A Way to Spot Cancer Early

Protein pull: A prototype scanner (top) detects cancer-specific proteins present in low concentrations in the blood by capturing them on magnetic sensors and tagging them with magnetic nanoparticles. The heart of the scanner is a silicon chip arrayed with magnetic sensors called spin valves (below). Credit: Sebastian Osterfeld (top); PNAS (bottom)

A new system for detecting cancer proteins uses the same magnetic phenomenon that lets computer hard drives read and write data. The Stanford University researchers developing the system hope that it will detect cancer in its earlier stages, when it's easier to treat. MagArray, a startup in Sunnyvale, CA, will commercialize the technology.

Well before cancers are visible on medical imaging scans, their cells release small amounts of telltale proteins into the blood. Researchers are developing ways to detect those proteins, frequently by tagging them with fluorescent labels. But while all biological samples have some background fluorescence, they have virtually no magnetic background. Magnetic protein detection could thus yield a clearer signal, says Shan Wang, a professor of materials science and engineering and electrical engineering at Stanford University.

Another approach to early cancer detection involves devices that catch cancer proteins on the tips of vibrating nanostructures and measure how they affect the flow of electrical current. But since the Stanford device exploits a physical phenomenon that is already the basis for consumer electronics, it could prove easier to mass-produce. "This is one of the things that will make this technology a success: there's no need to prove manufacturability," Wang says. "The challenge is to combine it with biochemistry."

Wang's device takes advantage of giant magnetoresistance, a phenomenon that won its discoverers the 2007 Nobel Prize in physics. The device is built on a silicon chip arrayed with 64 magnetic sensors called spin valves. Each valve is coated with a different kind of antibody, a molecule primed to latch on to a particular cancer protein. When the chip is exposed to blood serum, the target proteins stick to the antibodies. Wang then adds a solution of magnetic nanoparticles, also attached to antibodies, that stick to the captured proteins. The magnetic field of the captured nanoparticles measurably changes the resistance of the underlying spin valve, allowing Wang to determine the concentration of cancer proteins in the serum.

In tests where the Stanford prototype scanned for cancer proteins, including a marker of colon cancer, it was two orders of magnitude more sensitive than the standard technique for detecting blood proteins, which uses a similar antibody capture sandwich in combination with fluorescent tags.

The idea of using magnetic sensing for biomarker detection originated with David Baselt, a researcher at the U.S. Naval Research Laboratory. But the Stanford group's use of magnetic nanoparticles "clearly speeds up the process," says David Walt, a chemistry professor at Tufts University. In the Proceedings of the National Academy of Sciences, Wang's group describes the detection of very low levels of seven cancer markers in serum in as few as 30 minutes. "The results with serum suggest the method has promise," Walt says.

Electronics giant Philips also plans to commercialize a handheld device that uses magnetic nanoparticles next year. The device captures molecules in saliva that indicate drug use and uses magnets to bring them to a simple imager.

But Marc Porter, a chemical-engineering professor at the University of Utah, is confident that magnetoresistance will be an important tool for diagnosing complex diseases like cancer and heart disease, where more sensitive readings of multiple proteins, not just one, will lead to better diagnoses. Along with Utah researcher Michael Granger, Porter is developing a biomarker scanner that works more like a computer hard drive than does Wang's, scanning a magnetoresistant head over magnetically labeled biological samples. Porter and Granger also plan to start a company to get their scanner on the market.

Wang adds that magnetic scanners should be much less expensive than standard biomarker scanners. The instrument that reads the output of Wang's chip is smaller than the optical systems required to read fluorescent signals, and it will probably cost less than $10,000. Other researchers working on biomarker-detection systems that use microfluidics are aiming for even smaller and cheaper systems that can go out into the field. But Wang says that his system will integrate well with existing hospital infrastructure.

Wang says that MagArray is writing up an application to the U.S. Food and Drug Administration to do clinical trials of the cancer scanner in order to compare the blood-protein levels of healthy people and those with cancer. "At this point, the field of biomarkers is still under development," says Wang. "What the relative abundance of biomarkers means is not yet clear."