|
| Home | Evidence Based Therapeutics | MeReC | MeReC Bulletin Volume 17 No 4 - Main Bulletin |
 |
|
The role of newer insulins in diabetes: Main Bulletin
Over the last five years, the prescribing costs for diabetes in both primary and secondary care have almost doubled. In 2006, 28.4 million prescription items for diabetes were dispensed in primary care, costing £561 million. Although insulin prescriptions still account for just under a fifth of all prescriptions for diabetes and almost half of diabetes prescription costs in primary care, the amount spent on insulin prescriptions has increased by about 70% in the last five years. In 2006, £238 million was spent on insulin prescriptions in primary care. This compares with £178 million spent on oral antidiabetic drugs and around £141 million on blood glucose testing reagents.2 The increase in prescribing costs could be explained by increased insulin doses in attempts to achieve normoglycaemia and the increasing prevalence of type 2 diabetes. However, there has also been a shift in the type of insulin chosen, with the newer, more expensive insulin analogues becoming the most commonly prescribed insulins. Does the current evidence-base for these insulins justify this wider adoption for significant numbers of people with diabetes who need insulin, or does the evidence justify use for a limited number of people on an individual basis? The long-term complications of diabetes are serious. Microvascular complications due to small blood vessel damage include retinopathy (which can lead to visual impairment and blindness), nephropathy (which can lead to progressive kidney failure) and neuropathy. Macrovascular complications due to damage to large arteries can also occur, including coronary heart disease (CHD), stroke and peripheral vascular disease (PVD), which can lead to the development of foot ulcers and amputation.3 Life expectancy is reduced considerably by having diabetes. It has been estimated that the life expectancy of men diagnosed with diabetes at age 40 years is reduced by about 12 years and for women it is reduced by around 14 years.4 Is there evidence, or at least reasonable experience, that the wider use of newer insulins will improve this situation? Is glycaemic control important? Type 1 diabetes Observational follow-up of 1,375 patients from the original DCCT cohort is ongoing and all patients have been advised to use intensive insulin treatment. Even though seven years after the end of DCCT, HbA1C narrowed between the original intensive and conventional therapy groups, the reduced risk of complications seen with initial intensive therapy in DCCT was maintained.6 A significant reduction in cardiovascular (CV) events with initial intensive therapy has also been reported 11 years after DCCT ended.7 This is further supported by a recent meta-analysis of RCTs, including eight comparisons in 1,800 patients with type 1 diabetes (duration <10 years) which suggested that attempting to improve glycaemic control reduces the risk of macrovascular events in patients with type 1 diabetes.8 The drawback of aiming for tight glycaemic control with intensive insulin treatment is an increased risk of hypoglycaemia and of being overweight. In DCCT, three times as many patients in the intensive insulin group experienced hypoglycaemia requiring assistance (61 vs. 19 per 100 patient years, p<0.001) compared with conventional treatment.9 This means that 36 patients would need to be treated with intensive insulin therapy instead of conventional treatment for 6.5 years for one patient to experience a hypoglycaemic episode requiring assistance (i.e. number needed to harm [NNH] 36). Type 2 diabetes The meta-analysis outlined above (see type 1 diabetes) assessed six comparisons in 4,472 patients with type 2 diabetes (duration <10 years). It suggests that attempting to improve glycaemic control reduces the risk of CV events in type 2 diabetes as well as type 1 diabetes. However, a more modest effect was seen in patients with type 2 diabetes.8 This may, partly, be explained by the greater relative importance of addressing other risk factors (e.g. hypertension, hyperlipidaemia and platelet activity) in patients with type 2 diabetes. Glycaemic control is just one aspect of care in all patients with diabetes and factors other than glycaemic control seem to be most important in patients with type 2 diabetes (see Figure).
