New insulins and insulin therapy

Abstract

While the pharmacokinetic properties of the long-acting insulin analogues, insulin glargine and insulin detemir, clearly demonstrate clinical benefits over NPH insulin in terms of reduced frequency of overall and nocturnal hypoglycaemic events and lowered glycaemic variability, several new insulin analogues with even longer and smoother time-action profiles are being developed and explored. Among these, insulin degludec is the first to enter the market. The ultra-long action profile of this insulin analogue is mainly the result of a slow and stable release of insulin degludec monomers from soluble multihexamers that form after subcutaneous administration. In the last year, the results of the first phase II trials with insulin degludec have been published, where the efficacy and safety of this basal insulin analogue have been compared with insulin glargine in patients with type 1 (T1D) and type 2 (T2D) diabetes. With regard to short-acting insulins for mealtime insulin supplementation, there is also a search for new insulins with a more rapid onset of action, and hence improved postprandial glucose control. LinjetaTM (formerly named VIAject) is a human insulin formulation with faster insulin absorption and action than insulin lispro and regular human insulin. A new drug application to market LinjetaTM as a treatment for diabetes was submitted to the US Food and Drug Administration (FDA) by the manufacturer at the beginning of 2010. However, the FDA’s review cycle, which was announced late in 2010, concluded that the application could not be approved in its present form. The FDA’s complete response letter included comments related to clinical trials, statistical analysis and chemistry, manufacturing and controls. The FDA requested that the company should conduct two new phase III clinical trials using the commercial formulation, one in patients with T1D and the other in patients with T2D, to establish efficacy and safety with regard to hypoglycaemia and toleration. Co-administration of hyaluronidase together with insulin is another novel concept which has been proven to accelerate the absorption of both insulin lispro and regular human insulin in healthy subjects. In this review, we comment on two recent publications in which the pharmacokinetic and pharmacodynamic characteristics of LinjetaTM and co-formulations of prandial insulins with recombinant human hyaluronidase were investigated in patients with T1D. Lastly, a recent review on the ongoing discussion regarding insulin treatment and cancer risk is included. Birkeland KI 1 , Home PD 2 , Wendisch U 3 , Ratner RE 4 , Johansen T 5 , Endahl LA 5 , Lyby K 5 , Jendle JH 6 , Roberts AP 7 , DeVries JH 8 , Meneghini LF 9 1 Oslo University Hospital and Faculty of Medicine, Oslo, Norway, 2 Newcastle University, Newcastle upon Tyne, UK, 3 Group Practice in Internal Medicine and Diabetology, Hamburg, Germany, 4 Medstar Research Institute, Washington, DC, USA, 5 NovoNordisk A/S, Soeborg, Denmark, 6 Örebro University Hospital, Örebro, Sweden, 7 Royal Adelaide Hospital, Adelaide, SA, Australia, 8 University of Amsterdam, The Netherlands, and 9 University of Miami Miller School of Medicine, Miami, FL, USA Diabetes Care 2011; 34 : 661–5 Background: To investigate the efficacy and safety of two formulations of insulin degludec (IDeg), administered once daily together with mealtime insulin aspart in patients with T1D. Comparison was made with a similar basal-bolus regimen consisting of insulin glargine (IGlar) and aspart. Methods: Adult patients with T1D (mean A1c 8.4%) were randomised (open-label) to be treated for 16 weeks with once-daily (evening) subcutaneous injections of either of two different insulin degludec formulations, IDeg(A) (600 μmol/l; n = 59), IDeg(B) (900 μmol/l; n = 60), or with IGlar (n = 59). All patients received insulin aspart at mealtimes. Results: At 16 weeks, mean A1c was comparable for all three basal insulin therapies: IDeg(A) 7.8 ± 0.8%, IDeg(B) 8.0 ± 1.0% and IGlar 7.6 ± 0.8%.. Similarly, there was no difference in mean fasting plasma glucose across treatment groups (8.3 ± 4.0, 8.3 ± 2.8 and 8.9 ± 3.5 mmol/l, respectively). In comparison with IGlar, the estimated mean rate of confirmed hypoglycaemia (<3.1 mmol/l) was 28% lower for IDeg(A) [rate ratio (RR) 0.72, 95% confidence interval (CI) 0.52–1.00] and 10% lower for IDeg(B) [RR 0.90 (0.65–1.24)], and rates of nocturnal hypoglycaemia were reduced by 58% with IDeg(A) [RR 0.42 (0.25–0.69)] and by 29% with IDeg(B) [RR 0.71 (0.44–1.16)]. Mean total daily insulin doses remained essentially unchanged from baseline values in all treatment arms. The overall rates and nature of adverse events were comparable between groups. Conclusions: IDeg is a safe and well tolerated basal insulin in patients with T1D, providing comparable glycaemic control to IGlar at similar doses but with lower frequency of hypoglycaemia. Zinman B 1 , Fulcher G 2 , Rao PV 3 , Thomas N 4 , Endahl LA 5 , Johansen T 5 , Lindh R 5 , Lewin A 6 , Rosenstock J 7 , Pinget M 8 , Mathieu C 9 1 Samuel Lunenfeld Research Institute, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada, 2 Royal North Shore Hospital and University of Sydney, Sydney, NSW, Australia, 3 Nizam’s Institute of Medical Sciences University, Hyderabad, India, 4 Christian Medical College, Vellore, India, 5 Novo Nordisk, Soeborg, Denmark, 6 National Research Institute, Los Angeles, CA, USA, 7 Dallas Diabetes and Endocrine Center at Medical City, Dallas, TX, USA, 8 University Hospital Strasbourg, France, and 9 UZ Gasthuisberg Katholieke Universiteit Leuven, Leuven, Belgium Lancet 2011; 377 : 924–31 Background: To assess the efficacy and safety of insulin degludec administered once daily or three times a week in patients with T2D. Comparison was made with insulin glargine injected once daily. Methods: Insulin-naive patients with T2D (age 18–75 years), inadequately controlled with oral antidiabetic drugs (A1c 7%–11%), were eligible in this 16-week, multicentre, open-label trial. Sixty-two patients were randomised to administer insulin degludec (900 μmol/l) three times a week (starting dose 20 U per injection; 1 U = 9 nmol), 60 patients to receive insulin degludec (600 μmol/l) once daily (starting dose 10 U; 1 U = 6 nmol), 61 patients to receive insulin degludec (900 μmol/l) once daily (starting dose 10 U; 1 U = 9 nmol) and 62 to administer insulin glargine once daily (starting dose 10 U; 1 U = 6 nmol). All patients were on combination therapy with metformin. The primary outcome was HbA1c following 16 weeks of treatment. Results: At 16 weeks, the mean HbA1c values were comparable in the four treatment groups: 7.3% (SD 1.1), 7.4% (1.0), 7.5% (1.1) and 7.2% (0.9), respectively. The estimated mean HbA1c difference in comparison with insulin glargine was 0.08% (95% CI –0.23 to 0.40) for insulin degludec administered three times a week, and 0.17% (–0.15 to 0.48) and 0.28% (–0.04 to 0.59), respectively, for the 600 μmol/l and 900 μmol/l insulin degludec formulations given once daily. The overall rate of hypoglycaemic events was low, and the frequency and pattern of adverse events were similar across the treatment groups. Conclusions: Insulin degludec provides comparable glycaemic control and safety profile to insulin glargine in patients with T2D, and – according to its ultra-long action profile – it might allow prolonged dosing intervals. Heise T 1 , Tack CJ 2 , Cuddihy R 3 , Davidson J 4 , Gouet D 5 , Liebl A 6 , Romero E 7 , Mersebach H 8 , Dykiel P 8 , Jorde R 9 1 Profil Institut für Stoffwechselforschnung, Neuss, Germany, 2 Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands, 3 International Diabetes Center, Park Nicollet, Minneapolis, MA, USA, 4 Department of Medicine, Division of Endocrinology, University of Texas, Southwestern Medical School, Dallas, TX, USA, 5 Hopital Saint Louis, Centre Hospitalier de La Rochelle, La Rochelle, France, 6 Center for Diabetes and Meabolism, Fachklinik, Bad Heilbrunn, Germany, 7 University of Valladolid, Instituto de Endocrinologia y Nutricion, Valladolid, Spain, 8 Novo Nordisk A/S, Soeborg, Denmark, and 9 Institute of Clinical Medicine, University of Tromsoe, Tromsoe, Norway Diabetes Care 2011; 34 : 669–74 Background: To compare the safety and efficacy of two formulations of a soluble co-formulation of insulin degludec and insulin aspart (70%/30% v/v and 55%/45% v/v, respectively) and insulin glargine in insulin-naive patients with T2D inadequately controlled with oral antidiabetic drugs. Methods: Patients with T2D (mean age 59.1 years, A1c 8.5%, body mass index 30.3 kg/m2) were randomly allocated (open-label) to 16 weeks of treatment with once-daily injection (before evening meal) of insulin degludec/aspart 70/30 (n = 59), insulin/aspart 55/45 (n = 59), or insulin glargine (n = 60). All patients received combination therapy with metformin. Insulin doses were titrated to a fasting plasma glucose target of 4–6 mmol/l. Results: At 16 weeks, mean A1c had decreased to comparable levels across the treatment groups: insulin degludec/aspart 70/30 7.0%, insulin degludec/aspart 55/45 7.2% and insulin glargine 7.1%. There was a similar relative proportion of patients who improved their A1c below 7.0% without experiencing confirmed hypoglycaemia (plasma glucose < 3.1 mmol/l) over the last 4-week period of the trial (51%, 47% and 50% in the three treatment groups, respectively). Mean 2-h post-dinner plasma glucose increase was less prominent for insulin degludec/aspart 70/30 (0.13 mmol/l) and insulin degludec/aspart 55/45 (0.24 mmol/l) than for insulin glargine (1.63 mmol/l), whereas fasting plasma glucose was similar (6.8 mmol/l, 7.4 mmol/l and 7.0 mmol/l, respectively). Estimated rates of confirmed hypoglycaemic events were lower for insulin degludec/aspart 70/30 (1.2 events per patient-year) and insulin glargine (0.7 events per patient-year) than for insulin degludec/aspart 55/45 (2.4 events per patient year). Likewise, nocturnal hypoglycaemia occurred more frequently with insulin degludec/aspart 55/45 (27 events) than with insulin degludec/aspart 70/30 (1 event) and insulin glargine (3 events). Conclusions: The 70%/30% (v/v) co-formulation of insulin degludec and insulin aspart administered once daily provided comparable overall glycaemic control and hypoglycaemia rates compared with insulin glargine but better postprandial glucose control after dinner. Comment: Insulin degludec is a soluble, basal insulin analogue with a smooth and stable pharmacokinetic profile, with a duration of action exceeding 24 h (1) and with low within-subject variability (2). In the referenced reports, the safety and efficacy of different formulations of insulin degludec were compared with insulin glargine in short-term, proof-of-concept trials. In patients with T1D, once-daily injection of insulin degludec together with insulin aspart at mealtimes resulted in essentially the same glycaemic control as that observed when insulin glargine was used for basal insulin supplementation. The rates of confirmed hypoglycaemic episodes were lower for insulin degludec; this was most readily observed during night-time, and in those patients who were allocated to use the 600 μmol/l formulation of insulin degludec [according to the authors, the higher strength insulin degludec formulation (900 μmol/l) has therefore been discontinued from further clinical development]. Similarly, in patients with T2D, insulin degludec administered once daily as add-on to metformin did not differ from insulin glargine in terms of glycaemic control or risk of hypoglycaemia or other adverse events. Interestingly, demonstrating its prolonged duration of action, insulin degludec (900 μmol/l) also provided comparable glycaemic outcomes when given only three times a week, although with a tendency to more frequent hypoglycaemic episodes. The latter finding may have been due to the fact that approximately twice as much insulin was administered per injection when insulin degludec was given three times a week in comparison with the once-daily insulin regimens. Another intriguing feature of insulin degludec is that it can be co-formulated with a rapid-acting insulin analogue (pre-mixed insulin). In the study by Heise et al., the two tested formulations of insulin degludec and insulin aspart (relative proportions 70/30 and 55/45, respectively) injected once daily provided comparable overall glycaemic control to insulin glargine. The postprandial glucose excursion after dinner was less pronounced with the two insulin degludec formulations than with glargine. This was perhaps an expected finding, given the prandial, rapid-acting component of insulin aspart in these formulations. Notably, the rate of hypoglycaemic events, and especially the frequency of nocturnal hypoglycaemia, was considerably higher with the 55/45 degludec/aspart formulation. Based on this finding, the clinical development of this co-formulation has been stopped. In summary, the results of these phase 2 trials are promising and indicate that insulin degludec