Patients with diabetes can show vast improvement after receiving transplants of insulin-producing islets from cadavers. Though they must take drugs to stall rejection of the transplanted cells, several hundred patients with the most severe type of diabetes have benefited from the procedure since it first became established in 2000. But the effects don't last. After two years, islet function begins to decline, and unless more cells are transplanted, patients eventually return to full insulin dependency1,2,3.

Islet from cadaver organ donors. Credit: Prodo Laboratories

Also, donor pancreatic cells are in short supply — there are too many patients who need advanced treatment and too few cell sources. To solve this problem companies like Novocell in San Diego and Geron in Menlo Park, California, have turned to stem cells. By differentiating islets from human embryonic stem cells (hESCs), an unlimited supply of therapeutic cells can potentially be generated.

From there, the rest should be a snap. Unlike other therapeutic cells, transplanted islets don't have to integrate into a host organ. Indeed, cadaver islets are often delivered to the liver — a location far from their native home in the pancreas. Even so, the first trials are optimistically scheduled for 2011, assuming preclinical studies go as planned.

The problem starts with making the so-called beta cells that can sense blood sugar levels and react properly to them. Researchers can now prompt hESCs to differentiate into beta cells in the laboratory, but these do not respond correctly to cues in the bloodstream. "We can make cells that are loaded with insulin, but they won't release it — at least not in response to glucose," says Alan Lewis, executive director of the New York City–based Juvenile Diabetes Research Foundation.

Until January 2009, Lewis was chief executive of biotech company Novocell. He led the company when its scientists discovered a way to circumvent the problem of glucose responsiveness. Instead of differentiating hESCs into beta cells in the lab and then putting them into diabetic mice, the team tried transplanting less-mature cells. "The islets differentiated in vivo. The body provided the additional factors we were somehow missing," explains Lewis.

Two months after transplant, the cells matured into structures that looked and behaved like human pancreatic islets, and the mice showed no signs of diabetes4. When the cells were removed from one set of mice, the diabetes returned. In the other mice, the cells are still functioning well more than a year after the results were published.

Although these were groundbreaking results, a problem soon arose that limited their applicability to humans: tumours developed in more than 5% of the treated mice, suggesting that some of the progenitor cells never stopped dividing.

The mice used in the Novocell study lacked functioning immune systems and were probably less able to suppress wayward cells than normal mice would be. But even if the immune system were intact, researchers worry about what might happen in patients. Tumourigenicity is a black cloud hanging over the entire hESC field, says Lewis. "We simply don't know what would happen if this experiment were repeated in humans," he says.

Another concern in moving cells to humans is that an intact immune system would also attack properly functioning transplanted cells, weakening any therapeutic benefit. So even if in vivo differentiation works in humans, putting cells into a patient's body doesn't mean the cells will survive. Many worry that the immune response, or a requirement to suppress it, would prohibit the wide use of stem cell therapies.

This is why the diabetes community is placing its hopes in encapsulation technology, says Henrik Semb, director of the stem cell research institute at Lund University in Sweden. "At the moment, encapsulation is the only way to ensure prevention of tumours and autoimmune attack."

'Soft' encapsulation for immune protection

Soft, permeable coatings have been developed to hide cells from the immune system. This protects cells from an immune attack while allowing them to sense glucose and secrete insulin in response. And even if cells don't survive for a patient's lifetime, biodegradable coatings that eventually break down in the body could theoretically allow new cells to be transplanted without requiring the buildup or removal of capsule skeletons.

Several animal studies have shown encouraging results, and human trials are inching forward. The most extensive human trials were conducted in Moscow under the auspices of Living Cell Technology. The Auckland, New Zealand company uses beads made of alginate — a gel-forming biopolymer utilized in a variety of food and pharmaceutical products. In February, the company reported positive interim results of a trial involving seven patients with type 1 diabetes who received transplants of alginate-encapsulated neonatal pig islet cells.

Most of Living Cell Technology's patients received their first treatment more than a year ago, and no safety problems have been reported. During the study, glycated haemoglobin levels (a measure of blood glucose levels) dropped from a mean of 8% to 6.8%, which approaches the 6.2% considered normal by the American Diabetes Association. But because no control group was used, these results could also be attributed to increased patient monitoring and support, which can yield similar benefits.

Another human study, conducted by Novocell, delivered encapsulated human cadaver islets and low doses of the immunosuppressant cyclosporine to two type 1 diabetes patients. This study, conducted in 2005, was one of the first to use encapsulation technology in humans.

Novocell researchers ran tests for c-peptide — a component of proinsulin made only by islet cells and a clear sign of the body's insulin-producing capabilities. These tests indicated that the transplanted cells made and released insulin. Yet the patients had to keep taking almost as much insulin as they did before they received treatment.

"The results were marginal," says Novocell cofounder David Scharp, who left the company in 2006 to start Prodo Laboratories — a diabetes company based in Irvine, California, that supplies cadaver donor islet cells for research. Scharp helped guide the study, and although the findings have not yet been published, select results were presented in 2006 at the 66th meeting of the American Diabetes Association in Washington DC.

"The problem is that the present encapsulation strategies have yet to show efficient performance of islet cells," says Semb. Although encapsulation technologies seem to allow cells to survive and make insulin, they have not yet 'worked', in clinical terms, by enabling patients to decrease their insulin shots.