What target HbA1C? The goals of treating diabetes are to manage symptoms, prevent acute and chronic complications, improve quality of life and prevent premature diabetes-related death.15 Insulin is necessary and, indeed, lifesaving for all patients with type 1 diabetes.16 It is also an option for those patients with type 2 diabetes who have inadequate blood glucose control on oral glucose-lowering drugs14 and who wish to attempt to achieve better control. Insulin therapy always should be considered as a ‘package of care’ including self-monitoring of blood glucose, dose adjustment, full diabetes education, lifestyle, emotional support and paediatric, dietetic, medical and nursing expertise.13 Back
to top Type 1 diabetes
The insulin regimen chosen requires a careful assessment of the needs of the individual patient (e.g. lifestyle, level of glycaemic control required, risk/experience of hypoglycaemia, ability to adhere to the regimen, individual preference).13 In type 1 diabetes, the two most common insulin regimens consist of:15,18
A twice daily injection regimen usually involves injecting both a short-acting or rapid-acting insulin and an intermediate-acting insulin before breakfast and the evening meal. Often premixes containing various proportions of short-or rapid-acting insulin and intermediate-acting insulin are used for simplicity. NICE recommends this regimen for adults who consider the number of injections important to their quality of life. This regimen also might be helpful for other people, such as those who find it difficult adhering to their lunch-time injection and those with learning difficulties, who may need help giving their injections.13 A multiple daily injection regimen (sometimes known as the ‘basal bolus regimen’) consists of injections of short-acting or rapid-acting insulin before meals, together with one or more separate injections of intermediate-acting or long-acting insulin.13 The intermediate-acting or long-acting insulin is usually given at bedtime, but sometimes it is also needed in the morning. Type 2 diabetes Problems with insulin treatment Despite the use of multiple daily injections, it is difficult to mimic the body’s normal insulin profile and there are currently several problems with insulin treatment.21 Hypoglycaemia is common and can occur due to unpredictable absorption at injection sites, or the prolonged action of the mealtime insulin. Therefore, to avoid this, people using short-acting insulins before meals often have to snack between meals, which increases their risk of weight gain. Serious episodes of hypoglycaemia were considered to be rare in people with type 2 diabetes.22 However, a population-based study in the UK suggests that the frequency of severe hypoglycaemia requiring emergency treatment in patients with type 2 diabetes who are using insulin is similar to that seen in patients with type 1 diabetes (7.3% vs. 7.1% over 12 months).22 Hypoglycaemia can be a barrier to effective glycaemic control and it is a particular concern in elderly patients.23 Night-time control is also a problem because low levels of basal insulin are required, followed by an increase in the morning to counteract a rise in blood glucose.21 It is not always possible to maintain blood glucose control without causing nocturnal hypoglycaemia (thought to result from the peak of NPH and the long-tail of short-acting insulin). However, reducing the dose of insulin to avoid nocturnal hypoglycaemia can lead to hyperglycaemia in the morning, known as the ‘dawn phenomenon’. Adherence of some patients with diabetes is known to be poor. A small study of 89 young patients (<30 years) who needed insulin found that a quarter collected less insulin from the pharmacy than their prescribed dose.24 In a worldwide survey of 5,104 adults with type 1 or type 2 diabetes, only 19.4% of patients with type 1 diabetes and 16.2% of people with type 2 diabetes reported that they completely complied with all aspects of their prescribed regimens (diet, exercise, medication, self-monitoring of blood glucose and appointment keeping).25 One reason why patients might not want to take insulin is fear of weight gain.20 Weight gain with insulin is usually a result of improved glycaemic control.19 However, this can be a particular concern for patients with type 2 diabetes who might already be overweight. There is a lack of strong evidence showing that one insulin preparation is safer or more effective than another. However, the collaborating centre for the NICE type 1 diabetes guideline acknowledged that the use of insulin injections is not based on evidence from RCTs and never could be.12 Treatment choices are always a balance between effectiveness, safety, cost and individual patient factors.26 The following data shows limited differences in RCTs on effectiveness and safety when populations are compared. Therefore, individual patient circumstances and cost are likely to dominate decision making. Several studies have compared the different insulins in diabetes. However, their interpretation is often limited by various factors relating to their design. For example, it is not always possible to blind studies as insulins often look different and are not all given at exactly the same time