Although superior to other insulin therapies, hybrid closed-loop systems differ in efficacy
Key takeaways:
- All hybrid closed-loop systems improved time in range and time below range compared with other insulin therapies.
- MiniMed 780G achieved the greatest improvement in time in range of the systems.
Hybrid closed-loop systems outperformed other types of subcutaneous insulin treatment for diabetes, but some systems were more efficient than others, according to results published in Diabetes/Metabolism Research and Reviews.
“While there is plenty of evidence that hybrid closed-loop systems (HCLs) are more effective in improving time in range with similar or even improved time spent in hypoglycemia as compared with more traditional approaches, including multiple daily injection therapy, standalone insulin pumps and sensor augmenting pumps, there are no head-to-head trials directly comparing the different systems,” Sergio Di Molfetta, PhD, assistant professor at University of Bari Aldo Moro and endocrinologist in the Endocrinology Unit of the Policlinico Consorziale University Hospital of Bari, Italy, told Healio.
Also, although these systems are similar in overall structure, the researchers noted they include different algorithms, components, settings and additional functions, which may lead to disparate glycemic outcomes.
This inspired the researchers to perform a systematic review and network meta‐analysis to compare the safety and efficacy of seven commercially available alternative HCL systems: MiniMed 670G and MiniMed 780G (Medtronic); t:slim X2 with Control-IQ technology (Tandem Diabetes Care); CamAPS Fx (CamDiab Ltd.); Diabeloop Generation 1 (DBLG‐1; Diabeloop) and Diabeloop for highly unstable diabetes (DBLHU; Diabeloop); and Omnipod 5 (Insulet Corporation).
The researchers identified 28 randomized controlled trials that compared how 2,446 adults, adolescents and/or children with type 1 diabetes responded to an HCL compared with another type of subcutaneous insulin therapy treatment and reported 24-hour time in range (TIR) as an endpoint. The eligible trials examined five HCL systems (MiniMed 670G, Minimed 780G, Control-IQ, CamAPS Fx and DBLG1).
Across the trials, baseline HbA1c, duration of diabetes, duration of intervention and age of the participants were similar.
The primary endpoint of the analysis was TIR, with the comparators grouped as subcutaneous insulin therapy without continuous glucose monitoring (SIT); subcutaneous insulin therapy with continuous glucose monitoring (SITCGM); and low-glucose suspension or predictive low-glucose suspension (LGSPLGS).
Di Molfetta and colleagues found that all the HCL systems significantly increased TIR compared with SIT. Specifically, MiniMed 780G had the greatest improvement in TIR (mean difference [MD] = 21.59%; 95% CI, 17.64-25.53) compared with SIT, followed by Control IQ (MD = 16.49%; 95% CI, 11.72-21.25), MiniMed 670G (MD = 14.1%; 95% CI, 10.99-17.22), CamAPS Fx (MD = 12.64%; 95% CI, 8.06-17.23) and DBLG1 (MD = 10.9%; 95% CI, 5.72-16.08). Also, the improvement in TIR was significantly greater with HCL systems compared with LGSPLGS (except for DBLG1) and SITCGM and with MiniMed 780G compared with the other HCL systems.
Further, the researchers found that all the HCL systems decreased time below range compared with SIT. DBLG1 achieved the greatest reduction in time below range compared with SIT with an MD of –3.69% (95% CI, –5.2 to –2.19), followed by MiniMed 670G (MD = –2.9%; 95% CI, 3.77 to 2.04), MiniMed 780G (MD = 2.79%; 95% CI, 3.94 to 1.64), Control IQ (MD = –2.5%; 95% CI, 3.73 to 1.27) and CamAPS Fx (MD = –2.0%; 95% CI, 3.44 to 0.59). Reduction of time below range with HCL systems was numerically greater than with LGSPLGS and with SITCGM, except for CamAPS Fx.
All the HCL systems significantly decreased time above range and all but DBLG1 significantly reduced mean glucose compared with SIT and SITCGM. MiniMed 780G achieved the greatest reductions in time above range (MD = 18.82%; 95% CI, 24.3 to 13.34) and mean glucose (MD = 26 mg/dL; 95% CI, 36 to 15.9) compared with SIT.
Concerning safety, risk for hypoglycemia and diabetic ketoacidosis was similar between patients using HCL systems and patients using other types of insulin therapy.
Subgroup analyses also revealed that “Minimed 780G, differently from other HCL systems, was even more effective in improving TIR in studies with baseline HbA1c greater than or equal to 8%,” Di Molfetta told Healio.
Further, “HCL systems with purely predictive algorithms, including Control‐IQ, CamAPS Fx and DBLG1, achieved greater improvement in TIR in studies with mean age of participants younger than 18 years and shorter diabetes duration (lower than 10 years), and therefore are possibly more effective in these subgroups of patients.”
The researchers noted several limitations to this study, including low certainty of evidence for many comparisons and their inability to apply these results to patients with type 2 diabetes, pregnant women and other “special” populations with diabetes.
“MiniMed 780G is probably the most effective system for improving TIR, while DBLG1 is the most effective in reducing time below range,” Di Molfetta told Healio.
“The MiniMed 780G's algorithm combines elements from proportional-integral derivative, model predictive control and fuzzy logic control and delivers autocorrection boluses up to one every 5 minutes on the top of basal insulin regulation. That's why, in my opinion, it is possibly more efficient in tackling hyperglycemia, particularly in the postprandial period.
By contrast, the DBLG1 algorithm allows a higher hypoglycemia threshold to be set for insulin delivery and further recommends calibrated preventive sugaring when hypoglycemia is predicted despite basal rate reduction,” Di Molfetta said.
“We believe that our results may be useful for clinicians facilitating patient-tailored decision-making,” he added. “Specifically, we think that, alongside with other device features, results from clinical trials should be part of the decision process for orienting toward certain systems rather than toward others based on glycemic goals and patient clinical characteristics.”
For more information:
Sergio Di Molfetta, PhD, can be reached at sergio.dimolfetta@policlinico.ba.it.
Collapse