European Journal of Internal Medicine

Transcription

1 European Journal of Internal Medicine 20 (2009) S329 S339 Contents lists available at ScienceDirect European Journal of Internal Medicine journal homepage: Clinical Application of Incretin-Based Therapy: Therapeutic Potential, Patient Selection and Clinical Use David M. Kendall, MD, Robert M. Cuddihy, MD, Richard M. Bergenstal, MD International Diabetes Center at Park Nicollet, Minneapolis, Minnesota, USA article info abstract Keywords: GLP-1 receptor agonist DPP-4 inhibitor liraglutide exenatide sitagliptin vildagliptin Incretin-based therapies address the progressive nature of type 2 diabetes mellitus, not only by addressing glucose control but also with weight-neutral (i.e., dipeptidyl peptidase 4 inhibitors sitagliptin and vildagliptin) and weight-reducing effects (i.e., glucagonlike peptide 1 [GLP-1] receptor agonists exenatide and liraglutide). Preclinical data suggest that incretin-based therapies may also preserve β-cell function, holding promise of a truly disease-modifying therapy. This article examines clinical trial data and accepted algorithms with a view toward elucidating the application of these agents in routine clinical practice. We propose a systematic approach to treatment, addressing (1) patient selection, (2) optimal treatment combinations, and (3) timing and guidance for both initiation and intensification of therapy. The GLP-1 receptor agonists, for example, could be particularly beneficial in patients whose weight significantly increases cardiovascular risk. Early use of these agents may be effective in preventing diabetes in those at risk, or in halting or retarding disease progression in patients with frank diabetes. Additional clinical investigation will be required to test such hypotheses. Given the ever-increasing incidence of diabetes worldwide, the link between obesity and the development of type 2 diabetes, and the need for more effective, weight-focused, convenient and sustainable treatments, the data from such studies will be invaluable to further clarify the role of the incretins in the management of patients with type 2 diabetes European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved. Incretin-based therapies have recently been introduced into clinical practice and are currently the newest class of glucose-lowering agents available for the treatment of type 2 diabetes mellitus. These novel therapies exert their glucose-lowering effect by myriad mechanisms and are able to achieve improved glucose control with either no weight gain (dipeptidyl peptidase 4 [DPP-4] inhibitors, or with weight loss (glucagon-like peptide 1 [GLP-1] receptor agonists. These clinical effects, combined with unique mechanisms of action, suggest that incretin-based therapies may be appropriate for many patients with type 2 diabetes, particularly those in whom treatment may be limited by hypoglycemic risk, progressive weight gain, elevated postmeal blood glucose levels, or increased food intake. 1. A Pathophysiologic Perspective Progressive impairment of β-cell function and increased insulin demand as tissue becomes insulin resistant are core pathophysiologic This article is a copublication with The American Journal of Medicine, 122, S37-S50. For citation purposes please use European Journal of Internal Medicine, 20, S329-S339. DOI of original article: /j.amjmed Requests for reprints should be addressed to Richard M. Bergenstal, MD, International Diabetes Center, 3800 Park Nicollet Boulevard, Minneapolis, Minnesota address: Richard.Bergenstal@parknicollet.com (R.M. Bergenstal). defects in the development and progression of hyperglycemia in type 2 diabetes [1,2]. However, other important factors are also known to further exacerbate this pathology. Excess glucagon secretion, abnormal gastric emptying during hyperglycemia, obesity, and increased food intake all contribute to, or worsen, hyperglycemia in type 2 diabetes. Impaired release or action of incretin hormones, particularly GLP-1, and to a lesser degree glucose-dependent insulinotropic polypeptide (GIP), also play a role in the development and/or progression of type 2 diabetes. More recently, this essentially glucocentric view of the genesis and progression of type 2 diabetes has been supplemented by a more lipocentric perspective. Here the major mechanism is progressive ectopic lipid deposition (e.g., in myocytes and hepatocytes, rather than in adipocytes). Build-up of ectopic fat in those tissues ultimately induces insulin resistance, β-cell lipotoxicity, and diminished β-cell function, leading to metabolically inadequate insulin secretion [3]. 2. Incretin-Based Therapy: A Clinical Perspective Traditional treatment strategies for type 2 diabetes use single therapies, often added in a sequential fashion. As additional therapies are added only when specific treatment targets are not achieved, this failure-based approach often results in a delay in advancing therapy [4-6]. Current treatment recommendations for the management of type 2 diabetes advise lifestyle changes moderate weight loss (5% to /$ see front matter 2009 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved. doi: /j.ejim

2 S330 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 10%) and physical activity for 30 minutes per day [7,8]. If appropriate, metformin is recommended as initial therapy, with other agents such as thiazolidinediones, insulin secretagogues (sulfonylureas, meglitinides), and exogenous insulin added sequentially thereafter [9]. This approach to treatment has been broadly used, but it has also been shown to result in a gradual deterioration in glucose control and progressive elevation of hemoglobin A 1c (HbA 1c ) [10]. Traditional therapies have also been associated with significant weight gain [11] and an increased risk for hypoglycemia [12]. Current therapies approved for the treatment of type 2 diabetes exert their effects by means of unique mechanisms of action (e.g., targeting insulin resistance, improving insulin secretion, affecting gut absorption of glucose, enhancing the effect of incretin hormones). However, no single therapy has been convincingly shown to prevent the progressive deterioration in blood-glucose control associated with diabetes. Treatment-limiting side effects and/or intolerance to medication diminish the likelihood that even effective therapies will continue to remain so over time. The likely need for multidrug therapy also makes it difficult to achieve and sustain good glycemic control [10]. Titration issues, side effects, and problems with patient adherence to therapy become magnified as agents are added [13,14]. Treatment burden increases as complications and comorbidities develop, each of which may also require multidrug