Abstract
Introduction
E
Long-Term Insulin Pump Therapy
Insulin pump—long-term effects of glycemic control: an observational study at 10 diabetes clinics in Sweden
Carlsson B-M1,2, Attvall S2,3, Clements M4, Gumpney SR5, Pividic A6, Sternemalm L3, Lind M2,7
1Department of Medicine, SAS-Hospital Organization, Skene, Sweden; 2Insitute of Medicine, University of Gothenburg, Gothenburg, Sweden; 3Diabetes Section, Sahlgrenska University Hospital, Gothenburg, Sweden; 4Section of Endocrinology, Children's Mercy Hospitals & Clinics, Kansas City, MO; 5Endocrine and Diabetes Centre, Visakhapatnam, India; 6Statistika Konsultgruppen, Gothernburg, Sweden; and 7Department of Medicine, NU-Hospital Organization, Trollhatan, Sweden
Diabetes Technol Ther 2013;
Background
There is comparatively little information on the glycemic effectiveness of CSII over periods of longer than 1–2 years, particularly when combining results from more than one center, with their differing uptake of pump therapy. The aim of this study was to retrospectively survey the glycemic control achieved in long-term pump patients versus multiple daily injection (MDI)–treated subjects.
Methods
Adult type 1 diabetic subjects (n=272) from 10 centers in Sweden who had been treated by CSII for at least 5.5 years were studied. A group of patients treated by MDI (n=2437) over the same period acted as controls.
Results
At baseline, the pump patients were younger, were more often female, had shorter duration of diabetes than MDI subjects, and had worse control (HbA1c 8.4% vs. 8.1%, p<0.001). HbA1c improved for both groups at 5 years, but more so for CSII (unadjusted difference from baseline: 0.44% vs. 0.11%, CSII vs. MDI). Adjusting for differences in baseline HbA1c, there was a significant difference in change from baseline, favoring CSII over the years, but this effect lessened as duration of CSII increased: at 1 year, −0.42%; 2 years, −0.43%; and 5 years, −0.20%.
Conclusions
CSII achieves significantly lower HbA1c levels than MDI, even in the long-term, but the effect diminishes with time.
A 7-year follow-up, retrospective, international, multicenter study of insulin pump therapy in children and adolescents with type 1 diabetes
Mameli C1, Scaramuzza AE1, Ho J2, Cardona-Hernandez R3, Suarez-Ortega L3, Zucotti GV1
1Department of Pediatrics, University of Milan, Milan, Italy; 2Department of Pediatrics, University of Calgary, Calgary, Canada; and 3Department of Endocrinology and Diabetes, Hospital Saint Joan de Deu, Barcelona, Spain
Acta Diabetol 17 May 2013: [Epub ahead of print]; DOI: 10.1007/s00592-013-0481-y
Background
Studies of CSII lasting longer than about 4 years are limited, and the aim of this study was therefore to retrospectively analyze data from a large cohort of pediatric subjects with type 1 diabetes treated by CSII for 7 years at three centers in Canada, Italy, and Spain.
Methods
Demographic, clinical, pump usage, and diabetes-control data from type 1 diabetic subjects treated by CSII were collected by chart review (n=121; age 5–20 years; mean follow-up 6.9 [range 5–12 years]).
Results
HbA1c had significantly improved at 1 year, but showed a trend to worsen slightly during subsequent follow-up. Although the baseline HbA1c was similar for males and females, only the males showed significant improvement over the 7 years. Severe hypoglycemia (SH) decreased during follow-up, though not significantly (8.2 vs. 1.1 events/100 patient-years). A subgroup of patients who used advanced pump features (bolus calculator, different bolus profiles, and temporary basal rates) had a better HbA1c at the end of follow-up (7.8% vs. 8.5%, p=0.003).
Conclusions
CSII was safe and effective in the long-term for this group of pediatric diabetic patients, but the main benefit was seen in males, and those who use pump features achieved the best control.
Long-term metabolic effects of continuous subcutaneous insulin infusion therapy in type 1 diabetes
Cohen ND1, Hong ES2, Van Drie C3, Balkau B1,4,5, Shaw J1
1Baker Heart and Diabetes Institute, Melbourne, Australia; 2Cheongju St. Mary's Hospital, Cheongju, Korea; 3University of Amsterdam, Amsterdam, The Netherlands; 4INSERM, U1018, Villejuif, France; and 5University of Paris Sud, Villejuif, France
Diabetes Technol Ther 2013;
Background
The aim of this study was to determine the long-term effects of CSII in comparison with MDI, where both groups had undergone specialized education programs.
