Résumé
Insulin delivery in a closed-loop mode has been a dream for many patients with insulin-treated diabetes since bedside artificial pancreas (AP) systems were developed in the 1970s. Beside safe near-normal glucose levels, the goal of AP is to alleviate patients’ burden and fear of continual adjustment of insulin delivery needed to cope with daily activities and events. Portable pumps using subcutaneous (SC) insulin infusion and ‘needle-type’ enzymatic sensors allowing continuous glucose monitoring (CGM) in the interstitial SC fluid are typically used in the current AP prototypes. Model predictive control algorithms which take into account the delays inherent with SC insulin infusion and glucose sensing have shown improved glucose control in hospital setting. Currently, pilot trials are performed in home-like conditions to assess the technical feasibility, safety and efficacy of glucose control, and patients’ ability to manage AP. Recently developed wearable smart phone-based platforms connect wirelessly to the insulin pump and the CGM, run control algorithms, provide online information to/from the patient, and allow remote monitoring reaching a new frontier – first outpatient experiments. The future holds expansion of home trials supporting the approval of systems which could revolutionize diabetes treatment and make easier the daily life of patients with diabetes.