Automated insulin delivery lowers HbA1c, improves time in range in type 2 diabetes
Key takeaways:
- Use of the Omnipod 5 led to an HbA1c reduction of 0.8 percentage points at 13 weeks for adults with type 2 diabetes.
- FDA cleared the device for use in adults with type 2 diabetes in August.
Adults with insulin-treated type 2 diabetes had a significant HbA1c decline at 13 weeks with use of the Omnipod 5 automated insulin delivery system, according to findings from the SECURE-T2D trial.
As Healio previously reported, researchers presented initial data from SECURE-T2D during the American Diabetes Association Scientific Sessions in June showing the Omnipod 5 (Insulet) could lower HbA1c for adults with type 2 diabetes. In August, the FDA granted the Omnipod 5 clearance in adults with type 2 diabetes based on the presented data. In the full trial findings published in JAMA Network Open, researchers discussed how the automated insulin delivery system reduced HbA1c in type 2 diabetes regardless of sex, race and ethnicity, prior insulin therapy and SGLT2 or GLP-1 use.
“The performance of automated insulin delivery among individuals on different treatment regimens, including those also taking GLP-1s or SGLT2 inhibitors, highlights the flexibility and synergy of automated insulin delivery with current pharmacotherapies,” Francisco J. Pasquel, MD, MPH, associate professor of medicine and global health at Emory University, told Healio. “By simplifying insulin dosing and improving glucose control across diverse subgroups, this technology has the potential to reduce complications and enhance quality of life for adults with type 2 diabetes.”
Researchers conducted a 13-week single-arm trial at 21 U.S. centers. Adults aged 18 to 75 years treated with a stable insulin regimen for at least 3 months used the Omnipod 5 along with a Dexcom G6 continuous glucose monitor. Baseline data were obtained during a 14-day standard therapy phase prior to starting automated insulin delivery. Change in HbA1c from baseline to 13 weeks was the trial’s primary outcome. CGM data and patient-reported outcomes were also collected.
Omnipod 5 reduces HbA1c
There were 305 adults who started automated insulin delivery (mean age, 57 years; 57% women; 50% white; 24% Black; 22% Hispanic) and 289 participants who completed the 13-week trial. The study group had a decrease in HbA1c from 8.2% at baseline to 7.4% at 13 weeks (difference, –0.8 percentage points; 95% CI, –1 to –0.7; P < .001 for noninferiority and superiority). Noninferior HbA1c declines were observed for adults receiving multiple daily injections before the trial (mean HbA1c change; –0.8 percentage points) and those receiving basal insulin only (mean HbA1c change; –1.2 percentage points; P for noninferiority < .001 for both).
Significant HbA1c declines were seen for all baseline HbA1c groups except for those with a baseline HbA1c of less than 7%. All racial-ethnic groups had significant HbA1c decreases, though Black adults had the lowest mean change at –0.5 percentage points. Adults using a GLP-1 had an HbA1c change of –0.9 percentage points, and those using an SGLT2 had a –0.8 percentage point change in HbA1c. The percentage of adults achieving an HbA1c of less than 7% rose from 14% at baseline to 37% at 13 weeks.
Time in range rose from 45% at baseline 66% at 13 weeks (P < .001). The proportion of adults with a time in range of more than 70% increased from 19% at baseline to 42% at 13 weeks.
BMI increased from 34.9 kg/m2 at baseline to 35.1 kg/m2 at 13 weeks. Mean daily insulin requirements dropped from 0.8 U/kg at baseline to 0.57 U/kg with automated insulin delivery.
No serious adverse device effects were reported during the trial. There were 19 nonserious adverse device effects reported by 17 participants. One severe hypoglycemic event and 13 nonglycemic serious adverse events occurred during the trial, all of which were unrelated to the Omnipod 5.
Most users would recommend system
At the end of the study, 90% of participants said they would recommend using the Omnipod 5 to family or friends, 82% said it was easy to wear and 78% stated they would continue to use the system after the trial ended.
“It was noteworthy how effectively the automated insulin delivery system worked across a heterogeneous population, including those on adjunct therapies like GLP-1s or SGLT2 inhibitors, and regardless of education, income or insurance status,” Pasquel told Healio. “The high acceptance rate and ease of use, even among participants who may initially be considered less tech-savvy, underscore the accessibility and broad applicability of automated insulin delivery.”
Pasquel said future studies should examine outcomes over a long-time period and assess the efficacy automated insulin delivery in specific subgroups such as older adults.
For more information:
Francisco J. Pasquel, MD, MPH, can be reached at fpasque@emory.edu.
Perspective
Back to Top
The SECURE-T2D trial with Omnipod 5 showed consistent improvements in people with type 2 diabetes regardless of carb counting skills, race, education, income level and use of GLP-1s or SGLT2s. This confirms what we see in practice that Omnipod 5 (and other automated insulin delivery systems) work very well for people with type 2 diabetes that need insulin.
It was great to see that this was studied in a more diverse population as one of the common criticisms of insulin pump studies is that they are performed in highly-educated, tech-savvy users that don’t necessarily represent the average person with diabetes. This trial included higher percentages of people identifying as Black or Latino; and 25% had never worn CGM before. They also included a high percentage of participants not previously carbohydrate counting. Results demonstrated that fixed insulin doses or small, medium and large meal entries had similar HbA1c reductions compared to carbohydrate entries. This bodes well for its broad use. Carbohydrate counting and insulin pump tubing are two barriers that are easily overcome with Omnipod 5.
There may be some opportunity for algorithm refinement or additional education practices as final mean time in range was 66%, falling a little shy of the 70% goal for most people with diabetes, but it is still a very clinically significant improvement from baseline of 20 percentage points.
Collapse