Bringing closed-loop home: recent advances in closed-loop insulin delivery

Transcription

1 REVIEW C URRENT OPINION Bringing closed-loop home: recent advances in closed-loop insulin delivery Hood Thabit and Roman Hovorka Purpose of review To highlight the recent advances in closed-loop research, the development and progress towards utilizing closed loop outside of the clinical research setting and at patients homes. Recent findings In spite of the modern insulin therapy in type 1 diabetes, hypoglycaemia is still a major limiting factor. This often leads to suboptimal glycaemic control and risk of diabetes complications. Closed loop has been shown to improve glycaemic control whilst avoiding hypoglycaemia. Incremental progress has been made in this field, from the use of automated systems and bihormonal closed-loop systems in clinical research facility settings under close supervision to the use of ambulatory closed-loop prototype at patients homes in free-living conditions. Different population of patients with type 1 diabetes and control algorithm approaches have been studied, assessing the efficacy and safety. Transitional and home studies present different challenges at achieving better glycaemic outcome with closed loop. Improved glucose sensor reliability may accelerate the clinical use and faster insulin analogues increase the clinical utility. Summary Results and experience with closed-loop insulin delivery have been encouraging, leading the way for future improvements and development in the field, to make closed loop suitable for use in clinical practice. Keywords algorithm, ambulatory, automated, closed-loop, home studies INTRODUCTION The ultimate goal of diabetes therapy is to maintain normoglycaemia and avoid diabetes-related complications. Glycaemic control with insulin therapy in type 1 diabetes is influenced by insulin-dependent factors such as insulin kinetics, dosage and timing of insulin delivery [1,2], as well as physiological factors, such as effects of circadian rhythm and illness on insulin sensitivity [3 6]. Unfortunately, one of the known risks of tightening glycaemic control with insulin is hypoglycaemia [7]. The complexities of managing insulin therapy in daily life can, therefore, be overwhelming for some patients with type 1 diabetes, resulting in suboptimal control. Utilization of real-time continuous glucose monitoring (CGM) and modern continuous subcutaneous insulin infusion (CSII) therapy has been reported to improve the overall glycaemic control [8,9,10 ]. However, studies have shown that the benefits of these devices are limited to certain age groups and those who are motivated [11 13]. Closed-loop insulin delivery, also known as the artificial pancreas, may potentially be beneficial in improving the glycaemic control in type 1 diabetes, until a biological cure can be found. Closed-loop systems deliver insulin in a continued glucose-responsive manner by coupling subcutaneous CGM and subcutaneous insulin pump delivery [14]. This novel approach differs from the conventional pump therapy by the use of a control algorithm, which autonomously directs insulin delivery based on real-time sensor glucose levels. Significant advances have been made in this field over the last 2 years. Closed-loop systems have progressed from manual input, in which researchers Wellcome Trust-MRC Institute of Metabolic Science, University of Cambridge, Cambridge, UK Correspondence to Roman Hovorka, PhD, University of Cambridge Metabolic Research Laboratories, NIHR Cambridge Biomedical Research Centre, Wellcome Trust-MRC Institute of Metabolic Science, Box 289, Addenbrooke s Hospital, Hills Road, Cambridge, CB2 0QQ, UK. Tel: ; fax: ; rh347@ cam.ac.uk Curr Opin Endocrinol Diabetes Obes 2014, 21: DOI: /MED X ß 2014 Wolters Kluwer Health Lippincott Williams Wilkins

