Insulin was discovered in 1921. For most of the century since, "managing diabetes" meant injecting twice a day and hoping for the best.
We are not in that century anymore. The gap between "lab idea" and "clinic" is narrowing fast. Several treatments that would have sounded like science fiction ten years ago are now in late-stage trials or already entering practice. Here's the short, honest map.
1. Once-weekly insulin
The most immediately practical change for type 2 diabetes is insulin icodec — a long-acting insulin you inject once a week instead of once a day. Approved in Europe in 2023; rolling out in clinics now.
It works by binding to a protein in your blood that releases the insulin slowly over seven days. Trials show it controls blood sugar as well as daily insulin, with similar hypoglycaemia rates. Big deal? It cuts injections by 85%. For patients whose biggest barrier to starting insulin is the daily needle and the daily reminder, that's a meaningful change.
2. Beyond Ozempic
GLP-1 drugs (Ozempic, Wegovy, Mounjaro) have already changed the landscape of type 2 diabetes and obesity. They mimic a gut hormone that boosts insulin release, slows digestion, and quiets appetite.
But the pipeline has already moved past them. Retatrutide targets three hormone receptors at once. In Phase 2 trials, the highest dose produced 16.8% body weight loss in 24 weeks — meaningfully more than current dual agonists. Phase 3 trials are running. If they hold up, this becomes the most effective non-surgical weight-loss and glucose-lowering drug ever made.
Cost and access — not effectiveness — will be the constraint.
3. Lab-grown beta cells (the path to a cure for type 1)
Type 1 diabetes happens because the immune system destroys the cells in your pancreas that make insulin. The dream has always been to put those cells back.
Vertex Pharmaceuticals' zimislecel takes embryonic stem cells, differentiates them into functional insulin-producing cells, and transplants them. In the FORWARD trial, 10 of 12 patients who received a full dose were insulin-independent at one year.
Insulin-independent. Not "controlled better." Not "less insulin." None.
These are small numbers from an early trial — but the trajectory is striking. Phase 3 is running. Regulatory submissions could land around 2026.
The catch: recipients still need immune-suppressing drugs to stop their bodies rejecting the new cells. So zimislecel will mostly be appropriate, in the short term, for patients whose diabetes is already extremely hard to control. The next research step — engineering cells the immune system can't see — is in earlier stages.
4. Smart insulin
Today's insulin is a fixed dose. Too much → hypo. Too little → hyperglycaemia. The patient (or an algorithm) does the calibration.
Glucose-responsive "smart" insulin would skip that calculation entirely. Engineered with a switch that activates only when blood sugar is elevated and shuts off as it normalises — meaning, in principle, an insulin that can't cause hypoglycaemia.
Animal data is promising. Human trials are still several years out. But this would be the largest safety advance in insulin therapy since 1921.
What this means today
If you have type 1 diabetes and you're struggling with control, ask your diabetes team about hybrid closed-loop systems (the artificial pancreas) — already in clinical use, with real evidence behind them.
If you have type 2 diabetes and haven't been assessed for GLP-1 therapy, ask your doctor. The eligibility criteria have widened a lot in the last few years.
If you're watching the beta-cell pipeline, temper optimism with realism: clinical availability in most countries is at minimum three to five years out, even if trials succeed.
The hundred years since insulin added decades to the lives of people with type 1 diabetes. The next decade may begin to remove the disease itself.
Curiosity first. — Dr. Brugal