Tzield doesn’t cure type 1 diabetes. It doesn’t end the insulin pump. What it does, at a wholesale list price that runs roughly two hundred thousand dollars for a fourteen-day course, is reach into the autoimmune fire in a newly diagnosed kid’s pancreas and turn the flame down for a while. On June 12, the FDA granted accelerated approval for that use in children eight to seventeen diagnosed with Stage 3 type 1 diabetes within the previous six weeks. The catch is that the surrogate number the agency leaned on cleared its bar, while the outcomes families would actually feel on a glucose meter did not.
The immunology is the part I keep coming back to. Type 1 diabetes happens because the immune system mistakes the insulin-producing beta cells in the pancreas for invaders and kills them off, and by the time a child shows up in clinic with classic symptoms there is usually still a flicker of working beta cells left, a fading honeymoon. The longer those cells keep producing even a trickle of insulin, the easier it is to keep blood sugar from crashing or spiking. Teplizumab is a humanized anti-CD3 monoclonal antibody that binds to a receptor on T cells and tells the autoimmune attack to stand down, at least for a while. That mechanism has been the dream of T1D immunotherapy researchers for twenty years. Showing it slows beta-cell death in actual children, in a phase 3 trial, is a biological win.
Now the data. The new approval rests on the PROTECT trial, published in the New England Journal of Medicine in October 2023 by Eleanor Ramos and an industry-funded team. Three hundred twenty-eight kids aged eight to seventeen, all diagnosed with Stage 3 T1D within the previous six weeks, were randomized two-to-one to teplizumab or placebo and given two twelve-day courses of daily IV infusions twenty-six weeks apart. At week seventy-eight, the teplizumab kids had a stimulated C-peptide level 0.13 pmol/mL higher on average than the placebo kids, which the authors describe as roughly fifty-nine percent relative preservation of beta-cell function. Just under ninety-five percent of teplizumab kids still cleared the clinically meaningful 0.2 pmol/mL threshold, versus seventy-nine percent on placebo. Primary endpoint hit, p less than 0.001.
Here is where I came in skeptical and stayed there. C-peptide is a number on a lab printout. The endpoints a parent of a kid with T1D actually lives by are different: time-in-range on the glucose monitor, hemoglobin A1c, frequency of severe hypoglycemic events, total daily insulin dose. In PROTECT, the teplizumab arm had numerical trends in the right direction on those things, and not one of them reached statistical significance. The press release language is “favorable trends,” which is the kind of phrasing you reach for when you wanted a clean win and didn’t get one. The kids in the treated arm were still on insulin pumps. They were still counting carbs. They were still doing fingersticks. Their A1c, the number their endocrinologist actually looks at every three months, was not meaningfully different from the kids who got a saline drip.
That matters for how to read the approval. The FDA didn’t grant a regular approval here; it granted accelerated approval, the pathway used when a drug clears a surrogate endpoint that is “reasonably likely” to predict actual clinical benefit. C-peptide preservation is the surrogate. Whether that surrogate translates into kids living longer, hospitalizing less, going blind less, or losing fewer kidneys over a forty-year horizon is, in regulatory terms, a hypothesis. Sanofi has to run a confirmatory trial called BETA-PRESERVE to test it. That trial is enrolling now. Its results are years away. In the meantime the label is live and the bills will go out.
About those bills. When Provention launched Tzield, the wholesale acquisition cost ran around $14,411 per vial, or roughly $194,000 for the fourteen-day course at the original Stage 3-delay dosing; the PROTECT regimen of two twelve-day cycles is a longer course again. Most of that lands on commercial insurance, on employer-sponsored health plans, and on state Medicaid programs, which is to say on the rest of us. Sanofi paid $2.9 billion in 2023 to acquire Provention Bio for this single asset, betting that broadening the label, from delaying Stage 3 in at-risk Stage 2 patients to treating newly diagnosed Stage 3 kids, would justify the price tag. Commercial uptake has been thin enough that each new approval matters to the company as a survival event for the franchise, not a clinical update. The label widens. New insurance codes follow. Read uncharitably, the strategy is to make the price stick before the confirmatory trial reads out and questions get harder to wave off.
The safety picture deserves a look, because it is part of the calculation a family is being asked to make. In PROTECT, the most common adverse reactions in the teplizumab arm were lymphopenia (a drop in immune cells, expected for a drug whose entire job is to suppress T-cell activity), vomiting, rash, leukopenia, diarrhea, neutropenia, elevated liver enzymes, and headache. Serious adverse events ran 5.5 percent on teplizumab and 5.4 percent on placebo, basically a wash, though cytokine release syndrome, the brief immune storm the drug can provoke during the infusion, showed up in 1.4 percent of treated kids and zero placebo kids. The post-marketing picture has begun to fill in: a FAERS pharmacovigilance analysis published in Clinical Therapeutics in 2025 flagged signals for lymphopenia, abnormal liver enzymes, cytokine release syndrome, and injection-site infections, mostly tracking what the label already warns about, and an ADA Scientific Sessions abstract from the same year is moving in the same direction. These are not catastrophic findings. They are also not weightless for a drug given to children to delay a number on a lab report.
There is a version of this story where the regulators got it right. T1D is a serious, lifelong autoimmune disease, the unmet need is severe, the residual beta cells are worth saving, the immunology is sound, and an accelerated approval lets families with a newly diagnosed kid reach a disease-modifying option years before the confirmatory trial reads out. That version is defensible. There is another version, the one Sanofi’s commercial story implies, where a $2.9 billion acquisition that has been slow to pay back finally gets the pediatric label expansion that unlocks a wider prescribing base, on the strength of a surrogate endpoint that won and clinically meaningful endpoints that didn’t. Both versions are true at the same time. The FDA gets credit for opening earlier access to a mechanism that does what its biology promises, and the company gets credit for pricing that mechanism at a level the benefit, as currently measured, has not earned.
My honest take: I would want my kid in this trial if the alternative was nothing. I would not want my kid on this drug at this price, on this evidence base, if the confirmatory trial has not yet shown the C-peptide curve translates into anything I would actually notice with a glucose meter. And I would want to know, before signing the consent form, that a number going up on a lab printout is not the same as a child being meaningfully less sick. The accelerated-approval pathway exists precisely to bridge that gap with a follow-on trial. Until BETA-PRESERVE reports, the honest reading is that the mechanism won a surrogate at a price the family-level evidence has not yet earned.
Sources
- FDA – accelerated approval announcement for Tzield in pediatric Stage 3 T1D, June 12, 2026
- Sanofi – press release on the US Tzield Stage 3 pediatric approval, June 2026
- NEJM – Ramos et al., PROTECT trial of teplizumab in newly diagnosed type 1 diabetes, October 2023
- Clinical Therapeutics – real-world FAERS pharmacovigilance assessment of teplizumab, 2025
- Diabetes (ADA Scientific Sessions) – Lakshmi et al., pharmacovigilance abstract on Tzield adverse reactions, 2025
- Pharmaphorum – Provention Bio sets Tzield list price at ~$194,000 for a 14-day course
- Sanofi – agreement to acquire Provention Bio for $2.9 billion, March 2023