2 Diabetes and the endocrine pancreas I KEY POINTS Hypoglycaemia remains a significant risk of conventional insulin therapy in type 1 diabetes. Closed-loop system utilizes a control algorithm that directs insulin delivery in a glucose-responsive manner by coupling continuous glucose monitoring with insulin pump delivery. In controlled research facility settings, closed-loop systems are superior to conventional insulin pump therapy at improving the glycaemic control, whilst reducing the risk of hypoglycaemia. Closed-loop use outside of hospital settings is feasible, demonstrating safety and usability in transitional studies. Home studies are currently underway to assess closedloop safety and performance during longer periods of unregulated daily activities. provide manual data input to the algorithm [15,16,17], to automated systems as early prototypes or aimed for home use [18,19,20,21 ]. Control algorithms and sensor glucose performance can be tested in in-silico environments, accelerating progression to clinical studies and avoiding the need for preclinical animal testing [22,23]. Presently, several technological, pharmacological and physiological hurdles still remain. Amongst them are the delayed pharmacokinetics of current rapid-acting insulin analogues to reach its maximum glucose-lowering capacity [24], the occasional large glucose-sensing errors caused by incorrect calibration or sensor drift [25 ], and to a minor extent the physiological time lag between plasma and interstitial glucose levels because of the transport of glucose from blood to the interstitial fluid, which may contribute to sensor deviation and inaccuracy [26]. Initial closed-loop studies performed under close supervision in clinical research facilities (CRFs) are proof of concept studies, assessing the safety and efficacy. More recently, transitional studies have brought closed loop from controlled research environments to out-of-hospital settings, assessing the usability and reliability as well [19,21,27 ]. Although close monitoring and supervision is still needed in transitional studies, it is nevertheless a first step towards patients using closed loop under free-living conditions, which may eventually lead to closed-loop use in standard clinical practice. Progress is currently underway with longer duration unsupervised closed-loop studies at patients homes. This review article highlights the developments and advances made over the last 2 years, as well as the future direction of this field. STUDIES IN CLINICAL RESEARCH FACILITY Formative evaluation in CRFs is important to assess the safety and efficacy of closed-loop prototypes in a controlled environment. The following section outlines the recent findings from CRF-based closedloop studies in the paediatric and adult patients with type 1 diabetes. Closed-loop studies in the paediatric age group Children with type 1 diabetes are vulnerable to the adverse effects of neuroglycopenia, an extreme example being hypoglycaemia-related seizures at night [28]. By linking CGM readings to insulin delivery, closed-loop insulin delivery can potentially limit the exposure to overnight hypoglycaemia in this vulnerable group. The first randomized crossover manual closed-loop study in adolescents with type 1 diabetes by our group established that improved blood glucose control with minimal hypoglycaemia was possible using closed loop, compared with conventional pump therapy [29]. Automated closed-loop systems currently demonstrate similar safety and efficacy overnight. In a multicentre study, an automated closed-loop with feedback control algorithm was compared with conventional CSII on two sequential nights [30]. Closed loop significantly increased the time spent in nearnormal sensor glucose range ( mmol/l ¼ mg/dl) by more than three-fold. Closed loop also led to significantly less glucose variability overnight (P < 0.001). Despite the lower mean overnight glucose levels, no nocturnal hypoglycaemia events occurred during closed loop. In a recent overnight closed-loop study of very young type 1 diabetes children (<7 years old), no difference in time within the target glucose levels was observed [31 ]. However, closed loop reduced the time spent overnight with hyperglycaemia (16.7 mmol/l or 300 mg/dl, 0.18 vs. 1.3 h; P ¼ 0.035) without causing further hypoglycaemia. Accuracy and reliability of glucose sensors during closed loop are important [25,32]. Wearing additional CGM sensors has been considered as a possible mitigation against sensor performance failure. In a recent study, patients wore two CGM sensors during closed loop [20]. A fault detection system evaluated the signal deviations between the two sensors. Larger deviations, which are of particular concern, will prompt the patient to perform a 96 Volume 21 Number 2 April 2014

