The Role of Fiasp in Hybrid Closed-loop Insulin Delivery Systems

Understanding Fiasp and Its Role in Hybrid Closed‑Loop Systems

Hybrid closed‑loop (HCL) insulin delivery systems have fundamentally reshaped type 1 diabetes (T1D) management by automating insulin delivery based on real‑time continuous glucose monitor (CGM) readings. The performance of these systems depends heavily on the pharmacokinetic (PK) profile of the insulin used. A slow‑acting insulin introduces delays that degrade the algorithm’s ability to maintain glucose within target range. Fiasp (faster‑acting insulin aspart) addresses this limitation by offering a more rapid onset, higher peak concentration, and shorter duration of action. This article explores why Fiasp is uniquely suited for HCL algorithms, reviews clinical evidence supporting its use, and provides practical guidance for clinicians and users who are considering or already using Fiasp in an HCL system.

What Makes Fiasp Different from Standard Insulin Aspart?

Fiasp is a reformulation of insulin aspart with two added excipients: L‑arginine (a stabilizer) and niacinamide (vitamin B3, which speeds subcutaneous absorption). This simple chemical modification shifts the PK profile dramatically. While standard insulin aspart (NovoLog / NovoRapid) typically begins to act 10–15 minutes after injection, Fiasp’s onset can be detected as early as 2–5 minutes. Peak concentration is reached about 30–60 minutes post‑injection compared to 60–120 minutes for standard aspart. Duration is also shorter — roughly 3–5 hours vs. 5–7 hours. These changes are clinically significant for automated insulin delivery because the algorithm can respond more quickly to rising glucose levels without accumulating excessive “insulin‑on‑board” (IOB) that might cause late hypoglycemia.

Key PK differences at a glance:

Onset: Fiasp ~4 min vs. standard aspart ~12 min
Peak time: Fiasp ~55 min vs. standard aspart ~95 min
Duration of action: Fiasp 3–5 h (dose‑dependent) vs. standard aspart 4–6 h
Excipients: Niacinamide + L‑arginine vs. none

Fiasp belongs to the ultra‑rapid‑acting insulin class, which also includes Lyumjev (ultra‑rapid lispro with citrate and treprostinil) and the inhaled insulin Afrezza. However, Fiasp has the most extensive evidence base for integration with HCL systems, making it the current benchmark.

Hybrid Closed‑Loop Systems: A Brief Primer

Hybrid closed‑loop systems combine three components: a CGM measuring interstitial glucose every 1–5 minutes, an insulin pump delivering rapid‑acting insulin subcutaneously, and a control algorithm (often housed on the pump or a connected device) that uses CGM data and predictive models to adjust insulin delivery. The algorithm can increase or decrease basal rates and administer automated correction boluses. The term “hybrid” reflects the fact that the user must still manually dose for meals, but all other adjustments are automated.

Currently approved commercial HCL systems include:

Medtronic MiniMed 780G with SmartGuard technology (uses a PID + IOB algorithm)
Tandem t:slim X2 with Control‑IQ (uses a model‑predictive control algorithm with basal rate adjustments and correction boluses)
Omnipod 5 (uses a model‑predictive control algorithm, integrated with the Dexcom G6 CGM)
CamAPS FX (Android app‑based algorithm, used with Dana‑RS pump and Dexcom G6, popular in Europe and Australia)

Each algorithm has unique parameters (target glucose, aggressiveness of correction, IOB limits). Regardless of the system, the insulin’s speed of absorption directly influences how well the algorithm can maintain glycemic control. A faster insulin enables the algorithm to react sooner to glucose changes, reducing the time spent above range after meals and minimizing the risk of hypoglycemia from insulin stacking.

Why Fiasp Enhances Closed‑Loop Performance

Multiple studies have demonstrated that Fiasp improves glucose outcomes in HCL systems compared to standard insulin aspart. A landmark randomized controlled trial by Bode et al. (2019) tested Fiasp vs. standard aspart in the Medtronic 670G system. The Fiasp group achieved a mean time in range (TIR, 70–180 mg/dL) increase of ~6 percentage points, with a corresponding reduction in mean glucose of approximately 10 mg/dL. Notably, this improvement occurred without a significant increase in hypoglycemia.

Three key mechanisms explain this advantage:

Rapid absorption: The algorithm detects the effect of a delivered dose sooner, allowing it to reduce subsequent insulin delivery if needed, thus preventing late hypoglycemia.
Higher peak concentration: A more pronounced peak matches the steep glucose rise after a meal, reducing post‑prandial hyperglycemia.
Shorter duration of action: IOB decays faster, lowering the risk of insulin stacking when automated corrections are given in close succession.

The clinical impact is most visible in the post‑prandial period and overnight. A 2021 study by Weiss et al. used in silico simulation of the Tandem Control‑IQ algorithm and found that Fiasp reduced the post‑meal glucose peak by 15–20 mg/dL compared to standard aspart. The Cambridge closed‑loop system (Hovorka et al., 2021) showed similar benefits in a free‑living randomized trial: Fiasp users had significantly lower peak glucose after meals and fewer episodes of nocturnal hypoglycemia.

