Islet Transplants Face Test of Time

See allHide authors and affiliations

Science  01 Oct 2004:
Vol. 306, Issue 5693, pp. 34-37
DOI: 10.1126/science.306.5693.34

Will the Edmonton protocol, hailed as a major step toward a cure for type I diabetes, hold up in the long run?

Ellen Berty was driving home from her special-education job when the call came, on the cell phone she'd bought expressly for this purpose. The caller spoke the magical words every person needing a transplant dreams of hearing: “We have a match.” In her Mazda convertible, Berty let out a yell of triumph. “I'd won the contest of my life,” she recalls thinking on that sunny June day 3 years ago.

Ten hours later, Berty lay sedated in a radiology suite at the National Institutes of Health (NIH) in Bethesda, Maryland, while doctors delicately injected a yellowish green solution into a vein feeding into her liver. The mix held hundreds of thousands of islets, cells from the pancreas of a man who'd died suddenly. These cells were supposed to supply Berty with the critical hormone insulin she'd lacked for 40 years, ever since being diagnosed with type I diabetes at the age of 13.

Berty's islet-cell transplant is part of a vast global experiment, a test of a therapy that's been hailed as the greatest hope for curing type I diabetes. Five years after physicians in Edmonton began transplanting islets under a new and widely celebrated protocol, the long-term results of this strategy are beginning to emerge. They paint a nuanced and still unfinished picture of a treatment that some doctors concede is riskier than they expected and less effective than they had hoped.

The NIH trial in which Berty enrolled reflects the promise and peril of these transplants. Berty has been one of the lucky ones. She stayed off insulin injections for 2 years after her transplant. Today, she's back on a low dose, but she has relatively few side effects from the immunosuppressive drugs she takes to prevent islet rejection. Like most islet recipients, Berty also has none of the diabetes complications she suffered before.

Still, Berty was NIH's last islet-transplant patient. After treating her and five others, NIH stopped accepting new volunteers, its physicians increasingly anxious that antirejection drugs, which must be taken for life, were spawning problems worse than those the transplanted islets were solving.

Other centers disagreed. They continued testing the procedure, and today more than 300 patients have received islets under the protocol crafted by the Edmonton team. NIH, the Juvenile Diabetes Research Foundation (JDRF), other nonprofit organizations, and several European governments have poured hundreds of millions of dollars into coaxing these transplants to work. But as islet transplants expand and less experienced centers launch islet programs, it's become less clear what “work” really means.

All smiles, in this case.

Ellen Berty and her NIH doctor David Harlan both say her islet transplant was a success. But Harlan worries that not everyone has been so lucky.


The original goal of islet transplants has been met: Lifelong diabetics receiving new islets have been able to abandon, at least for a time, insulin shots. According to an NIH survey published last month, 22 of 38 islet recipients were still off insulin a year after their transplant. Those numbers sag with time, though, and it's not known how long transplanted islets can thrive, or what's killing them when they fail.

A more pressing question is whether insulin independence is enough. A sizable minority of islet recipients struggle with new health problems, from painful mouth ulcers to anemia to kidney disease, largely attributed to the combination of antirejection drugs prescribed by the Edmonton protocol. And no one knows whether patients given islets actually live longer than they would have without them. A controversial study from some of the NIH scientists who treated Berty hints that the risk of a shortened life span might be real.

Physicians are launching clinical trials to improve the safety and effectiveness of islet transplants, but they're far from offering this experimental therapy to all but the most severely affected diabetes patients. For one, there aren't enough cadaver pancreases to go around. Although many are looking at stem cells as a renewable source of islets, that's still a distant prospect.

Aldo Rossini, director of the diabetes division at the University of Massachusetts Medical School in Worcester, compares the current state of islet transplants to the Wright brothers' first flight. “They flew a couple hundred feet”—a remarkable accomplishment at the time, he notes. Still, says Rossini, “no one could have expected us to fly to California in that plane.”

Measures of success

Since 1972, when Paul Lacy, a researcher at Washington University in St. Louis, cured diabetic rats by giving them healthy islets, transplanters have sought to extend that success to humans. The approach seemed obvious: In type I diabetes, the body's immune system mistakenly attacks insulin-producing islet cells in the pancreas, and by the time the symptoms of diabetes surface, most of these islet cells are gone. But in more than 400 human islet transplants beginning in the 1970s, doctors couldn't get transplanted cells to stick. Many suspected that, ironically, the steroid drugs given to prevent islet rejection were also toxic to islet cells.