3 Bringing closed-loop home Thabit and Hovorka blood glucose calibration. Compared with conventional CSII, closed-loop systems significantly improved the CGM sensor time in target ( mmol/l ¼ mg/dl) from 46.7 to 84.5% (P < ), with less hypoglycaemia (<3.3 mmol/l or 59 mg/dl, 0.9 vs. 3%; P < ). However, the use of the additional sensor did not improve sensor accuracy when compared with the reference plasma glucose. Exercise is known to cause significant perturbation of glycaemia in type 1 diabetes [33]. A potential benefit of closed-loop systems is to reduce the risk of nocturnal hypoglycaemia following daytime exercise. In a randomized controlled study, the frequency of nocturnal hypoglycaemia following an afternoon exercise (60 min of brisk walk on treadmill) was reduced from eight events during conventional CSII to one during closed loop [34 ]. A greater portion overnight was also spent with normoglycaemia whilst on closed loop. Currently, however, daytime use of closed-loop systems during meals and exercise remains challenging. A 36-h hybrid closed-loop study was recently performed in adolescents with type 1 diabetes [35 ]. During the study, prandial insulin bolus was given before main meals, and participants underwent two sessions of moderate unannounced exercise. In comparison with conventional CSII, closed-loop systems increased the time spent in glucose target range by 35% and improved the mean plasma glucose levels. Hypoglycaemia events related to exercise persisted during closed loop (9 vs. 10 occasions). This highlights the limitations faced by the current closed-loop systems in relation to the delay in action of current subcutaneous insulin analogues. Closed-loop studies in adults As a result of the inherent lag of sensor glucose readings relative to the plasma glucose levels following meals and the delay in insulin analogue absorption and action, immediate postprandial hyperglycaemia and late postprandial hypoglycaemia are amongst the challenges faced by closed loop during mealtimes. Two control algorithm approaches were recently studied, in which information regarding meal composition and times were unannounced to both algorithms, and no premeal insulin boluses were given [36]. The authors reported comparable postprandial glycaemic excursions ( vs mg/dl), with notable reduction in the risk of hypoglycaemia in one of the study arms using an enhanced control approach (8 vs. 0 events). Although the study number was small (n ¼ 4) and no control arm available, it nevertheless highlights that with further optimization of control algorithms, improved outcomes can be obtained. Others have evaluated closed-loop performance at achieving different ranges of glucose levels. Utilizing two different control algorithm approaches at preventing extremes of glycaemic excursions, known as control to range (CTR), investigators reported incremental improvements in glycaemic outcomes when an enhanced CTR approach was compared with standard CTR, with decrements in hypoglycaemia risk [37 ]. Several closed-loop algorithms are currently being studied, but few have been compared to each other. A recent multicentre study compared two model predictive control (MPC) algorithms and conventional CSII with each other in 48 adult patients with type 1 diabetes [17]. Each patient had three 24-h study visits, in which blood glucose was controlled using conventional CSII, an MPC algorithm from the University of Cambridge, and an MPC algorithm resulting from a multinational collaboration (University of Pavia, Italy, and University of Virginia and University of California at Santa Barbara). Although the time spent in target glucose range was comparable in all three study arms, an almost three-fold reduction of the time spent in hypoglycaemia was observed during closed loop. This was achieved by a lower rate of insulin delivery and measured plasma insulin levels in both closed-loop algorithms. The glycaemic outcomes between the two MPC algorithms were comparable, with less insulin delivered by the Cambridge algorithm. However, sensor overreading during the study necessitated appropriate adjustments to the control algorithm to avoid insulin overdelivery and may have caused suboptimal performance of both closed-loop arms. This highlights a further need for improvements in sensor accuracy. Bihormonal close-loop system Closed-loop insulin delivery has been shown to be superior to conventional insulin pump therapy in improving the glycaemic control, whilst reducing the risk of hypoglycaemia. The risk of hypoglycaemia is still present primarily because of delayed insulin kinetics. To mitigate against this further, dual-hormone (also known as bihormonal) closedloop systems are under investigation, combining the delivery of insulin with subcutaneous glucagon [38]. In the longest bihormonal closed-loop study to date, participants spent 51-h in a CRF, with high carbohydrate meals accompanied by meal-priming dose and structured exercise to evoke glycaemic perturbations [39 ]. Although participant numbers were small and there was no control arm, hypoglycaemia (plasma glucose <3.9 mmol/l ¼ 70 mg/dl) X ß 2014 Wolters Kluwer Health Lippincott Williams Wilkins 97

4 Diabetes and the endocrine pancreas I occurred rarely (0.7% of the 576 h of closed-loop control). Overnight, closed loop was able to maintain plasma glucose between 3.9 and 10.0 mmol/l (i.e., mg/day) over 90% of the time, in spite of the exercise period during the day. Recently, a randomized crossover controlled study of a bihormonal closed-loop system was performed in 15 adults with type 1 diabetes [16 ]. Patients stayed overnight in the CRF for 15 h, during which they exercised for 30 min and had an evening meal. The bihormonal system improved the overall proportion of time [70.7 (interquartile range ¼ IQR ) vs. 57.3% (IQR ); P ¼ 0.003] within the target range ( mmol/l between hours of 16:00 and 23:00, and mmol/l between hours of 23:00 and 07:00). Despite no differences in the amount of insulin delivered, no episodes of hypoglycaemia (<4.0 mmol/l ¼ 72 mg/ dl) were observed during closed loop (0.0 vs. 10.2%). As a result of the sensor under-reading, however, no improvements in the time spent above the target glucose range or mean plasma glucose levels were reported. Apart from the issues related to sensor accuracy and reliability, other limitations to bihormonal closed-loop systems are currently present. An additional pump device is currently needed for glucagon delivery, contributing to device burden to patients [16,39 ]. Current glucagon aqueous solution preparation is not stable for extended pump use; it forms amyloid fibrils hours after reconstitution, leading to the formation of insoluble gels which can occlude pump catheters [40]. Studies are underway to prevent glucagon fibrillation, that is, using alkaline ph medium, and reducing its biodegradability, thus making it more feasible for bihormonal closed-loop use [41]. TRANSITIONAL PHASE STUDIES: FIRST STEPS OUTSIDE The objective of transitional phase studies is to assess the feasibility of ambulatory closed-loop systems outside of hospital clinical research settings. These studies represent an intermediate phase between the CRF and home. Although patients are studied in a real-world environment outside the CRF, closed monitoring by medical and research personnel are still implemented. A smartphonebased closed-loop prototype was recently evaluated in a short-term transitional feasibility study, involving research groups from Italy, France and the USA [21 ]. Patients were studied and monitored for 42-h in a hybrid hospital hotel setting in Italy and France, and an outpatient setting in the USA. The primary objective was to test the communication system reliability of the portable close-loop platform, which is hosted on a commercially available smartphone, with its other components in an outpatient setting. Together with the smartphone, each patient wore two Dexcom Seven Plus sensors (Dex- Com, San Diego, California, USA) and an Insulet OmniPod insulin pump (Bedford, Massachusetts, USA). Remote monitoring was implemented during the outpatient utilization of the study, with closed contact available to the research physicians at all times. Closed-loop operation was reported to be successful for h, representing 97.7% of total possible time from the start to the end of the study for each patient. The study was not designed to test the algorithm performance and post hoc analyses of closed-loop vs. open-loop nights did not show any significant differences in the glycaemic control parameters. A randomized crossover study in a diabetes youth camp was performed in three countries, utilizing a portable closed-loop