From the Journals

Artificial pancreas treatment improves glycemic control in T1DM: Meta-analysis

View on the News

Policy makers will need more (and better) evidence

Closed-loop glucose control systems have much to offer patients, but better evidence will likely be needed to convince policy makers who have increasing demands and scarce resources, Dr. Norman Waugh and his coauthors said in an editorial.

Users of closed-loop systems did spend approximately 10% more time near normoglycemia when compared with control subjects in the present systematic review and meta-analysis of randomized clinical trials by Dr. Bekiari and colleagues.

The overall base of evidence is weak, however, and many studies in the review were low quality, according to Dr. Waugh and his coauthors. Moreover, the trials were short, with the majority lasting 7 days or less, they noted.

The current review does demonstrate that closed-loop systems improve control overnight and reduce the burden of self-management during the day, but it’s unknown if the systems can reduce long-term diabetes complications, according to the editorial authors, and evidence to date is insufficient for policy makers to perform cost-effectiveness analyses.

“We need longer and larger trials, in both adults and children, to compare closed-loop systems with self-management using continuous glucose monitoring,” Dr. Waugh and his colleagues wrote. “These trials should measure HbA1c for modeling the effects on complications, blood glucose variability, hypoglycemia, quality of life, and cost effectiveness.”

Norman Waugh is professor of public health medicine and health technology assessment at the University of Warwick, England. These comments are derived from an editorial by Dr. Waugh and his coauthors BMJ. 2018;361:k1613. doi: 10.1136/bmj.k1613 . They reported advising Roche and Novo Nordisk on matters not related to the topic of this editorial.



Use of artificial pancreas devices added nearly 2.5 hours of time in near normoglycemia over 24 hours in patients with type 1 diabetes, a meta-analysis of randomized clinical trials showed.

The improvement versus control subjects was primarily because of the favorable effect of the closed loop glucose control systems in the overnight period, authors of the meta-analysis reported in the BMJ.

Both single and dual hormone systems had robust results in the meta-anaysis, said lead researcher Eleni Bekiari, MD, PhD, of Aristotle University of Thessaloniki, Greece, and her coinvestigators.

Results were likewise robust for an analysis restricted to trials conducted under normal living conditions and no remote monitoring, supporting the convenience and ease of use of these systems, according to Dr. Bekiari and her colleagues.

“Overall, our results reflect the progress made over recent decades of extensive research and development in artificial pancreas use,” they wrote.

Despite the findings, more research needs to be done, they added, since the individual clinical trials supporting use of closed-loop systems in type 1 diabetes have included relatively few patients and have had short follow-up durations.

The systematic review and meta-analysis by Dr. Bekiari and her colleagues was based on 40 randomized controlled trials involving a total of 1,027 participants. The primary outcome of the analysis was proportion of time that sensor glucose level was in the normoglycemic range of 3.9-10 mmol/L.

Overall, use of the systems was associated with 140 additional minutes in near normoglycemia over 24 hours, with a 9.62% mean weighted difference (95% confidence interval, 7.54%-11.7%), reported data show.

The favorable effect was even more evident on overnight measures, the investigators said, with a weighted mean difference of 15.15% (95% CI, 12.21%-18.09%).

Results were similar even when the analysis was limited to studies that had a low risk of bias, and also when the analysis was limited to studies of unsupervised patients in normal living conditions, according to Dr. Bekiari and her associates.

Artificial pancreas use also had favorable impacts on time in hyperglycemia over the entire day. Compared with controls, time with glucose concentrations less than 10 mmol/L were shortened by about 2 hours, the investigators said.

Likewise, mean levels of sensor blood glucose over 24 hours fell by 0.48 mmol/L compared with control treatment (95% CI, 0.3-0.66 mmol/L), they reported.

Taken together, these findings suggest artificial pancreas systems are efficacious and safe for patients with type 1 diabetes, the investigators concluded.

“Further research with rigorous studies, cooperation of research groups in terms of outcome reporting, and cost-effectiveness data are required to verify these findings and support adoption of artificial pancreas systems in clinical practice,” they wrote.

Dr. Bekiari reported no disclosures. Her coauthors reported disclosures related to Medtronic, Novo Nordisk, Sanofi, AstraZeneca, Boehringer Ingelheim and others outside of the submitted work.

SOURCE: Bekiai E et al. BMJ 2018;361:k1310.

Next Article: