COPYRIGHT MEDINEWS (DIABETES) LIMITED

Transcription

1 Levy&Ahmad P /2/04 14:58 Page 1 Pharmacology of insulin BUSHRA AHMAD Abstract Diabetes is associated with microvascular and macrovascular complications leading to significant morbidity and mortality. Glycaemic control is important to prevent and delay the progression of these complications. An ideal insulin regimen in patients with diabetes would mirror the 24-hour insulin profile of a non-diabetic person and thereby prevent hyperglycaemia without inducing hypoglycaemia. This has, until recently, proved difficult to reproduce by regular subcutaneous insulin injections due to the inherent pharmacokinetic properties of the available insulins. Normoglycaemia was rarely achieved without hypoglycaemia compromising the quality of patients lives. The advent of the new long- and short-acting insulin analogues are expected to both improve glycaemic control leading to a reduction in diabetes-related complications and reduce the incidence of hypoglycaemia thereby offering patients a better lifestyle. Br J Diabetes Vasc Dis 2004;4:10 14 Key words: insulin, pharmacology, analogues, lispro, aspart, glargine, detemir. Introduction Insulin, from the pancreatic beta-cells, plays a crucial role in glucose homeostasis. It acts on the liver to inhibit glycogenolysis and gluconeogenesis, and in the periphery to stimulate glucose utilisation by muscle and fat. An ideal insulin treatment regimen for people with diabetes would reinstate the normal daily insulin profile (figure 1) and prevent hyperglycaemia without inducing hypoglycaemia. However, in contrast to the physiological delivery of insulin into the hepatic portal circulation, subcutaneously injected insulin provides relatively less insulin to the liver than the periphery. Hence, the normal balance of hepatic and peripheral effects of insulin cannot be exactly restored by conventional injections. Correspondence to: Dr Bushra Ahmad Department of Diabetes, Oxford Centre for Diabetes, Endocrinology and Metabolism, Churchill Hospital, Oxford, OX3 7LJ, UK. Tel: +44 (0) ; Fax: +44 (0) [email protected] Bushra Ahmad Thus, a selection of different insulin structures, formulations and delivery regimens have been designed to address this issue. Structure The insulin molecule consists of two polypeptide chains, the A chain (21 amino acids) and the B chain (30 amino acids) (figure 2). It exists as monomers in dilute solutions such as the circulation but as hexamers (six monomers which self-associate in conjunction with two zinc ions) in crystals and in the beta-cell secretory granule. Human insulin differs from porcine insulin at position B30 (threonine in man versus alanine in pig), and from bovine insulin Figure 1. The normal daily profile of plasma insulin Plasma insulin mu/l Breakfast Lunch Dinner Snack Adapted with permission from Day et al. Br J Cardiol 2003;10: THE BRITISH JOURNAL OF DIABETES AND VASCULAR DISEASE

2 Levy&Ahmad P /2/04 14:58 Page 2 Figure 2. Amino acid sequences of native human insulin, the long-acting analogues glargine and detemir, and the rapid-acting monomeric analogues lispro and aspart. Sequences shown as single letter amino acid code However, there is no measurable difference in biological potency of human and animal insulins. 3 Human insulin is absorbed slightly faster from a subcutaneous injection site 1 than porcine insulin, which may be explained by greater hydrophilicity of the human insulin molecule. 2 Human insulin FVNQHLCGSHLVEALYLVCGERGFFYTPKT Glargine GIVEQCCTSICSLYQLENYCG FVNQHLCGSHLVEALYLVCGERGFFYTPKTR R Detemir FVNQHLCGSHLVEALYLVCGERGFFYTPK Lispro FVNQHLCGSHLVEALYLVCGERGFFYTK PT Aspart Tetradecanoyl FVNQHLCGSHLVEALYLVCGERGFFYTDKT Reproduced with permission from Day et al. Br J Cardiol 2003;10: at positions A8 and A10 (respectively threonine and isoleucine in man versus alanine and valine in the ox). Animal and human insulin are therefore nearly homologous and the differences are not crucial to the receptor binding or action of insulin. The carboxyl end of the B chain of the insulin molecule is involved in self-association of two insulin molecules into dimers. Therefore, the self-association properties of human and porcine insulin may be different. 