Automated Insulin Delivery in Elderly With Type 1 Diabetes (AIDE T1D)

A Randomized Cross-over Trial Evaluating Automated Insulin Delivery Technologies on Hypoglycemia and Quality of Life in Elderly Adults With Type 1 Diabetes

A multi-center, randomized, crossover trial consisting of three sequential 12-week periods, with the HCL feature used during one period, the PLGS feature used during one period and SAP therapy (control) during one period. The crossover trial will be preceded by a run-in phase in which participants will receive training using the study devices (Dexcom G6 and Tandem t:slim X2 pump). After the last crossover period, participants will be given the opportunity to use study devices for an additional 12 weeks to assess preference of system use (PLGS, HCL or SAP) and associated characteristics, durability and safety in a more real-world setting with less frequent study contact.

Automated insulin delivery (AID) technologies hold the promise of optimizing glycemic control and reducing the burden of diabetes care for patients with Type 1 Diabetes (T1D). However, clinical trials of lower burden AID technologies have not included older adults in sufficient numbers to allow for focused evaluation of efficacy and quality of life (QOL) impacts that may differ from those observed in younger age groups. Most notably, primary endpoints have focused on reducing hyperglycemia, while avoidance of hypoglycemia is of upmost concern for older adults with T1D. T1D Exchange clinic registry data have shown severe hypoglycemia (SH) occurs more commonly in older adults with longstanding T1D than in younger individuals with events occurring just as often with HbA1c levels >8.0% as with HbA1c levels <7.0%. These data do not support the strategy of "raising the HbA1c" as being an effective approach for hypoglycemia prevention in older adults with T1D. In addition to acutely altered mental status, hypoglycemia is associated with an increased risk for falls leading to fractures, car accidents, emergency room (ER) visits, hospitalizations, and mortality resulting in substantial societal costs. The occurrence of hypoglycemia, hypoglycemia unawareness and fear of hypoglycemia have adverse effects on overall QOL of both individuals with T1D and their families.

While continuous glucose monitoring (CGM) technology alone has the potential to be beneficial in reducing hypoglycemia in older patients, our preliminary data from the Wireless Innovations for Seniors with Diabetes Mellitus (WISDM) trial shows a majority of patients still have frequent hypoglycemia even when using CGM. Thus, knowledge of CGM alone may not be sufficient to avoid hypoglycemia in this population. Predictive low-glucose suspend algorithms have particular promise when the primary goal is hypoglycemia avoidance rather than glucose reduction. Whether the added complexity of closed loop systems provides additional glycemic benefit is not known. There is a critical need to determine whether automated insulin delivery can reduce hypoglycemia in the older adult population with T1D.

Inclusion Criteria:

1. Clinical diagnosis of type 1 diabetes
2. Age ≥ 65 years old
3. T1D Duration of at least 1 year
4. HbA1c < 10.0% from point of care or local lab within the past 6 months
5. Insulin regimen involves basal/bolus insulin via insulin pump or multiple daily injections
6. Most recent GFR ≥ 30 ml/min/m^2 from local lab within the past 6 months
7. Willingness to use a rapid acting insulin compatible with the Tandem t:slim X2 pump (currently aspart and lispro; other rapid acting insulins likely to be approved for pump use prior to study initiation such as Fiasp)
8. Familiarity with and willingness to use a carbohydrate ratio for meal boluses
9. Willing to use study devices and automated insulin delivery features
10. Ability to download study devices at home or if not able to download at home willing to come into clinic to bring devices for download of data at visits and as needed for safety
11. Participant is independently managing his/her diabetes with respect to insulin administration and glucose monitoring (may include assistance from spouse or other caregiver)
12. Participant understands the study protocol, agrees to comply with it and is able to successfully pass the consent understanding assessment with no more than 2 attempts
13. Participant comprehends written and spoken English
14. At least 240 hours of CGM readings available during the end of run-in assessment
15. At least 1.5% of time with CGM glucose levels < 70 mg/dL prior to SAP initiation
16. Active prescription for glucagon and willing and able to have glucagon available

Exclusion Criteria:

1. Use of PLGS technology or HCL insulin delivery in the past 1 month
2. History of 1 or more Diabetic Ketoacidosis episodes in the previous 6 months
3. Clinical diagnosis by a primary care provider, neurologist or psychiatrist of dementia, in the investigator's opinion a suspected severe cognitive impairment such that it would preclude ability to understand the study or use devices, or a score of 6 or less out of 15 on the 5 min MoCA (5-min T MoCA Version 2.1) (mild cognitive impairment is not an exclusion)
4. A condition, which in the opinion of the investigator or designee, would put the participant or study at risk, including severe vision or hearing impairment and any contraindication to the use of any of the study devices per FDA labeling
5. Known adhesive allergy or skin reaction during the run-in pre-randomization phase or previous difficulty with pump and CGM insertions that would preclude participation in the randomized trial
6. Concurrent use of any non-insulin glucose-lowering agent other than metformin (including GLP-1 agonists, Symlin, DPP-4 inhibitors, SGLT-2 inhibitors, sulfonylureas)
7. Stage 4 or 5 renal disease
8. The presence of a significant medical or psychiatric condition or use of a medication that in the judgment of the investigator may affect completion of any aspect of the protocol, or is likely to be associated with life expectancy of <1 year