Most research encapsulates islet cells directly. No one has yet tested encapsulated progenitor cells in humans or even demonstrated that encapsulated stem cells can differentiate into islets. Similar research, however, is progressing in other stem cell fields. Kristi Anseth's lab at the University of Colorado at Boulder has shown that coated mesenchymal stem cells can still differentiate in coated form. Her team encapsulated the cells in a hydrogel that contained small molecules known to direct differentiation. The encapsulated mesenchymal stem cells became adipocytes and other cell types. What's more, the hydrogel seems completely harmless when injected subcutaneously in animals, and previous studies have shown that it allows encapsulated islets to release insulin in response to glucose5.

Many researchers believe encapsulation can be made to work with islet cells6. "We anticipate our stem cell–derived therapies will be introduced to the clinic in encapsulated form," says Liz Bui, Novocell's director of intellectual property and corporate development. Novocell hopes to file an investigational new drug application in 2011 with the US Food and Drug Administration for its stem cell–derived beta cells. The company was founded in 1999 as an encapsulation company, and it has developed a 25–50 micrometre–thick coating that is made of polyethylene glycol, a substance also used in cosmetics and drugs. Geron is also developing a therapy for diabetes that relies on encapsulation technology, although it won't release details.

Trapping tumours

But no one thinks that all the transplanted cells will mature as desired. "If you transplant precursor cells, there will be some percentage that don't differentiate into the final cell type, and they will keep reproducing," says Scharp. "There is also a possibility that the cells can de-differentiate."

And coatings that hide cells from the immune system may not prevent the cells from escaping into the body. "With the softer technologies, you get breakage," explains Pamela Itkin-Ansari of the University of California, San Diego. Not only could tumour-forming progenitors escape, but they would trigger an immune response, threatening the transplanted cells.

More durable encapsulation might help, but cells tend to survive poorly within such structures, where diffusion distances are longer, and cells in the center don't get enough oxygen. This problem has been so persistent that researchers have tended to focus on soft capsules, which coat small groups of cells.

However, in April this year, Itkin-Ansari and her colleagues showed that there may be a way to boost cell survival within durable capsules. Again, the trick was to start with immature cells. Using bioluminescent imaging she showed that progenitor cells taken from one mouse can survive after transplantation into a second, immune compromised mouse. In a separate experiment, progenitor cells taken from human foetal pancreases differentiated inside durable capsules, giving rise to functional beta islets6. When mature beta cells were encapsulated, they died after transplantation.

Although Itkin-Ansari did not use stem cells, she considers the findings proof of concept for stem cell applications. To fund the project, Itkin-Ansari and Novocell received a joint grant from the California Institute of Regenerative Medicine. "We want to transplant stem cell–derived progenitors when they are lineage committed but not terminally differentiated," she says.

But these projects still have a long way to go before they are ready for the clinic. Though researchers are optimistic that encapsulated precursor cells can be made to differentiate in vivo, they debate whether it will happen by 2011, the touted date for clinical trials. "If you ask Novocell or [any one] who is working on this, they are under pressure from their investors and always give you a too optimistic view," says Semb. "My opinion is that the first human tests will take more than 5 years, probably between 5–10 years."

Wrangling lawyers

The race to bring a stem cell therapy for diabetes to the clinic may come down to getting the right cells with the right coating, but the technology may also stall in the courtroom. Novocell was issued a patent on 31 March for human definitive endoderm (DE) cells — the precursor for potentially therapeutic and lucrative beta cells. The company hopes its new intellectual property will bring in clinical trial partners and financial backers.

But Geron says its intellectual property comes first, both chronologically, because it is based on agreements with the University of Wisconsin made in the 1990s, and developmentally, because human DE cells are derived from hESCs. Geron claims exclusive commercial rights to make islets from hESCs. The company's license was granted by the Wisconsin Alumni Research Foundation, which owns multiple patents that, controversially, cover the use of any hESC application.

"Our view is that the patent office erred in granting the Novocell patent," says Geron company spokesperson Anna Krassowska, adding that Geron has previously described the same method Novocell claims.

If it stands up, Novocell's patent will be far-reaching. It covers any use of cells that arise from a culture containing 15% or more DE cells. "Instead of patenting a particular preparation method, our intellectual property protects any use of DE cultures," says Bui. Even if the DE cells are not part of the final product, their use will be covered by the new patent.

DE cells also differentiate into other potentially lucrative cells — including those in the thyroid, stomach, pancreas and even the middle ear. Thus, the patent could allow Novocell to absorb therapeutic programs from other companies, says Bui.

But Sean O'Connor, a law professor at the University of Washington School of Law in Seattle, says that with Geron gunning to challenge Novocell's patent, potential 'infringers' know they have a ready ally. "Novocell has to know this as well and must act very carefully in suing any infringer," he says. Bui is hoping that Novocell's patent estate will instead encourage alliances between Novocell and a big pharmaceutical company. "We are looking for a licensing partner that will bring us to the clinic." (Though Novocell signed a deal with Pfizer, based in New York City, in December, it is a drug-screening deal that does not include the diabetes clinical program.)

Any cell therapy for diabetes will likely need support from a large, experienced company to reach the market. So while scientists are working out how to make cells survive and function inside the body, businesses must work out how make expensive clinical programs go forward in a market with uncertain – and disputed – intellectual property.