1 B30 threonine adds an extra hydroxyl group to the human insulin molecule, increasing its hydrophilic properties and decreasing the lipophilic properties when compared to pork insulin. Therefore, human insulin is more soluble in aqueous solutions than porcine insulins. 1 The spherical structure of the insulin monomer is important for its interaction with the insulin receptor: porcine insulin has a slightly different threedimensional structure to human insulin. 2 Short-acting insulins Standard Insulin in neutral concentrated solutions in vials is mostly zinccontaining hexamers. For absorption to occur after subcutaneous injection, insulin needs to be in a monomeric form. As dissociation of the insulin hexamers to monomers is slow, so is the rate of absorption from the subcutaneous site. 4 For example, it can take approximately 1 2 hours for absorption of subcutaneous regular short-acting insulin. 5 This results in a slow rise to peak insulin concentrations and hence suboptimal plasma insulin levels in the early phase of glucose absorption after a meal, leading to postprandial hyperglycaemia. The insulin levels also fall slowly after the peak, extending the period of elevated insulin concentrations with the risk of later hypoglycaemia. 5 The influence of short-acting insulins on post-prandial hyperglycaemia is improved by subcutaneous administration 30 minutes to one hour before eating 6 but the delayed risk of hypoglycaemia remains. Therefore, the extent of the insulin crystallisation and the ease of dissociation play an important role in determining the rate of absorption after subcutaneous injection. Rapid-acting analogues Insulin hexamer formation is an obstacle to mimicking physiological insulin profiles after subcutaneous injection. This led to the hypothesis that a reduced tendency of insulin to self-associate would enable faster absorption and shorter duration of action. Using this rationale monomeric insulin analogues were created by alteration of selected amino acid residues in regions vital for self-association. 7 After subcutaneous injecting they are absorbed 2 3 times faster than the regular non-dissociating hexameric insulins 7,8 giving a prompt rise in circulating insulin and the opportunity to inject just before a meal. The rates of disappearance are also 2 3 times faster with no lag phase, 8 so monomeric insulins result in a more physiological prandial plasma insulin profile. The two currently available rapid-acting insulin analogues are insulin lispro (LysB28, ProB29) and insulin aspart (AspB28) (figure 2). Insulin lispro (LysB28, ProB29-human insulin) Insulin lispro, the first insulin analogue to be used clinically, has the B28 (proline) and B29 (lysine) residues inverted, resulting in reduced self-association. 7,9 This inverted sequence was found and thought to be responsible for the lower degree of self-association of insulin-like growth factor (IGF-1) and insulin lispro was engineered to incorporate this pharmacological advantage. 10 Both human insulin and insulin lispro exist in their respective formulations as hexamers that are stabilised with zinc and phenolic preservatives to assure two years of shelf life at 4 C. However, insulin lispro differs in that its hexamer complex rapidly dissociates VOLUME 4 ISSUE 1. JANUARY/FEBRUARY

