• Insulin
• Cardiology
• Hyperglycemia/Hypoglycemia
• Diabetes Mellitus
• Diabetes & Endocrinology Clinical Trials

Glycemic control is difficult to maintain in children with diabetes aged <7 years for several reasons. Children of this age are at increased risk of hypoglycemia (particularly at night), and there is the potential for hypoglycemia-related neurocognitive outcomes. In addition, at this age, children often have unpredictable eating patterns and variable levels of activity. Closed-loop insulin delivery is a recent medical innovation that aims to achieve tight glucose control while reducing the risk of hypoglycemia, but it has not been tested in young children. Andrew Dauber, MD, Boston Children's Hospital, Boston, Massachusetts, USA, presented data from the Closed-Loop Insulin Delivery in Children <7 years of Age Study that showed how closed-loop therapy (CLT) has the potential to improve diabetes care in young children [NCT01421225].

CLT combines glucose-sensing and insulin-delivery components with real-time glucose-responsive insulin administration. A disposable sensor measures interstitial glucose levels, which are fed automatically into an algorithm that is used to control the delivery of a rapid-acting insulin analog into the subcutaneous tissue by an insulin pump. This was a randomized crossover trial that compared CLT with standard (open-loop) pump therapy (SPT) in children aged <7 years who were diagnosed with type 1 diabetes for more than 6 months and had been treated with insulin pump therapy for more than 6 weeks. Study participants (n=10) had a mean age of 5.1 years (range 2.0 to 6.8 years) with a mean duration of diabetes of 2.1 years (range 0.5 to 4.7 years). Mean HbA1C was 8.1% (range 7.1% to 8.9%), and the average daily insulin dose was 0.72 units/kg (range 0.61 to 1.0 units/kg). All subjects had fasting c-peptide levels <0.1 ng/mL.

In this study, glucose values were transmitted from 2 Freestyle Navigator® sensors (one placed in each thigh) to a bedside receiver. The values were retrieved from the receiver and entered manually into a control algorithm to calculate insulin recommendations. The control algorithm was a physiological insulin delivery algorithm, developed by one of the investigators, that utilized proportional-integral-derivative terms that were modified by insulin feedback. All recommendations generated by the algorithm were approved by the physician and then entered into the Animus OneTouch® Ping® insulin pump remote, which transmitted the insulin order to the insulin pump attached to the patient. There were 2 periods of control: overnight (10:00 PM to 8:00 AM) when basal rates were adjusted every 20 minutes based on sensor readings, and daytime (8:00 AM to noon), when mini-boluses of insulin in increments of 0.05 units were given up to every minute, based on sensor readings. Target blood sugars were 150 mg/dL during the night and 120 mg/dL during the day.

Subjects were admitted to the clinic for 48 hours. Meals and snacks were provided on a regular schedule. Participants were free to choose from a standardized menu but received the identical meals/snacks on Days 1 and 2. All meals were weighed pre- and post-consumption. The Freestyle sensors were placed on the morning of admission. That afternoon patients were switched to the study insulin pump. An intravenous line was also placed to allow for frequent blood sampling. The study period began at 10:00 PM and ran until noon the following day. During that time, subjects were randomized to receive either CLT or SPT. From noon until 10:00 PM, subjects received their standard therapy. At 10:00 PM on the second evening, subjects were randomized to the opposite therapy from the prior night.

Time at overnight target was increased with CLT but was not significantly different from SPT; however, time in extreme hyperglycemia was significantly reduced as was the total glycemic excursion overnight. There was no difference in the number of interventions for hypoglycemia or in daytime peak postprandial glucose (despite the absence of a pre-meal bolus), while the pre-lunch was significantly decreased (Table 1).

• © 2012 MD Conference Express®