Diabetes Prevalence and Disease Burden
Diabetes Prevalence and Disease Burden
The prevalence of obesity and obesity-related complications, such as diabetes, is increasing at an alarming rate in both industrialized and developing countries. While it is currently estimated that 7.8% of the US population has diabetes, of whom 5.7 million are still undiagnosed,1 it is expected that by 2030 the worldwide prevalence of diabetes will reach 366 million,2 and even that figure may be conservative. Of even greater concern is the increasing prevalence of type 2 diabetes in young adults in the US, whose numbers are expected to double by 2050.3 Of the 246 million people affected worldwide, the International Diabetes Federation (IDF) calculates that 46% fall into the 40–49-year-old age group.4
Not only is earlier age at diagnosis of diabetes associated with greater glycemic exposure and consequently a higher risk for the development of nephropathy, retinopathy, and neuropathy, but also the associated comorbidities unique to type 2 diabetes and the metabolic syndrome could potentially translate into vascular events occurring at a much younger age than currently expected. For example, a recent publication projected that adults with early-onset type 2 diabetes (18–44 years of age) had a 14-fold higher risk for myocardial infarction (MI) than a control population, compared with the four-fold increased risk observed in usual-onset type 2 diabetes.5
Diabetes carries with it significant morbidity and mortality, and consequently a large economic burden. In the US, the projected cost of care for patients with diabetes has increased from $92 billion in 2002 to $174 billion in 2007,6 with worldwide estimates that in 10–15 years 7–13% of the world’s healthcare budget will be spent on diabetes treatment.4
Current Status of Diabetes Management
While hyperglycemia has been associated with an increased risk for microvascular and macrovascular complications, improvement in glycemic control has been clearly shown to reduce the risk for retinopathy, nephropathy, and neuropathy, the latter mostly in type 1 diabetes, with an estimated 20–30% relative risk reduction for every 1% fall in glycated haemoglobin (HbA1c).7–9 Additionally, intensive insulin therapy in patients with type 1 diabetes during the Diabetes Control and Complications Trial (DCCT) decreased the future risk for both microvascular and macrovascular complications during the 17 years of follow-up. Specifically, the intensively controlled subjects of the DCCT experienced a statistically significant 57% relative risk reduction for the occurrence of a non-fatal MI, stroke, or death from cardiovascular (CV) disease.10
In patients with type 2 diabetes, on the other hand, the impact of tight glycemic control on CV risk is less clearly defined. While at the conclusion of the United Kingdom Prospective Diabetes Study (UKPDS) the intensively managed group had a reduction in MI of borderline significance,11 with a clear association between lower HbA1c and reduced heart attacks,12 recently published trials have failed to show a beneficial effect of tight glycemic control on CV end-points.13,14 Although overall mortality rates were much lower than the predicted figure of 5% per patient per year (1.4 versus 1.1% per patient per year in the intensively treated versus the conventionally treated groups, respectively), subjects in the intensively treated arm of the ACCORD trial experienced a 22% (p=0.02) relative risk increase in mortality compared with conventionally treated patients. Subjects with no prior CV events and subjects with baseline HbA1c ≤8% had statistically significant reductions in the primary outcome of non-fatal MI, stroke, or CV death.13
The interim conclusion that might be drawn from the currently available evidence is that while tight glycemic control reduces microvascular complications in type 2 diabetes, it has considerably less impact on reducing CV mortality in this patient population. Additionally, subjects who are older, have had prior macrovascular events, or are at increased risk for severe hypoglycemia might not be ideal candidates for tight glycemic control.
Published guidelines from professional organizations recommend near-normal glycemic targets for reducing microvascular complications in diabetes. While the American Diabetes Association (ADA) suggests an HbA1c level <7% and the American Association of Clinical Endocrinologists (AACE) or the European Association for the Study of Diabetes (EASD) recommend an HbA1c target ≤6.5%, the common consensus is that patients with diabetes should strive to achieve and maintain an HbA1c as close to normal as possible without ‘significant’ hypoglycemia.15,16 Subjects with diabetes should also maintain strict blood pressure control, as well as optimal lipid management, to reduce the risk for a CV event or demise.
Available data from the US show there is ample opportunity to improve the treatment process and outcomes. Of patients thought to have diabetes, hypertension, or dyslipidemia, 35–50% have not been diagnosed with the condition, possibly as a result of lack of access to medical services, failure of providers to screen for the condition, and/or the silent nature of the disease.17–19 Of patients thought to have diabetes, hypertension, and dyslipidemia, fewer than half are receiving clinically indicated treatment to manage the condition and prevent future vascular complications.20 In addition, even when these medical conditions are appropriately diagnosed, fewer than half of patients achieve or maintain individual metabolic targets, and fewer than 10% have achieved appropriate treatment goals for blood glucose, blood pressure, and cholesterol. Contributing to this inability to achieve and maintain metabolic goals is the ‘failure’ of the provider to initiate and/or intensify therapy when clinically indicated, described in the medical literature as ‘clinical inertia.’21 Delaying the intensification of therapy can result in prolonged exposure to hyperglycemia, with subjects on dual oral agent therapy spending over two years with HbA1c values >8% (over four years >7%), and intensification of therapy, mostly in the form of insulin initiation, on average occurring when the HbA1c had exceeded 9.5%.22 Explanations from providers as to the most common reason for not modifying therapy when metabolic goals were not being met ranged from impression that metabolic control was improving (34%), to medication non-compliance (16%), to dietary noncompliance (10%), to no particular reason (18%).23
Available Treatment Options and Consensus Recommendations
The primary goal for blood glucose control remains the prevention of microvascular damage. With this in mind, providers should strive to individualize glycemic goals on the basis of patient age, hypoglycemia risk, presence of CV risk factors, and the presence and severity of microvascular complications. Once glycemic goals have been individually defined, a number of treatment algorithms are available to safely and effectively achieve those metabolic targets.
A combined consensus statement from the ADA/EASD published a few years ago recommended starting metformin along with lifestyle interventions in asymptomatic patients with a new diagnosis of type 2 diabetes, as long as no contraindications—such as renal insufficiency—were present.24 Additional therapies could be implemented based on individual needs, such as insulin replacement for greatest effectiveness, sulfonylureas as an inexpensive option, or thiazolidinediones if the goal was to minimize the risk for hypoglycemia. A recent systematic review comparing the effectiveness and safety of oral antiglycemic agents supports the choice of metformin as first-line therapy, given its equivalent glucose-lowering potential to other non-insulin therapies, its low risk for hypoglycemia, its weight-neutral effect, and its potential for CV risk reduction in overweight patients with type 2 diabetes.11,25 Compared with other monotherapies, thiazolidinediones have been shown to result in better long-term glycemic control (or maintenance of control) than glyburide or metformin, although this appears be due to improved insulin sensitivity and not enhanced beta-cell function.26 In A Diabetes Outcome Progression Trial (ADOPT), glyburide was associated with the greatest hypoglycemia risk but the fewest CV events, while rosiglitazone was associated with the most fluid retention and an increased risk of extremity fractures in women.26,27 While comparative long-term data on the safety and effectiveness of incretin-based therapies—such as exenatide or sitagliptin—are lacking, given the low risk of hypoglycemia and the weight-beneficial effect of these therapies it becomes conceivable that they may also be associated with improved maintenance of glycemic control over the long term.
Role of Insulin Therapy in the Management of Hyperglycemia
Management of hyperglycemia in the setting of worsening beta-cell function in type 2 diabetes will require, for many patients, the introduction of insulin replacement therapy. With the exception of patients who are highly symptomatic or ketotic, the debate as to when to initiate insulin treatment in patients with type 2 diabetes continues. Most physicians would agree on starting insulin therapy when other approaches have failed to reduce glycemic exposure to a level that would minimize diabetes-related complications. Discussions with patients regarding insulin therapy should occur long before the need to implement it. Patients tend to be concerned about starting insulin therapy for a number of reasons, including misconceptions about the consequences of insulin therapy, discomfort related to administration and monitoring, lifestyle intrusion, implications about disease state, and doubts about the beneficial effects of insulin replacement.28,29 Additionally, worries about insulin-related weight gain and hypoglycemia pervade any conversation regarding insulin management between patient and healthcare provider.
