Background and Aims

Prevalence of type 2 diabetes (T2D) is rapidly increasing worldwide. Although clinical trials of insulin treatment show good results, real world outcomes are poor. This is mainly caused by lack of adherence due to complexity of the treatment. Dose need is highly individual, dose adjustments are empirical, and glycemic targets should be set by clinicians based on physiological risk. In many cases, basal insulin dose adjustments are only performed during clinic visits, and reaching the glycemic target can therefore take years in practice.

Methods

We propose an adaptive dose guidance algorithm with automated glycemic target setting. The algorithm uses a dose estimation approach, developed using clinical data from 1.925 insulin naïve people with T2D. Based on self-monitored blood glucose and insulin injection data, the adaptive individual dose estimate and its uncertainty is used to propose a next safe and efficient dose. The glycemic target is automatically chosen to minimize risk of hypoglycemia. We test the performance in silico and compare to a simple standard of care algorithm.

Results

In a simulated scenario with low adherence to dose adjustments, 55% and 79% of participants reached the glycemic target after one year using the standard of care and proposed algorithm, respectively. The number of hypoglycemia events remained the same. Results of a simulated high adherence scenario indicate that the performance of the proposed and standard of care algorithms is similar.

Conclusions

The proposed algorithm has the potential to improve glycemic outcomes in a real-world setting where adherence is sub-optimal.

Background and Aims

We evaluated a new insulin pump designed to stop insulin delivery when a hypoglycemic value is predicted in young children with type 1 diabetes during a 4-day camp and after 10-15 days.

Methods

Children with type 1 diabetes (n = 28, age 6-8 years, from 14 Italian pediatric centers) participated in a 4-day camp to evaluate the efficacy and safety of a Tandem t:slim X2 pump with Basal-IQ integrated with a Dexcom G6 sensor, after a specific educational path. Data were compared with 15 days before and after the camp. The primary outcome was time in range (TIR). Secondary outcomes were time in hypo (<70 mg/dl) and time in hyper (>180 mg/dl), number of hypos during the camp, number of hypos needing glucose supplementation.

Results

Mean TIR before camp was 61.2±11.7 and increased to 67.1±12.9 during the camp (p=0.019), remaining stable 15 days after camp 65.9±10.4 (p=0.031 vs before and p=0.684 vs camp). Time in hypo was similar during the three periods (3.1±2.7 vs 3.4±2.8 vs 3.3±3.4, p=0.527), while time in hyper significantly decreased (35.7±12.6 vs 29.5±13.6, p=0.022, and vs 31.1±11.5, p=0.034 vs before and p=0.469 vs camp). No severe hypoglicemia occurred. During the camp a total of 49 hypoglycemic events have been reported, only 14 of them (28.6%) needing a glucose supplementation. Indeed, during a high intensity, long duration exercise (3-hour mountain climbing) 26/28 children had hypoglycemia, but only 6 (23%) needed glucose.

Conclusions

The Tandem t:slim X2 Basal-IQ system significantly improved TIR in young children, without increasing time spent in hypoglycemia.

Background and Aims

Safe glycaemic control during and after exercise is a challenge in type 1 diabetes(T1D). We aimed to analyse the performance of a new Closed-Loop(CL) controller under announced and unannounced exercise.

Methods

Adults with T1D and HbA_1c between 6.0-8.5% were eligible. Every subject performed three supervised inpatient studies randomized order and different days. CL included announced and unannounced exercise while Open-Loop(OL) included announced exercise only. Three sets of 15 minutes of cicloergometer at 70% of VO₂max with 5 minutes of rest between them were performed. CL is based on a glucose controller built in Android platform. The controller receives glucose measurements every 5 minutes from a Continuous Glucose Monitor(CGM) and calculates insulin delivery by automatically changing, at every time interval, basal rate and making fast-acting carbohydrate(CH) intake suggestions in case of impending hypoglycaemia. During announced CL operation the controller includes a mitigation module, triggering feed-forward actions for hypoglycaemia prevention. Primary endpoint was the number of hypoglycaemic episodes during and after exercise. Secondary outcomes included Time-in-Range(TIR), CH ingestion and insulin infusion.

Results

Ten patients with T1D were included: 7 men; 40.4±7.0 years-old; 23.8±13.2 years of disease duration and HbA_1c 7.3±0.8%. Announced and unannounced CL were associated with half of hypoglycaemic episodes compared to OL, with less time hypoglycaemia and lower CH consumption. Outcome details are shown in Table 1.

