The objective of insulin replacement is to mimic the insulin secretion pattern in the person without diabetes with multiple subcutaneous injections. In the person without diabetes, there is normally a rapid increase in plasma insulin after meals, triggered by glucose absorption into the bloodstream. This surge in insulin limits postprandial glycaemia by stimulating hepatic and peripheral glucose uptake. During fasting and between meals, insulin measurements drop to much lower levels (often called basal or steady state) which are sufficient to maintain blood glucose in the range 3.5-5.5 mmol/L. Even after a prolonged fast, it is possible to detect circulating insulin.
Basal insulin levels tend to be highest in the early morning, probably in response to the well-described surge in growth hormone and cortisol at that time of day (Figure 10.1). These counter-regulatory hormones tend to increase blood glucose and this has been termed the ‘dawn phenomenon’.
For practical reasons, insulin is usually injected subcutaneously and regimens comprise short-acting (soluble, regular or analogue) insulin to simulate the normal mealtime surge, together with a longer acting insulin which is used to provide the background or basal concentration. This combination is called the ‘basal-bolus’ regimen or multiple daily injection (MDI) therapy (Figure 10.2).
Other routes of insulin administration such as intravenous infusion or intramuscular injection have not proven practical in the long term and despite intensive research, oral insulin preparations are not yet available.
Until the 1980s, insulin was extracted and purified from animal sources. Porcine and bovine insulins are still available but have been largely replaced by human sequence insulin produced from genetically engineered bacteria. Recently modified human insulin molecules (so-called analogues) have now been developed (Figure 10.3).
Essentially insulins can be divided up into short-acting and intermediate- to long-acting preparations and those currently available in the UK are listed in Table 10.1.
Achieving normoglycaemia with insulin injections is frustrated by several pharmacological problems. Firstly, subcutaneously injected insulin is absorbed into the peripheral rather than the portal bloodstream; thus, effective insulinisation of the liver can be achieved only at the expense of systemic hyperinsulinaemia. Moreover, conventional short-acting insulins are absorbed too slowly to mimic precisely the normal prandial peaks, and must therefore be injected about 30 minutes before the meal so that the peak of blood insulin corresponds with postprandial glycaemia. Human insulin is absorbed more quickly and can be injected closer to eating (Figure 10.4). It is therefore much more convenient than animal insulins, although a systematic review has failed to demonstrate significant improvements in overall glycaemic control. Some patients who were well controlled on animal insulin felt less well when switched to the human preparation. Many complained that their warning signs of hypoglycaemia were lost. However, carefully controlled studies have failed to show significant differences in the glycaemic response to the two insulins and under blinded conditions patients were unable to distinguish between them.
The delay in absorption of subcutaneous insulin is a result of the formation of hexamers following injection (Figure 10.4). The hexamers need to dissociate into dimers or monomers so that insulin can be absorbed into the bloodstream. In order to get around this problem, the insulin molecule has been modified using genetic and protein engineering techniques. These changes in amino acid sequence reduce the tendency to self-associate into hexamers and therefore speeds absorption. The first short-acting analogue to be marketed was lispro closely followed by aspart and glulisine (see Figure 10.3). Their peak action occurs 1-2 hours after injection (compared to 2-4 hours for conventional soluble and slightly less for unmodified human) and they can therefore be injected at the start or even during a meal. Although this is highly convenient for patients, systematic reviews of short-acting analogues versus unmodified human insulin have failed to demonstrate consistent advantages in terms of glycaemic control, although a fairly consistent pattern of reduced nocturnal hypoglycaemia, lower postprandial and higher preprandial blood glucose levels is seen.
There are three main types of intermediate- and long-acting insulins. Isophane (or NPH, neutral protamine Hagedorn, named after its inventor) is an insoluble suspension of insulin made by combining it with the highly basic protein protamine, together with zinc, at a neutral pH. NPH insulin can be derived from animal, unmodified human or analogue insulins. Lente insulins are made by adding excess zinc to soluble insulin. There is also a combination of protamine and zinc suspended insulin available. NPH and lente insulins have a duration of action of between 8 and 16 hours after injection. So-called ultralente insulins are no longer available in the UK.
