Fuzzy Logic Automated Insulin Regulation (FLAIR)

National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), Medtronic, Jaeb Center for Health Research, Schneider Children's Medical Center, Israel, University of Ljubljana, Yale University, Joslin Diabetes Center, University of Florida, Stanford University, International Diabetes Center at Park Nicollet, Kinderkrankenhaus auf der Bult, University of Minnesota

Adolescents and young adults with type 1 diabetes often have a difficult time achieving good glucose control, which is so important in reducing the risk for diabetes complications. Despite the use of multiple daily injections or insulin pumps and glucose sensors, there is still a need for many individuals to further improve glucose levels without causing low blood glucose levels (hypoglycemia) or adding to the daily burden of living with diabetes. Today an insulin pump can receive glucose readings from a continuous glucose monitor and adjust the insulin delivery in an attempt to keep glucose levels in a more optimal range. These systems are called hybrid closed loop (HCL). This means that much of the insulin delivery is automated, yet the patient still interacts regularly with the system, particularly to help determine the insulin dose to deliver to cover a meal. Results of early studies using HCL systems in adolescents and adults with type 1 diabetes are encouraging. The objective of this study is to compare the efficacy and safety of the automated insulin delivery (AID) system with proportional integral-derivative (PID) algorithm (Minimed 670G 3.0 HCL) to an AID system with combined PID and Fuzzy Logic Algorithm (Minimed 670G 4.0 Advanced Hybrid Closed-Loop (AHCL)). The trial will test the hypothesis that the Minimed AHCL can reduce daytime hyperglycemia, currently the biggest challenge for AID systems, without increasing hypoglycemia. Up to 124 adolescents and young adults (ages 14-<30) will be recruited to test each system for three months in a randomized crossover trial. Investigators will compare how effective each hybrid closed loop system is at preventing high blood glucose readings during the day. The investigators will also evaluate the safety of each system and how participants adjust to the daily use of the technology.