Conclusions

CL controller outperforms OL therapy for both, announced and unannounced exercise, in terms of number of hypoglycaemic episodes and TIR. CL is able to perform better glucose control with less supplemental CH compared to OL.

Background and Aims

A prototype of our integrated and fully automated bihormonal artificial pancreas was successfully tested in a pilot study at home. Subsequently, a product version was developed and locked for CE-mark registration. The aim of this pivotal CE mark trial was to assess the performance and safety of this artificial pancreas.

Methods

Twenty-three adult patients with type 1 diabetes completed this randomized open cross-over study during daily life at home. Two weeks of closed loop glucose control with the artificial pancreas (Figure) were compared with two weeks of insulin pump therapy with Flash Glucose Monitoring (n=12), Predictive Low Glucose Suspend (n=5) or SMBG (n=6) and blinded CGM for data collection. A training period of four days preceded the closed loop period.

Results

The median [IQR] time in range (3.9-10 mmol/l) was higher during closed loop (86.6 [84.9-88.3]%) compared to the control treatment (53.9 [49.7-67.2]%, p<.001). The median glucose level decreased from 9.3 [8.3-9.9] mmol/l to 7.2 [7.0-7.4] mmol/l with the artificial pancreas (p<.001). Time in hypoglycemia was 2.0 [0.7-3.6 ]% during control treatment and 0.4 [0.1-0.8]% with the artificial pancreas (p<.001). No severe hypoglycemia or ketoacidosis occurred during the study.

Conclusions

This study demonstrates that our bihormonal artificial pancreas provides better glucose control than standard open loop therapy in adult patients with type 1 diabetes. The artificial pancreas met the predefined performance and safety requirements for the clinical validation and therefore we expect that the system will receive a CE-mark in 2020.

Background and Aims

Combining the bionic pancreas (iLet^TM) with fast-acting insulin aspart (faster aspart) is of interest given the improved pharmacological and glycaemic profile of faster aspart versus conventional rapid-acting insulin analogues. We investigated the safety and glycaemic control of the insulin-only configuration of the iLet^TM delivering faster aspart using different algorithm t_max settings in adults with type 1 diabetes.

Methods

We performed a single-centre, single-blind, crossover (two 7-day treatment periods) escalation trial over three sequential cohorts in which subjects were randomised to a default t_max (t₆₅) setting followed by a non-default t_max setting (t₅₀ [cohort 1], t₄₀ [cohort 2], t₃₀ [cohort 3]), or vice versa, all with faster aspart. Each cohort randomised eight new subjects if escalation stopping criteria were not met in the previous cohort.

Results

Two subjects discontinued treatment, one due to “low blood glucose” during the first treatment period of cohort 3 (t₃₀). No severe hypoglycaemic episodes were reported and there were no clinically significant differences in adverse events between groups. Mean time in low (sensor) glucose (<54 mg/dL, primary endpoint) was <1.0% for all t_max settings (Table). Except for the default t_max setting in cohort 1, the mean time in range (70–180 mg/dL) was >70%. Mean glucose in cohorts 1 and 2 was significantly lower at non-default versus default t_max settings, with comparable insulin dosing.

Conclusions

There were no safety concerns with faster aspart in the iLet^TM at non-default t_max settings. Improvements in glycaemic control at non-default t_max settings were observed.

Background and Aims

People with type 1 diabetes (T1D) have personal strategies for maintaining euglycemia while exercising which will serve them well during training. Their strategy often yields wildly differing results during a competition. The influence of competition stress (notably, epinephrine causing prompt elevations in glucose) and high-intensity and/or long-duration efforts can yield dysglycemia.

Methods

Recreational athletes with T1D were recruited for the study. A total of 5 runners (2M/3F) with 9 races have been studied to date. Participants completed an exercise stress test, an athletic competition, and a training run at the same pace as the athletic competition for a shorter duration and without the stress of the race. The carbohydrate and insulin intake prior to the athletic competition and the non-competitive race-intensity run were matched.

Results

The observed competition glucose trends are similar or elevated compared to the non-competition glucose trends of the same intensity at all distances studied. Within competitions for the same individual, in some cases the elevation observed during competition is not consistent despite reports of a similar level of stress in both competitions. When comparing these two cases, elevated heart rate in the race compared with the training session was correlated with elevated glucose in some individuals. Fewer hypoglycemic events were observed during athletic competitions than in non-competitive exercise sessions.

Conclusions

Competition stress may lead to an elevated glucose trend. Perceived stress level as reported by the individual is insufficient for determination of glycemia whereas an elevated heart rate may indicate an increase in glycemia at same running pace.

Background and Aims

Previous trials demonstrated a 8.6% per protocol (PP) improvement with the single hormone closed-loop system Diabeloop’s DBLG1 (2017, with Cellnovo pump), then a 14.0% PP improvement (2018 with the Kaleido pump). The objective of the present study was to evaluate the efficacy and safety of the closed-loop system DBLG1 with an improved algorithm and the Kaleido pump, in patients with type 1 diabetes (T1D) in real life situation.

Methods

25 non selected T1D patients in Corbeil and Grenoble diabetes centers, after a 2-week run-in period with G6 sensor and usual pump, were provided with the commercial setting of DBLG1 System with an improved algorithm. They were taught to manage it , during a day-hospitalization, and were checked during a visit, the day after. There was no structured remote monitoring.

Results

After 20 weeks follow-up, compared to run-in period, time in range (TIR) 70-180 mg/dl, improved by 17.1% (53.7% v 70.8%), TIR 50 – 70 mg/dl was halved (2.38% v 1.20%). TIR<50 mg/dl was very low (0.15% v 0.15%), TIR> 250 mg/dl decreased by 2/3 (17.7% v 7.3%). Mean blood glucose decreased from 179.4 mg/dl to 157.5 mg/dl. Satisfaction of patients was great, there was no adverse events due to the algorithm, and the only complaint was about frequent system disconnection. Nevertheless, closed loop was in functional mode for 88% of time during the 20 average weeks of the survey.

Conclusions

The DBLG1 System is able to largely improve glycemic control in real life situation, without serious adverse events.

Background and Aims

Studies have increasingly identified sleep disturbances in people with T1D. Older adults and young children may be a particularly vulnerable population to sleep disturbances. Anecdotally, it is reported that use of CLC improves sleep quality and quantity, but objective data on this effect are not available. This study assesses sleep outcomes of CLC compared to sensor-augmented pump therapy (SAP).

Methods

Participants with T1D on insulin pump therapy were enrolled in two age groups: older adults (ages 65+) and young children (ages 6-10). Participants completed an initial 4-week study with SAP using their own pump and a study CGM (Dexcom G6) followed by a 4-week phase of CLC (Control-IQ, Tandem Diabetes Care). Sleep was assessed using Actigraphy watches worn the last 10 days of each study phase. The Pittsburgh Sleep Quality Index (PSQI) questionnaire was administered at baseline and following each study phase.

Results

A total of 18 older adults (mean age 69) and 15 children (mean age 9) and their parents were enrolled in this ongoing study. Preliminary data analysis was performed on a subset of 4 older male adults. Sleep duration increased an average of 19 minutes during CLC in 2 subjects and remained virtually unaltered in the others. Sleep efficiency did not change. PSQI score improved in the two participants with worse scores on SAP. Results for each participant are shown below.

Conclusions

Though preliminary, these results suggest that the use of CLC may have a positive impact on the quality of sleep for some older adults with T1D.

Background and Aims

Due to significant glycemic excursions, exercising with T1D can be challenging. Closed loop (CL) has been shown to improve glycemia in people with T1D.

The aim was to evaluate CL performance compared with standard therapy in people with T1D undertaking exercise.

Methods

Ten adults with T1D were randomly recruited from a large randomised controlled trial investigating CL versus standard therapy. Four participants were randomised to the standard therapy and 6 were randomised to CL. Participants undertook high-intensity-interval-exercise (HIIE) and moderate-intensity-exercise (MIE), 45-minutes duration, in random-order. Frequent venous samples measured glucose, ketones, lactate, free insulin, and counter-regulatory hormones. The primary outcome was CGM time-in-target range from beginning of exercise to 24 hours post exercise.

Results

Compared with standard therapy, CL resulted in improved time-in-range (66.5% vs 42%) and reduced time in hypoglycemia (11.3% vs 8.6%) for aerobic exercise and improved mean sensor glucose (9.8mmol/L vs 8.2mmol/L, p=0.0237) with no increase in hypoglycemia for anaerobic exercise.

Glucose metrics from beginning of exercise to 24hrs post exercise

Conclusions

Closed loop results in improved glycaemia compared with standard therapy during and after exercise without increase in hypoglycemia.