Molecular Imaging and the "New Radiology" Paradigm
Editor's Note:
Molecular imaging is poised to revolutionize medicine, particularly in the arena of targeted therapies and the follow-up assessment for recurrence of disease in patients with cancer.
Sanjiv Sam Gambhir, MD, PhD, is the Director of the Molecular Imaging Program at Stanford (MIPS). The primary objectives of the MIPS program are to foster innovations in the early diagnosis of disease through the use of imaging modalities such as positron emission tomographic (PET) scanning and PET/computed tomographic (CT) fusion studies and to monitor the impact of emergent molecular therapies on clinical outcomes.
Dr. Gambhir sat down with Pippa Wysong of Medscape to detail the recent advances in molecular imaging that have the greatest relevancy to oncologists and other referring physicians.
Medscape: Can you tell Medscape readers why molecular radiology is such an exciting area?
Dr. Gambhir: There is a fundamental shift in what imaging can tell you when you bring in molecular details. I like the following analogy: Imagine you were from another planet studying the planet Earth. You might send satellites to take pictures of the planet, and you could zoom in to get more detail. However, you'd eventually be faced with the fact that no matter how much zooming you do, you can't figure out how the complicated cities across the planet work. You'd want to beam fellow aliens down to live in the cities and observe events, blending in so that they're acting as spies. This is what molecular imaging is all about.
Molecular imaging is geared to truly understanding the body. It's like sending little molecular spies into the body so they can live in the environment of molecules and observe what goes on around them. And those spies can then relay that information to the external world.
Medscape: How long has this approach been in place?
Dr. Gambhir: Nuclear medicine is well over 50 years old, and we've always had this concept of using molecules to observe molecular events in the body. They communicate with the external world by being radioactive. We've gotten better at making molecules that zero in on specific targets, and attaching things to them so they can signal back to you. The most primitive form of this concept is injecting simple radioactive iodine for thyroid disease.
Traditionally, radiologists found molecular images tricky to interpret because they were not very anatomically meaningful. You could get a signal that tells you there's a problem at the molecular level, but there was no superimposed anatomy to clarify where the signal was coming from. Initially, radiologists were frustrated in that they could see either just anatomy or just molecular information. Now, the two fields of anatomical and molecular imaging are merging, and we can appreciate the power each has to offer.
Medscape: What are some practical examples of this phenomenon?
Dr. Gambhir: FDG-PET and FDG PET-CT are used by almost every medium- to large-sized hospital, and these are perfect current-day examples. They are being used in cancer patients to stage tumors, monitor recurrence, and monitor effectiveness of therapies. The technologies and applications are used by radiologists, nuclear medicine physicians, and oncologists who are all working hand-in-hand.
At the molecular level, they specifically image the levels of glucose transporters and the hexokinase enzyme -- both of which are present in cancer cells at high levels. These molecular detectives hone in on cancer cells (more so than on normal, healthy cells) and you see a signal coming from tumors that you wouldn't see in healthy tissue, nor in conventional anatomic images.
Medscape: There is some discussion about using imaging to screen for cancer, before clinical symptoms become apparent. Where is molecular radiology in terms of that?
Dr. Gambhir: It's tough applying molecular techniques to population screening, largely because the tools are expensive and are not as easy to use as other screening methods, such as a blood test. Molecular techniques require making probes that have to be injected into the patient, and there is a cost associated with making those probes. Plus, if the probes are radioactive molecules, you have to figure out the logistics of dealing with radioactivity. These tools are not currently used in the asymptomatic screening setting because of the prohibitive costs, and cost-effectiveness for general population screening is yet to be determined.
However, if you go to a high-risk population -- for instance, smokers who have an increased risk for lung cancer -- then PET-CT might be more feasible. This sort of application may soon be tested in trials around the country.
Traditionally, molecular imaging was first used when it was already known that disease was present. The earliest applications of FDG PET-CT were for staging, which is an easier application than for early diagnosis. Eventually, it was thought it could be used to monitor for recurrence -- because now you had a high-risk population to work with.
Medscape: One of the hot topics in molecular imaging ties in with gene-based therapy. Where are we now with that, and where are we headed?
Dr. Gambhir: This is something my own lab is working on. The idea with gene therapy is that you deliver 1 or more genes to treat a disease, and you give the gene either once or infrequently. And gene therapy might even cure the disease.
The problem is, we don't know what happens to genes once they are delivered. We don't know where they go, whether the genes turn on, and, if they do, for how long. Imaging helps you answer each of these questions. The tracking of a gene's activity is done through use of an imaging probe that is injected into the body and acts as a molecular detective. This probe goes all over your body and looks for cells in which the desired gene has been turned on. It accumulates in those cells and sends signals to the external world. We have such probes and have used them in human gene therapy already. The hope is that as these become more routine, and as gene therapy itself starts to become more routine, we'll have better ways to image it. Good imaging will let us optimize the therapy.
The problem, as with any therapy, is that you need to be able to measure whether it's working so you don't waste your time (possibly months) with a therapy that doesn't work. With molecular imaging, things are shifting so that you can give a drug and then very quickly know whether the therapy is working. That lets you decide whether to stop or continue the therapy. This is important in cancer treatment, because we don't want to keep giving therapy that isn't working. Even if it's a nontoxic therapy, you want to stop if it's ineffective so that a patient doesn't become resistant to another therapy that might help. Molecular imaging is good at solving those kinds of issues.
Medscape: What kinds of probes are there?
Dr. Gambhir: Probes vary. Some probes are radioactive, that is, they can send a signal through a radioactive atom. There are optical probes, but these are used mostly in animal studies right now, not in clinics. Optical probes may start being used clinically in 3 to 5 years. Quantum dots are in the optical category, and are not likely to be in humans for any routine application for a few years. One problem with quantum dots, and other optical probes, is that light doesn't do a great job getting through thick tissues.
The question is, Why doesn't light do so well? The reason is because most visible light is absorbed by hemoglobin. Near-infrared light, which is beyond the visible range, does a better job of getting through tissue. Quantum dots, which emit in the near-infrared range, will do better, but they're still nothing like radioactive probes, where signals can travel through much larger depths.
One of the ongoing debates in molecular imaging is, How do you build probes that can get signals from anywhere within your body out to the external world? One way is through the use of radioactive probes. Another type is optical probes, which might be useful in breast imaging and a few other areas. And some probes are detectable with an MRI camera. There are also ultrasound probes, whereby you inject various types of bubbles that contain molecules that hone in on molecular events. So, ultrasound can be used for molecular imaging, too. Signals can be from almost any part of the physical spectrum, but which signal is better will depend on the results of ongoing studies for different applications.
Medscape: Are there specific compounds that radiologists should look for in the future?
Dr. Gambhir: One type is compounds that target new blood vessels. There is a lot of excitement about the possibility of slowing down cancer by cutting off the blood supply to tumors. Let's say you give a patient an anti-angiogenic agent. How do you know it has truly inhibited blood vessel growth in the target area? There are receptors on the surface of blood vessels, such as growth factors, that molecular probes can hone in on. There are several probes coming down the pipeline that measure angiogenesis receptors.
There are also molecular probes in the works that measure different genes expressed only in cancer cells; other probes are being developed that detect Alzheimer's disease early on via beta-amyloid receptors. There are probes in the pipeline for other areas of neuroscience, as well as some cardiovascular areas. Radiologists will see these over the next several years.
Medscape: About how many of these probes are in the pipeline?
Dr. Gambhir: I'd say about 6 to 12. The reason there aren't hundreds is that there are so many regulatory issues to contend with and so many safety tests that have to be performed. The US Food and Drug Administration is looking at mechanisms to accelerate movement of these probes into the marketplace so they can be given to humans, especially since many of them don't need to be given at pharmacologic levels. Often, they are given at trace levels, so there's really no measurable pharmacologic effect. If development can be speeded up, I would guess this would allow for dozens of probes to move up the pipeline.
Medscape: What are some specific examples of innovations coming down the pipeline?
Dr. Gambhir: There are the RGD peptides, which image angiogenesis. There is the fluorothymidine (FLT) probe, which measures DNA replication or cell proliferation. There is FDDNP, which is a probe that binds to beta-amyloid in patients with Alzheimer's disease. There is the annexin probe, which measures cell death. Measuring cell death is potentially useful in cancer therapies, to measure the point at which cells die, or even in the treatment of heart attacks or infarcts.
Medscape: How much promise does optical imaging have?
Dr. Gambhir: There are some breast-imaging systems being put into the clinics that use optics and are breaking new ground. Traditionally, breast imaging is done with mammography, which is just a low-energy x-ray. MRI has been considered for this, but provides mostly anatomic pictures. However, the breast doesn't have that much tissue. You can inject optical probes that hone in on breast tumors and then emit a light signal. There are a lot of exciting developments in optical imaging. Some of the key optical system developers are Advanced Research Technologies, Philips, Fuji, and others.
Medscape: Nuclear medicine and diagnostic imaging must be assuming ever more important roles. Where is this all going?
Dr. Gambhir: We're certainly seeing an acceleration in the development of new kinds of molecules that are going to be utilized for very specific applications. We're going to see a lot more customized imaging. We're going to see that one imaging study doesn't fit everyone -- different imaging probes measure different things and are appropriate for different subpopulations. And so, customized imaging matched to customized therapy is definitely one direction.
Medscape: I hear the word "custom" and I hear a cash register in the background. Is all this going to be really expensive?
Dr. Gambhir: We don't know yet. It's very hard for companies to make probes that can be applied only to small numbers of people. That's why you don't see many drug companies developing drugs for diseases that afflict only 50 people in the world. The economics of customized imaging is tough. How do you make probes that are effective for a small group of people and make a sufficient quantity of probes when your sales are small? There are ways around this that are being considered, such as more publicly based, molecular library-based strategies to try to minimize the cost. But initially this won't be cheap. On the other hand, PET-CT wasn't initially cheap, either, and people thought it was too pricey to use clinically. Now it's a routine procedure that's Medicare-reimbursed. Long term, I don't think cost will be the big issue, but it will be in the short term.
Medscape: One of your areas of interest is nanotechnology. How does this fit in with molecular imaging?
Dr. Gambhir: Our center, along with several others, have become National Institutes of Health centers of excellence for nanotechnology. We're building nanoparticles that go into the body to image different events. Initially, nanotechnologies will be more expensive. But, as nanotechnology develops and grows, we'll be able to build these things more quickly and efficiently, and prices will come down.
We're looking at nanoparticles -- the definition being a particle that is on the order of 20 to 100 nanometers in size. This is very small, but still much bigger than, for example, an FDG molecule. Nanoparticles will be man-made devices -- molecules, really -- that are injected into a person so they can basically live inside that person for a while and provide information that can be translated into images. Whether they'll live inside the person for just a few days and then be excreted through the urine, or whether they'll stay in longer, remains to be seen.
Medscape: Can you explain how nanotechnology imaging for disease would work? Is this different from traditional probes?
Dr. Gambhir: Nanoparticles involve both inorganic and organic components. Examples of inorganic components include gold, cadmium, and zinc. Organic elements such as carbon or nitrogen are attached to these inorganic materials. But nanoparticles can be more sophisticated than that. A nanoparticle can, for example, bind to 6 or 7 different receptors instead of just 1. Your nanotechnology probe can image several things at once.
Medscape: Is there a specific area in the molecular imaging field that really gets you excited?
Dr. Gambhir: I think the most exciting thing for me personally is that molecular imaging is a multidisciplinary field. It takes the best of chemistry, engineering, medicine, and pharmacology, and brings them together. It's not about any one field. What gets me excited about it is not any one application either -- it's this general concept of being able to put detectives, or spies, into the human body. It's a very powerful concept. Going back to the alien race analogy, you need to have something that can live inside the body. That's what these tools allow us to do.
Editor's Note:
Molecular imaging is poised to revolutionize medicine, particularly in the arena of targeted therapies and the follow-up assessment for recurrence of disease in patients with cancer.
Sanjiv Sam Gambhir, MD, PhD, is the Director of the Molecular Imaging Program at Stanford (MIPS). The primary objectives of the MIPS program are to foster innovations in the early diagnosis of disease through the use of imaging modalities such as positron emission tomographic (PET) scanning and PET/computed tomographic (CT) fusion studies and to monitor the impact of emergent molecular therapies on clinical outcomes.
Dr. Gambhir sat down with Pippa Wysong of Medscape to detail the recent advances in molecular imaging that have the greatest relevancy to oncologists and other referring physicians.
Medscape: Can you tell Medscape readers why molecular radiology is such an exciting area?
Dr. Gambhir: There is a fundamental shift in what imaging can tell you when you bring in molecular details. I like the following analogy: Imagine you were from another planet studying the planet Earth. You might send satellites to take pictures of the planet, and you could zoom in to get more detail. However, you'd eventually be faced with the fact that no matter how much zooming you do, you can't figure out how the complicated cities across the planet work. You'd want to beam fellow aliens down to live in the cities and observe events, blending in so that they're acting as spies. This is what molecular imaging is all about.
Molecular imaging is geared to truly understanding the body. It's like sending little molecular spies into the body so they can live in the environment of molecules and observe what goes on around them. And those spies can then relay that information to the external world.
Medscape: How long has this approach been in place?
Dr. Gambhir: Nuclear medicine is well over 50 years old, and we've always had this concept of using molecules to observe molecular events in the body. They communicate with the external world by being radioactive. We've gotten better at making molecules that zero in on specific targets, and attaching things to them so they can signal back to you. The most primitive form of this concept is injecting simple radioactive iodine for thyroid disease.
Traditionally, radiologists found molecular images tricky to interpret because they were not very anatomically meaningful. You could get a signal that tells you there's a problem at the molecular level, but there was no superimposed anatomy to clarify where the signal was coming from. Initially, radiologists were frustrated in that they could see either just anatomy or just molecular information. Now, the two fields of anatomical and molecular imaging are merging, and we can appreciate the power each has to offer.
Medscape: What are some practical examples of this phenomenon?
Dr. Gambhir: FDG-PET and FDG PET-CT are used by almost every medium- to large-sized hospital, and these are perfect current-day examples. They are being used in cancer patients to stage tumors, monitor recurrence, and monitor effectiveness of therapies. The technologies and applications are used by radiologists, nuclear medicine physicians, and oncologists who are all working hand-in-hand.
At the molecular level, they specifically image the levels of glucose transporters and the hexokinase enzyme -- both of which are present in cancer cells at high levels. These molecular detectives hone in on cancer cells (more so than on normal, healthy cells) and you see a signal coming from tumors that you wouldn't see in healthy tissue, nor in conventional anatomic images.
Medscape: There is some discussion about using imaging to screen for cancer, before clinical symptoms become apparent. Where is molecular radiology in terms of that?
Dr. Gambhir: It's tough applying molecular techniques to population screening, largely because the tools are expensive and are not as easy to use as other screening methods, such as a blood test. Molecular techniques require making probes that have to be injected into the patient, and there is a cost associated with making those probes. Plus, if the probes are radioactive molecules, you have to figure out the logistics of dealing with radioactivity. These tools are not currently used in the asymptomatic screening setting because of the prohibitive costs, and cost-effectiveness for general population screening is yet to be determined.
However, if you go to a high-risk population -- for instance, smokers who have an increased risk for lung cancer -- then PET-CT might be more feasible. This sort of application may soon be tested in trials around the country.
Traditionally, molecular imaging was first used when it was already known that disease was present. The earliest applications of FDG PET-CT were for staging, which is an easier application than for early diagnosis. Eventually, it was thought it could be used to monitor for recurrence -- because now you had a high-risk population to work with.
Medscape: One of the hot topics in molecular imaging ties in with gene-based therapy. Where are we now with that, and where are we headed?
Dr. Gambhir: This is something my own lab is working on. The idea with gene therapy is that you deliver 1 or more genes to treat a disease, and you give the gene either once or infrequently. And gene therapy might even cure the disease.
The problem is, we don't know what happens to genes once they are delivered. We don't know where they go, whether the genes turn on, and, if they do, for how long. Imaging helps you answer each of these questions. The tracking of a gene's activity is done through use of an imaging probe that is injected into the body and acts as a molecular detective. This probe goes all over your body and looks for cells in which the desired gene has been turned on. It accumulates in those cells and sends signals to the external world. We have such probes and have used them in human gene therapy already. The hope is that as these become more routine, and as gene therapy itself starts to become more routine, we'll have better ways to image it. Good imaging will let us optimize the therapy.
The problem, as with any therapy, is that you need to be able to measure whether it's working so you don't waste your time (possibly months) with a therapy that doesn't work. With molecular imaging, things are shifting so that you can give a drug and then very quickly know whether the therapy is working. That lets you decide whether to stop or continue the therapy. This is important in cancer treatment, because we don't want to keep giving therapy that isn't working. Even if it's a nontoxic therapy, you want to stop if it's ineffective so that a patient doesn't become resistant to another therapy that might help. Molecular imaging is good at solving those kinds of issues.
Medscape: What kinds of probes are there?
Dr. Gambhir: Probes vary. Some probes are radioactive, that is, they can send a signal through a radioactive atom. There are optical probes, but these are used mostly in animal studies right now, not in clinics. Optical probes may start being used clinically in 3 to 5 years. Quantum dots are in the optical category, and are not likely to be in humans for any routine application for a few years. One problem with quantum dots, and other optical probes, is that light doesn't do a great job getting through thick tissues.
The question is, Why doesn't light do so well? The reason is because most visible light is absorbed by hemoglobin. Near-infrared light, which is beyond the visible range, does a better job of getting through tissue. Quantum dots, which emit in the near-infrared range, will do better, but they're still nothing like radioactive probes, where signals can travel through much larger depths.
One of the ongoing debates in molecular imaging is, How do you build probes that can get signals from anywhere within your body out to the external world? One way is through the use of radioactive probes. Another type is optical probes, which might be useful in breast imaging and a few other areas. And some probes are detectable with an MRI camera. There are also ultrasound probes, whereby you inject various types of bubbles that contain molecules that hone in on molecular events. So, ultrasound can be used for molecular imaging, too. Signals can be from almost any part of the physical spectrum, but which signal is better will depend on the results of ongoing studies for different applications.
Medscape: Are there specific compounds that radiologists should look for in the future?
Dr. Gambhir: One type is compounds that target new blood vessels. There is a lot of excitement about the possibility of slowing down cancer by cutting off the blood supply to tumors. Let's say you give a patient an anti-angiogenic agent. How do you know it has truly inhibited blood vessel growth in the target area? There are receptors on the surface of blood vessels, such as growth factors, that molecular probes can hone in on. There are several probes coming down the pipeline that measure angiogenesis receptors.
There are also molecular probes in the works that measure different genes expressed only in cancer cells; other probes are being developed that detect Alzheimer's disease early on via beta-amyloid receptors. There are probes in the pipeline for other areas of neuroscience, as well as some cardiovascular areas. Radiologists will see these over the next several years.
Medscape: About how many of these probes are in the pipeline?
Dr. Gambhir: I'd say about 6 to 12. The reason there aren't hundreds is that there are so many regulatory issues to contend with and so many safety tests that have to be performed. The US Food and Drug Administration is looking at mechanisms to accelerate movement of these probes into the marketplace so they can be given to humans, especially since many of them don't need to be given at pharmacologic levels. Often, they are given at trace levels, so there's really no measurable pharmacologic effect. If development can be speeded up, I would guess this would allow for dozens of probes to move up the pipeline.
Medscape: What are some specific examples of innovations coming down the pipeline?
Dr. Gambhir: There are the RGD peptides, which image angiogenesis. There is the fluorothymidine (FLT) probe, which measures DNA replication or cell proliferation. There is FDDNP, which is a probe that binds to beta-amyloid in patients with Alzheimer's disease. There is the annexin probe, which measures cell death. Measuring cell death is potentially useful in cancer therapies, to measure the point at which cells die, or even in the treatment of heart attacks or infarcts.
Medscape: How much promise does optical imaging have?
Dr. Gambhir: There are some breast-imaging systems being put into the clinics that use optics and are breaking new ground. Traditionally, breast imaging is done with mammography, which is just a low-energy x-ray. MRI has been considered for this, but provides mostly anatomic pictures. However, the breast doesn't have that much tissue. You can inject optical probes that hone in on breast tumors and then emit a light signal. There are a lot of exciting developments in optical imaging. Some of the key optical system developers are Advanced Research Technologies, Philips, Fuji, and others.
Medscape: Nuclear medicine and diagnostic imaging must be assuming ever more important roles. Where is this all going?
Dr. Gambhir: We're certainly seeing an acceleration in the development of new kinds of molecules that are going to be utilized for very specific applications. We're going to see a lot more customized imaging. We're going to see that one imaging study doesn't fit everyone -- different imaging probes measure different things and are appropriate for different subpopulations. And so, customized imaging matched to customized therapy is definitely one direction.
Medscape: I hear the word "custom" and I hear a cash register in the background. Is all this going to be really expensive?
Dr. Gambhir: We don't know yet. It's very hard for companies to make probes that can be applied only to small numbers of people. That's why you don't see many drug companies developing drugs for diseases that afflict only 50 people in the world. The economics of customized imaging is tough. How do you make probes that are effective for a small group of people and make a sufficient quantity of probes when your sales are small? There are ways around this that are being considered, such as more publicly based, molecular library-based strategies to try to minimize the cost. But initially this won't be cheap. On the other hand, PET-CT wasn't initially cheap, either, and people thought it was too pricey to use clinically. Now it's a routine procedure that's Medicare-reimbursed. Long term, I don't think cost will be the big issue, but it will be in the short term.
Medscape: One of your areas of interest is nanotechnology. How does this fit in with molecular imaging?
Dr. Gambhir: Our center, along with several others, have become National Institutes of Health centers of excellence for nanotechnology. We're building nanoparticles that go into the body to image different events. Initially, nanotechnologies will be more expensive. But, as nanotechnology develops and grows, we'll be able to build these things more quickly and efficiently, and prices will come down.
We're looking at nanoparticles -- the definition being a particle that is on the order of 20 to 100 nanometers in size. This is very small, but still much bigger than, for example, an FDG molecule. Nanoparticles will be man-made devices -- molecules, really -- that are injected into a person so they can basically live inside that person for a while and provide information that can be translated into images. Whether they'll live inside the person for just a few days and then be excreted through the urine, or whether they'll stay in longer, remains to be seen.
Medscape: Can you explain how nanotechnology imaging for disease would work? Is this different from traditional probes?
Dr. Gambhir: Nanoparticles involve both inorganic and organic components. Examples of inorganic components include gold, cadmium, and zinc. Organic elements such as carbon or nitrogen are attached to these inorganic materials. But nanoparticles can be more sophisticated than that. A nanoparticle can, for example, bind to 6 or 7 different receptors instead of just 1. Your nanotechnology probe can image several things at once.
Medscape: Is there a specific area in the molecular imaging field that really gets you excited?
Dr. Gambhir: I think the most exciting thing for me personally is that molecular imaging is a multidisciplinary field. It takes the best of chemistry, engineering, medicine, and pharmacology, and brings them together. It's not about any one field. What gets me excited about it is not any one application either -- it's this general concept of being able to put detectives, or spies, into the human body. It's a very powerful concept. Going back to the alien race analogy, you need to have something that can live inside the body. That's what these tools allow us to do.
SHARE