Purpose of the study: A randomized crossover trial involving up to four clinical sites in the United States and three sites outside the US (Germany, Israel and Slovenia) will compare the efficacy and safety of an AID system with a PID algorithm versus an AID system with a PID algorithm enhanced with a fuzzy logic algorithm. Study Objectives: EFFICACY: The co-primary outcomes are difference in continuous glucose monitoring (CGM)-measured metrics between periods: Superiority for percent of time >180 mg/dL (10.0 mmol/L) from 6 AM to 11:59 PM; and Non-inferiority of percent of time <54 mg/dL (3.0 mmol/L) during the entire 24-hour period. KEY SAFETY OUTCOMES: Percentage of sensor glucose readings <54 mg/dL (overall is a co-primary outcomes) and <70 mg/dL (3.0 and 3.9 mmol/L, respectively) Diabetic Ketoacidosis (DKA) events Severe hypoglycemia events Study design: Randomized crossover trial with two 12-week crossover periods in automode, preceded by a run-in phase. Population: A maximum of 124 individuals may be enrolled and start the run-in phase. Approximately 112 are expected to enter the crossover trial, with a goal of at least 100 people completing the trial. Maximum duration of a study for a subject: Approximately 28-36 weeks. Recruitment: Subjects will be recruited through the clinical sites. Consent: Written consent/assent will be obtained for all subjects and/or guardians, in accordance with human subjects and regulatory requirements. Screening Assessments: Informed consent will be signed and eligibility will be assessed Medical history and physical examination HbA1c measurement Urine pregnancy test (if applicable) Surveys investigating participants' quality of life, psychosocial and cognitive functioning, and response to their current treatment will be distributed. Study Training: Run-In Period: (2-8 weeks) * Eligible participants will use the study 670G 3.0 HCL pump during the run-in. Participants who were pump users at screening may skip the Pump Run-In period (but must participate in the Pump+CGM Run-In period) per investigator discretion. 670G auto mode users may use the 670G pump in auto mode. Screening and start of run-in training visits may occur on the same day or separate days, but no more than 14 days apart. Standardized pump training will be provided to study participants and their diabetes care partners (for participants <18 years old). The study team will assist the participant in study pump infusion site initiation and will start the participant on the study pump. For current pump users, the study pump will be programmed with the participants usual basal rates and pump parameters. The participants personal pump will be removed. Participants may continue to use their personal CGM if applicable. The pump will be used for at least two weeks during the Pump Run-In, with the option of repeating Pump Run-In for an additional two weeks per investigator discretion. Contact will be made each week with additional contacts as needed. Prior to each contact, participants will be asked to upload device data for study staff to review. After completion of Pump Run-In, participants will proceed to use the study CGM along with the study 670G HCL pump during the Pump+CGM Run-In period. * Pump+CGM Run-In (670G 3.0 HCL + Guardian Sensor (3)) All participants must complete a two-week run-in period with the use of the study pump and CGM before being randomized into the crossover trial. During Pump+CGM Run-In, the predictive low glucose suspend feature will be turned on and auto mode will be off (i.e. manual mode). Participants who were 670G auto mode users at screening may use the pump in auto mode. Standardized device training will be provided to study participants and their diabetes care partners (for participants <18 years old). Personal pumps and CGMs will be removed during the Pump+CGM Run-In period as applicable. Contact will be made each week with additional contacts as needed. Prior to each contact, participants will be asked to upload device data for study staff to review. Run-In Assessment Successful completion of Pump Run-In is per investigator discretion. Pump Run-In may be repeated once. Successful completion of the Pump+CGM Run-In requires CGM data to be collected on at least 80% of the possible time in the prior 14 days of use. An average of at least three blood glucose meter (BGM) tests per day also will be required. If these are not achieved, the Pump+CGM Run-In period may be repeated once. Randomization into the Crossover Trial Eligible participants who successfully complete the Pump+CGM Run-In will be randomly assigned to begin with one Automated Insulin Delivery (AID) system during Period 1 and then crossover to the other AID system during Period 2. The two study AID systems (treatments) are: 670G 3.0 Hybrid Closed-Loop (HCL) (PID) insulin pump + Guardian Sensor (3) CGM 670G 4.0 Advanced Hybrid Closed-Loop (AHCL) (PID + Fuzzy Logic) insulin pump + Guardian Sensor (3) CGM Home Use of AID System during the Crossover Trial Period 1 (~13 weeks) Participants and their diabetes care partners (for participants <18 years old) will be trained by qualified personnel on the use of the assigned pump and on auto mode feature use including meal announcement, meal bolusing, and exercise. Training will also be provided in performing specific tasks including the following: Confirming pump parameters When not to use or rely on auto mode particularly during significant illness or acetaminophen use CGM calibration instructions Meal bolus procedures What to do when exercising while using the system How to react to safety/alert notifications How to perform fingerstick blood glucose measurements in accordance with the labeling of the study CGM device When and how to contact study staff to ask questions during the study Study staff will discuss the visit and contact schedule with the participant and will make arrangements for follow-up appointments. Participants will be asked to upload data before each contact and at least every two weeks. After auto mode training has been completed, participants will proceed with home use of the AID system (meaning free-living use at work, home, etc.) during Period 1 with either the 670G 3.0 HCL or 670G 4.0 AHCL pump. The predictive low glucose suspend feature will be on. The system will initially be used with auto mode deactivated (except for 670G auto mode users at screening who may activate auto mode if using the 670G 3.0 HCL pump) until participants are contacted 6-10 days into Period 1 with instructions to activate auto mode. Participants will be instructed to obtain an overnight fingerstick blood glucose measurement (between 2-3AM) for 2-3 nights following auto mode initiation and if fingerstick blood glucose is <70 mg/dL to treat with carbohydrate. Participants will then continue using the AID system for 12 weeks after auto mode is initialized. Participants will be expected to use auto mode at all times at home with some exceptions (e.g. times of illness, acetaminophen use). Participants on the 670G 4.0 AHCL will begin with an auto mode target glucose set point of 120 mg/dL (6.7 mmol/L) that may be lowered to 100 mg/dL (5.6 mmol/L) if participant meets the safety criteria per protocol and investigator discretion. HbA1c, C-peptide, and glucose levels will be collected for central lab analysis at the beginning of Study Period 1. Human Factors and Diabetes Technology Attitude Surveys will be administered at the end of Study Period 1. Period 2 (~13 weeks) At the beginning of Period 2, a urine pregnancy test will be completed as applicable. Eligible participants will then use the other AID system during Period 2. The procedures in Period 1 will be repeated in Period 2. At the end of the 12-weeks of AID use in auto mode at home, the participant will complete a final study visit. That visit may be completed in-person in clinic or in an alternate location such as the participant's home. The study visit may occur remotely via phone or videoconferencing. Certain procedures, such as the measurement of height, weight, vitals, and collection of the central HbA1c sample, may be missed if the visit is not completed in-person. Participants requiring a remote final visit will be transitioned off of the study device during the remote contact and an arrangement will be made between site staff and the participant to return all required study devices either in-person or via mail.