Ideally, any decision regarding initiation of insulin replacement therapy should take into consideration glycemic goals, risks for treatment-related adverse effects, and specific patient issues and concerns. The agreement to start insulin treatment should be followed by a discussion regarding the advantages and drawbacks of specific insulin replacement approaches. Factors such as current HbA1c level, timing of glycemic burden (fasting, post-prandial, or both), frequency of insulin administration, complexity of insulin regimen, and common side effects of therapy—namely hypoglycemia risk and weight gain— need to be considered.
In patients with type 2 diabetes, the choice of initial insulin treatment usually falls between basal insulin replacement and pre-mixed insulin therapy. In subjects with baseline HbA1c >8.5%, studies comparing these two options have in general demonstrated greater glycemic improvement with a premixed approach, at the expense of more weight gain and more frequent nonsevere hypoglycemia.30–32 The use of metformin with pre-mixed insulin therapy can minimize weight gain.33 Basal insulin replacement may be preferred to pre-mixed insulin therapy in patients whose HbA1c is not excessively high, whose fasting blood glucose levels are elevated, who require lifestyle flexibility and a minimum of daily injections, and whose tolerance to weight gain or hypoglycemia is low (see Table 1). Simple and practical algorithms to initiate, monitor, and adjust basal insulin therapy make it a very effective strategy in the insulin-naïve patient with type 2 diabetes. Most practitioners would agree that basal insulin therapy should be initiated much earlier than current practice,22 with maximum glycemic benefit evident before the HbA1c level exceeds 8.5%.32
General Concepts in Initiating and Adjusting Basal Insulin Therapy
Basal insulin analogs have been developed to more ‘physiologically’ replace basal insulin needs in patients with beta-cell insufficiency. Modifications in their amino acid composition result in alterations in their pharmacokinetic (insulin concentration in plasma) and pharmacodynamic (insulin time–action) properties. This allows delayed absorption into the circulation, a longer duration of action, a lower peak of insulin concentration, and a more consistent absorption profile and predictable glycemic effect.34–36 Comparisons of basal insulin replacement with human versus analog preparations have consistently shown similar efficacy, but a lower risk for hypoglycemia with the insulin analogs.37–40 Moreover, insulin detemir has been consistently associated with less weight gain than other basal preparations.38,41,42
Algorithms that have been developed for the initiation and adjustment of basal insulin therapy focus on simplicity, safety, and ease of use. In most cases they recommend initiating basal insulin as a once-daily evening dose, either with 10–12 units or based on a patient’s weight (using 0.2 units/kg/day).39,41 Patient self-monitoring of fasting blood glucose (FBG) levels is then used by the physician to make periodic adjustments to the evening basal insulin dose until pre-set FBG targets are met. FPG goals can range from <110–110mg/dl in the various treat-to-target studies to <130mg/dl based on ADA guidelines.15 While in general most of the basal insulin dose adjustments are performed by health professionals, many patients can be provided with simple titration rules and schedules and thereby effectively and safely self-adjust their own insulin doses, in collaboration with their treating physician (see Table 2).43–45 Most of the glycemic reduction from basal insulin initiation in insulin-naïve subjects with type 2 diabetes is achieved within the first three months of treatment, with final basal insulin doses averaging anywhere from 0.4 to 0.6 units/kg/day.