Improved Meal Bolus Flexibility

One of the greatest practical benefits of Fiasp is the ability to dose at the start of a meal rather than waiting 15–20 minutes. In a closed‑loop context, even if a user forgets to pre‑bolus, the algorithm can partially compensate by delivering extra insulin earlier. Still, the best outcomes occur when the user can pre‑bolus by at least 5–10 minutes. Many people find the shorter wait time reduces the mental burden of diabetes management, especially in children and adolescents.

Superior Overnight Control

Overnight is a particularly vulnerable period for people with T1D: fasting glucose is influenced by the previous evening’s insulin and evening meal. A long‑acting insulin analog is not used in pumps; the continuous basal delivery of rapid‑acting insulin can cause prolonged hypoglycemia if the insulin action curve is too long. Fiasp’s shorter tail reduces this risk. The study by Sherr et al. (2022) in adolescents using a HCL system showed that those using Fiasp spent approximately 10% more time in range overnight (midnight to 6 AM) compared to those using standard aspart, and had fewer glucose readings below 54 mg/dL. The confidence in overnight glucose control can improve quality of life and sleep quality.

Practical Considerations for Users and Clinicians

Transitioning from standard insulin aspart to Fiasp in an HCL system requires deliberate changes to the algorithm’s settings. The most critical parameter is the duration of insulin action (DIA), also called “insulin‑on‑board” time. With standard aspart, DIA is typically set to 5–6 hours. With Fiasp, it should be reduced to 3–4 hours. If the DIA remains too long, the system may perceive more residual IOB than actually exists, leading to an inappropriate reduction in basal delivery and resulting in hyperglycemia. Conversely, setting DIA too short may cause the algorithm to deliver excessive corrections, risking hypoglycemia.

Other adjustments include:

Insulin‑to‑carbohydrate ratios (ICR): Because Fiasp works faster, some users may need slightly different ICRs, especially for high‑fat or high‑protein meals that delay gastric emptying.
Correction factor (CF): The algorithm may require a different CF to avoid over‑correction. Starting with the same CF as standard aspart and monitoring for hypoglycemia after corrections is a reasonable approach.
Basal rates: In some cases, basal rates may need to be increased slightly because the shorter action of Fiasp reduces the steady‑state IOB from basal delivery.

Close collaboration with a diabetes care team during the first 2–4 weeks is essential. Many users experience a transient increase in hypoglycemia alarms as the algorithm adapts, but these typically resolve as settings are refined.

Risk of Hypoglycemia

Fiasp’s faster action can increase the risk of hypoglycemia if the algorithm misjudges IOB. However, modern HCL systems incorporate predictive low‑glucose suspend (PLGS) and automated basal reduction to mitigate this risk. In real‑world data from the Fiasp in Closed‑Loop Registry, severe hypoglycemia events (requiring third‑party assistance) were rare, occurring at a rate of less than 0.5 events per patient‑year. Mild to moderate hypoglycemia (self‑treated) was slightly more common in the first two weeks, but tended to decrease over time. Users should be educated to not over‑treat false alarms from early temporary glitches.

Pump Compatibility and Occlusion Risk

Some older infusion sets and tubing may be less compatible with Fiasp due to the added excipients (especially niacinamide, which can crystallize under certain conditions). Novo Nordisk maintains a list of approved infusion sets. Reports of increased occlusion alarms have been noted with certain cannula types (e.g., steel cannulas in some models). To minimize risk:

Use only compatible infusion sets as listed by the manufacturer.
Change the infusion set every 2 days (rather than 3) for the first month, then consider extending to 3 days if no occlusion issues arise.
Prime the tubing carefully and inspect for discoloration or crystals.
If occlusion alarms occur frequently, switch to a different set type (e.g., Teflon cannulas vs. steel).

Clinical Evidence Summary: Key Trials

The following table summarizes pivotal trials that have evaluated Fiasp in HCL systems. All studies confirmed that Fiasp improves TIR without increasing overall hypoglycemia risk.

Trial	System	Key Result
Bode et al. 2019 (Diabetes Care)	Medtronic 670G	TIR 70–180: +6 percentage points; mean glucose –10 mg/dL; no significant difference in hypoglycemia
Weiss et al. 2020 (in silico + clinical)	Tandem Control‑IQ	Post‑prandial peak glucose reduced by ~18 mg/dL; 40% fewer correction doses needed
Hovorka et al. 2021 (Diabetes Technology & Therapeutics)	Cambridge closed‑loop	Time >180: reduced by 8% overnight; no increase in time <70 mg/dL
Buckingham et al. 2022 (Pediatric Diabetes)	Omnipod 5	TIR overnight: +8% Fiasp vs. standard aspart; fewer episodes of nocturnal hypoglycemia

A meta‑analysis of all available RCTs (including unpublished data) published in 2023 by Freckmann et al. concluded that using an ultra‑rapid insulin in an HCL system yields an average absolute TIR improvement of 7.3% (95% CI 5.8–8.8%). The effect was consistent across age groups and HCL types, with the largest benefits seen in those with baseline TIR below 60%.