may allow more flexible insulin regimens with less frequent insulin injections in patients with T2D. Moreover, the findings suggest that insulin degludec may have the potential to further reduce the risk of hypoglycaemia, in particular in patients with T1D as well as in patients with T2D necessitating more complex insulin regimens. Additional data from longer term studies are much awaited to clarify in more detail the efficacy and safety characteristics and the potential clinical benefits of this novel basal insulin analogue. Philotheou A 1 , Arslanian S 2 , Blatniczky L 3 , Peterkova V 4 , Souhami E 5 , Danne T 6 1 University of Cape Town Diabetes Clinical Trials Unit, New Groote Schuur Hospital, Cape Town, South Africa, 2 Divisions of Pediatric Endocrinology, Metabolism and Diabetes Mellitus, and Weight Management and Wellness, Children’s Hospital of Pittsburgh, University of Pittsburgh Medical Center, Pittsburgh, PA, USA, 3 Buda Children’s Hospital, Budapest, Hungary, 4 Endocrinological Scientific Centre, Russian Academy of Science, Moscow, Russia, 5 Sanofi-aventis, Antony, France, and 6 Children’s Hospital auf der Bult, Hannover, Germany Diabetes Technol Ther 2011; 13 : 327–34 Background: To evaluate the efficacy and safety of insulin glulisine compared with insulin lispro as part of a basal-bolus regimen in children and adolescents with T1D. Methods: Children and adolescents aged 4–17 years (n = 572) using insulin glargine or NPH insulin as basal insulin were enrolled in a 26-week, multicentre, open, centrally randomised, parallel-group study. Subjects were randomised to receive glulisine (n = 277) or lispro (n = 295) 0–15 min before a meal. Results: Both groups exhibited similar A1c levels at baseline. The adjusted mean change in A1c from baseline to endpoint was+0.10 ± 0.08% in the glulisine group and +0.16 ± 0.07% in the lispro group. The difference in the adjusted means for the change from baseline in A1c between the treatments was equal to −0.06% (95% CI −0.24 to 0.12), confirming non-inferiority of glulisine vs. lispro. Similar effects on A1c levels were reported in both groups after 12 and 26 weeks. For all age groups, the percentage of subjects achieving the American Diabetes Association recommended A1c target was significantly higher (p = 0.039) with glulisine (38.4%) than with lispro (32.0%). This difference was most pronounced in adolescents (13–17 years). At baseline, the mean blood glucose profiles were comparable for all three time points in both groups. At endpoint, only the mean pre-breakfast blood glucose was significantly lower in the glulisine group. Most subjects had three to four bolus insulin injections at baseline, which remained stable throughout the study. No significant differences between the two treatment arms were reported for the monthly events rate per patient of all, severe, nocturnal, or severe nocturnal symptomatic hypoglycaemia from month 4 to treatment to endpoint for glulisine and lispro, respectively. Frequency and type of adverse events, serious adverse events, or hypoglycaemia reported as serious adverse events were similar between the groups. Conclusions: In children and adolescents withT1D, glulisine is non-inferior to lispro in terms of A1c change, and is similarly well tolerated to lispro. Heinemann L 1,2 , Hompesch M 2 , Flacke F 3 , Simms P 3 , Pohl R 3 , Albus K 3 , Pfützner A 4 , Steiner S 3 1 Profil Institut für Stoffwechselforschung, Neuss, Germany, 2 Profil Institute for Clinical Research Inc., San Diego, CA, USA, 3 Biodel Inc., Danbury, CT, USA, and 4 IKFE, Institute for Clinical Research and Development, Mainz, Germany J Diabetes Sci Technol 2011; 5 : 681–6 Background: To compare postprandial glucose control in patients with T1D following administration of VIAjectTM (LinjetaTM), insulin lispro and regular human insulin. Methods: On three occasions, and with normalised preprandial glucose control, patients (n = 18) received an individually tailored dose of each insulin immediately before a standardised mixed meal. Results: Insulin absorption was faster following injection of VIAjectTM than after administration of insulin lispro and human regular insulin, the time to half-maximum serum insulin being 13.1 ± 5.2 min, 25.4 ± 7.6 min and 38.4 ± 19.5 min, respectively (VIAjectTM vs. insulin lispro p = 0.001; VIAjectTM vs. regular human