Do Automated Insulin Delivery Systems Have a Place in T2D Treatment?

Automated insulin delivery (AID) systems using continuous subcutaneous insulin infusion (CSII) and continuous glucose monitors (CGMs) are safe and effective for hospitalized or closely monitored patients with type 2 diabetes (T2D), a recent review found.

Patients with T2D often need insulin for glycemic control, but the risks of hypoglycemia and weight gain increase with insulin treatment. AID systems were developed to mitigate the risk of hypoglycemia and to deliver insulin in a more physiologic manner.

Combining CSII and CGMs with a predictive control algorithm that regulates insulin infusion forms an AID called the artificial pancreas (AP) due to its more physiologic function vs traditional insulin injections. Fully closed loop systems function without prandial dosing, and hybrid closed loop systems (HCL) require preprandial insulin administration either based on preset doses or carbohydrate counting.

AID systems have become more common in the last decade and have been studied thoroughly in type 1 diabetes (T1D) management. They have been shown to improve glycemic outcomes, and 3 HCL systems are FDA approved for T1D. However, fully closed systems have shown difficulty achieving postprandial control in studies of patients who have either T1D or T2D, despite showing excellent overnight glucose control. For patients with T1D, HCL systems are being used at increasing rates and have shown benefits compared with traditional injections.

“With this increased adoption by clinicians, patients with T2D are another group of people with diabetes who may benefit from AID systems,” the authors wrote. “Specifically, T2D patients using multiple daily injections (MDI) seeking a balance of improved glycemic control, increased quality of life (QOL), and a reduction in hypoglycemic risk may consider an AID system as a treatment option.”

The review, published in Endocrine Practice, summarized recent data on insulin pump use and the potential for AID implementation in patients with T2D.

Across all FDA-approved AID devices, postmarketing observational studies have shown glycemic improvement and safety for patients with T1D. In 5 randomized controlled trials (RCTs) of AID systems in T2D, patients experienced increased time in range. AID systems also reduced glycemic variability and the risk of hypoglycemia.

Relatively high out-of-pocket costs and a theoretical concern of increased insulin administration, which would make patients more likely to gain weight, are both potential drawbacks of AID systems. In the RCTs assessed in the review, AID systems did not increase insulin requirements.

“The improvements in glycemic control and reductions in glycemic variability noted with AID to date in small trials are modest. At present, it has not been demonstrated that these devices improve morbidity or mortality from T2D,” the authors wrote. “However, hypoglycemia reduction, which is a main driver for hospitalization and medical cost (estimates of $120 million per year in united states), could justify the expenses of AID devices.”

Another potential barrier is usability, considering many patients with T2D are older and may be less technologically inclined. Patient training would be an integral part of successful implementation and minimize discontinuation rates if AID systems for T2D become commercially available. T2D is heterogenous and patients have a significant risk of comorbid health issues, meaning personalized glycemic targets may also be needed.

The majority of AID system trials for patients with T2D have also been done in supervised environments, not in the home setting. The studies therefore often include smaller, heterogenous populations with high comorbidity rates and may not be generalizable to the overall T2D population or home settings. Larger studies are needed to confirm the potential benefits of AID systems in T2D, and home studies, such as those currently being led by the pan-European consortium CLOSE (Automated Glucose Control at Home for People with Chronic Disease) are necessary to determine the feasibility of use in unsupervised environments.

Reference

Karol AB, O'Malley G, Fallurin R, Levy CJ. Automated insulin delivery systems as a treatment for type 2 diabetes: a review. Endocr Pract. Published online October 11, 2022. doi:10.1016/j.eprac.2022.10.001