Ultra-rapid lispro (Lyumjev)

Ultra-rapid lispro (URLi), sold as Lyumjev, is a rapid-acting mealtime insulin — taken at the start of a meal to cover the carbohydrates in it.¹ It is the same insulin molecule as Humalog (insulin lispro) but reformulated with two extra ingredients: treprostinil, which widens the small blood vessels under the skin, and citrate, which loosens the vessel walls so insulin can cross into the blood faster.⁵¹⁰ The goal is to chase the one thing standard analogs do poorly — get insulin working quickly enough to blunt the spike from a meal.

How fast is it? After a subcutaneous injection, insulin appears in the blood faster than Humalog (about 6 minutes faster in the clamp study) and reaches half its early concentration at roughly 13 minutes (versus ~25 minutes for Humalog).¹ In a euglycemic clamp study, insulin action started 11-12 minutes sooner than Humalog and was three-fold greater over the first 30 minutes.³ So it is fast at the front end, but it is still subcutaneous insulin — far slower than a healthy pancreas, which releases insulin into the blood the moment glucose rises.

The tail matters too. Because more of the dose is delivered early, less lingers late: exposure after 3 hours is cut by roughly 39-41%, and total duration is 72-88 minutes shorter than Humalog.³¹¹ Insulin action beyond 4 hours is 44-54% lower as well.³ A shorter tail means less "insulin stacking," with potentially fewer late post-meal lows.

Absorption variability and head-to-head speed. In a direct comparison against Fiasp (faster aspart), Humalog and NovoRapid, URLi reached early half-maximal concentration 6 minutes faster than Fiasp and 13-14 minutes faster than the standard analogs — and its early glucose curve most closely matched healthy controls.⁴ Like all injected insulin, absorption varies with site, dose and blood flow, so real-world timing is less tidy than clamp numbers.

Behaviour around exercise. For years the exercise case for ultra-rapid lispro was inferential — a shorter tail should mean less insulin on board when you start moving. A July 2026 Diabetes Care crossover RCT tested it head-on in 25 adults with type 1 diabetes on pump therapy, comparing URLi and lispro across two pre-exercise basal-rate-reduction strategies before an hour of walking.⁸ URLi significantly blunted the exercise glucose drop under both: -26.8 vs -39.0 mg/dL (-1.5 vs -2.2 mmol/L) with a 50% basal cut made 60 minutes ahead, and -46.9 vs -60.5 mg/dL (-2.6 vs -3.4 mmol/L) with a full 100% cut made 15 minutes ahead. Hypoglycemia was numerically less frequent with URLi — 6% of exercise sessions versus 16% with lispro — but that difference was descriptive, not statistically significant, so don't read it as a proven reduction. After a post-exercise meal, URLi's faster absorption also produced earlier post-meal glucose lowering. The trial was small (25 participants), so treat the size of the effect with caution — but the direction is now measured rather than assumed.

Meal timing. Usefully, the label allows dosing at the start of a meal or up to 20 minutes after it begins — handy for unpredictable appetites or kids.¹

Delivery and approvals. Lyumjev is given by injection (U-100 vials, cartridges or pens; U-200 pens only), by insulin pump (U-100), and intravenously with U-100 only under medical supervision.¹ The FDA approved it on 15 June 2020 for adults with type 1 and type 2 diabetes;⁶ the EU authorised it on 24 March 2020, now covering ages 1 and up.⁵ In the 26-week phase 3 PRONTO-T1D trial (1,222 adults), mealtime URLi was non-inferior to lispro on HbA1c and superior on 1- and 2-hour post-meal glucose; injection-site reactions were more common (2.9% vs 0.2%).²

Access and cost. Lyumjev is a branded insulin; Lilly's monthly out-of-pocket cap and savings cards lower the cost paid by many for Lilly insulins, but no Lyumjev-specific list-price or biosimilar source was verified in this review.⁷

What's coming. URLi is being studied for use inside automated insulin delivery (AID) systems, where its speed could tighten closed-loop control — in-silico work suggests pump settings must be re-tuned (faster insulin needs adjusted ratios) to capture the benefit.¹² But the core ceiling remains: even the fastest subcutaneous insulin can't fully solve unannounced meals, which is why we still track this against the insulin-speed gap.