prototype known as the MD-Logic AP system [19 ]. In comparison to the transitional phase study described previously, the primary objective of this study was to assess the frequency and duration of nocturnal hypoglycaemia, as well as mean overnight glucose levels during closed loop. Fifty-six patients used closed loop for one night, and a sensor-augmented insulin pump on another night as control. Study clinicians monitored each patient s glucose continuously overnight, utilizing a remote-monitoring system as well as hypoglycaemia alert alarm. On closed-loop nights, the number of episodes of hypoglycaemia with sensor glucose values below 3.5 mmol/l (63 mg/dl) was significantly reduced by 30%, compared with the control arm. The median glucose values overnight were not significantly different between closed-loop and control arms [7.0 (IQR ) vs. 7.8 mmol/l (IQR )]. The number of interventions overnight requiring rescue carbohydrates for hypoglycaemia by study clinicians was reportedly similar between the two arms. This study showed that ambulatory use of closed-loop systems outside the CRF settings is feasible, albeit still dependent on the research clinicians providing remote monitoring and carbohydrate interventions when needed. The evaluation of only a single night on closed loop in the study also needs to be taken into perspective, as daily and overnight glycaemia fluctuations are commonplace in individuals with type 1 diabetes [42]. The next challenge is implementing closed loop at home under real-life uncontrolled conditions, whilst improving on the safety and usability elements Volume 21 Number 2 April 2014

5 Bringing closed-loop home Thabit and Hovorka BRINGING CLOSE-LOOP HOME Several groups are currently assessing the closedloop safety and performance during unregulated daily activities at home. As the overnight period has been the main concern for many patients and clinicians [28,43], this has been the initial focus of closed-loop home studies. In spite of the benefits of real-time CGM, benefits of real-time CGM devices on nocturnal hypoglycaemia remain limited, mainly because of the challenges of responding to glucose alarms overnight and applying proper intervention [44,45]. Thus, achieving persistent euglycaemia overnight at home in an autonomous manner will be of significant benefit. An interim analysis of 15 patients (5 adults and 10 adolescents with type 1 diabetes) using the MD-Logic AP system with remote monitoring for four nights at home compared to conventional sensor-augmented pump therapy was recently published [46 ]. The primary endpoint analyses, based on the CGM readings, demonstrated over a ten-fold reduction in the time spent below 3.9 mmol/l (70 mg/dl) on closed-loop compared with control nights per patient [3.8 (0 11.6) vs min ( ); P ¼ ], but with correcting for simultaneous use of glucose sensor to control glucose and measure outcome [47], the difference was reduced [7.6 ( ) vs ( ) min; P ¼ NS]. No significant improvement was seen in the time spent overnight between 5.0 and 7.8 mmol/l ( mg/dl). We recently evaluated the home use of unsupervised closed loop in a group of 16 adolescents with type 1 diabetes over 3 weeks compared to sensor-augmented pump therapy [48 ]. Each patient used the closed loop overnight at home on their own, without any remote monitoring. They were provided with a 24-h telephone number to contact the study team in the event of any clinical or technical issues. Overnight closed loop was utilized on 311 nights (93%), turned on at 21 : 34 (20 : : 35) and turned off at 07 : 37 (07 : : 09) operating over 10.0 ( ) hours. Closed loop significantly increased the time spent in glucose target range ( mmol/l) by a median of 15% (P < 0.001). Mean CGM glucose overnight was also reduced by 0.8 mmol/l (P < 0.001). Time when glucose was below 70 mg/dl was less during both interventions, but nights with glucose below 63 mg/dl for at least 20 min were less frequent during closed loop (10 vs. 17%, P ¼ 0.01). Despite the lower total daily insulin doses by a median of 2.3 units, P ¼ 0.009, overall 24-h glucose was reduced by a mean 0.5 mmol/l (P ¼ 0.006) during closed loop. The feasibility of a portable bihormonal fully closed-loop system (no prandial boluses) was assessed in a randomized crossover 48-h home study [27 ]. The median glucose levels, time