Assessment of a New Closed-loop Algorithm in Type 1 Diabetes (Saddle Point Model Predictive Control : SP-MPC) (PPA)

Prospective, Monocentric, Crossover Randomized Trial, to Assess a New Closed-loop Algorithm in Type 1 Diabetes (Saddle Point Model Predictive Control : SP-MPC)

The purpose of the study is to assess a new closed-loop algorithm in type 1 diabetes (Saddle Point Model Predictive Control : SP-MPC)

Ten patients will be assessed during a closed-loop night and an open-loop night in random order.

The primary objective is to assess the ability of SP-MPC controler to maintain euglycemia and avoid hypoglycaemia during the night, compared to open loop in type 1 diabetic patients treated by insulin pump.

Inclusion Criteria:

• Patients with type 1 diabetes (C peptide negative) :
• 18 to 70 year old,
• with type 1 diabetes for more than 3 years,
• treated by insulin pump for more than 3 months,
• with previous CGM holter allowing preliminary settting of some of the SP-MPC parameters
• able to provide written informed consent.

Exclusion Criteria:

• pregnancy or breastfeeding,
• severe non stabilized diabetic retinopathy,
• cardiovascular event in the previous 6 months,
• infectious diseases with recent fever,
• anemia (hemoglobin < 11g/l),
• HbA1c > 10%,
• Current or recent (less than 4 weeks) oral or injectable corticosteroid treatment,
• creatinine clearance <40ml/mn,
• chronic alcoholism (weekly consumption of alcohol > 280g for men and > 140g for women),
• adults legally protected (under judicial protection, guardianship, or supervision), persons deprived of their liberty