Research Profile - Medical Imaging, What's Next? - A Panel Discussion
[ Back to main article ]The Panelists:
Dr. Robert Beanlands
Dr. Beanlands, an international leader in cardiovascular nuclear imaging, is Director of the National Cardiac Positron Emission Tomography Centre. He is Chief of Cardiac Imaging and Director of the Molecular Function and Imaging Program at the University of Ottawa Heart Institute and a Professor in the Divisions of Cardiology and Radiology at the University of Ottawa's Department of Medicine.
Dr. François Bénard
Dr. Bénard is the British Columbia Leadership Chair in Functional Cancer Imaging and a Professor of Radiology, based at the University of British Columbia and the BC Cancer Agency in Vancouver. His research focuses on advanced cancer imaging techniques for a range of research applications.
Dr. Ravi Menon
Dr. Menon is a Tier 1 Canada Research Chair in Functional and Molecular Imaging at the Robarts Research Institute in London, Ontario and Director of the Centre for Functional and Metabolic Mapping. He is a Professor of Medical Biophysics, Diagnostic Radiology and Nuclear Medicine, Physics and Psychiatry at the University of Western Ontario.
Dr. John Valliant
Dr. Valliant is Scientific Director and CEO of the Centre for Probe Development and Commercialization and an Associate Professor in the Departments of Chemistry and Medical Physics and Applied Radiation Sciences at McMaster University. His work focuses on the chemistry of imaging probes derived from medical isotopes while the Centre focuses on the production, translation and distribution of imaging probes.
Dr. Robert Beanlands
Dr. François Bénard
Dr. Ravi Menon
Dr. John Valliant
CIHR: To begin, what role does medical imaging play in your everyday work? And how do you see it changing in the next decade?
Bénard: Imaging is at the core of my research program and my clinical activities. What we do see is a greater need for more specific imaging tools that will address very clear clinical questions. For example: predicting or measuring treatment response. That's a shift away from traditional, purely anatomical imaging. Clinicians need to measure treatment response very clearly and quickly. So, it's a growing field and we're seeing a growing clinical need every day. As to where I see the role changing, I think there will be improved awareness from the public – they will want the greatest imaging tools available. So it will be a tug of war between our ability to deliver top-notch care and the ability of governments to finance the imaging tools. At the end of the day we'll save by reducing the use of futile therapeutic approaches but it will cost money up front to invest in new technologies.
Menon: I primarily use imaging as a research tool, although I have many clinical research collaborators who use our equipment. Where I see this going is some degree of convergence of technologies, but I see some roadblocks related to differences in the modalities. There are positron emission tomography (PET) people and there are magnetic resonance imaging (MRI) people and ultrasound people and so on. For example, computed tomography (CT) perfusion (injecting contrast agent into the bloodstream to reach an organ or tissue), can do many things that we currently use nuclear medicine for. But, because it involves different practitioners, at this point you don't see the transfer of this kind of technology into clinical care. There are potentially a lot of modalities – all of which are good for certain things and what they're good at is changing – but because of the way many diagnostic imaging departments are set up, the actual transfer of the technology to the patients does not always happen.
What's a PET? What's CT? What's MRI? – see our glossary of terms.
Valliant: I'm a chemist working on developing the next generation of medical isotope-based imaging probes. Our group works on all facets of the field, from basic science to production for clinical use and clinical research. If you look to the future of medical isotopes and the associated imaging probes, I believe you're going to see more specific tools that help physicians stratify patients and select therapies. You're going to see more imaging tools being used to guide biopsies, to help with pathology assessment and surgical guidance. So you're going to find that the imaging and probe development fields will not be standalone entities used solely for early diagnosis; they're going to be linked to pathologists and surgeons who routinely will use imaging technologies. An example is groups that are working towards one-stop breast cancer screening programs that include both imaging, biopsy and tissue analysis, which is going to require coordination between radiologists, nuclear medicine specialists and the people who do the pathology on the tissue samples. It will also require the creation of new probes. Right now, particularly in the world of Technetium-99m (the main isotope produced at Chalk River for medical imaging), there are a handful of agents that are used a lot. As we diversify and start producing more specialty agents, the question is: how do you produce them cost-effectively for a small number of patients? Because of the way that we manufacture agents today, the cost to make one dose is really not that much greater than making 10 doses. We have to find newer production paradigms to meet the future needs of the field in a cost-effective manner.
