Artificial pancreas may improve glucose control in type 1 diabetics

Artificial pancreas may improve glucose control in type 1 diabetics

A study published in the New England Journal of Medicine on October 16, 2019, describes how patients could benefit from the use of newly available automatic closed-loop systems to administer insulin, popularly called “the artificial pancreas”.

The new system could help type 1 diabetics to keep their blood glucose levels under better control for an additional 2.4 hours a day, while reducing the risk of hypoglycemia (low blood sugar levels), especially during the dangerous early morning hours when it is most common. This, in turn, could help prevent adverse outcomes as a result of unstable blood glucose levels.

Diabetes control is a difficult task, and very few patients manage to meet blood glucose targets fixed by various clinical guidelines, such as those of the American Diabetes Association. The artificial pancreas is a complete system that monitors glucose levels constantly and pumps in insulin as and when required.

Sensing low blood glucose levels

The study reports the results from the third phase of the International Diabetes Closed Loop (iDCL) trial. This was aimed at evaluating how well a closed-loop system (Control-IQ, Tandem Diabetes Care) worked and how safe it was in patients with type 1 diabetes.

The system feeds glucose levels into an algorithm to decide basal insulin dosage throughout the day and larger doses towards the second half of the night, to prevent early morning hyperglycemia.

It also has a specialized module to prevent hypoglycemia, and can deliver larger corrective doses of insulin besides the ongoing basal insulin delivery if required when hyperglycemia is detected.

The study included 168 patients with type 1 diabetes, who were randomly assigned to two treatment groups: the closed-loop group, and the control group who used an insulin pump with a sensor, for 26 weeks each. The patients ranged in age from 14 to 71 years.

All patients had been receiving insulin treatment with an insulin pump or daily insulin injections (more than once a day) for a year or more. The duration of diabetes ranged from 1 to 62 years.

The researchers measured glycated hemoglobin (HbA1C), which reflects the level of blood glucose over the past few months, first at baseline, and then at 13 and 26 weeks. The HbA1C levels in the patient population ranged from 5.4 to 10.6% at baseline.

A period of 2-8 weeks was first allowed for baseline monitoring of the patients, and training in the use of the closed-loop system, unless they had already been using a continuous glucose monitor and insulin pump.

The primary outcome was the amount of time in which the blood glucose level was between 70 and 180 mg/dL. The researchers also looked at other outcomes in a defined order, and found that all improved significantly with the closed-loop system.

Outcomes of the study

Patients in the test group spent over 90% of the time in the closed-loop system mode. The researchers found that closed-loop group patients had blood sugar levels in the target range for 71 ± 12% of the time, compared to only 61 ± 17% at the start.
The control group showed no change from the initial value of 59±14%. The 11 percentage point difference translates into more than 2.6 more hours of good control per day. The difference was noted from the first month onwards and persisted to the end of the study.

The secondary outcomes are given below, with the mean percentage difference between the groups for each outcome:

  • How long the glucose went above 180 mg/dL , in percentage of study time: -10 percentage points (2.4 hours less per day)
  • Average glucose level: -13 percentage points
  • HbA1C level  after 6 months: -0.33 percentage points
  • How long it went below 70 mg/dL: -0.88 percentage points (13 minutes less per day)
  • How long it went below 54 mg/dL : -0.10 percentage points

Overall, daytime control (6 am to midnight) was achieved for 70% and night control (midnight to 6 am) for76% of the time with the closed-loop system.

For the control group, both daytime and nighttime control was achieved for 59% of the time.
The greatest difference in mean glucose levels was at 5 am in test and control group, namely: 139 mg/dL vs. 166 mg/dL. This reflects better control in the second half of the night.

Hypoglycemia occurred for 1.6% and 1.4% of the time in the day and at night, with the closed-loop system. For the control group, this was seen for 2.2% and 2.4% of the day and night periods.

Is the system safe?

The closed-loop system compared well with the control group’s use of the insulin pump and continuous glucose monitor, given that adverse event reporting was more stringent with the test group since this is still investigational in status.

Patients in both groups spent equal number of days in ketosis, but serious hypoglycemia (blood sugar below 54 mg/dL) occurred in neither group. One patient had diabetic ketoacidosis (DKA) in the closed-loop group as a result of equipment failure.

Infusion set failure also caused 13 and 2 occurrences of high blood sugar or high ketone levels, but not enough to cause DKA, in the test and control groups respectively.

Is the artificial pancreas on the horizon for diabetics?

This trial found outcomes that agree well with prior research using the same and different algorithms. The current trial included very motivated patients, most of whom were already trained in and using insulin pumps. However, analysis of the results from the subgroup (about 20%) that was not already using a device to manage their diabetes failed to show any difference.

Overall, the closed-loop system helps patients to keep blood sugar levels steady within the target range and also avoid undesirable lowering of blood glucose. This helps achieve long-term control as shown by improvement in HbA1C levels over the 26 weeks of the trial. This view is echoed by Daniela Bruttomesso who accompanied the NEJM article with an editorial, saying:

These results are impressive and clinically relevant, since it has been shown that for each 10% reduction in the time spent in the glucose target range, the risk of development or progression of retinopathy increases by 64% and the risk of development of microalbuminuria by 40%.”

Journal reference:
Brown, S. A., et al. (2019). Six-month randomized, multicenter trial of closed-loop control in type 1 diabetes. New England Journal of Medicine. October 16, 2019. DOI: 10.1056/NEJMoa1907863.

No comments