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.

MI addresses conundrum of investigational drug trials

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 would that have been lost in failed clinical trials.

 National Institutes of Health researchers have successfully broken through the cell membrane, a barrier that has frustrated the efforts of many gene therapy strategies. Their study, published in the May 2005 issue of Radiology, found that high-intensity focused ultrasound significantly increases the number of exogenous genes that permeate the cells of subcutaneous squamous cell carcinomas. Ultrasonic waves were delivered in short pulses to dissipate heat and generate mechanical power that increased a tumor cell's permeability by therapeutic genes.

 Under the guidance of Dr. King C.P. Li, Dr. Kristin Dittmar and colleagues in the diagnostic imaging department of the NIH Clinical Center applied ultrasonic waves to one of two tumors growing on both flanks of mice, which immediately received injections of the reporter gene encoded for fluorescent green protein. Fluorescence microscopy detected green fluorescent protein exclusively in tumors that received exposure to the ultrasound. Western blot analysis revealed that the ultrasound-exposed cells emitted fluorescent signals nine times more intense than signals from cells in the tumors that did not receive ultrasound.

 Stained histologic sections reveal the nuclei of tumor cells in blue as seen with fluorescence microscopy. Green fluorescent protein reporter gene expression was not seen in control tumor that was not exposed to pulsed ultrasound (A) but does appear in green in a tumor exposed to pulsed ultrasound (B).

 This procedure is hypothetically generic for enhancing delivery to all tissues, according to coauthor Victor Frenkel, Ph.D. Li previously demonstrated that tumor cells increase their uptake of chemotherapeutic drugs with a similar ultrasound treatment.

"Now we've shown that it works for genes, and we're making the case that there's a connection between the two," Frenkel said.

Focused ultrasound helps exogenous genes permeate targeted cells' outer membrane

 Under the supervision of Dr. Lily Wu, researchers at the University of California, Los Angeles created a prostate-specific two-step transcription amplification (TSTA) technique to deliver suicide gene therapy to hormone-refractory prostate cancer (HRPC), an aggressive cancer that is stubbornly resistant to conventional therapy.

 In a study published in the May 15, 2005, issue of Clinical Cancer Research, lead author Dr. Makoto Sato tested a TSTA-driven adenovirus vector on three androgen-dependent and six HRPC cancers in mice. Optical imaging and PET/CT were used to monitor real-time gene expression. After three weeks of treatment with the adenoviral vector, optical imaging revealed vector activity in the androgen-dependent lines and four of the HRPC cell lines, all of which possess androgen receptors. These HRPC tumors displayed about seven times more luciferase expression than did androgen-dependent tumors. In addition, the HRPC tumors emitted PET signals that were more robust than those from the other tumors. While serum amounts of prostate-specific antigen leveled off in mice that received suicide gene therapy, PSA steadily increased in mice that received only saline. The researchers also treated a set of mice with a similar gene construct under the control of a constitutive viral cytomegalovirus (CMV) promoter. In contrast to mice treated with the TSTA construct, mice that received the CMV promoter revealed a high expression of thymidine kinase in the liver during PET/CT. Liver damage from the thymidine kinase was evidenced by elevated serum levels of liver transaminase.

 PET/CT records results of suicide gene therapy. LAPC-4 tumors were injected with 109 infectious units of prostate-targeted AdTSTA-sr39tk or constitutive AdCMV-sr39tk vector. PET/CT imaging performed seven days later and prior to ganciclovir (GCV) treatment shows (top) tumor-limited expression in the TSTA-treated animal (left), but the CMV-treated animal shows strong expression in the liver as well. After GCV treatment (80 mg/kg per day from day eight to 15), fluorine-18-FHBG PET/CT signals at day 22 (bottom) are diminished in the tumors and the liver of the CMV animal. Histology performed at the day 22 endpoint (right) revealed extensive apoptosis, as shown in TUNEL-positive brown staining in the tumor and liver of the CMV-treated animal.

Data from the study support the conclusion that androgen receptor function is activated in HRPC despite low levels of androgen. Because most recurrent cancers express androgen receptors and PSA, the authors see the TSTA approach as a promising way to administer gene therapy to treat advanced prostate cancer.

Two-step amplification technique delivers gene therapy to refractory prostate cancer

Under the supervision of Dr. Lily Wu, researchers at the University of California, Los Angeles created a prostate-specific two-step transcription amplification (TSTA) technique to deliver suicide gene therapy to hormone-refractory prostate cancer (HRPC), an aggressive cancer that is stubbornly resistant to conventional therapy.

National Institutes of Health researchers have successfully broken through the cell membrane, a barrier that has frustrated the efforts of many gene therapy strategies. Their study, published in the May 2005 issue of Radiology, found that high-intensity focused ultrasound significantly increases the number of exogenous genes that permeate the cells of subcutaneous squamous cell carcinomas. Ultrasonic waves were delivered in short pulses to dissipate heat and generate mechanical power that increased a tumor cell's permeability by therapeutic genes.

Molecular medical researchers no longer have to fly blind as they develop ways to navigate therapeutic genes or stem cells toward disease sites. By harnessing the natural behavior of viruses and cells, molecular imaging illuminates the path of therapeutic genes and cells and serves as the guiding light in human clinical trials. The technology is influencing basic research now and is making the transition to human trials that could have sweeping implication for radiological practice in the future.

The importance of reporter gene imaging stems from MI's incredible demands on image resolution. Ideally, MI should enable a radiologist to directly observe genes as they switch on or off in response to genetically based therapies. In reality, in vivo imaging does not generate enough resolution to delineate individual cells, much less the activity of nanometric structures inside their nuclei. Reporter gene imaging involves a clever set of strategies most frequently using PET, SPECT, or optical imaging to indirectly observe intracellular or genomic events.

Researchers are beginning to appreciate the insights possible with this new technology, according to Dr. Ronald G. Blasberg, director of the Center for Multidisciplinary In Vivo Molecular Imaging at Memorial Sloan-Kettering Cancer Center.

"What you basically have is the people doing gene therapy agreeing to link their gene therapy protocols to reporter gene imaging," Blasberg said.

Reporter gene imaging almost always involves two components: a targeted reporter gene capable of penetrating and occupying cells that exhibit specific gene expressions, and a reporter probe that carries a radioisotope and can accumulate in great quantities only in those cells occupied by the reporter gene. The PET or PET/CT presentation of reporter gene imaging is similar to that of standard FDG-PET. Cells expressing the desired genetic behavior glow orange on the screen. Hyperenhancing stem cells carrying the reporter gene and accumulated probes can be observed as they arrive and begin to proliferate in targeted tissue.

One of the more well-known reporter genes is herpes simplex virus type 1 thymidine kinase (HSV1-tk). After being fused into a construct with a constitutive promoter, HSV1-tk is packaged into a vector, usually an adenovirus (see illustration, page 2). As in many recent studies, the viral vector is injected at a target site such as a tumor. The adenovirus transduces the reporter gene construct into cells, which transcribe the reporter genes into messenger RNA (mRNA) in the nucleus. Once translated, HSV1-tk specifically phosphorylates thymidine analogs, such as the drugs ganciclovir and penciclovir, and traps them in the cell. As thymidine analogs, they can interfere with DNA replication and halt cell division. The use of HSV1-tk to stop cancerous growths has thus earned the name "suicide gene therapy."

With positron-emitting fluorine isotopes attached, low concentrations of these thymidine analogs also reveal their location in the body on PET. Other varieties of reporter probes were designed to abet phosphorylation and/or uptake into the cell. These include 2«-deoxy-2«-[F-18]fluoro-5-fluoro-1-b-D-arabinofuranosyluracil (FFAU) and 9-(4-[F-18]-fluoro-3-hydroxymethyl-butyl)guanine (FHBG). Iodine isotopes such as I-124 also emit positrons and can be incorporated into fluorodeoxyuridine derivatives to produce the probe 5-[I-124]-iodo-2«-fluoro-2«-deoxy-1-b-D-arabinofuranosyluracil (FIAU), which the HSV1-tk also phosphorylates.

"Originally, these radiolabeled nucleosides weren't developed for the purposes of suicide gene therapy but for imaging viral infection in humans," said Dr. Peter S. Conti, director of the PET Imaging Science Center at the University of Southern California. "But when gene therapy hit the presses, it was immediately obvious that a radioactively labeled version of ganciclovir could also be used as a reporter probe."

For years, Conti focused on developing tracers that would allow him to spot viral infections that tend to lay hidden in the body. This approach would be a far better option than, for example, the risky procedure of taking a biopsy of the brain in search of viral encephalitis. To study the infection of herpes simplex virus, Conti and others in the field attached a radioactive tracer to drugs such as ganciclovir, which was commonly used to treat viral infection.

Many systems are now available. Nearly all rely on PET, SPECT, or optical cameras such as cooled charge-coupled devices. Optical imaging is a valuable tool in the preclinical world of cells and small animals, which are small enough to allow detection of fluorescent and bioluminescent light. PET, SPECT, and gamma cameras can detect radionuclide-labeled probes anywhere in the human body, which is too large for optical signals to be detected. Although compact optical breast imagers have been developed, researchers disagree about whether optical images will find a place in clinical practice.

"We use a lot of optically based systems for monitoring small animals, but it doesn't translate well to the clinic," said Harvey Herschman, Ph.D., vice chair of molecular and medical pharmacology at the University of California, Los Angeles.

"For clinical applications, we're strongest at imaging radionuclides with PET or SPECT."

Reporter genes for optical imaging include green and red fluorescent proteins, firefly luciferase, and sea pansy Renilla luciferase. The sophistication of cameras has extended the use of optical imaging, used for decades to study gene expression in cultured cells, to small animals. Fluorescence and bioluminescence are easy to use, have a low price tag, generate images with low background, and do not require radioactive probes. These advantages keep optical imaging popular. It is the preferred modality in preclinical studies, and researchers are developing fiber-optic scopes packaged into catheters and endoscopes to reach the inner organs and cavities of the human body.

"No single imaging modality can address all the questions that we can answer with molecular imaging. Each of the correlating reporters has its strengths and weaknesses," said Dr. David Piwnica-Worms, director of the Molecular Imaging Center at Washington University in St. Louis.

Integrated into human clinical trials, the reporter genes are proving valuable in providing essential feedback. Most are fused into a gene construct with the therapeutic gene, which shares the same promoter. The expression of the reporter gene thus reflects the expression of the therapeutic gene. But because of the additional regulatory requirements for adding the reporter gene, most existing gene therapy protocols use viral thymidine kinase as the suicide gene and as a reporter gene, Blasberg said.

That was the strategy for Dr. Sanjiv "Sam" Gambhir, director of the Molecular Imaging Program at Stanford University (MIPS) and his colleague, Dr. Ivan Penuelas at the University of Navarra in Spain. Their study, scheduled to be published in the Journal of Gastroenterology, provides details of their pioneering efforts in conducting the first human clinical trial of adenoviral gene therapy with reporter gene imaging using FHBG and PET/CT. The results showed active gene expression at sites of the tumor in seven patients with hepatocellular carcinoma who received injections of the viral vector. By comparing images taken on days two and nine, the researchers found that ganciclovir treatment caused the tumors to shrink.

Gambhir is also planning a study that will use reporter gene imaging to follow gene therapy for rheumatoid arthritis.

Researchers at Henry Ford Health System in Detroit plan to employ imaging in gene therapy for prostate cancer. Their study will use an adenovirus to deliver a gene construct that includes the suicide genes cytosine deaminase and HSV1-tk, which leave cancerous cells more vulnerable to the effects of radiation therapy. The gene construct, also carrying the reporter gene NIS, will be injected into the prostate.

"It works great in dogs, but does it work in people?" asked Steve Brown, director of preclinical studies and a staff scientist in the department of radiation oncology at Henry Ford.

At the University of Alabama at Birmingham, ovarian cancer and head and neck cancer will be the targets for two human clinical trials slated to begin this year, said Kurt R. Zinn, Ph.D., codirector of molecular imaging. The studies will use adenoviruses, modified to increase their infectivity, to deliver the HSV1-tk suicide gene to kill ovarian tumor cells or the cytosine deaminase gene to fight head and neck cancer. Both studies, led by Dr. David Curiel, gene therapy director, and Dr. Ronald Alvarez, head of gynecologic oncology, will track gene expression using the somatostatin receptor hSSTr2 reporter gene.

"This is an evolutionary process," Zinn said. "We make improvements, and then we learn something else and improve it more."

While positive results are inspiring researchers, they are also facing a barrage of obstacles. Most challenging is the interference of the immune system, which is programmed to neutralize foreign particles such as adenoviruses, other viral vectors, nonviral vectors, therapeutic and reporter proteins, and gene constructs and transcripts.

In Gambhir's trial, the expression of HSV1-tk lasted only a few weeks. A second injection of the gene construct, which was packaged into the same adenovirus, failed to result in the type of robust gene expression previously seen.

"It's probably because of the immune system's response to the virus," Gambhir said. "In cancer, it's a catch-22: You want long-term gene expression and a good immune response, but it's hard to get both."

Researchers, however, would be hard-pressed to give up on viral vectors.

"Viruses will be the vehicles that we'll be using for the near future," Herschman said. "Viruses are naturally built to target cells and bring in information. At least for now, we're not going to be better than nature at designing cell-targeting agents."

Gambhir and many others are studying lentiviruses, which can integrate into the cellular genome and provide gene expression that lasts for months. The long-term expression opens the possibility of creating an effective therapy with just a single administration of virus. As members of the retrovirus family, however, lentiviruses will be intensely scrutinized before they are declared safe for human trials. Gambhir is working with lentiviruses containing a reporter gene that is exclusively expressed in the prostate.

"I think the outlook is positive as we prove their safety more and more, and as we get more experience," he said.

Another promising strategy may save investigators from having to wrestle with viruses. If the patient's own cells are used to deliver therapeutic genes, the immune system won't see the delivery vehicle as a foreign object. And by inserting a reporter gene, cellular immunotherapists can finally gather the information they need to improve the protocols.

"There are so many unknowns in this process," Piwnica-Worms said. "The opportunity to directly visualize and assess the migration of the cells in animal models is where imaging reporter genes are having a tremendous impact."

That's what oncologists at the City of Hope National Medical Center are counting on. The institution in Duarte, CA, is collaborating with Gambhir to treat glioblastoma with the aid of reporter gene imaging. Researchers will take patients' T cells and transduce them with the gene encoding the particular antigen on the tumor and a reporter gene.

"You harvest your own cells, teach them how to do a better job of killing the tumor, and reinject them, and they kill the tumor," Gambhir said. "So now when we have the cells ex vivo, we will also inject a reporter gene. We're marrying the best of both worlds."

Cardiologists are also embracing reporter genes in their study and treatment of heart disease.

Over the years, efforts to stimulate new blood vessel growth in the heart by administering the gene for vascular endothelial growth factor (VEGF) have been hampered by the lack of a reliable way to monitor expression of VEGF, said Dr. Joseph C. Wu, a cardiologist at the MIPS. To solve the problem, Wu has been working on ways to introduce molecular imaging. He recently completed a study in which he injected rats with adenoviruses carrying both VEGF and the reporter gene HSV1-tk. While PET imaging allowed Wu to monitor and quantify the activity of VEGF, such expression was limited to two weeks because of the immune system's response to the adenovirus.

To circumvent this problem, Wu has begun using the rat's own embryonic stem cells to deliver VEGF along with a reporter gene. His studies in rats have laid the groundwork for monitoring the activity of transplanted cells in rats and understanding their survival patterns. Wu transduced both firefly luciferase and HSV1-tk reporter genes into rat embryonic cardiomyoblasts, which he then injected into the rats' hearts. Optical imaging and microPET revealed that the cells remained in the heart myocardium and expressed both reporter genes. But because these cells lasted for only two weeks, Wu's lab is investigating ways to prolong cell survival.

"The advantage of molecular imaging is that you can transduce these cells with multiple reporter genes," he said. "With optical bioluminescence, you can do high-throughput studies with small animals and figure out which cell types are best in vivo, and then translate it into clinical applications using PET."

Gene and cell therapists face the additional challenge of gaining acceptance from patients and the FDA. Researchers in the field are constantly guarding against the potential for gene constructs to travel to sites outside the target area and to integrate into the genome of stem cells, Brown said.

Still, the FDA's regulatory hurdles can be overwhelming, Conti said. Approaching the approval of therapeutic gene constructs and their respective probes as drugs, the federal agency has slowed the path to human trials. Unlike pharmaceutical companies, universities, which are usually the source of such innovative agents, don't have the wherewithal to complete the tedious obstacle course of FDA approval.

But as the protocols mature and professional societies push the issue, the path may widen, said Conti, who is president-elect of the Society of Nuclear Medicine.

"There's momentum in the right direction in reducing the regulatory requirements," he said. "Within a year or two, there will be a much clearer and well-defined path, and people will move quickly into human studies."

That doesn't mean that pharmacies will be stocking gene or cell therapy kits any time soon, however.

"In reality, we're a long way from thinking about the routine use of diagnostic imaging genes to look at function or for therapy," Piwnica-Worms said. "While promising work is under way for human application, it's way in the future."

Laura Lane

Articles

Educators debate ways to integrate MI curricula

Molecular Imaging Briefing

MI addresses conundrum of investigational drug trials

Molecular Imaging Briefing

MI addresses conundrum of investigational drug trials

Focused ultrasound helps exogenous genes permeate targeted cells' outer membrane

Two-step amplification technique delivers gene therapy to refractory prostate cancer

Two-step amplification technique delivers gene therapy to refractory prostate cancer

Focused ultrasound helps exogenous genes permeate targeted cells' outer membrane

Reporter gene imaging illustrates molecular events