Methods
Records were studied from adults with type 1 diabetes who commenced CSII (n=126) or an intensive MDI program (n=121) between 2000 and 2011. MDI patients attended a Dose Adjustment for Normal Eating (DAFNE)–style education program. Follow-up was for 48 months in the pump patients and 39 months in the MDI patients (ns). HbA1c was similar at baseline (8.0% vs. 7.8%).
Results
In the pump patients, HbA1c was significantly lower than at baseline at every time point up to 3 years. The peak reduction was at 6 months (−0.64% for CSII, −0.15% for MDI), but the reduction declined over time. The difference in HbA1c from baseline for CSII versus MDI was significant to 2 years. According to blood glucose meter data, the standard deviation (SD) of glucose was decreased at 6 months in the CSII group but not the MDI group.
Conclusions
The HbA1c improvement seen with CSII versus MDI declines over time.
All three of these studies confirm that CSII achieves good glycemic control over many years, but the best effect of pumps on the mean HbA1c level is seen after 6 months to 1 year from pump start and then the improvement gradually declines, for both adult and pediatric/adolescent groups. We have comparatively little information still on how the long-term efficacy of pumps varies from one person to another, though Mameli et al. provide evidence that young females do worse. Perhaps this is related to the insulin resistance of puberty (mean age of the cohort was 13.5 years) or other psychosocial problems associated with adolescence. Others have previously shown that “brittle diabetes” associated with poor glycemic control and multiple admissions to the hospital in adolescent females is not effectively managed by CSII (1), and that teenage diabetic patients are those most likely to have poor control on pumps and to discontinue this therapy (2).
Mameli et al. also report that those who tended to use advanced features of modern pumps such as different bolus profiles for different meal types, the bolus calculator, and temporary basal rates achieved the best control. This is interesting because, although the logic and potential advantages of these features are quite apparent, trial evidence to support their use in clinical practice is still very limited.
The reasons why control deteriorates in some patients on long-term CSII are not known—loss of motivation and psychological issues, illness, increasing insulin resistance, and dietary factors might all play a part. Many clinics find that re-education in pump procedures, together with remotivation, is helpful in restoring good control in many patients whose control has been worsening after a period on CSII, though the number who respond to these measures has not been well documented.
It is interesting to note that the MDI patients in the study of Cohen et al. all underwent a DAFNE-style structured education program. There is ongoing debate about whether the improvement in control associated with CSII is attributable to the technology or to the education and attention linked to pump programs. This study indicates that, at least for some subjects, additional benefit can be obtained from pump therapy.
Sensor-Augmented Insulin Pump (SAP) Therapy
Threshold-based insulin-pump therapy interruption for reduction of hypoglycemia
Bergenstal RM1, Klonoff DC2, Garg SK4, Bode BW5, Meredith M6, Slover RH4, Ahmann AJ7, Welsh JB3, Lee SW3, Kaufman FR3
1International Diabetes Center, Park Nicollet, Minneapolis, MN; 2Diabetes Research Institute, Mills-Peninsula Health Services, San Mateo, CA; 3Barbara Davis Center for Childhood Diabetes, University of Colorado, Denver, CO; 4Atlanta Diabetes Associates, Atlanta, GA; and 5Department of Medicine, University of Wisconsin, Madison, WI; 6Harold Schnitzer Diabetes Health Center, Oregon Health and Science University, Portland, OR; 7Medtronic, Northridge, CA
N Engl J Med 2013;
Background
Hypoglycemia remains a major problem in type 1 diabetes, even during insulin pump therapy. SAP therapy with automatic suspend of the basal rate for up to 2 hours when a preset hypoglycemic threshold is reached (LGS) has been available commercially in Europe since 2009, but trials have been of short duration and have involved small numbers of subjects. The aim of this study—the ASPIRE In-Home Study—was to investigate the safety and compare the effect on hypoglycemia of SAP+LGS with SAP alone in type 1 diabetic patients prone to nocturnal hypoglycemia.
Methods
Type 1 diabetic subjects (aged 16–70 years) who had documented nocturnal hypoglycemia and used a glucose sensor for at least 80% of the time during a run-in phase (n=247) were randomly allocated to SAP+LGS or to SAP alone for 3 months. The threshold was initially set at 3.9 mmol/L (70 mg/dL). For the test group, LGS was used from 10 pm to 8 am but was optional at other times.
