The Math Behind the Medicine — Why Making Replacement Beta Cells Is Only Half the Challenge
By Terri D. Goldson, Sr.
A note before you read: This is an AI-generated demo exploring possibilities in stem-cell gene editing for Type 1 diabetes — not a finished study, and not peer-reviewed research. The numbers in it are modeled examples, not trial results. I am sharing it to invite further analysis by professionals. This is information, not medical advice.
The Big Picture
In an early study of zimislecel, a treatment made from stem cell–derived islets, 10 of 12 people who received the full dose were off injected insulin at one year. Everyone in the study received drugs to suppress the immune system. [1]
Researchers face two challenges. First, they need to make enough cells that release insulin in response to blood sugar. Then they need those cells to survive and keep working after transplant.
The Biological Typo
In Type 1 diabetes, the immune system attacks the pancreatic beta cells that make insulin. PDX1 helps guide pancreatic development. Some variants cause rare forms of single-gene diabetes, so PDX1 serves as an example of how a specific DNA change could be corrected in the lab. A study aimed at autoimmune Type 1 diabetes would choose a target suited to that goal. [2, 3]
Think of that kind of DNA change as a biological typo. A conventional CRISPR-Cas9 approach can cut DNA at the target and give the cell a template to copy during repair. [4]
The Invisible Bottleneck — Editing and Cell Survival
Conventional Cas9 acts like microscopic scissors, cutting both strands of DNA at the target. The editing rate tells you how often the intended change occurred among the surviving cells. The survival rate tells you how much of the starting batch made it through. [4]
Start with one million cells in a hypothetical batch.
If 60% survive the editing step, 600,000 cells remain.
If 88% of those survivors have the intended edit, multiply 600,000 by 0.88.
That leaves 528,000 surviving, correctly edited cells from the original million. Both percentages are assumed for the demonstration.
Small gains can add up. In this hypothetical batch of one million cells, raising survival from 60% to 65% while keeping the intended-edit rate at 88% among survivors would leave 44,000 more living cells with the intended DNA change. A lab could use its cell counts to pick one step to improve, such as how soon cells are moved into recovery medium after editing. Researchers could then split each starting batch and compare the usual process with that one change. The practical payoff would be more correctly edited beta cells at the end of the process that still release insulin when blood sugar rises.
A Different Approach — The Chemical Eraser
Base editing offers another way to make certain changes to DNA. [4]
Think of an adenine base editor as a chemical eraser and pencil. At a compatible site, it can change an A–T pair to G–C without cutting both strands of DNA. [4]
The demo compares two hypothetical batches of one million cells. These numbers are modeled examples.
Cas9 scenario. 52.8% of the starting batch survives with the intended edit, or 528,000 cells.
Base-editing scenario. About 96.3% survives with the intended edit, or about 963,000 cells.
A higher yield would leave more cells available for the next stages. Researchers would still need to show that the cells develop into functioning beta cells and can respond to changes in blood sugar.
What Early Human Studies Show
The studies below looked at different approaches. Each answers a different question about keeping transplanted cells working.
A different anti-rejection regimen. Tegoprubart is an investigational drug that suppresses part of the immune response. Researchers are testing it alongside other anti-rejection drugs in a regimen that avoids tacrolimus. Long-term safety and effectiveness are still being studied. [5]
An implant in the forearm. In one adult, genetically modified donor islets implanted in the forearm muscle showed signs of function at 14 months without immunosuppression. The result shows that the graft could function in that participant. [6, 7]
A teenage participant. In a separate three-person stem cell–derived islet report, a 15-year-old’s daily insulin dose dropped from 56.5 to 16 units at 52 weeks. A 45-year-old participant in the same report reached insulin independence by week 36. [8]
The Takeaway
Early studies show that transplanted islets can restore enough insulin production for some people with Type 1 diabetes to go without injected insulin for a period. The next questions are how long that benefit lasts, who can benefit, and how to protect the cells with fewer side effects. [1, 7, 8]
References
- Reichman et al., NEJM (2025), Stem cell–derived islets
- NIDDK, Type 1 Diabetes
- NIDDK, Diabetes in America, Monogenic Forms of Diabetes
- Gaudelli et al., Nature (2017), Adenine base editing
- Breakthrough T1D (October 2026), Tegoprubart and islet transplantation
- Carlsson et al., NEJM (2025), Forearm donor-islet transplant
- Carlsson et al., NEJM (2026), 14-month follow-up
- Shi et al., The Lancet Diabetes & Endocrinology (2026), Three islet transplant recipients
Shared for professional analysis. Please pick it apart — what’s sound, what’s overstated, and where should the analysis go next?
— Terri D. Goldson, Sr.
