A fully artificial pancreas versus a hybrid artificial pancreas for type 1 diabetes: a single-centre, open-label, randomised controlled, crossover, non-inferiority trial

An automatic deep reinforcement learning bolus calculator for automated insulin delivery systems

In hybrid automatic insulin delivery (HAID) systems, meal disturbance is compensated by feedforward control, which requires the announcement of the meal by the patient with type 1 diabetes (DM1) to achieve the desired glycemic control performance. The calculation of insulin bolus in the HAID system is based on the amount of carbohydrates (CHO) in the meal and patient-specific parameters, i.e. carbohydrate-to-insulin ratio (CR) and insulin sensitivity-related correction factor (CF). The estimation of CHO in a meal is prone to errors and is burdensome for patients. This study proposes a fully automatic insulin delivery (FAID) system that eliminates patient intervention by compensating for unannounced meals. This study exploits the deep reinforcement learning (DRL) algorithm to calculate insulin bolus for unannounced meals without utilizing the information on CHO content. The DRL bolus calculator is integrated with a closed-loop controller and a meal detector (both previously developed by our group) to implement the FAID system. An adult cohort of 68 virtual patients based on the modified UVa/Padova simulator was used for in-silico trials. The percentage of the overall duration spent in the target range of 70-180 mg/dL was 71.2 % and 76.2 % , < 70 mg/dL was 0.9 % and 0.1 % , and > 180 mg/dL was 26.7 % and 21.1 % , respectively, for the FAID system and HAID system utilizing a standard bolus calculator (SBC) including CHO misestimation. The proposed algorithm can be exploited to realize FAID systems in the future.

Automated Insulin Delivery Systems in Pediatric Type 1 Diabetes: A Narrative Review

This narrative review assesses the use of automated insulin delivery (AID) systems in managing persons with type 1 diabetes (PWD) in the pediatric population. It outlines current research, the differences between various AID systems currently on the market and the challenges faced, and discusses potential opportunities for further advancements within this field. Furthermore, the narrative review includes various expert opinions on how different AID systems can be used in the event of challenges with rapidly changing insulin requirements. These include examples, such as during illness with increased or decreased insulin requirements and during physical activity of different intensities or durations. Case descriptions give examples of scenarios with added user-initiated actions depending on the type of AID system used. The authors also discuss how another AID system could have been used in these situations.

The role of automated insulin delivery technology in diabetes

The role of automated insulin delivery systems in diabetes is expanding. Hybrid closed-loop systems are being used in routine clinical practice for treating people with type 1 diabetes. Encouragingly, real-world data reflects the performance and usability observed in clinical trials. We review the commercially available hybrid closed-loop systems, their distinctive features and the associated real-world data. We also consider emerging indications for closed-loop systems, including the treatment of type 2 diabetes where variability of day-to-day insulin requirements is high, and other challenging applications for this technology. We discuss issues around access and implementation of closed-loop technology, and consider the limitations of present closed-loop systems, as well as innovative approaches that are being evaluated to improve their performance.

Insulin's Legacy: A Century of Breakthroughs and Innovation

The clinical use of insulin to treat diabetes started just over 100 years ago. The past century has witnessed remarkable innovations in insulin therapy, evolving from animal organ extracts to bioengineered human insulins with ultra-rapid onset or prolonged action. Insulin delivery systems have also progressed to current automated insulin delivery systems. In this review, we discuss the history of insulin and the pharmacology and therapeutic indications for a variety of available insulins, especially newer analog insulins. We highlight recent advances in insulin pump therapy and review evidence on the therapeutic benefits of automated insulin delivery. As with any form of progress, there have been setbacks, and insulin has recently faced an affordability crisis. We address the challenges of insulin accessibility, along with recent progress to improve insulin affordability. Finally, we mention research on glucose-responsive insulins and hepato-preferential insulins that are likely to shape the future of insulin therapy.

A model of subcutaneous pramlintide pharmacokinetics and its effect on gastric emptying: Proof-of-concept based on populational data

Pramlintide, an amylin analog, has been coming up as an agent in type 1 diabetes dual-hormone therapies (insulin/pramlintide). Since pramlintide slows down gastric emptying, it allows for easing glucose control and reducing the burden of meal announcements. Pre-clinical in silico evaluations are a key step in the development of any closed-loop strategy. However, mathematical models are needed, and pramlintide models in the literature are scarce. This work proposes a proof-of-concept pramlintide model, describing its subcutaneous pharmacokinetics (PK) and its effect on gastric emptying (PD). The model is validated with published populational (clinical) data. The model development is divided into three stages: intravenous PK, subcutaneous PK, and PD modeling. In each stage, a set of model structures are proposed, and their performance is assessed using the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). In order to evaluate the modulation of the rate of gastric emptying, a literature meal model was used. The final pramlintide model comprises four compartments and a function that modulates gastric emptying depending on plasma pramlintide. Results show an appropriate fit for the data. Some aspects are left as open questions due to the lack of specific data (e.g., the influence of meal composition on the pramlintide effect). Moreover, further validation with individual data is necessary to propose a virtual cohort of patients.

