Gough's team also had to deal with the formation of scar tissue around the device. Scar tissue reduces permeability of glucose and oxygen and can interfere with detection over time. Researchers solved this by adding a reference sensor--a platinum wire oxygen sensor without the enzyme. This sensor allows the device to correct for changes in permeability.

"I was impressed that the device operated continuously [in pig experiments] as long as it did--almost two years--and by the fact that it had fairly stable function," says Steven Russell, a researcher and physician at Massachusetts General Hospital, who was not involved in the study. "One concern I had was accuracy. They reported it is as good as some subcutaneous sensors out there, which is true, but it's not as good as the best of them."

But commercializing the technology may prove difficult. "I think this is significant progress and an important proof of principle, but there are still substantial challenges to overcome," says Roman Hovroka, a researcher at the University of Cambridge, who was not involved in the research. "It's still a long journey to a commercial product."

He points out that because the studies take a year or two, commercial development and clinical testing can be very expensive. Other companies developing continuous-sensing technologies have shifted efforts from long-term implanted devices to short-term technologies, says Hovroka, possibly because of these high development and testing costs.

Still, Hovroka and others say, a device capable of functioning over the long term is an important goal. Studies have shown that the frequency with which patients monitor their blood sugar affects how well they can control it, which in turn is tied to long-term health. "So presumably an implantable sensor would have a bigger impact on outcomes than a sensor that has to be replaced often," says Hovroka. "But that needs to be offset with the need to perform local surgery on a yearly basis and the reliability of such devices."