Results
HbA1c did not change from baseline to study end in either group. The area under the curve for nocturnal hypoglycemia was 37.5% less in the SAP+LGS groups versus the SAP-alone groups (p<0.001). The mean sensor glucose at the start of nocturnal hypoglycemia was 3.5 mmol/L (62.6 mg/dL), and 9.0 mmol/L (162.3 mg/dL) 4 hours later. The median duration of nocturnal suspends was 11.9 minutes; 19.6% were for 2 hours (mean glucose 4 hours after suspend 9.4 mmol/L [168.8 mg/dL]). Scores for the Hypoglycemia Fear Survey did not differ between groups. No episodes of diabetic ketoacidosis (DKA) occurred. Four episodes of SH occurred in the SAP-alone group, none in the SAP+LGS group.
Conclusions
SAP with an LGS feature is safe to use in the home setting and reduced nocturnal hypoglycemia without increasing HbA1c.
Routine sensor-augmented pump therapy in type 1 diabetes: the INTERPRET Study
Nørgaard K1, Scaramuzza A2, Bratina N3, Lalić NM4, Jarosz-Chabot P5, Kocsis G6, Jasinskiene E7, DeBlock C8, Carrette O9, Castaneda J10, Cohen O11
1Hvidovre University Hospital, Hvidovre, Denmark; 2Luigi Sacco Hospital, Milan, Italy; 3University Medical Center and Children's Hospital, Ljubljana, Slovenia; 4Clinic for Endocrinology, University of Belgrade Faculty of Medicine, Belgrade, Serbia; 5Medical University of Silesia, Katowice, Poland; and 6Péterfy Hospital, Budapest, Hungary; 7Lithuanian University Hospital of Health Sciences, Kaunas, Lithuania; 8Antwerp University Hospital, Antwerp, Belgium; 9Medtronic International Trading, Tolochenaz, Switzerland; 10Medtronic Bakken Research Center, Maastrict, the Netherlands; 11Chaim Sheba Medical Center, Tel Hashomer, Israel
Diabetes Technol Ther 2013;
Background
Although the efficacy of SAP in reducing HbA1c without increasing hypoglycemia has been demonstrated in recent RCTs, there is a lack of information about the real-life value of SAP when used in the long-term. This multicenter, observational study aimed to document usage and safety of SAP and which variables are associated with improvement in HbA1c.
Methods
Data from 263 type 1 diabetes patients who switched from CSII alone to SAP at 27 centers in 15 countries in Europe and Israel were analyzed over a 12-month period. Subjects were children, adolescents, and adults.
Results
The main indications for SAP were glycemic instability (38%), high HbA1c (30%), recurrent hypoglycemia (8%), and hypoglycemia unawareness (8%). Mean sensor usage (30%) decreased over time (37–27%). Factors associated with improvement in HbA1c were baseline HbA1c, older age, and more frequent sensor use. Diabetes-related hospital admissions and length of stay were less on SAP than for the 12 months before SAP start. There was no significant change in the frequency of SH, but patients reported reduced fear of hypoglycemia and increased treatment satisfaction.
Conclusions
The results on the factors linked to the effectiveness of SAP in everyday practice—frequency of sensor usage, elevated baseline HbA1c, and older age of the patients—are consistent with information from recently published RCTs in volunteers.
Outpatient safety assessment of an in-home predictive low-glucose suspend system with type 1 diabetes subjects at elevated risk of nocturnal hypoglycemia
Buckingham BA1, Cameron F2, Calhoun P3, Maahs DM4, Wilson DM1, Chase P4, Bequette BW2, Lum J3, Sibayan J3, Beck RW3, Kollman C3
1Stanford University, Stanford, CA; 2Rensselaer Polytechnic Institute, Troy, NY; 3Jaeb Center, Tampa, FL; and 4Barbara Davis Center for Childhood Diabetes, University of Colorado, Denver, CO
Diabetes Technol Ther 2013;
Background
Commercially available SAP with LGS is based on suspension of the basal rate when the CGM glucose falls below a set threshold. This pilot study tests in the home setting a prototype SAP system where the LGS is activated for up to 2 hours when hypoglycemia is predicted, rather than when the actual threshold value is reached.
Methods
Subjects with type 1 diabetes (n=19, aged 18–56 years) were randomized to 21 nights of study, either 14 with the LGS prediction on or 7 with the feature off. The pump communicated with a bedside laptop computer that contained the algorithm for predictive insulin suspend (three algorithm versions were tested) and the randomization procedure.