Then in the summer of 2000, the dreary world of islet transplants changed forever. A team at the University of Alberta in Edmonton, Canada, reported in The New England Journal of Medicine that they'd given islets to seven diabetes patients under a new regimen, and after roughly a year, all seven were still off insulin.

Unlike earlier islet transplants, the Edmonton protocol didn't involve steroids. Led by James Shapiro, the Edmonton team combined three antirejection drugs, one of which, sirolimus, had recently begun human testing. It also gave patients islet cells from multiple pancreases.

The group's report instantly became medical legend. “Here,” says David Nathan, director of the diabetes center at Massachusetts General Hospital in Boston, “was this absolute miracle.”

Research funders quickly responded to Edmonton's success. JDRF, one of the country's wealthiest and most powerful disease advocacy groups, declared islet transplants a top priority, and since 2000 it has poured $225 million into the field. Hospitals in the United States and Europe raced to set up islet-transplant centers, and patients flocked to them in droves. Emory University's 18-month-old islet-transplant program has fielded 5500 inquiries from patients, says surgeon Christian Larsen, its director. Constrained by strict entry criteria and a tight budget, Emory has given transplants to just six.

Like others in the field, Larsen believes that ideal islet-transplant candidates are patients who, despite their best efforts, cannot control their blood sugar. More dangerously, their bodies have lost the ability to sense blood sugar lows, resulting in sudden fainting spells, seizures, and even comas or death. For patients like Berty, who suffered middle-of-the-night seizures and blackouts while driving, the condition is terrifying and profoundly disruptive. It's these patients—maybe 1% of type I diabetics—who islet transplanters welcomed into clinical trials. “Every patient we take on, they're near death's door or in desperate straits,” says Shapiro.

Out and in.

After extracting islets from a pancreas, doctors infuse them into a diabetes patient.


Transplanters quickly found, however, that the success of the Edmonton protocol is tough to sustain; the new islets seem to fade over time. Experienced islet-transplant centers like Edmonton, the University of Miami, and the University of Minnesota, Twin Cities, boast insulin independence rates of 80% to 90% a year after transplant, far higher than the rates of many smaller centers. After 3 years, that falls to 60% among Miami's patients, says Camillo Ricordi, scientific director of the Diabetes Research Institute there. Mark Atkinson, a pathologist who studies diabetes at the University of Florida, Gainesville, and research chair of JDRF, recently reviewed unpublished data on patients from Edmonton, 3 to 4 years after their transplants. Between 12% and 25% were insulin independent, he says. Among the original Edmonton seven, only two remain off insulin, says Shapiro.

“Something is not going in the right direction long term,” says Ricordi. One possibility, he says, is that the antirejection drugs, although less toxic to islets than steroids, still harm the cells. Some nondiabetic patients taking the drugs after receiving liver, heart, or kidney transplants have developed diabetes, notes David Sutherland, chief of transplantation at the University of Minnesota.

A more fundamental problem may be that the immunosuppressive drugs can't erase the underlying autoimmune response that killed a patient's original islets. “These people don't like islets, no matter whose they are,” says Peter Senior, an endocrinologist at the University of Alberta.

Another explanation for islet failure is that patients may be receiving too few islets, even if they get cells from multiple donors. A normal pancreas has roughly 1 million islets, but current techniques allow only about 400,000, at most, to be extracted from a donor pancreas. Moreover, unknown numbers die soon after they're transplanted, forcing the rest to labor unusually hard to supply enough insulin. The islet cells may just “poop out” over time, says Sutherland.

Edmonton found that giving patients islets from as many as three pancreases could sustain insulin production longer. But pancreases are a scarce and costly resource. Fewer than 2000 are donated each year, and most go toward whole-organ pancreas transplants for diabetes. In the United States, they also cost from $15,000 to $25,000 each.

Increasingly, however, transplanters are wondering whether insulin independence, a goal pushed heavily by islet-transplant centers, funders, and many patients, is the only yardstick by which to measure islet-transplant success. Patients like Ellen Berty and others who have gone back on insulin have found that partial islet function can stave off the hypoglycemia they experienced before their transplants. This has doctors hoping that islet transplants might prevent long-term complications of diabetes, even if recipients still need some insulin. “Even if they're not off insulin,” says Shapiro, “their problems go away.”

Walking a tightrope

But what if the therapy is as bad as the disease? Last month, the risky nature of these transplants was underscored by NIH's first report from its Collaborative Islet Transplant Registry. None of the 86 islet recipients NIH surveyed died from the procedure. But the agency cataloged 20 serious adverse events linked to islet transplants. They include four cases of life-threatening neutropenia, a depletion of white blood cells caused by antirejection drugs. “Islet transplants are still incredibly experimental,” says Ricordi.

Amy Parker learned that the hard way. Parker, who asked that her real name not be used, was diagnosed with type I diabetes as a teenager. As her disease became progressively more unmanageable, she began having seizures from low blood sugar, and blood vessels behind her eyes started to leak. She needed multiple laser eye surgeries to preserve her vision.

In 1999, soon after Edmonton began its revolutionary set of islet transplants for patients like her, she applied. In November and December 2002, Parker underwent two separate islet transplants.

Then, her new ordeal began.


James Shapiro pioneered the Edmonton protocol, in which more than 300 patients with diabetes have participated.


Since receiving the transplants, her insulin requirements have dropped to a quarter of what they once were, and she no longer suffers seizures or hypoglycemia. But every day she experiences “deathly horrible” headaches, a result of the antirejection drugs, she learned. Two summers ago, she began having trouble breathing while on a family vacation in British Columbia. In July, she was switched from the drug sirolimus, a possible culprit, to mycophenolate, another immunosuppressant. If that fails to help her, says Parker, she may drop out of the study and lose her islets.

The experimental nature of islet transplants was further driven home last June at the American Diabetes Association meeting in Orlando, Florida, where the Edmonton team released troubling kidney function data on its first 45 islet-transplant patients. Of the five patients Edmonton has followed for 4 years, two have “quite bad renal outcomes,” including one who has required dialysis, says Senior. Overall, a third of the 45 have high levels of a protein in their urine that's normally a harbinger of declining kidney function.

On the other hand, about a fifth of diabetes patients typically develop kidney disease. Says Senior, “These people may well have ended up with kidney failure irrespective of transplant. The question is, are these drugs hastening that?”

Changing course

It's mixed news like this that has dampened enthusiasm among a handful of doctors who once believed islet transplants were ready for patients. One is David Harlan, a diabetes specialist at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) in Bethesda, Maryland, who treated Berty. Like his colleagues around the world, Harlan was enthralled by the Edmonton protocol when it first appeared. In late 2000, he pulled together a transplant team and more than $1 million in NIH funding to launch an islet- transplant program at NIH. From 29 December 2000 through 14 June 2001—the date of Berty's transplant—he and his colleagues performed transplants in six women with severe type I diabetes.

The team quickly grew troubled by what it was seeing. Looked at through the lens of diabetes, the picture was relatively rosy: Four of six patients became insulin independent, and three stayed that way for at least a year and a half. Even those who still needed some insulin no longer suffered the hypoglycemic episodes that had driven them to this experimental trial in the first place.

But problems abounded. Two patients, including one off insulin, had to discontinue immunosuppressants because of the intolerable side effects, such as deteriorating kidney function, and their bodies rejected the islet cells. Even Ellen Berty, the NIH success story, ran into some trouble. In her first year after the transplant, the antirejection drugs contributed to a severe foot infection and caused mouth ulcers so large that NIH dentists photographed them for use in a textbook. For Harlan, the price NIH islet recipients were paying didn't seem worth it.

“When you expand the experience, you find problems that were not expected,” says Antonio Secchi, head of the transplant program at Milan's University Vita-Salute San Raffaele, one of about four major European islet-transplant centers. Two of his center's 10 islet recipients who became insulin independent have since dropped out of the program because of drug side effects.

One central question that preoccupies Harlan is whether islet recipients will live longer than those in comparable health who don't receive transplants. It's too early to answer that question directly, so Harlan turned to data on pancreas transplants. They have been used for years in much the way islet transplants are now, although most are given to diabetes patients who also need kidneys.