spent in target range ( mmol/l ¼ mg/dl) and below target range during closed loop, and conventional pump therapy were comparable in 11 out of 16 patients, but five patients could not be analysed because of the technical problems with the system. Median (IQR) glucose on the second day of closed-loop intervention was significantly reduced in the closed-loop arm [7.70 (2.29) vs mmol/l (0.87); P ¼ 0.027]; however, this came at the expense of greater time spent in the hypoglycaemia range [0.0 (11.0) vs. 2.8% (9.8); P ¼ ]. Insulin infusion by closed loop was higher compared with open loop during the first day [52.0 (29.0) vs IU (22.8); P ¼ 0.001], but was comparable on the second day [50.7 (39.9) vs IU (31.7); P ¼ ]. The daily glucagon dose during closed loop was comparable on both days [day 1 vs. day 2: 2.7 (2.0) vs. 1.7 mg (1.5)], and no differences in the oral carbohydrate rescue treatments were reported in both arms. CONCLUSION Improving the glycaemic control whilst reducing the risk of hypoglycaemia, especially at night, is achievable with closed loop. Recent results from the early transitional and home studies are encouraging, demonstrating that the roadmap for real-life clinical use in type 1 diabetes is making significant progress. Current closed-loop systems are still limited by factors such as CGM reliability and delay in insulin absorption; therefore, further developments in this field would improve outcomes. Harnessing the use of this technology may help reduce the burden and improve the quality of life of patients with this chronic condition. Acknowledgements This study was supported by the Juvenile Diabetes Research Foundation (# , # , # ), Diabetes UK (BDA07/ and BDA07/ ), European Commission Framework Programme 7 (247138), NIDDK (DK085621), and NIHR Cambridge Biomedical Research Centre. Conflicts of interest H.T. reports no conflicts of interest. R.H. reports having received speaker honoraria from Minimed Medtronic, Lifescan, Eli Lilly, BBraun, and Novo Nordisk; serving on advisory panel for Animas, Minimed Medtronic, and Eli Lilly; receiving license fees from BBraun and Beckton Dickinson; and having served as a consultant to Beckton Dickinson, BBraun, Sanofi, and Profil, and patent applications X ß 2014 Wolters Kluwer Health Lippincott Williams Wilkins 99

6 Diabetes and the endocrine pancreas I REFERENCES AND RECOMMENDED READING Papers of particular interest, published within the annual period of review, have been highlighted as: of special interest of outstanding interest 1. Liebl A. Challenges in optimal metabolic control of diabetes. Diabetes Metab Res Rev 2002; 18 (Suppl. 3):S36 S Ferrannini E. Physiology of glucose homeostasis and insulin therapy in type 1 and type 2 diabetes. Endocrinol Metab Clin North Am 2012; 41: Saad A, Dalla Man C, Nandy DK, et al. Diurnal pattern to insulin secretion and insulin action in healthy individuals. Diabetes 2012; 61: Ali NA, O Brien JM, Dungan K, et al. Glucose variability and mortality in patients with sepsis. Crit Care Med 2008; 36: Polonsky KS, Given BD, Van Cauter E. Twenty-four-hour profiles and pulsatile patterns of insulin secretion in normal and obese subjects. J Clin Invest 1988; 81: Hinshaw L, Dalla Man C, Nandy DK, et al. Diurnal pattern of insulin action in type 1 diabetes: implications for a closed-loop system. Diabetes 2013; 62: Epidemiology of severe hypoglycemia in the diabetes control and complications trial. The DCCT Research Group. Am J Med 1991; 90: Bergenstal RM, Tamborlane WV, Ahmann A, et al. Effectiveness of sensoraugmented insulin-pump therapy in type 1 diabetes. N Engl J Med 2010; 363: Slover RH, Welsh JB, Criego A, et al. Effectiveness of sensor-augmented pump therapy in children and adolescents with type 1 diabetes in the STAR 3 study. Pediatr Diabetes 2012; 13: Pickup JC. Management of diabetes mellitus: is the pump mightier than the pen? Nat Rev Endocrinol 2012; 8: A review article summarizing the currently available evidence from clinical studies of insulin pump therapy. 