Next-Generation Insulins: Better While the Cure Develops

What “next-generation” actually means

If you live with T1D long enough, you collect a small archive of insulins. The bottle in the fridge in 2005 is not the bottle in the fridge in 2026. Each step changed something concrete: the meal you could move, the run you could finish, the night you could sleep through. “Next-generation” is doing a lot of work in headlines about T1D, so it is worth being precise about what falls under it and what does not.

There are two parallel tracks. The first is prandial, the insulin you take with food. Over the past decade the meaningful change has been from rapid-acting analogues (aspart, lispro, glulisine) to faster-acting analogues (Fiasp, Lyumjev). Both are licensed and routine in many countries; the difference is measured in minutes, not hours, but those minutes change how the curve sits over a meal. The second track is basal, the insulin underneath. Here the arc has run from twice-daily NPH to once-daily glargine to ultra-long degludec, and most recently to a once-weekly basal (icodec) that has been trialled in T1D. Beyond both tracks sits the future-track: glucose-responsive (smart) insulins, designed to release more when the glucose is high and less when it is low. Smart insulins are early-phase; the families using them in 2026 are on trial pathways only.

The once-weekly basal that landed in T1D

When you have given yourself a basal injection every night for years, the idea of one injection per week sounds either very tempting or very foolhardy, depending on the day. The everyday case for it is obvious. Fewer injections is fewer interruptions; fewer interruptions is more space for the rest of life. The harder question is what a once-weekly basal does in T1D specifically, where insulin sensitivity changes meal-to-meal and day-to-day, and where overshooting once locks the dose in for a week.

The trial that puts numbers on this in T1D is ONWARDS 6 (Russell-Jones 2023, The Lancet). Five hundred and eighty-two adults with T1D, randomised to either once-weekly icodec or once-daily degludec on a treat-to-target protocol, both with mealtime aspart. After twenty-six weeks, HbA1c reduction was non-inferior between the two arms; once-weekly icodec held its own on glucose lowering. The headline that matters for a person living with T1D is the second result. Combined clinically significant or severe hypoglycaemia ran at 19.9 events per patient-year on icodec versus 10.4 on degludec. Roughly twice as much hypoglycaemia, in the same population, over the same period.

The mechanism makes sense. A basal insulin that lasts seven days is harder to titrate without overshooting in a condition where the insulin need can move ten or fifteen percent between Monday and Friday. Icodec is approved in many places for T2D, where the insulin requirement is generally steadier; the T1D regulatory conversation is ongoing. The honest read is that once-weekly basals are a real option that the field is now learning how to use, not a finished answer. For people for whom the daily injection is the wall, the trade-off may still favour weekly. For people whose hypoglycaemia is already where the daily life pinches, daily basal continues to make more sense. This is exactly the conversation to have with the diabetes care team, not to settle from a headline.

Faster-acting prandials, what they actually changed

The first time I gave a meal bolus with Fiasp instead of NovoRapid, the difference was small, then noticeable, then routine. Fiasp adds niacinamide to insulin aspart; Lyumjev adds treprostinil and citrate to insulin lispro. In both cases, the chemistry shifts the time-action curve forward by a few minutes. On a euglycaemic clamp the difference is measurable; in real life it shows up in two places. The post-meal peak is a little tighter, because more of the bolus is on board when the meal arrives at the gut. The pre-meal pause that the older rapid-acting insulins required (the ten or fifteen minutes between the bolus and the first bite) gets shorter, sometimes to nothing.

For families using bolus-driven mealtime dosing, those minutes are not academic. The pre-meal pause is one of the parts of T1D that does not show up on guideline tables but lands hard at every breakfast, every school dinner, every restaurant. Faster prandials soften that pause. They also fit closed-loop systems better, because the algorithm has less of a delay to predict around. They are not a different drug and they are not a fix; they are a shift in the curve. Some people switch and feel the difference inside a week; some switch and prefer to switch back, because the sharper early peak does not match how they eat. Insulin choice is a conversation with the team, not a one-way upgrade.

Smart insulins, why they matter even at early-phase

The hardest thing about insulin is that it does not know what you are doing. A bolus given for a slice of pizza behaves the same way whether you then sit down to a film or go for a hill run. The dose is fixed; the body’s need is not. Glucose-responsive insulins (often called smart insulins) are designed to dissolve that asymmetry. The molecule is engineered to release more insulin when glucose is high and less when glucose is low, so the dose self-modulates inside the body rather than being committed at the moment of injection.

It is important to be clear where this work sits in 2026. Smart insulins are in early-phase development. The molecules in the pipeline are at preclinical and Phase 1 stages; the families using glucose-responsive insulin in 2026 are on trial pathways only, not in routine clinic. That said, the concept matters even before clinical use, because of what it would change. A glucose-responsive insulin shifts the work from dose calculation to dose initiation. The mental load that runs underneath every meal, every workout, every late evening, the constant question of “is this dose right for what is about to happen”, would move from the person to the molecule. That is a different shape of T1D from the one any of us learned to live with. It is still some way off; the direction is honest.

What the parallel track means while the cure develops

You can hold two things at once. The cure is on its arc, with screening and staging carving out new windows for intervention, immunotherapy buying time at Stage 2, and beta-cell replacement starting to show what insulin independence looks like in trial cohorts. While that work is happening, the insulin underneath the daily life of T1D has not stopped moving. Faster-acting prandials are routine. Ultra-long basals are routine. A once-weekly basal has been trialled in T1D. Glucose-responsive insulins are in early-phase development. Both tracks are real; both are progressing; neither replaces the other.

The practical part of this is the part to take to clinic. Insulin choice is one of the few levers in T1D that genuinely belongs to the person and the team together. The diabetes team is set up for the conversation about which prandial fits how you eat, which basal fits how you sleep, and what trade-offs are worth thinking about between dosing convenience and hypoglycaemia. Ask early, ask in writing if you need to, and ask again if the first answer feels like a default rather than a fit. The daily life of someone with T1D in 2030 may look different even before any of the cure-research routes lands; some of that difference is already on the prescription pad.

References for this part