Automatic Insulin Delivery System Cuts A1c, With or Without Bolusing

SAN DIEGO -- An automatic insulin delivery system bolstered type 1 diabetes outcomes even among users who didn't follow recommendations on boluses for meals and other corrections, according to a retrospective study.

Among patients with uncontrolled type 1 diabetes, use of the Tandem Control IQ automated insulin delivery system overall lowered hemoglobin A1c (HbA1c) and increased time spent in target range, said Viral N. Shah, MD, of the University of Colorado Anschutz Medical Campus in Aurora.

After 12 months, patients who were "high bolusers" -- only using the system in auto-mode between 10% to 49% of the time -- saw a median HbA1c drop of 1.5% from an average baseline of 9.0%, Shah said during a presentation at the American Association of Clinical Endocrinology (AACE) annual meeting.

This group was bolusing "too much" and "not allowing the Tandem Control IQ to do auto-correction boluses," Shah explained.

Patients who were not bolusing -- using the closed loop system in auto-mode more than 90% of the time -- also saw a median 1.5% drop in HbA1c from a baseline value of 9.7%.

The sweet spot appeared to be in between the two. Using the system in auto-mode 50% to 90% of the time was associated with a median 2% drop in HbA1c from an average baseline of 9.4%. This group of "intermediate bolusers" saw the largest drop in HbA1c during the study.

"I think the intermediate group is doing better because they are balancing what they're supposed to do versus letting the system do its job," Shah suggested.

"When you do a bolus by yourself, you are turning off the auto-correction," Shah said. "Some people are micro-bolusers. They take a small amount of the boluses, which turns off the clock, and that bolus is not sufficient enough to really improve that kind of a glycemic control."

Bolusing included both meal-time insulin bolusing and correction-related doses.

"The Tandem Control IQ gives an auto-bolus every hour," Shah pointed out. It delivers 60% of the total correction bolus calculated based on the users personal profile correction factor, current continuous glucose monitor reading, and insulin on board.

All bolusing groups started to see drops in HbA1c within the first 3 months of using the automatic insulin delivery system, with values steadily improving throughout 12 months.

As for time spent in range, there were significant improvements across all three groups at month 12:

• High bolusers (auto 10-49%): 20% increase in time-in-range
• Intermediate bolusers (auto 50-90%): 25% increase in time-in-range
• Low to no bolusing (auto >90%): 15% increase in time-in-range

Looking at secondary outcomes, all bolusing behavior groups saw a slight uptick in body weight by around 6.6 lb (3 kg) by month 12, which Shah called "expected."

Insulin dose remained stable for all groups throughout the study, as well as time spent below range (70 mg/dL).

"Control IQ can be used even in patients who are not good at bolusing and those who have a very high A1c," Shah said, noting that he recommends it particularly for older patients with dementia. "It can be used very safely as a fully automated system."

The small study included 30 adults: 10 in each type of bolus group. The average age and duration of diabetes was 44 and 23 years, respectively. At baseline, the average HbA1c was 9.4% and insulin dose was 0.54 units/kg/day. Bolus-behavior groups were matched for age, sex, and diabetes duration.

All participants had type 1 diabetes for more than 2 years and were using a continuous glucose monitoring system for more than 6 months prior to Tandem Control IQ initiation.

The FDA first cleared this device in December 2019, making it the first interoperable, automated insulin dosing controller on the market. It works by connecting to an alternate controller-enabled insulin pump (ACE pump) and integrated continuous glucose monitor in order to automatically adjust insulin delivery to the user.

Because of its interoperable design, it can be used with compatible continuous glucose monitors.

In the original trial underpinning Tandem Control IQ's approval, published in The New England Journal of Medicine in 2019, closed-loop users utilized a Tandem t:slim X2 insulin pump with Control-IQ Technology with a Dexcom G6 continuous glucose monitor.

Shah pointed out that patients like those included in the current study are often excluded from trials due to their poorly controlled HbA1c. Therefore, these patients tend to not be seen as "ideal candidates" for insulin pump therapy.

"In my opinion, everyone with type 1 diabetes should get an automatic insulin delivery system," Shah said. "The objective here is that the people who did not do well in managing their diabetes still can use this kind of technology and improve their diabetes control."

However, he added: "I do not recommend to use the Control IQ as a fully automated system ... Ideally, I still encourage my patients to do the boluses prior to their meals, if possible, and let the system work in a different way."

Comment