3 Levy&Ahmad P /2/04 14:58 Page 3 into monomeric subunits, as the phenolic ligands diffuse away in the interstitial space resulting in a plasma absorption profile indistinguishable from that of pure monomeric insulin. 4,11 Immunogenicity, receptor binding and metabolic effects of insulin lispro are similar to human insulin Lispro does however have a slightly elevated IGF-1 receptor affinity. 14 Toxicology studies, which included reproductive and developmental periods in animals, showed no long-term toxicity 4 and pooled data from clinical studies showed no increase in adverse events or progression of diabetic complications. 15 The rapid absorption, onset and short duration of action of lispro results in a more precise insulin profile at mealtimes with a peak approximately one hour after subcutaneous injection and rapid elimination. After intravenous injection 'regular' human insulin and insulin lispro have identical pharmacokinetics and pharmacodynamics. 11,16 Glycaemic control on insulin lispro is comparable 17 or improved 18,19 compared to 'regular' short-acting insulin. Improved glycaemic control appears to be best achieved with multiple injection regimens where the basal dose is optimised, 20,21 although this can be associated with increased hypoglycaemia. 22 Insulin lispro can be mixed with neutral protamine Hagedorn (NPH) without any pharmacokinetic changes. 23 Several studies have also shown improved post-prandial glucose levels in type 1 17,24,25 and type 2 diabetes 26 including adolescents. 27 This improvement was at the expense of increased fasting and preprandial glucose levels in one study, 25 although fasting glycaemic control was not significantly different in a meta-analysis. 17 Insulin lispro reduced hypoglycaemia in several studies compared to regular human insulin, especially at night-time. 21,24,28 Severe hypoglycaemia was also reduced, 29 possibly due to purportedly enhanced hepatic sensitivity to glucagon with insulin lispro. 30 However, the counterregulatory hormone responses after insulin lispro compared to human insulin and porcine insulin were similar during a stepwise euglycaemic/hypoglycaemic clamp. 31 Further studies showed the counterregulatory hormone responses on insulin lispro were improved 21 or relatively conserved and equivalent to physiological responses to hypoglycaemia in patients on regular insulin. 32,33 Although the optimal time for injecting insulin lispro is immediately before the meal, it can be injected up to 15 minutes after a meal. 34 Treatment satisfaction and flexibility scores on insulin lispro were improved compared to regular human insulin. 35 Insulin lispro appears to be as safe as regular human insulin in pregnant women with type 1, type 2 and gestational diabetes. 36 It does not cross the placenta after a single standard dose 37 and does not cause adverse outcomes in pregnancy. 36 Concerns about progression of retinopathy have not been demonstrated, 38 but further prospective trials are required to confirm these findings. Insulin aspart (Novorapid) Insulin aspart is formed by replacing the proline at B28 with a negatively charged aspartic acid residue. This switch results in reduced self-association of the insulin molecule. Insulin aspart initially exists as hexamers, which after subcutaneous injection Table 1. Approximate pharmacokinetic parameters of traditional insulin preparations after subcutaneous injection of an average patient dose Human Onset of Peak of Duration of insulin type action action action Regular minutes 2 4 hours 6 8 hours NPH 1 3 hours 5 7 hours hours Lente 1 3 hours 4 8 hours hours Ultralente 2 4 hours 8 14 hours hours Regular human insulin is shown for comparison. Adapted from Hirsch I. Med Clin N Am 1998;82: rapidly dissociate into dimers and monomers. 4 Aspart resembles human insulin in receptor binding, potency 39 and metabolic and mitogenic effects. 14 No safety issues were raised by a range of experiments in which insulin aspart and human insulin exhibited similar pharmacological profiles. 40 It was associated with an increase in cross-reactive insulin antibodies, which subsequently fell towards baseline values, without any apparent effects on efficacy or safety. 41 After subcutaneous administration insulin aspart is absorbed faster, than 'regular' human insulin, with higher peak insulin concentrations and a reduced duration of action. 