Silicon Isn't Just for Computers. It Can Make a Pretty Good Kidney, Too

The world needs more kidneys. Here's how researchers are making artificial ones—in silicon.
Image may contain Green
Getty Images

Every week, two million people across the world will sit for hours, hooked up to a whirring, blinking, blood-cleaning dialysis machine. Their alternatives: Find a kidney transplant or die.

In the US, dialysis is a roughly 40-billion-dollar business keeping 468,000 people with end-stage renal disease alive. The process is far from perfect, but that hasn't hindered the industry's growth. That's thanks to a federally mandated Medicare entitlement that guarantees any American who needs dialysis—regardless of age or financial status—can get it, and get it paid for.

The legally enshrined coverage of dialysis has doubtlessly saved thousands of lives since its enactment 45 years ago, but the procedure’s history of special treatment has also stymied innovation. Today, the US government spends about 50 times more on private dialysis companies than it does on kidney disease research to improve treatments and find new cures. In this funding atmosphere, scientists have made slow progress to come up with something better than the dialysis machine-filled storefronts and strip malls that provide a vital service to so many of the country's sickest people.

Now, after more than 20 years of work, one team of doctors and researchers is close to offering patients an implantable artificial kidney, a bionic device that uses the same technology that makes the chips that power your laptop and smartphone. Stacks of carefully designed silicon nanopore filters combine with live kidney cells grown in a bioreactor. The bundle is enclosed in a body-friendly box and connected to a patient’s circulatory system and bladder—no external tubing required.

The device would do more than detach dialysis patients—who experience much higher rates of fatigue, chronic pain, and depression than the average American—from a grueling treatment schedule. It would also address a critical shortfall of organs for transplant that continues despite a recent uptick in donations. For every person who received a kidney last year, 5 more on the waiting list didn’t. And 4,000 of them died.

There are still plenty of regulatory hurdles ahead—human testing is scheduled to begin early next year1—but this bioartificial kidney is already bringing hope to patients desperate to unhook for good.

Innovation, Interrupted

Kidneys are the body’s bookkeepers. They sort the good from the bad—a process crucial to maintaining a stable balance of bodily chemicals. But sometimes they stop working. Diabetes, high blood pressure, and some forms of cancers can all cause kidney damage and impair the organs' ability to function. Which is why doctors have long been on the lookout for ways to mimic their operations outside the body.

The first successful attempt at a human artificial kidney was a feat of Rube Goldberg-ian ingenuity, necessitated in large part by wartime austerity measures. In the spring of 1940, a young Dutch doctor named Willem Kolff decamped from his university post to wait out the Nazi occupation of the Netherlands in a rural hospital on the IJssel river. There he constructed an unwieldy contraption for treating people dying from kidney failure using some 50 yards of sausage casing, a rotating wooden drum, and a bath of saltwater. The semi-permeable casing filtered out small molecules of toxic kidney waste while keeping larger blood cells and other molecules intact. Kolff's apparatus enabled him to draw blood from his patients, push it through the 150 feet of submerged casings, and return it to them cleansed of deadly impurities.

In some ways, dialysis has advanced quite a bit since 1943. (Vaarwel, sausage casing, hello mass-produced cellulose tubing.) But its basic function has remained unchanged for more than 70 years.

Not because there aren’t plenty of things to improve on. Design and manufacturing flaws make dialysis much less efficient than a real kidney at taking bad stuff out of the body and keeping the good stuff in. Other biological functions it can’t duplicate at all. But any efforts to substantially upgrade (or, heaven forbid, supplant) the technology has been undercut by a political promise made four and a half decades ago with unforeseen economic repercussions.

In the 1960s, when dialysis started gaining traction among doctors treating chronic kidney failure, most patients couldn't afford its $30,000 price tag—and it wasn’t covered by insurance. This led to treatment rationing and the arrival of death panels to the American consciousness. In 1972, Richard Nixon signed a government mandate to pay for dialysis for anyone who needed it. At the time, the moral cost of failing to provide lifesaving care was deemed greater than the financial setback of doing so.

But the government accountants, unable to see the country’s coming obesity epidemic and all its attendant health problems, greatly underestimated the future need of the nation. In the decades since, the number of patients requiring dialysis has increased fiftyfold. Today the federal government spends as much on treating kidney disease—nearly $31 billion per year—as it does on the entire annual budget for the National Institutes of Health. The NIH devotes $574 million of its funding to kidney disease research to improve therapies and discover cures. It represents just 1.7 percent of the annual total cost of care for the condition.

But Shuvo Roy, a professor in the School of Pharmacy at UC San Francisco, didn’t know any of this back in the late 1990s when he was studying how to apply his electrical engineering chops to medical devices. Fresh off his PhD and starting a new job at the Cleveland Clinic, Roy was a hammer looking for interesting problems to solve. Cardiology and neurosurgery seemed like exciting, well-funded places to do that. So he started working on cardiac ultrasound. But one day, a few months in, an internal medicine resident at nearby Case Western Reserve University named William Fissell came up to Roy and asked: “Have you ever thought about working on the kidney?”

Roy hadn’t. But the more Fissell told him about how stagnant the field of kidney research had been, how ripe dialysis was for a technological overhaul, the more interested he got. And as he familiarized himself with the machines and the engineering behind them, Roy began to realize the extent of dialysis' limitations—and the potential for innovation.

Limitations like the pore-size problem. Dialysis does a decent job cleansing blood of waste products, but it also filters out good stuff: salts, sugars, amino acids. Blame the polymer manufacturing process, which can’t replicate the 7-nanometer precision of nephrons—the kidney's natural filters. Making dialysis membranes involves a process called extrusion, which yields a distribution of pore sizes—most are about 7nm but you also get some portion that are much smaller, some that are much larger, and everything in between. This is a problem because that means some of the bad stuff (like urea and excess salts) can sneak through and some of the good stuff (necessary blood sugars and amino acids) gets trapped. Seven nanometers is the size of albumin—a critical protein that keeps fluid from leaking out of blood vessels, nourishes tissues, and transports hormones, vitamins, drugs, and substances like calcium throughout the body. Taking too much of it out of the bloodstream would be a bad thing. And when it comes to the kidney’s other natural functions, like secreting hormones that regulate blood pressure, dialysis can’t do them at all. Only living cells can.

“We were talking about making a better Bandaid,” Roy says. But as he and Fissell looked around them at the advances being made in live tissue engineering, they started thinking beyond a better, smaller, faster filter. “We thought, if people are growing ears on the backs of mice, why can’t we grow a kidney?”

It turned out, someone had already tried. Sort of.

Dialysis, Disrupted

Back in 1997 when Fissell and Roy were finishing up their advanced training at Case Western, a nephrologist named David Humes at the University of Michigan began working to isolate a particular kind of kidney cell found on the backend of the nephron. Humes figured out how to extract them from cadaver kidneys not suitable for transplant and grow them in his lab. Then he took those cells and coated the inside of hollow fibre-membrane filled tubes similar to the filter cartridge on modern dialysis machines. He had invented an artificial kidney that could live outside the human body on a continuous flow of blood from the patient and do more than just filter.

The results were incredibly encouraging. In clinical trials at the University of Michigan Hospital, it improved the mortality rates for ICU patients with acute renal failure by half. There was just one problem. To work, the patient had to be permanently hooked up to half a hospital room’s worth of tubes and pumps.

The first time Roy saw Humes’ set-up, he immediately recognized its promise—and its limitations. Fissell had convinced him to drive from Cleveland to Ann Arbor in the middle of a snowstorm to check it out. The trip convinced them that the technology worked. It was just way too cumbersome for anyone to actually use it.

The bioartificial kidney is a compact, surgically implanted, free-standing device to treat end stage renal disease (ESRD). It performs the vast majority of the biological functions of the natural kidney.Majed Abolfazli

In 2000, Fissell joined Humes to do his nephrology fellowship at Michigan. Roy stayed at the Cleveland Clinic to work on cardiac medical devices. But for the next three years, nearly every Thursday afternoon Fissell hopped in his car and drove three hours east on I-90 to spend long weekends in Roy’s lab tackling a quintessentially 21st century engineering problem: miniaturization. They had no money, and no employees. But they were able to ride the wave of advancements in silicon manufacturing that was shrinking screens and battery packs across the electronics industry. “Silicon is the most perfected man-made material on Earth,” Roy says from the entrance to the vacuum-sealed clean room at UCSF, where his grad students produce the filters. If they want to make a slit that’s 7 nanometers wide, they can do that with silicon every time. It has a less than one percent variation rate.

