Professor May Ng from Edge Hill University in Liverpool, England, speaks of an “explosion in diabetes technology in the past decade”. In particular, automatic insulin dosing systems (AID – automated insulin delivery) are a breakthrough in diabetes therapy, said the pediatric endocrinologist at the 17th Diabetes Technology Congress ATTD (Advanced Technologies & Treatments for Diabetes) in Florence.
It is time to present the available evidence to political decision-makers and to implement these AID systems in everyday practice.
This is still a long way away in Europe, said Professor Roman Hovorka from the University of Cambridge in England.
Improved glucose metabolic control in T1D and T2D
Hovorka and other speakers at the ATTD Congress were convinced that AID systems can significantly contribute to improving glucose metabolism control in people with type 1 diabetes (T1D) as well as insulin-dependent type 2 diabetes (T2D). Interestingly, patients with T2D seem to benefit even more significantly than those with T1D in terms of glucose metabolism control.
A distinction must be made between hybrid closed-loop systems (HCL) and fully closed-loop systems (FCL).
HCL systems automatically cover the basal insulin requirement, which corresponds to around 40 to 60 percent of the daily insulin requirement. Insulin boluses must be injected manually or carbohydrates must be consumed with meals. The systems have different approvals for some age groups. Five HCL systems are currently approved in Europe (as of March 2024):
- CamAPS FX, a non-commercial AID system developed at the University of Cambridge,
- Tandem Control IQ,
- Omnipod 5 HCL,
- MiniMed 670G and 780G as well
- Diabeloop DBLG1.
FCL systems, on the other hand, carry out both basal and bolus injections fully automatically – these AIDs are not yet commercially available.
Most scientific data on AID systems currently exist for T1D treatment. Last year, an international committee with 75 experts published consensus recommendations on application issues (Endocrine Reviews 2023; 44: 254-280).
Hybrid closed-loop in T2D
There is significantly less evidence for its use in T2D. Dr. Anders Carlson from the Health Partners Institute in Park Nicollet, Minnesota, cited a feasibility study in T2D patients who had unsatisfactory metabolic control with HbA despite therapy with basal or basal-bolus insulin1cvalues of 9 percent on average.
The 24 participants initially recorded their glucose profiles using CGM (continuous glucose monitoring) for two weeks of standard therapy and then progressed to eight weeks of treatment with a commercially available AID system (HCL).
The time when blood sugar was above 250 mg/dL decreased by an absolute 17 percent. “That equates to four fewer hours per day that patients were previously hyperglycemic,” Carlson said.
More time in the target area
Overall, hypoglycemia was rare in the study, which did not change with AID use. What changed, however, was the time in range (TIR), which increased by almost 22 percent.
The HbA1cvalue fell by 1.3 percentage points. Patients who were additionally treated with GLP-1 receptor agonists or SGLT2 inhibitors benefited the most in comparison (Diabetes Care 2023; 46:742-750). The results of this study are currently being attempted to be replicated in a large multicenter US study (SECURE-T2D) and is nearing completion.
Real-world research shows that both T2D patients who only require basal insulin and those on a basal-bolus regimen can benefit from automated insulin administration, Carlson said. Two thirds to three quarters of patients reach the optimal glucose target corridor (TIR), around 20 percent of the time the glucose level is above 180 mg/dl (TAR – time above range), with phases of significant hyperglycemia (>250 mg/dl). about another five percent.
Severe hypoglycemia, on the other hand, is very rare (0.1 to 0.3 percent), so strict target value settings can be practiced. In fact, optimal glucose metabolism settings with AID could apparently be achieved more quickly in T2D than in T1D, explained Carlson, but pointed out that experience is still limited.
Switching to HCL system could be useful
Can patients with T2D who need insulin injections multiple times a day benefit from AID? Carlson’s answer is: Yes! Switching to an HCL system has been shown to be beneficial in older home-care individuals with inadequately controlled T2D and multiple daily insulin injections.
With standard insulin therapy, only a third achieved the TIR (70-180 mg/dl). This rate doubled under AID therapy within twelve weeks, with correspondingly favorable effects on HbA1c-Value (Diabetes Obes Metab 2024;26:622-630).
“This means that AID can not only reduce the therapeutic burden for patients, but also for the nursing staff,” said the diabetologist. But that also means: If you want to implement AID in everyday practice, close cooperation between endocrinologists, family doctors and nursing staff is required.
Fully closed loop in T2D
A British pilot study shows that fully automated insulin delivery (FCL) also works in T2D and can therefore improve blood sugar control without increasing hypoglycemia events (Nat Med. 2023; 29: 203-208).
The system developed in Cambridge is called CamAPS HX. According to Hovorkas, behind this is an improved predictive algorithm that reacts more quickly to changes in glucose levels, a learning system that takes over all bolus insulin administrations.
The system is being tested on both inpatient and outpatient T1D patients, as well as on outpatient patients with T2D and on dialysis patients with T2D. “If the system works in very challenging situations, it will also be safe in others,” said Hovorka, citing a British-Swiss study.
In 136 hospitalized patients with T2D or with severe hyperglycemia, the FCL system was significantly more successful in achieving the target corridor of 100 to 180 mg/dl than with standard insulin therapy. A TIR of 66 percent compared to a rate of 42 percent in the control group.
The phases below 100 mg/dl decreased and the risk of hypoglycemia did not increase. On average, patients treated with the FCL system were in target glucose range for six more hours per day (N Engl J Med. 2018; 379: 547-556).
Values got better day by day
It was observed how the values improved day by day within one to two weeks. “The algorithm learns very quickly,” said Hovorka.
Furthermore, the system has been successfully tested in hospital patients who required artificial parenteral or enteral nutrition, as well as in hemodialysis patients. Attempts are currently underway to implement the FCL system into standard care in several regions of England and Scotland.
A surprising result of the studies with the FCL system is the very different daily variability of insulin requirements in T1D and T2D. For adults with T1D, this is on average less than 20 percent, but for T2D it is sometimes over 40 percent. This speaks even more of the fact that patients with T2D in particular need modern diabetes technology, says Hovorka.
“But what we were also interested in was what people think about the fully closed loop system, whether they accept it, whether they feel safe.” Hovorka quoted from structured interviews. Improved mood and quality of life are reported. “The diabetes depends on my life and not my life on the diabetes.”
“I’m less stressed”
You feel “normal” and are no longer afraid of nighttime hypoglycemia. “My quality of life has improved dramatically, I am less stressed,” is one exemplary statement. Another: “Freedom to eat without the stress of counting carbs, especially in social situations.”
However, many questions still remain to be answered: How to deal with parallel treatment with GLP-1 receptor agonists or SGLT2 inhibitors? When is the right time to start AID-based insulin therapy for T2D? Does this fit with primary care? How easy to use do the systems have to be? To what extent will artificial intelligence influence the underlying algorithms? And: How do the respective payers assess the benefits of AID systems?
The technology experts’ wish lists include even faster-acting insulins, improved algorithms or the combined application of insulin with, for example, GLP-1 receptor agonists or an amylin analogue.