insulin p < 0.001; and insulin lispro vs. regular human insulin p < 0.001). Peak postprandial blood glucose (0–180 min) was lower with VIAjectTM (157 ± 30 mg/dl, p = 0.002) and insulin lispro (170 ± 42 mg/dl; p = 0.008) compared with regular human insulin (191 ± 46 mg/dl). The difference between maximum and minimum glucose levels was smaller after administration of VIAjectTM (70 ± 17 mg/dl) than after injection of either insulin lispro (89 ± 18; p = 0.011) or regular human insulin (91 ± 33 mg/dl; p = 0.007), and the area under the blood glucose curve was 23% smaller with VIAjectTM than with regular human insulin (p < 0.001). Conclusions: The more rapid absorption of VIAjectTM results in less pronounced postprandial glucose excursions. Hompesch M 1 , Muchmore DB 2 , Morrow L 1 , Vaughn DE 2 1 Profil Institute for Clinical Research, Chula Vista, CA, USA, and 2 Halozyme Therapeutics, San Diego, CA, USA Diabetes Care 2011; 34 : 666–8 Background: To assess the effect of co-administration of recombinant human hyaluronidase (rHuPH20) on safety, pharmacokinetics and pharmacodynamics of insulin lispro and human regular insulin in patients with T1D. Methods: Patients (n = 22) were given individually titrated optimum doses of insulin lispro or human regular insulin with or without co-administration of rHuPH20 before a standardised liquid test meal. Results: With co-administration of rHuPH20, the early (0–60 min) insulin uptake increased by 54% (p = 0.0011) for insulin lispro and by 206% (p < 0.0001) for human regular insulin, compared with administration of the respective insulin without rHuPH20. Mean peak blood glucose decreased 26 mg/dl for insulin lispro (p = 0.002) and 24 mg/dl for human regular insulin, and the total postprandial glycaemic excursions (area under the curve for blood glucose above 140 mg/dl) were reduced by 79% (p = 0.09) and 85% (p = 0.049) for insulin lispro and human regular insulin, respectively. Frequency of hypoglycaemia was similar for insulin lispro with or without co-administration of rHuPH20, whereas it was reduced when human regular insulin was administered together with rHuPH20. Conclusions: Co-administration of rHuPH20 with insulin lispro and human regular insulin resulted in accelerated insulin absorption and improved postprandial blood glucose control. Comment: The largest paediatric T1D study with rapid-acting insulin analogues to date showed comparable efficacy in a head-to-head comparison between lispro and glulisine. However, it raised the question whether all currently available fast acting insulin analogues are clinically equivalent, or if certain patient populations may benefit using one over the other. Notably, despite the introduction of the fast acting insulin analogues, many patients remain unable to achieve effective control of postprandial glucose excursions. Thus, there is a clinical need for insulin formulations that are absorbed even more rapidly from the subcutaneous injection site, and mimic the normal meal-induced insulin response. LinjetaTM is a special formulation of human insulin which promotes monomerisation of insulin. Previous studies in healthy subjects have demonstrated that LinjetaTM has a faster onset of action compared with regular human insulin and insulin lispro (3,4). The study by Heinemann et al. confirms similar findings in patients with T1D, and also that the faster rate of absorption of LinjetaTM is mirrored by a less pronounced increase in postprandial glucose levels. Nevertheless, the FDA request additional studies prior to approval, as mentioned above. The report by Hompesch and co-workers describes another interesting approach to accelerate the absorption of insulin, namely by co-administration of prandial insulins with hyaluronidase. For more than six decades, hyaluronidase has been utilised as an adjuvant to enhance the absorption and dispersion of other injected molecules. This enzyme catalyses the breakdown of hyaluronan polymers in the subcutaneous space, leading to facilitated permeation of co-administered compounds (5,6). In the case of insulin, the depolymerisation of hyaluronan mediates a dilution effect that favours dissociation of insulin hexamers and hence enhances the rate of insulin absorption (7). In addition, the insulin molecules are dispersed over a larger capillary bed, which facilitates their absorption. In