| |
New Inhaled Insulin for Diabetes 2
|
| |
| |
"In April, 2008, there came reports that, in clinical trials of Exubera, the incidence of new primary lung cancer was five cases over 3900 patient-years (0·13 per 100 patient-years) compared with one case over 4100 patient-years (0·02 per 100 patients-years) for patients treated with a comparator (injected).11 Because these patients diagnosed with lung cancer had a history of cigarette smoking, association with the insulin therapy was regarded as inconclusive. From a scientific perspective, much will depend on the extent of exposure of the alveolar epithelium and underlying connective tissue to biologically active insulin, which in turn reflects the manner in which the inhaled insulin traverses from the alveolus into the pulmonary circulation.12 Until concerns about the effects of high concentrations of insulin on local IGF-I receptors are resolved, this aspect of safety will require special vigilance.13"
New Inhaled Insulin Matches Injections for Type 2 Diabetes - full text Lancet article follows below
MedPage Today
Published: June 24, 2010
Action Points
* Explain to interested patients that a study has found that a new inhaled insulin preparation given at meals in addition to a bedtime insulin shot worked just as well as twice-daily insulin shots in controlling HbA1c levels.
* Explain that longer-term studies are necessary to insure inhaled insulin's safety.
ORLANDO -- A novel formulation of inhaled insulin (Technosphere), when given at meals with bedtime insulin glargine (Lantus), equals twice-daily premixed biaspart insulin (NovoLog) for glucose control in type 2 diabetes, a randomized trial showed.
Glycosylated hemoglobin fell by an absolute 0.68% with the inhaled insulin combo and 0.76% with biaspart insulin in the per-protocol analysis, which passed the 0.4% margin for noninferiority, found Julio Rosenstock, MD, of the Dallas Diabetes and Endocrine Center at Medical City in Dallas, and colleagues.
The potentially more convenient regimen also reduced weight gain and hypoglycemia compared with biaspart insulin, they reported online in The Lancet in conjunction with a presentation at the American Diabetes Association meeting.
Dropping from two injections daily to one plus inhaled insulin -- administered from a palm-sized device -- may make initiation of insulin therapy more palatable to patients with type 2 diabetes and improve compliance as well, noted Clifford J. Bailey, PhD, of Birmingham (England) Children's Hospital, and Anthony H. Barnett, MD, of the University of Birmingham, England.
However, given the checkered past of inhaled insulins in both safety and acceptance, Bailey and Barnett suggested a "proceed with caution" tack in their editorial accompanying the Lancet paper.
The Exubera brand of inhaled insulin briefly entered the market but was withdrawn by Pfizer for financial reasons after doctors and patients showed a lack of interest (See Citing Lack of Acceptance, Pfizer Pulls Plug on Inhaled Insulin).
Following the demise of this "blockbuster that never was," pharma companies abandoned several other promising delivery systems for inhaled insulin as well, the editorialists explained.
The real problem, though, was an apparent excess of lung cancer with Exubera compared with injected insulin, they said (See Inhaled Insulin Associated with Excess of Lung Cancer Cases).
While the association with inhaled insulin was inconclusive, this issue is going to require special vigilance for any inhaled insulin delivery system "until concerns about the effects of high concentrations of insulin on local IGF-I receptors are resolved," they wrote in the editorial.
In Rosenstock's trial, inhaled insulin showed no elevated incidence of cancer or cardiovascular or cerebrovascular events, but more extensive and longer evaluation would likely be necessary.
The international, open-label study included 677 adults with type 2 diabetes uncontrolled (mean hemoglobin A1c 8.7%) on insulin therapy, with or without oral antidiabetes drugs.
Patients were randomized to 52 weeks of treatment with either prandial inhaled insulin plus bedtime insulin glargine or twice daily premixed biaspart insulin (30% short-acting insulin analog, 70% intermediate-acting NPH insulin), both titrated to perceived treatment goals.