Results
With the final algorithm, LGS occurred on 53% of the nights and the occurrence of overnight hypoglycemia was reduced from 30% of nights with SAP alone to 16% using the SAP+predictive LGS. Mean morning glucose was 8.0 mmol/L (144 mg/dL) versus 7.4 mmol/L (133 mg/dL) on intervention versus control nights. Morning blood ketone levels were >0.6 mmol/L after LGS on only one occasion.
Conclusions
Predictive LGS in the home setting is safe and shows promise as a technology for substantially reducing nocturnal hypoglycemia.
The study of Bergenstal et al. is important RCT evidence that confirms the safety and efficacy of SAP with LGS, though the technology has now been in use and shown to be safe and effective in Europe for about 4 years. Note that patients with frequent SH (more than one episode in the previous 6 months) were excluded, so the use of SAP+LGS in the group that is arguably the most in need of it was not tested. Indeed, as previously discussed in the Yearbook, the limited evidence for a reduction in SH with SAP (as opposed to mild-to-moderate hypoglycemia) is mostly because of the exclusion in trials of those with frequent hypoglycemia at baseline. Observational and RCT evidence for a reduction in SH with SAP, with or without LGS, has now been reported (3,4) and should be available in published form soon.
It is clear from both RCTs and now from the large, real-life study of SAP by Nørgaard et al. that the efficacy of SAP depends on frequent use of the sensor and is best applied in those with poor glycemic control at baseline. Note also in this study the interesting fact that hospital admissions and length of stay were reduced during SAP therapy—presumably this and the best application of SAP in those with poor control will be important determinants in any cost-effectiveness calculations, essential for reimbursement of CGM.
The study of Buckingham et al. on predictive LGS pumps documents their potential value, and pumps using this feature are set to become commercially available in the near future. We will need to know how effective they are at reducing SH in comparison with current threshold SAP+LGS.
Reduction in Blood Glucose Variability and Hypoglycemia with Insulin Pump Therapy
Glucose variability assessed by low blood glucose index is predictive of hypoglycemic events in patients with type 1 diabetes switched to pump therapy
Crenier L, Abou-Elias C, Corvilain B
Department of Endocrinology, ULB-Erasme Hospital, Brussels, Belgium
Diabetes Care 2013;
Background
Mild-to-moderate hypoglycemia is common in type 1 diabetes, may limit efforts to improve control, and may have adverse psychosocial effects in patients. Little is known about the factors that predict any reduction in this type of hypoglycemia on switching from MDI to CSII. The aim of this study was to identify such factors, particularly to test whether glycemic variability at baseline identifies those who have the best response to CSII in reducing hypoglycemia.
Methods
Adults with long-standing type 1 diabetes (n=50) treated by CSII because of persistent elevated HbA1c or frequent hypoglycemia on MDI were studied. In addition to HbA1c, various glycemic variability indices were calculated from meter-downloaded self-monitored blood glucose profiles at baseline and after 6 months of CSII.
Results
The patients had high glycemic variability at baseline, which was reduced by CSII: glucose SD was reduced from 5.0 mmol/L (90.8 mg/dL) on MDI to 4.6 mmol/L (83.5 mg/dL) on CSII (p=0.006). HbA1c reduced from 8.04% to 7.48% (p=0.00001). The baseline HbA1c on MDI was independently correlated with change in HbA1c on switching to CSII. Hypoglycemic events were in general not significantly different on MDI and CSII. In a multivariate analysis, baseline low blood glucose index (LBGI) correlated with change in hypoglycemia on switching to CSII, and those patients with the highest tertile of LBGI on MDI had a significant (23.3%) reduction in hypoglycemic events and a reduction in LBGI when treated by CSII.
Conclusions
Glycemic variability is reduced on CSII versus MDI. Those patients with the highest glycemic variability as measured by LBGI had the largest reduction in hypoglycemic events when switching from MDI to CSII.
Glycaemic variability in paediatric patients with type 1 diabetes on continuous subcutaneous insulin infusion (CSII) or multiple daily injections (MDI): a cross-sectional cohort study
Schreiver C1, Jacoby U1, Watzer B2, Thomas A3, Haffner D1,4, Fischer D-C1
1Department of Paediatrics, University Hospital Rostock, Rostock, Germany; 2Department of Paediatrics, Philips University, Marburg, Germany; 3Medtronic GmbH, Meerbusch, Germany; and 4Department of Paediatrics, Liver and Kidney Disease, Hannover Medical School, Hannover, Germany
Clin Endocrinol 2013;79:641–7
Background
The aim of this study was to test the hypothesis that glycemic variability is less in type 1 diabetic children treated by CSII than in those treated by MDI and to relate variability to markers of oxidative stress and cyclo-oxygenase activity.
Methods
In a cross-sectional design, 22 patients using CSII (mean age 12.6 years) and 26 using MDI (mean age 13.1 years) underwent real-time CGM for 3 days in order to assess blood glucose variability, including by the indices of mean amplitude of glycemic excursions (MAGE) and glucose SD. Two consecutive 24-hour urine collections were used to measure F2-isoprostanes and PGF2.
Results
Patients on CSII were of comparable age, sex, body–mass index, and diabetes duration to those on MDI. MAGE was lower in CSII versus MDI patients (4.96 vs. 6.50 mmol/L, p<0.01), as was SD of glucose levels (2.54 vs. 3.15 mmol/L, p<0.05). Mean glucose levels, number of episodes of glucose <3.9 mmol/L, and isoprostanes and PGF2 did not differ between groups.
Conclusions
Insulin pump therapy is associated with less glycemic variability than MDI, even when mean blood glucose levels are comparable between the therapies.
A population-based study of risk factors for severe hypoglycemia in a contemporary cohort of childhood-onset type 1 diabetes
Cooper MN1,2, O'Connell SM2, Davis EA1–3, Jones TW1–3
1Telethon Institute for Child Health Research, University of Western Australia, Perth, Australia; 2Department of Endocrinology and Diabetes, Princess Margaret Hospital, Perth, Australia; and 3School of Paediatrics, University of Western Australia, Perth, Australia
Diabetologia 2013;56:2164–70
Background
SH is the single most important barrier to achieving strict glycemic control in type 1 diabetes. Changes in the epidemiology of SH may have occurred in recent years as a result of improved therapies such as insulin pump therapy. This study aimed to investigate the rates and risk factors for SH in a contemporary cohort of children and adolescents with type 1 diabetes.
Methods
Children and adolescents attending a diabetes clinic in Perth, Australia, between 2000 and 2011 were studied (n=1,770; mean age 8.6 years; range 0–17.8 years). Data on management, demographics, and complications such as SH were prospectively recorded.
Results
The proportion of subjects using CSII increased in all age groups over the years 2000–2011, for example, from about 1% to 30% in those >12 years of age. SH rates declined over the years from a peak of 21.8 in 2002 to 5.5 in 2006, and 6.2 events/100 patient-years in 2011. Compared to those on injection regimens, those aged 12–18 years on CSII were at reduced risk of SH: incidence risk ratio (IRR) of CSII versus twice-daily regimen 0.58, IRR of MDI versus twice-daily regimen 1.45.
Conclusion
SH rates have declined in recent years, and the usage of pumps and MDI has increased. At least in older children, reduced SH is more often associated with CSII than with either MDI or twice-daily insulin regimens.
High blood glucose variability is important because it is related to a high frequency of hypoglycemia in type 1 diabetes (5), and it is a source of much frustration for patients who are trying to manage their diabetes more effectively. Whether variability is a risk factor for diabetic macrovascular or microvascular disease is still under discussion. Two of these studies confirm that blood glucose variability is lower on CSII than on MDI, a fact recorded in previous studies (6). Crenier et al. show that those patients with the highest LBGI, a measure of variability with weighting for hypoglycemia, enjoy the largest reduction in hypoglycemia on switching to CSII. This highlights the fact that although some studies may show no change in hypoglycemia frequency with CSII versus MDI, as with the whole group in the Crenier report here, the significant and notable effects of pump therapy in reducing hypoglycemia emerge when a high-risk, hypoglycemia-prone subgroup is treated.
Cooper et al. report some interesting facts on the contemporary epidemiology of SH, which is declining in incidence, possibly related to the increasing use of CSII. This is indicated by SH rates in the oldest children being lowest in those on pumps. Recent data from the Type 1 Diabetes Exchange Clinic Registry in the United States also show that pump users have the lowest SH rates (6).
Author Disclosure Statement
J.P. has received speaker and/or consultancy fees from Medtronic, Roche, CeQur, Novo Nordisk, and Eli Lilly.