NIH Fifth Artificial Pancreas Workshop 2023: Meeting Report: The Fifth Artificial Pancreas Workshop: Enabling Fully Automation, Access, and Adoption

The Fifth Artificial Pancreas Workshop: Enabling Fully Automation, Access, and Adoption was held at the National Institutes of Health (NIH) Campus in Bethesda, Maryland on May 1 to 2, 2023. The organizing Committee included representatives of NIH, the US Food and Drug Administration (FDA), Diabetes Technology Society, Juvenile Diabetes Research Foundation (JDRF), and the Leona M. and Harry B. Helmsley Charitable Trust. In previous years, the NIH Division of Diabetes, Endocrinology, and Metabolic Diseases along with other diabetes organizations had organized periodic workshops, and it had been seven years since the NIH hosted the Fourth Artificial Pancreas in July 2016. Since then, significant improvements in insulin delivery have occurred. Several automated insulin delivery (AID) systems are now commercially available. The workshop featured sessions on: (1) Lessons Learned from Recent Advanced Clinical Trials and Real-World Data Analysis, (2) Interoperability, Data Management, Integration of Systems, and Cybersecurity, Challenges and Regulatory Considerations, (3) Adaptation of Systems Through the Lifespan and Special Populations: Are Specific Algorithms Needed, (4) Development of Adaptive Algorithms for Insulin Only and for Multihormonal Systems or Combination with Adjuvant Therapies and Drugs: Clinical Expected Outcomes and Public Health Impact, (5) Novel Artificial Intelligence Strategies to Develop Smarter, More Automated, Personalized Diabetes Management Systems, (6) Novel Sensing Strategies, Hormone Formulations and Delivery to Optimize Close-loop Systems, (7) Special Topic: Clinical and Real-world Viability of IP-IP Systems. "Fully automated closed-loop insulin delivery using the IP route," (8) Round-table Panel: Closed-loop performance: What to Expect and What are the Best Metrics to Assess it, and (9) Round-table Discussion: What is Needed for More Adaptable, Accessible, and Usable Future Generation of Systems? How to Promote Equitable Innovation? This article summarizes the discussions of the Workshop.

Mitigating iftar-related glycemic excursions in adolescents and young adults with type 1 diabetes on MiniMed™ 780G advanced hybrid closed loop system: a randomized clinical trial for adjunctive oral vildagliptin therapy during Ramadan fasting

BACKGROUND

Ramadan Iftar meal typically causes glucose excursions. Dipeptidyl peptidase-4 inhibitors increase glucagon-like peptide-1 and thus, decrease blood glucose levels with low risk of hypoglycemia.

AIM

To investigate the efficacy and safety of vildagliptin as an add-on therapy on glucose excursions of Iftar Ramadan meals among adolescents and young adults with type 1 diabetes mellitus (T1DM) using advanced hybrid closed-loop (AHCL) treatment.

METHODS

Fifty T1DM patients on MiniMed™ 780G AHCL were randomly assigned either to receive vildagliptin (50 mg tablet) with iftar meal during Ramadan month or not. All participants received pre-meal insulin bolus based on insulin-to-carbohydrate ratio (ICR) for each meal constitution.

RESULTS

Vildagliptin offered blunting of post-meal glucose surges (mean difference - 30.3 mg/dL [- 1.7 mmol/L] versus - 2.9 mg/dL [- 0.2 mmol/L] in control group; p < 0.001) together with concomitant exceptional euglycemia with time in range (TIR) significantly increased at end of Ramadan in intervention group from 77.8 ± 9.6% to 84.7 ± 8.3% (p = 0.016) and time above range (180-250 mg/dL) decreased from 13.6 ± 5.1% to 9.7 ± 3.6% (p = 0.003) without increasing hypoglycemia. A significant reduction was observed in automated daily correction boluses and total bolus dose by 23.9% and 16.3% (p = 0.015 and p < 0.023, respectively) with less aggressive ICR settings within intervention group at end of Ramadan. Coefficient of variation was improved from 37.0 ± 9.4% to 31.8 ± 7.1%; p = 0.035). No severe hypoglycemia or diabetic ketoacidosis were reported.

CONCLUSION

Adjunctive vildagliptin treatment mitigated postprandial hyperglycemia compared with pre-meal bolus alone. Vildagliptin significantly increased TIR while reducing glycemic variability without compromising safety. Trial registration This trial was registered under ClinicalTrials.gov Identifier no. NCT06021119.

Fully Closed-Loop Glucose Control Compared With Insulin Pump Therapy With Continuous Glucose Monitoring in Adults With Type 1 Diabetes and Suboptimal Glycemic Control: A Single-Center, Randomized, Crossover Study

OBJECTIVE

We evaluated the safety and efficacy of fully closed-loop with ultrarapid insulin lispro in adults with type 1 diabetes and suboptimal glycemic control compared with insulin pump therapy with continuous glucose monitoring (CGM).

RESEARCH DESIGN AND METHODS

This single-center, randomized, crossover study enrolled 26 adults with type 1 diabetes using insulin pump therapy with suboptimal glycemic control (mean ± SD, age 41 ± 12 years, HbA1c 9.2 ± 1.1% [77 ± 12 mmol/mol]). Participants underwent two 8-week periods of unrestricted living to compare fully closed-loop with ultrarapid insulin lispro (CamAPS HX system) with insulin pump therapy with CGM in random order.

RESULTS

In an intention-to-treat analysis, the proportion of time glucose was in range (primary end point 3.9-10.0 mmol/L) was higher during closed-loop than during pump with CGM (mean ± SD 50.0 ± 9.6% vs. 36.2 ± 12.2%, mean difference 13.2 percentage points [95% CI 9.5, 16.9], P < 0.001). Time with glucose >10.0 mmol/L and mean glucose were lower during closed-loop than during pump with CGM (mean ± SD time >10.0 mmol/L: 49.0 ± 9.9 vs. 62.9 ± 12.6%, mean difference -13.3 percentage points [95% CI -17.2, -9.5], P < 0.001; mean ± SD glucose 10.7 ± 1.1 vs. 12.0 ± 1.6 mmol/L, mean difference -1.2 mmol/L [95% CI -1.8, -0.7], P < 0.001). The proportion of time with glucose <3.9 mmol/L was similar between periods (median [interquartile range (IQR)] closed-loop 0.88% [0.51-1.55] vs. pump with CGM 0.64% [0.28-1.10], P = 0.102). Total daily insulin requirements did not differ (median [IQR] closed-loop 51.9 units/day [35.7-91.2] vs. pump with CGM 50.7 units/day [34.0-70.0], P = 0.704). No severe hypoglycemia or ketoacidosis occurred.

CONCLUSIONS

Fully closed-loop insulin delivery with CamAPS HX improved glucose control compared with insulin pump therapy with CGM in adults with type 1 diabetes and suboptimal glycemic control.

Integrating metabolic expenditure information from wearable fitness sensors into an AI-augmented automated insulin delivery system: a randomised clinical trial

BACKGROUND

Exercise can rapidly drop glucose in people with type 1 diabetes. Ubiquitous wearable fitness sensors are not integrated into automated insulin delivery (AID) systems. We hypothesised that an AID can automate insulin adjustments using real-time wearable fitness data to reduce hypoglycaemia during exercise and free-living conditions compared with an AID not automating use of fitness data.

METHODS

Our study population comprised of individuals (aged 21-50 years) with type 1 diabetes from from the Harold Schnitzer Diabetes Health Center clinic at Oregon Health and Science University, OR, USA, who were enrolled into a 76 h single-centre, two-arm randomised (4-block randomisation), non-blinded crossover study to use (1) an AID that detects exercise, prompts the user, and shuts off insulin during exercise using an exercise-aware adaptive proportional derivative (exAPD) algorithm or (2) an AID that automates insulin adjustments using fitness data in real-time through an exercise-aware model predictive control (exMPC) algorithm. Both algorithms ran on iPancreas comprising commercial glucose sensors, insulin pumps, and smartwatches. Participants executed 1 week run-in on usual therapy followed by exAPD or exMPC for one 12 h primary in-clinic session involving meals, exercise, and activities of daily living, and 2 free-living out-patient days. Primary outcome was time below range (<3·9 mmol/L) during the primary in-clinic session. Secondary outcome measures included mean glucose and time in range (3·9-10 mmol/L). This trial is registered with ClinicalTrials.gov, NCT04771403.

FINDINGS

Between April 13, 2021, and Oct 3, 2022, 27 participants (18 females) were enrolled into the study. There was no significant difference between exMPC (n=24) versus exAPD (n=22) in time below range (mean [SD] 1·3% [2·9] vs 2·5% [7·0]) or time in range (63·2% [23·9] vs 59·4% [23·1]) during the primary in-clinic session. In the 2 h period after start of in-clinic exercise, exMPC had significantly lower mean glucose (7·3 [1·6] vs 8·0 [1·7] mmol/L, p=0·023) and comparable time below range (1·4% [4·2] vs 4·9% [14·4]). Across the 76 h study, both algorithms achieved clinical time in range targets (71·2% [16] and 75·5% [11]) and time below range (1·0% [1·2] and 1·3% [2·2]), significantly lower than run-in period (2·4% [2·4], p=0·0004 vs exMPC; p=0·012 vs exAPD). No adverse events occurred.

INTERPRETATION

AIDs can integrate exercise data from smartwatches to inform insulin dosing and limit hypoglycaemia while improving glucose outcomes. Future AID systems that integrate exercise metrics from wearable fitness sensors may help people living with type 1 diabetes exercise safely by limiting hypoglycaemia.

FUNDING

JDRF Foundation and the Leona M and Harry B Helmsley Charitable Trust, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases.

The Artificial Pancreas and Type 1 Diabetes

Diabetes technologies represent a paradigm shift in type 1 diabetes care. Continuous subcutaneous insulin infusion (CSII) pumps and continuous glucose monitors (CGM) improve glycated hemoglobin (HbA1c) levels, enhance time in optimal glycemic range, limit severe hypoglycemia, and reduce diabetes distress. The artificial pancreas or closed-loop system connects these devices via a control algorithm programmed to maintain target glucose, partially relieving the person living with diabetes of this constant responsibility. Automating insulin delivery reduces the input required from those wearing the device, leading to better physiological and psychosocial outcomes. Hybrid closed-loop therapy systems, requiring user-initiated prandial insulin doses, are the most advanced closed-loop systems commercially available. Fully closed-loop systems, requiring no user-initiated insulin boluses, and dual hormone systems have been shown to be safe and efficacious in the research setting. Clinical adoption of closed-loop therapy remains in early stages despite recent technological advances. People living with diabetes, health care professionals, and regulatory agencies continue to navigate the complex path to equitable access. We review the available devices, evidence, clinical implications, and barriers regarding these innovatory technologies.

Enabling fully automated insulin delivery through meal detection and size estimation using Artificial Intelligence

We present a robust insulin delivery system that includes automated meal detection and carbohydrate content estimation using machine learning for meal insulin dosing called robust artificial pancreas (RAP). We conducted a randomized, single-center crossover trial to compare postprandial glucose control in the four hours following unannounced meals using a hybrid model predictive control (MPC) algorithm and the RAP system. The RAP system includes a neural network model to automatically detect meals and deliver a recommended meal insulin dose. The meal detection algorithm has a sensitivity of 83.3%, false discovery rate of 16.6%, and mean detection time of 25.9 minutes. While there is no significant difference in incremental area under the curve of glucose, RAP significantly reduces time above range (glucose >180 mg/dL) by 10.8% (P = 0.04) and trends toward increasing time in range (70-180 mg/dL) by 9.1% compared with MPC. Time below range (glucose <70 mg/dL) is not significantly different between RAP and MPC.

Bolus Insulin calculation without meal information. A reinforcement learning approach

In continuous subcutaneous insulin infusion and multiple daily injections, insulin boluses are usually calculated based on patient-specific parameters, such as carbohydrates-to-insulin ratio (CR), insulin sensitivity-based correction factor (CF), and the estimation of the carbohydrates (CHO) to be ingested. This study aimed to calculate insulin boluses without CR, CF, and CHO content, thereby eliminating the errors caused by misestimating CHO and alleviating the management burden on the patient. A Q-learning-based reinforcement learning algorithm (RL) was developed to optimise bolus insulin doses for in-silico type 1 diabetic patients. A realistic virtual cohort of 68 patients with type 1 diabetes that was previously developed by our research group, was considered for the in-silico trials. The results were compared to those of the standard bolus calculator (SBC) with and without CHO misestimation using open-loop basal insulin therapy. The percentage of the overall duration spent in the target range of 70-180 mg/dL was 73.4% and 72.37%, <70 mg/dL was 1.96 and 0.70%, and >180 mg/dL was 23.40 and 24.63%, respectively, for RL and SBC without CHO misestimation. The results revealed that RL outperformed SBC in the presence of CHO misestimation, and despite not knowing the CHO content of meals, the performance of RL was similar to that of SBC in perfect conditions. This algorithm can be incorporated into artificial pancreas and automatic insulin delivery systems in the future.

