11.1 Denominators and study populations
Consider the statement: “Ten participants became insulin-independent.”
The statement is incomplete without the following information:
- Ten out of how many?
- Were all participants treated at the same dose?
- How were participants selected?
- For how long were they followed?
- Were they taking immunosuppressive drugs?
- Were serious harms observed?
The numerator is the number who reached an outcome. The denominator is the relevant total. “10 people” means something different in a study of 12 than in a study of 1,000.
The evidence ladder
- 1
Cells in a dish
- 2
Animal model
- 3
First person
- 4
Small cohort
- 5
Controlled trial
- 6
Regulatory approval
- 7
Long-term real-world use
11.2 Zimislecel: the strongest scalable replacement signal
Zimislecel is an allogeneic, stem-cell-derived, fully differentiated islet-cell therapy. Cells are infused into the portal vein. Participants receive immunosuppression.
The 2025 peer-reviewed report included 14 participants who had at least 12 months of follow-up: 2 received a half dose and 12 received a full dose. All had undetectable C-peptide before treatment.
The significance of this result lies in the renewable starting cell supply and the observation that the grafts produced meal-responsive endogenous insulin at a dose that often replaced external insulin.
Why is this not yet the broad cure? The high-efficacy version still requires transplant-style immunosuppression. Manufacturing consistency, pivotal completion, longer durability, and regulatory review remain important.
11.3 UP421: initial human evidence of immune evasion
The UP421 study tested gene-edited donor islets, not yet a mass-manufactured stem-cell product. A low dose was placed into the forearm muscle of one person with long-standing T1D without immunosuppressive drugs.
At 12 weeks, the peer-reviewed report showed stable, glucose-responsive C-peptide and no detected immune response against the cells. Later investigator and company updates reported continued function through 14 months.
This case provides evidence that gene-edited allogeneic islets can survive and produce glucose-responsive insulin in a human participant without systemic immunosuppression.
It does not yet show that a therapeutic dose of edited, stem-cell-derived islets can safely replace insulin in many people.
Concept questionWhy is one participant scientifically important but clinically insufficient?
Explanation
A single person can disprove “this is biologically impossible.” It cannot estimate how often the result occurs, who benefits, which rare harms appear, or whether manufacture is reproducible.
11.4 Autologous reprogrammed cells
In 2024, researchers reported that a woman received islets made from her own chemically reprogrammed pluripotent cells. The cells were implanted under abdominal muscle, and she achieved insulin independence during the reported year.
This result demonstrated another route from an adult person’s cells to a functioning islet graft.
But the participant had previously received a liver transplant and was already taking immunosuppression. The case therefore cannot prove that an autologous product escapes recurrent autoimmunity or eliminates the need for immune control.
What the case supports
Patient-derived reprogrammed cells can be manufactured into a graft capable of restoring insulin independence.
What it cannot establish
That bespoke autologous manufacture is scalable, broadly reproducible, or safe without immunosuppression.
11.5 Negative and mixed results are evidence too
VX-264 did not produce enough C-peptide for clinical benefit and was discontinued. That weakens confidence in that closed-device implementation.
The antigen-specific DIAGNODE-3 program had a major 2026 setback. That weakens confidence in the tested intervention and subgroup strategy.
Early cell and immune programs can also be terminated, redesigned, or moved into follow-up studies. A registry status change does not explain the biological reason by itself, but it is a reminder to verify whether a program is recruiting, paused, completed, or terminated.
Science advances by eliminating attractive designs as well as confirming successful ones.
Review questions
Review the chapter concepts.
- Why should “10 participants became insulin-independent” always include the denominator?
- What did the zimislecel study establish?
- Which treatment burden prevents the leading zimislecel result from being a broad cure?
- What can and cannot be inferred from the one-participant UP421 report?
- Why does the autologous case not establish freedom from immune control?
Use these questions to identify concepts that require additional review.