42,43 Aspart improves post-prandial glucose control 44 and in some studies it improves glycaemic control 44,45 compared to regular insulin in patients with type 1 diabetes. It has been associated with reduced night-time and severe hypoglycaemia in some 45 but not all studies. 44 The counterregulatory and symptomatic responses to acute hypoglycaemia between patients on insulin aspart and regular human insulin are similar. 46 Aspart can be administered immediately prior to a meal or postprandially 47 and treatment satisfaction and quality of life scores were significantly improved compared to 'regular' human insulin. 45 No adverse effects were seen with insulin aspart 44 but it is of note that no safety data in pregnancy are available. It is likely that insulin aspart, like insulin lispro, will require multiple injection regimens where the basal dose is optimised. Insulin aspart can be mixed with NPH without alteration of its pharmacokinetics. 48 Long-acting insulins The first successful insulin preparation with a prolonged action was protamine insulin followed by the Lente crystalline zincinsulin series. Classically long-acting insulins have been formulated as crystalline or amorphous suspensions obtained by addition of small amounts of zinc that at high insulin concentrations and neutral ph increase the self-association into dimers and hexamers. 8 The resulting insulin is absorbed slowly from the injected depot. The current long-acting insulins are Ultralente (Ultratard), the longest acting human insulin-zinc crystalline suspension, NPH (or isophane) insulin, an intermediate acting insulin protamine suspension and Lente (70/30 mixture of ultralente with semilente), an intermediate acting insulin (table 1) THE BRITISH JOURNAL OF DIABETES AND VASCULAR DISEASE

4 Levy&Ahmad P /2/04 14:58 Page 4 Key messages Physical and chemical properties of insulin have been used to alter subcutaneous absorption Recombinant DNA technology has led to the production of novel insulin analogues The new short-acting insulin analogues allow more flexibility with rapid absorption, onset and short duration of action The new long-acting insulin analogues allow a more physiological basal insulin profile Crystalline insulin suspensions have a number of disadvantages. Ultratard, although it has a long duration of action, has a high intersubject and intrasubject variation of absorption. This is partly due to suboptimal mixing of the insulin suspension. NPH and Lente are usually too short-acting for a once a day dosing. 4,49 NPH peaks in action 5 7 hours after injection and so the night-time dose can cause nocturnal hypoglycaemia in the early hours when insulin sensitivity is often increased. Reducing the evening dose, however, results in raised morning blood glucoses due to a relative reduction in insulin sensitivity between hours (known as the dawn phenomenon). 50 Long-acting analogues To achieve physiological insulin profiles ideal basal insulins should have a 24-hour peakless activity profile allowing oncedaily dosage, with little variability in blood glucose excursions. 49 Two approaches are being used to develop new long-acting insulin analogues. The first approach changes the isoelectric point (the ph at which insulin is least soluble and precipitates) resulting in an insulin molecule that is soluble at acidic ph (and hence eliminates the variability problems with suspensions) but precipitates in subcutaneous tissues where the ph is near neutral. Insulin glargine (A21 Gly, B31 Arg, B32 Arg) was developed using this method (figure 2). It is a long-acting human insulin analogue that has a near peakless action for at least 24 hours with reduced incidence of hypoglycaemia. 51 The addition of two positively charged arginine molecules to the C-terminus of the B- chain results in a more neutral ph. To confer stability asparagine at position 21 in the A chain was replaced by glycine. 4,49 Insulin glargine is well tolerated with improved treatment satisfaction scores in both type 1 and type 2 diabetes. It elicits less hypoglycaemia, especially nocturnal hypoglycaemia, than NPH insulins with similar levels of improved glycaemic control. 52,53 Initial safety concerns arose after a study in human osteosarcoma cells showed a 6 8-fold IGF-1 receptor binding affinity and mitogenic potency compared to regular human insulin. 14 However, further studies in human skeletal muscle cells found insulin glargine was equivalent to regular human insulin for metabolic responses, and mitogenic effects were not demonstrated. 54 The second approach involves the addition of a fatty acyl chain, which allows the insulin to bind to non-esterified binding sites on albumin. As the modified insulin dissociates slowly from albumin and favours the hexameric state the resultant absorption and duration of action are longer. This is the rationale behind the current development of insulin detemir. 4,49 Insulin detemir is a soluble long-acting analogue with a fatty acyl chain attached to B29 Lys (figure 2). In healthy subjects detemir has a less pronounced peak compared to NPH insulin. Maximum concentrations were reached after 4 6 hours, 55 and detemir appears to produce a more predictable glycaemic control, a smoother plasma glucose profile with a significant reduction in hypoglycaemia. 56 Further human pharmacodynamic and clinical studies are awaited. Conclusions New long- and short-acting insulin analogues are providing greater opportunity for patients to improve glycaemic control. This should lead to a reduction in diabetes-related complications and reduce the incidence of hypoglycaemia, thereby offering patients a better lifestyle. References 1. Heinemann L, Richter B. Clinical Pharmacology of human insulin. Diabetes Care Dec 1993;16(suppl 3): Chawdhury SA, Dodson EJ, Dodson GG et al. The crystal structures of three non-pancreatic human insulins. Diabetologia 1983;25: Brogden RN, Heel RC. Human insulin. A review of its biological activity, pharmacokinetics and therapeutic use. Drugs 1987;34(3): Bolli GB, Di Marchi RD, Park GD, Pramming S, Koivisto VA. Insulin analogues and their potential in the management of diabetes mellitus. Diabetologia 1999;42: Binder C, Lauritzen T, Faber O, Pramming S. Insulin pharmokinetics. Diabetes Care 1984;7: Dimitriadis GD, Gerich JE. Importance of timing of preprandial subcutaneous insulin administration in the management of diabetes mellitus. Diabetes Care 1983;6: Brange J, Ribel U, Hansen JF et al. Monomeric insulins obtained by protein engineering and their medical implications. Nature 1988;333: Brange J, Owens DR, Kang S, Volund A. Monomeric insulins and their experimental and clinical implications. Diabetes Care 1990;13: Brems DN, Alter LA, Beckage MJ et al. Altering the association properties of insulin by amino acid replacement. Protein Eng 1992;5: Di Marchi RD, Chance RE, Long HB, Shields JE, Slieker LJ. Preparation of an insulin with improved pharmokinetics relative to human insulin through consideration of structural homology with insulin-like growth factor 1. Horm Res 1994;41(suppl 2):S93-S Radziuk JM, Davis JC, Pye WS et al. Bioavailability and bioeffectiveness of subcutaneous human insulin and two of it s analogues- LysB28, ProB29- human insulin and AspB10, LysB28, ProB29-human insulin-assessed in a conscious pig model. Diabetes 1997;46: Somwar R, Sweeny G, Ramlal T, Klip A. Stimulation of glucose and amino acid transport and activation of the insulin signalling pathways by insulin lispro in L6 skeletal muscle cells. Clin Ther 1998;20: Slieker LJ, Brooke GS, Di Marchi RD et al. Modifications in the B10 and B26-30 regions of the B chain of human insulin alter affinity for the human IGF-1 receptor more than for the insulin receptor. Diabetologia 1997;40(suppl 2):S54-S61. VOLUME 4 ISSUE 1. JANUARY/FEBRUARY

5 Levy&Ahmad P /2/04 14:58 Page 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(6): Glazer NB, Zalani S, Anderson JH Jr, Baastyr EJ 3rd. Safety of insulin lispro: pooled data from clinical trials. Am J Health Syst Pharm 1999;56(6): Howey DC, Bowsher RR, Brunelle RL, Woodworth JR. [Lys(B28), Pro(B29)]-human insulin. A rapidly absorbed analogue of human insulin. Diabetes 1994;43: Davey P, Grainger D, MacMillan J, Rajan N, Aristides M, Gliksman M. Clinical outcomes with insulin lispro compared with human regular insulin: a meta-analysis. Clin Ther 1997;19(4): Anderson JH Jr, Brunelle RL, Koivisto VA et al. Improved mealtime treatment of diabetes mellitus using an insulin analogue. Clin Ther 1997;19: Chase HP, Lockspeiser T, Peery B, Shepherd M, MacKenzie T, Anderson J, Garg SK. The impact of the diabetes control and complications trial and humalog insulin on glycohemoglobin levels and severe hypoglycemia in type 1 diabetes. Diabetes Care 2001;24(3): Del Sindaco P, Ciofetta M, Lalli C et al. Use of the short-acting insulin analogue lispro in intensive treatment of Type 1 diabetes: importance of appropriate replacement of basal insulin and time interval injection meal. Diabetic Medicine 1998;15: Lalli C, Ciofetta M, Del Sindaco P et al. Long-term intensive treatment of type 1 diabetes with the short-acting insulin analogue lispro in variable combination with NPH insulin at mealtime. Diabetes Care 1999;22: Stades AM, Hoekstra JB, van den Tweel I, Erkelens DW, Holleman F Stability Study Group. Additional lunchtime basal insulin during insulin lispro intensive therapy in a randomized, multicenter, crossover study in adults: real-life design. Diabetes Care 2002;25(4): Joseph SE, Korzon-Burakowska A, Woodworth JR et al. The action profile of lispro is not blunted by mixing in the syringe with NPH insulin. Diabetes Care 1998;21: Anderson JH Jr, Brunelle RL, Koivisto VA et al. Reduction in postprandial hyperglycaemia and frequency of hypoglycaemia in IDDM patients on insulin-analog treatment. Diabetes 1997;46: Gale EA. A randomized, controlled trial comparing insulin lispro with human soluble insulin in patients with Type 1 diabetes on intensified insulin therapy. The UK Trial Group. Diabet Med 2000;17(3): Anderson JH Jr, Brunelle RL, Keohane et al. Mealtime treatment with insulin analog improves postprandial hyperglycaemia and hypoglycaemia in patients with non-insulin dependent diabetes mellitus. Arch Int Med 1997;157: Holcombe JH, Zalani S, Arora VK, Mast CJ. Lispro in Adolesecents Study Group. Comparison of insulin lispro with regular human insulin for the treatment of type 1 diabetes in adolescents. Clin Ther 2002;24(4): Heller SR, Ameil SA, Mansell P. Effect of the fast-acting insulin analog lispro on the risk of nocturnal hypoglycemia during intensified insulin therapy. UK Lispro Study Group. Diabetes Care 1999;22(10): Brunelle RL, Llewelyn J, Anderson JH, Gale EAM, Koivisto VA. Metaanalysis of the Effect of Insulin Lispro on Severe Hypoglycaemia in Patients with Type 1 Diabetes. Diabetes Care 1998;21: Launay B, Zinman B, Tildesley HD, Strack T, Chiasson JL. Effect of Continuous Subcutaneous Insulin Infusion With Lispro on Hepatic Responsiveness to Glucagon in Type 1 Diabetes. Diabetes Care 1998;21: Jacobs MA, Salobir B, Popp-Snijders C, Ader H, Heine RJ. Counterregulatory hormone responses and symptoms during hypoglycaemia induced by porcine, human regular insulin and Lys(B28), Pro(B29) human insulin analogue (insulin Lispro) in healthy male volunteers. Diabet Med 1997;14(3): Torlone E, Fanelli C, Rambotti AM et al. Pharmacokinetics, pharmacodynamics and glucose counterregulation following subcutaneous injection of the monomeric insulin analogue [Lys(B28),Pro(B29)] in IDDM. Diabetologia 1994;37: McCrimmon RJ, Frier BM. Symptomatic and physiological responses to hypoglycaemia induced by human soluble insulin and the analogue lispro human insulin. Diabet Med 1997;14: Schernthaner G, Wein W, Sandholzer K et al. Post prandial insulin lispro. A therapeutic option for type 1 diabetic patients. Diabetes Care 1998;21: Kotsanos JG, Vignati L, Huster W et al. Health-related quality-of-life results from multinational clinical trials of insulin lispro. Assessing