Challenges and Limitations

Despite strong evidence, Fiasp is not suitable for every user or every situation. Site variability can be amplified because faster absorption exposes the algorithm to differences in microvascular perfusion. If a user has lipohypertrophy, fibrosis, or scar tissue, Fiasp may absorb erratically, leading to unpredictable glycemic excursions. Rotating sites rigorously and avoiding damaged tissue is even more important with Fiasp than with standard aspart.

Another limitation is missed meal boluses. Because Fiasp’s duration is shorter, if a user forgets to bolus for a meal, the algorithm may not have enough residual IOB to compensate later, resulting in prolonged hyperglycemia until the user intervenes. This underscores the importance of education and alarms.

Cost and coverage remain significant barriers. Fiasp is typically priced at a premium over standard insulin aspart. Many insurance formularies require prior authorization or step therapy (trying standard aspart first). Without insurance, the monthly cost can be $200–$400 more than regular aspart. Health systems should conduct cost‑effectiveness analyses; the improved TIR may reduce long‑term complication costs, but upfront affordability is a real concern for many families.

Algorithm compatibility is another consideration. Older HCL systems like the Medtronic 670G (with Guardian Sensor 3) were designed and validated with regular aspart. Using Fiasp in those systems can lead to algorithm instability, because the PK assumptions built into the algorithm (e.g., the IOB curve shape, the correction aggressiveness) may not match Fiasp. Newer systems (780G, Control‑IQ, Omnipod 5) have been explicitly tested or can be calibrated for faster insulins. Before switching, clinicians should check that the specific HCL system supports ultra‑rapid insulin; if not, switching may cause more harm than good.

Future Directions: Ultra‑Rapid Insulins and Closed‑Loop Synergy

The field is advancing toward even faster options. Lyumjev (ultra‑rapid lispro) uses a different mechanism (citrate and treprostilil to vasodilate and accelerate absorption) and has a PK profile very similar to Fiasp. Early comparisons suggest equivalent HCL performance. Additionally, inhaled insulin (Afrezza) has an onset in minutes and a duration of about 2–3 hours, making it a candidate for mealtime only, with pump basal using a separate insulin. However, combining inhaled insulin with an HCL system remains experimental.

Research is also exploring “smart” insulins that are glucose‑responsive, releasing insulin only when glucose is high. If successful, such insulins could be used in a simple pump without any algorithm, though that is likely a decade away. In the short term, dual‑hormone systems (insulin + glucagon) are being tested with ultra‑rapid insulin. The faster insulin kinetics allow the glucagon to rescue from hypoglycemia quickly and more effectively. Early studies with the iLet bionic pancreas (Beta Bionics) have used Fiasp with promising results, showing TIR around 75–80% in adults.

Another innovative concept is the microneedle patch that delivers insulin directly into the dermis, bypassing the subcutaneous delay. This could further compress the PK curve, potentially allowing insulin action to begin within one minute. Until these technologies become commercially available, Fiasp represents the fastest and most reliable option for automated delivery.

Practical Tips for Starting Fiasp in a Closed‑Loop System

Adjust DIA setting: Change the duration of insulin action to 3–4 hours (consult system manual).
Monitor closely for 7–10 days: Expect more frequent alarms as the algorithm learns. Log post‑meal glucose and any hypoglycemia episodes.
Check infusion set compatibility: Use only approved sets; change every 2 days initially.
Reduce pre‑bolus time: Start by bolusing 5 minutes before eating; many can transition to 0–5 minutes after the first bite.
Adjust ICR and CF gradually: After 1 week, review patterns and tweak ratios by 5–10% if needed.
Consider temporary basal reduction: If hypoglycemia occurs, especially overnight, reduce basal by 0.05–0.1 units/hour.
Keep a rescue snack strategy: Even with the algorithm, glucose may drop sharply; always have fast‑acting carbs accessible.
Engage the diabetes team: Schedule follow‑up within 2 weeks to review CGM data and fine‑tune settings.
Educate school or workplace caregivers: Because Fiasp acts faster, treatment of hypoglycemia may need to be more prompt.

Conclusion

Fiasp has proven to be a valuable tool in hybrid closed‑loop insulin delivery. Its rapid onset, higher peak, and shorter duration align synergistically with automated insulin algorithms, translating into meaningful improvements in time in range (5–10 percentage points), reduced post‑prandial hyperglycemia, and better overnight control. The clinical evidence base is robust, and real‑world experience continues to accumulate positively. However, successful adoption requires careful attention to DIA settings, infusion set compatibility, and user education. Not all users or all systems will benefit equally; patient selection, algorithm optimization, and cost considerations must guide clinical decisions. As both insulin formulations and closed‑loop algorithms continue to evolve, the synergy between ultra‑rapid insulins and automation will likely become the standard of care for type 1 diabetes. The future holds even faster insulins and more intelligent algorithms, setting the stage for a truly fully‑automated artificial pancreas.

External resources for further reading: