during CSII, has received comparatively little attention in ... 1Department of Endocrinology and Diabetes, Princess. Margaret Hospital for ... Trust, London, UK.
DIABETES TECHNOLOGY & THERAPEUTICS Volume 17, Supplement 1, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/dia.2015.1503
ORIGINAL ARTICLE
Insulin Pumps John Pickup
Introduction
T
hree of the four themes concerning insulin pump therapy that we focus on this year have been the subject of ongoing research in previous yearbook reports: lowglucose insulin-suspend (LGS) pumps, long-term continuous subcutaneous insulin infusion (CSII), and CSII in type 2 diabetes. This year there have been notable advances in all three areas. Robust evidence for the efficacy of LGS in reducing severe hypoglycemia sets the scene for reimbursement, and the realization that excellent control on CSII is not maintained in all patients in the long-term will restimulate research on pump education throughout the treatment on CSII and why some patients resist these measures. Insulin pump therapy in poorly controlled type 2 diabetes is set to become one of the major research themes of the next years, with the publication of a major randomized controlled trial (RCT) of CSII versus multiple daily insulin injections (MDI). The fourth theme, adverse events and complications during CSII, has received comparatively little attention in recent decades. There is increasing note being given to the performance and safety of medical devices such as insulin pumps, not only at the point of market entry but also during their ongoing use. Reports that the reliability of pumps and CSII components seems to be not much better than during the early days of CSII will therefore be of great interest to many parties. LOW-GLUCOSE INSULIN-SUSPEND PUMPS Effect of sensor-augmented insulin pump therapy and automated insulin suspension vs standard insulin pump therapy on hypoglycemia in patients with type 1 diabetes. A randomized controlled trial Ly TT 1–3, Nicholas JA1,2, Retterath A 2, Lim EM 4, Davis EA1–3, Jones TW 1–3
1
Department of Endocrinology and Diabetes, Princess Margaret Hospital for Children, Perth, Australia; 2Telethon Institute for Child Health Research and 3School of Paedia-
trics and Child Health, University of Western Australia, Perth, Australia; and 4PathWest Laboratory Medicine, Queen Elizabeth II Medical Centre, Nedlands, Western Australia, Australia JAMA 2013; 310: 1240–47
This manuscript is also discussed in article on Continuous Glucose Monitoring in 2014, p. S-15. Background LGS insulin pump therapy has the potential to reduce hypoglycemia in type 1 diabetes, but its effect in patients with moderate and severe hypoglycemia lacks trial evidence. The aim of this study was to compare hypoglycemic outcomes with LGS therapy versus standard CSII in patients at risk of significant hypoglycemia. Methods Type 1 diabetes patients (n = 95, mean age 18.6 years) with hypoglycemia unawareness were randomized to standard CSII or CSII with LGS (Medtronic Paradigm Veo) for 6 months. The LGS threshold was 60 mg/dL (3.3 mmol/L). Results The combined severe (hypoglycemia causing seizure or coma) and moderate (hypoglycemic event requiring thirdparty assistance) hypoglycemia rate was 34.2 for the CSII group and 9.5 events/100 patient-months for the LGS group, with a rate ratio of 3.6 (95% confidence interval [CI]: 1.7–7.5, p < 0.001). The severe hypoglycemia rate ratio was 1.5 (0.3– 2.7, p = 0.02). HbA1c did not differ in either group (7.4% vs. 7.5% at 6 months [57 vs. 59 mmol/mol]; change - 0.06% vs. - 0.1%, CSII vs. LGS). Counterregulatory hormone responses to hypoglycemia did not change, and hypoglycemia awareness did not differ between groups at study completion. Conclusions LGS insulin pump therapy reduces the rate of moderate and severe hypoglycemia in type 1 diabetes.
Diabetes Research Group, King’s College London School of Medicine, Guy’s Hospital, London, UK
S-21
S-22 Real-time continuous glucose monitoring significantly reduces severe hypoglycemia in hypoglycemia-unaware patients with type 1 diabetes Choudhary P 1,2, Ramasamy S 3, Green L 2, Gallen G 2, Pender S 3, Brackenridge A 2, Amiel SA1,2, Pickup JC 1,2 1
Division of Diabetes and Nutritional Sciences, King’s College London, London, UK; 2King’s College Hospital, London, UK; and 3Guy’s and St. Thomas’s Hospitals NHS Trust, London, UK
PICKUP Background LGS pumps to date have been based on basal rate suspension of CSII at a preset threshold of CGM-measured hypoglycemia, but triggering insulin suspension when the sensor glucose is predicted to reach the hypoglycemic range could be even more effective at preventing hypoglycemia. This study investigated potential benefits of predictive LGS versus threshold LGS using computer simulation (in silico), in addition to a feasibility clinical trial in children and adolescents.
Diabetes Care 2013; 36: 4160–62
Methods
This manuscript is also discussed in article on Continuous Glucose Monitoring in 2014, p. S-14 and article on Diabetes Technology and the Human Factor, p. S-112.
In an in silico study using the University of Virginia glucose simulator, 100 virtual patients received simulations of no pump suspension, predictive LGS (30 min glucose prediction), or threshold LGS after an insulin bolus designed to induce hypoglycemia. A predictive LGS system based on the Medtronic Veo pump and Enlite sensor was then tested in a clinical trial on 22 type 1 diabetic subjects who exercised until the system suspended or the reference blood glucose reached the predicted value.
Background CGM used with CSII or MDI may benefit people with type 1 diabetes who are at risk of severe hypoglycemia, but trials to date have not shown convincing evidence of a reduction in severe hypoglycemia with this technology. The aim of the study was to audit changes in severe hypoglycemia in patients with problematic hypoglycemia in spite of best attempts with intensified insulin regimens. Methods Type 1 diabetes patients with hypoglycemia unawareness and problematic hypoglycemia (n = 35, 33 on CSII) who had failed to improve with structured diabetes education were switched to CGM and studied after 12 months of treatment; 23 used an LGS pump (Medtronic Veo). Results The median (interquartile range [IQR]) severe hypoglycemia rate was reduced from a baseline of 4.0 (0.75–7.25) to 0 (0–0) episodes/patient-year after 12 months in the LGS group, and to 0 [0–2.0] episodes/patient-year in the nonLGS group on CGM. The final HbA1c in the LGS group was reduced to 7.7% (61 mmol/mol), with a baseline of 8.1% (65 mmol/mol). The mean Gold score for hypoglycemia awareness was not changed at study completion, though 54% of subjects reported subjective improvement in awareness. Conclusions In selected patients with type 1 diabetes, LGS pump therapy reduces severe hypoglycemia while improving HbA1c. The PILGRIM Study: in silico modeling of a predictive low-glucose management system and feasibility in youth with type 1 diabetes during exercise Danne T1, Tsioli C 1, Kordonouri O 1, Blaesig S1, Remus K 1, Roy A 2, Keenan B 2, Lee SW 2, Kaufman FR2 1
Children’s Hospital auf der bult, Hannover, Germany; and Medtronic Inc., Northridge, CA
2
Diabet Technol Ther 2014; 16: 1–10
This manuscript is also discussed in article on Diabetes Technology and Therapy in the Pediatric Age Group, p. S-100.
Results The average duration of hypoglycemia in the computer simulation study was less with predictive LGS than the other systems: 65.7 vs. 93.7 vs. 130.4 min (predictive vs. threshold LGS vs. no suspension). In the clinical study, hypoglycemia was prevented in 80% of occasions with the predictive LGS pump when hypoglycemia occurred after exercise. The mean duration of suspension was 90 min, the sensor glucose at predictive suspension was 92 mg/dL (5.1 mmol/L), and the glucose nadir was 77 mg/dL (4.3 mmol/L). Conclusions Predictive LGS technology may further reduce the severity of hypoglycemia compared to threshold LGS pumps. Comment Although LGS pumps have been in use for about 5 years in many centers in Europe and have found special use in clinical practice for reducing severe hypoglycemia in type 1 diabetes, randomized trial evidence to support this indication has been lacking. Ly et al. present the first randomized controlled trial of LGS versus standard CSII in subjects at special risk of hypoglycemia. The subjects were young and the results still need to be confirmed in adults with long-standing diabetes. The report of Choudhary et al. did study adult subjects, though it was observational in design; however, this study supports the notion that LGS insulin pump therapy reduces severe hypoglycemia in those who have failed other measures, notably CSII and/or MDI in combination with structured diabetes education. Cost-effectiveness analysis can now be performed with the intention of deciding evidence-based reimbursement policies for LGS pumps in different countries. Interestingly, hypoglycemia awareness did not improve in parallel with the reduction in hypoglycemia in
INSULIN PUMPS
either the Ly et al. or Choudhary et al. studies (though there were individuals in whom awareness was improved). Possibly, more rigorous avoidance of hypoglycemia is necessary for awareness to be restored in a greater proportion of subjects using LGS. The next development of control-to-range pumps with automatic basal rate suspend is likely to be the commercialization and clinical use of predictive LGS insulin pumps. Danne et al. provide the first evidence (although using an in silico study) that the reduction in hypoglycemia with this technology is likely to be somewhat better than for threshold LGS. We look forward to seeing RCTs comparing threshold and predictive LGS pumps with standard CSII in patients who are likely to benefit from the technology—those with disabling hypoglycemia. LONG-TERM CSII Cross-sectional survey and retrospective analysis of a large cohort of adults with type 1 diabetes with long-term continuous subcutaneous insulin infusion treatment Joubert M 1, Morera J 1,2, Vicente A 1, Rod A 1, Perienti J-J 2,3, Reznik Y1,2 1
Endocrinology Department, Caen University Hospital; University of Caen; and 3Research and Biostatistics Department, Caen University Hospital, Caen, France
2
J Diabet Sci Technol 2014; 8: 1005–10
Background There is relatively little published information on the longterm clinical benefit of CSII and associated quality of life and treatment satisfaction. This study is a retrospective and crosssectional audit of adult patients receiving CSII for more than 1 year at a center in France. Methods A survey and case-note analysis was performed on 295 patients using long-term CSII; the median duration of CSII was 5 (IQR 3–8) years.
S-23 Variations in the quality and sustainability of long-term glycemic control with continuous subcutaneous insulin infusion Nixon R, Folwell R, Pickup JC Diabetes Research Group, King’s College London School of Medicine, Guy’s Hospital, London, UK Diabetic Med 2014; 31: 1174–77
Background Previous studies on long-term CSII are difficult to interpret because some subjects with a relatively normal HbA1c at baseline are included (probably because the indication for pump treatment was hypoglycemia), and mean responses in HbA1c disguise individuals or groups with differing patterns of response. This is a study of 5-year changes in HbA1c and subgroups of responses in people with poorly controlled type 1 diabetes on MDI at baseline. Methods Adults with type 1 diabetes and with elevated HbA1c (‡8.0% [64 mmol/mol], n = 35) at baseline on MDI were studied using clinic records, changes in HbA1c were recorded during at least 5 years of CSII, and mean yearly HbA1c was calculated. Using a significant change in HbA1c from 1 year to the next as – 0.5% (5 mmol/mol) when the change occurred from a value of ‡7.5% (58 mmol/mol), three patterns of HbA1c were identified. Results Improved HbA1c was maintained by 88% of the subjects during 5 years of CSII. However, three subgroups were identified, all with a similar baseline HbA1c of 9.2–9.3% (77–78 mmol/mol) on MDI. Group A (57%) had improvement for 1–2 years after starting CSII, followed by a deterioration; group B (31%) had sustained improvement over 5 years of CSII; and group C (12%) had no significant improvement in HbA1c from baseline at any year of pump treatment. The groups did not differ by age, sex, duration of diabetes, or fear of hypoglycemia score, but group C patients had a higher mean body mass index (BMI) than the other two groups: 31.0 vs. 25.9 vs. 25.2 kg/m2 (C vs. A vs. B, p = 0.02). Conclusions
Results Overall satisfaction with CSII was high for 93% of subjects, and 88% would not want it removed. Discomfort scores were low, though somewhat higher for sport and sexual activity. Despite a high level of diabetes education, only 24% used the bolus calculator and 43% the temporary basal rates. Safety was good, with a hospitalization rate of 0.18 patients/ year and scarce cannula infections. HbA1c fell from a baseline of 8.2% (66 mmol/mol) by 0.5% (5 mmol/mol) at years 1–3, and remained fairly stable at years 4–6 and beyond. Conclusions Safety, satisfaction, and efficacy of CSII are maintained at long-term follow-up, but many patients do not use the advanced features of the pump.
A high percentage of people enjoy sustained improvement in control over many years of CSII that is better at all times than on MDI, but there are variations in efficacy, with some patients improving and then deteriorating, some maintaining the level of good control over the years, and some not improving at any point. Comment Both these studies show that most people receiving longterm CSII maintain control that is significantly better than MDI. It is not clear why the HbA1c in some people on CSII in the study of Nixon et al. improves and then deteriorates; possibly, some fail to implement therapeutic and lifestyle changes recommended during renewed
S-24
CSII education at the time of first worsening, or they have a psychological profile that is not consistent with the commitment required for successful long-term CSII. Nonresponders did not have an excess fear of hypoglycemia that might prevent tightening of control, but the higher BMI may indicate insulin resistance and/or dietary noncompliance as contributory factors. In any case, practitioners should be alert that control may start to worsen on CSII after about 2 years but, with renewed education and therapeutic advice, optimal control can be restored or deterioration minimized in the majority. Of interest in the study of Joubert et al. is that only about one-third of the subjects used the advanced features of insulin pumps such as bolus calculators, indicating that in many cases there may be scope for further improvement in control on CSII by taking advantage of all the pump features.
PICKUP talization or emergency department admission was reported by 8%, mostly due to set and site problems and pump malfunctions associated with hyperglycemia/ketosis; there was no hypoglycemia as a cause of hospitalization. Regarding family confidence in the CSII system, 87% were confident or very confident in the reliability of their pump and 71% in the reliability of the infusion sets. Conclusions There is a high self-reported rate of adverse events, even with modern CSII, but a notably high proportion of families still have confidence in CSII. Nonmetabolic complications of continuous subcutaneous insulin infusion: a patient survey Pickup JC 1, Yemane N 2, Brackenridge A2, Pender S 2 1
Diabetes Research Group, King’s College London School of Medicine, Guy’s Hospital, London, UK; and 2Guy’s and St. Thomas’s Hospitals NHS Foundation Trust, London, UK
COMPLICATIONS AND ADVERSE EVENTS ASSOCIATED WITH CSII
Diabet Technol Ther 2014; 16: 145–49
Family perceptions of insulin pump adverse events in children and adolescents
Although much is known about potential clinical complications of CSII such as hypoglycemia or ketoacidosis, little is known about the frequency of nonmetabolic technical problems of contemporary insulin pump therapy when studied by standardized questionnaire.
Wheeler BJ 1,2, Donaghue KC 1,3, Heels K 1, Ambler GR1,3 1
Institute of Endocrinology and Diabetes, Children’s Hospital at Westmead, Sydney, Australia; 2Department of Women’s and Children’s Health, University of Otago, Dunedin, New Zealand; and 3Discipline of Paediatrics and Child Health, University of Sydney, Sydney, Australia
Diabet Technol Ther 2014; 16: 204–7
Background CSII is widely used in children and adults but, in contrast to clinical outcomes such as HbA1c and hypoglycemia, there has been little reported recently on adverse events. The aim of this study therefore was to investigate family perceptions of adverse events during insulin pump therapy. Methods All families of people with type 1 diabetes £ 19 years of age who were receiving CSII at a diabetes clinic were approached to complete a survey of retrospective perceptions of pump-related adverse events occurring over the previous 12 months, as well as information pertaining to CSII-related confidence and education.
Background
Methods Adult patients with type 1 diabetes treated by CSII for ‡ 6 months (median CSII duration 3.3 [range 0.5–32] years) completed a self-report questionnaire on the frequency and nature of pump malfunctions, infusion set and site problems, set duration, pump insulin use, and patient-related issues such as weight change and coping, and other psychological issues. Results The commonest infusion set problems were kinking (64% of subjects) or blockage (54%), with leaking reported in 16%. Blockage was associated with >3 days set use plus lispro insulin use in the pump. Lipohypertrophy occurred at the infusion site in 26% and was more common with long-duration CSII. Infection at the infusion site occurred in 17%. Pump malfunction had occurred in 48% of subjects (mostly in the first year of pump use). There was no consistent pattern of weight change associated with CSII. Conclusions Pump, infusion set, and infusion site problems are still common with CSII, in spite of modern technology.
Results Responses were obtained from 235 participants (99%), with a mean age of 12.3 years and duration of CSII 2.6 years. Pump malfunction occurred in 56%; 23% required pump replacement by the manufacturer; and infusion set or site failure occurred in 45%. An adverse event requiring hospi-
Comment These two studies on the adverse events reported with insulin pump therapy agree that about one half of CSII patients have a pump malfunction in the first year of
INSULIN PUMPS
treatment. Pumps had to be replaced by the manufacturer in 23% of subjects in the study of Wheeler et al. These proportions are disturbingly high because an early survey of CSII in 1986 (1) reported that 25% of patients had a pump breakdown in the first year. Pump reliability then does not seem to have improved with the increased sophistication of technology in the last 28 years, and indeed there are more opportunities for malfunction with the intricacies of modern pump software, electronic keyboard controls, computer download functions, and so on. There is an increasing demand for regulated postmarket surveillance of efficacy and safety of any new pumps that are introduced. It is notable that in both surveys the frequency of infusion set complications was high—about 45–65% of the subjects experienced some problems—with blockage, kinking, and leakage being the main issues. Insulin infusion sets are considered to be one of the main weak points of the CSII system (2), and much could be done to improve this component of the technology.