of day. Also, surrogate markers or disease-oriented outcomes (DOOs) e.g. blood glucose concentrations and HbA1C are often assessed, instead of more important outcomes that really matter to patients (known as patient-oriented outcomes [POOs]), such as quality of life or development of complications (e.g. blindness or amputations). HbA1C is the primary endpoint used in most insulin studies as a surrogate measure of diabetic complications. It is the most widely accepted measure of overall long-term blood glucose control in patients with both type 1 and type 2 diabetes.23 Many RCTs have also assessed the effects of insulins on hypoglycaemia (a POO). However, this was frequently a secondary endpoint and different definitions of hypoglycaemia were often used in trials, making comparison of treatments difficult. NICE allows for an individualised approach when choosing insulin preparations.13,14 Their type 1 diabetes guideline states that adults with type 1 diabetes should have access to the types of insulin they find allow them optimal well-being.13 They also recommend that children and young people with type 1 diabetes are offered the most appropriate insulin preparations according to their individual needs.13 NICE guidance for type 2 diabetes suggests that the choice of insulin should be informed by local experience, patient preference and relative costs.14 This guideline is due to be updated in 2008. The collaborating centre that informed the NICE type 1 diabetes guideline has cautioned against switching patients with control problems onto the newer more expensive insulins without proper assessment of underlying causes.12 Such causes of poor control might include problems with the patient’s injection technique or inadequate knowledge and poor self-management skills.12 Also, this year, several older insulin preparations will be discontinued by the manufacturer.27 Reviewing patients’ insulin regimens needs to be done carefully by assessing each individual’s requirements. Before switching a patient with poor glycaemic control onto different insulins, it is important to ensure that they have been educated and understand how to manage their condition and treatment. A RCT (n=169) found that educating patients with type 1 diabetes and moderate or poor control to adjust their insulin to the desired carbohydrate intake at each meal reduced HbA1C by 1.0% at six months (95% CI 0.5 to 1.4, p<0.0001). An improvement was also seen in the impact of diabetes on overall quality of life (p<0.01).28 Specific guidance on educating patients with diabetes has been issued by NICE (See Resources box). Mealtime insulin replacement is usually provided by either a short-acting, soluble insulin or a rapid-acting insulin analogue (See Panel). Rapid-acting insulin analogues have been developed with a pharmacokinetic profile that more closely resembles insulin requirements after meals i.e. they have a more rapid onset and shorter duration of action compared with soluble, short-acting insulin. Therefore, they might be expected to improve glycaemic control and, perhaps, reduce the incidence of hypoglycaemia. Over the last five years insulin aspart (NovoRapid®) has become the leading mealtime insulin prescribed and short-acting soluble insulin has become the least. Prescriptions for insulin lispro (Humalog®) have risen slowly during this time.2 Type 1 diabetes The HbA1C reductions with insulin lispro and insulin aspart were slightly more pronounced when insulin pumps were used, but the size of the effect compared with soluble insulin was still modest (WMD between –0.28 and –0.31%).29 Few fully-published RCTs of the more recently introduced rapid-acting insulin analogue, insulin glulisine (Apidra®▼), are available. A 12-week open label RCT (n=866) compared insulin glulisine used either before or after meals with soluble insulin used before meals. Insulin glargine was the basal insulin used in all patients. Only a modest statistically significant reduction in HbA1C was seen when premeal insulin glulisine was compared with soluble insulin (difference in HbA1C change from baseline 0.13%, p=0.02). In addition, no statistically significant difference was found when postmeal insulin glulisine was compared with soluble insulin. This trial was initially designed to show whether insulin glulisine was non-inferior to soluble insulin.30 The Canadian HTA identified some RCTs suggesting that rapid-acting insulin analogues reduce postprandial blood glucose compared with short-acting soluble insulin.29 This might be important for specific patients. Considering mean differences in HbA1C in isolation has some limitations. It does not tell us how many patients achieved an HbA1C less than a certain value. Also, HbA1C is not adjusted for episodes of hypoglycaemia. For example, it can be low in patients with poor control and frequent episodes of hypoglycaemia. There is some evidence that rapid-acting insulin analogues, particularly insulin lispro, are associated with a lower rate of nocturnal hypoglycaemia than soluble short-acting insulin in patients with type 1 diabetes. However, there is little evidence that they reduce episodes of overall or severe hypoglycaemia (the most important outcome). When the Canadian HTA assessed 8 RCTs