GNTI-122 engineered Tregs (POLARIS)
GentiBio
Promising idea, first-in-human, and open only to a rare genotype.
A one-time autologous engineered regulatory T-cell therapy made from a person's own blood cells, designed to restore immune tolerance to the cells that make insulin in recent-onset T1D. Phase 1 POLARIS is recruiting a small, genotype-restricted group (HLA-DRB1*04:01-positive adults); no efficacy data yet.
The scorecard
The goal is beta-cell protection, and POLARIS measures biomarkers including C-peptide, but no human efficacy data are posted yet.[2]
Engineered Tregs are intended as a durable immune-tolerance reset, but persistence and durability in T1D remain unproven until follow-up reports. The largest POLARIS cohort (10 of 16 participants) receives GNTI-122 together with rapamycin, specifically to try to extend how long the engineered cells survive — an admission that persistence is the open question.[2]
This is first-in-human autologous cell therapy in T1D; safety and tolerability are the primary endpoints, so risk is deliberately scored conservatively.[2]
Extremely narrow. POLARIS enrolls just 16 adults aged 18-55, diagnosed within 180 days, at US specialist sites — and participants must be HLA-DRB1*04:01 positive, a genotype restriction that rules out most people with T1D. This is a proof-of-mechanism population, not a preview of who could be treated.[2]
The full picture
GNTI-122 is the most concrete engineered-Treg entry now in human testing for type 1 diabetes. The product is made from the participant's own blood cells and engineered to counter the autoimmune imbalance that destroys beta cells. POLARIS is a small phase 1, open-label study, so every score is still about plausibility and maturity rather than demonstrated disease modification. The reason it matters is strategic: if engineered Tregs can survive, traffic correctly, and preserve C-peptide without broad immune suppression, they could become the immune-protection half of a future cure stack.
The trial itself is deliberately tiny. POLARIS (NCT06919354) plans 16 adults aged 18 to 55, diagnosed within 180 days of screening, at US specialist sites. It runs as three sequential dose cohorts: three people get a low dose of GNTI-122, three get a high dose, and the largest group — 10 of the 16 — gets the high dose together with rapamycin, an immunosuppressant included because it may help the engineered cells survive longer. It is a 78-week study, with primary completion listed as February 2028.
One caveat the headlines skip: to join POLARIS you must carry the HLA-DRB1*04:01 gene variant, because the engineered receptor is built around it. Most people with type 1 diabetes do not carry it. Whatever POLARIS shows, it will not immediately generalise to everyone with type 1 diabetes — a wider product would need receptors designed for other genotypes, which is more work, not a formality.
Two honest reads on that rapamycin cohort. The optimistic one: the team is being pragmatic about a known weakness of cell therapy, and pairing the cells with a drug that buys them time is sensible engineering. The sceptical one: needing rapamycin at all cuts against the central promise of engineered Tregs, which is protection without broad immune suppression. Which read is right is exactly what the trial exists to answer, and there are no efficacy data yet either way.
Coming soon
ETA · Phase 1 POLARIS recruiting (16 participants); 78-week study, primary completion February 2028
- →First POLARIS safety, cell-persistence and C-peptide findings · Primary completion February 2028 (registry)
Sources
- [1]GentiBio — GNTI-122 for new-onset type 1 diabetes · manufacturer
- [2]POLARIS phase 1 study of GNTI-122 in adults recently diagnosed with T1D (NCT06919354) · registry
- [3]GentiBio announces first participant dosed in POLARIS phase 1 clinical trial of GNTI-122 · manufacturer · 2026-04-14
- [4]Helmsley Trust and Breakthrough T1D support POLARIS trial of GNTI-122 · news · 2026-02-13