regimens. As such, it is critical that practitioners understand the benefits and drawbacks of each new antidiabetic therapy. Two classes of therapy, the GLP-1 receptor agonists (exenatide, liraglutide, taspoglutide, AVE0010, albiglutide) and the DPP-4 inhibitors (sitagliptin, vildagliptin, alogliptin, saxagliptin), seek to leverage the pharmacologic effects of incretin hormones. The development and introduction of these agents, combined with their unique mechanisms of action, have generated substantial clinical interest. However, many physicians and other healthcare providers have limited experience with these therapies; moreover, their most effective clinical application has not yet been fully elucidated. In this review, we discuss the clinical effect of the incretin-based therapies and their application in routine clinical practice. We describe a systematic approach to treatment that addresses appropriate patient selection, optimal treatment combinations, and timing and guidance for both initiation and intensification of therapy. We also review the unique characteristics of incretin-based therapies on postmeal blood-glucose control, body weight, cardiovascular risk factors, and β-cell secretory function. 3. Current Clinical Challenges in the Treatment of Diabetes Mellitus The vast majority of patients with type 2 diabetes currently are treated with 1 oral hypoglycemic agent with increasing numbers of patients now using insulin (most commonly basal insulin) earlier in the course of treatment. Despite the introduction of several new classes of antidiabetic agents in the United States and worldwide between 1995 and 2004, slightly >50% of patients with type 2 diabetes achieve HbA 1c values 7% (the current community standard advised by the American Diabetes Association [ADA]) [15]. That almost half of the population of patients with type 2 diabetes still have HbA 1c values >7%, even with stepped addition of oral glucoselowering therapies, suggests a significant unmet therapeutic need. Data from the landmark United Kingdom Prospective Diabetes Study (UKPDS) suggest that within 3 years of diagnosis 50% of patients with type 2 diabetes will require multiple therapies ( 1 oral agent, including a combination of oral agents with insulin) to achieve glycemic target [16]. Koro and colleagues [17] have reported a decline in the percentage of patients with diabetes who are able to achieve HbA 1c target on a broad array of therapies from 44.5% for the National Health and Nutrition Examination Survey (NHANES) III (1988 to 1994) to 35.8% for NHANES 1999 to More recently, Saaddine and colleagues [18] analyzed NHANES data (1988 to 1994 and 1999 to Figure 1. Postulated role of insulin resistance, β-cell dysfunction, and an impaired incretin effect in the pathogenesis of type 2 diabetes mellitus. (Adapted from J Clin Endocrinol Metab, [23] Diabetes, [24,27] Eur J Clin Endocrinol, [25] and J Clin Invest [26].) 2002), as well as Behavioral Risk Factor Surveillance System data and found that over the past decade the proportion of patients with poor glycemic control declined slightly (by 4%) but nonsignificantly. When Hoerger and coworkers [19] included NHANES 2003 to 2004 data in the analysis, mean HbA 1c levels showed a slight but significant decrease (0.5%; P = 0.03) when compared with 1999 to 2000 data, and for the first time reported that just over half of the population surveyed (55.7%) had an HbA 1c level b7%. However, in the recently completed Action to Control Cardiovascular Risk in Diabetes (ACCORD) and Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) studies, even conventional treatment regimens required 2 therapies to achieve mean HbA 1c values of 7.4% [20,21]. The efficacy of earlier implementation of more aggressive multidrug treatment approaches, combined with lifestyle modification, while promising, needs to be evaluated in clinical trials that also address many of the other challenges associated with the clinical management of type 2 diabetes. The need for multidrug therapeutic regimens to treat comorbid conditions, mealtime dosing of many medications, treatment-related adverse effects (including weight gain, hypoglycemia, and the potential for drug drug interactions), equitable access to quality team care, and the high cost of multiple medications represent only some of the challenges faced by clinicians and patients alike. Accordingly, the complexities of appropriate treatment design, combined with the progressive natural history of type 2 diabetes and the presence of other comorbidities, underscore the need for new, unique, and effective therapies that can reduce or eliminate some of the barriers to effective care [1,2,22]. The introduction of incretin-based therapies (including GLP-1 receptor agonists and DPP-4 inhibitors) represent one such advance, and the successful clinical application of these therapies has significantly expanded the clinician's options for helping patients manage type 2 diabetes more effectively. 4. A Pathophysiologic Perspective: The Natural History of Type 2 Diabetes Mellitus Intelligent use of antidiabetic drugs is critically dependent on a sound understanding of the natural history of diabetes and of the underlying pathophysiologic defects that give rise to the disease. Type 2 diabetes is a complex metabolic disorder characterized by elevated blood glucose and a marked increase in the risk of cardiovascular disease. The increase in cardiovascular disease risk is a consequence of a cluster of metabolic and vascular abnormalities including hyperglycemia, dyslipidemia, and hypertension. An important pathophysiology in patients with type 2 diabetes is the presence of tissue insulin resistance. Insulin resistance (often tied

3 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 S331 Figure 2. Representative depiction of the natural history of type 2 diabetes mellitus highlighting the role of insulin resistance, insulin deficiency, and impaired incretin effect. Both the time course and relative function are descriptive. These 3 core pathophysiologic defects likely combine to contribute to the progressive nature of the disease, and may account for much of the deterioration in glucose control observed clinical in patients with type 2 diabetes. IFG = impaired fasting glucose; IGT = impaired glucose tolerance. For glucose, 1 mg/ dl = mmol/l. (Courtesy of the International Diabetes Center 2008.) to obesity and/or adiposity) and defects in insulin secretion (relative insulin deficiency) have long been discussed as the key factors that give rise to the hyperglycemia characteristic of type 2 diabetes. More recently, investigators have reported that impairments in the secretion levels and/or the activity of key incretin hormones (i.e., gut