The primary endpoint of change in hemoglobin A1c from baseline to week 52 showed no difference between groups that exceeded the 0.4% threshold for noninferiority in the following analyses:
* Modified intent-to-treat population (all randomized patients who got at least one dose of study drug and one follow-up hemoglobin A1c measurement), with a difference of 0.06% favoring premixed insulin.
* Last observation carried forward, with a difference of 0.12% favoring premixed insulin.
* Per protocol, with a difference of 0.07% favoring premixed insulin.
Among the secondary outcomes, mean fasting plasma glucose at week 52 actually improved more with the inhaled insulin regimen (7.8 versus 8.7 mmol/L for biaspart insulin, change from baseline 2.0 versus 1.0 mmol/L, P=0.0029).
The inhaled insulin group was just as likely to get to the 7.0% hemoglobin A1c goal by week 52 as were those on twice daily insulin injections (22% versus 27%, P=0.2793).
Mean weight gain was significantly lower with the inhaled insulin regimen than with premixed insulin (0.9 versus 2.5 kg, P=0.0002).
Overall adverse event rates were similar between groups, but hypoglycemia -- the most common treatment-associated event -- was less common in the inhaled insulin group for mild-to-moderate episodes and overall (both 47.99% versus 68.88%, P<0.0001).
Both the weight and hypoglycemia results "would not be unexpected if the reduction in HbA1c is slightly less," the editorialists wrote.
As expected, cough was common with inhaled insulin (32% versus 4% with biaspart insulin), typically early in treatment and soon after insulin inhalation. Measures of lung function such as vital capacity and diffusing capacity of carbon monoxide did not decrease more with inhaled insulin than with the premixed insulin injections.
Infections, particularly upper respiratory tract infections, were also elevated with inhaled insulin.
The researchers noted that the modest hemoglobin A1c improvements might have been greater if the trial had used forced titration. The study was also limited by the unblinded trial design.
The study was funded by Mannkind, which is the company developing the inhaled insulin used in this trial.
Rosenstock reported financial conflicts of interest with MannKind, Pfizer, Roche, sanofi-aventis, Novo Nordisk, Eli Lilly, MannKind, GlaxoSmithKline, Takeda, Daiichi Sankyo, Forest, Johnson & Johnson, Novartis, Amylin, Merck, Roche, Bristol-Myers Squibb, AstraZeneca, and Boehringer Ingelheim.
Co-authors reported conflicts of interest with MannKind, Biodel, Takeda, Johnson & Johnson, GlaxoSmithKline, Sirtris, and Servier. Six co-authors were employed by and held shares or stocks in MannKind.
Bailey reported having attended advisory boards for sanofi-aventis and Novo Nordisk, and delivered continuing medical education programs sponsored by Novo Nordisk and Eli Lilly. Barnett reported having attended advisory boards for sanofi-aventis, Novo Nordisk, Eli Lilly, Pfizer, and Mannkind, and given continuing medical education programs sponsored by sanofi-aventis, Novo Nordisk, Eli Lilly, and Pfizer.
The Lancet, Volume 375, Issue 9733, Pages 2244 - 2253, 26 June 2010
Prandial inhaled insulin plus basal insulin glargine versus twice daily biaspart insulin for type 2 diabetes: a multicentre randomised trial
Dr Julio Rosenstock MD a Corresponding AuthorEmail Address, Daniel L Lorber MD b, Luigi Gnudi MD c, Campbell P Howard MD d, David W Bilheimer MD d, P-C Chang MS d, Richard E Petrucci MD d, Anders H Boss MD d, Peter C Richardson MRCP d
a Dallas Diabetes and Endocrine Center at Medical City, Dallas, TX, USA
b Diabetes Control Foundation, Diabetes Care and Information Center, Flushing, NY, USA
c Unit for Metabolic Medicine, Cardiovascular Division, King's College London, London, UK
d MannKind Corporation, Valencia, CA, USA
Summary
Background
Insulin therapy is often a delayed strategy in patients with type 2 diabetes mellitus because it is associated with weight gain, hypoglycaemia, and the need for subcutaneous injections. We aimed to assess the efficacy and safety of prandial Technosphere inhaled insulin compared with twice daily biaspart insulin.
