Evolution of Insulin: From Human to Analog Joseph M. Tibaldi, MD Queens Diabetes and Endocrinology Associates, Fresh Meadows, New York, NY.

ABSTRACT The development of insulin analogs has made improved treatment of type 2 diabetes possible. In this article, structural alterations, pharmacokinetics and pharmacodynamics, clinical end points, and safety issues are reviewed for the currently available basal insulins, rapid-acting insulins, and premixes. The flatter activity profiles of insulin glargine and insulin detemir translate into good clinical efficacy with a lower risk of hypoglycemia relative to neutral protamine Hagedorn insulin. Weight gain is consistently lower with insulin detemir than with neutral protamine Hagedorn insulin. Insulin degludec, licensed in Europe and Japan but not yet in the United States, has a mean half-life of 25.4 hours, a duration of action of >42 hours, and low variability. In trials in type 2 diabetes, rates of nocturnal hypoglycemia were lower with insulin degludec than with insulin glargine, and more flexible; once-daily dose timing was shown to be possible. Insulin lispro, insulin aspart, and insulin glulisine are rapidly absorbed after injection and thus provide better coverage of the post-prandial glucose surge compared with human insulin. Trials and meta-analyses show that reductions in glycated hemoglobin are similar and control of postprandial glucose is better with the rapid-acting analogs versus human insulin. Convenience is greater for patients because the analogs can be injected just before a meal. In premix or biphasic insulins, a proportion of the rapid-acting analog is protaminated, providing both rapid-acting and intermediate-acting components in one formulation, thus reducing the number of injections required. Alterations to human insulin have resulted in improvements in safety, efficacy, tolerability, and convenience for patients. Ó 2014 Elsevier Inc. All rights reserved.
The American Journal of Medicine (2014) 127, S25-S38 KEYWORDS: Analog insulin; Human insulin; Hypoglycemia; Insulin evolution; Type 2 diabetes

The history of insulin therapy has been one of continually evolving improvement. Since insulin was first isolated in 1921, milestones in this evolution have included the development of a slower-acting preparation in the late 1940s, neutral protamine Hagedorn insulin, and the use of recombinant technology to enable production large amounts of insulin in 1977 (Figure 1). This synthetic insulin was named “human insulin” to distinguish it from the earlier preparations derived from animal sources. Further milestones were the

Funding: The publication of this manuscript was funded by Novo Nordisk Inc. Conflict of Interest: JMT is a member of the speakers’ bureau for Merck, Janssen, Daiichi, and Novo Nordisk, and a consultant for Novo Nordisk. Authorship: The author takes full responsibility for the content of this manuscript. Writing support was provided by Watermeadow Medical. Requests for reprints should be addressed to Joseph M. Tibaldi, MD, Queens Diabetes and Endocrinology Associates, 5945 161st St, Fresh Meadows, New York, NY 11365. E-mail address: [email protected] 0002-9343/$ -see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjmed.2014.07.005

introduction of rapid-acting insulin analogs in the 1990s and long-acting basal analogs in the early 2000s. The ever-increasing prevalence of type 2 diabetes and longer life expectancy of patients with diabetes mean that the use of and demand for insulin therapy will continue to grow. This article reviews recent developments in insulin that have improved the treatment of type 2 diabetes. We will focus first on the basal or long-acting insulin analogs, because these are the most widely prescribed insulins in type 2 diabetes. We will then briefly consider the rapid-acting analogs, used primarily in advanced type 2 diabetes, and premix insulins, which may be used to initiate or optimize insulin therapy.

BASAL INSULINS Where Are We Currently? Currently, widely used basal insulins are the intermediateacting neutral protamine Hagedorn insulin and the basal analogs insulin glargine (Lantus; Sanofi, Paris, France) and

S26

The American Journal of Medicine, Vol 127, No 10A, October 2014

Figure 1

Milestones in the evolution of insulin therapy. NPH ¼ neutral protamine Hagedorn.

insulin detemir (Levemir; Novo Nordisk Inc, Plainsboro, NJ). One additional analog, insulin degludec (Tresiba; Novo Nordisk Inc), has been licensed for use in Europe and elsewhere, but approval in the United States will be contingent on the results of a dedicated cardiovascular outcomes trial that is currently in progress. Use of the basal analogs has enabled patients with type 2 diabetes to achieve equivalent levels of glycemic control, with a lower associated risk of hypoglycemia, compared with neutral protamine Hagedorn insulin.1-4 However, only a proportion of patients achieve levels of glycemic control that meet current guidelines, suggesting there is still room for improvement. Indeed, although the aim is for exogenous basal insulin to match the body’s own physiologic basal insulin output as closely as possible, subcutaneous injection substantially changes the kinetics of insulin exposure. Endogenous insulin is secreted by the pancreas directly into the portal vein, from where it is transported to the liver. On the other hand, subcutaneously administered insulin is absorbed into the systemic circulation, so for sufficient insulin to reach the liver and suppress endogenous glucose output, peripheral tissues are inevitably overexposed. This overexposure has been suggested as one of the possible causes of insulin-associated weight gain, a concern that deters many patients from initiating insulin therapy.5 Furthermore, the pharmacokinetic profile of injected insulin differs from that of endogenous insulin secretion, in which a steady rate of basal output is augmented by rapidly produced peaks in response to meals. Absorption of an insulin injection tends to follow a profile that increases to a peak before “tailing-off.” In the case of basal insulins, this profile is protracted, but it may be only a poor approximation

of the smooth flat profile of basal endogenous insulin secretion seen in the healthy state. This discrepancy increases the risk of hypoglycemia when insulin levels are too high relative to food intake or hyperglycemia when levels are low.6 The profile of absorption also can vary between patients and even from injection to injection in the same patient. Therefore, the blood glucoseelowering effect may vary not only across the day but also day to day,7 making it difficult to calculate correct dosing. Finally, with once-daily dosing, subcutaneously injected basal insulin can become depleted in 24 hours, and most are single-dose studies because of constraints on the patients, who must fast for the duration of the clamp), they are at present the best method available for measuring insulin activity profiles. The activity profiles of insulin glargine are flatter than those of neutral protamine Hagedorn insulin in type 1 diabetes14 and type 2 diabetes.15 Likewise, insulin detemir exhibits flatter activity profiles than neutral protamine Hagedorn insulin in both type 1 diabetes14,16 and type 2 diabetes.17 Neither analog showed a completely peakless profile; rather, both showed a gentle increase and decrease in activity. Both analogs have shown a mean duration of action of approximately 24 hours in clamp studies using clinically relevant doses.6 Insulin degludec has been shown to have a smooth, stable pharmacodynamic profile over 24 hours at steady state; a mean half-life of 25.4 hours, twice that of insulin glargine (12.5 hours)18; and a duration of action of >42 hours.19 As mentioned earlier, within-subject variability (defined as inconsistency of the blood glucoseelowering profile from one injection to the next in a single subject) can make it difficult to dose insulin correctly and safely reach glycemic targets. Pharmacodynamic variability is assessed using repeat clamp studies. The area under the glucose infusion rate profiles over specific time periods, for example, 24 hours, and the maximum glucose infusion rate are measured, and a coefficient of variation determined. Insulin detemir showed significantly lower within-subject variability than both neutral protamine Hagedorn insulin and insulin glargine, with coefficient of variation values as follows: glucose infusion rate profiles over 24 hours 27% (insulin detemir), 68% neutral protamine Hagedorn insulin, 48% insulin glargine; maximum glucose infusion rate 23%, 46%, and 36%, respectively (P < .001 for all comparisons).14 Insulin degludec, in turn, has shown a lower day-to-day within-subject variability in total glucoselowering effect than insulin glargine in patients with type 1 diabetes.20

