This is the second of a two-part series on cancer imaging. This week's article discusses how new contrast agents can better evaluate cancer processes in animal models.

Chemistry is king in the land of molecular imaging, says Dr. Peter Choyke, chief of the Molecular Imaging Program (MIP) in NCI's CCR. MIP's mission is to perfect and demonstrate a concept that could revolutionize diagnostic imaging.

This strategy delivers a molecular "beacon" to the outside of a cell to guide and enhance more precise imaging, Dr. Choyke explains. The concept is similar to the "Trojan horse" strategy that delivers chemotherapeutic agents, loaded inside immunoliposomes, to specific cancer cells (NCI Cancer Bulletin, April 12).

Targeted imaging probes are now being tested in animal models, Dr. Choyke notes. For example, by labeling the cancer drug trastuzumab, researchers can monitor it as it binds to the HER-2 protein receptor site. All tumors expressing the HER-2 protein can be imaged and emphasized with this imaging agent. The underlying principle may also apply to other cancers if binding sites can be found that distinguish those particular cells, he says. This system might not only dramatically improve diagnostic imaging of tumors, but also help monitor for recurrence and deliver therapeutic agents.

In the past year, Dr. Choyke's team has assembled a chemical arsenal of different nanoparticles, each designed to find binding sites on the surface of specific target cells. The key is to find a biological receptor that will distinguish a particular kind of cell - often a cancer cell in a particular organ - to the exclusion of all others, he says. Once delivered and locked onto their target, these designer molecules provide a beacon to image the cell's location.

Small groups of MIP scientists specialize in a particular imaging method, including magnetic resonance imaging (MRI), usually enhanced with a contrast agent; positron emission tomography (PET), which picks up positron-emitting radionuclides and can be combined with computed tomography (CT); and optical imaging.

The first stage in the development of an imaging agent involves synthesizing the agent to target a particular tumor or process, and then tweaking the imaging instruments to take advantage of it, Dr. Choyke explains. Then the agent can move to preclinical animal studies and to early-phase human clinical studies, he adds.

Dr. Martin Brechbiel, chief of NCI's Radioimmune and Inorganic Chemistry Section (RICS), and his team work closely with MIP. "The work of our labs is totally integrated," says Dr. Choyke, of the relationship. "There's a true synergy between RICS' chemistry and MIP's knowledge of imaging technology. Dr. Brechbiel's resources were already on the ground and primed, which has enabled us to develop and begin testing a number of compounds fairly quickly."

The innovation that distinguishes these programs' nanoparticle constructs is the chelate, or linking compound. "The delivery vector ligand binds to one location on the chelate," says Dr. Brechbiel. "On the other side, you chemically attach the imaging beacon you want to deliver to that particular target cell, wherever it might be in the body." The researchers must balance many characteristics, preserving the binding accuracy while enhancing the reliability and versatility of the delivery vector.

After an agent is produced and the imaging instrumentation is optimized, preclinical studies in mice allow a variety of cancer processes to be evaluated, says Dr. Choyke. For instance, MIP has projects looking at tumor angiogenesis and lymphangiogenesis, the latter of which, he adds, "is an understudied aspect of cancer metastasis because there has been no good way to image the lymphatics."

MIP's ultimate goal is to move these new imaging agents into clinical trials. "FDA's investigational new drug process represents an enormous hurdle," says Dr. Choyke. Even the largest drug companies are wary of the costs involved in developing imaging agents, estimated at about $250 million per agent. "But we're getting their interest with a tantalizing possibility," he says. "If you could use our system to help preselect which patients are more likely to respond to a treatment, you could conceivably run smaller trials with larger response rates."

By Addison Greenwood