It will be a tug of war between our ability to deliver top-notch care and the ability of governments to finance the imaging tools. At the end of the day we'll save by reducing the use of futile therapeutic approaches but it will cost money up front to invest in new technologies. Dr. François Bénard
Beanlands: In terms of research, what's really impressed me is the wealth of potential in terms of how we can use imaging. It really has changed how we understand disease; we now understand different aspects of anatomy and physiology that we really had no idea about 20 years ago. So I think it has revolutionized human physiology research to an impressive degree. I was trained in nuclear cardiology and PET. What's clear now in cardiovascular care is that if you want to do imaging, you can't train in one type of imaging – you basically need to become an expert in more than one modality. The clinical community needs people who are expert in multiple modalities. Not in the next 10 years but somewhere beyond that, there may well be an independent speciality that's not cardiology and that's not imaging, but is cardiovascular imaging. You might see that in other fields as well, a merging of specialties around a disease process or a group of disease processes with multiple modality imaging to evaluate it.
The clinical community needs people who are expert in multiple modalities. Not in the next 10 years but somewhere beyond that, there may well be an independent speciality that's not cardiology and that's not imaging, but is cardiovascular imaging. You might see that in other fields as well, a merging of specialties around a disease process or a group of disease processes with multiple modality imaging to evaluate it. Dr. Robert Beanlands
CIHR: Are there workable solutions to replace or complement what Chalk River was supplying?
Bénard: There are multiple potential solutions. One of them is we can diversify the sources of Technetium, which remains an extremely important isotope. It provides important diagnostic information for hundreds of thousands of patients per week. It's nominally cheap so we can't move away from needing this isotope. We received a grant that was funded by the Natural Sciences and Engineering Research Council (NSERC) and the Canadian Institutes of Health Research (CIHR) to explore the production of Technetium using cyclotrons (circular particle accelerators). If we can make large quantities of Technetium using cyclotrons instead of nuclear reactors, then that provides multiple backup sites for production and diversifies the source of Technetium. It's been known since 1971 that this could be done but it wasn't competitive in the marketplace because the nuclear reactors were subsidized. Also, there are new reactor technologies being explored, so I think the marketplace will correct the deficit within a period of three to four years and we will have a much healthier supply chain for Technetium.
Want to learn more about some of the key milestones in imaging? Check out our imaging timeline.
Valliant: We have to attack this problem from multiple sides, so the cyclotron production is an excellent strategy for complementing the world reactor base supply. I think the one concern we have is that a lot of remote sites use Technetium. We're going to have to be able to produce enough to make sure we have no regional disparity in care.
Technetium's not going anywhere, it's part of the scenario. But there are situations – cardiac perfusion, for example – where you can use a CT scanner and you can use the iodinated contrast. This is a commercial product available on thousands of CT scanners. I don't know if anybody's actually done a clinical trial to test that against Technetium – but those are the kinds of things that need to be done because then you could potentially take the cardiac stuff out of the loop and worry about other things. Dr. Ravi Menon
Beanlands: Another approach is to not rely on Technetium at all and use PET. One of my colleagues is leading a project that will see dissemination of PET technology using Rubidium as a perfusion agent, instead of Technetium-based perfusion agents to look at cardiac imaging. That has potential, if there's dissemination of technology, to significantly impact Technetium use.
Menon: Technetium's not going anywhere, it's part of the scenario. But there are situations – cardiac perfusion, for example – where you can use a CT scanner and you can use the iodinated contrast. This is a commercial product available on thousands of CT scanners. I don't know if anybody's actually done a clinical trial to test that against Technetium – but those are the kinds of things that need to be done because then you could potentially take the cardiac stuff out of the loop and worry about other things. This is the sort of thinking that needs to be encouraged – not just more agents or different agents. Those are important too, but they're not the only solutions.
Bénard: Usually what we see is that the physician acceptance has been overwhelming and quick when something is clearly superior and more cost effective. This happened, for example, with the widespread adoption of CT as soon as it came out for brain tumour imaging and for MR over CT for brain tumour. So the market will shift quickly.
Menon: MR was used for brain tumours long before there was any evidence that MR was actually better for brain tumours. Radiologists were saying, “I know it's better, I can see things better,” but the actual evidence didn't come out for seven years.
So MR was in use for brain scans long before clinical trials proved that it was worthwhile?
Bénard: It was so obvious that it was better that people were using it without waiting for a clinical trial. That happens – it's happened quite a few times in medicine's history.
So the actual clinical practice will sometimes race ahead of the research?
Menon: More than sometimes. Some things are so obvious you might say evidence is not always there. And evidence needs to include cost-effectiveness. MR is not always necessarily the most cost-effective; CTs can be much cheaper.
