New Optical Glucose Sensor

Continued from page 1

By Prachi Patel

Thursday, December 28, 2006

Meanwhile, Gerard Cote, a biomedical-engineering professor at Texas A&M University who first developed the chemical mechanism that Loeb's optical-fiber sensor employs, is attempting to make a similar glucose sensor using a slightly different design. Cote and Michael Pishko, a chemical engineer at Penn State, are developing 20-micrometer-wide polyethylene-glycol beads that could be implanted just under a person's skin, like a tattoo. A light-emitting diode on a wristwatch-type analyzer would shine light on the beads and measure the fluorescence. "The beads would be totally implantable, and the skin would heal over it," Cote says, so there is nothing "that penetrates the skin and opens the body to infection."

Current implanted glucose sensors require inserting an electrode under the skin; a metal wire connects the electrode to a monitoring device. The electrode has to be held in place with an adhesive tape. However, such devices can quickly cause infection and must be replaced every three days, which can get expensive. "They are looked upon as a foreign body, a splinter, and the body tries to reject them," says George Wilson, a chemistry professor involved in biosensor research at the University of Kansas.

Loeb says that it should be possible to leave his glucose-sensing fiber inside human skin for a few months. He has implanted his fiber in pigs for up to three months without causing an inflammation.

So far, Loeb has tested his fiber-optic system in glucose solutions and has found that it is sensitive to the range of glucose concentrations that are found in the human body. But the system's response inside living animals will be crucial to letting researchers know if it works. "There is an enormous difference between a sensor that operates in solution and a sensor that operates in animals," Wilson says.

Once the optical glucose sensor is tested in animals, it would need to go through clinical trials, which can take four to five years. If everything goes as planned, both Loeb and Cote believe that their sensors could be available as products in five to ten years.