Because of the duration of action of NPH insulin there was a tendency for patients to develop nocturnal hypoglycaemia. As a result of this, two longer-acting insulin analogues have been developed which have a flatter absorption profile. The first of these (insulin glargine) was made by adding two arginine molecules to the C-terminal of the B chain and substituting a glycine for alanine at A21 (Figure 10.3). This modification altered the isoelectric point (which is when proteins are least soluble) from pH 5.4 to 7.4. This means that at a slightly acidic pH in the vial, glargine is soluble and clear (in contrast with NPH and lente which are both cloudy solutions) but after subcutaneous injection it precipitates as microcrystals and is gradually absorbed (Figure 10.5). Detemir is human insulin where the C-terminal amino acid on the B chain is substituted with a C14 fatty acid (see Figure 10.3). This binds to albumin which slows absorption and also prolongs circulation.
The duration of action of both of these analogues is longer than that of NPH and lente (Figure 10.5), but despite this no consistent benefit in terms of HbA1c has been found. However, fasting glycaemia and rates of nocturnal hypoglycaemia tend to be reduced and current NICE guidance suggests that insulin glargine be considered instead of NPH if there are problems with night-time glycaemic control.
A number of these are available (see Table 10.1) but those most commonly in use in UK have 30% human soluble and 70% human NPH although the newer analogues come in 25:75, 30:70 or 50:50 mixtures. These fixed rate combinations give less flexibility than MDI. Moreover, the early evening injection of the NPH component may not be adequate to provide overnight glycaemic control, particularly if the patient exhibits the dawn phenomenon.
The recommended injection sites are the subcutaneous tissue of the abdomen, upper outer thighs, upper outer arms, and buttocks (Figure 10.6). Disposable plastic syringes with a fine needle can be reused for several injections, although these have been largely superseded in the UK by insulin pens (see below). There is no need to pinch up the skin prior to injection (in fact, this probably causes more discomfort). Care should be taken to avoid inadvertent intramuscular injection which can be a particular risk in the upper arms and legs of slim people or children.
Insulin absorption is fastest in the abdomen and slowest in the thigh and buttocks although it can be accelerated from these sites by exercise or taking a sauna or warm bath. Short-acting insulin is usually given into the abdomen, which is less affected by exercise, and longer acting insulins into the thigh.
Repeated injection into the same subcutaneous site may, in the long term, give rise to an accumulation of fat (lipohypertrophy) because of the local trophic action of insulin (Figure 10.7). Lipohypertrophy can be unsightly and can affect insulin absorption. In order to prevent lipohypertrophy, patients should be advised to rotate the site of injection.
It is important to remember that lipohypertrophic areas become relatively painless and are thus often favoured by patients who may inadvertently make the problem worse. For this reason inspection of injection sites is an important part of the annual patient review.
Pen needle devices have become a popular option for insulin therapy in recent years (Figure 10.8). There are at least eight currently available in the UK (Table 10.2). Whilst they differ slightly, the principles are the same. Insulin is contained in a 3 mL cartridge in the barrel of the pen. There is an adjustable dosage device which drives a plunger in the cartridge and insulin is delivered through a removable fine needle which screws on at the opposite end of the pen. The advantages of pens are their convenience, the needles are often finer than conventional syringes and needles, and they tend to keep their sharpness longer because they are not being continually inserted through the rubber bung of a 10 mL vial. It is also possible to get pen needles of different lengths which means it is easier to avoid inadvertent intramuscular injection.
Multiple daily injections (MDI) of insulin are only part of intensive or optimised treatment. The other components are patient education, dietary advice and carbohydrate counting and continual insulin adjustment. Moreover, patients need to be in a systematic programme of care and medical surveillance. These aspects will be covered later (see Chapter 31).
Multiple daily injections consist of short-acting, regular or rapid-acting analogue insulins given with meals at a variable dose, depending on the carbohydrate content. Patients are encouraged to monitor carefully and learn by checking the postprandial blood glucose 11h-2 hours after meals to see if their first estimation was correct.
Isophane or lente insulins or long-acting analogues can be given at night or twice daily. As previously mentioned, insulin sensitivity decreases in the few hours before breakfast because of surges of growth hormone and cortisol during sleep. The effect of this, together with the waning of the previous evening’s insulin dose, results in fasting hyperglycaemia (the ‘dawn phenomenon’) (Figure 10.9). This problem can often be countered by moving the longer-acting insulin injection to bedtime.
Deciding which insulins to use can often be a case of trial and error based upon the patient’s lifestyle and needs, employment, etc. Current evidence would suggest that NPH plus a short-acting human insulin (or possibly an analogue) is the logical starting point.