CIHR: What about biomarkers? Researchers are identifying new biomarkers every day that can indicate the presence of disease or a person's suitability for a particular therapy – such as whether a breast cancer patient is likely to respond well to Herceptin. But we don't seem to have the capacity to put these biomarkers to the test for disease diagnosis and more personalized treatments. Are innovation and application out of sync?
Biomarker: A biomarker is simply a sign of the presence or progress of a disease or condition, such as a raised temperature signalling a fever. It can also be a substance introduced into an organ to see how it functions – for example, using rubidium chloride as an isotope to assess blood flow to the heart (myocardial perfusion). In health research, biomarkers are specific biochemicals used to measure the status of a disease or the effects of a treatment.
Valliant: The genesis for the creation of our Centre for Probe Development and Commercialization was the fact that people at universities and hospital research centres are developing imaging agents as surrogate biomarkers and as alternatives to Technetium-based agents. The major challenge was to get them translated and in the hands of physicians so that clinical trials can be done to demonstrate that they can change the course of treatment and improve outcomes. That's the part that's been missing. The mechanism to evaluate biomarkers has not really met up with the pace of biomarker discovery.
Bénard: The delay from innovation to actual protocols is linked to the capacity to quickly do large-scale trials to evaluate these tools and biomarkers. There is a high cost. And regulatory bodies like Health Canada have no idea what to do with the explosion of new tools and biomarkers and how to regulate them or to accept them in clinical trials. Clinical trials is one of the areas where funding is lacking. Most of the money that is available from private companies is for therapeutic trials. That's a different mindset than trying to develop biomarkers that will help save money to select the most appropriate treatment.
Menon: It's simply not sexy to do a clinical trial to prove something is either cost-efficient or is more efficacious. That funding simply doesn't exist. If pharma won't do it, then nobody will do it.
CIHR: It's been pointed out that the limited ability to coordinate and conduct multi-centre clinical trials is restricting adoption of new imaging technologies. How big a challenge is it and could a network – a clinical trials network – address the problem?
Beanlands: There's no question that there's a need for better ways to evaluate imaging and to translate imaging developments, new developments, new biomarkers from the investigation stage to the application stage. A trials network is one of the pieces that would help to pull that together. I agree with the fact that doing that kind of research is often not sexy or popular, but organizations like CIHR have now identified that this is a direction to pursue.
Just to be clear, we're talking about CIHR's two-year, $10-million commitment to a clinical trials network initiative?
Menon: Yes, but I don't think that in two years – you might be able to start a trial or two, as a demonstration project, but for the most part that's as far as you're going to get in two years.
Beanlands: To me it's almost a bridge to establish some kind of network that would enable folks to raise the level of what's being done in the country.
Bénard: It's a great first start. The challenge will be trying to find a way to create a network that will be sustainable over the long run given the short-term nature of the current funding.
Menon: Right, because you're going to have to set up the infrastructure and data farms and quality assurance, and all these things and that's non-trivial. This is why we have so few clinical trials networks, period, in any field.
Valliant: One comment about the network idea, which is something that we've been pushing pretty hard for: we want to be cautious that we don't get caught up like clinical trials networks in other countries – where they become cumbersome entities. You'd like to have multiple centres working together and not have to spend a lot of the clinical trials money associated with management of the network. You want to make sure the resources have lots of impact.
CIHR: Part of the challenge is that medical imaging spans such a spectrum of scientific disciplines – from biology to mechanical engineering. You have experts in cancer working with experts in photonics. Is it possible to construct a network that covers all that, that you can have everything under that umbrella?
Menon: The discipline is different from the infrastructure, and the infrastructure, a good clinical trial's infrastructure, will allow you to answer questions in all these different disciplines. So I think you'd have to build that and then ask interesting questions in each of those areas. At that point you have to bring the right clinicians and scientists on board.
Bénard: You need to ask specific labs if they can be virtual core labs, but include the imaging expert, the physicist and so on to standardize the procedures. The basic infrastructure of the clinical trial is common regardless of the modality. The specificity will come from the trial or specific modality or groups of modalities. The American College of Radiology Imaging Network model has been pretty good for that.
Beanlands: I would agree that you need, potentially, a network that has principles and methodologies that can be followed that would be common to all sub-specialties. And then, basically, the various centres could provide the links to the technical experts required to develop and implement standardization and so on. So it's definitely possible.
CIHR: When it comes to imaging, the whole medical paradigm has shifted from the detection of disease to the detection of the possibility of disease. This is a huge change in how health care can be practised. Can this kind of transformational change be smoothly integrated into health care?
Menon: Completely as a cost-benefit thing, you don't treat probabilities. You don't start treatment for Alzheimer's for example, until the disease manifests. Unless someone shows otherwise – that it's cheaper to treat everybody with a genetic marker than a few people after they get the disease – you're not going to change the paradigm.
Bénard: There are examples in cardiovascular disease, for example, where there's widespread treatment of people. But there we have inexpensive biomarkers – cholesterol levels and risk factors. It's not the case yet in Alzheimer's disease. It will not work with very expensive diagnostic tools. We have no treatment, but if we did then everybody would want some type of imaging scan to detect Alzheimer's early. People would be requesting these expensive tests and this would crash the health care system. So detecting the possibility of disease will work as long as the tool that we use to identify who needs preventive measures or treatments has sufficiently low cost.
Beanlands: Something that's low-cost – that's a fundamental aspect for screening. Most of the advanced imaging techniques would not fall into this category. A genetic test or a simple biomarker, for example, could be much more widely applied. Then you're looking at preventive strategies which may or may not involve drugs. It could be non-pharmacologic; it could be diet-related and so on. So if there is this paradigm shift in terms of imaging, I see it as secondary. If someone was screened as positive, you might start the prevention process and use imaging to confirm diagnosis or follow the disease. But you wouldn't be providing it over a broad population base.
You don't start treatment for Alzheimer's for example, until the disease manifests. Unless someone shows otherwise – that it's cheaper to treat everybody with a genetic marker than a few people after they get the disease – you're not going to change the paradigm. Dr. Ravi Menon
Bénard: There are private clinics where people pay to get their PET and CT scans for cancer screening. But you need to do thousands of scans with many false positives to detect individual cancers. The same with breast MRIs: if you used MRIs on a widespread basis to screen for breast cancer you'd end up overwhelming all of the biopsy facilities with false positives. So these tools are best used in the proper clinical context – which is a very high-risk population either after some initial screening tool has been used or because of the genetic markers, family history or other. Those are much more appropriate uses of expensive technology.
CIHR: What do you see as the cutting-edge technologies that are likely to have the biggest impact on health care in the future?
Beanlands: If you're considering PET and MR as alternatives to Technetium-based imaging, then those could have a very significant impact in the near future. In terms of long-term impact, if you have genetic testing for diseases that you have therapies for, that could have far-reaching impact. But I don't know that we have those yet, and I don't know that any of them are actually emerging except for a few small, uncommon conditions.
Menon: That's probably the future. If you could just do a blood test, determine somebody has a disease – you don't care where it is in the body. You inject a retrovirus and – presto – they're cured. And there is no imaging in that whole process unless you want to look at response to therapy. But these kinds of things are a long, long way off; most of the therapies don't exist and the sensitivities in blood tests for a lot of these things don't exist either.
Valliant: From my perspective, it's going to be tools. Whether it's the agent or the scanner that guides the selection of therapy, most of the targeted therapies we're seeing at the moment are incredibly expensive. We need to have a better way of selecting the specific therapy for patients. The technology that improves that process is going to have an impact both in terms of outcomes for patients and the economics of health care.
If you're considering PET and MR as alternatives to Technetium-based imaging, then those could have a very significant impact in the near future. Dr. Rob Beanlands
Bénard: I see much greater dissemination and use of imaging tools. But that will have to come at a drastic price reduction. I think manufacturers have been trying to keep high prices and fancy equipment; they've added hybrid devices that are very powerful. But as we've seen with CT scans – where prices have fallen and where they will keep on falling – we will see the same thing with MR and PET scanners that are much more sensitive.
Whether it's the agent or the scanner that guides the selection of therapy, most of the targeted therapies we're seeing at the moment are incredibly expensive. We need to have a better way of selecting the specific therapy for patients. The technology that improves that process is going to have an impact both in terms of outcomes for patients and the economics of health care. Dr. John Valliant
Valliant: What about the return to dedicated scanners, so that cardiac or breast scanning doesn't necessarily have to be done in a hospital's nuclear medicine department? It could be in radiology or a breast screening clinic or in the cardiac clinic. Do you think that's going to make a change?
Beanlands: I can speak to the cardiac side and I think that's going to have to happen.
Bénard: I think we'll see a return of dedicated PET scanners, for example, for doing general bone imaging and cardiac imaging – or even dedicated brain units. I think that will come back at some point, but not in the very short term. I think there are plenty of ways that the cost can get reduced for more dedicated equipment to be installed in less centralized locations.