A Comparison of Postprandial Glucose Control in the Medtronic Advanced Hybrid Closed-Loop System Versus 670G

Background: We recently reported that use of an "advanced" hybrid closed-loop system reduced hyperglycemia without increasing hypoglycemia compared to a first-generation system. The aim of this analysis was to evaluate whether this improved performance was specifically related to better mealtime glycemic control. Methods: We conducted a secondary analysis of postprandial glycemic control in an open-label, multinational, randomized crossover trial of 112 participants with type 1 diabetes, aged 14-29, of the Medtronic MiniMed™ 670G hybrid closed-loop system (670G) versus the Medtronic advanced hybrid closed-loop (AHCL) system, for 12 weeks each. We compared glycemic and insulin delivery metrics over a 3 h horizon across all meals to assess system performance and outcomes. Results: Overall meal size and premeal insulin on board were similar during run-in and between 670G and AHCL arms. Compared with 670G arm, premeal, peak, and mean glucose levels were numerically lower in the AHCL arm (167 ± 23, 231 ± 23, and 177 ± 20 mg/dL vs. 175 ± 23, 235 ± 23, and 180 ± 19 mg/dL, respectively), with a trend to lower hyperglycemia level 2 in AHCL arm. Adjusting for premeal glucose level, all postmeal outcomes between 670G and AHCL were statistically similar. Prandial insulin delivery also was similar in both treatment arms (21 ± 9 vs. 23 ± 10 U), with a shift in basal/bolus ratio from 28%/71% in 670G arm to 20%/80% in AHCL arm. Conclusions: Reduced hyperglycemia with AHCL compared to 670G was not related to early postprandial glycemic excursions after adjusting for premeal glucose level (<3 h after meal), but likely to later (>3 h) postprandial or overnight improvements. Further refinements to mealtime bolus algorithms and strategies may more optimally control prandial glycemic excursions.

Glucagon-like peptide-1 receptor agonists as adjunctive treatment for type 1 diabetes: Renewed opportunities through tailored approaches?

Exogenous insulin has been the mainstay treatment for individuals living with type 1 diabetes (T1D). Although there has been tremendous growth in both pharmacological and technological advancements, insulin monotherapy has proven to be insufficient for maintaining optimal glycaemic targets for most adults with T1D. At present, there is still no breakthrough for the treatment of T1D. Adjunctive pharmacotherapies might therefore complement insulin management to achieve better glycaemic control, while possibly offering additional benefits. Recent interest in re-purposing glucagon-like peptide-1 receptor agonists (GLP-1RAs), a leading antihyperglycaemic medication class approved for type 2 diabetes, has prompted the field to seek extended potential for the T1D population. The adjunctive use of GLP-1RAs has been at the forefront of T1D research, albeit with some conflicting trial findings to date. However, the potential of GLP-1 agonism for T1D may have been underestimated, possibly from missed opportunities or categorized effects. Moreover, some GLP-1RAs have demonstrated extra-pancreatic potential with emerging multi-organ protection involving the heart, kidneys, liver and brain in varied cohorts, which may bode well for the growing T1D profile of comorbid complications. This narrative review aims to summarize and critically appraise the current evidence-based literature from large-scale randomized controlled trials and closed-loop system pilot studies that examined GLP-1RAs as adjunctive therapy for T1D. Furthermore, we outline uncharted opportunities with GLP-1 agonism using versatile approaches in selected T1D populations that may inspire and re-direct future research in this field.

Current Status and Emerging Options for Automated Insulin Delivery Systems

Combining technologies including rapid insulin analogs, insulin pumps, continuous glucose monitors, and control algorithms has allowed for the creation of automated insulin delivery (AID) systems. These systems have proven to be the most effective technology for optimizing metabolic control and could hold the key to broadly achieving goal-level glycemic control for people with type 1 diabetes. The use of AID has exploded in the past several years with several options available in the United States and even more in Europe. In this article, we review the largest studies involving these AID systems, and then examine future directions for AID with an emphasis on usability.