Glycaemic targets published in the UK and the USA are shown in Table 10.3. These will need to be discussed and agreed with individual patients. For example, those with existing microvascular complications may need a lower target HbA1c (NICE recommends <6.5%) whereas patients suffering from regular hypoglycaemia (particularly if their warnings are blunted) may need more relaxed targets. Recently Diabetes UK has come up with the phrase ‘4 is the floor’, suggesting a minimum blood glucose of 4 mmol/L.
Continuous insulin delivery systems can be either ‘open loop’, in which insulin infusion rates are preselected by the patient, or ‘closed loop’ in which there is continuous glucose sensing and a computer-regulated feedback control of insulin delivery (the so-called ‘artificial pancreas’).
Continuous subcutaneous insulin infusion (CSII) was developed over 30 years ago. This is an open-loop delivery system in which a portable pump infuses insulin subcutaneously at variable rates via an implanted cannula (Figures 10.10 and 10.11). This is meant to mimic basal insulin. At mealtimes patients activate a bolus dose. This combination is meant to reproduce the insulin secretory pattern seen in non-diabetic people. Because only short-acting (usually analogue) insulin is used, the problem of variable absorption of intermediate-acting preparations is overcome. Moreover, basal rates can be changed hour by hour and this is particularly useful in individuals who have the dawn phenomenon.
Closing the loop using subcutaneous interstitial glucose sensing connected to an insulin pump is now available but at the time of writing was not yet funded on the NHS in the UK (see Chapter 9).
There are now several different insulin pumps available; they cost between £2500 and £3000 with running costs of approximately £800-1000 per year.
All pumps comprise a chamber where usually 3 mL of insulin is drawn up into a special syringe. A motor drives a plunger which delivers insulin from the syringe via a cannula under the skin.
A typical strategy for commencing pump therapy is to reduce the patient’s total daily insulin dose on injections by 20% and then allocating the remainder, one half to the basal rate and the other split equally between the three main meals. In order to fine-tune insulin delivery, multiple capillary blood glucose tests are required. In addition, patients are taught how to estimate the carbohydrate content of their meals and generally one unit of insulin is given for every 10 g portion. Higher doses may be required in patients who may be insulin resistant (e.g. adolescents, pregnancy, obesity, where 1.25 units/10 g may be required). Conversely, for smaller children who tend to be more insulin sensitive, 0.5 units/10 g is a useful starting dose.
The best-established clinical indication for CSII is in patients who have failed to achieve glycaemic targets on MDI or who have problematic, frequent and unpredictable hypoglycaemia. In addition, there may be particular indications for patients who have a marked dawn phenomenon or for those women who are considering pregnancy or who are pregnant and have difficulty achieving glycaemic control, and for those rare patients who have a true insulin allergy. Psychological problems should not preclude a trial of CSII; one centre reported significant benefit in terms of reduction of hospital admission rates with diabetic ketoacidosis, together with a reduction in HbA1c even though the level on CSII was less than ideal at >9% (75 mmol/mol). Estimates of the numbers of patients in the UK who might want or benefit from CSII vary but it is probable that 10-20% would accept the treatment and achieve benefit.
As is the case with many medical technologies, the evidence base for the effectiveness or otherwise of CSII is not as strong as we would like. A Health Technology Appraisal from NICE in the UK was updated in 2008 and approved CSII for adults and children with type 1 diabetes provided certain conditions have been met (see Box 10.1). As a result, CSII has become much more widely available in the UK, although absolute numbers of patients using pumps are far fewer than those in other European countries and North America.
A meta-analysis in 2008 confirmed glycaemic improvement in terms of both HbA1c (average reduction 0.61%) and severe hypoglycaemic rates (4.19 times less likely) on pump therapy compared to MDI (Figure 10.12). NICE found that patients with a higher HbA1c on MDI obtained most benefit in terms of glycaemic control on CSII. Moreover, CSII was effective in terms of cost per QALY (quality adjusted life year) with a range of £16,842 to £34,330; this fell to less than £29,000 when improvement associated with the reduction in severe hypoglycaemic rates was taken into account. Totally implantable pumps that deliver insulin into the peritoneal cavity have been used in Europe and the USA for many years. Overall glycaemic control tends to be similar to that during MDI or CSII but with reduced fluctuations and fewer hypoglycaemic episodes. These pumps need to be implanted subcutaneously with a catheter placed in the peritoneal cavity; requiring a general anaesthetic. The insulin pump reservoir is topped up by using a syringe and needle via an injection port. The pump is programmed by a small external device which can be used to deliver variable basal rates and boluses. These devices are only funded in the UK on an individual patient basis and their use is really confined to patients with severe brittle diabetes who have not responded to CSII.
