Disease-Modifying Therapy: Stopping the Immune Attack

What “disease-modifying therapy” means in T1D

In paediatric diabetes clinic, the conversations about T1D have always sat in two boxes. The first box is about how the body is behaving today: insulin doses, hypos, sick days, school plans, the AID system overnight. The second box, until recently, was about the future of the condition itself, and that box was almost empty. Insulin kept the person alive; nothing on the shelf changed the underlying immune attack on the insulin-producing cells. The first medicine to crack that second box open is teplizumab.

Disease-modifying therapy is the field’s name for treatments that act on the autoimmune process itself, rather than on the consequences of it. It is a careful phrase, because it draws a deliberate line between what these medicines can do and what they cannot. They do not cure T1D. They do not regenerate insulin-producing cells. What they appear to do, on the evidence to date, is dampen the immune attack for long enough that the pancreas keeps more of its insulin-producing capacity for longer. The TrialNet TN-10 trial published in the New England Journal of Medicine (Herold 2019) randomised 76 relatives at high risk (two or more autoantibodies plus dysglycaemic OGTT, Stage 2) to a single 14-day course of teplizumab or placebo and found a median delay to clinical T1D of around two years versus the placebo arm. That trial is the basis on which the FDA approved teplizumab (Tzield) in November 2022 to delay the onset of Stage 3 T1D in people aged 8 and older with Stage 2 disease.

The reason the language matters is that families read the difference. “Delay” is honest; it tells you what the medicine has been shown to do. “Prevent” is not yet supported by the evidence. “Cure” is a different category of claim and belongs to a different part of the pipeline. The conversation in clinic gets clearer when the words match the data.

The wider pipeline, in plain English

Teplizumab is the medicine furthest along the regulatory road, and the one most families hear about first. It is not the only medicine in the conversation. The wider pipeline contains agents at different stages of evidence, each acting on a different point in the immune system, with different routes (an infusion in clinic, a subcutaneous injection at home, a tablet at breakfast) and different cost profiles. The shape of the pipeline matters, because what is licensed today is a snapshot, and the snapshot is changing.

Each agent is doing something different to the immune system. Teplizumab is a humanised monoclonal antibody against CD3, a marker on T cells; it is given as a single 14-day course of IV infusions. Golimumab blocks TNF-alpha, an inflammatory signal involved in beta-cell destruction, and it is given as a subcutaneous injection over a year (Quattrin 2020 reported preserved C-peptide at week 52 in newly diagnosed paediatric T1D). Verapamil is an off-patent oral calcium-channel blocker that has been used for high blood pressure for decades; the CLVer trial in newly diagnosed paediatric T1D (Forlenza 2023) reported around a 30% relative increase in C-peptide at one year on a tablet that costs pennies a day. Abatacept blocks the co-stimulation signal T cells need to activate; anti-thymocyte globulin depletes T cells more broadly. The historic Phase 2 trials of those last two are the field’s earlier evidence base, with newer combination work ongoing.

What “Phase 2” and “Phase 3” mean for a family is worth being plain about. Phase 2 trials test whether a medicine appears to do what it is supposed to in a small group, and whether the side-effect profile is broadly acceptable. Phase 3 trials are larger, longer, and designed to support regulatory approval. “Licensed” means a regulator has reviewed the trial evidence and authorised the medicine for use in a defined population. Teplizumab is the only agent on this page in the licensed box, and only for one indication (Stage 2 T1D, age 8 and over, in the United States). The rest sit further back along the road.

What “delay” actually means, and what it does not

The TN-10 result is the one that gets quoted most often, and it is worth being careful with the numbers. In the original trial, the median time from randomisation to clinical T1D was around 24 months on placebo and around 48 months on teplizumab; a single 14-day course of the medicine roughly doubled the median diabetes-free period. The extended follow-up of the same cohort (Sims 2021) reported the gap holding open: median time to T1D of around 60 months on teplizumab versus 27 months on placebo, with half the treated participants still diabetes-free at the longer follow-up against around 22% in the placebo group.

“Median” is a number that is easy to misread. It is the middle of the distribution, not a promise to any one person. Some participants in the placebo arm progressed within months; some teplizumab participants progressed quickly too. Some teplizumab participants were still diabetes-free at five years and beyond. The median delay describes the population; what happens to one child sits inside that population somewhere, and the position is not predictable in advance. Sims and colleagues did report a mechanistic correlate, a specific pattern of partially-exhausted CD8 T cells that tracked with how much benefit each individual got, and that gives the field a starting point for thinking about who responds best. That work is ongoing, not yet a clinic-ready test.

The side-effect profile on TN-10 was what you would expect from a T-cell-modulating therapy. Lymphocyte counts dropped sharply in the first week and recovered by around six weeks in all but one participant. Around a third of participants had a transient skin rash. Around half of the teplizumab participants who were already EBV-seropositive had detectable EBV DNA in the weeks after the infusion, all of which resolved without treatment. None of these were trivial; none were catastrophic in the trial cohort either. Long-term safety in real-world Stage-2 populations is still being mapped through post-marketing surveillance; “we know what happens for ten years across hundreds of thousands of people” is not yet what we have.

Where the access lives, and where it does not

Where the medicine is licensed and where it is available are not the same map. Teplizumab is FDA-approved in the United States under the brand name Tzield (November 2022) for delaying Stage 3 T1D in people aged 8 and older with Stage 2 disease. In the United Kingdom and Europe, it is not yet licensed for this indication. Access in the UK at the moment runs through trial pathways (where a child meets eligibility criteria for a research study) or through named-patient routes (where a clinician applies for individual access on the basis of clinical need). Both routes exist; both are decided case by case, with input from a specialist paediatric diabetes service.

The honest gap is between evidence and access. The evidence has moved faster than the regulatory and commissioning systems have. That gap is not a reason for any one family to delay a screening conversation or a Stage-2 conversation, but it is a reason to ask in writing, ask early, and ask again if the first answer is “not yet”. The team is set up to work through this; the system rewards families who advocate.

What I would want to know, before deciding

I will say this directly. I have not been offered teplizumab. My children have not been screened at Stage 2. So I am writing this from inside the conversation, not from inside the decision, and I want to be honest about that. Until you are sitting with the actual letter from the clinic, the actual side-effect leaflet, the actual access route in your country, the actual age and antibody profile of your child, you are guessing. So is everyone else. The point of this page is not to tell you what I would do; it is to lay out the things I would want to know, if it were me.

Things I would ask the team. What does the access pathway look like here, today, for someone at this stage and this age? What does the day of the infusion feel like for a child, and what do the two weeks after it look like at home, at school, with siblings? What is the realistic side-effect profile in the first month, and what is being monitored at six months, twelve months, three years out? What does “did not respond” look like, and what is the next conversation in that case? If we proceed and progression to Stage 3 happens at year two anyway, what does the team think we will have gained by then, given the rest of the technology and adjunctive landscape?

The other piece of honesty is that “delay” is not a small thing. An extra year before a four-year-old’s family has to absorb pumps and CGMs and carbohydrate counting is a different family year. An extra two years on the technology curve is more time for the next-generation pivotal to land. Both readings are reasonable. Neither is automatic. The team that has known your child’s pancreas longest is the team whose job it is to help you weigh this.

References for this part