Renal and hepatic impairment can affect the metabolism of many drugs, including insulin, which is partly metabolized through the kidneys. The prescribing information for both insulin glargine and insulin detemir advises possible reductions in dose in patients with renal or hepatic impairment, because of reduced insulin metabolism.21,22 In this author’s opinion, the ability to use a basal insulin analog without needing to be concerned about renal or hepatic impairment would be particularly advantageous in treating patients with diabetes hospitalized for other reasons, who are at major risk of experiencing hypoglycemia.

Clinical End Points Although preclinical evidence is useful, it is the clinical efficacy and safety that ultimately determine whether a new insulin represents a step forward from existing products. In diabetes, the key criteria for judging clinical efficacy are glycemic control and hypoglycemia; other outcomes of interest include weight change and insulin dose. Because glycemic control and hypoglycemia are linked, modern clinical trials are “treat-to-target” trials in which basal insulin is titrated to achieve a similar fasting plasma glucose value. This should result in similar glycated hemoglobin (HbA1c) values at the end of the trial, which means that any differences in the incidence/severity of hypoglycemia can be interpreted without having to try to correct for different levels of glycemic control.

Clinical Trials in Type 2 Diabetes Patients with type 2 diabetes who used insulin glargine1,4 or insulin detemir2,3 in basaleoral therapy achieved similar glycemic control to those using neutral protamine Hagedorn, with lower rates of hypoglycemia, particularly nocturnal hypoglycemia (Table 1). A comparison of insulin glargine with insulin detemir in basaleoral therapy showed similar efficacy for both analogs.23 When used in basalebolus therapy in type 2 diabetes, insulin detemir was shown to effectively reduce HbA1c without increasing hypoglycemia, relative to neutral protamine Hagedorn insulin.24 Insulin glargine and insulin detemir have been compared in basalebolus therapy in type 2 diabetes.25,26 Overall, for use in type 2 diabetes (basaleoral or basalebolus), glycemic control was similar with the 2 analogs, which was not surprising because the trials were treat-to-target in design. Weight gain has been consistently lower with insulin detemir. On the other hand, there is evidence that twice-daily dosing—required by some patients—may be needed more frequently with insulin detemir than with insulin glargine, and that unit doses may be higher with insulin detemir. These are overall impressions based on mean results and supported by a Cochrane review of the literature.27 In practice, therapy needs to be individualized for each patient according to factors such as their age and comorbidities.

Tibaldi

Evolution of Insulin

Table 1

S29

Key Clinical Trials of the Basal Insulin Analogs in Type 2 Diabetes

Treatments Compared

Key Trials in BasaleOral Therapy

Glargine vs NPH

Yki-Jarvinen et al 20004

Glargine vs NPH

Riddle et al 20031

Detemir vs NPH

Hermansen et al 20062

Detemir vs NPH

Raslova et al 20043 (degludec þ aspart vs NPH þ human soluble insulin)

Glargine vs detemir

Rosenstock et al 200823

Glargine vs detemir

Swinnen et al 201127 (Cochrane review)

Degludec OD vs glargine OD

Onishi et al 201229

Results Analog vs NPH: HbA1c similar Hypoglycemia lower: NPH 43%, glargine 34% of patients (P ¼ .04) Nocturnal hypoglycemia lower: NPH 24%, glargine 9.9% of patients (P ¼ .0001) Analog vs NPH: HbA1c similar Hypoglycemia lower: NPH 13.9, glargine 17.7 events/patient-y (P < .02) Nocturnal hypoglycemia lower: NPH 6.9, glargine 4.0 events/patient-y (P < .001) Analog vs NPH: HbA1c similar Hypoglycemia lower: RR detemir/NPH, 0.53 (P < .001) Nocturnal hypoglycemia lower: RR detemir/NPH, 0.45 (P < .001) Weight gain 1.6 kg lower with detemir (estimated) (P < .001) Analog vs NPH: HbA1c similar Hypoglycemia similar Nocturnal hypoglycemia lower: RR detemir/NPH, 0.54 (P ¼ .04) Weight gain lower with detemir: Weight gain 0.51 kg (detemir), 1.13 kg (NPH), P ¼ .038 HbA1c similar Hypoglycemia similar Weight gain lower with detemir: Weight gain 2.7 kg (detemir), 3.5 kg (glargine), P ¼ .03 Detemir dose higher: 0.78 U/kg (detemir), 0.44 IU/kg (glargine) (not analyzed statistically) HbA1c similar Hypoglycemia similar Weight gain lower with detemir: detemir glargine 0.91 kg (95% CI, 1.21 to 0.61) Detemir dose 0.26 U/kg higher (95% CI, 0.11-0.41) Degludec OD vs glargine OD: HbA1c similar Hypoglycemia similar Nocturnal hypoglycemia similar

Key Trials in BasaleBolus Therapy

Results

Not performed

-

-

-

Haak et al 200524

Detemir vs NPH: HbA1c similar Hypoglycemia similar Weight gain 0.79 kg lower (P ¼ .017)

-

-

Hollander et al 200826

HbA1c similar Hypoglycemia similar Weight gain 1.04 kg lower with detemir, P < .05 Insulin dose similar

Raskin et al 200925

HbA1c similar Hypoglycemia similar Weight gain 1.2 kg (detemir), 2.7 kg (glargine), P ¼ .001 Insulin dose similar

Garber et al 201230

HbA1c similar Hypoglycemia lower: Estimated RR degludec/glargine, 0.82 (P ¼ .0359) Nocturnal hypoglycemia lower: Estimated RR degludec/glargine, 0.75 (P ¼ .0399)

S30

The American Journal of Medicine, Vol 127, No 10A, October 2014

Table 1

Continued

Treatments Compared

Key Trials in BasaleOral Therapy 28

Degludec OD vs glargine OD

Zinman et al 2012

Degludec dosed flexibly vs glargine OD

Meneghini et al 201331

Results Degludec OD vs glargine OD: HbA1c similar Hypoglycemia similar Nocturnal hypoglycemia lower: estimated RR, degludec/glargine 0.64 (P ¼ .038) Flexibly dosed degludec vs glargine: HbA1c similar Hypoglycemia similar Nocturnal hypoglycemia similar

Key Trials in BasaleBolus Therapy

Results

-

-

-

-

Differences were statistically significant unless stated otherwise. Estimated refers to values estimated from reported final values or from figures. CI ¼ confidence interval; HbA1c ¼ glycated hemoglobin; NPH ¼ neutral protamine Hagedorn; NS ¼ not significant; OD ¼ once daily; RR ¼ rate ratio.

Phase 3 treat-to-target studies of insulin degludec for basaleoral therapy in type 2 diabetes showed that glycemic control with insulin degludec was similar to that with insulin glargine, and rates of nocturnal hypoglycemia were lower28 or similar.29 In a treat-to-target trial of basalebolus therapy in type 2 diabetes, glycemic control was similar with insulin degludec and insulin glargine, with lower rates of overall confirmed hypoglycemia and nocturnal confirmed hypoglycemia in the insulin degludec group.30 A further study showed that insulin degludec can be dosed once daily in a more flexible manner, at intervals of between 8 and 40 hours, without losing efficacy or increasing hypoglycemia.31

Real-Life Observational Studies Real-life observational studies have lent support to the expectations of good efficacy and safety raised by the clinical trials of insulin glargine and insulin detemir. For example, an observational study showed that 12,216 patients with type 2 diabetes who received insulin glargine as add-on treatment to oral antidiabetic agents demonstrated improved glycemic control without weight gain and with a low incidence of hypoglycemia at 9 and 12 months.32 In the observational study Predictable Results and Experience in Diabetes through Intensification and Control to Target: An International Variability Evaluation (PREDICTIVE), 14-week follow-up of patients who switched to insulin detemir from other insulin regimens showed improved glycemic control, a reduction in major hypoglycemia, and a small decrease in mean body weight in type 2 diabetes (n ¼ 12,981).33

Long-term Safety In modifying biological molecules such as insulin, it is essential to ensure that their potencies with respect to other endocrinological actions are not altered so as to be potentially harmful. With insulin, the main concern is that molecular alterations could increase the risk of mitogenic activity by increasing residence time on the insulin receptor or increasing the affinity of the ligand for insulin-like growth factor-1 receptors.34,35

Concerns were recently expressed that insulin glargine could increase the incidence of cancer, possibly as a result of an increased affinity ratio for insulin-like growth factor-1 receptors versus insulin receptors, relative to human insulin.34,36,37 However, these concerns were based on epidemiologic analyses with some methodological shortcomings, for example, the confounding effects of cancer incidence being associated with obesity, diabetes, and increasing age. Subsequently, concerns about cancer risk with insulin glargine appear to have been refuted by the publication in 2012 of the Outcome Reduction With Initial Glargine Intervention (ORIGIN) study, one of only a few studies reporting on the long-term safety of a basal insulin analog.38 The study examined the use of insulin glargine to target normal fasting plasma glucose levels in more than 12,000 people with cardiovascular risk factors plus impaired fasting glucose, impaired glucose tolerance, or type 2 diabetes. At 6-year follow-up, insulin glargine had a neutral effect on the incidence of cancer.38 Although increased affinity for the insulin-like growth factor-1 receptor is an unwanted characteristic in an insulin analog, it would not necessarily result in mitogenic effects. The increased affinity of the analog is still likely to be orders of magnitude lower than its affinity for the insulin receptor or the affinity of insulin-like growth factor-1 for its own receptor, and the analog would need to reach cells expressing the receptor in high enough concentrations to elicit a mitogenic response.36 In addition, insulin glargine is quickly transformed to its M1 metabolite, which has shown a lower affinity for insulin-like growth factor-1 than insulin glargine or, indeed, human insulin, after subcutaneous injection in male subjects with type 1 diabetes.39 In vitro molecular characterization studies showed that insulin detemir has an insulin-like growth factor-1 receptor:insulin receptor binding affinity ratio of 1 relative to human insulin and a similar dissociation pattern from insulin receptors to that of human insulin.34 Similarly with insulin degludec, insulin-like growth factor-1 receptor:insulin receptor affinity and mitogenic:metabolic potency ratios are not increased relative to human insulin.40 For a discussion of the possibility of cardiovascular

Tibaldi

Evolution of Insulin

S31

effects arising from long-term use of insulins, see the article by Moghissi and King41 in this supplement.

RAPID-ACTING INSULINS Where Are We Currently? A major disadvantage of exogenous human insulin is that it cannot match the rapid surge observed with physiologic insulin secretion after meals, which suppresses increases in postprandial glucose. Human insulin has a slower onset of action (with a peak effect 3 hours after dosing) and longer duration of action (extending >8 hours) compared with this endogenous postprandial surge.42 The rapid-acting analogs were developed to speed up insulin absorption relative to human soluble insulin, and thereby more effectively minimize postprandial glucose excursions and reduce the risk of hypoglycemia due to exogenous insulin concentrations remaining high for longer than needed. Three rapidacting analogs are currently available: insulin lispro (Humalog; Eli Lilly and Company, Indianapolis, Ind), insulin aspart (NovoLog; Novo Nordisk Inc), and insulin

Figure 3

glulisine (Apidra; Sanofi). In the United States, the use of rapid-acting analogs is widespread and human soluble insulin is now rarely used. Rapid-acting insulin analogs are generally prescribed in type 2 diabetes only when basal insulin or (less commonly) use of oral agents has failed to achieve glycemic control. The analogs can then be used as 1, 2, or 3 times per day additions to basal therapy43 or to provide so-called complementary insulin therapy or supplementary insulin therapy, in which oral antidiabetic therapy is supplemented with prandial insulin only. The latter approach is favored in some countries but has not been widely adopted in the United States. Rapid-acting analogs also are used in continuous subcutaneous insulin infusion. Premix formulations of the rapid-acting analogs, used to initiate insulin therapy in type 2 diabetes, are described separately.

Structure of the Rapid-Acting Insulin Analogs In all 3 rapid-acting analogs, small changes have been made in the amino acid sequence of the insulin B chain (Figure 3). These amino acid substitutions destabilize the

Structure of insulin lispro, insulin aspart, and insulin glulisine.

S32

The American Journal of Medicine, Vol 127, No 10A, October 2014

natural tendency for insulin monomers to self-associate as dimers and hexamers. As a result, the analogs are absorbed into the circulation more rapidly after subcutaneous injection.44,45

Pharmacokinetics and Pharmacodynamics of the Rapid-Acting Insulin Analogs Because all 3 rapid-acting analogs are designed to achieve rapid dissociation into monomers in the subcutaneous depot, their pharmacokinetic and pharmacodynamic properties do not differ much from each other. The key feature of all 3 rapid-acting analogs is an earlier insulin peak and a more rapid tailing-off of activity compared with human insulin. The pharmacokinetic and pharmacodynamic of the rapidacting analogs have been described in a recent systematic review,44 and only a few examples will be cited. Three studies of insulin aspart in healthy volunteers showed that after subcutaneous injection, the median time to maximum concentration was 40 to 50 minutes, versus 105 to 150 minutes for human insulin (P < .05 in all 3 studies).46-48 Studies in nondiabetic subjects have shown similar patterns of release for insulin aspart, insulin lispro, and insulin glulisine. Two studies that compared insulin lispro directly with insulin aspart or insulin glulisine showed similar pharmacokinetic and pharmacodynamic profiles for all 3 analogs.49,50 Likewise, in studies in patients with type 1 diabetes, all 3 analogs were released more rapidly and reached a higher maximum insulin concentration compared with human insulin.51-53 Figure 4 shows the pharmacokinetic profiles of insulin aspart and insulin glulisine, which, after injection, reach their maximum plasma concentrations at approximately 40 and 55 minutes, respectively. Insulin lispro exhibits a similar profile to both these analogs, reaching its peak plasma concentration approximately 68 minutes after injection. The results obtained in these studies in patients with type 1 diabetes could not be directly compared because of interstudy variability in the comparator (human insulin) and various other parameters. However, within each study, when compared with human insulin administered at the same time (ie, immediately before a meal), maximum glucose concentrations were significantly lower for each of the insulin analogs, insulin lispro (9.9  1.4 mmol/L vs 11.9  2.8 mmol/L [mean  standard deviation {SD}], P < .05), insulin aspart (13.5  3.5 mmol/L vs 16.4  3.4 mmol/L [mean  SD], P < .0001), and insulin glulisine (10.0 mmol/L vs 11.6 mmol/L, 95% confidence interval [CI]).51-53 Similar results were seen for glucose excursion measured by area under the curve, with significantly lower values for all 3 analogs when compared with human insulin: insulin lispro: 29.3  5.7 mmol/L*h vs 37.7  11.3 mmol/L*h (mean  SD) (P < .01); insulin aspart: 14.9  8.7 mmol/L*h vs 21.9  8.5 mmol/L*h (mean  SD) (P < .0001), and insulin glulisine: 15.5 mmol/L*h vs 18.6 mmol/L*h (95% CI).51-53 In the studies with insulin aspart and insulin

Figure 4 Serum insulin levels after subcutaneous injection of (A) insulin aspart, and (B) insulin glulisine, compared with human insulin in patients with type 1 diabetes. Adapted with permission from Lindholm et al51 and Rave et al.52

glulisine, human insulin also was administered at an additional time point, 30 minutes before a meal, and this resulted in a smaller difference between analog and human insulin.51,52 A study in type 2 diabetes compared insulin aspart injected just before a meal with human insulin injected just before or 30 minutes before a meal.54 The insulin profile of insulin aspart was similar to that observed in the studies in healthy volunteers and patients with type 2 diabetes. The postprandial glucose excursion was smaller in the insulin aspart group (899  609 mmol/L$min) compared with the group that received human insulin injected just before a meal (1102  497 mmol/ L$min, P < .01), but did not differ for insulin aspart compared with human insulin injected 30 minutes before a meal (868  374 mmol/L$min). Dosing at least 30 minutes before a meal is a requirement for human insulin to achieve adequate control of postprandial glucose levels, but many patients are unable to follow this requirement. The fact that all the rapid-acting

Tibaldi

Evolution of Insulin

S33

analogs can be administered closer to the start of a meal translates into greater convenience for patients. When insulin lispro and insulin glulisine were compared in obese patients with type 2 diabetes, both were rapidly absorbed. Insulin glulisine showed a faster increase in insulin concentration than insulin lispro, but the 2 analogs produced similar plasma glucose profiles.45 Renal impairment, hepatic impairment, and obesity do not affect the pharmacokinetic of insulin aspart to a clinically meaningful degree.55 However, the prescribing information for all 3 rapid-acting analogs recommends that, as with all insulin medicinal products, glucose monitoring should be intensified and the bolus insulin dose adjusted on an individual basis in patients with renal or hepatic impairment.56-58

Long-term Safety

Clinical End Points Insulin lispro, insulin aspart, and insulin glulisine have been compared with human insulin in trials of basalebolus therapy in type 2 diabetes. All 3 analogs showed similar or superior decreases in HbA1c and superior postprandial glucose control versus human insulin (Table 2).59-61 Between-group rates of hypoglycemia were similar, with a trend toward reduced nocturnal hypoglycemia in the study with insulin lispro.61 Several meta-analyses of the results from clinical trials of the rapid-acting analogs are available, some of which included premix formulations (Table 3).62-66 The studies included were heterogeneous and inconsistent in methods of reporting postprandial glucose or in the definition of hypoglycemia. Overall, the rapid-acting analogs did not differ significantly from human insulin in decreasing HbA1c, but postprandial glucose (when analyzed) was significantly lower with the analogs, in particular for post-breakfast and post-dinner values (likely reflecting the fact that premixes are frequently administered in the morning and evening).

Table 2

The meta-analyses did not find any significant difference in hypoglycemic risk between the analogs and human insulin. A more recent meta-analysis of patient-level data from trials comparing insulin aspart with human insulin in type 1 diabetes and type 2 diabetes concluded that nocturnal hypoglycemia was significantly lower with insulin aspart (unfortunately, the results are not reported by diabetes type).63 In line with the patient-centered approach recommended by the American Diabetes Association/European Association for the Study of Diabetes, the rapid-acting analogs are a useful choice for patients with type 2 diabetes whose HbA1c remains high despite fasting plasma glucose being at target after 3 to 5 months of basal insulin titration.43 In these patients, more effective control of postprandial glucose is needed.

All 3 rapid-acting analogs have a binding affinity for, and residence time on, the insulin receptor that is similar to or lower than that of human insulin.67,68 In different experiments, insulin aspart was shown to have 81% of the affinity of human insulin for the insulin-like growth factor-1 receptor, insulin glulisine 20% to 25%, and insulin lispro 156%. These results suggest that increased mitogenicity should not be a concern, and experiments in cell lines showed that the mitogenic potency of each analog was lower than that of human insulin.67,68 These findings apply to premixes derived from insulin lispro or insulin aspart.

PREMIX INSULINS Where Are We Currently? Premix or biphasic insulins were designed to maximize patient convenience and reduce the number of daily injections required by providing both rapid-acting and intermediate-acting components in one formulation. Until

Key Clinical Trials of Insulin Lispro, Insulin Aspart, and Insulin Glulisine in BasaleBolus Therapy in Type 2 Diabetes

Treatments Compared Lispro þ basal insulin vs human insulin þ basal insulin Aspart  NPH vs human insulin  NPH* Glulisine þ NPH vs human insulin þ NPH

Key Trial Ross et al 2001

Results 61

Bretzel et al 200459

Dailey et al 200460

HbA1c similar Post-breakfast and post-dinner glucose 1.3-1.4 mmol/L lower (estimated) (P < .02) Hypoglycemia similar HbA1c similar Mean postprandial glucose lower by 0.44 to >1.67 mmol/L with aspart vs human insulin (no significance testing reported) Hypoglycemia similar HbA1c lower with glulisine (0.46% vs 0.30% with human insulin; P ¼ .0029) Post-breakfast glucose 0.36 mmol/L lower, post-dinner glucose 0.51 mmol/L lower with glulisine (estimated values) Hypoglycemia similar

Estimated refers to values estimated from reported final values or from figures. HbA1c ¼ glycated hemoglobin; NPH ¼ neutral protamine Hagedorn. *A third arm, human premix insulin (70% NPH/30% regular), is not reported.

S34

The American Journal of Medicine, Vol 127, No 10A, October 2014

Table 3

Selected Meta-analyses of the Efficacy and Safety of the Rapid-Acting Analogs Compared with Human Insulin in Type 2 Diabetes

Reference Siebenhofer et al 200665

Singh et al 200966

Mannucci et al 200962

Included Studies in Type 2 Diabetes No. of Studies for Each Analog

Result: Rapid-Acting Analog Compared with Human Insulin

Studies published up to September 2005 Lispro (6 trials) Aspart (1 trial) No glulisine studies Studies published up to 2007 Included premix formulations Lispro (11 trials) Aspart (4 trials) No glulisine studies Studies published up to January 2008 Lispro (7 trials) Aspart (4 trials) Glulisine (2 trials) Included premix formulations

HbA1c: no significant difference Hypoglycemia: no significant difference

HbA1c: no significant difference Risk of hypoglycemia: no significant difference

Rys et al 201164

Studies published up to July 2009 Included premix formulations Only studies on aspart (6 trials) or BIAsp (4 trials)

Heller et al 201363

Patient-level data from trials comparing aspart with human insulin. Trials in type 1 diabetes (6 trials, n ¼ 1909), type 2 diabetes (3 trials, n ¼ 219), and types 1 and 2 diabetes combined (1 trial, n ¼ 110)

HbA1c: 0.10% in favor of analogs (95% CI, 0.01 to 0.19) In 3 studies, blood glucose was significantly lower with analogs after breakfast (by 0.7 mmol/L) and dinner (by 0.6 mmol/L) (both P < .001) Severe hypoglycemia: no significant difference HbA1c: no significant difference PPG: daily mean PPG lower by 1.18 mmol/L (95% CI, 1.88 to 0.47) with aspart* Risk of hypoglycemia: no significant difference HbA1c: 0.10% in favor of aspart (95% CI, 0.15 to 0.04), P < .001 PPG: significantly lower by 0.47 mmol/L (95% CI, 0.70 to 0.25) (P < .001) with aspart Hypoglycemia: no significant difference Nocturnal hypoglycemia: significantly lower with aspart: RR, 0.76 (95% CI, 0.67-0.85) (P < .001) These results are for types 1 and 2 diabetes combined

BIAsp ¼ biphasic insulin aspart; CI ¼ confidence interval; HbA1c ¼ glycated hemoglobin; PPG ¼ postprandial glucose. *Including studies of aspart and BIAsp.

recently, the theoretically ideal combination of a mixture of a rapid-acting and a basal analog has not been possible in practice because the basal analog formulations cannot be mixed with other insulins. Instead, in the premixes, a proportion of the rapid-acting analog is protaminated so that it becomes an intermediate-acting insulin. Biphasic human insulin (BHI) contains 30% regular human insulin in solution and 70% protaminated regular human insulin. Currently, the most widely used analog premixes are biphasic insulin aspart 30 (BIAsp 30) (NovoLog Mix 70/30; Novo Nordisk Inc) and biphasic insulin lispro (Mix 25) (Humalog Mix 75/25; Eli Lilly and Company). These consist of 70%/75% intermediate-acting protaminated analog suspension and 30%/25% rapid-acting analog solution, respectively. A mixture of insulin lispro with 50/50 proportions of protaminated and rapid-acting components is available but is used less frequently. The challenges of combining a basal and a rapid-acting insulin in a single formulation recently have been overcome in the form of insulin degludec/insulin aspart (Ryzodeg; Novo Nordisk), a soluble co-formulation of insulin degludec (70%) and insulin aspart (30%). Insulin degludec/ insulin aspart has been licensed in Europe and elsewhere but is not yet widely available, and licensing in the United

States will await the results of the cardiovascular outcomes trial of insulin degludec. Premix insulins are typically prescribed as an initiation regimen for patients with type 2 diabetes who do not achieve glycemic control using oral antidiabetic medications or as an optimization regimen for those whose HbA1c values remain too high despite use of a basal analog.

Pharmacokinetics and Pharmacodynamics of the Premixes A recent review of 10 years of experience with biphasic insulin aspart (BIAsp) included a discussion of the pharmacokinetic and pharmacodynamic profiles.69 The pharmacokinetic profile of BIAsp 30 is closer to that of normal physiologic insulin secretion than the pharmacokinetic profile of BHI, with faster absorption and a higher peak concentration.70 In healthy subjects, the glucose-lowering effect of BIAsp was earlier and more pronounced than that of BHI, whereas the duration of action of the basal component was similar for the 2 preparations.71 A 3-way study that compared the pharmacokinetic and pharmacodynamic of BIAsp, Mix 25, and BHI showed that measures of insulin pharmacokinetic and glucose concentrations were

Tibaldi

Evolution of Insulin

S35 1.24; 95% CI, 1.08-1.43; P < .01).75 Qayyum et al74 did not find any difference in hypoglycemia between BIAsp or Mix 25 compared with BHI. A study comparing BIAsp 30 and Mix 25 in patients with type 2 diabetes concluded that efficacy and safety are similar for the 2 premixes.76 Numerous studies have compared the premix insulins with basal insulins as initiation regimens in patients with type 2 diabetes. Because the current article is focused on the evolution of insulins, from human to analog preparations, these studies are outside its scope, and we refer the reader to a recent review by Vaag and Lund.77 These authors concluded that there is a lack of consensus on whether insulin therapy should be initiated with a basal or a premix insulin and emphasized the importance of considering reallife factors, such as patient lifestyle and compliance with complex treatment regimens, when prescribing insulin for initiation.

Dosing Convenience

Figure 5 Mean postprandial serum glucose (A) and insulin (B) profiles for patients with type 2 diabetes who received an injection of BIAsp or Mix 25 immediately before a meal or BHI 30 to 15 minutes before a meal. Reproduced with permission from Hermansen et al.72 BHI ¼ biphasic human insulin; BIAsp ¼ biphasic insulin aspart.

significantly closer to the physiologic ideal with BIAsp versus human insulin (Figure 5).72 The measures did not differ significantly between BIAsp and Mix 25. Specific studies of the effects of renal or hepatic impairment on the pharmacokinetic and pharmacodynamic of the premix insulins have not been conducted. As with other insulins, doses of the premix insulins may need to be reduced in such patients.56,73

Clinical End Points Studies that have compared BIAsp with BHI in type 2 diabetes are listed in the recent review.69 In general, HbA1c control was similar and postprandial glucose control was better with BIAsp, as would be expected on the basis of the pharmacokinetic profile. This finding also was reported in meta-analyses of trials of rapid-acting analogs74 and trials of BIAsp versus BHI.75 Rates of overall hypoglycemia did not differ between treatments in the individual studies or in the meta-analysis by Davidson et al.75 However, the metaanalysis showed that nocturnal hypoglycemia was significantly lower with BIAsp versus BHI (rate ratio, 0.50; 95% CI, 0.38-0.67; P < .01) and that daytime hypoglycemia was significantly higher with BIAsp than with BHI (rate ratio,

BHI, like human insulin, needs to be given 30 minutes before a meal, a requirement that is impractical for many patients. The premixes, like the rapid-acting analogs, offer greater convenience because they can be administered closer to the start of a meal. Premixes also offer patients the convenience of easy intensification. If control is not achieved with 1 injection daily, second and third injections can be added. A disadvantage of the premixes is the fact that doses of the rapid- and intermediate-acting components cannot be adjusted individually. For most patients, this constraint is outweighed by the convenience of fewer daily injections, but some patients may prefer to change to basalebolus therapy as their type 2 diabetes progresses and insulin therapy changes from being used essentially as just a supplement to a full replacement regimen.

CONCLUSIONS The need for effective insulin therapies is ever increasing as the incidence of type 2 diabetes continues to increase. Two primary concerns of patients are whether an insulin preparation controls their blood glucose adequately without causing hypoglycemia and whether the regimen is simple, convenient, and forgiving if there is a delay in administering it. In terms of the basal analogs, great progress has been made in meeting these requirements. Insulin glargine and insulin detemir offer equivalent glycemic control, with less risk of hypoglycemia, compared with neutral protamine Hagedorn. Insulin degludec, in turn, offers the prospect of once-daily dosing for all patients, greater flexibility of dosing, and even further reductions in the risk of nocturnal hypoglycemia. For the rapid-acting analogs, research efforts are focused on developing molecules that are absorbed into the bloodstream even more quickly after subcutaneous injection. Other approaches to the future of insulin therapy are covered in this supplement.13 With

S36

The American Journal of Medicine, Vol 127, No 10A, October 2014

increasing recognition of the importance of patientreported outcomes, more recent trials of insulin therapy are including measures of patient quality of life. The expectation is that features such as a reduced risk of hypoglycemia and increased convenience will be reflected in improved quality of life scores. The story of insulin is one of a molecule that continues to evolve.

17.

18.

19.

ACKNOWLEDGMENTS The author thanks Grace Townshend and Gabrielle Parker of Watermeadow Medical for writing and editing assistance, supported by Novo Nordisk.

References 1. Riddle MC, Rosenstock J, Gerich J. The treat-to-target trial: randomized addition of glargine or human NPH insulin to oral therapy of type 2 diabetic patients. Diabetes Care. 2003;26:3080-3086. 2. Hermansen K, Davies M, Derezinski T, et al. A 26-week, randomized, parallel, treat-to-target trial comparing insulin detemir with NPH insulin as add-on therapy to oral glucose-lowering drugs in insulin-naive people with type 2 diabetes. Diabetes Care. 2006;29:1269-1274. 3. Raslova K, Bogoev M, Raz I, et al. Insulin detemir and insulin aspart: a promising basal-bolus regimen for type 2 diabetes. Diabetes Res Clin Pract. 2004;66:193-201, Erratum in Diabetes Res Clin Pract. 2006;72:112. 4. Yki-Jarvinen H, Dressler A, Ziemen M. Less nocturnal hypoglycemia and better post-dinner glucose control with bedtime insulin glargine compared with bedtime NPH insulin during insulin combination therapy in type 2 diabetes. HOE 901/3002 Study Group. Diabetes Care. 2000;23:1130-1136. 5. Russell-Jones D, Khan R. Insulin-associated weight gain in diabetesecauses, effects and coping strategies. Diabetes Obes Metab. 2007;9:799-812. 6. Heise T, Pieber TR. Towards peakless, reproducible and long-acting insulins. An assessment of the basal analogues based on isoglycaemic clamp studies. Diabetes Obes Metab. 2007;9:648-659. 7. Vora J, Heise T. Variability of glucose-lowering effect as a limiting factor in optimizing basal insulin therapy: a review. Diabetes Obes Metab. 2013;15:701-712. 8. Felig P. Landmark perspective: Protamine insulin. Hagedorn’s pioneering contribution to drug delivery in the management of diabetes. JAMA. 1984;251:393-396. 9. Heinemann L, Linkeschova R, Rave K, et al. Time-action profile of the long-acting insulin analog insulin glargine (HOE901) in comparison with those of NPH insulin and placebo. Diabetes Care. 2000;23:644-649. 10. Bolli GB, Owens DR. Insulin glargine. Lancet. 2000;356:443-445. 11. Havelund S, Plum A, Ribel U, et al. The mechanism of protraction of insulin detemir, a long-acting, acylated analog of human insulin. Pharm Res. 2004;21:1498-1504. 12. Jonassen I, Havelund S, Hoeg-Jensen T, et al. Design of the novel protraction mechanism of insulin degludec, an ultra-long-acting basal insulin. Pharm Res. 2012;29:2104-2114. 13. Sorli C. New developments in insulin therapy for type 2 diabetes. Am J Med. 2014;127(Suppl 1):S39-S48. 14. Heise T, Nosek L, Ronn BB, et al. Lower within-subject variability of insulin detemir in comparison to NPH insulin and insulin glargine in people with type 1 diabetes. Diabetes. 2004;53:1614-1620. 15. Luzio S, Dunseath G, Peter R, et al. Comparison of the pharmacokinetics and pharmacodynamics of biphasic insulin aspart and insulin glargine in people with type 2 diabetes. Diabetologia. 2006;49: 1163-1168. 16. Plank J, Bodenlenz M, Sinner F, et al. A double-blind, randomized, dose-response study investigating the pharmacodynamic and

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

pharmacokinetic properties of the long-acting insulin analog detemir. Diabetes Care. 2005;28:1107-1112. Klein O, Lynge J, Endahl L, et al. Albumin-bound basal insulin analogues (insulin detemir and NN344): comparable time-action profiles but less variability than insulin glargine in type 2 diabetes. Diabetes Obes Metab. 2007;9:290-299. Heise T, Nosek L, Bottcher SG, et al. Ultra-long-acting insulin degludec has a flat and stable glucose-lowering effect in type 2 diabetes. Diabetes Obes Metab. 2012;14:944-950. Kurtzhals P, Heise T, Strauss HM, et al. Multi-hexamer formation is the underlying mechanism behind the ultra-long glucose-lowering effect of insulin degludec. Diabetes. 2011;60(Suppl 1A):LB12. Heise T, Hermanski L, Nosek L, et al. Insulin degludec: four times lower pharmacodynamic variability than insulin glargine under steady-state conditions in type 1 diabetes. Diabetes Obes Metab. 2012;14:859-864. Sanofi-aventis USA. Last updated 2013. Lantus Prescribing Information. Available at: http://products.sanofi.us/lantus/lantus.html. Accessed August 8, 2014. Novo Nordisk A/S. Last updated 2013. Detemir Prescribing Information. Available at: http://www.novo-pi.com/levemir.pdf. Accessed August 8, 2014. Rosenstock J, Davies M, Home PD, et al. A randomised, 52-week, treat-to-target trial comparing insulin detemir with insulin glargine when administered as add-on to glucose-lowering drugs in insulinnaive people with type 2 diabetes. Diabetologia. 2008;51:408-416. Haak T, Tiengo A, Draeger E, et al. Lower within-subject variability of fasting blood glucose and reduced weight gain with insulin detemir compared to NPH insulin in patients with type 2 diabetes. Diabetes Obes Metab. 2005;7:56-64. Raskin P, Gylvin T, Weng W, Chaykin L. Comparison of insulin detemir and insulin glargine using a basal-bolus regimen in a randomized, controlled clinical study in patients with type 2 diabetes. Diabetes Metab Res Rev. 2009;25:542-548. Hollander P, Cooper J, Bregnhoj J, Pedersen CB. A 52-week, multinational, open-label, parallel-group, noninferiority, treat-to-target trial comparing insulin detemir with insulin glargine in a basal-bolus regimen with mealtime insulin aspart in patients with type 2 diabetes. Clin Ther. 2008;30:1976-1987. Swinnen SG, Simon AC, Holleman F, et al. Insulin detemir versus insulin glargine for type 2 diabetes mellitus. Cochrane Database Syst Rev. 2011:CD006383. Zinman B, Philis-Tsimikas A, Cariou B, et al. Insulin degludec versus insulin glargine in insulin-naive patients with type 2 diabetes: a 1-year, randomized, treat-to-target trial (BEGIN Once Long). Diabetes Care. 2012;35:2464-2471. Onishi Y, Iwamoto Y, Yoo SJ, et al. Insulin degludec compared with insulin glargine in insulin-naïve patients with type 2 diabetes: a 26-week, randomized, controlled, Pan-Asian, treat-to-target trial. J Diabetes Invest. 2013;4:605-612. Garber AJ, King AB, Del Prato S, et al. Insulin degludec, an ultralongacting basal insulin, versus insulin glargine in basal-bolus treatment with mealtime insulin aspart in type 2 diabetes (BEGIN Basal-Bolus Type 2): a phase 3, randomised, open-label, treat-totarget non-inferiority trial. Lancet. 2012;379:1498-1507. Meneghini L, Atkin SL, Gough SCL, et al. The efficacy and safety of insulin degludec given in variable once-daily dosing intervals compared with insulin glargine and insulin degludec dosed at the same time daily: a 26-week, randomized, open-label, parallel-group, treat-totarget trial in people with type 2 diabetes. Diabetes Care. 2013;36: 858-864. Schreiber SA, Haak T. Insulin glargine benefits patients with type 2 diabetes inadequately controlled on oral antidiabetic treatment: an observational study of everyday practice in 12,216 patients. Diabetes Obes Metab. 2007;9:31-38. Dornhorst A, Luddeke HJ, Sreenan S, et al. Safety and efficacy of insulin detemir in clinical practice: 14-week follow-up data from type 1 and type 2 diabetes patients in the PREDICTIVE European cohort. Int J Clin Pract. 2007;61:523-528.

Tibaldi

Evolution of Insulin

34. Hansen BF, Glendorf T, Hegelund AC, et al. Molecular characterisation of long-acting insulin analogues in comparison with human insulin, IGF-1 and insulin X10. PLoS One. 2012;7:e34274. 35. Pollak M, Russell-Jones D. Insulin analogues and cancer risk: cause for concern or cause celebre? Int J Clin Pract. 2010;64:628-636. 36. Gough SC, Belda-Iniesta C, Poole C, et al. Insulin therapy in diabetes and cancer risk: current understanding and implications for future study: proceedings from a meeting of a European Insulin Safety Consensus Panel, convened and sponsored by Novo Nordisk, held Tuesday October 5, 2010 at The Radisson Edwardian Heathrow Hotel, Hayes, Middlesex, UK. Adv Ther. 2011;28(Suppl 5):1-18. 37. Hemkens LG, Grouven U, Bender R, et al. Risk of malignancies in patients with diabetes treated with human insulin or insulin analogues: a cohort study. Diabetologia. 2009;52:1732-1744. 38. ORIGIN Trial Investigators. Basal insulin and cardiovascular and other outcomes in dysglycemia. N Engl J Med. 2012;367:319-328. 39. Bolli GB, Hahn AD, Schmidt R, et al. Plasma exposure to insulin glargine and its metabolites M1 and M2 after subcutaneous injection of therapeutic and supratherapeutic doses of glargine in subjects with type 1 diabetes. Diabetes Care. 2012;35:2626-2630. 40. Nishimura E, Sorensen AR, Hansen BF, et al. Insulin degludec, a new generation ultra-long acting basal insulin designed to maintain full metabolic effect while minimising mitogenic potential. Diabetologia. 2010;53(Suppl 1):S388-S389. 41. Moghissi E, King AB. Individualizing insulin therapy in the management of type 2 diabetes. Am J Med. 2014;127(Suppl 1):S3-S10. 42. Bolli GB, Luzio S, Marzotti S, et al. Comparative pharmacodynamic and pharmacokinetic characteristics of subcutaneous insulin glulisine and insulin aspart prior to a standard meal in obese subjects with type 2 diabetes. Diabetes Obes Metab. 2011;13:251-257. 43. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycaemia in type 2 diabetes: a patient-centered approach. Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetologia. 2012;55:1577-1596. 44. Home PD. The pharmacokinetics and pharmacodynamics of rapidacting insulin analogues and their clinical consequences. Diabetes Obes Metab. 2012;14:780-788. 45. Luzio S, Peter R, Dunseath GJ, et al. A comparison of preprandial insulin glulisine versus insulin lispro in people with type 2 diabetes over a 12-h period. Diabetes Res Clin Pract. 2008;79: 269-275. 46. Heinemann L, Kapitza C, Starke AA, Heise T. Time-action profile of the insulin analogue B28Asp. Diabet Med. 1996;13:683-684. 47. Heinemann L, Weyer C, Rauhaus M, et al. Variability of the metabolic effect of soluble insulin and the rapid-acting insulin analog insulin aspart. Diabetes Care. 1998;21:1910-1914. 48. Mudaliar SR, Lindberg FA, Joyce M, et al. Insulin aspart (B28 aspinsulin): a fast-acting analog of human insulin: absorption kinetics and action profile compared with regular human insulin in healthy nondiabetic subjects. Diabetes Care. 1999;22:1501-1506. 49. Heise T, Nosek L, Spitzer H, et al. Insulin glulisine: a faster onset of action compared with insulin lispro. Diabetes Obes Metab. 2007;9: 746-753. 50. von Mach MA, Brinkmann C, Hansen T, et al. Differences in pharmacokinetics and pharmacodynamics of insulin lispro and aspart in healthy volunteers. Exp Clin Endocrinol Diabetes. 2002;110: 416-419. 51. Lindholm A, McEwen J, Riis AP. Improved postprandial glycemic control with insulin aspart. A randomized double-blind cross-over trial in type 1 diabetes. Diabetes Care. 1999;22:801-805. 52. Rave K, Klein O, Frick AD, Becker RH. Advantage of premealinjected insulin glulisine compared with regular human insulin in subjects with type 1 diabetes. Diabetes Care. 2006;29:1812-1817. 53. Heinemann L, Heise T, Wahl LC, et al. Prandial glycaemia after a carbohydrate-rich meal in type I diabetic patients: using the rapid acting insulin analogue [Lys(B28), Pro(B29)] human insulin. Diabet Med. 1996;13:625-629.