Methods
In this randomised, open-label, parallel-group study, adult patients with type 2 diabetes mellitus and poor glycaemic control despite insulin therapy, with or without oral antidiabetes drugs, were enrolled from ten countries between Feb 23, 2006, and Aug 8, 2007. Patients were randomly allocated in a 1:1 ratio to receive 52 weeks' treatment with: prandial Technosphere inhaled insulin powder plus bedtime insulin glargine; or twice daily premixed biaspart insulin (70% insulin aspart protamine suspension and 30% insulin aspart of rDNA origin). The primary endpoint was a comparison of change in glycosylated haemoglobin (HbA1c) from baseline to week 52 between treatment groups; the non-inferiority margin was 0·4%. Analysis was by per protocol for non-inferiority testing of the primary endpoint. This study is registered with ClinicalTrials.gov, number NCT00309244.
Findings
334 patients were allocated to inhaled insulin plus insulin glargine, and 343 to biaspart insulin; 107 patients on inhaled insulin plus insulin glargine and 85 on biaspart insulin discontinued the trial. 211 patients on inhaled insulin plus insulin glargine and 237 on biaspart insulin were included in per-protocol analyses. Change in HbA1c with inhaled insulin plus insulin glargine (-0·68%, SE 0·077, 95% CI -0·83 to -0·53) was similar and non-inferior to that with biaspart insulin (-0·76%, 0·071, -0·90 to -0·62). The between-group difference was 0·07% (SE 0·102, 95% CI -0·13 to 0·27). Patients had significantly lower weight gain and had fewer mild-to-moderate and severe hypoglycaemic events on inhaled insulin plus insulin glargine than on biaspart insulin. The safety and tolerability profile was similar for both treatments, apart from increased occurrence of cough and change in pulmonary function in the group receiving inhaled insulin plus insulin glargine.
Interpretation
This study is part of a large clinical development programme addressing the efficacy and tolerability of use of Technosphere inhaled insulin in a wide variety of patients.
Funding
MannKind.
Introduction
The metabolic abnormalities of type 2 diabetes include insulin resistance and progressive insulin deficiency affecting peripheral glucose uptake and suppression of hepatic glucose production in the postabsorptive and early postprandial periods.1 Although oral antidiabetes drugs might control hyperglycaemia early in disease, progressive β-cell insufficiency can eventually lead to the need for insulin therapy to achieve glycaemic control.2-4
Biphasic insulin can provide greater reduction in glycosylated haemoglobin (HbA1c) than can basal insulin glargine, but at the expense of increased insulin doses, hypoglycaemia, and weight gain.5 In the 4-T study,6 treatment with biphasic or prandial insulin for 1 year in patients with type 2 diabetes produced similar reductions in HbA1c, both of which were greater than with basal insulin detemir, but at the expense of increased hypoglycaemia and weight gain. At 3-year follow-up, most patients eventually needed intensified treatment with a basal prandial regimen, consistent with consensus recommendations.3, 7
Present prandial insulin therapy has a late onset and extended duration of action, leading to early postprandial hyperglycaemia and increased risk of late postprandial hypoglycaemia. Technosphere (MannKind, Valencia, CA, USA) is an inhaled insulin formulation containing recombinant human insulin adsorbed onto powder (fumaryl diketopiperazine).8 Insulin and fumaryl diketopiperazine are almost completely cleared from the lungs of healthy individuals within 12 h of dosing. With Exubera inhaled insulin, the insulin concentration in epithelial lung fluid of patients with diabetes 12 h after the dose was about 8-9% of the concentration immediately after the dose. By comparison, 12 h after healthy individuals were given 60 U Technosphere inhaled insulin, the lungs contained 0·3% of the insulin concentration and 0·4% of the fumaryl diketopiperazine concentration recorded at 30 min after inhalation.9, 10
The Technosphere system allows for pulmonary delivery of peptide hormones, resulting in safe and rapid absorption of large molecules.11 Once inhaled, the insulin dissolves immediately on contact with the lung surface, and is rapidly absorbed into the systemic circulation.8 Inhaled insulin reaches a maximum blood insulin concentration within 15 min and produces rapid glucose lowering with a short duration of action (about 2-3 h),12-16 allowing suppression of endogenous glucose production earlier than do rapid-acting insulin analogues.17
We aimed to compare the efficacy and safety of prandial Technosphere inhaled insulin plus bedtime insulin glargine versus twice daily biaspart insulin for treatment of type 2 diabetes in patients previously treated with insulin, with or without oral agents.
