An important way of determining whether a tumor has metastasized is to detect cancer cells that circulate in the bloodstream after breaking free from the original tumor. But at most only a few of these cells are present among the tens of billions of normal cells in a one-­milliliter sample of blood. An MIT aeronautical engineer and a Harvard bioengineer have now created a device that can isolate them with unprecedented sensitivity.

The new device builds on an earlier version developed by Mehmet Toner, SM ‘85, PhD ‘89, a professor of biomedical engineering at Harvard Medical School. In his original setup, blood taken from a patient flowed past tens of thousands of tiny silicon posts coated with antibodies that trapped tumor cells. The problem was that some cells never encountered the posts at all, allowing them to escape.

Toner thought if the posts were porous instead of solid, blood could flow right through them, making it more likely that cancer cells would stick. To achieve that, he enlisted the help of Brian Wardle, SM ‘95, PhD ‘98, an MIT associate professor of aeronautics and astronautics and an expert in designing nano-engineered advanced composite materials to make stronger aircraft parts.

The result is a dime-size microfluidic device that collects cancer cells eight times better than the original version. The secret is carbon nanotubes—tiny, hollow cylinders whose walls are lattices of carbon atoms.

The researchers replaced the silicon posts in the original device with clusters containing 10 billion to 100 billion carbon nanotubes per square centimeter. The clusters are less than 1 percent carbon and 99 percent air, leaving plenty of space for fluid to flow through.

As in the original device, the surface of each tube can be coated with antibodies specific to cancer cells. However, because the fluid can go through the nanotube clusters as well as around them, the target cells or particles are much more likely to get caught.

The device, which could also be adapted to detect viruses such as HIV, could become the basis of low-cost diagnostic equipment for use in developing countries where expensive alternatives are hard to come by, says Toner.

The researchers are now tailoring the device for HIV diagnosis. Toner’s original cancer-cell detector is being tested in several hospitals; the nanotube version will probably need a few more years of development before it’s ready for human testing.