The central problem in beta cell replacement therapy has never been whether transplanted cells can produce insulin. They can. The problem is that the immune system finds them and destroys them, often within months. Every promising result in this field eventually hits that wall. At the ADA Scientific Sessions in New Orleans last week, a small trial from the University of Chicago produced evidence that the wall is not as permanent as it looked.

In type 1 diabetes, the immune system destroys the insulin-producing beta cells in the pancreas. That process cannot be stopped. So the field has spent decades trying a different strategy: replace the destroyed cells, then prevent the immune system from destroying the replacements.

Replacing the cells has never been the hard part. Islet cells, the clusters of pancreatic tissue that contain beta cells, have been shown to produce insulin reliably once transplanted. The sticking point has always been what comes next. Standard immunosuppressive drugs, particularly calcineurin inhibitors, carry significant long-term costs: elevated infection risk, kidney damage, and a heightened likelihood of cancer. Patients who have received liver or kidney transplants accept those risks because the alternative is worse. For people managing T1D with insulin injections, the calculation has historically been harder to justify.

Last week at the ADA Scientific Sessions, Eledon Pharmaceuticals and the University of Chicago Medicine reported updated results from a 12-patient pilot study funded by Breakthrough T1D. Every participant had T1D of approximately 33 years' duration on average. Each had a history of recurrent severe hypoglycemic events, the episodes that can cause seizures, loss of consciousness, or death. After receiving cadaveric donor islet transplants alongside tegoprubart, an investigational anti-CD40L antibody, all 12 achieved insulin independence. Average HbA1c across the cohort fell from 8.0% before transplantation to 5.4% at most recent follow-up. There were no rejection episodes and no severe hypoglycemic events after transplantation.

Follow-up ranged from four to 22 months, with a median of eight months. That is encouraging, but durability will need to be demonstrated across years, not months.

The mechanism is what makes this study worth paying attention to beyond the headline number. Tegoprubart does not suppress the immune system broadly. It blocks a specific costimulation signal that T cells need in order to recognise and attack foreign tissue. Rather than damping immunity across the board, it disrupts the precise pathway the immune system uses to declare transplanted cells an enemy. Consistent with that, the researchers found no donor-specific antibodies across the cohort, which are the markers that indicate the immune system is actively building a targeted response against a graft.

Two important constraints. These are cadaveric donor islets, not stem cell-derived. The Reader Lens below explains why that distinction matters. And this is a 12-person pilot at a single academic centre. Replication at scale will determine whether the result holds.

What the study contributes is evidence that the immune rejection problem can be addressed with more precision than the field previously had. That conclusion extends directly to every cell therapy program in development, including zimislecel. The transplanted cells work when the immune system leaves them alone. The work now is figuring out how to make that happen reliably.

Published in Nature (May 2026) | Human stem cells, xenograft and disease cohort analysis | University Health Network, Toronto / University of Oxford

What they did: Researchers identified a previously undescribed population of human blood stem cells that retain altered gene activity after repeated exposure to infection or stress. These cells, which they called HSC-iM, carry reprogrammed inflammatory signals long after the original inflammation resolves and pass that programming on to the immune cells they generate.

What they found: HSC-iM cells accumulated with age and appeared at elevated levels in people who had severe COVID-19, in those with inherited inflammatory blood disorders, and in those with clonal haematopoiesis, an age-related condition associated with blood cancer risk. Cancer-associated mutations concentrated preferentially in HSC-iM cells compared with other blood stem cells.

Evidence level: Human stem cell analysis combining laboratory experiments with direct human disease cohorts. Not a clinical trial.

Key caveat: Identifying the memory is not the same as knowing how to erase or redirect it. The therapeutic applications are not yet defined.

Why it matters anyway: If prior illness permanently reprograms the stem cells that generate immune responses, that begins to explain why older immune systems tend to be less tolerant of foreign cells. For a field built on the premise of immune tolerance, understanding where that intolerance is rooted matters.

The tegoprubart trial used islet cells recovered from deceased organ donors: real pancreatic tissue, extracted, purified, and infused into the recipient's liver.

Stem cell-derived islets, like Vertex's zimislecel, are manufactured in a laboratory. Induced pluripotent stem cells are guided to develop into functional beta cells, then harvested and infused. No donor pancreas required.

The distinction is about scale. Donor islets are limited by how many compatible organ donations occur each year. In the United States, that means a few thousand patients at most, reserved for the most severe cases. Stem cell-derived islets, if the manufacturing can be made consistent and the immune rejection problem solved, could in principle reach everyone who needs them. That is why the stem cell route is not an alternative to donor islet therapy. It is the version of the same idea that could actually scale.

Twelve people went off insulin last week. All of them still need a drug to stop their immune systems from destroying the replacement cells. The science can now do something remarkable. The body hasn't agreed to accept it yet.