CSII IN TYPE 2 DIABETES Insulin pump treatment compared with multiple daily insulin injections for the treatment of type 2 diabetes (OpT2mise): a randomized open-label controlled trial Reznik Y1, Cohen O 2, Aronson R 3, Conget I 4, Runzis S5, Castaneda J 6, Lee SW 7 for the OpT2mise Study Group 1
Department of Endocrinology, University of Caen, Coˆte de Nacre Regional Hospital Center, Caen, France; 2Institute of Endocrinology, Chaim Sheba Medical Center, Tel Hashomer, Israel; 3LMC Diabetes & Endocrinology, Toronto, ON, Canada; 4Diabetes Unit, Endocrinology and Nutrition Department, University Hospital Clinic, Barcelona, Spain; 5Medtronic International Trading Sa`rl, Tolochenaz, Switzerland; 6Medtronic Bakken Research Center, Maastricht, The Netherlands; and 7Medtronic Diabetes, Northridge, CA Lancet 2014; 384(9950): 1265–72
Background Previous RCTs comparing CSII versus MDI in type 2 diabetes have produced inconsistent results. The aim of this study was to perform a large-scale RCT of these treatments in type 2 diabetic patients but targeted at those who were still poorly controlled after a period on MDI with active insulin titration to try and optimize glycemic control. Methods With referral centers in Canada, Europe, Israel, South Africa, and the United States, patients with type 2 diabetes underwent a 2-month run-in period of intensified MDI with a standardized titration protocol. Those who were still poorly controlled (HbA1c 8–12% [64–108 mmol/mol], insulin dose
S-25 0.7–1.8 units/kg, n = 331) were randomized to insulin pump therapy or continued MDI for 6 months. Results The baseline HbA1c was 9% (75 mmol/mol) in both treatment groups. At 6 months, the mean HbA1c had declined by 1.1% (12 mmol/mol) in the CSII group and 0.4% (4 mmol/ mol) in the MDI group, with a between-group difference of 0.7 (95% CI: 0.9–0.4)% or 8 (10–4) mmol/mol, p < 0.0001. The between-group difference in HbA1c for those with the highest tertile of HbA1c at baseline (9.3–11.5%, 78–102 mmol/mol) was 1.1% (12 mmol/mol). The mean total daily insulin dose was 20% less in the CSII group at study completion, but there was no difference in weight gain or hypoglycemia frequency. Conclusions CSII can be considered a safe and valuable treatment for people with type 2 diabetes who remain poorly controlled on MDI. Comment Many national guidelines do not recommend CSII in type 2 diabetes and reimbursement is therefore limited. This is because of a lack of convincing evidence of efficacy of CSII versus MDI in controlled trials in type 2 diabetes to date. However, several observational studies in recent years have indicated that patients with an elevated HbA1c (and usually insulin resistance) on MDI can enjoy a substantial improvement in control when switched to CSII (3,4). This study by Reznik et al. rightly then targets treatment at those with poor control and/or insulin resistance, in spite of best attempts with MDI. Note that the between-group difference in HbA1c is even greater for those worst controlled at baseline—1.1% (12 mmol/mol). Formal cost-effectiveness studies will now be needed to decide whether CSII is an affordable treatment for type 2 diabetes for various health services and for which patient group it is most suited. This analysis will take into account not only the reduction in HbA1c but also the notable reduction and cost savings in insulin requirements. This study was performed with insulin pumps originally intended for type 1 diabetes, and it is possible that smaller and cheaper ‘‘patch’’ pumps with simpler insulin regimens will prove to be more suitable for type 2 diabetes (5). Cost-effectiveness might be further improved with these pumps. Author Disclosure Statement J.P. has received speaker and/or consultancy fees from Cellnovo, Cequr, Medtronic, and Roche, manufacturers of insulin pumps. References 1. Mecklenburg RS, Guinn TS, Sannar CA, Blumenstein BA. Malfunction of continuous subcutaneous insulin infusion
S-26 systems: a one-year prospective study of 127 patients. Diabetes Care 1986; 9: 351–55. 2. Heinemann L, Walsh J, Roberts R. We need more research and better designs for insulin infusion sets. J Diabet Sci Technol 2014; 8: 199–202. 3. Edelman SV, Bode BW, Bailey TS, Kipnes MS, Brunelle R, Chan X, Frias JP. Insulin pump therapy in patients with type 2 diabetes. Safely improved glycemic control using a simple
PICKUP insulin dosing regimen. Diabet Technol Therapeut 2010; 12: 627–33. 4. Leinung MC, Thompson S, Luo M, Leykina L, Nardacci E. Use of insulin pump therapy in patients with type 2 diabetes after failure of multiple daily injections. Endocr Pract 2013; 19: 9–13. 5. Pickup JC. Insulin pump therapy for type 2 diabetes mellitus. Nat Rev Endo 2014; 10: 647–49.
