes complex phenomena such as existential decisions, takes some pretty clever design and invariably will reflect—and I don’t see how it could not—the cultural orientation and biases of the experimenters developing it.
The way I value something might be very different from the way you value something at Vanderbilt. When we start to probe personhood, we’re unequivocally invoking issues of values and culture, ethnicity, so there’s a limitation right there.
Another limitation is, as we know, in the statistical processing of the data. The different kind of statistics you do may affect the results, and along with that we don’t have a very good handle yet on individual neuro-functionality in terms of blood flow, for example. We’re still using group averages, although people are really making some fantastic strides in that domain.
Some investigators are concerned that we’re only able to study subsets of certain disease populations. I think of autism, for example, where we have quite a few studies on high functioning autistic children but considerably fewer on those who are low functioning and more difficult to manage, certainly in the context of an MRI environment.
One that I’d add is in the day-to-day variability of physiology that goes into brain measurements. We know that brain physiology changes are different by gender, by stress level, by food level. Again, understanding individual variability is really crucial.
There are definitely limitations to using the knowledge we gain in the laboratory, especially about these incredibly complex phenomena in the real-world setting. But nonetheless, we are measuring them in the laboratory. These laboratory studies are being covered by the press and the word is out in the public domain.
And so we as neuroethicists really have to work closely with our neuroscience colleagues in the trenches to be proactive about aligning the ethical considerations of some of these studies up front in the research design and really trying to anticipate what kind of social impact or legal impact or ethical impact it might have down the road.
We’re accustomed to, in many ways, living in a very privileged environment where we publish in science journals and we speak with our colleagues at meetings, and a little bit filters out to the public. But I really believe and our data suggest that we’re in a different place now and we have to consider these ethical, legal, social issues to a far greater extent than we ever did before.