Harlan and his colleagues examined data from 124 transplant centers in the United States from 1995 to 2000 and arrived at an unsettling conclusion: Patients receiving a solitary pancreas or a pancreas after a kidney transplant were more likely to die within 4 years than those still on the waiting list.

Published last December in the Journal of the American Medical Association, the article touched off a furor. Many transplant surgeons disputed its results. Minnesota's Sutherland and his colleague Rainer Gruessner have reanalyzed the data, and Sutherland says they've arrived at a conclusion opposite to Harlan's. Some patients in Harlan's study, says Sutherland, were on the waiting list of more than one hospital and ended up being counted twice. The study also excluded patients awaiting pancreas transplants who had very poor kidney function; Harlan worried that that might produce misleading results, but Sutherland believes those patients should be included.

Concerns about long-term survival after an islet transplant, however, must be weighed against the improved quality of life that many transplant recipients experience, at least initially. “The psychological benefit of insulin independence is potentially enormous,” says Emory's Larsen, “and it's hard to understand for a nondiabetic.”

Rita Hart, 46, is off insulin after undergoing three transplants at Miami over 2 years, the last in July 2003. Before her transplant, diabetes was consuming her life and complications were piling up. “I was losing hope,” she says. Now, despite drug side effects that include anemia, she feels vastly more optimistic.

“It's striking how many patients ask for a third transplant,” says Senior. “Even with all the side effects and all the downsides, they still think it's a good thing.”

Donation in demand.

Islet cells such as these are in short supply for transplants.


And so Edmonton, like many other islet-transplant centers, continues to grow. Today, more than 25 hospitals have performed islet transplants that hew closely to the Edmonton protocol. NIH will soon announce $75 million in awards for a new clinical islet transplantation consortium in which centers will collaborate on islet studies. Although Harlan ended his islet-transplant trial early, the agency believes the treatment is worth pursuing. “This is not a black-and-white issue,” says Allen Spiegel, director of NIDDK.

Roadblocks to expansion

New money, however, will go only so far: Islet transplants are extraordinarily expensive, costing up to $200,000 in the United States for one patient in the first year. Antirejection drugs add another $30,000 annually after that. At centers like Miami, where most patients remain part of a protocol, the price of success—of supporting patients for years after a transplant—is becoming prohibitive, says Rodolfo Alejandro, an endocrinologist and director of the clinical islet-transplant program at the University of Miami. (Costs in Canada are somewhat lower because there's no charge for organs, and the Alberta health care system agreed in 2001 to pay for transplants for Alberta residents.) Because they're still considered experimental, most United States islet transplants are funded by NIH, JDRF, and sometimes by pharmaceutical companies that manufacture immunosuppressants.

Costs are one roadblock to performing the kind of large, controlled studies that some say are needed before islet transplants can shift from being an experimental therapy to being one approved by the U.S. Food and Drug Administration (FDA). Some islet transplanters, like Alejandro, believe that one option is for FDA to approve the therapy under its existing “orphan drug” category, making it available to essentially the same patients getting islets now—those with uncontrolled diabetes. That way, it could be covered by insurance. A year ago, FDA held a public advisory committee meeting in Gaithersburg, Maryland, and agency officials made clear they want certain issues addressed first. Those include consistency in how islets are processed and a better assessment of the risk-benefit balance.

No matter how FDA rules, major hurdles stand in the way of islet transplants going mainstream. First, the shortage of donor pancreases means scientists must find a renewable source of islets. One popular option would involve using some type of stem cell. This year, JDRF has committed more than $8 million to stem cell research, more than $6 million of it to human embryonic stem cell work. Yet creating islets from stem cells isn't imminent, according to Larsen and other transplanters.

Milder immunosuppressive regimens might come more rapidly. One study that's gearing up at Miami calls for giving islet recipients a dose of bone marrow cells culled from the donor's vertebrae, to try to help patients better tolerate the islet cells.

Current islet recipients, and the many more people with diabetes hoping for a transplant, are eagerly awaiting the day when islet transplants are easier to come by and gentler to receive. But Berty remains upbeat. A book she's written chronicling her experience came out this spring. Its title: I Used to Have Type 1 Diabetes: Kiss My Islets.

View Abstract

Stay Connected to Science

Navigate This Article