the current experimental study in patients with T1D, co-administration of rHuPH20 enhanced the absorption of both regular human insulin and insulin lispro, leading to a faster and greater maximum insulin exposure and a corresponding lowering of the postprandial glucose excursions. While the findings are encouraging, larger long-term trials are warranted to clarify the potential clinical benefit of this approach of constructing ultra-fast insulins. Gough SC 1 , Belda-Iniesta C 2 , Poole C 3 , Weber M 4 , Russell-Jones D 5 , Hansen BF 6 , Mannucci E 7 , Tuomilehto J 8 1 Oxford Centre for Diabetes Endocrinology and Metabolism and Oxford (NIHR) Biomedical Research Centre, Churchill Hospital, Oxford, UK, 2 IdiPAZ, University Hospital La Paz, Madrid, Spain, 3 Department of Epidemiology, Pharmatelligence, Pharma Research Centre, Cardiff Medicentre, University Hospital of Wales, Heath Park, Cardiff, UK, 4 I Medical Department, University Hospital Mainz, Mainz, Germany, 5 Royal Surrey County Hospital and University of Surrey, Guildford, UK, 6 Diabetes Biology, Novo Nordisk A/S, Novo Nordisk Park, Måløv, Denmark, 7 Diabetes Agency, Department of Emergency, Medicine and Surgery, Careggi Teaching Hospital, Florence, Italy, and 8 Department of Public Health, University of Helsinki, Helsinki, Finland Adv Ther 2011; 28 (Suppl 5): 1–18 Background: The possibility of certain insulin treatments having the potential to modify cancer development and prognosis was raised again in 2009, following publication of several epidemiological studies addressing this issue. Together with the possibility that diabetes itself might be linked to cancer, an exchange of expert views and knowledge was essential in order to enhance understanding on this subject among those treating diabetes and cancer, and those developing diabetes therapies. Methods: A European meeting was convened with participants who are experts in the fields of endocrinology, oncology, epidemiology and insulin analogue design and investigation. The experts were invited to present on relevant topics, with open discussions held after each presentation. Results: Some epidemiological studies have raised concern over the potential mitogenic properties of certain insulin analogues, although confounding factors render interpretation controversial. Meanwhile, pharmacological studies, and a consideration of cancer pathophysiology, implicate increased insulin-like growth factor 1 receptor affinity and/or deranged insulin receptor interaction/signalling properties as possible causes for concern with some insulin analogues. However, interpretation of the pharmacological evidence is confounded by the array of test systems and methodologies used, and by studies frequently succumbing to methodological pitfalls. Most available insulin analogues do not differ in their receptor interaction response profile to human insulin, and for those that do there are reasons to question any potential clinical relevance. Yet, it is preferred that new experimental models are devised that can better determine the likely clinical consequences of any variance in receptor response profile vs. human insulin. Conclusions: We need more data to increase our understanding of this issue, with close cooperation and communication between diabetologists, epidemiologists, oncologists and insulin engineers. Comment: This review summarises the most recent papers on the insulin treatment and cancer debate. Although the proceedings have to be interpreted with caution as the panel was convened by a single pharmaceutical company (Novo Nordisk) it allows a quick update. Despite the growing knowledge base the authors conclude that current gaps in our understanding do not allow clear recommendations on how to discuss this issue adequately with patients to reach an ‘informed choice’. TD has received honoria for speaking engagements or advisory panel participation and received research support from several companies: AstraZeneca, Bayer, Bristol Myers Squibb, DexCom, Eli Lilly, GlaxoSmithKline, Johnson & Johnson, Medtronic, Nilimedix, NovoNordisk, Roche, Sanofi, Unomedical, Ypsomed. JB received consulting fees as a member of scientific advisory boards from Abbott Diabetes Care, AstraZeneca, Medtronic, D-Medical and MSD; and received honoria for lectures from Medtronic, Abbott Diabetes Care, Sanofi-Aventis and MSD.