11. Pickup JC, Freeman SC, Sutton AJ. Glycaemic control in type 1 diabetes during real time continuous glucose monitoring compared with self monitoring of blood glucose: meta-analysis of randomised controlled trials using individual patient data. BMJ 2011; 343:d3805. doi: /bmj.d Chase HP, Beck RW, Xing D, et al. Continuous glucose monitoring in youth with type 1 diabetes: 12-month follow-up of the Juvenile Diabetes Research Foundation continuous glucose monitoring randomized trial. Diabetes Technol Ther 2010; 12: Tamborlane WV, Beck RW, Bode BW, et al. Continuous glucose monitoring and intensive treatment of type 1 diabetes. N Engl J Med 2008; 359: Hovorka R. Closed-loop insulin delivery: from bench to clinical practice. Nat Rev Endocrinol 2011; 7: Hovorka R, Kumareswaran K, Harris J, et al. Overnight closed loop insulin delivery (artificial pancreas) in adults with type 1 diabetes: crossover randomised controlled studies. BMJ 2011; 342:d1855. doi: /bmj.d Haidar A, Legault L, Dallaire M, et al. Glucose-responsive insulin and glucagon delivery (dual-hormone artificial pancreas) in adults with type 1 diabetes: a randomized crossover controlled trial. CMAJ 2013; 185: This is the first randomized crossover study of a bihormonal closed-loop system in a clinical research facility setting. 17. Luijf YM, Devries JH, Zwinderman K, et al. Day and night closed-loop control in adults with type 1 diabetes mellitus: a comparison of two closed-loop algorithms driving continuous subcutaneous insulin infusion versus patient self-management. Diabetes Care 2013; 36: Elleri D, Allen JM, Nodale M, et al. Automated overnight closed-loop glucose control in young children with type 1 diabetes. Diabetes Technol Ther 2011; 13: Phillip M, Battelino T, Atlas E, et al. Nocturnal glucose control with an artificial pancreas at a diabetes camp. N Engl J Med 2013; 368: This is the first randomized crossover transitional study to demonstrate the efficacy of short-term overnight closed-loop use in diabetes camp to reduce the frequency of nocturnal hypoglycaemia. 20. O Grady MJ, Retterath AJ, Keenan DB, et al. The use of an automated, portable glucose control system for overnight glucose control in adolescents and young adults with type 1 diabetes. Diabetes Care 2012; 35: Kovatchev BP, Renard E, Cobelli C, et al. Feasibility of outpatient fully integrated closed-loop control: first studies of wearable artificial pancreas. Diabetes Care 2013; 36: This study assesses the technical feasibility of an ambulatory closed-loop system in a transitional setting without changes to glucose control. 22. Wilinska ME, Chassin LJ, Acerini CL, et al. Simulation environment to evaluate closed-loop insulin delivery systems in type 1 diabetes. J Diabetes Sci Technol 2010; 4: Kovatchev BP, Breton M, Man CD, Cobelli C. In silico preclinical trials: a proof of concept in closed-loop control of type 1 diabetes. J Diabetes Sci Technol 2009; 3: Ellingsen C, Dassau E, Zisser H, et al. Safety constraints in an artificial pancreatic beta cell: an implementation of model predictive control with insulin on board. J Diabetes Sci Technol 2009; 3: Leelarathna L, Nodale M, Allen JM, et al. Evaluating the accuracy and large inaccuracy of two continuous glucose monitoring systems. Diabetes Technol Ther 2013; 15: This study provides a method of assessing the suitability and safety of continuous glucose monitor in a closed-loop system by evaluating the frequency and duration of large sensor inaccuracy of two continuous glucose monitoring devices. 26. Keenan DB, Mastrototaro JJ, Voskanyan G, Steil GM. Delays in minimally invasive continuous glucose monitoring devices: a review of current technology. J Diabetes Sci Technol 2009; 3: Van Bon AC, Luijf YM, Koebrugge R, et al. Feasibility of a portable bihormonal closed-loop system to control glucose excursions at home under free-living conditions for 48 hours. Diabetes Technol Ther 2013; doi: /dia [Epub ahead of print] The first study to assess the feasibility of a bihormonal fully closed-loop system (no prandial boluses) in a home setting. 28. Buckingham B, Wilson DM, Lecher T, et al. Duration of nocturnal hypoglycemia before seizures. Diabetes Care 2008; 31: Hovorka R, Allen JM, Elleri D, et al. Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet 2010; 375: Nimri R, Danne T, Kordonouri O, et al. The Glucositter overnight automated closed loop system for type 1 diabetes: a randomized crossover trial. Pediatr Diabetes 2013; 14: Dauber A, Corcia L, Safer J, et al. Closed-loop insulin therapy improves glycemic control in children aged <7 years: a randomized controlled trial. Diabetes Care 2013; 36: The use of closed loop in a very young paediatric cohort was demonstrated to be well tolerated and was able to decrease the risk of nocturnal hypoglycaemia. 