The purpose of this investigator-initiated trial is to compare the efficacy in terms of time to recovery from hyperglycemia as measured by time to arrest of hyperglycemic excursion ("glucose plateau point", primary endpoint) and return to premeal glucose target if feasible (secondary endpoint) between Fiasp and conventional insulin aspart when used as a correction bolus. These endpoints will be determined by CGM (Dexcom) from data exported from the Dexcom Clarity program.

Patients with type 1 DM using CSII require bolus insulin for two purposes: first, to cover carbohydrate intake to control postprandial glucose, and second, to correct episodes of hyperglycemia. The latter function is referred to as a "correction dose" or "correction bolus". Insulin pumps have bolus calculators which calculate correction doses based on the patient's individualized BG target and insulin sensitivity factor (ISF). Rapid-acting insulin analogues delivered by pump typically require 2 to 4 hours to fully correct an acute hyperglycemic episode, and sometimes multiple correction doses are needed to normalize the blood glucose level. This can be frustrating for patients, particularly when a correction bolus is necessary in addition to a meal bolus; frequently, patients must wait for the correction bolus to take effect and delay eating until the blood glucose has begun to normalize to avoid severe postprandial hyperglycemia. There is consequently an unmet need in insulin delivery, and in particular in CSII, for an insulin which can correct a hyperglycemic episode more rapidly than is currently possible with rapid-acting insulin analogues.

Faster insulin aspart (Fiasp) is a novel formulation of insulin aspart with an accelerated time-action profile which results in twice the exposure to insulin and 74% greater insulin action within the first 30 minutes after injection compared to conventional insulin aspart. This results in twice-as-fast onset of appearance in the bloodstream (4 vs. 9 min compared to conventional insulin aspart) which has been demonstrated to reduce postprandial glucose levels in patients with type 1 DM using CSII. Theoretically, this faster insulin action would be useful in correction dosing during acute episodes of hyperglycemia to normalize the blood glucose level more rapidly than is currently possible with conventional insulin aspart (NovoLog).

Many patients with type 1 DM using CSII now also use continuous glucose monitoring (CGM) for making insulin dosing decisions. Currently the FDA has approved 2 CGM systems for nonadjunctive use in bolus insulin dose calculations. Only one of these systems, the Dexcom, reads continuously to the patient and has alarms to warn of impending episodes of hyper- or hypoglycemia, and this is the system used most commonly by patients with type 1 DM using open-loop CSII. Patients now incorporate the Dexcom trend arrow, which depicts the rate and direction of glucose change, into the correction dose calculation, and recommendations on how to incorporate CGM information into correction dose calculations have recently been updated based on an expert consensus report. However, these guidelines were created for use with rapid acting insulin analogues. How they might need to be modified for use with Fiasp (the first and only ultra-rapid insulin analogue) is not known.

The purpose of this investigator-initiated trial is to compare the efficacy in terms of time to recovery from hyperglycemia as measured by time to arrest of hyperglycemic excursion ("glucose plateau point", primary endpoint) and return to premeal glucose target if feasible (secondary endpoint) between Fiasp and conventional insulin aspart when used as a correction bolus. These endpoints will be determined by CGM (Dexcom) from data exported from the Dexcom Clarity program.

Study hypothesis:

Compared to conventional insulin aspart, Fiasp will correct hyperglycemia (defined as arrest of rise of blood glucose, following correction bolus, ie, GPP ) faster than conventional insulin aspart in subjects with type 1 DM using CSII.

Inclusion Criteria:

• Male and female patients > 18 years of age
• Type 1 DM of > 1 year duration
• Use of any open loop insulin pump, Tandem T-Slim with Basal IQ, Insulet Omnipod Dash, or any other investigator-approved insulin pumps with Dexcom CGM G5, G6, or newer version for > 6 months
• Good baseline glycemic control (HbA1c < 7.5%; low risk of hypoglycemia by CGM as defined by Dexcom Clarity report)
• No episodes of severe hypoglycemia in the previous 3 months
• Pump download shows regular meal bolusing, accurate carbohydrate counting ability, and willingness to use exercise markers in Dexcom
• CGM download shows regular use (>85% of time) and regular calibration if using G5 sensor (G6 requires no calibration)
• Females using adequate contraception

Exclusion Criteria:

• Use of CGM other than Dexcom G5 or G6 or a newer Dexcom CGM version
• Suboptimal baseline glycemic control (HbA1c > 7.5%)
• Pump or CGM download shows suboptimal use of devices (lack of meal boluses, frequent overrides of pump, excessive pump suspension, inadequate calibration or inconsistent usage of CGM)
• Serious comorbidities including CVD with recent event, actively treated malignancy, renal dysfunction with eGFR < 45 ml/min, or any other condition which in the opinion of the investigator would preclude subject's ability to participate in trial
• Females unwilling to use contraception, planning pregnancy or breastfeeding
• Use of any other glucose-lowering agents than insulin
• Hypersensitivity to insulin aspart or one of the excipients in faster insulin aspart
• Known diabetic gastroparesis

Data and statistical analysis will be available to any researcher with appropriate request

post-study, up to 5 years

email Principal Investigator

This is an open-label crossover trial. Subjects will be randomized to treatment sequence (Fiasp then NovoLog or NovoLog then Fiasp), each for a 12-week treatment period, by CSII. Randomization sequence will be determined by computerized randomization program. All patients will receive both treatments unless they drop out. Dropouts are unlikely since the participants are all regular continuing patients of the site's clinical practice.