Drugs

, Volume 63, Issue 16, pp 1743–1778

Insulin Glargine

An Updated Review of its Use in the Management of Diabetes Mellitus

Authors

  • Christopher J. Dunn
    • Adis International Limited
  • Greg L. Plosker
    • Adis International Limited
  • Gillian M. Keating
    • Adis International Limited
  • Kate McKeage
    • Adis International Limited
    • Adis International Limited
Adis Drug Evaluation

DOI: 10.2165/00003495-200363160-00007

Cite this article as:
Dunn, C.J., Plosker, G.L., Keating, G.M. et al. Drugs (2003) 63: 1743. doi:10.2165/00003495-200363160-00007

Summary

Abstract

Insulin glargine is a human insulin analogue prepared by recombinant DNA technology. Modification of the human insulin molecule at position A21 and at the C-terminus of the B-chain results in the formation of a stable compound that is soluble at pH 4.0, but forms amorphous microprecipitates in subcutaneous tissue from which small amounts of insulin glargine are gradually released. The plasma concentration versus time profile of insulin glargine is therefore relatively constant in relation to conventional human insulins, with no pronounced peak over 24 hours. This allows once-daily administration as basal therapy.

Early randomised trials with insulin glargine generally showed greater reductions in fasting blood or plasma glucose levels and a reduced frequency of nocturnal hypoglycaemia relative to neutral protamine Hagedorn (NPH) insulin in patients with type 1 diabetes mellitus. In addition to this basal therapy, patients continued to use the regular mealtime insulin regimen to which they were accustomed. More recent data with insulin glargine have included evidence of improved glycaemic control, with improvements in satisfaction with treatment over NPH insulin. Furthermore, the time of day at which insulin glargine is injected has no clinically relevant effect on glycaemic control in these patients. There are also data from small, nonblind studies to suggest comparable glycaemic control with insulin glargine and continuous subcutaneous insulin infusion.

Results from comparative studies and meta-analyses in individuals with type 2 diabetes show lower incidences of nocturnal hypoglycaemia with insulin glargine than with NPH insulin, with two studies showing a significantly greater improvement in glycosylated haemoglobin levels with insulin glargine than with NPH. Insulin glargine is well tolerated, and is not associated with greater immunogenicity or increases in bodyweight than NPH insulin. Long-term data show maintenance of glycaemic control with insulin glargine for up to 39 months in adults and children with type 1 and adults with type 2 diabetes.

In conclusion, insulin glargine is an effective and well tolerated basal insulin therapy when given as a single daily subcutaneous injection to patients with diabetes, with benefits in terms of glycaemic control and reduced frequency of hypoglycaemia over regimens based on conventional basal insulins. Accumulating data and official recommendations show the suitability of insulin glargine for first-line use in selected patients with type 2 diabetes who require insulin treatment, as well as in patients with type 1 disease, and confirm its use in children and adolescents.

Pharmacodynamic Properties

Insulin glargine is a long-acting recombinant human insulin analogue made by modifying human insulin using recombinant DNA technology. These changes at position 21 of the A-chain and at the C-terminus of the B-chain enable insulin glargine to remain soluble in the acidic environment of the vial, but form amorphous microprecipitates in the neutral pH of subcutaneous tissue after injection. Insulin glargine is slowly released from these microprecipitates to provide basal insulin supplementation over its once-daily dosage interval. Isoglycaemic and euglycaemic clamp studies in patients with diabetes mellitus and healthy volunteers have been conducted to evaluate the glucose-lowering effects of insulin glargine. Results show a time profile of pharmacodynamic activity characterised by a slower onset but longer duration of activity than neutral protamine Hagedorn (NPH) insulin. Unlike NPH or ultralente insulin, insulin glargine has no peak effect and has almost constant glucose-lowering activity lasting 24 hours.

In general, at clinically relevant concentrations, insulin receptor binding kinetics of insulin glargine are similar to those of regular human insulin, and blood glucose levels are lowered by stimulation of peripheral glucose uptake and inhibition of hepatic glucose production. Physiological and biochemical responses to hypoglycaemia induced by insulin glargine in patients with type 1 diabetes mellitus and healthy volunteers were similar to those induced by regular human insulin. Mitogenic effects of insulin are thought to be primarily mediated via the insulin-like growth factor-1 (IGF-1) receptor, and insulin glargine appears to have a generally higher affinity for the IGF-1 receptor than regular human insulin in vitro . However, the clinical relevance of this is probably limited since the affinity of insulin glargine for the IGF-1 receptor is <0.5% that of IGF-1 in human skeletal muscle cells. Moreover, in most cell types tested in vitro, mitogenic activity was similar between insulin glargine and regular human insulin.

Pharmacokinetic Properties

The rate of absorption of insulin glargine appears to provide a basal insulin level that remains constant for at least 24 hours. Following subcutaneous injection, the absorption of insulin glargine (containing zinc 15 or 80 μg/mL) was significantly slower than that of NPH insulin in healthy volunteers, in terms of the time to disappearance of 25% radioactivity (8.8 and 11.0 vs 3.2h; p < 0.0001). Importantly, in healthy volunteers, absorption of the drug was similar irrespective of the site (arm, leg or abdomen) of administration of insulin glargine containing zinc 30 μg/ mL (standard formulation).

No accumulation of insulin glargine occurred with daily subcutaneous injections in patients with type 1 diabetes. On days 1, 4 and 11, the maximum free serum insulin concentration (Cmax) ranged from 192–250 pmol/L, trough free serum insulin concentrations ranged from 77–86 pmol/L and the time to Cmax ranged from 2.8–4.1 hours. Steady-state insulin glargine concentrations are achieved 2–4 days after the first dose.

Insulin glargine is partially degraded in the subcutaneous tissues to two active metabolites. Both unchanged drug and metabolites are present in the plasma.

There is a lack of data concerning the pharmacokinetics of insulin glargine in special patient populations, including paediatric patients and those with renal or hepatic impairment.

Numerous drugs may affect glucose metabolism resulting in the need for adjustment of the insulin glargine dosage.

Therapeutic Efficacy

Type 1 diabetes: Six previously reviewed randomised trials, each involving more than 250 adults (aged 18–70 years) with type 1 diabetes, compared insulin glargine given once daily at bedtime with NPH insulin once or twice daily over 4–28 weeks. In all these studies patients continued to use the regular mealtime insulin regimen to which they were accustomed, and dosages of basal insulin (glargine or NPH) were based on fasting blood glucose (FBG) targets ranging from around 4–7 mmol/L. In these studies, patients receiving insulin glargine showed greater reductions in mean FBG levels at study endpoint than NPH insulin recipients, with significant (p < 0.05) reductions from baseline in insulin glargine groups (but not the NPH insulin group) in one of the trials. Mean glycosylated haemoglobin (HbA1c) levels were reduced to a similar extent with both insulin types in all 16- and 28-week trials. The incidence of symptomatic hypoglycaemia was greater (p < 0.05) with NPH than with insulin glargine in two 28-week studies. Percentages of patients reporting at least one episode of nocturnal hypoglycaemia were higher with NPH than with insulin glargine in most trials, and significantly (p < 0.05) higher in two.

Since the last review in Drugs, data have become available from several further randomised comparisons of insulin glargine with NPH insulin in patients with type 1 diabetes. Study durations ranged from 28 weeks to 1 year, and all findings are currently available as abstracts. One trial recruited 585 patients whereas the others randomised approximately 50–120 patients each. Findings were broadly similar to those of the earlier comparisons, although smaller studies showed statistically significantly greater improvements in glycaemic control as shown by FBG and/or HbA1c levels with insulin glargine than with NPH insulin. Frequencies of moderate or severe episodes of nocturnal hypoglycaemia were reduced with insulin glargine relative to NPH insulin in one trial and the other reported statistically significant reductions with insulin glargine in incidences of overall and nocturnal hypoglycaemia. Additional German data from 378 patients have shown no effect on glycaemic control of the timing of the daily injection of insulin glargine (breakfast, dinner or bedtime). Long-term findings demonstrate maintenance of glycaemic control, with minimal effects on bodyweight, with insulin glargine for up to 3 years in patients with type 1 diabetes.

A 28-week analysis in 517 patients has shown improved levels of satisfaction with treatment according to the Diabetes Treatment Satisfaction Questionnaire and Well-Being Questionnaire with insulin glargine relative to NPH insulin treatment. Preliminary 1-year results from a further study in 121 Italian patients support these findings. Basal therapy with insulin glargine has also been compared with continuous subcutaneous insulin infusion (CSII) in three small studies. Results are currently inconclusive, although one parallel-group comparison in 32 patients has shown insulin glargine therapy and CSII to offer equivalent glycaemic control together with reductions in frequency of hypoglycaemia relative to previously used conventional multiple-dose insulin regimens.

Randomised comparisons in children and adolescents (one large 28-week parallel-group study and one small crossover trial with 16-week treatment periods) have shown greater antihyperglycaemic efficacy of insulin glargine than NPH insulin in terms of FBG but not HbA1c levels, with less nocturnal hypoglycaemia with insulin glargine. Maintenance of glycaemic control for up to 3 years has been reported in a long-term extension study in 129 children and adolescents. Recent nonrandomised trials and noncomparative investigations in which series of patients were switched to treatment with insulin glargine show results broadly concordant with the randomised studies.

Type 2 diabetes: Previously reviewed studies, each in over 100 patients with type 2 diabetes, were of up to 1 year’s duration and showed similar improvements in glycaemic control with either insulin glargine or NPH insulin. Lower incidences of nocturnal hypoglycaemia were found in most trials with insulin glargine relative to NPH insulin (ranges of 10–31.3% vs 24–40.2%). Two further large, randomised trials that compared insulin glargine with NPH insulin (with concurrent oral antihyperglycaemic therapy) have shown statistically significant (p < 0.05 all comparisons) reductions in risk of nocturnal hypoglycaemia with insulin glargine therapy. One study has also demonstrated a statistically significantly (p < 0.01) greater decrease from baseline in mean HbA1c level with insulin glargine given at breakfast time (reduction of 1.24%) than with NPH insulin (0.84%). In another similarly designed trial, there were no differences in the mean reduction in HbA1c levels or in the incidence of nocturnal, all symptomatic or severe hypoglycaemia with morning versus evening administration of insulin glargine.

Notably, two large (n = 2304 and 1786) meta-analyses of phase III and IV clinical trials indicated that there was a significantly (p ≤ 0.01 all comparisons) lower risk of all confirmed symptomatic or nocturnal hypoglycaemia with insulin glargine when compared with NPH insulin treatment. Poisson regression analysis showed that when the incidence of confirmed symptomatic hypoglycaemia or confirmed nocturnal hypoglycaemia was made equivalent in the two treatment groups, there was a clinically relevant and statistically significant reduction in HbA1c with insulin glargine versus NPH insulin.

Patients with type 2 diabetes showed improved treatment satisfaction during a 1-year randomised, multicentre trial in both the insulin glargine and NPH insulin groups. At weeks 36, the improvement was greater in the insulin glargine group than in the NPH insulin group.

Long-term data from a multicentre, nonblind, extension study in 239 patients with type 2 diabetes who received insulin glargine in combination with oral antidiabetic agents showed that he mean HbA1c level was reduced from 9.44% at baseline to 8.42% over a period of up to 39 months. Only two severely symptomatic episodes of hypoglycaemia were reported during this period.

Significant (p < 0.05 all comparisons) reductions from baseline in mean HbA1c levels after starting insulin glargine treatment have also been reported in patients with type 2 diabetes in several nonrandomised studies and in a small (n = 20; 16 with type 2 diabetes) retrospective analysis of patients with end-stage renal disease.

Effect on bodyweight: Bodyweight gains were no greater with insulin glargine than with NPH insulin in comparative studies in patients with type 2 diabetes, with one previously reviewed trial showing less weight gain with insulin glargine and another showing similar increases with either type of insulin. Although more recently reported trials have not yet reported details of effects on bodyweight of randomised treatment, several prospective case series have shown no clinically relevant increases in mean bodyweight in patients with type 2 disease receiving insulin glargine for up to 9 months, despite significant improvements in glycaemic control relative to baseline. Data collected over up to 36 months in patients with type 1 diabetes receiving insulin glargine during long-term follow-up have shown a minimal increase in mean bodyweight with insulin glargine treatment.

Tolerability

The incidence of adverse events with insulin glargine has been generally similar to that with NPH insulin in randomised clinical studies. Injection site reactions, most of which are minor, are the most common adverse events with insulin glargine, and are seen in around 3–4% of patients. Observations in 239 patients with type 2 diabetes who participated in a recent long-term extension study revealed no injection site reactions with insulin glargine for up to 39 months.

Evidence to date shows that insulin glargine is no more immunogenic than NPH insulin, and there have been no clinically relevant increases in levels of antibodies to Escherichia coli reported in clinical trials. There have also been no indications of any increases in risk of progression of diabetic retinopathy related to the generally higher affinity of insulin glargine than regular human insulin for IGF-1 receptors. Recent data from patients with end-stage renal disease who received insulin glargine have shown no specific tolerability concerns, and animal studies have indicated no effect of insulin glargine on fetal or postnatal development, and no evidence of carcinogenicity linked to interaction with IGF-1 receptors.

Dosage and Administration

Once-daily, self-administered, subcutaneous injections of insulin glargine provide basal insulin levels for the treatment of adults or children (aged >6 years) with type 1 diabetes and adults with type 2 diabetes. Insulin glargine should not be diluted or mixed with any other insulin or solution as this could alter its time-action profile because of the acidic nature of its formulation.

The dosage of insulin glargine is determined individually for each patient and adjustments to dosage are made according to blood glucose levels. In clinical trials, insulin-naive patients were started with a dose of 10 IU once daily and were maintained at dosages ranging from 2–100 IU once daily. In patients receiving once-daily NPH or ultralente insulin, the initial dose of insulin glargine was usually matched in terms of the number of international units administered, whereas the initial dose was reduced by approximately 20% for the first week and then adjusted according to blood glucose levels in patients previously treated with twice-daily NPH insulin. Paediatric patients should be managed in the same way. The dosage and timing of additional short-acting insulin or oral antidiabetic agents may need to be adjusted with the initiation of insulin glargine.

Introduction

Diabetes mellitus is estimated to affect approximately 150 million people worldwide and, of these, patients with type 2 diabetes account for approximately 85–95% of cases.[1] The prevalence of type 2 (non-insulin-dependent) diabetes is increasing rapidly, particularly in developed countries, as a result of aging populations, sedentary lifestyles, unhealthy diets and obesity.[1] Although type 2 diabetes is more common among the elderly, prevalence rates are rising rapidly among youth.[2] The incidence of type 1 (insulin-dependent) diabetes is highest in school children and adolescents but the disorder may occur at any age.[3]

Both types of diabetes are associated with excessive morbidity and mortality. Complications include coronary artery disease, stroke and peripheral vascular disease (macrovascular complications), retinopathy, nephropathy and neuropathy (microvascular complications). However, maintaining good glycaemic control with intensive therapy (i.e. treatment resulting in normalisation of blood glucose levels) reduces the risk of developing microvascular complications in both type 1[4] and type 2[5] disease. In patients with diabetes who require insulin replacement therapy, the aim is to induce metabolic effects similar to that produced by endogenous insulin secretion. Treatment usually includes a basal long-acting insulin together with short-acting insulin before meals, or together with an oral antidiabetic agent depending on the type and stage of diabetes.[6] Advances in recombinant DNA technology resulted in the manufacture of short- and long-acting human insulin analogues that are better able to induce physiological blood glucose levels than animal insulins, thereby enabling improvements in glycaemic control.[7]

Insulin glargine (Lantus®1) is a long-acting human insulin analogue that provides the basal component of insulin replacement therapy with once daily administration. This review focuses on its use in patients with type 1 and 2 diabetes, and updates clinical knowledge gained since the previous review in Drugs.[8]

Pharmacodynamic Properties

The pharmacodynamic properties of insulin glargine have been reviewed previously in Drugs.[8] This section provides a brief overview of the pharmacodynamic properties of insulin glargine, including recent studies published since the previous review (table I).
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab1.jpg
Table I

Overview of pharmacodynamic (PD) studies evaluating insulin glargine (IG; administered SC unless stated otherwise)

Insulin glargine is a long-acting recombinant human insulin analogue, with a duration of action of up to 24 hours.[8,9] The structure of insulin glargine differs from that of human insulin in that glycine replaces asparagine at position 21 of the A-chain, and two arginines are added to the C-terminus of the B-chain (figure 1). These changes enable insulin glargine to remain completely soluble in the acidic environment of the vial (pH 4.0) and precipitate in the neutral pH of subcutaneous tissue after injection. Insulin glargine is then slowly released from amorphous microprecipitates, achieving a relatively constant plasma insulin concentration over 24 hours with no pronounced peaks (section 3.1). This is consistent with the time profile of the pharmacodynamic activity of insulin glargine, as demonstrated in isoglycaemic and euglycaemic clamp studies in patients with type 1 diabetes and healthy volunteers (section 2.2). It also allows a flexible, once-daily subcutaneous administration schedule to provide the basal insulin requirements in patients with diabetes (section 6). As with other insulins, however, there may be inter- and intraindividual variation in the time profile of pharmacodynamic activity of insulin glargine.[9,10]
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Fig1.jpg
Fig. 1

Structural formulae of insulin glargine and regular human insulin.[8]

The commercially available formulation of insulin glargine contains a small amount of zinc (30 μg/mL), which is necessary for the formation of amorphous microprecipitates in the neutral pH of subcutaneous tissue after injection and further delays absorption.[8,9] Early-stage studies of insulin glargine used formulations with zinc concentrations of 15 or 80 μg/mL.[8]

Mechanism of Action

In general, at clinically relevant concentrations, insulin receptor binding kinetics of insulin glargine are similar to those of regular human insulin (table I),[11,12] and these agents are considered to mediate the same type of effect via the insulin receptor.[13] Thus, insulin glargine acts in the same way as other insulins in that its primary action is regulation of glucose metabolism.[9] Blood glucose levels are lowered through stimulation of peripheral glucose uptake, particularly by skeletal muscle and fat, as well as inhibition of hepatic glucose production.

Glucose-Lowering Effect

Insulin glargine has a glucose-lowering profile characterised by a relatively slow onset of action, no peak effect, and relatively constant pharmacodynamic activity lasting up to 24 hours (table I).[1518] This was demonstrated in a crossover trial using an isoglycaemic clamp technique to compare the time profiles of the pharmacodynamic activity of insulin glargine, neutral protamine Hagedorn (NPH) insulin, ultralente insulin (all given as single subcutaneous doses of 0.3 U/kg) and a continuous subcutaneous insulin infusion (CSII) of insulin lispro 0.3 U/kg/24h in 20 patients with type 1 diabetes.[15] As highlighted in figure 2, the onset of action was slower but the duration of action was longer with insulin glargine than NPH insulin. Moreover, the metabolic activity of insulin glargine was more constant than that of NPH or ultralente insulin and similar to that of insulin lispro CSII. From 3–24 hours, the glucose infusion rate was almost constant at 0.78 mg/kg/min for insulin glargine, whereas for NPH insulin it peaked at 3.4 mg/kg/min after 5.4 hours and decreased to 0 mg/kg/min by 20 hours. When defined as the time to an increase in plasma glucose level >8.3 mmol/L minus the time to onset, duration of action was significantly longer for insulin glargine than for NPH insulin (20.5 vs 13.2h, p < 0.05). Likewise, end of action (defined as the time at which the plasma glucose level increased to >8.3 mmol/L) occurred later with insulin glargine than NPH insulin (22 vs 14h, p < 0.05) [table I].
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Fig2.jpg
Fig. 2

Profile of glucose-lowering effect of insulin glargine and comparators in a randomised crossover trial using an isoglycaemic clamp technique.[15] Rates of glucose infusion needed to maintain plasma glucose at a target value of 7.2 mmol/L following subcutaneous injections of insulin glargine (0.3 U/kg), neutral protamine Hagedorn (NPH) insulin (0.3 U/kg), ultralente insulin (0.3 U/kg) and continuous subcutaneous insulin infusion (CSII) of insulin lispro (0.3 U/kg/24h) in 20 patients with type 1 diabetes mellitus. Reproduced from Lepore et al.[15] with permission from the American Diabetes Association Inc.

In a more recent study in 20 patients with type 1 diabetes, the time profile of pharmacodynamic activity was evaluated using an isoglycaemic clamp technique after the first and seventh dose of insulin glargine 0.3 U/kg subcutaneously at bedtime.[18] Comparing results on day 7 (which were more likely to reflect activity at steady state) versus those on day 1, multiple-dose administration was associated with an earlier onset of action (0.7 vs 1.5h, p < 0.05), longer duration of action (23.2 vs 20.5h, p < 0.05), lower interindividual variability (assessed by the standard deviation of glucose infusion rates, p < 0.05) and more constant glucose-lowering effect (table I).

Results in patients with diabetes are supported by euglycaemic clamp studies in healthy volunteers (table I), which showed a time profile of pharmacodynamic activity of about 24 hours for insulin glargine[16] and more constant metabolic activity (i.e. glucose consumption over time) than that observed with NPH insulin.[17] In a more recent report,[20] differences between insulin glargine and regular human insulin were assessed in terms of activation/deactivation of endogenous glucose output and stimulation of peripheral glucose disposal. The activation period involved intravenous infusion of equimolar doses of the respective insulins for 4 hours, and this was followed by a 3-hour deactivation period. Results of this randomised crossover study (n = 12) showed that suppression of endogenous glucose output (of which the hepatic contribution is ≈80%) was similar with insulin glargine or regular human insulin at all time points during the activation and deactivation periods. Likewise, the maximum rate of glucose disposal was similar for both types of insulin throughout the study period.

In patients with diabetes evaluated overnight with a continuous glucose monitoring system, those who received insulin glargine subcutaneously as basal insulin therapy (n = 8) had blood glucose levels outside the target range (3.9–11.1 mmol/L) for a greater proportion of time than those who received insulin lispro CSII (n = 11) [50.7% vs 20.9%, p = 0.04].[21] It was not stated whether patients had type 1 or 2 diabetes in this brief report.

Other Metabolic and Physiological Effects

Like other insulins, insulin glargine inhibits lipolysis and proteolysis, and increases protein synthesis.[9] The effects of insulin glargine and regular human insulin on lipid metabolism were compared in a randomised crossover study in healthy volunteers (n = 6) and patients with type 1 diabetes (n = 13) using a euglycaemic clamp technique for 2 hours.[22] Insulin glargine and regular human insulin significantly (p < 0.05) reduced plasma levels of nonesterified fatty acids from baseline, and the magnitude of reduction was similar between the agents. For example, in patients with type 1 diabetes, levels were decreased from 171 μmol/L at baseline to 55 μmol/L with insulin glargine (p = 0.04), and from 257 to 43 μmol/L with regular human insulin (p = 0.02).

Using a clamp technique, these investigators also evaluated the symptomatic response to hypoglycaemia induced by intravenous insulin glargine or regular human insulin in healthy volunteers (n = 6) and patients with type 1 diabetes (n = 13).[23] Blood glucose levels were progressively lowered over a 5-hour period (hourly target levels of 4.7, 4.2, 3.6, 3.1 and 2.5 mmol/L). Peak hypoglycaemic symptom scores (19.08 vs 17.46) and peak plasma adrenaline (epinephrine) levels (321.8 vs 332.5 ng/L) in patients with diabetes were similar with insulin glargine and regular human insulin. Results in healthy volunteers were also similar for the two treatment groups.

In a long-term study (briefly reported as an abstract) evaluating blood flow responses to intra-arterial infusions of vasodilators in 11 patients with type 2 diabetes treated with bedtime injections of insulin glargine for 3.5 years, marked improvements from baseline were observed for both endothelium-dependent (+86%, p < 0.01) and -independent (+72%, p < 0.01) vasodilation.[29] These preliminary results suggest a beneficial rather than deleterious effect of insulin glargine on vascular function.

Activation of Insulin-Like Growth Factor 1 and Studies of Mitogenic Potential

Mitogenic effects of insulin appear to be mediated primarily through the insulin-like growth factor-1 (IGF-1) receptor, although mitogenic responses may also be activated through the insulin receptor.[8,11] Changes to the structure of insulin, particularly at the C-terminus of the B-chain, can alter the way in which the insulin molecule interacts with the IGF-1 receptor.[11] Several in vitro studies have therefore compared insulin glargine with regular human insulin in terms of their affinity for the IGF-1 receptor and mitogenic or growth-promoting activity (table I).

In human skeletal muscle cells from healthy volunteers and patients with type 2 diabetes, the affinity of insulin glargine for the IGF-1 receptor was <0.5% that of IGF-1.[11] Although the affinity of insulin glargine for this receptor was generally higher than that of regular human insulin, a statistically significant difference was achieved only at very high (clinically irrelevant) ligand concentrations.[11] In other studies, insulin glargine also appeared to have greater affinity than regular human insulin for IGF-1 receptors,[14,25,30] although these results are complicated by differences in the relative expression of IGF-1 receptors on the various cell types used.[11] Moreover, because the affinity of insulin glargine for the IGF-1 receptor is much lower than that observed for endogenous IGF-1,[11,24] any differences between regular human insulin and insulin glargine in this regard may be of limited clinical significance.[11]

In most cell types tested, with the exception of human Saos/B10 osteosarcoma cells,[14] mitogenic activity was similar between insulin glargine and regular human insulin in vitro.[11,12,24,25] For example, uptake of [3H]thymidine in DNA of human skeletal muscle cells from healthy volunteers and patients with type 2 diabetes was similar between insulin glargine and regular human insulin.[11] Compared with IGF-1, both insulin glargine and regular human insulin had markedly lower sensitivities and potencies (<1% IGF-1) for thymidine uptake (sensitivities were calculated against the maximal possible response due to IGF-1 in each set of cells; potencies were derived from the highest response seen for each agent).[11]

Histological evidence of tumour development was not observed in mammary glands of rats or mice receiving insulin glargine (up to 12.5 U/kg/day) for up to 24 months, and no other neoplastic changes that may affect humans were detected.[27] Standard 2-year tests in mice and rats receiving insulin glargine at several times the recommended human subcutaneous dose found malignant fibrous histiocytomas at injection sites in male rats (statistically significant) and male mice (not statistically significant) in acid vehicle-containing groups, although the clinical relevance of these findings is unclear.[9,27] These tumours (which are associated with the permanent irritation of local connective tissue due to the acid pH of the vehicle solution) did not develop in female animals, or those in the saline control group or in insulin comparator groups using a different vehicle. Evaluation of reproductive toxicity and embryotoxicity of insulin glargine in rats and rabbits found no deleterious effects other than those induced by hypoglycaemia in response to relatively high doses of insulin glargine and NPH insulin.[9,31]

Pharmacokinetic Properties

The pharmacokinetic properties of insulin glargine have been reviewed previously.[8] This section provides a brief overview of the pharmacokinetics of the drug; data from clinical trials are supplemented by information from the manufacturer’s prescribing information.[9,13]

Absorption

The rate of absorption of insulin glargine appears to provide a basal plasma insulin level which remains constant for at least 24 hours.[9] Following subcutaneous injection, the absorption of insulin glargine was significantly slower than that of NPH insulin in healthy volunteers, in terms of the time to disappearance of 25% radioactivity from the injection site (T75%) and residual radioactivity at 24 hours.[32] In a randomised, single-blind, crossover study, significantly longer mean T75% values (8.8 and 11.0 vs 3.2h; p < 0.0001) and significantly more mean residual radioactivity at 24 hours (43.8% and 52.2% vs 21.9%; p < 0.0001) occurred with administration of 125I-labelled insulin glargine containing zinc 15 or 80 μg/mL versus NPH insulin to healthy volunteers (commercially available insulin glargine contains zinc 30 μg/mL).[32] When plasma exogenous insulin profiles were compared, a distinct peak concentration was reached 4–6 hours after administration of NPH insulin. In contrast, absorption of insulin glargine from the injection site occurred at a relatively constant rate with no prominent peak in plasma insulin concentration.[32]

Furthermore, there was significantly less fluctuation in insulin levels over 24 hours after single or multiple doses of insulin glargine than with NPH insulin or ultralente in three studies in healthy volunteers or patients with type 1 diabetes (reported in a single abstract).[33] Fluctuation was defined as the percentage deviation around the average serum insulin concentration over 24 hours (PF24), with lower values indicating a more physiological basal insulin profile.[33] In a single-dose study in healthy volunteers (n = 36), mean PF24 values with insulin glargine, NPH insulin or ultralente were 19.8%, 31.9% and 47.2%, respectively (p < 0.0001 both comparisons).[33] There was also less fluctuation in mean PF24 values in patients with type 1 diabetes (n = 20) receiving a single dose of insulin glargine than in recipients of a single dose of NPH insulin (14.2% vs 25.8%; p < 0.0001).[33] At steady state, mean PF24 values after 2, 5 and 12 days remained constant (19.8%, 20% and 20.1%, respectively) in 15 patients with type 1 diabetes receiving insulin glargine plus insulin lispro for 11 days and were comparable to those seen after single doses.[33] Due to the decreased fluctuation of serum insulin levels following single and multiple doses of insulin glargine, insulin glargine more closely approaches a physiological basal insulin profile and should increase the predictability of insulin action.

The site of administration had no effect on absorption of insulin glargine, with similar T75% values and mean residual radioactivity at 24 hours in healthy volunteers in a randomised, nonblind, crossover study.[32] Mean T75% values ranged from 11.9–13.2 hours and mean residual radioactivity at 24 hours ranged from 47.7–57.2% following administration of subcutaneous insulin glargine containing zinc 30 μg/mL in the arm, leg or abdomen.

In patients with type 2 diabetes (n = 14), there was a trend for slower absorption of insulin glargine compared with NPH insulin (statistical analysis not reported in abstract).[34] Median T75% was 15.0 hours with insulin glargine and 6.5 hours with NPH insulin and mean residual radioactivity at 24 hours was 54.4% and 27.9%, respectively.[34]

The residual radioactivity of insulin glargine at the injection site and the finding that the duration of action of the drug is >24 hours in the majority of individuals (section 2.2) led to concern that accumulation of circulating insulin may occur resulting in increased metabolic effects.[35] However, no accumulation of insulin glargine occurred with daily subcutaneous injections in patients with type 1 diabetes (n = 15) [table II].[35] On days 1, 4 and 11, the maximum free serum insulin concentration (Cmax) ranged from 192–250 pmol/L and the time to Cmax ranged from 2.8–4.1 hours;[35] trough free serum insulin concentrations ranged from 77–86 pmol/L (table II).[35] Steady-state insulin glargine concentrations are achieved 2–4 days after the first dose.[13]
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab2.jpg
Table II

Multiple-dose pharmacokinetics (mean values) of subcutaneous insulin glargine in patients with type 1 diabetes mellitus (n = 15) who received a constant once-daily dose for 12 days (mean dosage 24 U/day) [pharmacokinetic parameters derived from free serum insulin concentrations].[35] On days 1, 4 and 11, patients fasted for 24h after each dose of insulin glargine and did not receive any other insulin preparations. On all other days, patients received preprandial insulin lispro

Metabolism

Partial degradation of insulin glargine occurs in the subcutaneous tissues.[36] The drug is degraded at the carboxyl terminal end of the B-chain to two active metabolites: M1 (21A-Gly-insulin) and M2 (21A-Gly-des-30B-Thr-insulin).[36,37] Both unchanged drug and metabolites are present in the plasma.[36]

Special Patient Populations

There is a lack of data concerning the pharmacokinetics of insulin glargine in special patient populations (e.g. paediatric patients, patients with renal or hepatic impairment). However, insulin glargine requirements may be reduced in patients with impaired renal function (because of reduced metabolism of the drug) or impaired hepatic function (because of a diminished capacity for gluconeogenesis and reduced metabolism of the drug) [section 6].[13]

Potential Drug Interactions

Numerous drugs may affect glucose metabolism resulting in the need for adjustment of the insulin glargine dosage (section 6). For example, the blood glucose lowering effect of insulin glargine may be potentiated by oral antidiabetic agents, ACE inhibitors, dextropropoxyphene, disopyramide, fibrates, fluoxetine, monoamine oxidase inhibitors, pentoxifylline, salicylates and sulfonamide antibacterials.[13] Moreover, the blood glucose lowering effect of insulin glargine may be attenuated by corticosteroids, danazol, diazoxide, diuretics, glucagon, isoniazid, estrogens, progestogens, phenothiazine derivatives, somatropin, sympathomimetics (e.g. adrenaline, salbutamol [albuterol]) and thyroid hormones.[13] Other agents which may affect the blood glucose lowering effect of insulin glargine include β-blockers, clonidine, lithium salts or alcohol.[13]

Therapeutic Efficacy

Insulin glargine has been studied in both patients with type 1 and those with type 2 diabetes. Type 1 disease is associated with absolute insulin deficiency attributed to autoimmune destruction of pancreatic β-cells, while type 2 diabetes results from defective insulin secretion that progresses against a background of impaired utilisation of insulin by tissues (i.e. insulin resistance).[6,38]

The American Diabetes Association (ADA)[38] recommends three criteria for the diagnosis of diabetes:

  • symptoms of disease and random plasma glucose level of at least 11.1 mmol/L (200 mg/dL)

  • fasting plasma glucose (FPG) level of at least 7.0 mmol/L (126 mg/dL)

  • 2-hour postprandial plasma glucose level of 11.1 mmol/L (200 mg/dL) or more during an oral glucose tolerance test.

Note that the ADA defines ‘fasting’ as no caloric intake for at least 8 hours.[38]

British guidelines[6] state that the most important outcome measures in the assessment of management of diabetes are glycaemic control (as shown by fasting blood glucose [FBG] or FPG and glycosylated haemoglobin [HbA1c] levels); prevention of acute episodes of hypo- and hyperglycaemia; reduction in other macrovascular risk factors such as dyslipidaemia, high blood pressure and increased bodyweight; short-term quality of life, adverse events and treatment tolerance; and long-term effects on the incidence of diabetic complications, quality of life and mortality. In general, good glycaemic control is considered to be shown by an HbA1c level of less than 7%, a preprandial plasma glucose level of 5.0–7.2 mmol/L (90–130 mg/dL) and a peak postprandial plasma glucose level of below 10.0 mmol/L (<180 mg/dL).[38] Treatment goals should be individualised, however, with special populations (e.g. children and the elderly) requiring special consideration, and less intensive control targets set for patients with severe or frequent hypoglycaemia.

It should be noted that HbA1c levels show the extent to which circulating erythrocytes have been exposed to glucose, and thus give an indication of the patient’s overall glycaemic control over the preceding 2–3 months (the average life span of an erythrocyte is 120 days), whereas plasma or blood glucose measurements give a ‘snapshot’ view only.[39] Glucose can be measured in whole blood, serum or plasma, although plasma is recommended for diagnostic purposes. Because of differences in water content, glucose concentrations are approximately 11% higher in plasma than in whole blood, and around 5% lower than in serum.[39]

At the time of the last review of insulin glargine in Drugs,[8] large and well designed randomised trials of up to 52 weeks’ duration had established that a single daily subcutaneous injection of this agent provides effective basal glycaemic control, with a statistically and clinically significantly reduced risk of nocturnal hypoglycaemia relative to conventional NPH insulin (see McKeage and Goa[8]). Most data had been obtained in adults with type 1 or type 2 diabetes, with paediatric findings restricted to a single 28-week study in 349 children with type 1 disease.[40]

Recent studies of insulin glargine have continued to focus on glycaemic control and incidence of hypoglycaemia relative to patients receiving NPH insulin, particularly in patients with type 1 diabetes. Other developments have included the emergence of long-term efficacy data in both types 1 and 2 diabetes, early results from a number of analyses in patients with type 2 disease receiving insulin glargine in combination with oral antihyperglycaemic agents, and comparisons of insulin glargine therapy with CSII. Further data obtained from children and adolescents have also become available, as have additional findings relating to the effect of insulin glargine on bodyweight, particularly in patients with type 2 disease. Patients with end-stage renal disease who were prescribed insulin glargine therapy have also been studied.

In Adults with Type I Diabetes

Randomised Comparisons

At the time of the previous review in 2001,[8] six randomised trials involving more than 250 adult patients with type 1 diabetes,[4145,47] all of which compared insulin glargine (once daily at bedtime) with NPH insulin (once or twice daily), had been published or presented at meetings (table III). NPH has been the most frequent comparator to date because of its widespread use: it is estimated that this formulation accounts for over 80% of currently prescribed basal insulin therapy in the UK.[6] In all these studies, which were carried out over 4–28 weeks, patients continued to use the regular mealtime insulin regimen to which they were accustomed, and dosages of basal insulin (glargine or NPH) were based on FBG targets ranging from around 4–7 mmol/L. Patients’ ages ranged from 18–70 years and the duration of diabetes from 1–48 years, and body mass index (BMI) overall was around 25 kg/m2 (note that the ADA states that, as a broad guideline, individuals with a BMI of 25 kg/m2 should be considered as overweight, although this may need adjustment for some ethnic groups[38]). Baseline levels of FPG and FBG were approximately 11–14 mmol/L and 8–10 mmol/L, respectively, and HbA1c levels ranged from 7.6–8.1% in these previously reviewed studies.
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab3.jpg
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab3A.jpg
Table III

Randomised, parallel-group trials comparing insulin glargine (IG) with neutral protamine Hagedorn (NPH) insulin in adults with type 1 diabetes mellitus. Regimens shown are basal insulins given subcutaneously; these were used in conjunction with subcutaneous mealtime insulin in all cases

In the above studies,[4145,47] patients receiving insulin glargine showed greater reductions in mean FBG levels at study endpoint than NPH insulin recipients, with significant reductions from baseline in the insulin glargine groups (but not the NPH insulin group) in one[42] of these trials (table III). These improvements were evident from as early as week 1 of study therapy and were sustained throughout the treatment periods. In one of the 28-week trials,[47] the mean FBG level was decreased to a statistically significantly (p < 0.05) greater extent with insulin glargine than with twice-daily NPH insulin, but to a similar extent to once-daily NPH insulin. In addition, variability in self-monitored FBG levels was reduced to a greater extent in patients receiving insulin glargine than in those receiving NPH insulin in a 16-week study.[44] Mean HbA1c levels were reduced to a similar extent with both types of insulin in all 16- and 28-week studies (table III).[41,44,45,47]

Substantial percentages of patients (40–100%) reported at least one episode of symptomatic hypoglycaemia with both types of insulin in all studies (table III);[4145,47] in two 28-week trials the incidence of symptomatic hypoglycaemia was greater (p < 0.05) with NPH insulin than with insulin glargine.[41,45] The incidence was higher (p = 0.03) with insulin glargine in one other trial;[43] however, this study was of 4 weeks’ duration only. Percentages of patients reporting at least one episode of nocturnal hypoglycaemia were generally higher in the NPH insulin groups than in the insulin glargine groups in most previously reviewed trials;[42,44,45,47] this difference was statistically significant in two analyses.[42,45]

Patients with type 1 diabetes who received insulin glargine reported improved levels of satisfaction with treatment over 28 weeks in an analysis of 517 patients participating in a randomised, controlled European trial.[52] NPH insulin therapy was associated with a slight reduction in overall level of satisfaction with treatment, however. Patients were assessed with the Diabetes Treatment Satisfaction Questionnaire status version (DTSQ), which measures treatment satisfaction (six items), perceived frequency of hyperglycaemia (one item) and perceived frequency of hypoglycaemia (one item),[53] and the Well-Being Questionnaire (W-BQ) which measures general wellbeing in terms of 22 items, with four subscales to measure depression, anxiety, energy and positive wellbeing.[54] This analysis, which was reported as being ‘in press’ at the time of the last evaluation of insulin glargine in Drugs,[8] has now been published fully.[52]

Since the publication of the last review,[8] data from several further randomised comparisons of insulin glargine with NPH insulin have been reported (table III).[46,4851] Study durations ranged from 28–52 weeks. In two of the studies, mean BMIs were 24.9[46] and 26.6 kg/m2,[49] and mean age was around 40 years. These details were not available for the other trials.[48,50] Mean durations of diabetes were 15.5[46] and 17.5 years[49] in the two studies for which these details were reported.

In the largest study (n = 585), carried out by the European Insulin Glargine Study Group over 28 weeks,[46] all patients continued with their usual short-acting mealtime insulin after randomisation to insulin glargine or NPH insulin. Mean baseline HbA1c (7.9% and 8.0% for insulin glargine and NPH insulin, respectively) and FBG (9.3 and 9.2 mmol/L) levels were similar in each group. The 125 patients in the study carried out by the Australian Lantus Investigators[49] had poor glycaemic control, with mean FBG levels of at least 11 mmol/L at baseline in both treatment groups. A statistically significant difference between groups in baseline HbA1c levels was found in this study (9.18% for insulin glargine vs 9.72% for NPH insulin; p < 0.02), and results were analysed in terms of least square means from an analysis of covariance with baseline HbA1c level as a covariate. These poorly controlled patients were treated intensively in this trial, with a target FBG level of less than 5.5 mmol/L being set. Insulin lispro was given at mealtimes. Intensive treatment was also used in a 1-year Italian study in 121 patients,[50] although these individuals were already on long-term intensive therapy (baseline HbA1c level of 7.1% in both groups) with insulin lispro and NPH insulin mixtures at mealtimes and NPH insulin alone at bedtime. Patients randomised to insulin glargine continued to receive insulin lispro (without NPH insulin) at mealtimes, but with the addition of insulin glargine at dinnertime.

Although data were not reported in the abstract presented, DTSQ results from this trial showed greater treatment satisfaction with insulin glargine-based therapy than with NPH insulin (p = 0.001).[50] These findings are in accordance with previously reported European data[52] summarised earlier.

Differences in the ways in which results were reported, together with a scarcity of details, hinder the interpretation of these studies at this time, but the findings reported to date are nevertheless broadly in accordance with the randomised trials reviewed earlier.[8] The large European study[46] reported trends in favour of insulin glargine treatment in terms of reduction in FBG and HbA1c levels over 28 weeks from baseline, although statistical significance was not attained (table III). There was no increase in risk of hypoglycaemia with insulin glargine over NPH insulin. Interestingly, a substantial decrease in mean FBG level from baseline relative to that seen with NPH insulin (1.42 vs 0.81 mmol/L) was noted for insulin glargine in the subgroup of patients who had received twice-daily basal insulin before randomisation. No statistical analysis was reported, but the difference was stated to be clinically relevant. This observation concurs with previously reviewed results obtained by Rosenstock et al.,[43] and it has been suggested that a higher tolerance to insulin in such patients allows the use of higher dosages of insulin glargine without the onset of severe hypoglycaemia.[8] However, details of insulin dosages used were not reported in the abstract.

In contrast to the above study and the large trials previously reviewed in Drugs,[8] the smaller studies of intensive insulin therapy presented in this update indicate improvements relative to NPH insulin in long-term overall glycaemic control as indicated by HbA1c levels in patients randomised to treatment with insulin glargine (table III).[4851] The Australian investigators showed statistically significant differences between treatments that favoured insulin glargine in both FBG and HbA1c levels in terms of final readings expressed as least square means at study endpoint (table III).[49] The Italian study also showed statistically significant differences in favour of insulin glargine in these variables when mean differences between treatments were analysed (table III).[50] Similarly, in a crossover study, recipients of insulin glargine had significantly lower mean HbA1c levels after 32 weeks than those receiving NPH insulin (table III).[48] The Australian investigators reported reduced frequency of moderate or severe episodes of nocturnal hypoglycaemia with insulin glargine relative to NPH insulin, but did not give further details or definitions.[49] The Italian group, however, showed statistically significant reductions with insulin glargine over NPH insulin in incidences of overall and nocturnal hypoglycaemia defined as a blood glucose level of below 4 mmol/L (table III).[50] These were accompanied by a higher mean total insulin dosage in insulin glargine patients than in the NPH insulin group (0.69 vs 0.67 U/kg/day; p < 0.05). These data were supported by the other small studies conducted in Europe (table III).[48,51]

An additional multicentre, randomised and parallel-group study, carried out in 378 German patients aged 18–68 years and with baseline HbA1c levels ranging from 6.5–9.8%, has shown that glycaemic control with insulin glargine is not affected by the time at which a single daily injection is given.[55] These individuals were transferred from conventional basal therapy (typically NPH insulin) to treatment with insulin glargine given at breakfast, dinner or bedtime for 24 weeks, after which adjusted mean HbA1c levels were reduced from 7.6% (all groups) to 7.4%, 7.5% and 7.5%, respectively. Symptoms of nocturnal hypoglycaemia were noted in significantly (p = 0.005) fewer patients in the breakfast group (59.5%) than in the dinner (71.9%) or bedtime (77.5%) groups.[55] Furthermore, in 275 evaluable patients, mean DTSQ treatment satisfaction scores (particularly those for convenience and ‘wish to continue’) were improved in all three groups regardless of when insulin glargine was administered.[56] At study end, overall mean DTSQ scores had increased from baseline (scores ranged from 27.4–28.1) by 1.4, 2.5 (p = 0.0002 vs baseline) and 1.8 (p = 0.009), respectively, with breakfast, dinner or bedtime administration.[56] Corresponding changes in mean DTSQ convenience scores were 0.5, 0.8 and 0.7 (p ≤ 0.005 all comparisons vs baseline), with scores for ‘wish to continue’ improving by 0.4, 0.8 (p = 0.0001) and 0.6 (p = 0.0007).

Results from the extension phase of a multinational randomised comparative trial show that the antihyperglycaemic efficacy of insulin glargine is maintained in the long term in adult patients with type 1 diabetes.[57] Data obtained from 218 individuals (mean age 39.5 years; mean BMI 24.64 kg/m2) treated with insulin glargine at bedtime (with regular human insulin at mealtimes) who were followed for up to 36 months show a further reduction in mean HbA1c level from 8.2% at the end of the controlled study period to 7.79%. A 0.75kg increase in mean bodyweight was described as minimal. There were 124 episodes of hypoglycaemia, of which only 11 were described as severe.

Basal therapy with insulin glargine provided significantly greater improvements in HbA1c levels than NPH insulin in a retrospective analysis of ≈500 patients with type 1 diabetes who had received at least 6 months’ treatment (available as an abstract).[58] Mean HbA1c levels decreased from 8.1% at baseline to 7.7% at study end (p < 0.01 vs baseline) in the insulin glargine group versus a reduction from 8.2% to 8.0% in the NPH insulin group. There was also a significant (p < 0.01) between-group difference in HbA1c levels favouring insulin glargine at study end; mean duration of treatment in the insulin glargine group was 11.5 versus 13.2 months in the NPH insulin group (not statistically different).[58]

Other Studies

Other recent trials of insulin glargine in type 1 diabetes have focused on the effect of change of treatment on the frequency of hypoglycaemia[59] and on comparison with CSII.[6062] As with recent comparisons with NPH insulin, all results were available as abstracts or letters to journals at the time of this update.

Kelly et al.[59] have carried out a retrospective review of outpatient clinic experience in 80 patients, 64 of whom had type 1 diabetes. These individuals were aged from 15–83 years, had durations of disease ranging from less than 10 to over 30 years, and had used insulin glargine for at least 3 months. Previous basal insulin therapy included NPH and/or lente or ultralente. Most patients (51) were using insulin lispro for mealtime glycaemic control. Over the period of observation (3 months in 80 patients and 6 months in 30), a nonsignificant trend towards HbA1c improvement was noted (figures not reported) with no gain in bodyweight. However, a statistically significant decrease from baseline in the incidence of clinically significant (p = 0.002) hypoglycaemia (blood glucose level below 3.3 mmol/L) was apparent over the first 3 months (figure 3). There were also numerically fewer episodes of hypoglycaemia during the subsequent 3-month period, although this reduction was not significantly different from baseline (figure 3). There was an increase from baseline in frequency of self-monitoring of blood glucose levels after transition to insulin glargine, but this was not statistically significant (from 3.42 to 3.72 times daily at 3 months; p = 0.07).[59]
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Fig3.jpg
Fig. 3

Incidence of hypoglycaemia with insulin glargine treatment in patients with diabetes mellitus. Retrospective analysis of the incidence of clinically significant hypoglycaemia episodes (blood glucose <3.3 mmol/L) after 3 and 6 months’ basal treatment with insulin glargine in 64 outpatients with type 1 diabetes and 16 recipients with type 2 disease.[59] Previous basal treatment included neutral protamine Hagedorn and/or lente or ultralente insulin. * p = 0.002 vs baseline.

Three small studies have recently been carried out to compare basal insulin therapy with insulin glargine with CSII.[6062] The largest of these trials[62] was carried out over 1 year in 32 individuals with type 1 diabetes poorly controlled (HbA1c >8%) with NPH and regular or lispro insulin. The investigators concluded that both insulin glargine and CSII were superior in terms of glycaemic control and frequency of hypoglycaemia to the conventional multiple-dose regimens from which the patients were transferred (table IV). There were no significant differences between insulin glargine and CSII in any of the clinical endpoints analysed, and no relevant change in mean BMI with either treatment. Daily insulin requirements were, however, reduced from baseline to a considerably greater extent with CSII than with insulin glargine (by 10.3 vs 0.9U; p < 0.001). This study, which was reported in letter form, was carried out in a nonblind fashion in parallel patient groups, but the authors did not state what method of randomisation, if any, had been used.
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab4.jpg
Table IV

Clinical endpoints in a nonblind, parallel-group comparison of insulin glargine (IG) with continuous subcutaneous insulin infusion (CSII).[62] Patients had type 1 diabetes mellitus poorly controlled with previous conventional multiple-dose (MDI) treatment with neutral protamine Hagedorn insulin basal therapy

Data from the two remaining studies, one a randomised crossover trial in seven patients[61] and the other a prospective analysis in 19 individuals,[60] show conflicting results. The first-mentioned study[61] was conducted in patients already controlled with CSII therapy (mean HbA1c level at baseline = 6.6%), and involved partial replacement of basal insulin treatment with insulin glargine with a corresponding reduction in subcutaneous infusion rates. Measurement of the rate of metabolic decompensation for 360 minutes on day 28 of each treatment period suggested that this partial basal replacement conferred protection against ketosis, with no adverse effect on the stability of glycaemic control. The second study[60] involved patients recruited from an outpatient clinic and matched for HbA1c level, age, gender, BMI, frequency of hypoglycaemia and self-monitored blood glucose readings, eight of whom were receiving insulin glargine and 11 of whom were being treated with CSII. The authors stated that CSII conferred better control (in terms of glycaemic stability) than insulin glargine, as determined by 3-day glycaemic profiles obtained using a continuous glucose monitoring system and determined over a treatment period of several weeks or months.

In addition, two randomised, nonblind trials have compared insulin-containing multiple dose regimens with CSII using insulin aspart[63] or insulin lispro[64] in patients with type I diabetes (available as abstracts). In the first study, a multiple-dose injection regimen of insulin glargine at bedtime plus insulin aspart immediately before each meal was compared with CSII using insulin aspart in patients with type 1 diabetes previously treated with CSII.[63] Glucose exposure was significantly lower with CSII than with the multiple-dose regimen, as assessed by continuous subcutaneous glucose monitoring over a 48-hour period (2059 vs 2687 mg ⋅ h/dL; p < 0.01). There was no between-group difference in the percentage of patients experiencing hypoglycaemia (94% with the multiple-dose regimen vs 92%) or nocturnal hypoglycaemia (72% vs 73%). After a one week washout-period during which all patients received CSII with insulin aspart, 50 patients were randomised to receive a multiple-dose injection regimen (insulin glargine plus insulin aspart) and 50 to receive CSII with insulin aspart. After 5 weeks’ treatment, patients in each group were switched to the alternate regimen for a further 5 weeks’ treatment.[63] A second 7-day study (n = 38) also using continuous subcutaneous glucose monitoring demonstrated that insulin glargine (ratio 1 : 1 or 1 : 1.2 of total basal CSII daily dose) once daily plus bolus insulin lispro prior to meals effectively maintained glycaemic control in patients using CSII infusion, providing an opportunity to change from the use of CSII for short periods.[64] After switching to a multiple dose regimen, the incidence of hypoglycaemia was comparable to that observed at baseline, with no episodes of severe hypoglycaemia or hyperglycaemia leading to ketosis and no serious adverse events occurring.[64]

In Children and Adolescents with Type 1 Diabetes

Previously reviewed[8] data on insulin glargine in children were restricted to a single abstract from a large (n = 349), 28-week, randomised and parallel-group comparison with NPH insulin. This study has now been reported fully,[65] and results are shown together with those from another fully reported but smaller (n = 25) randomised trial[66] in table V. Note that the latter study[66] was carried out according to a crossover design, and included adolescents only. The large parallel-group study[65] included a run-in period of up to 4 weeks, and bedtime doses of insulin glargine were titrated to achieve a target FBG level of 4.4–8.8 mmol/L. Patients continued with the regular mealtime insulins to which they were accustomed. In the crossover study,[66] patients received insulin lispro at mealtimes while receiving basal insulin glargine, whereas regular human insulin was used while on NPH insulin. Each treatment period lasted for 16 weeks, and the target FBG was 4.5–9.0 mmol/L.
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab5.jpg
Table V

Randomised trials comparing insulin glargine (IG) with neutral protamine Hagedorn (NPH) insulin in children and adolescents with type 1 diabetes mellitus. Regimens shown are basal insulins given subcutaneously; these were used in conjunction with subcutaneous mealtime insulin in all cases

Both trials showed greater antihyperglycaemic efficacy of insulin glargine than NPH insulin in terms of effect on FBG levels, but not on levels of HbA1c (table V).[65,66] As well as laboratory-monitored FBG, self-monitored blood glucose levels at various times of the day were also presented in the crossover study of Murphy et al.[66] Insulin glargine-based treatment was associated with significantly lower levels than NPH insulin 2 hours after breakfast (8.1 vs 10.7 mmol/L; p < 0.0005), before lunch (8.9 vs 10.1 mmol/L; p < 0.01), and 2 hours after lunch (8.0 vs 9.5 mmol/L; p < 0.002); there was no difference between groups in blood glucose levels either before or after dinner. Overnight symptomatic hypoglycaemia was noted in two patients only in this trial (once in each treatment period), and nocturnal hypoglycaemia was observed during 8 of 25 nights (32%) with insulin glargine and 15 of 25 nights (56%) with NPH insulin. The difference was significant (p < 0.05) when analysed by unpaired χ2 test.[66] Although statistical significance was not attained, there were strong trends towards lower incidences of severe overall and nocturnal hypoglycaemia in favour of insulin glargine in the large parallel-group study of Schober et al.[65]

Glycaemic control was maintained with insulin glargine-based therapy for up to 36 months in a long-term, nonblind, multicentre extension study in 129 evaluable children and adolescents.[67] The mean HbA1c level increased slightly from a baseline value of 8.65% to 9.0%.

A nonrandomised prospective trial in children and adolescents aged 5 years and over, reported as an abstract, has compared insulin glargine given once daily in the evening (n = 67) with NPH insulin given once, twice or three times daily (n = 62).[68] Respective mean durations of treatment were 10.9 and 11.7 months. HbA1c levels increased slightly from baseline with both treatments (from 7.3% to 7.7% with insulin glargine and from 7.6% to 7.9% with NPH insulin). There was a trend towards less symptomatic hypoglycaemia with insulin glargine (no data or statistical analysis reported), and the incidence of severe hypoglycaemia was slightly lower in the insulin glargine than in the NPH insulin group (0.31 vs 0.36 episodes/patient-year). Perceived quality of life, which appeared to be based on patient preference, was stated to be better with insulin glargine than with NPH insulin, although details of the quality-of-life instrument used were not given.[68]

Addition of insulin glargine at bedtime to the existing insulin lispro plus NPH insulin (given in the morning) regimen of 114 children (mean age 12.6 years) with type 1 diabetes significantly improved glycaemic control in a noncomparative study (available as an abstract).[69] Relative to baseline (i.e. during the 9 months prior to addition of insulin glargine), the mean HbA1c level (9.6% vs 9.3% after 9 months treatment) and daily insulin dose (1.02 vs 0.96 U/kg) were significantly (p < 0.01 both comparisons) reduced after 9 months’ treatment, as was the number of non-severe hypoglycaemic episodes (2 vs 1.3 episodes/week; p < 0.001; defined as blood glucose levels <3.3 mmol/L). Furthermore, the overall number (22 vs 9 episodes/9 months) and the number of nocturnal (14 vs 4) severe hypoglycaemia episodes were also reduced, although the frequency of these episodes was too low to permit statistical analysis for this number of patients. The mean duration of diabetes in these children was 5.4 years, with all patients having had the disease for at least 1 year.[69]

Transition from conventional combinations of short- and intermediate-acting insulin to treatment based on insulin glargine resulted in a reduction in frequency of episodes of severe hypoglycaemia despite intensification of insulin treatment in a retrospective study in 140 patients aged from 2–21 years.[70] Most of these individuals were switched to insulin glargine because of poor glycaemic control with their previous treatment (HbA1c ≥ 8.5%) and/or because of problems with hypoglycaemia requiring assistance. After 3 months, the mean HbA1c level was reduced only slightly (from 9.1% to 8.9%; p = 0.4), but the number of severe hypoglycaemic episodes requiring assistance was reduced from 11 in the 3 months preceding insulin glargine therapy to three during the 3-month study period. No statistical analysis was reported for this latter endpoint.[70]

Replacement of NPH or lente insulin given two to four times daily by insulin glargine once daily reduced the frequency of asymptomatic nocturnal hypoglycaemia detected by continuous subcutaneous glucose monitoring in a further study in 30 patients aged from 4.5–18.3 years.[71] The proportion of patients showing asymptomatic nocturnal hypoglycaemia (blood glucose level <3.3 mmol/L) was reduced by 45% (p = 0.002) 4–8 weeks after switching to basal insulin glargine therapy. There was no significant change from baseline in mean HbA1c level or in the mean total daily basal or mealtime insulin dose. Both this study and the retrospective trial discussed above[70] were available as abstracts only.

In Patients with Type 2 Diabetes

Type 2 diabetes can be managed with diet and exercise, at least in its early stages. However, the condition is progressive, and nearly all patients will require oral antidiabetic treatment after a time. Further progression leads to the need for exogenous insulin to maintain satisfactory glycaemic control in most patients, and current guidelines recommend a ‘step-up’ approach to treatment as the patient proceeds through these stages in the development of the disease.[6] In patients with type 2 diabetes, a basal insulin supplement is often given at bedtime while oral therapy is continued; alternatively, oral agents may be discontinued and a premixed insulin be given before breakfast and dinner, or a basal-bolus regimen be implemented.[72]

Randomised Comparisons with NPH Insulin

The previous review[8] described six large (each in over 100 patients) randomised comparisons of insulin glargine with NPH insulin in patients with type 2 diabetes (table VI).[74,7678,80,81] These trials ranged in duration from 4 weeks to 1 year, and two[80,81] were reported in full. Most studies recruited patients who were not dependent on insulin and who were being treated with oral antidiabetic agents; two trials, however, recruited patients who were already receiving insulin therapy alone.[74,80] Studies of longer duration (28 weeks to 1 year) were also of parallel-group design,[74,76,80,81] and all trials were carried out in a nonblind manner (similar to those conducted in patients with type 1 diabetes). In addition, as reviewed previously,[8] treatment satisfaction (assessed using DTSQ) and patients’ wellbeing (W-BQ) were assessed in patients enrolled in one[81] of these trials (available as an abstract[82]).
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Tab6.jpg
Table VI

Randomised, nonblind, parallel-group, multicentre (except one single-centre[73] ) trials comparing insulin glargine (IG) with neutral protamine Hagedorn (NPH) insulin in patients with type 2 diabetes mellitus. Regimens shown are basal insulins given subcutaneously in conjunction with oral antihyperglycaemic agents

As would be expected for type 2 disease, patients recruited ranged in age from 40–80 years, and had higher bodyweights overall than the individuals followed in the studies in type 1 diabetes (mean BMIs of 27–29 kg/m2). HbA1c levels ranged from 7–12% and patients had had diabetes for at least 3 years. In most of these previously reported studies, basal insulin was given once daily at bedtime, and changes in insulin requirements throughout study periods were similar for insulin glargine and NPH insulin in two trials.[74,81] A reduction in insulin requirement was noted in a single study in patients who were transferred to insulin glargine from conventional basal regimens given more than once daily.[74]

Most trials (including all those of 28 weeks’ to 1 year’s duration) showed similar reductions from baseline in mean HbA1c levels with either basal insulin regimen (from between 0.35–0.8%) [table VI].[74,76,77,80,81] In addition, the single 4-week study that reported FPG levels showed similar improvements in glycaemic control with insulin glargine and NPH insulin (table VI).[78] However, in most trials the incidence of nocturnal hypoglycaemia (proportion of patients reporting at least one episode) was statistically significantly lower with insulin glargine (10–31.3%) than with NPH insulin (24–40.2%).[74,76,80,81] Mean overall incidences of symptomatic hypoglycaemia were also lower with insulin glargine (7.3–61.4%) than with NPH insulin (19.1–66.8%), although the differences were not as great as for nocturnal hypoglycaemia and were statistically significant in three trials (table VI).[74,77,81]

More recent studies include two further large, randomised comparisons of insulin glargine with NPH insulin (table VI).[75,79] Both studies recruited patients poorly controlled by oral antihyperglycaemic therapy. As might be expected in a study in this type of patient, mean BMI values showed a tendency for patients to be overweight, particularly in one of the trials (mean BMIs at baseline were 28[75] and 32[79] kg/m2). In addition, patients were predominantly middle-aged (mean ages 60.8[75] and 55[79] years). Targets for glycaemic control were set in both trials, with Fritsche et al.[75] specifying an FBG goal of below 5.5 mmol/L and target HbA1c level of below 8%. Riddle and Rosenstock[79] specified a target FPG level of 5.56 mmol/L or less, but with tighter control in terms of HbA1c levels (target ≤7%) than in the other study. Across both studies, patients had had type 2 diabetes for 8–10 years. In one of the trials,[75] patients on insulin glargine were assigned to breakfast or bedtime injections. Both studies combined basal insulin treatment with oral antihyperglycaemic therapy: Fritsche et al.[75] gave all patients single daily doses of the sulphonylurea glimepiride, whereas Riddle and Rosenstock[79] continued previous oral treatment.

Similar to results of previously reviewed randomised trials,[8] Riddle and Rosenstock found no difference between bedtime insulin glargine and NPH insulin in terms of effect on glycaemic control, but did note that patients on insulin glargine were able to achieve glycaemic improvements over baseline with statistically significantly less risk of nocturnal hypoglycaemia than patients randomised to NPH insulin (table VI).[79] Of particular note were the proportions of patients who achieved the target HbA1c level without experiencing any episodes of nocturnal hypoglycaemia (33% with insulin glargine vs 27% with NPH insulin; p < 0.05).

The findings of Fritsche et al.[75] are of particular interest because they show a greater improvement in glycaemic control in terms of glycosylated haemoglobin levels in insulin glargine recipients than was seen in patients receiving NPH insulin (table VI). It was found that insulin glargine reduced the mean HbA1c level over 24 weeks to a statistically significantly greater extent when given before breakfast (by 1.24%) than when given at bedtime (0.96%), and that breakfast administration only (not bedtime administration) resulted in statistically significantly better control than NPH insulin (reduction of 0.84%) [table VI]. Both insulin glargine groups showed significantly lower incidences of nocturnal hypoglycaemia than the NPH insulin group (table VI).

These results should be compared and contrasted with the German findings (reported earlier; section 4.1.1) in patients with type 1 diabetes,[55] which showed no effect of injection timing on glycaemic control but did demonstrate an advantage in terms of nocturnal hypoglycaemia for insulin glargine given at breakfast rather than at lunch or dinner. The data from this German study are supported by a large (n = 624), 28-week, randomised, nonblind, multicentre study in patients with type 2 diabetes (available as an abstract).[83] Patients with poorly controlled glucose levels whilst receiving oral glimepiride were also given insulin glargine either in the morning or at bedtime. There was a similar incidence of nocturnal hypoglycaemia in the morning and bedtime group (13% vs 14.9%; on-treatment analysis; 95% CI −100%, 2.84%), with 51.3% and 54.8% of patients in the morning and bedtime groups experiencing only one episode during the treatment period. The corresponding incidence of symptomatic (42.8% vs 38.1%) and severe (1.3% vs 0.7%) hypoglycaemia episodes was similar in both groups. The decrease from baseline in mean HbA1c was also comparable in the morning (8.82% vs 7.18%) and bedtime (8.81% vs 7.23%) groups.[83]

A recent meta-analysis of 2304 patients with type 2 diabetes enrolled in phase III clinical trials confirmed the significantly lower risk of hypoglycaemic episodes with insulin glargine treatment compared with that of NPH insulin (available as an abstract) [figure 4].[84] The incidence of all symptomatic and symptomatic nocturnal hypoglycaemia episodes, both confirmed and unconfirmed, was significantly (all p < 0.001) lower in the insulin glargine than the NPH insulin group, with a 29% risk reduction of confirmed nocturnal hypoglycaemia in the insulin glargine group (figure 4).[84] In terms of mean HbA1c levels, both treatment groups achieved identical reductions over the study period (1% decrease from baseline levels; mean baseline HbA1c level of 8.8% in the insulin glargine group vs 8.7% in the NPH insulin group).[84] Patients received insulin glargine once daily or NPH insulin once or twice daily for 16–28 weeks (52 weeks’ treatment in one study, with interim 20-week data being used).
https://static-content.springer.com/image/art%3A10.2165%2F00003495-200363160-00007/MediaObjects/40265_2012_63161743_Fig4.jpg
Fig. 4

Comparative incidence of hypoglycaemia with insulin glargine (IG) or neutral protamine Hagedorn (NPH) insulin treatment. Meta-analysis of patients with type 2 diabetes mellitus enrolled in phase III clinical trials of 16–28 weeks’ duration, with one being of 52 weeks’ duration (interim 20-week data were used from this trial) [abstract presentation].[84] Patients received insulin glargine once daily (n = 1142) or NPH insulin once or twice daily (n = 1162). The number in brackets above each set of data is the insulin glargine percentage risk reduction. con = confirmed; noc = nocturnal; PG = plasma glucose; sym = symptomatic; * p < 0.001, ** p < 0.0001 vs NPH insulin.

In another meta-analysis, once-daily insulin glargine was predicted to be associated with a clinically relevant (not defined in abstract) and statistically significant reduction in HbA1c compared with once-daily NPH insulin when the incidence of confirmed symptomatic hypoglycaemia or confirmed nocturnal hypoglycaemia were made equivalent in the two treatment groups.[85] In insulin glargine recipients, the overall incidence of confirmed symptomatic hypoglycaemia was reduced by 23% (mean 420 vs 552 episodes/100 patient-years; p = 0.015) and that of confirmed nocturnal hypoglycaemia by 39% (mean 142 vs 238 episodes/100 patient-years; p < 0.001) compared with NPH insulin. A Poisson regression analysis predicted that HbA1c levels would be 0.44% lower with insulin glargine treatment at a given rate of confirmed symptomatic hypoglycaemia and that this would be statistically significant over the range of 100–2000 episodes/100 patient-years (p values not reported). With a given incidence of confirmed nocturnal hypoglycaemia, mean HbA1c levels were predicted to be 0.87% lower and would be statistically significant between a mean of 15–255 episodes/100 patient-years (p values not reported). This meta-analysis evaluated data from 1786 patients with type 2 diabetes enrolled in a phase III and two phase IV trials.[85]

Long-term data have become available from the extension phase of a randomised, multicentre, nonblind comparative study, in which 239 patients continued to receive individually titrated basal therapy with bedtime insulin glargine in combination with oral antihyperglycaemic therapy.[86] The mean HbA1c level was reduced over a period of up to 39 months from 9.44% at baseline to 8.42%, with 23 episodes of hypoglycaemia reported in 5.4% of patients. Only two of these episodes were severely symptomatic. No details regarding the duration of previous treatment were reported in the abstract presentation; nor was the type of oral antihyperglycaemic therapy clearly indicated, although some patients were on glibenclamide (glyburide) or gliclazide.[86]

Additionally, preliminary 18-month data are available from a single-centre, nonblind trial in 103 evaluable patients with type 2 diabetes who were receiving a multiple injection insulin regimen (table VI).[73] At study end, mean HbA1c levels in the insulin glargine group were significantly reduced relative to baseline (7.4% reduced to 7%; p < 0.003), whereas the reduction in the NPH insulin was not significant (7.4% to 7.2%; p < 0.06). There were no significant increases in the daily insulin doses in either the insulin glargine (0.28 IU/kg at baseline vs 0.29 IU/kg final dose) or NPH insulin (0.35 vs 0.38 IU/kg) group, nor did patients in either treatment group experience a significant increase in BMI. Notably, there were significantly fewer episodes of mild hypoglycaemia in the insulin glargine (not defined, nor were data reported in the abstract). No episodes of severe hypoglycaemia were reported during the study.[73]

As in patients with type 1 disease (section 4.1.1), DTSQ results showed greater treatment satisfaction with insulin glargine-based therapy than with NPH insulin in patients with type 2 diabetes in a large, randomised, multicentre trial.[82] The improvement in treatment satisfaction was significantly greater with insulin glargine than with NPH insulin at week 36 (p = 0.03 vs NPH insulin), with a trend to better treatment satisfaction in the insulin glargine group at study end (p = 0.06) [no data reported in abstract].[82] Furthermore, within each treatment group, treatment satisfaction increased over time for all timepoints (p < 0.01 vs baseline for all comparisons; no data reported in abstract). In addition, insulin-experienced patients (n = 114) who were randomised to insulin glargine showed a greater increase in treatment satisfaction from baseline than those randomised to NPH insulin. There was no change over time in W-BQ scores within each treatment group, nor were there any between-group differences in these scores. The questionnaires were conducted at baseline and at weeks 8, 20, 36 and 52.

Analysis of a recent multicentre trial[75] discussed above (table VI) indicated that the investigator’s most common reason for not adjusting insulin doses to achieve the specified target FBG level was fear of overall hypoglycaemia (29% of investigators) and nocturnal hypoglycaemia (15%).[87] A comparable percentage of patients receiving morning insulin glargine (18.2%), bedtime insulin glargine (26%) or NPH insulin (25.4%) achieved target FBG levels during this study. At study end, significantly (p ≤ 0.002 both comparisons) fewer patients in the morning insulin glargine group (9%) than in the bedtime insulin glargine (16%) or NPH insulin group (19%) were not titrating their insulin dose because of fear of nocturnal hypoglycaemia.[87]

Other Studies

Two nonrandomised community-based studies and a retrospective analysis in patients with type type 1 or type 2 diabetes who were prescribed insulin glargine have been reported (all as abstracts) in addition to the above randomised studies since the previous review of this agent in Drugs.[8] The nonrandomised studies involved patients with either type 1 or type 2 diabetes who were followed for 6[88] or 9[89] months after initiation of insulin glargine therapy, whereas the retrospective analysis involved patients with end-stage renal disease.[90] In addition, 18-month data from 60 patients with type 2 diabetes enrolled in one[89] of the nonrandomised trials is also available as an abstract.[91]

Significant reductions in mean HbA1c levels after treatment with insulin glargine were reported in patients with type 2 diabetes in all these studies. In one,[89] addition of insulin glargine to established oral antihyperglycaemic therapy resulted in a reduction in this variable from 8.1% to 7.6% after 9 months in 60 individuals. Similarly, 6-month data showed that mean HbA1c levels were reduced by 0.79% (p < 0.05) in 208 patients transferred from other bedtime insulin regimens, and by 1.7% (p < 0.01) in 14 patients previously treated with oral antihyperglycaemic therapy only.[88] At 18 months, mean HbA1c levels had been reduced by 1.2% from baseline (p < 0.0003).[91]

Use of insulin glargine was associated with a decrease in mean HbA1c level from 7.7% to 6.8% (p = 0.005), with no episodes of severe hypoglycaemia, over a mean 9.9 months in 20 patients undergoing haemodialysis for end-stage renal disease.[90] Of these individuals, the majority (16) had type 2 diabetes, 19 had previously received conventional or intensified conventional NPH insulin therapy, and one had received oral antihyperglycaemic agents alone.

Patients with type 2 diabetes receiving insulin glargine therapy were also evaluated in two other nonrandomised studies;[92,93] at least one[92] of which was community-based. In 72 individuals, 32 of whom had been previously treated with oral antihyperglycaemic agents and 40 with insulin, mean HbA1c decreased from 8.14% to 7.58% (p < 0.001) over 9 months after being transferred to once-daily insulin glargine plus oral therapy with either glimepiride or glimepiride plus metformin (patients previously on oral therapy) or once-daily insulin glargine with mealtime regular or lispro insulin (patients previously on insulin).[92] A further 42 patients not responding adequately to oral antihyperglycaemic therapy showed a reduction in mean HbA1c level from 9.8% to 8.2% (p < 0.001) over 12 weeks after starting treatment with insulin glargine at bedtime and oral therapy with metformin 850mg twice daily.[93] There were no reports of hypoglycaemia.[93]

In patients with type 2 diabetes poorly controlled by conventional insulin therapy, switching to insulin glargine plus oral antihyperglycaemic agents reduced the daily insulin dose, with a trend to lower HbA1c and FPG levels, in a 16-week, randomised, nonblind, parallel-group trial (available as an abstract).[94] Seventeen patients continued with conventional pre-mixed insulin twice daily, 17 were switched to insulin glargine in the morning plus glimepiride once daily and 18 received the same insulin glargine plus glimepiride regimen and metformin twice daily; the baseline mean HbA1c level was 8.4% and the mean duration of diabetes and of insulin therapy was 15.5 and 4.2 years. There were no significant between-group differences in glycaemic control at study endpoint. Recipients of insulin glargine-containing dual (mean HbA1c reduced by 0.4%) or triple (0.7% reduction) therapy regimens achieved significant (p < 0.05 both comparisons) reductions from baseline in mean HbA1c levels over the study period, with those continuing with existing therapy having a 0.3% reduction in mean HbA1c level. Similarly there was a trend for greater reduction in mean FPG values with dual or triple therapy relative to conventional insulin therapy (reduced by 0.9, 1.9 and 0.05 mmol/L, respectively). Compared with baseline, daily insulin doses were reduced in the insulin glargine triple (−15.5 IU/day; p = 0.0005) and dual (−4.5 IU/day; not significant) therapy groups, whereas there was a significant increase in those continuing with conventional insulin therapy (+5.4 IU/day; p < 0.05).[94]

In addition, the efficacy of low-dose insulin glargine in patients with impaired fasting glucose, impaired glucose tolerance or early type 2 diabetes has been evaluated in a phase I, randomised, double-blind, placebo-controlled trial (available as an abstract).[95] Patients were on restricted diet (25 kcal/kg/day) and received insulin glargine 4 IU/day (n = 15) or placebo (n = 4) at bedtime for 12 days, with the dose titrated to achieve a fasting blood glucose of ≤ 5.3 mmol/L. Mean FBG (5.4 mmol/L reduced to 4.8 mmol/L) and mean daylong blood glucose (0.45 mmol/L reduction) decreased in the insulin glargine group. In contrast, mean FBG (5.8 mmol/L increased to 6.2 mmol/L) and daylong blood glucose levels (0.44 mmol/L increase) increased in the placebo group.[95] These preliminary data require confirmation in larger studies.

Effect on Bodyweight

The achievement of good glycaemic control without excessive gain in bodyweight when insulin therapy is prescribed is a challenge for those managing patients with type 2 diabetes. Previously reviewed data[8] suggested a potential advantage for insulin glargine over NPH insulin in this respect, with 28-week results from one trial in 518 patients with type 2 diabetes showing less weight gain in insulin glargine than in NPH insulin recipients (mean 0.4 vs 1.4kg; p = 0.0007).[80] However, 1-year data showed similar gains in mean bodyweight in 214 patients receiving insulin glargine (2.6kg) and 208 NPH insulin recipients (2.3kg).[81]

Recent randomised studies in individuals with type 2 diabetes[75,79] have not reported details of any effect of randomised treatment on bodyweight to date. However, long-term results also reported in this update have shown a gain in mean bodyweight of 2.02kg over up to 39 months in 239 patients being treated with insulin glargine in combination with oral antihyperglycaemic therapy.[86] This gain accompanied an absolute fall in mean HbA1c level of 1.02% (section 4.3.1). Improvements in mean HbA1c levels were not accompanied by any statistically significant bodyweight changes in prospective case series reviewed above (section 4.3.2).[88,89,9193] One of these analyses[93] involved 42 obese patients (mean baseline BMI = 33.2 kg/m2) who showed an increase in mean bodyweight of only 0.2kg over 12 weeks despite statistically and clinically significant improvements in HbA1c and FBG levels after commencement of insulin glargine therapy. Small mean bodyweight reductions (from 94.9 to 93.8kg) were reported over 9 months in another two of these studies in a total of 132 patients with type 2 diabetes.[89,92] Patients in one[89] of these studies (n = 60) experienced a nonsignificant reduction of 8.1kg over the entire 18-month insulin glargine-treatment period.[91]

Tolerability

The tolerability profile of insulin glargine was reviewed in detail previously in Drugs,[8] and is summarised in literature published by the manufacturer[9] and by the European Agency for the Evaluation of Medicinal Products.[13] A brief overview is presented here together with pertinent details from recent studies. Note that hypoglycaemia and bodyweight gain may be considered as adverse events of any insulin treatment, but their clinical significance is such that they have been treated as efficacy endpoints in clinical trials. These effects have therefore been discussed in the previous section.

In general, the incidence of adverse events reported in patients receiving insulin glargine in randomised clinical trials has been similar to that seen in NPH insulin recipients. Injection site reactions (redness, pain, itching, hives, swelling or inflammation) are the most common adverse events, and are seen in some 3–4% of patients who receive insulin glargine. Most reactions are minor, do not require discontinuation of therapy, and resolve within a few days or weeks.[13] Lipodystrophy, which may be seen with any insulin therapy, can be minimised or prevented by rotation of injection sites.[9,13] Of interest in this respect are observations in 239 patients with type 2 diabetes who participated in one of the long-term extension studies referred to earlier in this review (section 4.3.1):[86] no injection site reactions were noted with insulin glargine over treatment periods of up to 39 months.

Differences in structure between endogenous insulin and compounds formulated for medicinal use may lead to an increased risk of the development of insulin-binding antibodies.[96] Previously reviewed studies[8] showed no evidence of increases in this immunogenic response in patients receiving insulin glargine over those given NPH insulin. Indeed, in two 1-year studies,[76,81] increases in insulin antibody levels from baseline were higher with NPH insulin than with insulin glargine (p ≤ 0.0001). Recent data obtained in patients with type 1 or type 2 diabetes undergoing long-term therapy with insulin glargine have shown decreases[57,67] or very small and clinically irrelevant increases[86] in mean antibody titres over maximum treatment periods of at least 3 years. The clinical relevance of these observations remains unclear. No clinically relevant changes in titres of antibodies to Escherichia coli were reported in previously reviewed trials.[8] Increased titres were noted in 23–31% of patients participating in recent long-term studies,[57,67,86] but no relationship was found between these increases and the incidence of adverse events.

Following on from previously reviewed[8] evidence from 2207 patients treated for 28–52 weeks,[97] there continues to be no indication that insulin glargine is associated with increased risk of progression of diabetic retinopathy (related to the generally higher affinity of insulin glargine than regular human insulin for IGF-1 receptors). It should be noted that temporary alterations in turgidity and refractive index of the lens may accompany sudden changes in glycaemic control with any insulin treatment, and that intensification of therapy with abrupt improvements in glycaemic control may cause a temporary worsening of retinopathy.[13]

The effect of renal impairment on the pharmacokinetics of insulin glargine has not been studied specifically, but recent data from patients undergoing dialysis for end-stage renal disease who were treated with insulin glargine revealed no tolerability concerns.[90] No clinical data relating to the use of insulin glargine in pregnancy in humans are available,[9] but recent animal studies have shown no effect of the drug on fetal or postnatal development.[31] In addition, early concerns over potential mutagenicity related to altered interactions with IGF-1 receptors of insulin glargine (reviewed by McKeage and Goa[8] ) have not been borne out in histological studies. Recently published data obtained in rats and mice over up to 2 years show no evidence of carcinogenicity of insulin glargine, with no difference in incidence of mammary tumours between animals receiving insulin glargine (up to 5 U/kg/day in rats and 12.5 U/kg/day in mice), NPH insulin or normal saline.[27] The distribution of rodent-specific subcutaneous malignant fibrous histiocytoma was found not to be dose-dependent, and neuronal necrosis of the cerebrum in rats was attributed to persistent and repeated episodes of hypoglycaemia induced by high doses of insulin.

Dosage and Administration

Insulin glargine provides basal insulin levels after once-daily subcutaneous administration, and is indicated for the treatment of adults or children (aged >6 years) with type 1 diabetes and adults with type 2 diabetes.[9] The insulin glargine solution is clear and slightly acidic (pH 4.0) and should not be diluted or mixed with any other insulin or solution as this could alter its time-action profile.[9]

The dosage of insulin glargine should be determined individually for each patient based on blood glucose levels and any previous insulin therapy. In a clinical study of insulin-naive patients with type 2 diabetes receiving oral antidiabetic agents, insulin glargine was started with a dose of 10 IU once daily. Dosage was then adjusted based on the patient’s blood glucose levels and average doses ranged from 2–100 IU once daily.[9] In clinical studies of patients who were receiving once-daily NPH or ultralente insulin, the initial dose of insulin glargine was usually not changed, with patients receiving an equivalent number of international units of insulin glargine. However, when patients were previously treated with twice-daily NPH insulin, the dosage of insulin glargine was reduced by approximately 20% for the first week of treatment and then adjusted according to blood glucose levels.[9] Metabolic monitoring under medical supervision is recommended during transfer and the early weeks of therapy.[9] Paediatric patients should be managed in the same way.[9] The dosage and timing of additional short-acting insulin or oral antidiabetic agents may need to be adjusted. Dose adjustment of antihypoglycaemic agents may also be required should changes occur in the patient’s weight or lifestyle, and during intercurrent illness.[9]

As with all subcutaneous insulin injections, the site should be rotated within an area after each injection.[9] Similar absorption rates of insulin glargine were recorded after subcutaneous injection in the abdomen, thigh or deltoid regions (section 3.1). There was no evidence of accumulation of insulin glargine in patients receiving daily injections over 11 days (section 3.1).

Well controlled studies of the use of insulin glargine during pregnancy do not exist and the drug should be given to pregnant women only if clearly indicated.[9] Insulin glargine requirements may be reduced in patients with renal or hepatic impairment (section 3.3).[9] To avoid the risk of hypoglycaemia, elderly patients require cautious dose selection and titration.[9] Consideration should be given to any concomitant therapy that may affect glycaemic control (see section 3.4).

There have been some reports of errors resulting from confusion between the prescription of insulin glargine by its brand name Lantus™ and that of lente insulin.[98100] Care should be taken in prescribing and dispensing these agents, and measures taken to eliminate this confusion, for example, patient education and prescriptions written clearly without the use of abbreviations.[98]

Place of Insulin Glargine in the Management of Diabetes Mellitus

Good glycaemic control is a fundamental part of the management of diabetes. Large and authoritative prospective clinical trials, most notably the Diabetes Control and Complications Trial (DCCT) in type 1[101,102] and the United Kingdom Prospective Diabetes Study (UKPDS) in type 2 disease,[103105] have confirmed that improvements in glycaemic control are associated with sustained and clinically relevant reductions in rates of retinopathy, nephropathy and neuropathy. In these studies, treatment regimens that reduced mean HbA1c levels to around 7% (approximately 1% above the upper limit of normal) were associated with reduced frequencies of long-term microvascular complications, although intensive therapy was also found to increase the risk of severe hypoglycaemia and weight gain.[103,106] There are also data (obtained in patients with type 1 diabetes) to suggest that intensive insulin therapy stabilises macrovascular disease or prevents its progression in patients at risk.[106] Although the UKPDS showed no significant effect of the lowering of blood glucose levels on the incidence of cardiovascular complications in patients with type 2 disease,[104] epidemiological analysis showed a continuous association between risk of cardiovascular complications and glycaemia.[103]

The need for patients with diabetes to have access to treatments that induce a metabolic effect by imitating endogenous insulin secretion as closely as possible is therefore clear, but the injectable preparations available to date have been found wanting in this respect. Regular human insulins used to provide the postprandial ‘boost’ needed after meals cannot provide the almost immediate ‘on-demand’ insulin peaks of short duration that non-diabetic individuals experience after eating, and the use of a basal insulin (such as NPH) in conjunction with mealtime doses (or oral antihyperglycaemic agents in patients with type 2 disease) gives only a poor approximation of the normal physiological insulin profile, even in patients with excellent compliance (see review by Lindholm[107] and De Witt and Hirsch[108]). Problems associated with the excessively slow absorption and elimination of human regular insulin used at mealtimes have been addressed by the introduction of short-acting novel analogues such as insulin lispro and insulin aspart, but basal therapy has been hampered by the pharmacokinetic limitations of traditional medium- to long-acting preparations.[107,108] These limitations relate to the inability of these compounds to recreate the low overnight insulin profile with no pronounced peak that is seen in persons who do not have diabetes.[107,108] After the subcutaneous injection of NPH insulin, blood concentrations of the hormone rise to a peak after 4–6 hours and then steadily decline, and, although this provides prolonged exposure to insulin, it does not correspond to a physiologically normal ‘flat’ profile (reviewed by Barnett[72]). In addition, NPH insulin is characterised by low bioavailability and erratic absorption from the injection site,[107] and doses are limited in size because of the risk of hypoglycaemia associated with insulin peaks.[72] It may therefore be necessary to use twice-daily injections to provide basal cover with NPH insulin. Similar problems are encountered with ultralente insulin, which provides peak insulin concentrations in blood after 8–16 hours and has a long duration of action (20–36 hours). This long duration of action may confer an advantage over NPH insulin, but the absorption and elimination of ultralente insulin are even more variable than those of NPH.[109]

It is clear that if the benefits of the novel rapid-onset insulins are to be realised in full the problems inherent in the conventional insulins used for basal therapy must be overcome or at least minimised. The pharmacokinetic profile of insulin glargine is consistent with the requirement for physiological insulin profiles when used once daily: amorphous precipitation in the subcutaneous tissues after injection yields predictable blood insulin profiles that lack a peak effect and persist for approximately 24 hours (section 2). The previous review of insulin glargine in Drugs noted that the drug was well tolerated and at least as effective as once- or twice-daily NPH insulin when used as basal therapy in the control of fasting blood or plasma glucose levels in patients with type 1 or type 2 diabetes.[8] Variability in absorption between individuals was found to be lower with insulin glargine than with NPH or ultralente, and most studies showed clinically and statistically significant reductions relative to NPH in incidence of nocturnal hypoglycaemia when insulin glargine was injected at bedtime.[8] Data reviewed in the present update confirm these findings and allow the place of insulin glargine as basal therapy in the management of diabetes to be defined further.

A major development since the last review[8] has been the emergence of evidence that insulin glargine therapy has a significantly greater beneficial effect on HbA1c levels than NPH insulin in both patients with type 1[4850] (section 4.1.1) and those with type 2[75] diabetes (section 4.3.1). In all these studies, patients continued to use the regular mealtime insulin regimen (those with type 1 disease) or oral antihyperglycaemic agents (those with type 2 disease) to which they were accustomed. A recent meta-analysis of phase III and IV clinical trial data predicted that once-daily insulin glargine was associated with a clinically relevant and statistically significant reduction in HbA1c compared with once-daily NPH insulin when the incidence of confirmed symptomatic or confirmed nocturnal hypoglycaemia were made equivalent in the two treatment groups (section 4.3.1).[85] Interestingly, whether insulin glargine was administered in the morning or evening had no clinically relevant effect on glycaemic control in patients with type 1 or 2 diabetes. These observations are particularly pertinent when viewed in light of European[110] and US FDA[111] approval of administration of insulin glargine at any time of day. There are also early indications from small, nonblind studies that glycaemic control with daily subcutaneous injections of insulin glargine may be comparable to that obtained with basal therapy with CSII (section 4.1.2).[62] Results to date are inconclusive, however, and further comparison of these treatment options is needed.

Additional data obtained in children and adolescents with type 1 diabetes are in accordance with earlier findings. Although no further large, randomised comparisons with conventional basal insulin have yet been carried out in this group of patients, results from a small crossover trial[66] and noncomparative investigations[70,71] continue to show reductions relative to NPH and other conventional insulin therapies in severe or nocturnal hypoglycaemia when patients are transferred to insulin glargine treatment (section 4.2).

Preliminary results have also become available from long-term (up to 39 months) extension studies to show that the antihyperglycaemic efficacy of insulin glargine is maintained beyond the periods covered by the originally reviewed randomised comparisons with NPH insulin. Mean HbA1c levels continued to be reduced over the extension periods in patients with either type of diabetes (section 4.1.1 and 4.3.1).[57,86] Glycaemic control was maintained for up to 36 months in a further long-term investigation in children and adolescents (section 4.2).[67]

The long-term effect of insulin glargine relative to that of NPH insulin on bodyweight, particularly in patients with type 2 diabetes, has not been fully clarified to date. However, current data indicate that bodyweight gains are no greater overall with insulin glargine than with NPH, with one study showing less weight gain with insulin glargine[80] and another showing similar increases in each group.[81] More recent noncomparative studies showed improvements in glycaemic control over several months with either minimal weight gain only or small reductions in bodyweight in patients with type 2 diabetes (section 4.3.3).[88,89,92,93] Furthermore, long-term data in patients with type 1 diabetes have shown a minimal increase only in mean bodyweight for up to 36 months with insulin glargine treatment (section 4.1.1).[57]

Briefly summarised DTSQ results from one trial in patients with type 1 diabetes[50] concur with previously published and reviewed data that show greater patient satisfaction with insulin glargine treatment than with NPH (section 4.1.1).[8,52] Preliminary data from a 52-week randomised, multicentre trial in over 400 patients with type 2 disease also indicated that treatment satisfaction was greater with insulin glargine treatment than with NPH insulin (section 4.3).[82] Further details from these studies, together with additional quality-of-life analyses in patients using insulin glargine, will be needed to confirm the apparently high patient acceptability of this preparation.

At the time of the previous review,[8] randomised trials showed that the incidence of adverse events overall with insulin glargine was no greater than that with NPH insulin, with similar levels of immunogenic response to each compound. Recent data are in agreement with these earlier findings. In spite of earlier concerns that the precipitation of insulin glargine in subcutaneous tissues might give rise to high rates of injection site reactions, data continue to show these to be generally mild and of limited duration, and not to necessitate withdrawal of treatment in the large majority of patients. Indeed, no injection site reactions were reported for up to 39 months’ follow-up in one group of patients participating in a recently reported long-term study (section 5).[86] Limited data are also now available to show no clinically relevant tolerability concerns relating to the use of insulin glargine in patients with renal failure (section 5).[90]

Structural changes made to human insulin in the development of insulin analogues can alter interactions between the insulin molecule and its reaction with human insulin and IGF-1 receptors. This has given rise in the past to concern over the mitogenic potential of insulin analogues, but 2-year data obtained in rats and mice show no evidence of any systemic carcinogenicity of insulin glargine (section 2.4).[27]

It should be noted that official guidelines relating to the use of insulin glargine have been published by the National Institute for Clinical Excellence (NICE) in the UK.[6] On the basis of evidence of efficacy in terms of glycaemic control and potential avoidance of hypoglycaemic episodes, the Institute has recommended that insulin glargine be recommended as a treatment option in patients with type 1 diabetes who require insulin. This recommendation was also based on a cost-effectiveness estimate for which details are not available but which suggest that the incremental cost-effectiveness ratio for insulin glargine versus NPH insulin would be around £3500 per quality-adjusted life-year (QALY) for type 1 disease (figures reported in December 2002). The incremental cost per QALY for patients with type 2 diabetes was calculated as £32 500. In light of this last figure, the Institute concluded that it would be desirable to recommend the use of insulin glargine in patients with type 2 disease most likely to benefit. These include those who require assistance from a caregiver or healthcare professional when injecting insulin; those whose lifestyle is significantly restricted by recurrent symptomatic hypoglycaemia; and those who would otherwise need twice-daily basal insulin injections in combination with oral antihyperglycaemic drugs.

The Institute estimated the incremental cost (based on cost prices of insulin glargine and NPH insulin) to the UK National Health Service of switching all potentially eligible patients to treatment with insulin glargine to be around £16 million (as at December 2002).[6] There was no further discussion, however, of potential future healthcare savings that might accrue from improvements in glycaemic control relative to those obtained with previously used insulin regimens, and from the reduction in frequency of hypoglycaemia across the treated population. Well designed pharmacoeconomic analyses will be needed to identify and quantify these potential benefits.

In conclusion, insulin glargine is an effective and well tolerated basal therapy when given as a single daily subcutaneous injection to patients with diabetes, with benefits in terms of glycaemic control and reduced frequency of hypoglycaemia over regimens based on conventional basal insulins. Accumulating data and official recommendations show the suitability of insulin glargine for first-line use in selected patients with type 2 diabetes who require insulin treatment as well as in patients with type 1 disease, and confirm its use in children and adolescents.

Footnotes
1

Use of tradename is for identification purposes only and does not imply endorsement.

 

Copyright information

© Adis Data Information BV 2003