hormones contributing to the regulation of carbohydrate metabolism, pancreatic islet-cell function, and plasma glucose) may also play a significant role in the development and progression of hyperglycemia in type 2 diabetes (Figure 1) [23-27]. Studying the natural history of type 2 diabetes, it is clear that in response to insulin resistance, β-cells will secrete additional insulin, and the insulin levels will initially increase (Figure 2: lower panel, insulin level). This increase in insulin secretion actually still represents a relative deficiency of insulin because, as can be seen in the lower panel of Figure 2, early in the course of the natural history of type 2 diabetes, β-cell function starts to deteriorate. The existence of impaired β-cell function is a necessary and sufficient condition for the appearance of prediabetes and the onset of frank diabetes. Figure 2 shows that at the onset of diabetes (when insulin secretion no longer keeps pace with insulin resistance), there is already a significant reduction in β-cell function. By the time diabetes is clinically diagnosed (often 10 years after the onset of diabetes according to the UKPDS), β-cell function may be reduced by 50% [28]. β-cell impairment is now recognized to arise and worsen years before the onset of diabetes and its initial manifestation of postprandial hyperglycemia [29]. Whether as cause or consequence, an incretin defect or, as shown in the lower panel of Figure 2, the decline in incretin effect (level or action) is clearly present in type 2 diabetes [30]. Augmentation/correction of the impaired incretin effect, particularly augmentation of GLP-1, should therefore significantly improve blood glucose control in diabetes [31,32]. 5. Clinical Application of Incretin-Based Treatments Both GLP-1 receptor agonists and the DPP-4 inhibitors seek to leverage the pharmacologic effects of incretin hormones. Incretinbased agents provide clinicians with new interventional strategies for the improvement of fasting and postmeal glucose, islet function, and body weight Treatment Choices: A Rational Approach to Treatment Decisions Traditional treatment strategies for patients with type 2 diabetes have generally used a stepwise, failure-based approach to medication selection. Simply stated, this approach features the sequential addition of therapies only when glycemic control deteriorates significantly. This treatment paradigm was systematically evaluated in the landmark UKPDS [10]. Although most researchers and clinicians today would say that this overall approach is not optimal, it has shown some impressive benefits. Early, aggressive intervention is now known to exert a legacy effect, even years after patients have opted out of intensive therapy protocols, fostering a sustained reduction in the risk of both micro- and macrovascular complications [33,34]. However, even early, aggressive therapeutic intervention, at least as implemented in trials to date, does not halt the relentless progression characteristic of type 2 diabetes. The addition of multiple antidiabetic agents becomes necessary as glycemic control deteriorates, and many patients will eventually require insulin supplementation. In the years following publication of the UKPDS, there has been a greater focus on treatments that target the core pathophysiologic defects of type 2 diabetes, particularly impaired β-cell function. The availability of newer, more innovative therapies has resulted in the development of several algorithm-based treatment approaches designed to assist clinicians with earlier intervention and more effective therapy selection. In the remarks that follow, we will focus on the clinical use of incretin-based therapies the GLP-1 receptor agonists and the DPP-4 inhibitors and provide insight into their unique clinical benefits and their place in the treatment paradigm for type 2 diabetes. A number of consensus treatment algorithms for the management of type 2 diabetes have been developed, including those from the ADA and from the European Association for the Study of Diabetes (EASD), the American Association of Clinical Endocrinologists (AACE), and the Canadian Diabetes Association [35-37]. While each of these algorithms describe a possible role for new classes of blood glucoselowering medications, limited guidance is provided as to which therapy may be best suited for which individual patient. The increased number of choices available to practitioners and patients may heighten uncertainty regarding regimen design. There are few published studies with metabolic outcomes associated with specifically

4 S332 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 Figure 3. International Diabetes Center (IDC) treatment algorithm for the management of type 2 diabetes. This recommended approach to treatment underscores the central importance of established glycemic targets and the critical role of diabetes self-management education and nutrition and activity counseling. It further underscores the selection of drug therapies based on pathophysiologic characteristics and individual patient characteristics. To convert glucose measurements, 1 mg/dl = mmol/l. CV = cardiovascular; DPP-4 = dipeptidyl peptidase 4; FPG = fasting plasma glucose; GI = gastrointestinal; GLP-1 = glucagonlike peptide 1; OA = oral antidiabetic medication; RPG = random plasma glucose; SMBG = self-monitoring of blood glucose; SU = sulfonylurea; TZD = thiazolidinedione. *Saxagliptin and liraglutide pending US and EU regulatory review at time of this publication. Limited clinical data for combination therapy with insulin plus either DPP-4 inhibitor or GLP-1 agonist. (Adapted from Diabetes Care, [38,39,43,44] J Fam Pract, [40] J Clin Outcomes Manag, [41] Am J Manag Care, [42] Staged Diabetes Management: a Systematic Approach, [45] Endocr Pract, [46] and Popul Health Manag [47]. Courtesy of the International Diabetes Center 2009.) implementing any of these association consensus guidelines. The International Diabetes Center (IDC) has perhaps the longest history of algorithm development for glycemic control in diabetes, particularly focused on the needs of primary care providers who see >90% of the patients with type 2 diabetes. There are published outcomes describing the training, implementation, and outcomes achieved by multiple medical group practices implementing their own customized version of the IDC glycemic algorithm or decision path referred to as staged diabetes management [38-47]. Because the IDC glycemic algorithm (Figure 3) [38-47] has associated outcome data, is based on addressing glycemic targets (HbA 1c and selfmonitoring of blood glucose) and the pathophysiologic defects of type 2 diabetes, has been updated multiple times since 1994 to include new therapies