Sanjiv Sam Gambhir is the director of the Molecular Imaging Program and the Canary Center for Cancer Early Detection, both at Stanford University. GT's Jeanene Swanson spoke with him about how molecular imaging is changing the way disease is detected and treated. What follows are excerpts of that conversation, edited for space.
Genome Technology: What role does imaging play in early disease detection?
Sanjiv Sam Gambhir: We feel that molecular imaging will have a significant role in the early detection of cancer. Right now, we see it as the three I's: identify, isolate, and intervene. The 'identify' is a low-cost screening test that involves blood biomarkers or urine biomarkers. 'Isolate' is where imaging comes in, to isolate where in the body that disease is, if, in fact, the blood test was correct. Then you go to the third 'I,' which is to 'intervene.' You do it at such an early stage you're able to much more effectively treat the disease.
GT: What new techniques are you working on?
SSG: We're developing several new strategies that are designed to detect much smaller disease foci than what is currently possible. One strategy is using ultrasound. You can inject into the body bubbles and the bubbles are targeted to blood vessels growing in early cancer. The ultrasound is imaging the location molecularly of new blood vessels that are feeding the cancer.
Another strategy we're developing is based on photoacoustic imaging. The idea here is that light goes in, interacts with imaging agents, and produces sound. The advantage of that is you now can develop imaging agents that can leave blood vessels and go and interact with molecular targets on the surface of cancer cells.
A third technique is Raman molecular imaging. In Raman imaging, light goes to a target tumor site, interacts with an imaging agent that scatters the light, and we detect the Raman scatter. And that ends up being very sensitive. We can take advantage of systems biology in that we can do multiplexing, we can target many different molecular events and see a separate signal from each one. One of the problems with molecular imaging is it's not highly multiplexed — you can interrogate one thing at a time, maybe a couple things at a time. With Raman, we can interrogate 20 [or] 30 things simultaneously. In Raman, light gets scattered inelastically. Inelastic light scattering can allow you to interrogate different properties of the tumor, for example.
The hope is that by syncing the low-cost screening through blood testing to highly sensitive, highly multiplexable molecular imaging strategies, one can really make a significant shift in earlier detection of disease.
GT: How does molecular imaging make use of biomarker research?
SSG: Molecular imaging is a lot like the field of drug development; that is, you have to start with a molecular target. In order to build imaging agents specific for that target, you [have] got to know what target to go after. Now, we don't need the target to be able to control the cell because we just want to detect it — not treat it — but sometimes the targets are the same.
How do we go after different molecular targets? We do our own biomarker discovery: we look for, in tissues, what biomarkers are on the surface of cells [or] within cells that make them unique from normal cells. We sometimes leverage off of biomarkers being discovered by the pharmaceutical industry. In each case, the idea is that we're able to find biomarkers that we can build imaging agents towards, and a lot of the same techniques — proteomics for discovering biomarkers in the blood — have been also used for finding biomarkers in tissue.
Systems biology and other proteomics techniques are helping us also to understand which biomarkers we should target from an imaging perspective.