Discussion
Treatment with Technosphere inhaled insulin plus insulin glargine led to reductions in HbA1c concentrations that were maintained throughout the 52-week trial, and were similar to those of biaspart insulin. A study of similar design comparing Exubera inhaled insulin with subcutaneous injection of prandial insulin showed a decline in HbA1c of 0·4% and 0·5%, respectively, which was similar to the reduction in HbA1c that we recorded in this study.19 In our study, inhaled insulin plus insulin glargine also provided significantly greater reduction in fasting plasma glucose and 1 h postprandial glucose concentrations than did biaspart insulin. Furthermore, fewer hypoglycaemic episodes (both mild to moderate and severe) and lower weight gain occurred in patients on inhaled insulin plus insulin glargine than in those on biaspart insulin.
HbA1c reductions through 52 weeks and postprandial glucose AUC values for 0-360 min at 52 weeks were each similar between groups, but each group reached these AUC values differently. Consistent with the pharmacokinetic and pharmacodynamic properties of the treatments, AUC for 0-120 min was slightly lower with inhaled insulin plus insulin glargine, whereas AUC for 120-360 min was lower with biaspart insulin. However, the biaspart insulin group had long-term effects beyond 4 h with significantly more episodes of hypoglycaemia.
Insulin titration to glycaemic targets was encouraged and guided by an algorithm on the basis of self-monitoring blood glucose profiles. However, titration was not enforced, which explains the modest HbA1c reductions in both groups that were previously treated with different conventional insulin regimens. However, forced titration used in a study of another inhaled insulin resulted in robust and clinically meaningful reductions in HbA1c concentrations.20
Many studies have shown the efficacy of insulin therapy in type 2 diabetes.5,21-23 Basal insulin alone improves glycaemic control, but as HbA1c concentrations decrease, the effect of postprandial glucose control on HbA1c increases.24 Present regimens for intensified insulin therapy by use of a prandial insulin to achieve aggressive HbA1c targets are hampered by the occurrence of increased hypoglycaemia and weight gain, raising concerns about the long-term safety of this approach.25-27 Findings from the 4-T study6 at 1 year showed that prandial insulin aspart and biaspart insulin produced better HbA1c control than did insulin detemir, but with a greater risk of increased hypoglycaemia and weight gain. However, forced insulin titration in the APOLLO study28 led to similar lowering of HbA1c with insulin lispro thrice daily as with insulin glargine once daily.
Compared with subcutaneously administered rapid-acting insulin analogues, the pharmacokinetics of Technosphere inhaled insulin more closely mimic early phase insulin release.15, 29 The reduced risk of hypoglycaemia with inhaled insulin plus insulin glargine was correlated with the rapid onset and shorter duration of action of Technosphere inhaled insulin versus biaspart insulin, suggesting that Technosphere inhaled insulin might become an option to reduce risk of hypoglycaemia in patients with insulin-treated type 2 diabetes.
In our study, the standard meal challenge was associated with a lower 1 h postprandial glucose and fasting plasma glucose with inhaled insulin plus insulin glargine than with biaspart insulin, which was probably related to greater suppression of endogenous glucose production.17 The reduced weight gain seen with inhaled insulin plus insulin glargine relative to biaspart insulin might relate to better synchronisation between the ultrarapid pharmacokinetics of inhaled insulin and the mealtime blood glucose profile, and could also be a clinical indication of lower hypoglycaemia with inhaled insulin plus insulin glargine, but the mechanism is not well understood. Reduced weight gain with inhaled insulin plus insulin glargine was not attributed to metformin use, which was equal in both treatment groups.