Randomized crossover trial with two 12-week crossover periods in auto mode preceded by a run-in phase.

Participants will receive insulin delivered by the Medtronic Minimed 670G 3.0 HCL system using a PID algorithm..

Participants will receive insulin delivered by the Medtronic advanced hybrid closed loop system (Minimed 670G 4.0 AHCL) with Guardian Sensor (3) continuous glucose monitoring sensor.

The components of the intervention are the insulin pump with insulin delivery algorithm (PID + Fuzzy Logic) and Guardian Sensor (3).

Continuous Glucose Monitoring derived indices over the first 84 days of each treatment period for 24 hours (excluding time before auto mode is turned on). Glucose levels based on sensor glucose data for Mean glucose.

CGM derived indices over the first 84 days of each treatment period for 24 hours. Glucose levels based on sensor glucose data for coefficient of variation; percentage of sensor glucose readings in the range of 70 to 180 mg/dL (3.9-10.0 mmol/L) and 70 to 140 mg/dL (3.9 to 7.8 mmol/L); percentage of sensor glucose readings >180 mg/dL (daytime is a co-primary outcome) and >250 mg/dL (10.0 and 13.9 mmol/L, respectively)

Efficacy: Amount of Total, Basal and Bolus Daily Insulin Over the First 84 Days of Each Treatment Period

Amount of total, basal, and bolus daily insulin over the first 84 days of each treatment period (excluding time before auto mode is turned on) for 24 hours. Glucose levels based on sensor glucose data. Sum of daytime and nighttime values may not equal total daily values due to rounding.

Glycated hemoglobin (HbA1c) was measured at a central laboratory (Advanced Research and Diagnostic Laboratory University of Minnesota, MN, USA) at randomization and at the end of each period by use of an International Federation of Clinical Chemistry and Laboratory Medicine aligned method (Tosoh HPLC Glycohemoglobin Analyzer, Tosoh Medics, San Francisco, CA, USA; coefficient of variation range 1.4 1.9%).

Time Frame: End of crossover period 1 (Week 14 through 20, depending); and end of crossover period 2 (Week 26-32, depending)

Height and weight. Body mass index (BMI) is a person's weight in kilograms divided by the square of height in meters. BMI is interpreted using standard weight status categories. These categories are the same for men and women of all body types and ages. Below 18.5 : Underweight; 18.5 - 24.9: Normal or Healthy Weight; 25.0 - 29.9: Overweight; 30.0 and Above: Obese.

Time Frame: Screening visit, at initiation (Day 0); randomization (Week 2 through 8, depending); end of crossover period 1 (Week 14 through 20, depending); and end of crossover period 2 (Week 26-32, depending)

Time Frame: Screening visit, at initiation (Day 0); randomization (Week 2 through 8, depending); end of crossover period 1 (Week 14 through 20, depending); and end of crossover period 2 (Week 26-32, depending)

Key Safety Outcome 1) Percentage of Time Sensor Glucose Readings Were <54 mg/dL and <70 mg/dL (3.0 and 3.9 mmol/L, Respectively

DKA as defined by the Diabetes Control and Complications Trial (DCCT) and described below: Symptoms such as polyuria, polydipsia, nausea, or vomiting; Serum ketones >1.5 mmol/L or large/moderate urine ketones; Either arterial blood pH <7.30 or venous pH <7.24 or serum bicarbonate <15; and Treatment provided in a health care facility

Severe hypoglycemia event as defined by the Diabetes Control and Complications Trial (DCCT) and described below: The event required assistance of another person due to altered consciousness, and required another person to actively administer carbohydrate, glucagon, or other resuscitative actions. This means that the participant was impaired cognitively to the point that he/she was unable to treat himself/herself, was unable to verbalize his/ her needs, was incoherent, disoriented, and/or combative, or experienced seizure or loss of consciousness.

