Laura Lane

Molecular imaging enthusiasts would undoubtedly like to encounter more residents like Dr. Jinha M. Park. In his last year of residency at the University of California, Los Angeles, Park completed his fellowship at Stanford, where he spent a year studying optical imaging modalities.

Few would argue that Park's scholastic pursuits are above and beyond those of most residents. Yet radiologists are voicing a diversity of viewpoints regarding how residency programs should integrate MI curricula. Many experts are pushing for more focused and extensive training. Faculty members and policy makers recommend a more gradual approach. Meanwhile, the debut of MI questions on the American Board of Radiology's certification exam demonstrates that the molecular era of radiology has begun.

Across the country, program directors have taken the cue.

"A subject that's lab-based [like MI] and has clear and immediate clinical applications is important for residents to learn. But competing demands prevent us from requiring residents to spend a lot of time on MI," said Dr. Gautham Reddy, residency program director at the University of California, San Francisco. "In five or 10 years, as MI fulfills its great potential in becoming a clinically important technique, in-depth training is not going to be optional."

For now, required coursework encompasses nine review papers published as a primer on molecular biology for imagers in Academic Radiology between 2003 and 2004. The material forms the basis for MI questions that appear on the ABR test, according to senior author Dr. King Li, who chairs the National Institute of Health's Imaging Sciences Program. The questions focus on differences between the various imaging modalities, such as target specificities, penetration depths, and resolutions. Other questions cover basic molecular biology, genomics, proteomics, and molecular diagnostics.

Residents and practicing radiologists can master the information and then apply the knowledge to considering the molecular basis of morphologic images that they see every day, Li said.

"The first step toward becoming a molecular imager is to start thinking about the molecular basis of disease, even if you're looking at a mammogram or CT scan," he said.

Lectures are also a convenient way for residency programs to address MI. The American College of Radiology's Commission on Molecular Imaging has proposed a set of 14 lectures on DVD that cover the most relevant MI topics. The first three lectures focus on basic molecular biology, which serves as a review for residents, many of whom learned the material in medical school. The next three address probe design and pharmacokinetics that are unique to MI. The final eight lectures address the preclinical and clinical applications of MI in PET, SPECT, optical imaging, MRI, MR spectroscopy, and ultrasound. All members of the ACR will receive a free copy of the DVD with the August 2006 issue of the Journal of the American College of Radiology.

Such endeavors are providing more opportunities for residents to learn about MI. However, depending on the definition, MI has been part of radiology since the field began using radioiodine 60 years ago, said Dr. James H. Thrall, radiologist-in-chief at Massachusetts General Hospital in Boston.

An estimated 80% of nuclear medicine already meets the definition of MI, he said. With molecular imaging probes consisting of a component for localization and another for tracing, they're conceptually similar to traditional radiopharmaceuticals.

"If you understand the mechanism of localization, the target, and the mechanism of detection, then you can infer the clinical applications," he said.

On the other hand, much of MI can't be restricted to nuclear medicine, Thrall said. That fact has some educators in a quandary over how to integrate MI curriculum, especially with the constraint of a fixed four-year span. Others caution that curricula changes could be in vain, given the unpredictability of medicine.

If MI never gains a firm foothold in the clinic, residents would have wasted time that they could have spent on established modalities, said Dr. Jannette Collins, an associate professor of radiology at the University of Wisconsin in Madison.

But patience may not be a winning virtue, according to Dr. Gary J. Becker, assistant executive director of the ABR. In the past, wait-and-see strategies worked because new technologies were natural extensions of existing modalities. Other specialists lacked the basis to get involved. But given the interdisciplinary nature of MI, radiologists must be more proactive.

"We radiologists must become engaged in the basic science, and we must insinuate ourselves into the MI research at the translational interface," he said. "If we do not work to earn our place in directing and conducting the studies that introduce new MI methods into the clinics, we will have lost an opportunity forever."

But radiologists may have a difficult time embracing a new MI paradigm, one that involves examining images of fluorescent molecules or signals resulting from enzymatic reactions, according to Dr. Hedvig Hricak, chair of radiology at Memorial Sloan-Kettering Cancer Center in New York City.

That's partly why Hricak supports the use of training as a way for radiologists to prepare for a future in MI. While some educators suggest introducing MI through lectures, research seminars, and journal clubs, others advocate MI workshops now available at most conferences. Others are not only calling for MI integration but also for a reexamination of traditional approaches to training.

"Unless we're willing to radically change the way we are training our residents, we're not going to add all of the exciting things on the horizon," said Dr. Stanley Baum, emeritus chair of radiology at the University of Pennsylvania in Philadelphia. "I would take a very hard look at the rotations and shave a lot of them."

Eliminating the internship year has also been proposed to make room for MI training.

"We've been on the warpath to do away with the internship year as we know it," said Dr. Ronald Arenson, chair of radiology at UCSF. "Our residents still need clinical exposure, but they also need clinical time as part of their subspecialty training in an expanded fellowship."

At a minimum, residents should have the freedom to spend more time on their future subspecialty and less time on skills that they won't use in their careers, he said. For example, residents who plan to specialize in neuroradiology should spend less time perfecting their mammography skills.

This personalized approach can improve the cookie-cutter nature of current programs, said Dr. Jeremy Friese, a chief resident at the Mayo Clinic in Rochester, MN. While acknowledging the importance of understanding all the modalities and body systems, he points out that after two years, most residents have already chosen a subspecialty.

Another possibility is eliminating outdated modalities that are becoming obsolete. It's a controversial topic. Educators point to barium x-ray imaging of the gastrointestinal tract and nuclear studies using a gamma camera as possibilities. But residents say that board exams require them to learn the older modalities even though they are being used less frequently.

"If there's a way to cut those out, we could focus on the cutting-edge techniques," Park said.

Independent of the board exams, however, residents still need sufficient experience in all imaging techniques before they can choose a subspecialization, Collins said. Without such exposure, residents can't make informed decisions about which rotations will deliver the most value.

"This could lead to deficiencies in training that might be limiting in private practice," she said.

Most educators agree that MI programs should prepare residents for a lifetime of learning. But doing so requires a change in educational philosophy, Baum said. Far more important than maximizing the amount of material, programs should focus on providing a firm foundation in medicine, which residents can use to acquire new skills and learn new technologies.

Research experience has shown him MI's potential to lead the way into the age of personalized medicine. That radiology could be instrumental in bringing about that shift is a great motivator, Park said. In the end, the most exciting goal is very simple.

"Molecular imaging is good for the patient," he said.

Articles

Educators debate ways to integrate MI curricula

Four strategies for creating MR imaging probes came to light at the International Society for Magnetic Resonance in Medicine meeting in May. While differential accumulation of contrast agent may not occur, the agent appears only where it's turned on exclusively in target tissues, said Angelique Louie, Ph.D., an assistant professor of biomedical engineering at the University of California, Davis.

Selective imaging first surfaced to track the transfection and expression of exogenous genes. Researchers built a construct containing both the gene of interest and a marker gene. Reflecting the expression of the gene of interest, the marker gene communicates by serving as the catalyst for activating the contrast agent. The tyrosinase gene, for example, encodes for melanin that causes iron accumulation in the cell, which can be viewed with T2 enhancement. Constitutive expression of a modified transferrin receptor fills cells with abnormally high levels of iron.

Another marker gene encodes for the enzyme carboxypeptidase, which cleaves a polylysine tail from gadolinium chelate, allowing it to bind to human serum albumin. Bound to albumin in the target tissue, gadolinium distinguishes itself with the slower rotation and higher relaxivity characteristics of larger molecules. The reverse strategy uses agents that are attached to an agglomeration of moieties. A marker enzyme cleaves off these moieties to increase rotation and decrease relaxivity.

Water molecules also modulate the correlation time of the contrast agent. Increasing pH deprotonates phenol and frees it to bind gadolinium. The binding event displaces water from gadolinium and decreases its relaxivity. Increasing its relaxivity, the gadolinium chelate can bind to water once the marker enzyme beta-galactosidase cleaves off lactose.

Louie's research focuses on detecting a change in voltage using a gadolinium chelate. In response to such a change, the attached moiety takes on an open conformation, allowing water to bind and increase the relaxivity.

Altering the exchange rate of water can change relaxivity and contrast. When bound to molecular moieties that deprotonate in response to pH, both gadolinium and europium experience shifts in the rate at which they bind to and detach from water molecules.

Molecular Imaging Briefing

Molecular imaging tracers, agents, and probes have been cast in a new tactical assemblage designed to improve investigational drug evaluations.

By signaling that drugs have reached their pathologic targets or serving as a surrogate marker for drug activity, MI promises to help reduce the financial risks of therapeutic drug development. It may also increase the likelihood that agents selected for costly human trials will someday reach the market.

The pharmaceutical industry can use help: Only a small percentage of drugs have received FDA approval recently, and pain-relieving Cox II inhibitors such as Vioxx were withdrawn from the market after their implication in cardiovascular problems. With patents for several highly profitable drugs due to expire soon, the pharmaceutical industry is placing its hopes on MI's potential to speed replacement agents through the R&D pipelines.

"Big pharma needs to speed up clinical evaluation of drugs. That's where imaging comes in," said Michael J. Welch, Ph.D., a professor of radiology at Washington University in St. Louis.

Every imaging modality, with the exception of plain-film x-ray and fluoroscopy, may have a role to play in drug development. Pharmaceutical manufacturers are applying PET, SPECT, MR, ultrasound, and optical imaging technologies in animal studies because they can generate safety and efficacy data about new drugs faster than conventional experimental methods, Welch said.

Imaging is useful for testing not only potential blockbuster drugs but also targeted agents that treat small groups of patients who share similar molecular pathologies. MI can select these patients to improve clinical trial outcomes by providing direct measures of drug targets, defining the biological process of disease, and separating responders from nonresponders, said Michael Phelps, Ph.D., director of the Institute for Molecular Medicine and the Crump Center for Molecular Imaging at the University of California, Los Angeles.

Growing interest in MI will encourage universities and hospitals to engage in clinical trials, increasing the financial resources of radiology and nuclear medicine departments. Clinicians will increasingly rely on MI to deliver diagnoses with greater precision and foster more informed treatment decisions for cancer, neurodegenerative diseases, and cardiovascular conditions.

"In the future, radiologists may be called upon to make the gatekeeper decisions on whether or not therapeutics should be administered," said Dr. George Mills, director of the Center for Drug Evaluation and Research at the FDA.

Clinicians will eventually draw most of those conclusions based on a variety of indications and molecular signatures representative of drug action. These biomarkers act as surrogate evidence in the form of anatomic, physiological, or other biological changes. Rigorous validation studies verify that these changes are directly tied to the biological pathway of the drug target, said Homer Pien, Ph.D., managing director at the Center for Biomarkers at Massachusetts General Hospital.

"To get comprehensive assessment of any patient, you need to look at a combination of biomarkers," Pien said. "Some biomarkers show the toxicity profiles, some show efficacy, and others show distribution."

In contrast to CT's ability to gather anatomic measures of lesion size, prevalence, and location, MR and PET can measure physiological processes such as perfusion, diffusion, permeability, blood oxygenation, and glucose metabolism. As a result, they provide information that can serve as a good predictor of clinical outcome, Pien said.

Scanners capable of gathering such data from small animals during preclinical trials have been in use for about five years. During that time, pharmaceutical companies have created divisions or collaborations with academia to incorporate molecular imaging into their development strategies.

The physiologic responses of animals examined on micro-PET or CT scanners can be compared with the drug response of humans who receive the same agents before being scanned on full-scale equipment. The new technology makes comparisons of animals and humans possible, Mills said.

"These new uses set the stage for assessing how the drug will fare in human testing," he said.

That is true despite the fact that animals, cohorts of which are usually genotypically identical, rarely offer observations perfectly predictive of those in humans. Still, preclinical testing provides information reliable enough to weigh heavily in determining which drug candidates display the greatest potential, said Dr. Anwar Padhani, lead MR researcher at the Paul Strickland Scanner Centre at Mount Vernon Cancer Center in Northwood, U.K. Today's imaging capabilities bolster the confidence with which drug developers make decisions, saving money that would have been lost in failed clinical trials.

"When drug companies go from animal to human, things suddenly get incredibly expensive, so they want to pick a winner at an earlier stage. However, not all drugs that are predicted by imaging to win finally get licensed; that depends on other factors also," said Padhani, associate director of oncology at Synarc, a clinical imaging research organization that supports imaging studies for pharmaceutical companies.

Optical imaging could be the means to screen out the losers. Optical imaging, which does not involve the noise that hampers interpretation of PET and other modalities, uses luciferase reporter genes to provide precision in studying drug targets. Signals are detected only from cells expressing luciferase, which reflects the presence of receptors, cell surface markers, or enzymes that could be causing disease, as well as their response to drugs.

In addition to delivering clear-cut results, optical imaging is less expensive than other imaging tools. Its reliability means the end of sacrificing multiple cohorts of animals and performing tedious histology and cell protocols. The low cost of optical detectors, which are usually cooled charge-coupled device cameras, and the ease and convenience of using them make optical modalities attractive for drug development, said Christopher Contag, Ph.D., an assistant professor of pediatrics at Stanford University. Such advantages don't extend to studying humans, however. After passing through the large organs and other tissue, optical signals are often too weak to be detected.

These preclinical tactics are a welcome addition to the limited set of tools available for studying neurodegenerative conditions, said Dr. William E. Klunk, an associate professor of psychiatry at the University of Pittsburgh.

"The diseases of the brain are invisible at preclinical stages. That slows the development of drugs," Klunk said.

Bruce Jenkins, Ph.D., and other researchers are addressing this problem by translating MI results into guidelines for diagnosis and treatment. The interpretation of the hemodynamic changes as revealed by pharmacologic MR, for example, may lead to a better understanding of molecular events in the brain and drugs that act on neuronal receptors, said Jenkins, an instructor of radiology at Massachusetts General Hospital.

Instead of relying on PET, which uses tracers to bind to cell receptors or measure biochemical reactions, Jenkins' approach takes advantage of the inherent biology of the tissue of interest to magnify the signal of hemodynamic coupling with drug activity. This approach has been used to study antagonists of neuronal receptors for adenosine, dopamine, serotonin, and acetylcholine.

Dr. Clifford R. Jack, a radiology professor at the Mayo Clinic in Rochester, MN, is developing MRI protocols to examine Alzheimer's disease. His biomarker catalog includes MR spectroscopy for brain metabolites, functional MRI measurement of microperfusion, and volumetric MRI for brain mass measurements that track AD progression.

Observing biochemical changes with FDG-PET can also be advantageous. PET can detect anatomically specific decreases in glucose metabolism and enable detection of mild to moderate Alzheimer's with 93% accuracy as early as five years before the appearance of symptoms, Phelps said. PET also reveals specific metabolic signatures indicative of other forms of degenerative dementias, such as Lewy body dementia and Huntington's disease.

Klunk developed carbon-11 Pittsburgh Compound B (PIB), the most promising agent thus far for imaging amyloid plaques associated with Alzheimer's disease. PIB could help drug developers investigate the effect of anti-amyloid drugs, which can inhibit amyloid synthesis.

Researchers use PET to visualize brain pathology as well. At UCLA, Jorge R. Barrio, Ph.D., has created a tracer that binds specifically to amyloid plaques and neurofibrillary tangles. The agent is fluorine-18 2-dialkylamino-6-acylmalononitrile substituted naphthalenes, or FDDNP, which can reveal the earliest signs of Alzheimer's and serve as another indication of amyloid burden and its breakdown in response to drug candidates.

Sophisticated imaging probes are boosting efforts to perfect oncologic drugs. In the case of anti-angiogenesis drugs, for example, researchers once gathered results by sacrificing cohorts of mice, a procedure that obviously can't be repeated in humans, Pien said.

Functional MR and PET allow researchers to noninvasively examine vascular changes in animals and humans. A comparison of results enables them to better predict the efficacy of the drug and the value of continuing the clinical trial.

New PET tracers will increase the modality's value for drug testing. Radiotracers will focus on specific chemical compounds such as nuclides tethered to antibodies and peptides that bind exclusively to specific targets on tumors. With evidence that copper-60 ATSM accumulates in hypoxic tissue, PET may be used to predict the efficacy of radiation therapy and to steer clinicians toward more effective therapies, according to Welch.

Patrick M. Winter, Ph.D., an assistant professor of medicine at Washington University, is developing targeted MR agents, including alpha v beta 3-targeted paramagnetic nanoparticles. Tumor neovasculature and the angiogenic processes associated with atherosclerotic plaque development are known to express the anb integrin. It gave Winter targets for very early detection of cancer and atherosclerosis. Additionally, alpha v beta 3 integrin expression can potentially serve as a surrogate marker for the therapeutic response of anti-angiogenic agents. Studies in mice have thus far shown that signals from the nanoparticles can be detected despite the low concentration of integrin.

"By mapping angiogenesis, the nanoparticles will show how aggressive the tumor is and which parts are more aggressive," Winter said.

Such applications of targeted agents will not be common any time soon, however, according to Padhani.

"The problem is that the imaging agent is experimental, so you end up performing the drug trial in addition to validation of the imaging agent in patients. That won't fly with the FDA," he said.

Padhani anticipates that most drug developers will continue to rely on conventional imaging such as CT and MRI. Those modalities provide reliable information about toxicity, pharmacodynamics, blood flow, and other variables that can help drug companies decide which agents should be moved through the clinical trial process.

"It's extremely unlikely that the FDA will ever license an anti-angiogenesis drug on the basis of functional imaging changes alone," Padhani said. "They will carry only a little weight with the FDA."

The FDA, however, has plans to incorporate the advantages of MI into its clinical trials criteria. The agency's Critical Path Initiative, adopted in 2004, proposed streamlining the approval process for new imaging probes and related drug candidates. Investigational new drug applications can be approved with toxicology testing on just one animal model, Welch said.

"Before, you had to do at least two and often had to include dogs, which gets very expensive," he said. "It's going to be quicker and less expensive."

Through the Critical Path Initiative, the FDA is striving to allow sponsors to more easily use imaging biomarkers during the IND process, according to Mills. The public comment period for the FDA's proposed critical path rules was closed as of Sept. 1, but the final language for the regulations had not yet been published. Mills is confident that the increased use of imaging biomarkers will have desirable effects.

"This use of imaging biomarkers enables sponsors to move quickly into drug development, to assess drug candidates more readily, to save time and costs that would have been devoted previously to less promising candidates, and to focus energies and economic resources on the more promising drug candidates," he said.

Laura Lane

Articles

Educators debate ways to integrate MI curricula

Molecular Imaging Briefing

MI addresses conundrum of investigational drug trials

Molecular Imaging Briefing