In our study, inhaled insulin was well tolerated apart from an increased frequency of cough, and an increased risk of infections, especially upper-respiratory-tract infections. For all patients with type 2 diabetes who had been treated for up to 2 years, the integrated summary of safety of the Technosphere inhaled insulin development programme, reported to the US Food and Drug Administration, recorded infections in 30% (537/1795) of those on inhaled insulin and 32% (435/1345) of those receiving a comparator. Furthermore, upper-respiratory-tract infections were recorded in 8% (151/1795) of patients on inhaled insulin and 10% (131/1345) of those in comparator groups. No differences were shown in the occurrence of cancer, cardiovascular events, or cerebrovascular events between groups. Long-term safety studies are planned with Technosphere inhaled insulin after regulatory approval.
The occurrence of cough rarely led to study discontinuation. The pattern and characteristics of cough in the group receiving inhaled insulin plus insulin glargine was probably due to stimulation of the cough reflex by inhalation of a dry powder formulation, which abated with continued use. Changes in pulmonary function tests were small, asymptomatic, and not judged to be clinically significant by investigators. In a study assessing pulmonary function at 3 months after cessation of Technosphere inhaled insulin, the small, non-progressive differences from baseline FEV1 and DLCO disappeared in patients with type 1 and type 2 diabetes, indicating that changes in pulmonary function are reversible when treatment has ended.30 These results are consistent with another study of inhaled insulin in which differences in pulmonary function compared with conventional insulin were resolved when Exubera was discontinued after 2 years of treatment.31
Our study is limited by the fact that the design and the different administration times of the comparator drugs did not allow masking. However, because of the need to titrate insulin doses and monitor glycaemia, masking in studies assessing insulin is not judged to be feasible by some regulatory authorities. Greater reductions in HbA1c might have been achieved if insulin titration to glycaemic targets was systematically enforced during the study, rather than allowing clinicians to treat patients to their perceived glycaemic targets. This study was designed to compare a prandial-based inhaled insulin therapy with a premix-based therapy since the selected premixed therapy is widely used in patients with type 2 diabetes. With the increased use of basal bolus regimens for the treatment of patients with type 2 diabetes, future studies comparing such regimens are warranted. Safety considerations restricted the population in this study, but the safety data gathered in this trial will allow larger and more diverse populations to be included in future trials.
Our findings show that inhaled insulin plus insulin glargine, alone or in combination with an oral antidiabetes drug (eg, metformin), is an effective alternative to conventional insulin therapy (biaspart insulin) in uncontrolled type 2 diabetes. We believe that use of Technosphere inhaled insulin, an ultrarapid prandial insulin, along with a basal insulin, could provide improved glycaemic control with lower weight gain and rates of hypoglycaemia in many individuals with type 2 diabetes.
Results
Of 2064 patients screened for eligibility, 677 (33%) were randomly allocated to treatment (figure 1). Baseline characteristics for age, sex, body-mass index, HbA1c, fasting plasma glucose, and diabetes duration were well balanced between groups (table 1).
The mean total daily dose of inhaled insulin was titrated rapidly upwards during the first 8 weeks of treatment, and began to plateau after the first 12 weeks: 82 U (SD 42) at week 1, 189 U (72) at week 12, 190 U (71) from after 3 months up to 6 months, 193 U (73) from after 6 months up to 9 months, and 198 U (74) from after 9 months up to 12 months. Table 2 shows the dose conversion for inhaled insulin to rapid-acting insulin analogue. A similar temporal pattern was recorded with biaspart insulin: 61 IU (SD 32) at week 1, 81 IU (43) at week 12, 83 IU (43) from after 3 months up to 6 months, 85 IU (45) from after 6 months up to 9 months, and 88 IU (48) from after 9 months up to 12 months. Mealtime doses of inhaled insulin were evenly distributed between meals, but the mean daily dose of insulin glargine was 47 IU (SD 25) at study end. Concomitant use and doses of metformin and thiazolidinediones were similar between groups (table 1).
Mean changes in HbA1c from baseline to week 52 were similar across all analysis populations with all upper 95% CIs less than 0·4, showing that inhaled insulin is non-inferior to biaspart insulin (table 3). The treatment effect remained stable from week 14 though week 52 (figure 2A). Additional information is provided in webappendix p 6.
Mean fasting plasma glucose values at week 52 were 7·8 mmol/L (SD 3·1) for inhaled insulin plus insulin glargine and 8·7 mmol/L (3·3) for biaspart insulin. Changes from baseline were 2·0 mmol/L (SD 0·3, 95% CI -2·5 to -1·5) for inhaled insulin plus insulin glargine versus 1·0 mmol/L (0·2, -1·5 to -0·5) for biaspart insulin. The between-group difference was -1·0 mmol/L (SD 0·3, 95% CI -1·6 to -0·3, p=0·0029; figure 2B). Additional information is provided in webappendix p 6. Seven-point blood glucose readings were recorded by patients at home on any 3 days during each week preceding a clinic visit. The variability between timepoints tended to be higher in the biaspart insulin group than in the inhaled insulin group, and glucose concentrations in both groups increased during the day (webappendix p 7).
At week 52, postprandial glucose AUC during 0-360 min was similar between treatment groups: 59·8 mmol/h per L for inhaled insulin plus insulin glargine versus 56·7 mmol/h per L for biaspart insulin (figure 3). However, patients on inhaled insulin plus insulin glargine versus those on biaspart insulin had significantly lower mean 1 h postprandial glucose (9·5 mmol/L [SD 3·7] vs 11·6 mmol/L [3·9], p=0·0001), and lower 1 h glucose excursion (1·8 mmol/L [SD 2·4] vs 2·9 mmol/L [2·4]). 2 h postprandial glucose was similar between treatment groups: 11·8 mmol/L (SD 4·1) for inhaled insulin plus insulin glargine versus 11·8 mmol/L (4·3) for biaspart insulin. After 2 h, glucose excursions were higher in patients on inhaled insulin plus insulin glargine than in those on biaspart insulin (data not shown). In the biaspart insulin group, postprandial glucose was lowered to below baseline after 200 min and remained below baseline thereafter (figure 3). Additional information is provided in webappendix p 8.
The proportion of patients with HbA1c of 7·0% or less at week 52 was similar between patients on inhaled insulin and insulin glargine (47/213 [22%]) and those on biaspart insulin (65/243 [27%], p=0·2793). Mean weight gain was significantly lower with inhaled insulin plus insulin glargine 0·9 kg (SD 0·3, 95% CI 0·3-1·5) than with biaspart insulin 2·5 kg (0·3, 1·9-3·0), with a treatment difference of -1·6 kg (SD 0·4, 95% CI -2·4 to -0·7, p=0·0002).
In the safety population, adverse events occurred in 272 patients (84%) on inhaled insulin plus insulin glargine and 296 (89%) of those on biaspart insulin. Treatment-emergent adverse events and serious adverse events are listed in table 4, and study drug-related adverse events are listed in webappendix pp 9-10. Hypoglycaemia was the most frequent adverse event attributed to study drugs, occurring in 99 patients (31%) on inhaled insulin plus insulin glargine and 163 (49%) on biaspart insulin. In the safety population, mild-to-moderate and total hypoglycaemia was significantly lower with inhaled insulin plus insulin glargine than with biaspart insulin (table 5). Severe hypoglycaemia event rate during the late night was lower with inhaled insulin plus insulin glargine than with biaspart insulin (table 6).
In the safety population, 103 patients (32%) treated with inhaled insulin plus insulin glargine reported cough compared with 14 (4%) receiving biaspart insulin. 142 events of cough were reported in patients on inhaled insulin plus insulin glargine, and most were characterised as intermittent (91 [64%]) or as a single-defined episode (48 [34%]). Furthermore, most of these cough events occurred within 10 min of insulin inhalation (109 [77%]) and were non-productive (127 [89%]), and most coughing episodes were reported during the first week of treatment (41 [29%]) and declined to about two (1%) per week by week 6.
Between patients on inhaled insulin plus insulin glargine and those on biaspart insulin, mean changes from baseline to week 52 were similar for FEV1 (-0·13 L [SD 0·22] vs -0·09 L [0·2], p=0·2173), forced vital capacity (-0·12 L [SD 0·25] vs -0·09 L [0·22], p=0·5265), and DLCO (-0·80 mL/min per mm Hg [SD 2·45] vs -1·11 mL/min per mm Hg [2·51], p=0·0946). Mean changes in laboratory parameters were similar between groups, as were occurrences of cardiovascular events, cerebrovascular events, and cancer.
Of patients randomly allocated to treatment, 191 (28%) withdrew from the study, with a greater number in those on inhaled insulin plus insulin glargine (32%) than on biaspart insulin (24%; figure 1). In the safety population, the proportion of patients discontinuing because of adverse events was higher with inhaled insulin plus insulin glargine (29 [9%]) than with biaspart insulin (12 [4%]; figure 1), mainly as a result of treatment-emergent adverse events affecting the respiratory tract, which accounted for 19 (58%) discontinuations on inhaled insulin plus insulin glargine. The most common respiratory event leading to discontinuation in patients on inhaled insulin plus insulin glargine was cough (6 [2%]). Hyperglycaemia led to two discontinuations from inhaled insulin plus insulin glargine and none from biaspart insulin, whereas hypoglycaemia led to no discontinuations from inhaled insulin plus insulin glargine and three from biaspart insulin. Of patients randomly allocated to treatment, four (1%) on inhaled insulin plus insulin glargine died (haemorrhagic stroke, myocardial infarction, sepsis, and worsening ischaemic heart disease), compared with one (<1%) on biaspart insulin (cardiac arrest). No deaths were judged to be related to the study drug in either group.
Insulin antibody concentrations increased in both groups at week 14 and continued to increase until they reached a plateau, with a slight decline, at about 10 months. The size of the increase was greater in patients treated with inhaled insulin plus insulin glargine. Insulin antibody concentrations were not correlated with changes in key clinical parameters such as HbA1c, fasting plasma glucose, insulin dose, hypoglycaemia, or pulmonary function tests.
From the SF-36 QoL and insulin treatment questionnaires, diabetes worries decreased significantly from baseline in the group on inhaled insulin plus insulin glargine (p=0·0083) but not in the biaspart insulin group; the between-group difference was not significant (p=0·1825). Attitudes towards insulin therapy, treatment satisfaction, and treatment preference significantly improved from baseline in both treatment groups (p<0·0001), with no significant between-group difference (p=0·5276). No between-group differences (p=0·2452) or within-group differences (p=0·1105) were recorded in any SF-36 component scales at week 45. Initiation of inhaled insulin plus insulin glargine was not associated with deterioration in overall physical quality of life or quality of life related to mental health, and we recorded significant improvements in diabetes worries, attitudes toward insulin therapy, treatment satisfaction, and treatment preference. These findings suggest that patients accepted treatment with inhaled insulin.