where appropriate, and has been implemented in multiple US and international clinical settings, we will refer to this algorithm when discussing how incretin-based therapeutics may fit into clinical decision making in type 2 diabetes management Patient Selection and Patient Characteristics Impaired insulin secretion [1,48,49] and β-cell dysfunction, [50] insulin resistance, obesity, and an impaired incretin effect [30] all play a role in the pathogenesis and progression of type 2 diabetes (Figure 2). Understanding the interplay among these factors should enable clinicians to select the most appropriate therapy or therapies for the individual patient with type 2 diabetes. In addition, a better understanding of the effects of each therapy should improve clinical understanding of how best to use the newer therapies. The ability of GLP-1 receptor agonists to lower blood glucose and reduce body weight not only improves insulin secretion but also improves islet function (both α- and β-cell function) and may improve insulin action (insulin resistance), perhaps because of the reductions in body weight. Pleiotropic effects of these agents, although not yet definitively established, could open a new dimension in diabetes treatment (e.g., glycemic control agents that directly benefit the heart and the vascularity) [51-53]. Clinicians must also be aware of and work with patients to limit the risk of treatment-associated side effects. Although metformin is broadly advocated as first-line therapy for patients with type 2 diabetes, it is not well tolerated by some patients due to significant gastrointestinal side effects. Moreover, patients with impaired renal function and a history of heart failure are not appropriate candidates for metformin treatment [54,55]. Both sulfonylurea and insulin therapy are associated with significant weight gain and risk of hypoglycemia [56]. Thiazolidinedione

5 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 S333 therapy often used in patients with clinical evidence of cardiovascular disease and/or insulin resistance may also result in significant weight gain, fluid retention, and a higher risk of heart failure [57]. Newer concerns about thiazolidinediones and bone fractures have also been reported [58]. 6. Use of Incretin-Based Therapies: A Pathophysiologic Perspective Currently, metformin is widely recommended as the initial therapy of choice for patients with type 2 diabetes. Metformin is commonly used in combination with lifestyle interventions as the first step in treatment or as add-on therapy if patients are unable to achieve or maintain adequate glycemic control (HbA 1c 7%) with diet and exercise alone. Figure 3 [38-47] describes an approach to therapy selection based on the level of glycemic control at presentation, while emphasizing the central role of self-management, medical nutrition, and behavior changes in the management of type 2 diabetes. After metformin and lifestyle management, there are 4 reasonable choices of therapy, depending on individual patient characteristics as outlined in the algorithm (Figure 3) [38-47]. Afifth therapeutic option of basal insulin plus an oral agent may be considered at this stage in special circumstances, but in this algorithm insulin is normally suggested as a next step in combination with 2 drugs. GLP-1 receptor agonists and DPP-4 inhibitors make up 2 of the 4 options outlined as second choice add-on therapies, and deserve further discussion because they are relatively new therapies GLP-1 Receptor Agonists GLP-1 receptor agonists represent a novel class of therapies that leverage the glucoregulatory effects of the endogenous incretin hormone GLP-1. This class of agents includes twice-daily exenatide, originally derived from the salivary glands of the Gila monster, which has 53% homology with human GLP-1 [59]. Exenatide was first introduced for clinical use in It must be administered before morning and evening meals (up to 60 minutes) [60]. Several new GLP- 1 receptor agonists are currently in development; these include liraglutide, taspoglutide, albiglutide, and a once-weekly formulation of exenatide. GLP-1 receptor agonists significantly lower HbA 1c, whether used alone or in combination with other medications. Clinical trials with twice-daily exenatide as an add-on to 1 oral agent demonstrated HbA 1c reductions of 0.6% to 0.9% [61-63]. Subsequent open-label treatment [64] demonstrated sustained HbA 1c reductions averaging 1.1%. Exenatide therapy has also been compared with starter insulin in 4 separate open-label studies [65-68]; results indicate that exenatide at doses of 10 μg bid was similarly effective to either titrated basal or analog mix insulin (with HbA 1c reductions of 0.9% to 1.4%). A once-weekly (long-acting release) formulation of exenatide is in development and has been compared with the current twice-daily formulation in a 30-week trial [69]. Onceweekly exenatide therapy resulted in a significantly greater mean reduction in HbA 1c ( 1.9% vs. exenatide bid 1.5%). The once-weekly formulation was associated with lower rates of nausea the most common adverse effect seen with GLP-1 receptor agonists [66,69]. The once-daily human GLP-1 receptor agonist liraglutide is currently under review by the US Food and Drug Administration (FDA), and a significant amount of clinical data is available from reports of 6 large, randomized, placebo-controlled phase 3 trials. Liraglutide is 97% homologous to native GLP-1, with a single amino acid substitution extending its half-life to 13 hours [70]. Data from the Liraglutide Effect and Action in Diabetes (LEAD) program have demonstrated that liraglutide, when used as either as monotherapy or in combination with 1 oral agent, lowers HbA 1c by 0.84 to 1.48% [71,72]. In selected previously drug-naive subpopulations, representing the true initial monotherapy population, liraglutide reduced HbA 1c by 1.19% to 1.60% [71]. In LEAD-6, the head-to-head trial versus exenatide (with both GLP-1 based agents added to metformin and/or sulfonylurea), liraglutide 1.8 mg once daily resulted in a significantly greater reduction in HbA 1c than exenatide 10 mg bid ( 1.1% vs. 0.8%; P b0.0001)[73]. Taspoglutide (a once-weekly formulation of a human GLP-1 analog) and albiglutide (an albumin conjugated GLP-1 analog) are in earlier stages of clinical development [74,75]. Both agents have demonstrated glucose reductions, but few data on their clinical effect are currently available. Although no GLP-1 based agent has yet been approved for use in combination with insulin, clinical trial data on the addition of these agents to insulin have been reported in the literature. A casecontrolled, retrospective analysis of the addition of exenatide to insulin therapy reported