Surveys completed by participants. The Glucose Monitoring Satisfaction Survey is 15 items on a 1-5 scale; Total score calculated as mean of all item scores; higher scores indicate greater satisfaction. The Diabetes Distress Scale is 17 items on a 1-6 scale; Total score calculated as mean of all item scores; higher score denotes more distress. The Hypoglycemia Confidence Survey is 8 items on a 1-4 scale; Total score calculated as mean of all item scores; higher score denotes more confidence. The Diabetes Technology Attitudes Survey is 5 items on a 0-4 scale; Total score calculated as sum of all item scores; higher score denotes more satisfaction with diabetes technology. The Adult INSPIRE Survey is 22 items on a 1-5 scale; Total score calculated as mean of all item scores; higher score denotes more satisfaction with AID. The Adolescent INSPIRE Survey is 17 items on a 1-5 scale; Total score calculated as mean of all item scores; higher score denotes more satisfaction with AID.

Time Frame: Screening visit, at initiation (Day 0); end of crossover period 1 (Week 14 through 20, depending); and end of crossover period 2 (Week 26-32, depending)

Inclusion Criteria: Type 1 diabetes mellitus (as diagnosed clinically) for at least one year Age 14-<30 years at enrollment For females, not currently known to be pregnant, be breast-feeding or planning to become pregnant within the study duration. Using an insulin pump or multiple daily injections of insulin Participant must be able to obtain U-100 rapid acting insulin analogues, Aspart or Lispro, for use during the study (since these are the only insulins approved for the study pump and the study is not supplying insulin) MDI users must be on a basal/bolus regimen Participants must have a minimum total daily dose (TDD) of at least eight units HbA1c from an approved HbA1c point of care analyzer with a value 7.0%-11.0% Willingness or ability to do carbohydrate counting In the investigator's judgment, able to understand and likely to be adherent to the protocol For subjects <18 years old, living with one or more diabetes care partners (eg.g. parent/legal guardian), of whom at least one is committed to participating in study training for emergency procedures for severe hypoglycemia and able to contact the participant in case of an emergency. Have adequate internet access and a computer system that meets requirements for uploading data. For participants currently using CGM or insulin pump, willingness to discontinue personal CGM and pump when using the study CGM and pumps (note: including implantable CGMs). Exclusion Criteria: Individuals meeting any of the following exclusion criteria at screening will be excluded from study participation: Concomitant disease that influences metabolic control or HbA1c interpretation (e.g. anemia, significantly impaired hepatic function, confirmed gastroparesis, renal failure, history of adrenal insufficiency, sickle cell disease, haemoglobinopathy, or has received red blood cell transfusion or erythropoietin within three months prior to time of screening) or other medical condition which, in the Investigator's opinion, may compromise patient safety, affect outcome assessments, or affect the participant's ability to follow the protocol Oral or parenteral glucocorticoids taken within 1 month prior to enrollment, or plans to take oral or parenteral glucocorticoids within the planned study duration. Exceptions: Short term oral or parenteral glucocorticoids up to seven days Use of antidiabetic agents other than insulin Use of other medications, which in the judgment of the investigator would be a contraindication to participation in the study One or more episodes of severe hypoglycemia (hypoglycemia requiring treatment by another person) within the previous six months Known allergy to medical grade adhesives Participation in another study of a medical device or drug that could affect glucose measurements or glucose management or receipt of any investigational medical product within 1 month prior to enrollment Current eating disorder such as anorexia or bulimia Currently abusing illicit drugs, marijuana, prescription drugs, or alcohol Visual impairment or hearing loss, which may compromise the participant's ability to perform all study procedures safely, as determined by the investigator One or more episodes of diabetic ketoacidosis (DKA) requiring hospitalization within six months prior to screening Working night shifts Untreated celiac disease, hyperthyroidism, or hypothyroidism Clinically significant nephropathy (eGFR <45 mL/min) or on dialysis. Creatinine to determine eGFR must have been obtained as part of usual care within 12 months prior to enrollment (if not available, at time of enrollment, screening can proceed but it must be available prior to randomization)

Bergenstal RM, Tamborlane WV, Ahmann A, Buse JB, Dailey G, Davis SN, Joyce C, Peoples T, Perkins BA, Welsh JB, Willi SM, Wood MA; STAR 3 Study Group. Effectiveness of sensor-augmented insulin-pump therapy in type 1 diabetes. N Engl J Med. 2010 Jul 22;363(4):311-20. doi: 10.1056/NEJMoa1002853. Epub 2010 Jun 29. Erratum In: N Engl J Med. 2010 Sep 9;363(11):1092.