EDITORIAL
The Lancet, Volume 375, Issue 9733, Pages 2199 - 2201, 26 June 2010
Inhaled insulin: new formulation, new trial
Clifford J Bailey a bEmail Address, Anthony H Barnett c d
a Life and Health Sciences, Aston University, Birmingham B4 7ET, UK
b Birmingham Children's Hospital, Birmingham, UK
c Heart of England NHS Foundation Trust, Birmingham Heartlands Hospital, Birmingham, UK
d The Medical School, University of Birmingham, Edgbaston, Birmingham, UK
Fine tuning of insulin delivery remains a crucial challenge to better mimic the physiological control of glucose homoeostasis.1, 2 The present variety of insulins with different pharmacokinetic properties provides preparations that are rapid, short, intermediate, or longacting, and which partly fulfil this challenge. Multiple daily injections of basal (intermediate and longacting) and bolus (rapid and shortacting) insulins are widely accepted by patients with type 1 diabetes, but evidence and opinion about the use of complex regimens in type 2 diabetes remains inconsistent and controversial.3-5
Multiple daily injections are not regarded as a barrier to insulin therapy by most established users, but the prospect can deter starting treatment and might affect compliance, especially in patients with type 2 diabetes in whom insulin is needed later in life. The introduction of inhaled insulin to deliver a small pre-meal bolus of insulin is seen as a potential convenience to facilitate and possibly improve daytime glycaemic control.6, 7 Several inhaled preparations have been developed and one, Exubera, came briefly to the market (August, 2006, until October, 2007) before withdrawal because of economic constraints after poor adoption-a blockbuster that never was.8, 9 The demise of Exubera led to the discontinuation of several other promising delivery systems for inhaled insulin, although patients who have used inhaled insulins reported good acceptance and tolerability.7, 8 There were minor difficulties with dose equivalents to accommodate the low bioavailability of inhaled insulins, and various precautions and exclusions associated with respiratory diseases, infections, and smoking. The size and complexity of operation of the inhalation devices was an inconvenience that time would doubtless resolve, but the main issue concerned safety.9, 10
In April, 2008, there came reports that, in clinical trials of Exubera, the incidence of new primary lung cancer was five cases over 3900 patient-years (0·13 per 100 patient-years) compared with one case over 4100 patient-years (0·02 per 100 patients-years) for patients treated with a comparator (injected).11 Because these patients diagnosed with lung cancer had a history of cigarette smoking, association with the insulin therapy was regarded as inconclusive. From a scientific perspective, much will depend on the extent of exposure of the alveolar epithelium and underlying connective tissue to biologically active insulin, which in turn reflects the manner in which the inhaled insulin traverses from the alveolus into the pulmonary circulation.12 Until concerns about the effects of high concentrations of insulin on local IGF-I receptors are resolved, this aspect of safety will require special vigilance.13
In The Lancet today, Julio Rosenstock and colleagues14 report a 1-year open-label randomised parallel-group trial in which patients with type 2 diabetes who were poorly controlled (mean HbA1c 8·7%) had insulin therapy added with or without oral glucose-lowering agents. The insulin was either twice-daily premixed biaspart insulin or bedtime longacting insulin glargine plus prandial inhaled insulin powder (Technosphere). Although the study included several variables (insulin type, regimen, and route of delivery), the result was similar reductions in HbA1c, albeit with different dose titrations, with the biphasic injections (-0·76%) and the longacting plus inhaled (-0·68%) approaches. Use of the Technosphere inhalation afforded slightly lower early postprandial glucose excursions, as well as less weight-gain and fewer hypoglycaemic episodes, which would not be unexpected if the reduction in HbA1c is slightly less.
At this stage in 302 patients treated with the Technosphere inhalation, there were no apparent safety issues, although increased cough and respiratory events were noted. Thus inhaled bolus insulin-delivery as part of a basal-bolus regimen in patients with type 2 diabetes gave non-inferior efficacy and was tolerable compared against a conventional biphasic injection regimen. Potential safety concerns must await more extensive and longer evaluation, including further details about the manner in which the inhaled insulin traverses the alveoli.
The opportunity for convenient inhaled bolus insulin, to facilitate complex insulin-delivery regimens, will be welcomed by some patients. For now, we say: proceed with caution.
CJB has attended advisory boards for Sanofi-Aventis and Novo Nordisk, and delivered continuing medical education programmes sponsored by Novo Nordisk and Eli Lilly. AHB has attended advisory boards for Sanofi-Aventis, Novo Nordisk, Eli Lilly, Pfizer, and Mannkind, and given continuing medical education programmes sponsored by Sanofi-Aventis, Novo Nordisk, Eli Lilly, and Pfizer.
| |
| |
| |
|
|
|