32. Lane JE, Shivers JP, Zisser H. Continuous glucose monitors: current status and future developments. Curr Opin Endocrinol Diabetes Obes 2013; 20: Manohar C, Levine JA, Nandy DK, et al. The effect of walking on postprandial glycemic excursion in patients with type 1 diabetes and healthy people. Diabetes Care 2012; 35: Sherr JL, Cengiz E, Palerm CC, et al. Reduced hypoglycemia and increased time in target using closed-loop insulin delivery during nights with or without antecedent afternoon exercise in type 1 diabetes. Diabetes Care 2013; 36: This study highlights the benefit of closed loop in reducing the risk of nocturnal hypoglycaemia, even following antecedent exercise during the day. 35. Elleri D, Allen JM, Kumareswaran K, et al. Closed-loop basal insulin delivery over 36 hours in adolescents with type 1 diabetes: randomized clinical trial. Diabetes Care 2013; 36: The challenges of day and night closed-loop insulin delivery, especially during exercise and meal periods, are highlighted in this study. 36. Ruiz JL, Sherr JL, Cengiz E, et al. Effect of insulin feedback on closed-loop glucose control: a crossover study. J Diabetes Sci Technol 2012; 6: Breton M, Farret A, Bruttomesso D, et al. Fully integrated artificial pancreas in type 1 diabetes: modular closed-loop glucose control maintains near normoglycemia. Diabetes 2012; 61: Data from an international, multicentre, randomized, crossover, clinical research facility-based study is presented in this study. 38. Bakhtiani PA, Zhao LM, El Youssef J, et al. A review of artificial pancreas technologies with an emphasis on bi-hormonal therapy. Diabetes Obes Metab 2013; 15: Russell SJ, El-Khatib FH, Nathan DM, et al. Blood glucose control in type 1 diabetes with a bihormonal bionic endocrine pancreas. Diabetes Care 2012; 35: Bihormonal closed-loop system was shown to reduce hypoglycaemia in a 48-h clinical research facility study, despite the challenges related to meals and exercise. 40. Steiner SS, Li M, Hauser R, Pohl R. Stabilized glucagon formulation for bihormonal pump use. J Diabetes Sci Technol 2010; 4: Jackson MA, Caputo N, Castle JR, et al. Stable liquid glucagon formulations for rescue treatment and bi-hormonal closed-loop pancreas. Curr Diab Rep 2012; 12: Garg SK, Voelmle MK, Beatson CR, et al. Use of continuous glucose monitoring in subjects with type 1 diabetes on multiple daily injections versus continuous subcutaneous insulin infusion therapy: a prospective 6-month study. Diabetes Care 2011; 34: Buckingham B, Chase HP, Dassau E, et al. Prevention of nocturnal hypoglycemia using predictive alarm algorithms and insulin pump suspension. Diabetes Care 2010; 33: Deiss D, Hartmann R, Hoeffe J, Kordonouri O. Assessment of glycemic control by continuous glucose monitoring system in 50 children with type 1 diabetes starting on insulin pump therapy. Pediatr Diabetes 2004; 5: Volume 21 Number 2 April 2014

7 Bringing closed-loop home Thabit and Hovorka 45. Monaghan MC, Hilliard ME, Cogen FR, Streisand R. Nighttime caregiving behaviors among parents of young children with type 1 diabetes: associations with illness characteristics and parent functioning. Fam Syst Health 2009; 27: Nimri R, Muller I, Atlas E, et al. Night glucose control with MD-Logic artificial pancreas in home setting: a single blind, randomized crossover trial-interim analysis. Pediatr Diabetes doi: /pedi [Epub ahead of print] This study was an interim analysis of a short-term overnight closed-loop home study. It reports an improvement in the time spent with hypoglycaemia overnight compared with sensor-augmented pump therapy. 47. Hovorka R, Nodale M, Haidar A, Wilinska ME. Assessing performance of closed-loop insulin delivery systems by continuous glucose monitoring: drawbacks and way forward. Diabetes Technol Ther 2013; 15: Hovorka R, Elleri D, Thabit H, et al. Benefits of unsupervised home use of overnight closed loop insulin delivery in youth. Diabetes Care. doi: / DC This is the first randomized crossover study to demonstrate the safety and efficacy of unsupervised overnight closed-loop use at home for 3 weeks, with reduction in nights with hypoglycaemia X ß 2014 Wolters Kluwer Health Lippincott Williams Wilkins 101