Night glucose control with MD-Logic artificial pancreas in home setting: a single blind, randomized crossover trial-interim analysis.

BACKGROUND Artificial pancreas (AP) systems have shown an improvement in glucose control and a reduced risk of nocturnal hypoglycemia under controlled conditions but remain to be evaluated under daily-life conditions. OBJECTIVE To assess the feasibility, safety, and efficacy of the MD-Logic AP in controlling nocturnal glucose levels in the patient's home. METHODS Two-arm study, each covering four consecutive nights comparing the MD-Logic AP ('closed-loop' arm) with sensor-augmented pump therapy ('control' arm). Fifteen patients (mean age 19 ± 10.4 yr, A1c 7.5 ± 0.5% or 58 ± 5.9 mmol/mol, diabetes duration 9.9 ± 8.2 yr) were randomly assigned either to 'Group A' (first 'closed-loop', then 'control' arm) or to 'Group B' (vice versa). Investigators were masked to treatment intervention. Primary endpoints were the time spent with glucose levels below 70 mg/dL and the percentage of nights in which the mean overnight glucose levels were within 90-140 mg/dL. Endpoint analyses were based on unmodified sensor glucose readings of the four study nights. RESULTS Time of glucose levels spent below 70 mg/dL was significantly shorter on the closed-loop nights than on control nights, median and interquartile range 3.8 (0, 11.6) and 48.7 (0.6, 67.9) min, respectively; p = 0.0034. The percentage of individual nights in which mean overnight glucose level was within 90-140 mg/dL was 67 (33, 88), and 50 (25, 75), under closed-loop and control nights, respectively, with no statistical difference. Secondary endpoint analyses demonstrated significant improvements in hypoglycemia parameters. No serious adverse events were reported. CONCLUSION This interim analysis demonstrates the feasibility, safety, and efficiency of the MD-Logic AP system in home use, and demonstrates an improvement over sensor-augmented pump therapy. (ClinicalTrials.gov identifier NCT01726829).