Curing autoimmune diseases
Over 100 diseases arise when the immune system attacks the body; drugs only suppress it, while CAR-T cell resets show remission in small studies.
Open in the interactive tree →In autoimmune diseases such as lupus, multiple sclerosis, type 1 diabetes and rheumatoid arthritis, immune cells attack the body's own tissue. Standard treatment dampens the immune system for life and leaves patients open to infections. A cure would remove or retrain only the cells that cause the attack, and ideally stop disease before symptoms start.
As of October 2026
A Mayo-led study published in 2024 found 15.4 million Americans (4.6 percent) diagnosed with at least one of 105 autoimmune diseases, 63 percent of them women. In January 2026 Nature Medicine reported the CASTLE trial: 22 of 24 patients with lupus, systemic sclerosis or myositis responded at six months after one infusion of CD19 CAR-T cells, off immune drugs, and a May 2026 report on a pooled analysis of 145 lupus patients in 16 studies gave 84 percent drug-free remission, though relapses and one death from meningitis occurred. A 2025 NEJM letter described the first in-body CAR-T (mRNA in lipid nanoparticles) in five lupus patients, with B-cell depletion and better disease scores in most but little response in two at three months. Teplizumab, which delays the onset of type 1 diabetes in people with early-stage disease, was extended in April 2026 to children from age 1.
What is missing
- Proof that remission lasts for years, from larger trials with control groups
- Safer and cheaper ways to reset the immune system, without chemotherapy or cell factories
- Ways to switch off only the harmful cells and keep protection against infection
- Tests that find people at risk before organs are damaged
- Understanding what sets off autoimmunity in the first place
Becomes possible once solved
- Drug-free remission for diseases now managed for life
- Screening and early treatment before symptoms
- Fewer infections and side effects from lifelong immune suppression
- Treatment of rare diseases that have few options today
Open steps
- CAR-T made inside the body Medium AI leverageDeliver CAR instructions to T cells with injected particles, so that no cell harvesting, factory or chemotherapy is needed.
- Who relapses after a reset? Medium AI leverageFind early signs in blood, such as returning B-cell types or receptor sequences, that predict who stays in remission after CAR-T or other resets.
- Switch off only the harmful cells Medium AI leverageIdentify the self-antigens and receptors behind a disease and build treatments that remove or silence just those cells, keeping protection against infection.
- Treat before the first symptom Medium AI leverageUse autoantibody and risk tests to find people in the silent stage and test early treatment, with type 1 diabetes and teplizumab as the model.
- Why immunity turns on the body Medium AI leverageFind the triggers, such as infections, genes and environment, that start autoimmunity, so it can be prevented and not only treated.
Where AI could help
Medium AI leverage. AI sorts immune data and helps pick targets, but remission has to be shown by years of follow-up in patients.
- Reading T-cell and B-cell receptor sequences to find disease-driving cells
- Designing and screening cell-therapy targets and delivery particles
- Risk models that flag people before diagnosis
Shown so far
- A 2025 preprint (not peer reviewed) reported a deep-learning model that separated lupus and rheumatoid arthritis patients from blood T-cell receptor sequences, with AUCs of 98.95 and 97.76 percent. source