The next revolution in radiology: personalized cancer treatment

Robert Canfield

David Dick, PhD, demonstrates the new cyclotron, which creates radioactive isotopes used for imaging.

When diagnosing cancer, assistant professor of radiology Andrew Quon, MD, would like to know as much as possible about the tumor he's trying to describe—down to the very molecules that distinguish it from other types of cancer—so that he can help plan the most effective attack against it.

For years, cancer therapy has been indiscriminate warfare: "a relatively crude approach," Quon explained, "using chemotherapy that's toxic to both normal tissue and cancer and then waiting to see what lives and what dies." To fine-tune the treatment, Quon and colleagues in the division of nuclear medicine are using a set of techniques, collectively known as molecular imaging, that involve radioactive indicator molecules that can be imaged by positron emission tomography, or PET.

The group is particularly interested in identifying which breast cancers are likely to respond to hormone therapy. To do this, they use a radioactive form of estrogen to indicate which cancer cells have receptors for the hormone estrogen on their surfaces. If a cancer lights up during a PET scan, it has radioactive estrogen latched on to its receptors, making it a good candidate for estrogen therapy. If a cancer doesn't light up, it is not producing many estrogen receptors and hormone therapy would not combat it.

"It has long been known that different cancers behave differently and respond to treatment differently," said Quon. "Now we are starting to be able to image and treat tumors individually, not just by type."

Welcome to the brand new world of radiology.

Quon is one of a host of faculty members who are looking to make personal medicine a reality, and it's a focal point of the event Saturday celebrating a century of radiology and radiation oncology at Stanford.

"For us involved with molecular imaging," said Quon, "personalized medicine means going far beyond diagnosing cancer simply based on tumor location and gross physical appearance to attempting to tailor individual therapies based on what we find out about each tumor on a molecular level, while the tumor is still within the body rather than under a microscope."

At the start of the 20th century, doctors used X-rays to see what was going on inside the body for the first time. They also quickly realized that X-ray radiation could destroy diseases. With advances in technology, the images became more precise and the treatments more effective, but doctors still can't distinguish all the variations of a disease that might exist.

That's where molecular imaging comes in. "Every disease has a molecular signature," said Christopher Contag, PhD, associate professor of radiology and of microbiology and immunology. "Personalized medicine is the idea of diagnosing and typing a problem using these molecular markers."

Contag co-directs the Molecular Imaging Program at Stanford, which aims to use the molecular signatures of disease as a means to developing "molecular detectives." These snoops are injected into the body to home in and identify a specific problem—for example, a protein that a cell produces once it turns cancerous. The detectives then report what they find to sensors located outside the body.

The promise of molecular imaging is that disease-specific beacons can sound an alarm when only a few suspect cells are present rather than the billions needed by current methods, said Gary Glazer, MD, professor and chair of radiology.

Beyond detection, the targeted molecules could become vehicles for delivering therapeutic medications to malignant cells, said Sanjiv Sam Gambhir, MD, PhD, the director of the molecular imaging program and the head of nuclear medicine.

"We really aren't treating individuals yet, we are treating with therapies tailored for a population," said Gambhir, who is also a professor of radiology. "By having the next generation of therapies customized for a given individual's genetic makeup we have the opportunity to truly move towards personalized medicine."

Gambhir, who will be speaking about personalized medicine and imaging at the Saturday symposium, said, "Personalized medicine means that medical care will no longer have to be suboptimal."

Although there is a long way to go in the field, making the detection and eventually treatment of cancer personal is already on its way. Current molecular imaging methods combine a detectable radioactive beacon with a disease-indicating molecule. Once these tracers are injected into the body, physicians use a PET scanner to reveal the location of any abnormalities.

The most widely used molecular imaging tool today is called FDG, radioactive glucose, which can disclose areas that are using more glucose than normal—often a sign of cancer. "We will likely have many, many more such drugs in the next decade," said Gambhir. "There will be no disease process that will not be helped with personalized medicine."

Stanford now has the ability to produce molecular imaging agents like FDG on demand thanks to a new cyclotron located in the recently opened Lucas Expansion building. Fired up just last week, the cyclotron is a machine that creates the radioactive isotopes which are coupled to molecules to form molecular imaging agents sometimes called tracers.

Having a cyclotron available on-site will also allow scientists here to create completely new compounds, according to David Dick, PhD, the head of cyclotron physics. Especially exciting, he said, would be radioactive isotopes that are extremely short-lived; such isotopes could be created in the lab and then rushed to patients in the hospital in a matter of minutes.

No matter what treatment is ultimately used, the earlier a problem is caught, the greater the chance that any treatment has a chance of working. Glazer, who will be speaking at the symposium on the future of medical imaging, said, "Until something fundamentally changes with cancer, the best hope is earlier detection."

Amato Giaccia, PhD, professor of radiation oncology, will be speaking at the symposium on the future of targeted radiotherapy. "The key to personalizing treatment is the ability to visualize," he said, or to have some kind of assay that gives answers individually rather than stating broad probability ranges.

But seeing an individual's problem is only half of the battle; it doesn't help a patient much if there isn't something that can be done that makes the visualized disease go away.

Giaccia, who directs the Division of Radiation and Cancer Biology, summarized the problem for cancer: the doctors need to see it accurately, especially any metastases. Then they need to hit it accurately with deadly force, but at the same time protect the normal tissues of the body.

The marriage of radiology and radiation oncology in molecular imaging, said Giaccia, will allow the doctors to spot cancer cells—even far-flung tiny metastases—and to target those cells with a highly focused lethal dose of radiation. One upcoming technology he mentioned was quantum dots, particles on the nanometer (a billionth of a meter) scale that could potentially not only seek out specific targets on cancer, but also carry a toxic radioactive payload. Such "real-time radiotherapy" would allow treatment to be visualized as it is happening.

Molecular imaging is a great way to pair diagnosis and therapy, said Contag, which is especially useful in cancer because of the many different molecular mechanisms that can cause this disease. If scientists could identify specific targets, they could try to "hit" those spots and then monitor whether the targeted therapy had an effect on the tumor.

"Therapy and diagnostics are two sides of the same coin," said Gambhir. "In order to individualize one, you have to individualize the other. Without having imaging, therapy optimization is not fully possible. The two go hand-in-hand towards making personalized medicine a reality."

Of course all of the scenarios envisioning individually targeted cancers are completely dependent on identifying the targets—what the informative genetic markers are.

"The whole idea of personalized medicine is that it is personal, and there can be an enormous amount of variation in the diseases that each patient may have," said Quon. "So lots and lots of specific probes will be needed to find all the markers.

"The irony is that we have to take a shotgun approach to discovery of these probes so that we don't have to take a shotgun approach to therapy."