Detemir Insulin in the Management of Insulin-requiring Type 2 Diabetes
Insulin detemir was developed by removing the amino acid threonine at position 30 on the beta-chain of the human insulin molecule, along with acylation of myristic acid to the lysine in position B29 (see Figure 1). These modifications encourage reversible binding of insulin detemir to albumin, as well as self-association of insulin detemir molecules into hexamers and di-hexamers.46 The affinity of insulin detemir to albumin binding results in a lower insulin receptor affinity, contributing to the need for an increase in molar insulin concentration of insulin detemir in commercial preparations.47
Insulin detemir is a clear, sterile solution with a neutral pH. Following subcutaneous injection, insulin detemir binds to albumin and further selfassociates into di-hexamers. Even though in circulation approximately 98% of insulin detemir is albumin-bound,48 conditions of hypoalbuminemia or renal protein loss do not appear to affect its pharmacokinetics. Insulin detemir has an onset of action of between one to two hours, and in patients with type 2 diabetes has a small peak of action, comparable to insulin glargine, approximately nine to 11 hours after injection.49 In this latter euglycemic clamp study, insulin detemir had a duration of action of over 24 hours in most subjects. A recent review of euglycemic clamp studies in subjects with type 1 or type 2 diabetes receiving insulin detemir or insulin glargine concluded that these two insulin analogs had similar and near-peakless time–action profiles.35
Although both basal insulin analogs have a significantly longer duration of action than human neutral protamine Hagedorn (NPH) insulin,36,50 studies in type 2 diabetes comparing either analog versus NPH insulin have consistently shown similar efficacy in lowering HbA1c or FPG.38,39,41 Three basal-only insulin replacement studies involving detemir insulin in a treat-to-target format have been published in insulin-naïve subjects with type 2 diabetes.38,41,42
The first study was a 20-week randomized controlled trial (RCT) comparing evening NPH insulin (n=166) versus evening (n=170) or morning (n=168) detemir insulin, starting with mean baseline HbA1c levels of 9.2, 8.9, and 9.1%, respectively.41 Insulin doses were titrated at least every four weeks based on a pre-set adjustment algorithm using the mean of three consecutive plasma glucose results and aiming for a fasting plasma glucose ≤108mg/dl (≤6.0mmol/l) for the evening insulin doses, or a similar pre-dinner plasma glucose target for the morning detemir schedule. At study end, mean HbA1c reductions were similar across treatment groups (-1.74, -1.48, and -1.58% for evening NPH, evening detemir, and morning detemir, respectively; p=NS). Hypoglycemia frequency was significantly lower with detemir insulin. Specifically, overall and night-time hypoglycemia were significantly reduced by 53 and 65%, respectively, in subjects on evening detemir insulin compared with those on evening NPH insulin, with an 87% relative risk reduction (p<0.001) when comparing morning detemir insulin with evening NPH insulin. Weight gain was significantly lower for evening detemir insulin compared with evening NPH insulin (0.7 versus 1.6kg; p=0.005), with patients on morning detemir insulin gaining 1.2kg by study end. Mean daily doses of insulin at 20 weeks were similar at 0.4, 0.4 and 0.5 units/kg for evening NPH, evening detemir, and morning detemir insulin, respectively. The second treat-to-target study in insulin-naïve subjects with type 2 diabetes poorly controlled with oral antidiabetic agents (OADs) compared once-daily glargine insulin versus once-daily detemir insulin, with the option of twicedaily detemir insulin administration.42 This was a 52-week non-inferiority RCT (n=582) in which both insulin preparations were initiated as a 12-unit dose in the evening and periodically adjusted, according to a pre-set algorithm, based on a fasting plasma glucose target of ≤108mg/dl (≤6.0mmol/l). Additionally, the insulin detemir arm had the option of adding a second dose of insulin in the morning if fasting plasma glucose levels were <126mg/dl (<7.0mmol/l) but pre-dinner plasma glucose remained >126mg/dl (>7.0mmol/l). By study end, HbA1c levels had declined from a baseline of 8.6% in all groups to 7.1% in completers on once-daily detemir (n=104), twice-daily detemir (n=127), and insulin glargine (n=275). The HbA1c difference between all detemir and glargine patients was 0.05%, supporting the hypothesis of non-inferiority of insulin detemir compared with insulin glargine. While subjects on detemir insulin used larger insulin doses (especially those switched to twice-daily dosing), they nevertheless gained less weight than subjects on glargine insulin (see Table 3). This study appears to support the conclusion that basal insulin treatment in type 2 diabetes is optimally dosed once daily, and that splitting the basal insulin dose twice daily to control daytime hyperglycemia, presumably a result of insufficient prandial insulin replacement, is not an effective strategy.