S37 54. Rosenfalck AM, Thorsby P, Kjems L, et al. Improved postprandial glycaemic control with insulin Aspart in type 2 diabetic patients treated with insulin. Acta Diabetol. 2000;37:41-46. 55. Holmes G, Galitz L, Hu P, Lyness W. Pharmacokinetics of insulin aspart in obesity, renal impairment, or hepatic impairment. Br J Clin Pharmacol. 2005;60:469-476. 56. Lilly USA. Humalog Mix 75/25 Prescribing Information. Available at: http://pi.lilly.com/us/humalog7525-pi.pdf. Accessed August 8, 2014. 57. Sanofi-aventis USA. Apidra Prescribing Information. Available at: http://products.sanofi.us/apidra/apidra.pdf. Accessed August 8, 2014. 58. Novo Nordisk A/S. NovoRapid Product Information. Available at http://www.novonordisk.com.au/media/PIs/NovoRapid_NovoMix_ PI3a.pdf. Accessed August 8, 2014. 59. Bretzel RG, Arnolds S, Medding J, Linn T. A direct efficacy and safety comparison of insulin aspart, human soluble insulin, and human premix insulin (70/30) in patients with type 2 diabetes. Diabetes Care. 2004;27:1023-1027. 60. Dailey G, Rosenstock J, Moses RG, Ways K. Insulin glulisine provides improved glycemic control in patients with type 2 diabetes. Diabetes Care. 2004;27:2363-2368. 61. Ross SA, Zinman B, Campos RV, Strack T. A comparative study of insulin lispro and human regular insulin in patients with type 2 diabetes mellitus and secondary failure of oral hypoglycemic agents. Clin Invest Med. 2001;24:292-298. 62. Mannucci E, Monami M, Marchionni N. Short-acting insulin analogues vs regular human insulin in type 2 diabetes: a meta-analysis. Diabetes Obes Metab. 2009;11:53-59. 63. Heller S, Bode B, Kozlovski P, Svendsen AL. Meta-analysis of insulin aspart versus regular human insulin used in a basal-bolus regimen for the treatment of diabetes mellitus. J Diabetes. 2013;5: 482-491. 64. Rys P, Pankiewicz O, Lach K, et al. Efficacy and safety comparison of rapid-acting insulin aspart and regular human insulin in the treatment of type 1 and type 2 diabetes mellitus: a systematic review. Diabetes Metab. 2011;37:190-200. 65. Siebenhofer A, Plank J, Berghold A, et al. Short acting insulin analogues versus regular human insulin in patients with diabetes mellitus. Cochrane Database Syst Rev. 2006:CD003287. 66. Singh SR, Ahmad F, Lal A, et al. Efficacy and safety of insulin analogues for the management of diabetes mellitus: a meta-analysis. CMAJ. 2009;180:385-397. 67. Kurtzhals P, Schaffer L, Sorensen A, et al. Correlations of receptor binding and metabolic and mitogenic potencies of insulin analogs designed for clinical use. Diabetes. 2000;49:999-1005. 68. Stammberger I, Seipke G, Bartels T. Insulin glulisineea comprehensive preclinical evaluation. Int J Toxicol. 2006;25:25-33. 69. Liebl A, Prusty V, Valensi P, et al. Ten years of experience with biphasic insulin aspart 30: from drug development to the latest clinical findings. Drugs. 2012;72:1495-1520. 70. Jacobsen LV, Sogaard B, Riis A. Pharmacokinetics and pharmacodynamics of a premixed formulation of soluble and protamine-retarded insulin aspart. Eur J Clin Pharmacol. 2000;56:399-403. 71. Weyer C, Heise T, Heinemann L. Insulin aspart in a 30/70 premixed formulation. Pharmacodynamic properties of a rapid-acting insulin analog in stable mixture. Diabetes Care. 1997;20: 1612-1614. 72. Hermansen K, Colombo M, Storgaard H, et al. Improved postprandial glycemic control with biphasic insulin aspart relative to biphasic insulin lispro and biphasic human insulin in patients with type 2 diabetes. Diabetes Care. 2002;25:883-888. 73. Novo Nordisk A/S. Last updated 2013. NovoLog Mix Prescribing Information. Available at: http://www.novonordisk-us.com/Images/ PDF/Novolog_Jun_29_2011.pdf. Accessed August 8, 2014. 74. Qayyum R, Wilson LM, Bolen S, et al. Last updated 2008. Comparative effectiveness, safety, and indications of insulin analogues in premixed formulations for adults with type 2 diabetes. Comparative Effectiveness Review No. 14. (Prepared by the Johns Hopkins University Evidence-based Practice Center under Contract

S38

The American Journal of Medicine, Vol 127, No 10A, October 2014

No. 290-02-0018.) Rockville, MD: Agency for Healthcare Research and Quality. Available at: www.effectivehealthcare.ahrq.gov/reports/ final.cfm. Accessed August 8, 2014. 75. Davidson JA, Liebl A, Christiansen JS, et al. Risk for nocturnal hypoglycemia with biphasic insulin aspart 30 compared with biphasic human insulin 30 in adults with type 2 diabetes mellitus: a metaanalysis. Clin Ther. 2009;31:1641-1651.

76. Niskanen L, Jensen LE, Rastam J, et al. Randomized, multinational, open-label, 2-period, crossover comparison of biphasic insulin aspart 30 and biphasic insulin lispro 25 and pen devices in adult patients with type 2 diabetes mellitus. Clin Ther. 2004;26:531-540. 77. Vaag A, Lund SS. Insulin initiation in patients with type 2 diabetes mellitus: treatment guidelines, clinical evidence and patterns of use of basal vs premixed insulin analogues. Eur J Endocrinol. 2012;166:159-170.