Automated Insulin Delivery System Helps in Type 2 Diabetes

The use of an automated insulin delivery (AID) system reduced A1c and hyperglycemia without increasing hypoglycemia in people with insulin-treated type 2 diabetes (T2D), a new randomized trial showed.

The data, from Tandem Diabetes Care’s 2IQP trial, informed the February 2025 US Food and Drug Administration (FDA)’s added T2D indication of the company’s next-generation Control-IQ+ algorithm, following its prior clearance for type 1 diabetes.

The Control-IQ+ algorithm enables AID in conjunction with Tandem’s t:slim X2 and Mobi insulin pumps and a compatible continuous glucose monitor (CGM). The updated algorithm allows for a wider range of weight (maximum 440 lb vs 308 lb) and total daily insulin inputs (maximum 200 units vs 100 units) compared with the first-generation Control-IQ. Another AID system, the Omnipod 5, is also FDA-cleared for use in T2D.

The new results for Control-IQ+ were published on March 19, 2025, in The New England Journal of Medicine and simultaneously presented in an industry symposium at the Advanced Technologies & Treatments for Diabetes (ATTD) 2025 meeting taking place in Amsterdam, the Netherlands (March 19-22, 2025).

“In a diverse population of adults with insulin-treated type 2 diabetes, the use of AID safely reduced glycated hemoglobin levels and hyperglycemia without increasing hypoglycemia as compared with a control group using CGM,” Yogish C. Kudva, MD, of Mayo Clinic, Rochester, Minnesota, and colleagues wrote.

Asked to comment, Independent Industry Consultant Charles M. Alexander, MD, told Medscape Medical News, “It’s nice to see substantial improvement in glycemic control without any increase in hypoglycemia. All the subgroups showed the same improved glycemic efficacy compared to basal-bolus insulin.”

As for the overall role of the use of AID technology in T2D, Alexander said, “AID certainly could become standard of care for basal-bolus insulin-using type 2 diabetes, but not for type 2 generally, including those only on oral agents, on GLP-1 [glucagon-like peptide 1] injections or those only on basal insulin without prandial insulin.”

However, he also noted, “access and affordability will be a big issue for AID unless the industry can either have a comparable price point or show the health economic advantage of AID compared to traditional basal-bolus insulin.”

Lower A1c, Higher Time in Range, No Increased Hypoglycemia 

The 13-week, multicenter trial included 319 people with T2D who were using both basal and mealtime insulin, 96% with injections, and only 4% with pumps. At baseline, 75% were using fixed-dose insulin injections at mealtime — that is, without carbohydrate counting — and 71% were using CGMs.

Most were using other glucose-lowering medications, including 44% on a GLP-1 receptor agonist (RA)–based drug, 37% on a sodium-glucose cotransporter 2 (SGLT2) inhibitor, and 21% using both medications.

Participants were randomly assigned in a 2:1 ratio to receive the AID with Tandem’s t:slim X2 insulin pump with Control-IQ+ technology in conjunction with a Dexcom G6 CGM, or to continue their usual insulin delivery method (controls) while also using the Dexcom G6.

The mean A1c level at 13 weeks (primary outcome) decreased by 0.9 percentage points, from 8.2% to 7.3%. The mean adjusted difference was a −0.6-point difference between the AID and control groups (P < .001).

There was greater benefit among those with higher A1c levels at baseline. For example, among those with baseline A1c ≥ 9%, the reduction at 13 weeks went from 10.3% to 7.9% in the AID group vs 9.7% to 8.6% in the control group. Reductions were also seen in the subgroups using GLP-1 RA and/or SGLT2 inhibitors.

“Our trial results indicate that AID can further reduce glycated hemoglobin levels when added to a diabetes management regimen that includes one or more of these medications,” Kudva and colleagues wrote.

The treatment effect also appeared similar among participants using fixed doses of insulin and those who had been counting carbohydrates for meal boluses before the trial, and among those with both high and low baseline numeracy scores.

“Only 4% of the patients had experience using an insulin pump before the trial, and the patients did not receive formal training in traditional carbohydrate-counting methods or nutritional management of diabetes during the trial. Most patients elected to use a simple fixed-bolus regimen that allowed for minor adjustments for meal carbohydrate content. Consequently, the results of this trial suggest that previous experience with an insulin pump or in-depth training in carbohydrate-counting methods are not prerequisites to the successful and safe use of AID to improve glycemia in patients with type 2 diabetes,” the authors said.

The percentage of time patients spent in the target glucose range of 70-180 mg/dl rose from 48% at baseline to 64% at 13 weeks in the AID group vs from 51% to 52% in the control group (mean difference, 14 percentage points; P < .001). This difference represents a mean time in target glucose range of 3.4 hours per day longer in the AID group than in the control group, Kudva and colleagues pointed out.

Hypoglycemia assessed by CGM was low at baseline and remained so over the 13 weeks. The between-group difference in percentage of time spent with a glucose level < 70 mg/dl was an insignificant −0.1 percentage points.

One patient in the AID group had a severe hypoglycemia event, which was successfully treated with oral carbohydrate. There were 20 nonserious device-related hyperglycemia events in 13 patients in the AID group, primarily related to infusion set failures. There were no severe hypoglycemia events in the control group, and no cases of diabetic ketoacidosis or hyperosmolar hyperglycemic syndrome in either group.

Kudva’s institution had received grants/contracts from Dexcom, Medtronic Minimed, Novo Nordisk, and Tandem Diabetes Care. Alexander is a volunteer consultant to the 

diaTribe Foundation with the title of Scientific and Medical Advisor.

Miriam E. Tucker is a freelance journalist based in the Washington DC area. She is a regular contributor to Medscape Medical News, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X (formerly Twitter) @MiriamETucker and BlueSky @miriametucker.bsky.social.