A 40-year-old manager with type 1 diabetes had noticed a gradual deterioration in his glycaemic control. He developed diabetes aged 25 when working offshore and had to change career. He was managed on a twice-daily regimen of soluble and lente insulins. His doses of insulin had increased such that he was taking >120 units/day (1.4 units/kg bodyweight) and he periodically experienced severe hypoglycaemia with little warning. His HbA1c was 8.2%. He exercised regularly but was finding this more difficult because of general fatigue and hypoglycaemia. Investigations showed normal thyroid function and negative insulin antibodies.
He was commenced upon CSII using human insulin and his total daily dose fell to 60 units. His fatigue and hypoglycaemic episodes disappeared and his HbA1c fell to 6.9%. He has now been on CSII for 9 years and remained in stable control with minimal retinopathy only.
Comment: Lente insulins could be associated with unpredictable hypoglycaemia. Part of the problem is the excess zinc in these preparations and the potential for this to accumulate under the skin. The cause of the fatigue was probably a com- bination of poor glycaemic control and hyperinsulinaemia. Insulin doses often fall on CSII, on average 15%, so the 50% reduction in this case is exceptional. It is also noteworthy that hypoglycaemia rates diminished despite a reduction in HbA1c. This man was converted to CSII pre glargine. He declined a trial of this analogue although current NICE guidance now suggests it prior to commencing CSII.
Islet transplantation for type 1 diabetes was first attempted in the 1980s with very poor results – less than 10% of patients were insulin independent at 1 year. In the late 1990s a new protocol was developed by workers in Edmonton, Canada, which avoided the use of high-dose corticosteroids for immunosuppresion. The islets are infused directly into the hepatic portal vein percutaneously. The initial results were extremely promising with all of the first seven patients maintaining insulin independence at 1 year. However, more extensive experience from multicentre trials and transplant registries suggests that longer term insulin independence is much less good.
From 1999 to April 2008 the Collaborative Islet Transplant Registry (CITR) recorded 325 recipients of 649 islet infusions. At 3 years post first infusion, only 23% were insulin independent, with a further 29% showing some insulin secretion; 26% had lost function.
There are problems associated with longer term immunosuppression. Apart from the well-recognised complications of infection and cancer, the current protocols rely on agents that are nephrotoxic. Rates of decline in kidney function of 2-4 times that of non-transplanted controls have been reported. For this reason, patients with renal impairment are ineligible for islet transplantation. Moreover, islet recipients have developed sensitivity to HLA antigens, making subsequent kidney transplantation (if required) more difficult. Thus patients with nephropathy potentially needing a kidney graft in the future are also ineligible.
However, for patients with severe disabling hypoglycaemia with unawareness, islet transplantation has been life transforming and even those who lose their insulin independence continue to have fewer and much less severe hypoglycaemic episodes. In 2008 NICE approved allogeneic islet transplantation for patients with type 1 diabetes and severe hypoglycaemia with unawareness.
Alejandro R, BartonFB, Hering BJ, Wease S. 2008 update from the Collaborative Islet Transplant Registry. Transplantation 2008; 86: 1783-1788.
American Diabetes Association. Standards of medical care in diabetes – 2010. Diabetes Care 2010; 33(Suppl 1): S1-S100.
Daneman D. Type 1 diabetes. Lancet 2006; 367: 847-858.
Pickup JC, Sutton AJ. Severe hypoglycaemia and glycaemic control in type 1 diabetes: meta-analysis of multiple daily insulin injections versus continuous subcutaneous insulin infusion. Diabetic Med 2008; 25: 765-774.
Richter B, Neises G. ‘Human’ insulin versus animal insulin in people with diabetes. Cochrane Database Syst Rev 2004; 3: CD003816.
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 2005; 3: CD003287.
For more information and to purchase this book, just follow this link:
Note: The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting a specific method, diagnosis, or treatment by physicians for any particular patient. The publisher and the authors make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. Readers should consult with a specialist where appropriate. The fact that an organisation or website is referred to in this work as a citation and/or a potential source of further information does not mean that the authors or the publisher endorse the information the organisation or website may provide or recommendations it may make. Further, readers should be aware that Internet websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the authors shall be liable for any damages arising herefrom.