The silicon filters had another advantage, too. Because Roy and Fissell wanted to create a small implantable device, they needed a way to make sure there wasn’t an immune response—similar to transplant rejection. Stacks of silicon filters could act as a screen to keep the body’s immune cells physically separated from Humes’ kidney cells which would be embedded in a microscopic scaffold on the other side. The only thing getting through to them would be the salt and waste-filled water, which the cells would further concentrate into urine and route to the bladder.

By 2007 the three researchers had made enough progress to apply for and receive a 3-year $3 million grant from the NIH to prove the concept of their implantable bioartificial kidney in an animal model. On the line was a second phase of funding, this time for $15 million, enough to take the project through human clinical trials. Roy moved out west to UCSF to be closer to the semiconductor manufacturing expertise in the Bay Area. Fissell worked on the project for a few more years at the Cleveland Clinic before being recruited to Vanderbilt while Humes stayed at the University of Michigan to keep working with his cells. But they didn’t make the cut. And without money, the research began to stall.

By then though, their kidney project had taken on a following of its own. Patients from all over the world wanted to see it succeed. And over the next few years they began donating to the project—some sent in five dollar bills, others signed checks for a million dollars. One six-year-old girl from upstate New York whose brother is on dialysis convinced her mother to let her hold a roadside garden vegetable sale and send in the proceeds. The universities chipped in too, and the scientists started to make more progress. They used 3D printing to test new prototypes and computer models of hydraulic flow to optimize how all the parts would fit together. They began collaborating with the surgeons in their medical schools to figure out the best procedure for implanting the devices. By 2015 the NIH was interested again. They signed on to another $6 million over the next four years. And then the FDA got interested.

That fall the agency selected the Kidney Project to participate in a new expedited regulatory approval plan intended to get medical innovations to patients faster. While Roy and Fissell have continued to tweak their device, helped along by weekly shipments of cryogenically frozen cells from Humes’ lab, FDA officials have shepherded them through two years of preclinical testing, most of which has been done in pigs, and shown good results. In April, they sent 20 agency scientists out to California to advise on their next step: moving into humans.

The plan is to start small—maybe ten patients tops—to test the safety of the silicon filter’s materials. Clotting is the biggest concern, so they’ll surgically implant the device in each participant’s abdomen for a month to make sure that doesn’t happen. If that goes well they will do a follow-up study to make sure it actually filters blood in humans the way it’s supposed to. Only then can they combine the filter with the bioreactor portion of the device, aka Humes’ renal cells, to test the full capacity of the artificial kidney.

The scientists expect to arrive at this final stage of clinical trials, and regulatory approval, by 2020. That may sound fast, but one thing they’ve already got a jump on is patient recruiting. Nearly 9,000 of them have already signed up to the project’s waitlist, ready to be contacted when clinical trials get the green light.

These patients are willing to accept the risk of pioneering a third option, besides transplants, which are too expensive and too hard to get for most people, and the drudgery of dialysis. Joseph Vassalotti, a nephrologist in Manhattan and the Chief Medical Officer for the National Kidney Foundation says “the more choices patients have the better,” even though he’s skeptical the device will become a reality within the next few years. An implantable kidney would dramatically improve their quality of life and be a welcome innovation after so many years of treatment status quo. “During World War II we didn’t think dialysis would be possible,” Vassalotti says. “Now half a million Americans are being treated with it. It’s amazing the progress just a few decades makes.”

1Correction: 12:50pm ET The Kidney Project is now slated to begin clinical trials in early 2018. A previous version of this article incorrectly stated they would take place later this year. Changes have also been made to correctly identify the size and timing of grants to the Kidney Project.