- βAfrezza inhaled insulin already available and faster than any injectable option for qualifying patients
- βArecor's AT247 shows Phase 2 evidence of faster peak action than FiAsp with potential Phase 3 advancement
- βGlucose-responsive 'smart' insulins in development could eliminate hypoglycemia risk from meal-timing errors
The Problem With Even Our Fastest Insulins
FiAsp (faster aspart, made by Novo Nordisk) was a genuine leap forward when it launched β acting roughly 5 minutes faster at onset than standard NovoLog, with a steeper early peak. Lilly's Lyumjev (ultra-rapid lispro) followed, using citrate and treprostinil to speed absorption in a similar way. Both are excellent insulins. But here's the uncomfortable truth: even these "ultra-rapid" options still can't fully replicate what a healthy pancreas does.
In people without diabetes, beta cells sense rising glucose and release insulin directly into the portal vein within 1β2 minutes. Peak insulin action arrives in 30β60 minutes, precisely matching the glucose surge from a meal. FiAsp and Lyumjev, even injected 20 minutes before eating, still lag behind that curve. The result: post-meal glucose spikes that are higher and longer than they should be β a stubborn problem that drives A1C up and increases cardiovascular risk over time.
The race is now on to close that gap with an entirely new generation of formulations.
The Science of "Faster": What Researchers Are Actually Targeting
Injectable insulins face a fundamental bottleneck: they're injected into subcutaneous fat, where they must dissociate from hexamers (clumped molecules) into monomers before absorption. This dissociation step is the primary brake on speed. Current ultra-rapid insulins use additives like niacinamide (FiAsp) or citrate/treprostinil (Lyumjev) to accelerate this process β but only partially.
Next-generation approaches attack this problem from multiple angles:
- Glucose-responsive "smart" insulins that activate only when glucose rises
- Novel excipient technology to enhance subcutaneous dispersion and vascular uptake
- Inhaled insulin reformulations bypassing subcutaneous tissue entirely
- Intradermal delivery using microneedle arrays to deposit insulin into the highly vascularized dermis
Key Programs to Watch in 2025
Arecor's AT247 (Arefast)
UK-based Arecor Therapeutics has developed AT247, a concentrated ultra-rapid aspart formulation using their proprietary Arestat stabilization platform. In Phase 2 trials, AT247 demonstrated faster time-to-peak and lower 1-hour postprandial glucose compared to FiAsp in Type 1 patients. Arecor has published data showing AT247 achieves earlier insulin appearance in circulation, with the PK/PD profile suggesting meaningful clinical advantages over current ultra-rapid options. Arecor is actively seeking development partners for Phase 3 advancement.
Mannkind's Afrezza β The Inhaled Option That Already Exists
Afrezza (inhaled insulin, MannKind Corporation) is technically already available and is genuinely faster than any injectable option. Delivered via a small inhaler as Technosphere insulin, it reaches peak action in approximately 12β15 minutes and clears by 3 hours β closely mimicking physiological kinetics. The STAT study and multiple PK/PD analyses confirm it outperforms FiAsp on speed metrics. The limitations are real: lung function monitoring is required, it's not approved for smokers or those with asthma/COPD, dosing is in fixed 4-unit increments (limiting fine-tuning), and insurance coverage remains inconsistent. But for patients who qualify, it's the closest thing to physiological speed currently available in the US.
Glucose-Responsive Insulins: The Holy Grail
Multiple research groups, including teams at JDRF-funded academic centers and companies like Zafgen (now Larimar Therapeutics has pivoted) and Novo Nordisk's internal programs, are pursuing "smart" insulins that activate in response to elevated glucose and become inactive when glucose normalizes. This approach would eliminate hypoglycemia risk from meal-timing errors β arguably a bigger clinical win than raw speed alone. These programs remain largely in preclinical or early Phase 1 stages as of 2025. Realistic timelines for approval are 7β10 years minimum.
Microneedle and Intradermal Delivery
Academic and startup research into microneedle patch delivery continues to show promise in early studies, with intradermal insulin consistently showing faster absorption than subcutaneous injection. No product has reached Phase 3 as of 2025, but this remains an active space.
Current Status (2025)
Honestly assessed: FiAsp and Lyumjev remain the fastest approved injectable insulins for most patients. Afrezza is faster but comes with meaningful practical limitations. AT247 is the most clinically advanced next-generation injectable in the pipeline but has not completed Phase 3. Glucose-responsive insulins remain years away. No new ultra-rapid injectable has received FDA approval since Lyumjev in 2020.
Timeline
- Now (2025): FiAsp, Lyumjev, and Afrezza are available. Afrezza is the fastest option for qualifying patients.
- 2026β2028: AT247 or similar next-generation injectables could reach Phase 3 completion and regulatory submission if development partnerships are secured.
- 2030+: Glucose-responsive "smart" insulins remain a longer-horizon target.
What This Means for Patients
If post-meal spikes are your main challenge, have a direct conversation with your endocrinologist about pre-bolusing timing with your current rapid insulin β even FiAsp performs significantly better when injected 15β20 minutes before meals rather than at mealtime. If you're on a closed-loop system (like Control-IQ or Omnipod 5), the algorithm partially compensates for insulin lag, making ultra-rapid speed somewhat less critical.
Patients interested in participating in clinical trials for next-generation insulins can search ClinicalTrials.gov using terms like "ultra-rapid insulin" or "faster insulin formulation" to find enrolling studies. Staying connected through resources like mdsdiabetes.com can help you track what supplies and technologies are becoming available as this field moves forward.
The honest message: the tools we have today are genuinely good, and the pipeline is real β but don't hold off on optimizing your current regimen while waiting for the next breakthrough.
