Better Glycemic Control With Novel Hybrid Closed Loop System

An advanced hybrid closed loop system (AHCL) successfully reduced daytime hyperglycemia in patients with type 1 diabetes, according to the FLAIR trial.

In a crossover trial of adolescents and young adults, those using a Medtronic AHCL spent significantly less time in daytime hyperglycemia compared with those using a MiniMed 670G system, meeting the trial’s primary endpoint (34% vs 37%, adjusted difference -3%, 95% CI -4% to -2%, P<0.001), reported Richard Bergenstal, MD, of the International Diabetes Center at Park Nicollet & HealthPartners in Minneapolis.

Daytime hyperglycemia was defined as 6 a.m. to 11:59 p.m., Bergenstal explained during a presentation at the American Diabetes Association (ADA) virtual meeting.

Notably, improvement in daytime hyperglycemia did not come at the cost of hypoglycemia, as those on the AHCL also spent less time in hypoglycemia over a 24-hour period (0.46% vs 0.5%, adjusted difference -0.06%, 95% CI -0.11% to -0.02%, P<0.001).

“We were pleased we made our primary outcome of less daytime hyperglycemia without increasing hypoglycemia,” said Bergenstal.

The global, randomized trial took place at seven clinical sites (four in the U.S., two in Europe, one in Israel) and focused on patients ages 14 to 29 years. Bergenstal calling them the “toughest of the tough” in terms of patient populations, often seeing the highest HbA1c levels.

What set FLAIR apart from previous automatic insulin delivery trials was its broader entry criteria and participants with many characteristics, he noted. Additionally, most studies using automated insulin delivery have compared these devices to older diabetes management strategies, like sensor augmented pumps, while FLAIR’s comparator was the commercially available Medtronic 670G closed-loop insulin delivery system.

When compared with the 670G system, some of the main differences with the AHCL include a glucose target setpoint of 120 mg/dL or 100 mg/dL — not just 120 mg/dL — and a lower correction bolus target of 120 mg/dL rather than 150 mg/dL. It also has a “unique” feature for automatic correction boluses.

“Systems that have auto-correction are going to be essential in this tough age-group of ages 14 to 30,” Bergenstal said.

The crossover trial included 126 individuals with type 1 diabetes, who have previously used several different methods. Prior to the study, 20% were using multiple daily injections and almost 40% had never used a continuous glucose monitor (CGM) before, while only 13% were using the 670G system already. About 25% of the study population had a baseline HbA1c between 8.6% to 11%, while less than 20% of the population had an HbA1c between 7% to 7.4%.

All participants underwent a run-in phase using the 670G system in manual mode. Then, 56 participants received the AHCL first while the 57 individuals continued on the 670G system for 12 weeks. All participants crossed over to the other method for an additional 12 weeks. At baseline, all individuals used a CGM for 2 weeks prior to randomization, and continued throughout the trial. In total, 111 participants completed the trial.

Looking at 24-hour data, those on the AHCL always maintained average glucose levels below those on the 670G system, while the largest delta in glucose levels was usually seen during the early morning hours, around 6 a.m. However, both groups always remained below baseline levels during a 24-hour period.

People on the AHCL also tended to spend on average more time in range (70-180 mg/dL) during a 24-hour period than the 670G group, particularly during those early morning hours:

• Baseline: 57%
• 670G: 63%
• AHCL: 67%

“We started with a pretty low time in range in this group. They came up [for time in range] significantly in both groups,” Bergenstal pointed out. “There was a 10% increase in time spent in range [comparing baseline to AHCL], remembering that 5% is really clinically significant.”

The average percentage of daily time spent in hyperglycemia (over 180 mg/dL) was also significantly improved, dropping 10% for those on the AHCL (41% baseline, 34% 670G, 31% AHCL). A similar outcome was also seen for average percentage of time spent in severe hyperglycemia, or over 250 mg/dL (13% baseline, 10% 670G, 9% AHCL).

Both age groups — ages 14 to 20 and 21 to 29 — saw significant improvements in HbA1c and time spent in range. Also, results were consistent for those with baseline HbA1c at or below 8.5% versus higher.

Bergenstal pointed out that even patients who entered the trial with no technology experience — those who had never used a CGM and used multiple daily injections — also saw a significant benefit. This group in particular jumped from a 45% time in range at baseline to a 65% average time spent in range.

“Remember, these patients are usually excluded from trials,” Bergenstal explained. “I’m making a plea not to exclude people just because they haven’t previously used technology, particularly these adolescents.”

Both technologies were safe, with no reports of diabetic ketoacidosis in either group.

ADA session chair Timothy Bailey, MD, of the AMCR Institute in Escondido, California, asked which system the participants preferred. Bergenstal said that “pretty much across the globe” that everyone said they would opt for the AHCL. “There’s no extension for [the participants] so they’re going to have to just wait and see what the [regulatory] approval time [for the device] is,” he stated.