Stem cells and type 1 diabetes: where the research stands

Type 1 diabetes looks like a problem with a simple fix. The body has lost one kind of cell, so put that cell back. Researchers have known how to do a version of this for more than 25 years, and in a handful of people it has worked for a decade or longer. Yet almost everyone with type 1 diabetes still takes insulin several times a day or through a pump. The reasons why sit at the center of every stem cell approach now in trials.

What the disease takes away

Insulin is made by beta cells, which sit in small clusters called islets scattered through the pancreas. In type 1 diabetes the immune system's T cells attack and destroy those beta cells. The attack usually starts years before diagnosis, and by the time symptoms appear much of the insulin-making capacity is gone. Without insulin, blood sugar climbs; with injected insulin, people can live full lives, but dosing is imprecise and blood sugar can also drop dangerously low.

That second danger drives most of the work described below. Some people lose the warning signs of low blood sugar, a condition called hypoglycemia unawareness, and can pass out without notice. For them, replacing islets is not a convenience but a safety measure.

Any cell replacement has to clear three hurdles:

1. Supply: enough working islets to control blood sugar.
2. Rejection: donor cells are foreign tissue, so the immune system attacks them as it would a transplanted kidney.
3. Autoimmunity: the original attack on beta cells has not gone away, and new beta cells, even a person's own, are a target for it.

The donor islet era

In 2000 a team in Edmonton, Canada, reported in the New England Journal of Medicine that seven people with type 1 diabetes had all come off insulin after islet transplants. The islets came from deceased donors' pancreases, often two or more per patient, and were infused into the liver through the portal vein, where they lodge and start working. The team used a new drug regimen without steroids, which damage beta cells. It became known as the Edmonton protocol.

The five-year follow-up, published in 2005, was more sobering. About 80 percent of recipients still made some of their own insulin, measured as C-peptide, but only about 10 percent were still off insulin entirely, and the median time off insulin was about 15 months. The people who had to go back on insulin were still better off: their blood sugar was steadier and dangerous lows were far less common.

In June 2023 the FDA approved Lantidra (donislecel), the first donor islet cell therapy licensed in the United States. It is for adults who cannot reach their blood sugar targets because of repeated severe hypoglycemia despite intensive management. Across two studies of 30 people, 21 went without insulin for a year or more, and 10 of them for more than five years; 5 never became insulin-free. The FDA also noted that most participants had at least one serious adverse reaction tied to the infusion procedure or the immunosuppressive drugs, and that stopping those drugs meant losing the islets.

Donor islets prove the idea, but they cannot scale. There are far too few donor pancreases, quality varies, and every recipient needs lifelong immunosuppression.

Timeline

Islets grown from stem cells

Stem cells answer the supply problem. In 2014 two groups, one at Harvard and one in industry, published methods that walk human pluripotent stem cells through the stages of pancreas development until they become clusters that sense glucose and release insulin. Our pages on stem cell technology and induced pluripotent stem cells cover how that kind of directed differentiation works and how it is scaled for manufacturing.

The leading test of the idea is zimislecel (formerly VX-880), made by Vertex Pharmaceuticals from a donor stem cell line. Results from its phase 1/2 study appeared in the New England Journal of Medicine in 2025. The 12 participants who received a full dose, infused into the liver's portal vein, all started making their own insulin. All 12 were free of severe hypoglycemic episodes and had an HbA1c under 7 percent, and 10 of the 12 were off insulin at one year.

The caveats matter as much as the headline. Every participant took immunosuppressive drugs, so this approach swaps insulin for a different lifelong medication with its own risks. Two people in the study died: one from cryptococcal meningitis, a fungal infection that is a known danger when the immune system is suppressed, and one from progression of dementia that was present before the study. The study was small, follow-up was short, and the authors described the analyses as interim. The company had planned to apply for approval in 2026, but in a spring 2026 filing it reported pausing and then resuming dosing in the late-stage trial after an internal manufacturing review, and said it would update its timelines. Check the trial's ClinicalTrials.gov record (NCT04786262) for its current status.

Three ways around the immune problem

Put the cells behind a barrier

If cells sit inside a pouch or capsule that lets insulin and nutrients pass but keeps immune cells out, no drugs would be needed. The idea is decades old. The first human trial of stem-cell-derived pancreatic cells in a protective device started in 2014, run by the company ViaCyte. The body walled the devices off with scar tissue and too few cells survived. A later "open" version let blood vessels grow in but required immunosuppression, and reports in 2021 showed some participants making small amounts of insulin, far short of what they needed. Vertex bought ViaCyte in 2022 and in March 2025 stopped its own encapsulated program, VX-264, after it failed to raise insulin output enough. Other groups are still testing different materials.

Edit the cells so the immune system ignores them

Gene editing can remove the surface markers that immune cells use to recognize foreign tissue and add signals that tell them to stand down. In 2025 researchers in Uppsala, Sweden, reported in the New England Journal of Medicine on one man with long-standing type 1 diabetes who received gene-edited donor islet cells in his forearm muscle with no immunosuppressive drugs. Twelve weeks later his immune system showed no response to the cells, and they were releasing insulin in step with his blood sugar. The dose was deliberately small, so this tested survival, not freedom from insulin. It was one person, using donor islets rather than stem cells. Open questions include how long such cells last and what happens if an edited cell ever grows out of control, since the immune system would not catch it.

Use the patient's own cells

Making islets from a person's own reprogrammed cells avoids rejection, though not the autoimmune attack. In 2024 a team in China reported in Cell on a 25-year-old woman given islets made from her own chemically reprogrammed stem cells, placed under a sheath of abdominal muscle instead of in the liver. She was off insulin from day 75 and stayed off it at one year. She was already taking immunosuppressants for an earlier liver transplant, though, so the case cannot show whether her own cells would have escaped the immune attack on their own. Personalized manufacturing is also slow and expensive.

What is realistic right now

Stem-cell-made islets are available only inside clinical trials. Those trials mostly enroll adults with long-standing type 1 diabetes and severe, hard-to-control hypoglycemia, because for them the risks of immunosuppression can be worth taking.

Donor islet transplantation, including Lantidra, is available for a narrow group of adults at a small number of centers, and whole pancreas transplants are an established option, often done together with a kidney transplant. Teplizumab is not a cell therapy; it delays diagnosis in some people found through antibody screening.

To look for a trial, the usual starting point is your endocrinologist. You can also search ClinicalTrials.gov for "type 1 diabetes" and "islet," and use our guide to reading a ClinicalTrials.gov listing to check phase, sponsor, eligibility and whether it is recruiting. Our explainer on translational medicine shows why a promising phase 1/2 result can still take years to become an approved product.

Clinics advertising stem cells for diabetes are a different matter. Infusions of fat-derived, cord blood or "mesenchymal" cells offered for cash outside a trial are unproven. Even in a careful study at the University of Florida, children given their own stored cord blood did not keep their insulin production (Diabetes Care, 2011).

Type 2 is a different problem

Type 2 diabetes is not an autoimmune disease. It starts with the body responding poorly to insulin, and beta cells gradually wear out under the strain. Treatment rests on lifestyle changes and a long list of approved drugs, from metformin to GLP-1 medicines to insulin. Cell replacement is not where type 2 research is focused, and a clinic offering stem cells for type 2 diabetes is selling something no trial has shown to work.

The most useful single fact for anyone with type 1 diabetes following this field: every stem cell islet therapy that has freed people from insulin so far has also required immunosuppressive drugs or depended on them already being taken. Watch for results that remove that requirement; that is the step that would change who can benefit.