an incremental 0.6% reduction in HbA 1c. Use of exenatide enabled a reduction in insulin dose and was associated with weight loss [76]. Studies assessing the clinical impact of both exenatide and liraglutide in combination with insulin are currently under way. At present, it is recommended that prescription of GLP-1 receptor agonists in combination with insulin take into account the potential for increased risk of postprandial hypoglycemia, particularly if sulfonylureas or meal-related (prandial) insulin are also part of the treatment regimen. Sulfonylurea doses are typically reduced by 50% when combined with exenatide. Prandial insulin doses are usually reduced by 30% to 50%, depending on the pretreatment levels of glycemic control [60]. The question of whether insulin may be replaced by GLP-1 based therapy remains to be explored. A small study (N = 49) of short duration (16 weeks) found that such a switch resulted in a slight increase in HbA 1c (+0.3%) in patients switched to exenatide [77,78]. In addition to their effects on blood glucose, GLP-1 receptor agonists significantly reduce body weight and typically lower blood glucose without a risk for hypoglycemia. In the initial registration trials, exenatide reduced mean body weight by 0.9 to 2.8 kg; 2- and 3-year extension data [64,79] demonstrated a weight reduction of approximately 5 kg. More recently, treatment with liraglutide (LEAD trials) [71,72,80-82] has demonstrated similar reductions in body weight whether liraglutide was used alone or in combination with a variety of oral hypoglycemic agents. In a double-blind, placebocontrolled, 26-week study of liraglutide 0.6 to 1.8 mg once daily in combination with metformin, liraglutide demonstrated body weight reductions of 2.8 kg, and 4.4 kg in patients with a body mass index (BMI) >35 [81,83]. In a 52-week monotherapy study of liraglutide 1.2 and 1.8 mg once daily versus a sulfonylurea, liraglutide demonstrated weight reductions of 2.05 to 2.45 kg [71]. Comparable weight reductions (between 2.8 kg and 4.1 kg) were observed in LEAD-6 comparing liraglutide once daily to twice-daily exenatide, [73] where liraglutide demonstrated significantly greater HbA 1c reductions over 24 weeks of therapy (liraglutide 1.1% vs. exenatide 0.8%, exenatide; P b0.0001). Liraglutide therapy was also associated with significantly lower rates of minor hypoglycemia (P = 0.013) [73] GLP-1 Receptor Agonist Therapy: Other Therapeutic Implications Clinical use of the GLP-1 receptor agonists has not surprisingly been directed toward patients in need of additional blood glucose control and weight reduction (Figure 3) [38-47]. Reduction in body weight, beneficial in itself, may also exert beneficial effects on insulin action, cardiovascular disease risk factors such as blood pressure, and plasma lipids. In clinical studies, exenatide therapy decreased systolic blood pressure by as much as 2.6 mm Hg [78,84]. A recent analysis of pooled data [85] demonstrated systolic blood pressure reductions of as much as 1.7 mm Hg. A 52-week study with once-weekly exenatide showed a 6-mm Hg systolic blood pressure reduction and improvements in the lipid profile in patients with type 2 diabetes [86]. In open-label observational studies, exenatide therapy has been associated with improved triglycerides, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, and apolipoprotein B levels [84]. In three 26-week phase 3 trials, liraglutide reduced systolic blood

6 S334 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 pressure with a statistically significant difference (from 2.7 mm Hg to 4.5 mm Hg; P b0.05) [87]. A 26-week trial in which liraglutide 1.2 mg and 1.8 mg once daily were combined with metformin plus a thiazolidinedione reported systolic blood pressure reductions from baseline of 6.7 mm Hg (1.2 mg) and 5.6 mm Hg (1.8 mg) [72]. In a 14-week trial, liraglutide therapy reduced systolic blood pressure by 7.9 mm Hg; a 22% reduction in triglycerides levels was also noted [88]. Whether these favorable effects on cardiovascular risk factors are the result of treatment alone or are partially associated with body weight loss warrants further investigation. GLP-1, exenatide, and liraglutide have been reported to exert beneficial effects on β-cell function (insulin secretion) and (in animal studies) β-cell mass perhaps by prevention of cell loss through apoptosis (cell death) and by promotion of β-cell growth [89-92]. Whether the β-cell benefits of incretin hormones will translate into modification of disease progression in humans with type 2 diabetes is not currently known. However, measures of β-cell function (homeostasis model assessment [HOMA] B and HOMA-IR) are improved in the setting of both exenatide and liraglutide treatment [61,93-97] Safety and Tolerability of Incretin-Based Therapies The clinical effect of GLP-1 receptor agonists is now well established. Clinical experience with exenatide, along with the extensive clinical trial data for both exenatide and liraglutide, have resulted in increased interest in these agents, particularly because beneficial effects appear to be associated with a high degree of safety and tolerability. In clinical trials to date, b5% of patients have discontinued GLP-1 receptor agonist therapy because of adverse effects. Gastrointestinal side effects are the most common adverse events reported with exenatide and liraglutide use; and the case is similar in the early studies with taspoglutide. Exenatide treatment is associated with mild-to-moderate nausea in up to 44% of treated patients generally occurring concomitantly with drug initiation and diminishing over the first few weeks of therapy [60]. Nausea was somewhat lower with once-weekly exenatide (26%) [69]. In clinical studies with exenatide, 2% to 5% of treated individuals discontinued treatment due to severe or refractory nausea [60]. Treatment options for patients with diabetes must also take into account renal status and any underlying hepatic disease. GLP-1 receptor agonists have no known direct effect on renal function. However, exenatide is predominantly eliminated by glomerular filtration with subsequent proteolytic degradation; hence, it is contraindicated in patients with end-stage renal disease or severe renal impairment. Exenatide is not recommended in patients with severe gastrointestinal disease. There are no specific contraindications for the use of exenatide in patients with hepatic dysfunction [60].Ina small proportion of patients, formation of antibodies at high titers could result in the failure of exenatide to achieve adequate improvement in glycemic control [60]. Data from liraglutide trials suggest lower rates of nausea in most studies, with rates ranging from 11% to 30% in the phase 3 monotherapy trial. Overall, discontinuation due to nausea was uncommon in patients treated with liraglutide [71]. No single organ system accounts for significant amounts of extraction or metabolism of liraglutide. A study of liraglutide metabolism and elimination found similarities to the metabolism of endopeptidases i.e., fully degraded in the body with no significant renal or hepatic metabolism [98]. Therefore, liraglutide is unlikely to be contraindicated or require dosage adjustment in patients with severe renal [99] or hepatic disease [100]. This has been confirmed in a recent study of the effect of renal impairment on the pharmacokinetics of liraglutide, which indicated no effect of renal impairment of liraglutide exposure. The pharmacokinetics of liraglutide appear to be essentially independent of renal function [101]. Diarrhea has also been reported in patients treated with the GLP-1 receptor agonists liraglutide and exenatide. Approximately 10% to 16% of individuals treated with exenatide reported diarrhea, [61,62] as did approximately 8% to 19% of patients treated with liraglutide [71,81]. Lower gastrointestinal side effects rarely resulted in discontinuation of treatment with either agent (b1% to 4% of subjects) [62,71] Other Safety Considerations Although GLP-1 receptor agonist therapy is usually not discontinued because of gastrointestinal-related side effects, clinicians should be aware of other reported adverse events. In 2006, a case of acute pancreatitis was reported in a patient treated with exenatide. An additional 6 cases of severe hemorrhagic or necrotizing pancreatitis with 2 fatalities were reported in an update of exenatide use in late The manufacturers of exenatide (Byetta; Amylin Pharmaceuticals, Inc., San Diego, CA and Eli Lilly & Co., Indianapolis, IN) noted these were very rare case reports (with very rare defined as a rate b1 in 10,000 or b0.01%), and that the proportion of complications or fatalities reported appeared similar to that seen in the general population who develop pancreatitis [102]. It is also important to note that the incidence of acute pancreatitis is increased by approximately 3-fold in all patients with type 2 diabetes compared with the general population [103]. Risk factors associated with the development of pancreatitis (e.g., obesity, sulfonylurea use, [104] gallstones, [105] and hypertriglyceridemia [106]) are also associated with type 2 diabetes and its treatment. In the exenatide trials, the incidence of pancreatitis was lower in the study drug arm (1.7 cases per 1,000 subject-years) than in placebo-treated subjects (3.0 cases per 1,000 subject-years) or insulin-treated subjects (2.0 cases per 1,000 subject-years) [107]. Nevertheless, the FDA has cited evidence of a strong temporal association between exenatide [use] and acute pancreatitis [108] as reason for a new exenatide warning label that alerts clinicians to the potential for this adverse effect [109]. Early data suggest that the number of pancreatitis cases seen in patients treated with either liraglutide or comparator drugs in the phase 3 liraglutide development program is consistent with expectations for a population with type 2 diabetes. A pertinent medical history or alternative etiology was noted in patients who developed pancreatitis during treatment with liraglutide. At present, there are too few cases to be able to determine scientifically whether or not an association exists between development of pancreatitis and liraglutide treatment [71,80,81] Clinical Use of DPP-4 Inhibitor Therapy A number of inhibitors of the ubiquitous enzyme DPP-4, which regulates the bioactivity of native GLP-1, have been developed; 2 of these agents (sitagliptin and vildagliptin) are available for clinical use [32]. Sitagliptin has been approved in both the United States and Europe; vildagliptin has been approved by the European Medicines Agency (EMEA) for use in the European Union. As either monotherapy or an add-on to oral agents, DPP-4 inhibitors reduce mean HbA 1c by approximately 0.5% to 0.8%, a clinical effect somewhat less than that reported with the GLP-1 receptor agonists [ ]. DPP-4 inhibitors lower blood glucose without a significant increase or reduction in body weight ( 0.2 to +0.8 kg) [ ]. DPP-4 inhibitors are being used clinically in combination with most other oral antidiabetic agents (including sulfonylureas, thiazolidinediones, and metformin), typically in patients failing to achieve adequate control, or who wish to limit weight gain, or in whom adverse effects of other oral antidiabetic drugs have been reported. Clinical trial data suggest that DPP-4 inhibitors may provide greater glucose reductions when used in combination with metformin. The mechanism of this synergistic response has not been established. Goldstein and associates [112] found that initial combination therapy using various dosages of metformin and sitagliptin (e.g., sitagliptin 50 mg metformin 500 mg bid; sitagliptin 100 mg metformin 1,000 mg bid) resulted in HbA 1c reductions from

7 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 S335 Table 1 Summary of the Clinical Effects of Glucagon-like Peptide 1 (GLP-1) Receptor Agonists and Dipeptidyl Peptidase 4 (DPP-4) Inhibitors. DPP-4 Inhibitors (sitagliptin, alogliptin, saxaglitpin, vildagliptin) HbA 1c reduction % Weight neutral Oral administration No significant GI side effects Low rates of hypoglycemia Improved meal-related insulin secretion, reduced glucagon release Can reduce dose and use in renal insufficiency GLP-1 Receptor Agonists (exenatide, liraglutide, taspoglutide) HbA 1c reduction % Significant and sustained weight loss generally observed Injected therapy (once daily, twice daily, once weekly) GI side effects most common (nausea, diarrhea particularly with initiation) Low rates of hypoglycemia Multiple mechanisms of action Insulin secretion, glucagon release Reduced food intake, slowing of gastric emptying Weight loss GI = gastrointestinal; HbA 1c = hemoglobin A 1c ; = increased; = decreased. baseline of between 1.4% and 1.9% at 24 weeks. In this trial, patients with an HbA 1c of 7.5% to 11% at screening, and who were not taking an oral antidiabetic agent, entered a 2-week, single-blind placebo run-in period; patients with an HbA 1c >11% and not on an oral agent entered a diet and exercise run-in period of up to 6 weeks; patients on an oral antidiabetic with HbA 1c of 7% to 10% had the agent discontinued and entered a washout period of 6 to 10 weeks. Subsequent to washout, patients with an HbA 1c of 7.5% to 11% entered a 2-week, single-blind placebo run-in [112]. No data are available on the use of combination DPP-4 inhibitor therapy with GLP-1 receptor agonists. The mechanism of action of these therapies, and the related nature of the potential clinical effects, suggest that such a combination may not be appropriate for use in most patients [118]. Comparative data assessing the relative effects of exenatide versus sitagliptin on postprandial blood glucose [118] suggest that GLP-1 receptor agonists (perhaps due to higher mealtime concentrations of exenatide versus the levels of native GLP-1 achieved with sitagliptin) have a greater effect on postprandial glucose excursions (2-hour postprandial glucose 133 ± 6 mg/dl with exenatide versus 208 ± 6 mg/dl with sitagliptin [1 mg/dl = mmol/l]). Data demonstrating extraglycemic effects of DPP-4 inhibitors such as benefits on lipids, blood pressure, or markers of inflammation are very limited. There is no evidence that DPP-4 inhibitor therapy results in significant body weight, appetite, or food intake reductions. Trials with vildagliptin have shown modest improvements in triglycerides and high-density lipoprotein cholesterol ( 4.8% and +10.6%, respectively, in combination with a thiazolidinedione), [115] as well as reductions in systolic and diastolic blood pressure (significant for diastolic only, P = ) [116]. Modest improvements in measures of β-cell secretory function have been reported for both sitagliptin [110,119] and vildagliptin, [ ] although the long-term clinical implications of these improvements have not been determined. DPP-4 inhibitor therapy is generally well tolerated, with no significant gastrointestinal or systemic side effects having been reported in clinical trials. DPP-4 inhibitor therapy has demonstrated a low dropout rate in large clinical trials; therefore good patient adherence may be expected [111,123]. A favorable tolerability profile means that DPP-4 inhibitor therapy can be safely administered to patients with a range of decrements and comorbidities. Nevertheless, with Januvia (sitagliptin; Merck & Co., Inc., Whitehouse Station, NJ), the DPP-4 with the most published clinical data, dosage adjustments are recommended in patients with moderate- to end- stage renal disease, because Januvia is cleared by the kidney [124]. Both sitagliptin and vildagliptin are contraindicated in patients with severe hepatic dysfunction; however, no sitagliptin dosage adjustment is needed in patients with mild-to-moderate hepatic insufficiency. Sitagliptin is largely eliminated ( 80% of dose) unchanged via urinary excretion. Dosage adjustment is recommended in patients with moderate or severe renal impairment, and in patients with end-stage renal disease requiring dialysis. Postmarketing reports of sitagliptin-associated serious hypersensitivity reactions, including anaphylaxis, angioedema, Stevens-Johnson syndrome, and hepatic enzyme elevations have been noted [124]. Vildagliptin is primarily metabolized by liver hydrolysis and is contraindicated in patients with severe hepatic dysfunction [125]. Long-term risk of upper respiratory infection has also been noted with sitagliptin [124]. Vildagliptin has been associated with changes in liver enzyme levels [126]. Approval has not yet been granted in the United States pending submission of additional safety data [127]. 7. Clinical Comparison of DPP-4 Inhibitors and GLP-1 Receptor Agonists Table 1 provides a summary of the clinical effects of both GLP-1 receptor agonists and DPP-4 inhibitors. Although the list of effects is not exhaustive, it should help clinicians quickly identify the potential role for each of these therapies. Both classes of drug work well in combination with other diabetes therapies, although efficacy may be greater with GLP-1 receptor agonists. Both classes differ in route of administration and may have differing side effect and tolerability profiles. Thus clinicians must consider a composite of benefits, side effects, and patient preferences when making a therapy recommendation. A few considerations are worth additional emphasis. GLP-1 receptor agonists are generally regarded as demonstrating greater overall glucose-lowering effects compared with DPP-4 inhibitors, although clinical trial data supporting this contention are relatively limited. DPP-4 inhibitors are administered orally generally once daily (sitagliptin). Adherence and administration may be important considerations. Ease of use and reduced frequency of administration can affect clinical utility of chronic therapies. On the other hand, exenatide and liraglutide both achieve higher concentrations of the hormone, GLP-1, which induces insulin secretion and may preserve β-cell health; however, both exenatide and liraglutide require subcutaneous injection. Parenteral route of delivery may be perceived as a barrier by some patients. Yet, there are many effective strategies for overcoming the injection barrier, including early training of patients, use of simple administration devices, and adequate support from clinical staff. As with any therapy, appropriate patient selection is the key to successful adherence and successful disease management. There are differences warranting consideration within the GLP-1 receptor agonist class. A head-to-head trial has reported greater HbA 1c reduction with liraglutide versus exenatide when added to ongoing treatment with oral agents (metformin or a sulfonylurea) [73]. Timingof administration also differs among GLP-1 receptor agonists (i.e., exenatide twice daily, liraglutide once daily, exenatide once weekly, and taspoglutide once weekly). The importance of timing and frequency of administration, and the effect of such factors on tolerability and acceptability to patients, will require additional study and clinical experience. 8. Clinical Use of Incretin Based Therapy: A Practical Summary Effective therapies for diabetes exist, but therapies that are even more effective, safer, and more durable in their ability to control glycemia are clearly needed. Incretin-based therapies are among the more recent attempts to meet those needs. The appropriate use of such therapies should assist the clinician in addressing the progressive nature of type 2 diabetes not only preventing deterioration in

8 S336 D.M. Kendall et al. / European Journal of Internal Medicine 20 (2009) S329 S339 glucose control but also the weight increase commonly observed in this disease and often exacerbated by other therapies. Preclinical data suggest that GLP-1 based therapies may preserve β-cell function. Preserving the β-cell holds out the promise of a truly diseasemodifying therapy that would target the progressive nature of type 2 diabetes. Ultimately, the goal is to select therapies that are both well tolerated and that address the core defects in diabetes pathophysiology. Such therapies will improve the likelihood that patients will achieve and maintain glycemic control, thus reducing the risk of diabetes-related complications. Broadened experience with all incretin-based therapies, combined with additional clinical studies, will allow clinicians to better understand the ultimate role of these agents. Patient selection is critical; the GLP-1 receptor agonists, for example, could be particularly beneficial in patients whose weight significantly increases cardiovascular risk. Early use of these agents may be effective in preventing diabetes in those at risk, or in halting or retarding disease progression in those who have already developed type 2 diabetes. Analyses of follow-up data from the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions (DCCT/EDIC) and UKPDS, suggest that a kind of metabolic memory exists and may be associated with vascular damage not easily reversed even by effective treatment and subsequent good glycemic control [128,129]. If deleterious metabolic memory is indeed a real pathology, then early aggressive therapy, particularly therapy that might preserve β- cell health, could be even more important. Of course, additional clinical investigation will be required to test those hypotheses. Given the ever-increasing incidence of diabetes worldwide, the link between obesity and the development of type 2 diabetes, and the need for more effective, weight-focused, convenient and sustainable treatments, the data from such studies will be invaluable to further clarify the role of the incretins and of all new therapies in the management of patients with diabetes. 9. Author Disclosures The authors who contributed to this article have disclosed the following industry relationships: David M. Kendall, MD, has worked as a consultant to Amylin Pharmaceuticals, Inc., Daiichi-Sankyo Co. Ltd., Eli Lilly & Co., HealthPartners, Intarcia Therapeutics, Inc., Nektar Therapeutics, and Takeda Pharmaceuticals North America, Inc; has served on scientific/clinical advisory boards for Amylin Pharmaceuticals, Inc., Daiichi-Sankyo Co. Ltd., Eli Lilly & Co., HealthPartners, Intarcia Therapeutics, Inc., Nektar Therapeutics, and Takeda Pharmaceuticals North America, Inc.; has received research/grant support (to the International Diabetes Center [IDC]) from Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer Diabetes Care, DexCom, Inc., Eli Lilly & Co., Hoffman-La Roche Inc., MannKind Corp., Merck & Co., Inc., Medtronic Minimed, Inc., Novo Nordisk A/S, ResMed, and sanofi-aventis; has received education funding (to IDC) from Amylin Pharmaceuticals, Inc., Bayer Diabetes Care, Eli Lilly & Co., Merck & Co., Inc., Novartis Pharmaceuticals Corp., sanofi-aventis, and UnitedHealth Group; and has performed professional consulting services funded by grants from Amylin Pharmaceuticals, Inc., Bayer Diabetes Care, Eli Lilly & Co., Merck & Co., Inc., Novartis Pharmaceuticals Corp., sanofi-aventis, and UnitedHealth Group. All research activity, advisory/consultancy work and educational services and grants are performed under contract to the non-profit Park Nicollet institute and the International Diabetes Center. Dr. Kendall receives no personal compensation for these activities. Robert M. Cuddihy, MD, has worked as a consultant to Bayer, Novo Nordisk A/S, and Roche Diagnostics; has served as an investigator for Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer, Eli Lilly & Co., MannKind Corp., Medtronic, Inc., Novo Nordisk A/S, ResMed, and sanofi-aventis; and has received educational support (to IDC) from Eli Lilly & Co., LifeScan, Inc., and Novartis Pharmaceuticals Corp. All research activity, advisory/consultancy work and educational services and grants are performed under contract to the non-profit Park Nicollet institute and the International Diabetes Center. Dr. Cuddihy receives no personal compensation for these activities. Richard M. Bergenstal, MD, has worked as a consultant to Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer, Eli Lilly & Co., LifeScan, Inc., MannKind Corp., Medtronic, Inc., Merck & Co., Inc., Novartis Pharmaceuticals Corp., Novo Nordisk A/S, Pfizer Inc, ResMed, Roche, sanofi-aventis, UnitedHealth Group, and Valeritas Inc.; has served on scientific advisory boards for Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer, Eli Lilly & Co., LifeScan, Inc., MannKind Corp., Medtronic, Inc., Merck & Co., Inc., Novartis Pharmaceuticals Corp., Novo Nordisk A/S, Pfizer Inc, ResMed, Roche, sanofi-aventis, UnitedHealth Group, and Valeritas Inc.; has performed clinical trials for Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer, Eli Lilly & Co., LifeScan, Inc., MannKind Corp., Medtronic, Inc., Merck & Co., Inc., Novartis Pharmaceuticals Corp., Novo Nordisk A/S, Pfizer Inc, ResMed, Roche, sanofi-aventis, UnitedHealth Group, and Valeritas Inc.; has received educational support (to IDC) from Abbott Diabetes Care, Amylin Pharmaceuticals, Inc., Bayer, Eli Lilly & Co., LifeScan, Inc., MannKind Corp., Medtronic, Inc., Merck & Co., Inc., Novartis Pharmaceuticals Corp., Novo Nordisk A/S, Pfizer Inc, ResMed, Roche, sanofi-aventis, UnitedHealth Group, and Valeritas Inc.; and owns stock in Merck & Co., Inc. All research activity, advisory/consultancy work and educational services and grants are performed under contract to the non-profit Park Nicollet institute and the International Diabetes Center. Dr. Bergenstal receives no personal compensation for these activities. Acknowledgment We thank AdelphiEden for providing editorial services. References [1] Weyer C, Bogardus C, Mott DM, Pratley RE. The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest 1999;104: [2] Gallwitz B. The fate of Beta-cells in type 2 diabetes and the possible role of pharmacological interventions. Rev Diabet Stud 2006;3: [3] Unger RH. Reinventing type 2 diabetes: pathogenesis, treatment, and prevention. JAMA 2008;299: [4] Karter AJ, Moffet HH, Liu J, et al. Glycemic response to newly initiated diabetes therapies. Am J Manag Care 2007;13: [5] Cook MN, Girman CJ, Stein PP, Alexander CM. Initial monotherapy with either metformin or sulphonylureas often fails to achieve or maintain current glycemic goals in patients with Type 2 diabetes in UK primary care. Diabet Med 2007;24: [6] Salber PR, Bestermann W, Schwartz S, Marchetti A. Loss of confidence in diabetes management. Manag Care 2008;17: [7] Doggrell SA. Metformin & lifestyle intervention prevent Type 2 diabetes: lifestyle intervention has the greater effect. Expert Opin Pharmacother 2002;3: [8] Castaneda C. Diabetes control with physical activity and exercise. Nutr Clin Care 2003;6: [9] Nathan DM, Buse JB, Davidson MB, et al, American Diabetes Association and the European Association for the Study of Diabetes. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: A consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 2009;32: