Although still largely in the research realm, molecular imaging is beginning to emerge as an important component of the diagnosis of disease and the guidance of therapy. Many radiologists admittedly are somewhat mystified by what this field entails and how it will impact the practice.

Figure 1. Molecular imaging of multidrug resistance (MDR1) P-glycoprotein (Pgp) transport activity in vivo. Pgp has been a pharmacological target for cancer therapy wherein the transporter is inhibited by a variety of drugs known as MDR modulators. One such modulator in clinical trials is valspodar (Novartis). The figure shows an example of mapping in a living human the effect of valspodar on the pharmacokinetics of Tc-99m-Sestamibi, a radiopharmaceutical recognized as a transport substrate by Pgp. Two whole body posterior planar images of a patient pre- and post-treatment with valspodar are shown. Both images were obtained with a nuclear medicine gamma camera 120 minutes after intravenous injection of Tc-99m-Sestamibi. The pretreatment planar image (PRE) shows significant clearance of the radiotracer from the liver and kidneys, a result of prompt Pgp-mediated clearance of the tracer from these tissues. The post-treatment image (POST) was obtained following 24 hours of treatment with valspodar. This image now depicts significantly enhanced retention of the radiopharmaceutical in the liver and kidneys as their respective Pgp transport activities were blocked. The heart, a Pgp-negative tissue, serves as an internal control and shows little difference in retention of the radiopharmaceutical, a sign of molecular specificity. Imaging data such as these can be applied to assessment of Pgp transport activity in tumors to guide therapeutic choices and distribution of drugs in normal tissues for correlation with pharmacogenomic data. Image courtesy of David Piwnica-Worms, MD, PhD, Washington University School of Medicine, St Louis.

As defined by Ronald Blasberg, MD, head of the neuro-oncology PET program  at the Memorial Sloan Kettering Cancer Center (MSKCC), New York City, molecular imaging is the "visualization in space and time of normal as well as abnormal cellular processes at a molecular or genetic level of function." The radiologist can detect and localize these processes within the subject using an imaging device and a molecule, known as a molecular probe or beacon, tracer, or contrast agent, that homes in on specific targets within the body. The noninvasive, in vivo imaging technologies currently include radionuclide, magnetic resonance, and optical imaging.

Contrast agents, in the traditional sense, are not specific for disease processes, as are some radionuclides, eg, the use of iodine 131 for the detection of recurrent thyroid cancer, or the use of FDG-PET for the detection of metabolically active cancer cells. And clinical research in nuclear medicine is advancing in the direction of molecular medicine and targeted therapy. "We can be more specific in our approaches," mentions Blasberg. The goal here is to detect pathophysiological processes early and to target them specifically in appropriate patients.

Sanjiv Sam Gambhir, MD, PhD, director of the Crump Institute for Molecular Imaging at the University of California, Los Angeles, states that "the field is evolving such that we will see more customized imaging. We will probably see probes being developed that are based on a patient's specific molecular errors," he suggests. "For example, tumors will be characterized after surgery for the gene expression pattern by going to a library of molecular probes. This library will be appropriate for imaging their specific cancer. So, instead of just giving everyone an FDG-PET scan, we will probably customize PET scan procedures for individual cancer patients. This customized imaging will hopefully lead to improved diagnosis and disease management.

"In addition to conventional, anatomical approaches like CT and MRI in the traditional sense," Gambhir says, "radiologists will begin to see functional information put into play. When a cancer patient gets therapy, we will not just be looking to see if their tumor is getting smaller, we'll be looking to see if the tumor is responding by having our probes target specific molecules that are, for example, overexpressed in the tumor. This information would lead to much better detection of whether therapies are working."

PROBES AS MARKERS

Using a CT scan, a patient might receive a drug for months before imaging reveals that it is useless. Molecular imaging, however, could show very quickly whether the drug is working because the molecular probes can monitor not just the tumor size but also the tumor's ability to replicate, to express certain proteins on its surface, or to make intracellular enzymes. Molecular probes can be direct, indirect, or surrogate. FDG is an example of a surrogate probe that is detectable by PET. The marker of interest is the glucose transporter and hexokinase, whose increased activities are direct markers for increased cellular metabolism. The detection of increased metabolism is used as a surrogate marker (also known as a bio-marker) for the presence of a tumor, since many tumors have high levels of glucose utilization (Warburg effect).¹

Increased glucose metabolism is not a specific marker for cancer. Radiologists can expect to see the emergence of direct and indirect markers of specific molecular events that serve as signatures of diseases, most notably, individual cancer types, but, perhaps, also other disease types such as Alzheimer's disease. More specific probes include direct probes that are specific for cell surface receptors, intracellular molecules such as those that are synthesized in signal transduction pathways, or during gene expression. Direct probes target these molecules specifically.¹ Indirect imaging involves the use of a reporter gene, in which the probe accumulates in a given area and becomes detectible when the gene of interest is being expressed at that locus.

Table 1. A comparison of the currently available molecular imaging modalities.

CURRENT, EMERGING MODALITIES

Several review articles address the emergence of molecular imaging into the field of radiology.^2-5 In developing molecular probes, there are distinct advantages and disadvantages to each of the three imaging modalities: radionuclide, magnetic, and optical imaging. While radionuclide imaging currently leads the way, optical imaging is an exciting new area. MRI has provided exquisite resolution, yet issues regarding probe design remain a challenge.

1. Radionuclide Imaging. Molecular imaging with radionuclides is performed within the branch of nuclear medicine. A wide array of probes is available currently. Very low concentrations of probe are required, and therefore pharmacological effect is not usually a concern. The images provide quantitative information (expressed in units of radioactivity) in real time. However, image resolution is very low. Nuclear imaging with radiolabeled probes is still limited to the larger centers, ie, ones with PET scanners and radiochemistry wherewithal to develop the probes. This process requires a dynamic and interdependent process between clinical professionals and molecular biologists, as well as experts in chemistry and nuclear physics.

The initial applications of molecular imaging will probably be within the radionuclide field. PET imaging in oncology has exploded enormously within the last 5 years. This change has occurred largely due to reimbursement by the Centers for Medicare & Medicaid Services (formerly known as the Health Care Financing Administration). At MSKCC, PET scans, also referred to as "metabolic" scans, are conducted routinely prior to surgery. Many surgeons will not operate until they have this functional, metabolic scan done for the purpose of identifying and locating the foci of tumor metastases. The use of FDG-PET scans has been cost-effective due to the decreased cost of medical care through the reduction in the ineffective application of surgical procedures.

2. Magnetic Resonance. Radionuclide imaging has the distinct disadvantage of poor resolution; it can be difficult to pinpoint the locus of pathogenesis. MR imaging offers the highest imaging resolution of the three major molecular modalities. However, MRI contrast agents usually need to be given at high (pharmacological) concentrations. Therefore, it is difficult not to overwhelm the system under investigation. "The potential toxicities are going to be higher," explains David Piwnica-Worms, MD, PhD, professor of radiology at Washington University in St Louis School of Medicine. "And because you need more, you might swamp the very system you are trying to study...you can't get enough contrast agent in there before you have exceeded...low-density receptors."

Because each contrast agent needs to go through the FDA approval process, and the economic returns are nowhere close to those of drugs, it is just not cost-effective for vendors to invest in the development of new agents. Furthermore, the addition or substitution of large, paramagnetic atoms to molecules often changes the properties of the parent molecule. "The challenge for MR will be to develop probes and modify the instruments to improve sensitivity," explains Gambhir. "And it's not as simple as changing the coils. It will involve a fundamental change in the way the probes are developed, how they are magnetically labeled, so they are not altered in terms of their pharmacokinetics...MR is struggling with this, unlike PET where you can just swap out a carbon for a carbon or a hydrogen for a fluorine-18 like in FDG-PET. As soon as you swap out a molecule and put in gadolinium or try to chelate a gadolinium, which is an MR susceptible agent...then the molecule doesn't behave like the parent molecule anymore. MR is the most difficult area to adapt to molecular imaging, which is going to be an ongoing challenge. Hence, MRI will need to undergo fundamental changes in the way probes are labeled."

But the wide penetration of MRI is likely to play a role in the future, according to Blasberg. "A major advantage is the clinical MRI base that already exists in almost all hospitals," says Blasberg. It may be that the combined use of MRI and radionuclide imaging will become useful for pinpointing pathophysiological processes more specifically.

3. Optical Imaging. Optical imaging is based on the use of probes that emit radiation in the visible spectrum. Blasberg emphasizes that it is not just a light microscope or looking at the organism through a window in the skull or the skin. "Now with endoscopic technologies, with fluorescent and bioluminescent probes, this [field] is growing very rapidly, much more rapidly than I have ever expected," he notes. Although optical imaging is limited currently to research, he believes that the techniques will find their niche in the practice of radiology, especially through the coupling of endoscopic technology for the purpose of endoscopic imaging. For example, optical coherence tomography (OCT) could be performed through optical imaging for the visualization of individual cells or some molecular event.

Gambhir believes that optical imaging is useful because light is produced (allowing for direct visualization of probe locus), but the light cannot get very far through tissue. He emphasizes that red and near-infrared wavelengths produce the greatest penetration. Optical cameras might be used for imaging breast and skin, and for endoscopic procedures, but not for deep tissues. It offers great sensitivity, but a lot of probe (microgram to milligram quantities) is required. This creates a problem because of pharmacological or toxic side effects. Gambhir believes that more stringent documentations will be required for FDA approval. The FDA may need to institute an approval process for optical probes (as well as MR contrast agents) that is separate from that of radiolabeled probes.

EDUCATING RADIOLOGISTS

Which radiologists will need to bring themselves up to speed regarding the use of molecular imaging in their practices? To a certain extent, it may depend on the age of the radiologist. Older radiologists with established practices may well remain fiscally sound using long-established technologies. Radiologists will continue, for example, to see patients who present with bone fractures, and who require an x-ray or an MRI scan. "When older radiologists were in medical school, they didn't even have restriction enzymes," mentions Piwnica-Worms. Some of these older physicians may not find it worthwhile to embark on the tremendous learning curve required to expand their practice to incorporate molecular imaging. On the other hand, "younger radiologists out of school will need to just hang on to their biochemistry and molecular biology that they learned in medical school," Piwnica-Worms asserts.

Figure 2. Imaging gene expression in vivo in mouse leg muscle. The expression of the gene for dopamine transporters (DAT) was measured using [99mTc]TRODAT-1 and ultra-high resolution single photon emission tomography. In one leg, the gene for DAT was introduced using an adeno-associated viral vector, while the other leg had the same amount of viral vector containing the gene for lacZ as a control. Specific binding of the tracer to the transfected DAT was observed, with significantly higher uptake in the treated muscle (yellow arrow) compared to the control (red arrow). Also shown is a mouse that received 10% of the viral vector, exhibiting reduced uptake of [99mTc]TRODAT-1 in a dose-dependent manner. The image, which was selected as one of two Images of the Year by the Society of Nuclear Medicine (SNM), illustrated Abstract 218, by authors Acton PD, Auricchio A, Plossl K, Wilson J, Kung H, Department of Radiology and Institute for Human Gene Therapy, University of Pennsylvania, Philadelphia, in the proceedings of the SNM Meeting, Los Angeles, 2002.

The American College of Radiology (ACR) held a colloquium in April 2001 that was focused on the translation of molecular imaging research into the practice of radiology.⁶ In order to advance the application of molecular imaging into the practice of radiology, it will be important for universities and research centers to encourage the education of radiologists and radiology residents. Radiologists need to become involved with learning in the areas of molecular pathophysiology, probe design, and novel imaging technology. To a certain extent, education in this area will require that radiologists learn the language of professionals outside of their immediate area(s) of expertise. Communication between radiologists and these other professionals should foster the development and implementation of new ideas for translating the science of molecular imaging into effective clinical practice.

It was concluded at the ACR Colloquium that the Residency Review Committee should work toward including molecular imaging as a training requirement in the "Green Book."⁶ As with many medical specialties, it will be important for radiology residents to participate in molecular imaging research and training programs. As such programs are instituted, it may become mandatory for residents and fellows to spend a minimum of a 1-year period conducting research as part of their programs.

Hedvig Hricak, MD, chairman of the Department of Radiology at MSKCC, has established a training program for clinical radiologists and other molecular imaging professionals. A training grant from the National Cancer Institute provides these professionals with up to 2 years of salary for the purpose of developing an awareness of how molecular imaging relates to their careers.

ON THE CUTTING EDGE

The National Institute of Biomedical Imaging and Bioengineering (NIBIB), the newest member of the NIH, was established on December 29, 2000. Since January 2001, Donna J. Dean, PhD, served as the acting director. Roderic Pettigrew, MD, PhD, was due to assume the role of director on September 23, 2002.

In 2002, the two requests for applications (RFAs) from the NIBIB included: a broad announcement on imaging and a request for proposals concerning sensor technology. These initial RFAs were of a general nature to get a sense of what the community was thinking about. Dean believes that the vast majority, if not all, of the grant applications relevant to molecular imaging received by the NIBIB show promise toward the application to the field of radiology. Some of the key research areas have involved the development of new probes and contrast agents, novel imaging modalities, multiple modality imaging (eg, CT plus PET scan for the simultaneous visualization of structure and function, respectively), and bioinformatic applications to handle vast amounts of visual data.

The NIBIB will want to stimulate the community to do more in the area of developing imaging biomaterials, specifically ones that might serve as sensors of molecular and physiological events. The sensor should emit a signal that can be detected for the purpose of localizing the event and for collecting data from it. As Dean notes: "A picture is worth not a thousand words, but a million words, if you know what you are looking for." She emphasizes that physicians need to have knowledge of molecular and cell biology to understand the new technologies, and consequently how to most effectively and efficiently help the patient. Similarly, the nonclinical scientists need to understand how their work applies to health care delivery.

This prospect will require that radiologists meld their expertise in a collaborative, multidisciplinary approach with many other physicians and scientists. The translation of molecular imaging technology to clinical practice is expected to require collaboration between radiologists, surgeons, internists, clinical scientists, physicists, cell and molecular biologists, geneticists, chemists, mathematicians, engineers, and experts in genomics, proteomics, bioinformatics, and computer science. The design of molecular probes, contrast agents, novel imaging devices, and computational algorithms will be required to translate pathophysiological events at the molecular level into information that the radiologist can interpret.

In the future, there will be more collaborations between drug companies and imaging companiesmore interplay between imaging probe development and drug development. During the next 3 to 5 years, the practicing radiologist will primarily hear about research in small animal models. But most of these agents will fail, and then they will not hear about them anymore. For every 100 probes developed, perhaps a few will be useful. Another challenge that will be faced will be to change the way the FDA regulates probesto convince the agency to treat probes differently from drugs when they are not given in pharmacological amounts. The FDA is attempting to restructure to provide a special branch that will deal with radiopharmaceuticals and molecular probes. This change will be expected to accelerate the preclinical-to-clinical transition of probe usage.

Another challenge is one of economics. According to Gambhir, "Diagnostic probes are a drop in the bucket in terms of dollar revenue in comparison to drugs." He believes that drug companies are not going to spend as much on developing an imaging probe as they would on a drug. "FDG took a long time to get to the point of being FDA approved and reimbursed," he explains. "A practicing radiologist will not see super-FDG in action until Medicare reimburses it for certain applications." He emphasizes that regulatory and reimbursement issues are the key challenges that lie ahead for this field. "There are plenty of probes being developed," he notes. "It's getting them to the clinic that is limiting."

CONCLUSION

Molecular imaging is evolving to allow radiologists to visualize the disease process at cellular and molecular levels. Through an interdisciplinary approach with other physicians, basic scientists, and health care professionals, radiologists will play an increasingly important role in the diagnosis of disease and the management of therapy. Radionuclide imaging of molecular processes should emerge first, followed by MRI and optical imaging. The hope is that molecular imaging will provide for the guidance of targeted therapy for patients through the use of customized imaging, which would allow for the detection of specific pathophysiologic processes. As molecular imaging emerges from the research phase into practice, radiologists can keep abreast of this field through relevant societies and journals. Radiology residents and fellows can expect that molecular imaging will be a substantial part of their training and research requirements. Experts suggest that the emergence of molecular imaging into the practice of radiology would be enhanced by the restructuring of the FDA approval process for the use of probes in patients, as well as through increased interaction between professionals in the areas of health care, science, engineering, and informatics.

Chip Reuben, MS, is a contributing writer for Decisions in Imaging Economics.

References:

1. Blasberg RG, Gelovani J. Molecular-genetic imaging: a nuclear medicine-based perspective. Molecular Imaging. 2002;1:160-180.
2. Weissleder R, Mahmood U. Molecular imaging. Radiology. 2001;219:316-333.
3. Wunderbaldinger P, Bogdanov A, Weissleder R. New approaches for imaging in gene therapy. Eur J Radiol. 2000;34:156-65.
4. Phelps ME. PET: the merging of biology and imaging into molecular imaging. J Nucl Med. 2000;41:661-81.
5. Bremer C, Weissleder R. In vivo imaging of gene expression. Acad Radiol. 2001;8:15-23.
6. Hillman BJ, Neiman HL. Translating molecular imaging research into radiologic practice: summary of the proceedings of the American College of Radiology Colloquium, April 22-24, 2001. Radiology. 2002;222:19-24.