"With nanoparticles you're working with materials that are small enough to get across mucosal barriers, across vascular pores, and actually find things like tumor cells and get into them and change them directly," says Baker.

The dendrimers' small size also means that once they've done their work and the cancer cells break down, they can be safely removed from the bloodstream by the kidneys.

Despite their current success, though, Baker's method has drawbacks. For one, attaching many different molecules to one dendrimer is difficult, and, once completed, such a nanodevice will deliver only one drug to one kind of cancer cell.

To simplify manufacture, and add flexibility to the device, Baker is developing a way to make multiple dendrimers stick to each other, using complementary strands of DNA that act similar to the hooks and loops in Velcro. Each dendrimer would have just one function, such as drug delivery or targeting. Baker could then select dendrimers with the functions he needs and combine them in a solution.

At first, he will link just two dendrimers, but eventually group together several, making it possible to deliver multiple drugs along with fluorescing molecules.

A future application could involve tailoring drug combinations for specific people, based on observations of the uptake and effectiveness of each drug.

Some tumors have multiple types of cancer cells that require different drugs. The dendrimers could be used to detect which types of cells are present, and deliver the required drugs. They could also potentially track how individual cells respond to the drugs, allowing doctors to modify treatments.

Other groups are also working on multipurpose platforms for detecting and treating cancer. However, they have yet to show their devices can get out of the bloodstream and later out of the body. A group led by Mostafa El-Sayed of the Georgia Institute of Technology and his son Ivan El-Sayed of the University of California, San Francisco, has attached antibodies to gold nanoparticles, allowing the particles to selectively bind to certain cancer cells. Then light reflected from these particles reveals their location, and light absorbed by them causes the particles to heat up, destroying the cells they inhabit, but leaving healthy cells unharmed.

So far, the group has demonstrated its technology only in cell cultures. Although their particles are about eight times the size of Baker's molecules, Ivan El-Sayed is hopeful that they can be delivered through the bloodstream. Eliminating them from the body is another matter, he admits -- it may turn out that the particles can only be used in small doses to avoid a buildup of the gold.

Several other groups are also developing nanotech devices for cancer detection and drug delivery, including ones at MIT, Harvard, Virginia Commonwealth University, University of Missouri, University of Washington, and Rice University. "I think there's going to be more and more involvement," says Tomalia of Dendritic NanoTechnologies, because combining detection and treatment using nanotechnology "is a very appealing strategy."

What's more, delivering chemotherapy drugs using nanodevices is only the beginning, according to Baker. Researchers in his lab are also designing molecules that could act as artificial ion channels -- tunnels through the cell membrane that could replace defective channels in diseases such as cystic fibrosis.

"The idea [with nanodevices] is to take advantage of the size of this material, get it into cells, and literally fix them," says Baker.