of insulin lispro, four (n=1,780) found a statistically significant reduction in nocturnal hypoglycaemia compared with soluble insulin, whereas the other four RCTs (n=427) did not. After pooling data from four of these studies (n=1,377, including two studies showing a statistically significant difference and two that did not) the authors found that insulin lispro reduced nocturnal hypoglycaemia by 0.55 episodes/patient/month (95% CI 0.19 to 0.92).29 They were unable to show a significant difference between insulin lispro and short-acting soluble insulin on overall hypoglycaemia or on major or severe hypoglycaemia. Two (out of 16) RCTs did show a statistically significant reduction in severe hypoglycaemia with insulin lispro.29 The largest of these (n=1,184, duration 3 months), found that although patients on soluble insulin and insulin lispro both had 1.8 episodes of hypoglycaemia per month, a smaller proportion of these episodes was classed as severe in the insulin lispro group (13.8% vs. 18.7%, p<0.001). However, many patients injected their soluble insulin closer to meals than recommended.31 Only one RCT (out of five) found a statistically significant reduction in overall hypoglycaemia, and one RCT (out of seven) found a statistically significant reduction in severe hypoglycaemia with insulin aspart compared with soluble insulin.29 The latter was a relatively small 4-week crossover RCT in 90 male patients (episodes of major hypoglycaemia requiring third party help: 20 events in 16 subjects vs. 44 events in 24 subjects, p<0.002). However, as with the insulin lispro study (above) patients injected their soluble insulin closer to meals than is usually recommended.32 Two out of three RCTs comparing insulin aspart with soluble insulin found a statistically significant reduction in nocturnal hypoglycaemia with the analogue. These were relatively small 16-week RCTs. The first RCT (n=146) compared these insulins when used in an insulin pump and found that the rate of confirmed nocturnal hypoglycaemic episodes per patient per 30 days was 0.5 (+ SD 0.83) with insulin aspart and 0.9 (+ SD 0.97) with soluble insulin (p=0.004).33 The second study (n=155) reported 0.067 nocturnal hypoglycaemic events/patient/month with insulin aspart and 0.225 events/patient/month with soluble insulin (p=0.001) when they were used with NPH insulin once or twice daily.34 However, as seen in some other studies,31,32 patients in this study injected the soluble insulin closer to mealtimes than 15 to 30 minutes beforehand, which is usually recommended.17,34 In addition, the third RCT, a larger (n=753), three-year study, found no significant differences between insulin aspart and soluble insulin used with NPH on major nocturnal hypoglycaemia or all episodes of major hypoglycaemia. Instead, insulin aspart slightly increased the risk of having a minor hypoglycaemic episode (2.46 vs. 2.03 episodes per patient per month, RR 1.24, 95% CI 1.09 to 1.39, p=0.024).35 The only fully-published RCT (n=860) comparing insulin glulisine with soluble insulin in patients with type 1 diabetes did not demonstrate a statistically significant difference in the rates of symptomatic, severe, or nocturnal hypoglycaemia between treatments over 12 weeks.30 Type 2 diabetes When the authors of the Canadian HTA pooled 10 RCTs that compared the effect of insulin lispro and soluble insulin on HbA1C, the WMD in HbA1C was only –0.11 (95% CI –0.22 to 0.00), suggesting that there is no significant difference between treatments. They found a slight difference in favour of insulin lispro when they pooled eight studies of more than three months (WMD –0.14%, 95% CI –0.27 to –0.01), but no statistically significant differences when they pooled six crossover studies or four parallel studies. The limitations of using mean HbA1C values from trials have been discussed above.29 The authors of the Canadian HTA also pooled six RCTs comparing insulin aspart with soluble insulin in 750 patients and, again, found no statistically significant differences between treatments on HbA1C (WMD –0.09%, 95% CI –0.23 to 0.05). Similarly, no significant differences in HbA1C were seen when insulin glulisine was compared with soluble insulin in two other trials (n=1,768; WMD –0.03, 95% CI –0.18 to 0.11).29 As was seen in patients with type 1 diabetes, there is some evidence from RCTs that rapid-acting insulins reduce postprandial blood glucose compared with short-acting soluble insulin.29 There is little evidence that the rapid-acting insulin analogues produce large reductions in episodes of severe, overall or nocturnal hypoglycaemia compared with short-acting soluble insulin in most patients with type 2 diabetes. The Table summarises the relative risks (RR) of severe, overall and nocturnal hypoglycaemia with the rapid-acting insulin analogues compared with soluble insulin in type 2 diabetes. These figures were calculated by the authors of the Canadian HTA after pooling data from trials. Although the RRs often appear to be reduced with the rapid-acting insulin analogues, in all cases the confidence intervals cross ‘one’ suggesting that there are no statistically significant difference between treatments.29
Place in therapy of analogues Because of their short duration of action compared with short-acting soluble insulin, rapid-acting insulin analogues can be used instead of short-acting soluble insulin where patients need or wish to avoid snacking between meals. They may also be tried for patients who have problems with nocturnal or late interprandial hypoglycaemia. For patients on twice daily injection regimens, NICE has suggested that biphasic insulin analogues may be advantageous for those who are prone to hypoglycaemia at night.13 Some studies comparing rapid-acting insulin analogues with short-acting soluble insulin have found an improvement in patients’ perceived quality of life. Part of this might be due to the fact that they can be injected immediately before meals, rather than 15 to 30 minutes beforehand, as is required with soluble insulin.17,29 The flexibility of injection times (i.e. just before or after meals) might appeal to some people, especially as there is evidence that patients often inject their soluble insulin closer to mealtimes than recommended.12 It is important to ensure that all patients are properly educated about how to use their insulin. The suggestions above may also apply to patients with type 2 diabetes. However, the role of these insulins in this specific group may be further clarified by the NICE guidance on type 2 diabetes, which is expected next year. The aim of the basal (background) insulin is to provide a constant level of insulin between meals without increasing the risk of hypoglycaemia, particularly at night.15 This should be provided by use of the intermediate-acting insulin NPH (isophane) or a long-acting insulin analogue (e.g. insulin glargine Lantus®▼, or insulin detemir Levemir▼).13 The insulin analogues can have a more prolonged action than NPH insulin.13,36 In the last five years, insulin glargine has replaced NPH as the most commonly prescribed basal insulin.2 When used as part of the basal-bolus regimen (i.e. a multiple daily injection regimen), NPH should be given at bedtime. However, if daytime insulin requirements are increased by the use of rapid-acting insulin analogues at mealtimes, or if the midday insulin dose is small or lacking, changing the frequency to twice daily (or more often) should be considered.13 Insulin detemir should be given once or twice daily, depending on the patient, whereas insulin glargine, which has the longest duration of action, only needs to be given once a day.17,37 Type 1 diabetes For patients with type 1 diabetes, there appears to be some evidence that insulin glargine reduces hypoglycaemia, particularly hypoglycaemia at night. However, not all RCTs have shown a difference between insulin glargine and NPH in the incidence of hypoglycaemia. The authors of the UK HTA concluded that evidence for the control of hypoglycaemia with insulin glargine was equivocal. Whilst some studies found that the frequency of nocturnal hypoglycaemia was reduced by insulin glargine, the authors suggested that this may only be apparent when compared with once daily and not twice daily NPH. They also found no conclusive evidence that insulin glargine was superior to NPH insulin in controlling symptomatic hypoglycaemia or overall hypoglycaemia.16 Some RCTs in patients with type 1 diabetes that were published since the UK HTA have shown a reduction in hypoglycaemia, particularly symptomatic and nocturnal hypoglycaemia with insulin glargine compared to NPH insulin in patients with type 1 diabetes.39–43 However, other RCTs have failed to show a difference between treatments.38,44,45 The incidence of hypoglycaemia appears to be slightly less frequent with insulin glargine, but severe hypoglycaemia seems to occur at a rate similar to that seen with NPH insulin.46 There is no strong evidence that insulin detemir provides better HbA1C control than NPH insulin in patients with type 1 diabetes. A recent review assessed 5 RCTs (four to six months) of insulin detemir in type 1 diabetes and found that it had a similar efficacy to NPH insulin in treating such patients.36 In addition, whilst some studies have shown benefits with insulin detemir on certain definitions of hypoglycaemia, particularly nocturnal hypoglycaemia, not all studies have shown a difference compared with NPH. For example, a meta-analysis of four RCTs (n=1,336, duration 16 to 24 weeks) compared insulin detemir with NPH insulin injected once or twice daily in patients with type 1 diabetes.47 A difference in HbA1C of 0.15% (95% CI –0.22 to –0.08) in favour of insulin detemir was found,47 but this was only minor. In addition, only two48,49 of the four studies48–51 included in the meta-analysis demonstrated a statistically significant difference in HbA1C between insulin detemir and NPH. The meta-analysis reported a reduction of just one major hypoglycaemic event (requiring third party assistance) per year per 100 patients with insulin detemir (41 vs. 42 events/year/100 patients).47 However, it should be noted that no consistent benefit on any one definition of hypoglycaemia was found in the individual studies included in this meta-analysis.48–51 This is also true when RCTs not included in the meta-analysis are considered. Several studies have shown a reduction in nocturnal hypoglycaemia with insulin detemir,48–50,52–55 but not all studies have demonstrated this benefit.51,56 As with insulin glargine, it is unclear whether the frequency of the basal insulin is important. In RCTs, patients using insulin detemir appeared to lose weight, or gain less weight, compared with those on NPH. The mechanism for this is uncertain, but a modest weight difference of about one or two kg was seen in RCTs up to six months.36 However, the long-term clinical significance of this is unclear. Back
to top In patients with type 2 diabetes, insulin glargine appears to reduce hypoglycaemia, particularly nocturnal, and symptomatic hypoglycaemia. However, there is little strong evidence that it reduces severe hypoglycaemia compared with NPH insulin. The Cochrane Review (discussed above) found that, compared with NPH insulin, insulin glargine resulted in statistically significant reductions in the numbers of patients who reported all symptomatic or symptomatic nocturnal episodes of hypoglycaemia. Different definitions were used in trials, with some also reporting statistically significant reductions in the rate of these events. When three RCTs were pooled, the authors found a RR of 0.84 (95% CI 0.75 to 0.95, p=0.005) for symptomatic hypoglycaemia, and a RR of 0.66 (95% CI 0.55 to 0.80, p<0.0001) for symptomatic nocturnal hypoglycaemia, in favour of insulin glargine.57 It should be noted, however, that only one study in this systematic review included a treatment arm involving twice daily NPH. In addition, the Cochrane Review did not find any significant difference in severe hypoglycaemia between treatments, although only a small number of events were reported.57 Whilst the reported reductions in HbA1C seem to be similar between insulin glargine and NPH insulin, for certain patients a reduction in hypoglycaemia with insulin glargine might allow them to reach their target HbA1C, by enabling them to titrate their insulin dose more intensively.46,59 Insulin detemir does not appear to reduce HbA1C compared with NPH insulin in patients with type 2 diabetes. Whilst it might be associated with a lower incidence of nocturnal hypoglycaemia, this was not seen in all RCTs. In addition, there is little strong evidence that it reduces severe hypoglycaemia compared with NPH insulin. We identified four fully-published RCTs (n=1,880; duration 20–26 weeks) that compared these insulins in people with type 2 diabetes. None of the trials found a statistically significant difference in HbA1C reduction.60–63 Two RCTs (n=980) in patients who were taking oral hypoglycaemic agents reported a statistically significant reduction in nocturnal hypoglycaemia with insulin detemir.60,61 However, the other two studies found no difference between insulin detemir and NPH on hypoglycaemia when they were used with a mealtime insulin (regular soluble insulin or aspart).62,63 When the authors of the recent Cochrane Review (discussed above under insulin glargine) pooled the results from two of the RCTs60,63 they found that the RR of patients having at least one episode of nocturnal hypoglycaemia was 0.63 (95% CI 0.52 to 0.76, p<0.00001).57 However, they were unable to identify any significant difference between insulin detemir and NPH insulin on severe hypoglycaemia.57 Place in therapy of analogues In its type 1 diabetes guidance, NICE recommends that long-acting insulin analogues (insulin glargine) should be used when patients on NPH insulin have a problem with nocturnal hypoglycaemia, or when morning hyperglycaemia leads to difficult daytime blood glucose control. They can also be used when rapid-acting insulin analogues are used for mealtime blood glucose control.13 NICE guidance from 2002 does not recommend insulin glargine for patients with type 2 diabetes unless they suffer from recurrent episodes of hypoglycaemia, they require assistance with their insulin injections or they would otherwise need twice daily basal insulin injections in combinations with oral antidiabetic drugs.15 It should be noted, however, that all relevant NICE guidance was produced before insulin detemir was launched. Until further guidance on these insulins Inhaled insulin (Exubera®▼) is a fast-acting, dry powder form of insulin that is inhaled via a specially designed device. It was introduced in 2006 for use as a mealtime insulin for adults with type 1 diabetes, and for adults with type 2 diabetes who require insulin therapy.65 There is currently little evidence that inhaled insulin has any clinical advantages over short-acting soluble subcutaneous insulin injections. A recent meta-analysis assessed 16 open-label RCTs of inhaled insulin in patients with type 1 (n=1,534) and type 2 diabetes (n=2,489) and concluded that it has a comparable glycaemic efficacy to short-acting soluble insulin injections, in terms of its effect on HbA1C.66 A statistically significant difference in HbA1C in favour of the soluble insulin injection was found, but this was considered to be clinically unimportant. In addition, no difference was found between inhaled and subcutaneous insulin in the proportion of patients reporting at least one episode of severe hypoglycaemia. RCTs were for 12 to 24 weeks duration, but one study compared treatments for two years.66 Some RCTs have reported that patients may prefer inhaled insulin, but the comparisons were often with needles and syringes rather than patient-friendly pen devices commonly used in practice.67 Also, the methods used to obtain some of the patient-preference data have been questioned.68 The committee for the NICE technology appraisal expressed concern about the designs of the RCTs that were reported. They pointed out that none of the trials compared inhaled insulin with the current standard of care in the UK and this limited the ability to generalise the results to routine clinical practice.68 Non-productive cough has been reported more frequently with inhaled insulin than with subcutaneous insulin or oral hypoglycaemic agents in RCTs (16.9% vs. 5.0%). In addition, patients using inhaled insulin had a small statistically significant reduction in pulmonary function (assessed by FEV1) compared with insulin injections, which progressed slowly up to six months, but stabilised by two years in one study.66 Inhaled insulin is contraindicated in patients with lung disease or in recent smokers, and lung function should be checked during treatment.65 A cautious approach to using this product should be adopted until its long-term safety over several years has been established. Whilst inhaling insulin might seem like an exciting alternative to injections, it does not completely eliminate the need for injections. All patients with type 1 diabetes, and many with type 2 diabetes, will still need one or more subcutaneous injections each day of a longer-acting basal insulin. In addition it is expensive, the inhalation device is bulky and patients will still need to check their blood glucose. It is important to note that inhaled insulin is available only as 1mg and 3mg blisters (approximately equivalent to 3IU and 8IU of subcutaneous fast-acting insulin, respectively). In view of the above, NICE has restricted the use of inhaled insulin to specialist diabetes centres, only in people with type 1 or type 2 diabetes who have poor control and either an injection phobia (meeting DSM-IV criteria) or severe persistent problems with injection sites.68 Despite an increase in prescribing of the newer more expensive insulin analogues, there is currently no strong evidence that they result in large improvements in HbA1C compared with older insulins. Several studies that compared rapid-acting insulin analogues with short-acting soluble insulin, and long-acting insulin analogues with NPH insulin, have suggested that insulin analogues reduce hypoglycaemia, particularly at night. However, studies could not be completely blinded, this has not been seen in all RCTs, and definitions of hypoglycaemia have often varied between trials, making assessment of these benefits difficult. Therefore, in the absence of long-term safety data over many years, it is reasonable not to support a widespread policy of using insulin analogues first-line in most patients. Nevertheless, they still have a valuable role in managing specific patients, particularly those whose glycaemic control is suboptimal with older insulins, especially if they have problematic hypoglycaemia. Inhaled insulin does not appear to improve HbA1C or reduce hypoglycaemia compared with short-acting soluble insulin used at mealtimes. In addition, it does not completely eliminate the use of injections in most patients. Therefore, NICE have restricted its use to specialist diabetes centres, only in people with type 1 or type 2 diabetes who have poor control and either an injection phobia (meeting DSM-IV criteria) or severe persistent problems with injection sites.68 Any decision to start a newer insulin needs to be balanced carefully against the lack of long-term safety data over many years and increased prescribing costs. In addition, people with control problems should be properly assessed for underlying causes before these newer, more expensive insulins are considered. This includes ensuring that they have been properly educated and understand how to manage their disease and treatment.
The National Institute for Health and Clinical Excellence (NICE) is associated with MeReC Publications published by the NPC through a funding contract. This arrangement provides NICE with the ability to secure value for money in the use of NHS funds invested in its work and enables it to influence topic selection, methodology and dissemination practice. NICE considers the work of this organisation to be of value to the NHS in England and Wales and recommends that it be used to inform decisions on service organisation and delivery. This publication represents the views of the authors and not necessarily those of the Institute. NPC materials may be downloaded / copied freely by people
employed by the NHS in England for purposes that support
NHS activities in England. Any person not employed by
the NHS, or who is working for the NHS outside England,
who wishes to download / copy NPC materials for purposes
other than their personal use should seek permission
first from the NPC. National Prescribing Centre, The Infirmary, 70 Pembroke Place, Liverpool, L69 3GF Tel: 0151 794 8146 Fax: 0151 794 8139
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Freedom of Information | Accessibility | Terms and Conditions |
|||||
News & RSS