The third treat-to-target study compared twice-daily NPH (n=225) versus detemir (n=227) insulin in a 26-week RCT in insulin-naïve subjects with suboptimal glycemic control on OADs.38 Starting with six to 10 units of insulin before breakfast and dinner, doses were adjusted periodically based on similar pre-set algorithms, aiming for pre-breakfast and pre-dinner plasma glucose levels ≤108mg/dl (≤6.0mmol/l). Baseline HbA1c levels were 8.6 and 8.5% in the detemir and NPH insulin groups, respectively, decreasing to 6.8 and 6.6% by study end, respectively. Overall, 70% of participants in both groups achieved an HbA1c target of ≤7.0%, although significantly more subjects in the detemir group achieved glycemic goals without hypoglycemia during the last 12 weeks of treatment (26 versus 16%; p=0.008). Despite subjects on detemir insulin injecting larger insulin doses by study end (approximately 0.78 versus 0.53 units/kg/day), patients in the insulin detemir group gained significantly less weight over the course of the trial (1.2 versus 2.8 kg; p<0.001). Interestingly, subjects with a higher baseline body mass index (BMI) tended to gain less weight while on insulin detemir (see Figure 2).
There have been only a few published RCTs assessing the use of detemir in basal–bolus combination in type 2 diabetes. The first available trial compared analog (basal detemir with prandial aspart insulin replacement) versus non-analog (basal NPH with prandial regular insulin replacement) strategies for glycemic control in subjects (n=395) on prior insulin therapy in a 22-week RCT.51 Baseline HbA1c (8.2 versus 8.1% for analog and non-analog insulin groups, respectively) values decreased by -0.65 and -0.58%, respectively (p=NS for between-group differences), while there was significantly less weight gain in the group on the detemir plus aspart insulin combination (0.51 versus 1.13kg; p=0.038). There was a trend for less overnight hypoglycemia with the detemir plus aspart combination (relative risk reduction of 38%; p=0.14), which did not reach statistical significance. Mean total daily insulin doses at 22 weeks were 0.95 and 0.79 units/kg in the analog and non-analog insulin groups, respectively.
A second 26-week RCT compared the efficacy and safety of basal–bolus insulin replacement with detemir insulin versus NPH insulin, both in combination with prandial aspart insulin, in subjects (n=505) with type 2 diabetes previously on insulin therapy.52 Baseline HbA1c levels were close to goal in both the detemir and the NPH groups (7.9 and 7.8%, respectively) and decreased by -0.2% (p=0.004 compared with baseline) and -0.4% (p=0.0001 compared with baseline), respectively, by study end. Fewer than 2% of subjects in either treatment group experienced an episode of severe hypoglycemia; the frequency of overall and night-time hypoglycemia was similar in the two groups. Subjects in the detemir insulin group gained significantly less weight over the course of the trial than those on NPH insulin (1.0 versus 1.8kg, respectively; p=0.017). In both the Raslova and Haak studies, the subjects randomized to insulin detemir experienced less variability in self-measured fasting blood glucose, a result that reached statistical significance but is of unclear clinical significance.
Several observational studies have followed these RCTs and have consistently supported the findings of efficacy, low hypoglycemia risk, and weight advantage of detemir insulin when initiated in insulin-naïve subjects, or substituted to ongoing basal insulin treatment.45,53–56
In conclusion, the weight of the evidence supports detemir insulin as an effective option in the management of type 2 diabetes, associated with a lower risk for non-severe hypoglycemia than NPH insulin, and a weight-sparing effect, even in the setting of somewhat higher insulin doses, compared with either NPH or glargine insulin. Insulin analog preparations, because of their more physiological time–action profile and other characteristics, are possibly better choices than non-analog insulin options when trying to optimize blood glucose control in subjects approaching glycemic targets.■