The clean blood and watery filtrate are both shunted into the other half of the system: a separate cartridge. Here, they flow over more silicon membranes, these ones coated with a single type of human kidney cell, which helps the device reabsorb some of the water, sugars, and salts, as well as produce vitamin D and help prevent blood pressure from sinking too low--normal kidney functions that are not offered by dialysis. The waste that's not reabsorbed is shunted to a tube attached to the bladder and removed as waste in the urine--just like a normal kidney would do.

It's far from a complete system, and the researchers note that they don't ever expect it to replace kidney transplants. "Your kidney has 20 to 30 cell types in it, all of which accomplish different functions. But we'd like to overcome a critical issue that's emerged in renal failure," says Fissell. "If you're listed for a kidney transplant, you're far more likely to die on the waitlist than you are to get a kidney." He says the device could act as a bridge for patients awaiting transplant.

"From a general perspective, any implantable device would sharply reduce the burden that patients now experience," says Glenn Chertow, the chief of nephrology at the Stanford University School of Medicine. "And if some of the additional magic that a native kidney provides could be added to an implantable device, we could come closer to a restoration of good health."

The researchers have already worked out some of the more difficult issues: Humes has worked out how to culture kidney cells on the necessary scale (he can culture enough cells for 100,000 devices from a single kidney). He's also determined the best way to freeze them for future use. And Roy, a TR35 winner in 2003, has tested the implant in a dozen rats and a handful of pigs. They still have to scale up the implant's efficiency to something that could work effectively in humans, but they hope to start human trials in five-to-seven years. Right now the biocartridge can filter between 30 to 35 liters of water per day, and it needs to be able to filter at least 43. They also have to find ways to ensure that the devices don't cause blood clots or immune reactions.

Other groups are also working toward alternatives to thrice-weekly dialysis appointments, although most are concentrating on wearable dialysis devices--a difficult proposition in itself, given the challenge of constant filtration at such large volumes without an external pump. One such device is already in the second stage of clinical trials. But even constant dialysis can't take the place of the kidney's other functions.

Allen Nissenson, CMO of DaVita, one of the country's largest dialysis provider, says the implantable concept holds appeal. "It's a bioreactor kidney, an incredibly innovative concept, and really exciting if it proves to be workable on a larger scale," he says.