Phillip M, Battelino T, Atlas E, Kordonouri O, Bratina N, Miller S, Biester T, Stefanija MA, Muller I, Nimri R, Danne T. Nocturnal glucose control with an artificial pancreas at a diabetes camp. N Engl J Med. 2013 Feb 28;368(9):824-33. doi: 10.1056/NEJMoa1206881.

Bergenstal RM, Klonoff DC, Garg SK, Bode BW, Meredith M, Slover RH, Ahmann AJ, Welsh JB, Lee SW, Kaufman FR; ASPIRE In-Home Study Group. Threshold-based insulin-pump interruption for reduction of hypoglycemia. N Engl J Med. 2013 Jul 18;369(3):224-32. doi: 10.1056/NEJMoa1303576. Epub 2013 Jun 22.

Nimri R, Muller I, Atlas E, Miller S, Fogel A, Bratina N, Kordonouri O, Battelino T, Danne T, Phillip M. MD-Logic overnight control for 6 weeks of home use in patients with type 1 diabetes: randomized crossover trial. Diabetes Care. 2014 Nov;37(11):3025-32. doi: 10.2337/dc14-0835. Epub 2014 Jul 30.

Nimri R, Phillip M. Artificial pancreas: fuzzy logic and control of glycemia. Curr Opin Endocrinol Diabetes Obes. 2014 Aug;21(4):251-6. doi: 10.1097/MED.0000000000000073.

Nimri R, Muller I, Atlas E, Miller S, Kordonouri O, Bratina N, Tsioli C, Stefanija MA, Danne T, Battelino T, Phillip M. Night glucose control with MD-Logic artificial pancreas in home setting: a single blind, randomized crossover trial-interim analysis. Pediatr Diabetes. 2014 Mar;15(2):91-9. doi: 10.1111/pedi.12071. Epub 2013 Aug 15.

Bergenstal RM, Garg S, Weinzimer SA, Buckingham BA, Bode BW, Tamborlane WV, Kaufman FR. Safety of a Hybrid Closed-Loop Insulin Delivery System in Patients With Type 1 Diabetes. JAMA. 2016 Oct 4;316(13):1407-1408. doi: 10.1001/jama.2016.11708. No abstract available.

Nimri R, Bratina N, Kordonouri O, Avbelj Stefanija M, Fath M, Biester T, Muller I, Atlas E, Miller S, Fogel A, Phillip M, Danne T, Battelino T. MD-Logic overnight type 1 diabetes control in home settings: A multicentre, multinational, single blind randomized trial. Diabetes Obes Metab. 2017 Apr;19(4):553-561. doi: 10.1111/dom.12852. Epub 2017 Jan 19.

Dovc K, Battelino T, Beck RW, Sibayan J, Bailey RJ, Calhoun P, Turcotte C, Weinzimer S, Smigoc Schweiger D, Nimri R, Bergenstal RM. Impact of Temporary Glycemic Target Use in the Hybrid and Advanced Hybrid Closed-Loop Systems. Diabetes Technol Ther. 2022 Nov;24(11):848-852. doi: 10.1089/dia.2022.0153. Epub 2022 Aug 9.

Hood KK, Laffel LM, Danne T, Nimri R, Weinzimer SA, Sibayan J, Bailey RJ, Schatz D, Bratina N, Bello R, Punel A, Calhoun P, Beck RW, Bergenstal RM, Phillip M. Lived Experience of Advanced Hybrid Closed-Loop Versus Hybrid Closed-Loop: Patient-Reported Outcomes and Perspectives. Diabetes Technol Ther. 2021 Dec;23(12):857-861. doi: 10.1089/dia.2021.0153. Epub 2021 Oct 26.

Bergenstal RM, Nimri R, Beck RW, Criego A, Laffel L, Schatz D, Battelino T, Danne T, Weinzimer SA, Sibayan J, Johnson ML, Bailey RJ, Calhoun P, Carlson A, Isganaitis E, Bello R, Albanese-O'Neill A, Dovc K, Biester T, Weyman K, Hood K, Phillip M; FLAIR Study Group. A comparison of two hybrid closed-loop systems in adolescents and young adults with type 1 diabetes (FLAIR): a multicentre, randomised, crossover trial. Lancet. 2021 Jan 16;397(10270):208-219. doi: 10.1016/S0140-6736(20)32514-9.