番茄社区

ORLANDO -- Putting type 2 diabetes patients on an automated insulin delivery (AID) pump improved glycemic control, the SECURE-T2D pivotal trial showed.

Type 2 diabetes patients started on the Omnipod 5 System, currently approved only for use in type 1 diabetes, had a 0.8 percentage point lower HbA1c compared with baseline (7.4% vs 8.2%, P<0.001), reported Francisco J. Pasquel, MD, MPH, of Emory University in Atlanta, at the American Diabetes Association (ADA) annual meeting.

For most, it was their initial device: 73% had been on multiple daily injections, 21% on basal-only injections, and 5.6% on a non-automated pump. The findings were the same for those who previously had been on multiple daily insulin injections but even greater for those who had been taking only basal insulin (7.5% vs 8.6%, P<0.001).

"It's been already used a lot. I think this is going to give more reassurance to people that there [are] data out there," he said. "We know already that it works."

Reassurance was echoed by a member of the audience who said he works for another device company that has something similar in development. "I worry about potentially naïve people who have used insulin but haven't used a pump. I would be very interested to see what the safety data are like, but then clearly it works. Actually, these data are much stronger than we would have thought."

Co-author John Buse, MD, PhD, of the University of North Carolina at Chapel Hill, argued that safety isn't too much of a concern: "In general, the automated insulin delivery pumps are associated with better safety data than the old insulin pumps -- less hypoglycemia, in particular, which is a major safety concern ... If you were to ask me do I expect any new safety signals from using the Omnipod 5 in patients with type 2 diabetes as opposed to patients with type 1 diabetes, I'd say no."

Time in hypoglycemia was an identical 0.2% with both treatments. Only one case of severe hypoglycemia occurred during Omnipod 5 use, which was deemed unrelated to study device malfunction. No diabetes-related ketoacidosis or hyperosmolar hyperglycemic syndrome occurred.

The trial included 305 adults at 21 U.S. centers who had type 2 diabetes and were on a stable insulin regimen without prior AID use. They were monitored for 2 weeks on their existing therapy then switched to the Omnipod 5 system used in conjunction with a Dexcom G6 continuous glucose monitor (CGM) for 13 weeks. They were given flexible options for meal boluses based on carb counting, size of the meal, or another method.

Of the participants, 62% were current CGM users at baseline and 25% had never used CGM.

The key secondary endpoint of time in the target glycemic range of 70-180 mg/dL improved after starting on the AID (66% vs 45%, P<0.001), representing 4.8 more hours a day at target.

The researchers also found a significant reduction in the proportion of patients (55% vs 66% at baseline, P<0.001) with clinically meaningful distress about their diabetes as marked by a score of at least 2.0 on the Type 2 Diabetes Distress Assessment System scale, which assesses shame or stigma, management demands, care access, long-term health, hypoglycemia, and interpersonal issues.

HbA1c results were greater at higher baseline levels, with as much as a 2.1 percentage point decrease for those at 9.0% or greater at baseline (P<0.001 for trend). Outcomes were similar for those using and not using GLP-1 receptor agonists and across prior meal dosing methods, although the impact was significantly greater for Hispanic or Latino patients (P=0.004 for interaction by race or ethnicity), which Pasquel suggested was likely related to socioeconomic determinants of health.

Amount of insulin used was lower on the Omnipod 5 by about 23 units per day (0.57 vs 0.80 U/kg, P<0.001).

Buse noted, though, that the higher doses of insulin that tend to be used in type 2 diabetes as compared with type 1 diabetes could mean more frequent cartridge changes for these patients. At least one patient even needed new cartridges daily.

One thing that might be explored in other AID studies is use of more concentrated insulins like U200, Buse suggested. And replication with other AID systems and algorithms would add to the confidence for improved glycemic control, Pasquel noted.

"It is really exciting to see now the value of technology, both CGM and AID in type 2 diabetes," commented ADA press conference moderator Marlon Pragnell, PhD, ADA vice president for research and science.

Buse agreed: "Nobody wants to have diabetes, but this is a great time to have diabetes. It is incredible ... The natural history of the disease is changing in front of our eyes, and we just can't even imagine what the impact is of these GLP-1 receptor agonists, SGLT2 inhibitors, CGM, automated insulin delivery, and then the innovations that are still coming on the lives of people with diabetes."

Comment