benefits of a new diabetes therapy. Diabetes Care J 1997;20(6): Bhattacharyya A, Brown S, Hughes S, Vice PA. Insulin Lispro and regular insulin in pregnancy. QJM 2001;94(5): Boskovic R, Feig DS, Derewlany L, Knie B, Portnoi G, Koren G. Transfer of insulin lispro across the human placenta: in vitro perfusion studies. Diabetes Care 2003;26(5): Loukovaara S, Immonen I, Teramo KA, Kaaja R. Progression of retinopathy during pregnancy in type 1 diabetic women treated with insulin lispro. Diabetes Care 2002;26(4): Volund A, Brange J, Drejer K et al. In vitro and in vivo potency of insulin analogues designed for clinical use. Diabet Med 1991;8: Dall V. Preclinical safety pharmacology studies on the rapid-acting insulin analogue insulin aspart. Arzneimittelforschung 1999;49(5): Lindholm A, Jensen LB, Home PD, Raskin P, Boehm BO, Rastam J. Immune responses to insulin aspart and biphasic insulin aspart in people with type 1 and type 2 diabetes. Diabetes Care 2002;25(5): Mudaliar SR, Lindberg FA, Joyce M et al. Insulin aspart (B28 asp-insulin): 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(9): Kang S, Brange J, Burch A, Volund A, Owens DR. Absorption kinetics and action profiles of subcutaneously administered insulin analogues (AspB9, GluB27; AspB10, AspB28) in healthy subjects. Diabetes Care 1991;14: Raskin P, Guthrie RA, Leiter L, Riis A, Jovanovic L. Use of insulin aspart, a fast-acting insulin analog, as the mealtime insulin in the management of patients with type 1 diabetes. Diabetes Care 2000;23(5): Home PD, Lindholm A, Riis A. Insulin aspart vs human insulin in the management of long-term blood glucose control in Type 1 diabetes mellitus: a randomized controlled trial. Diabet Med 2000;17(11): Frier BM, Ewing FM, Lindholm A, Hylleberg B, Kanc K. Symptomatic and counterregulatory hormonal responses to acute hypoglycaemia induced by insulin aspart and soluble human insulin in Type 1 diabetes. Diabetes Metab Rev 2000;16(4): Brunner GA, Hirschberger S, Sendlhofer G et al. Post-prandial administration of the insulin analogue insulin aspart in patients with Type 1 diabetes. Diabet Med 2000;17(5): Halberg IB, Jacobsen LV, Dahl UL. A study on self-mixing insulin aspart with NPH insulin in the syringe before injection. Diabetes 1999;48(suppl 1):A104 (Abstract). 49. Barnett AH. A review of basal insulins. Diabet Med 2003;20: Perriello G, De Feo P, Torlone E et al. The dawn phenomenon in Type 1 (insulin-dependent) diabetes mellitus: magnitude, frequency, variability, and dependency on glucose counterregulation and insulin sensitivity. Diabetologia 1991;34: Lepore M, Pampanelli S, Fanelli C et al. Pharmacokinetics and pharmacodynamics of subcutaneous injection of long-acting human insulin analog glargine, NPH insulin, and ultralente human insulin and continuous subcutaneous infusion of insulin lispro. Diabetes 2000;49(12): Rosenstock J, Schwartz SL, Clark CM Jr, Park GD, Donley DW, Edward MB. Basal insulin therapy in type 2 diabetes: 28-week comparison of insulin glargine (HOE 901) and NPH insulin. Diabetes Care 2001;24(4): Pieber TR, Eugene-Jolchine I, Derobert E. Efficacy and safety of HOE 901 versus NPH insulin in patients with Type 1 Diabetes. The European Study Group of HOE 901 in type 1 diabetes. Diabetes Care 2000;23(2): Ciaraldi TP, Carter L, Seipke G, Mudaliar S, Henry RR. Effects of the longacting insulin analog insulin glargine on cultured human skeletal muscle cells: comparisons to insulin and IGF-1. J Clin Endocrinol Metab 2001;86(12): Heinemann L, Sinha K, Weyer C, Loftager M, Hirschberger S, Heise T. Time-action profile of the soluble, fatty acid acylated, long acting insulin analogue NN304. Diabet Med 1999;16(4): Vague P, Selam JL, Skeie S et al. Insulin detemir is associated with more predictable glycemic control and reduced risk of hypoglycaemia than NPH insulin in patients with type 1 diabetes on a basal-bolus regimen with premeal insulin aspart. Diabetes Care 2003;26(3): THE BRITISH JOURNAL OF DIABETES AND VASCULAR DISEASE